FAO GENEVA SYMPOSIUM

on

THE EXPERIENCE WITH IMPLEMENTING THE WTO AGREEMENT ON AGRICULTURE AND SPECIAL AND DIFFERENTIAL TREATMENT TO ENABLE DEVELOPING COUNTRIES TO EFFECTIVELY TAKE ACCOUNT OF THEIR DEVELOPMENT NEEDS, INCLUDING FOOD SECURITY AND RURAL DEVELOPMENT

Geneva, 2 October 2002

Paper No. 3

Developing country experience with the implementation of the Uruguay Round Agreement on Agriculture: Synthesis of the findings of 23 Country Case Studies

Commodities and Trade Division

FAO, Rome

Developing country experience with the implementation of the Uruguay Round Agreement on Agriculture:

Synthesis of the findings of 23 Country Case Studies ¹

1. INTRODUCTION

1.1 Context

1. As mandated under Article 20 of the Uruguay Round Agreement on Agriculture (AoA), negotiations to continue the reform process designed to achieve the long-term objective of substantial progressive reductions in agricultural support and protection resulting in fundamental reform began in early 2000. These negotiations take into account the experience from implementing the reduction commitments under the AoA; the effects of these commitments on world trade in agriculture; non-trade concerns; special and differential treatment to developing country members of the WTO and the objective to establish a fair and market-oriented trading system and other objectives mentioned in the Agreement’s preamble; and further commitments necessary to achieve the Agreement’s objectives.

2. The Doha Declaration sets a timeframe in which virtually all the linked negotiations, including agriculture, are to end by 1 January 2005. Modalities for a revised AoA must be agreed by March 2003. The Declaration also confirms that special and differential treatment for developing countries should be integral to the negotiations, both in countries’ new commitments and in any relevant new or revised rules and disciplines. It further states that the outcome should be effective in practice and should enable developing countries to meet their needs, in particular in food security and rural development.

3. Developing country participation in these negotiations should be grounded in an analytical and empirical understanding of the effects of the Agreement to date. FAO initiated a major exercise in 1999 to evaluate the experience with the implementation of the AoA in developing countries. Fourteen country case studies were commissioned and an overview paper synthesising these experiences and the lessons to be learned was prepared². FAO has now updated this exercise, drawing on a wider set of countries and more recent data. Sixteen country case studies were commissioned at the beginning of 2002 to review national experiences. In some cases, the case studies revisited countries included in the 1999 sample but the opportunity was also taken to widen the sample by including additional countries. This synthesis paper draws on the findings for countries included in both sets of studies – 23 in all – to develop a synthesis of their experiences. However, the emphasis in the textual commentary is on the second set of 16 case studies completed in 2002. References to the experiences of individual countries in this synthesis paper are based on the individual case studies unless otherwise stated.

1.2 The country sample

4. The selection of the 23 countries was based on a number of considerations, such as broad geographical balance, the need to include different categories of countries such as least-developed countries, net food-importing countries and agricultural exporters, and availability of national consultants to complete the studies by a specified deadline. Table 1 shows the countries included in the exercise and on whose experience this synthesis paper draws³.

5. Agriculture plays an important role in the economies of all the countries covered, as can be seen from Table 2. For all countries in the sample, agriculture contributes on average 20 percent of GDP, ranging from 4 percent in Botswana to 44 percent in Uganda. Agriculture remains much more important as a source of livelihoods accounting on average for 46 percent of total employment in the sample, ranging from 17 percent in Brazil to 80 percent in Uganda. Agriculture contributes on average just under 30 percent of all earnings from merchandise good exports for countries in the sample, with a range from under 3 percent in Bangladesh to 76 percent in Malawi.

Table 1. The sample countries included in this study

Table 2. Role of agriculture in the sample countries

Source: FAOSTAT; and World Development Indicators CD-Rom 2002.

1.3 Objectives of the study

6. This synthesis paper and the accompanying case studies set out to provide answers to four questions concerning the impact of the AoA on developing countries:

• Have the AoA commitments led to changes in domestic agricultural policy in developing countries? If so, what have been the nature of those changes? If not, why were the commitments not binding? How much flexibility do developing countries currently have under the current Agreement? What are likely to be the sensitive areas for individual countries in the current negotiations? Note that the case studies cover not just the commitments under the AoA but also those relating to the Agreement on Sanitary and Phytosanitary Standards (the SPS Agreement) as well as rights and obligations under Article 27(3) of the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) covering plant varieties protection.

• Did the AoA commitments have an impact on trade flows (imports and exports) of developing countries? To the extent that in answer to the first question the consultants find that the AoA has had a limited impact on policy implementation in developing countries to date, it follows that the impact on import flows will be correspondingly limited. However, it is interesting to analyse the experience of the case study countries with respect to food imports to identify sensitive issues. Have there been import surges in particular commodities, and how have countries dealt with these? Have developing countries faced problems with dumped or subsidised imports? What about import price volatility? It is also important to look for evidence of improved market access opportunities which were the quid pro quo for the commitments entered into by developing countries. Have developing countries been successful in increasing traditional agricultural exports or in diversifying into either new markets or new product lines as a result of the Agreement? What has their experience been with SPS barriers and the implementation of the SPS Agreement? Some cases studies also seek evidence on the relationship between more open agricultural trade policies and foreign investment flows into the agri-food sector, but not enough case studies address this issue to be able to draw any generalisations on this subject.

• The third question is whether the AoA commitments had an impact on food security? Were these impacts positive or negative? If the impact on trade flows to date has been limited, so will be the impacts on food security. The case studies investigate this issue from a number of angles. Trends in per caput dietary intake and in the incidence of under-nourishment provide some initial insights into the food security situation at household level. However, these indicators are influenced by a wide range of factors, of which the AoA, and even the Uruguay Round (UR) Agreements as a whole, are just one. Many of the case studies investigate particular channels through which the AoA might have impacted on food security, including changing self-sufficiency ratios and the capacity to pay for food imports; trends in relative food prices; or identifying those farm groups which have been able to take advantage of improved export opportunities or which have faced additional import competition. While, in the space available, this was inevitably a broad-brush approach, the case studies do provide useful insights into the AoA impact on food security to date.⁴

• Finally, each of the case studies identifies key negotiating issues and demands of the individual countries which, together, reflect the range of negotiating positions currently proposed by developing countries. The purpose of this section was to highlight the priorities and main concerns of the case study countries in the negotiations. In some cases, the case studies identified particular areas where technical assistance would be useful, as regards both analysis and policy formulation, as developing countries seek to adapt policy to the new trading rules.

1.4 Methodological approach

7. Assessing the impact of the AoA, or indeed any trade liberalisation agreement, faces a particular methodological problem. While a simple comparison of trends (whether in agricultural trade flows or food security indicators) before and after implementation of the Agreement is the most obvious approach to assessment, it runs into the difficulty that many other factors may be at work influencing these trends, and it may be very difficult to isolate the influence of the Agreement alone. Methodologically, there are a number of approaches to get around this problem.

8. Simulation modelling using country models is one possible approach.⁵ A simulation model purports to be a realistic representation of the structure and behaviour of a country’s agricultural sector. By using the model to simulate the impact of implementing the policy changes agreed under the AoA, it is possible to estimate its impact while holding all other factors constant. While conceptually this is an attractive approach, the outcome of model simulations is heavily determined by their built-in assumptions on economic behaviour. Another drawback of country models is that they are often too aggregative to capture the detail of tariff line negotiations.

9. Another approach is to use cross sectional statistical analysis to examine the average relationship between trade liberalisation, agricultural activity and food security. In a properly specified statistical model it is possible to take account of other factors in a way which allows the underlying relationship between these variables to be identified. The accuracy of the results will depend on whether the assumptions behind the statistical model are valid or not, and specification and data issues are often not easy to resolve.

10. A third approach which is used in this study is to rely on case studies to throw light on the impact of the Agreement. The case study approach has the advantages that it allows issues to be examined in their specific context, it is disaggregated and it makes use of the detailed country-specific knowledge of the individual consultants who prepare the case studies. However, the case study approach also has its limitations which it is important to be aware of when reading this synthesis report. First, it is important to choose a representative sample of case studies; many arguments drawn from case studies are based on just one or two well-chosen examples. Second, despite common terms of reference in broad terms, the details of each case study are inevitably different. Not all case studies allow answers to some of the questions raised above. The fact that problems with dumped imports are not discussed in a case study, for example, does not necessarily mean that the problem did not arise. Third, the conclusions drawn from a case study depend on the analyst and the data basis chosen for the analysis. Despite these caveats, however, the case study approach allows the questions raised earlier on the AoA impact to be discussed in great detail and with a richness that is not possible with either the simulation modelling or statistical approaches.

11. This paper is organised as follows. Section 2 which follows summarises the national experiences in implementing the main provisions of the AoA and country commitments, notably with respect to market access, domestic support measures and export subsidies. It also deals briefly with experiences related to the SPS and TRIPS Agreements. Section 3 reviews the experience of the sample countries with food and agricultural trade before and after implementation of the AoA. Section 4 looks at trends in food security and undernutrition during the AoA implementation period. Finally, Section 5 draws some general conclusions from the discussion on the specific interests and concerns of each of the case study countries in the context of the current Doha Round trade negotiations in agriculture.

2. IMPLEMENTATION OF AOA AND OTHER WTO AGREEMENTS

The focus of this section is on the experience with implementing commitments on market access, domestic support, export subsidies and the SPS Agreement – and not on the effects of their implementation. The experience with trade flows is reviewed in Section 3 and with food security in Section 4.

2.1 Market access

12. While bound tariffs are high on average, there are several exceptions. Not all developing countries have high bound tariffs on some or all agricultural products – contrary to prevailing views – despite the option they had in the UR to offer ceiling bindings generally. If 40 percent is arbitrarily taken as a threshold distinguishing between high and low bound rates, 12 of the 19 case study countries for which information is available had high rates (>40 percent) but 7 of the 19, or one-third of the total, had low rates (40 percent or less) (see Table 3). Bound rates in the group of countries with high bound rates varied between 50 percent (Sri Lanka) up to 200 percent (Bangladesh). Of the 7 countries with low bound rates, some are classified as net agricultural exporters (Brazil, 35 percent; Philippines, probably around 30-35 percent; Thailand 36 percent) but a surprising number are net food importers (Egypt, initially 62 percent but reducing to 28 percent in 2004; Fiji, 40 percent; Peru, 30 percent; Honduras, 35 percent).

13. Applied rates are, on average, much lower than bound rates. On a very approximate calculation, the bound rate for those countries in the sample for which data are available averaged 82 percent, while the applied rate averaged 19 percent. In only two countries in the sample for which information is available are average bound rates similar to applied rates; in Thailand (where the average applied rate of 32 percent compares to an average bound rate of 36 percent) and Egypt (which committed to reducing its bound rates from 62 percent initially to 28 percent in 2004 compared to an average applied rate of 22 percent). If the sample is again divided into two, this time arbitrarily using 15 percent as the threshold between countries with high and low average applied rates, the average value for the group of countries with high average applied rates is around 20-25 percent. For those countries in the group with low applied tariffs, the average value is around 12 percent. Indonesia and Botswana stand out with average applied tariffs of only 5 percent and 6 percent respectively.

14. There are at least four reasons why applied rates are so much lower than bound rates. First, in many cases, countries were deliberately cautious in committing the bound rates, to allow the maximum amount of flexibility in the future and possibly to maximise their negotiating leverage. By contrast, applied tariffs were low because of tariff reforms i.e. reduction and harmonization, had been ongoing for a decade or more prior to the conclusion of the Uruguay Round.

15. Second, the regional integration process involving many developing countries in the 1990s has been particularly important in lowering applied tariffs. For example, in the Mercosur negotiations the maximum tariff on agricultural products was set at 20 percent, compared to Brazil’s bound rate of 35 percent on most agricultural imports and 55 percent for a range of important commodity staples. In 1997, as a result of the revision of the Common External Tariff (CET) of the Central American Common Market (CACM), the maximum applied tariff was set at 15 percent compared to Honduras’ general WTO binding of 35 percent. Jamaica applies the CET of CARICOM where the maximum tariff is 40 percent compared to Jamaica’s WTO bindings of 100 percent. In Peru, where applied tariffs are already below bound tariffs, the gap will widen further when Peru implements the Andean Common Market CET in the near future. Indonesia announced in May 1995 a long-run tariff reduction package (Pakei’95) which was of far more significance for its tariff schedule than the Uruguay Round and which was, in part, a response to its commitments under APEC. A particularly striking case is Senegal, which as a member of WAEMU, since January 2000, has lowered its top rate of tariff to 20 percent compared to its average bound rate (including ODCs) of 180 percent.

16. Third, some countries with large populations at or near poverty levels have not found it politically feasible to maintain high domestic food prices through tariffs. Fourth, some countries have been obliged to set applied rates much below their WTO bound rates due to loan conditionality.

17. Another striking feature is that developing countries, in many cases, have opted to continue reducing applied tariffs in the post-AoA implementation period. In the AoA, developing countries (with the exception of the least-developed countries) committed to reducing their bound rates by an unweighted average of 24 percent over ten years, subject to a minimum reduction of 10 percent in each tariff line. However, given the gap between bound and applied rates, such a commitment need not have required any reduction in applied rates over the implementation period. In fact, however, developing countries have continued to reduce their applied tariffs over this period. Examples include Costa Rica (where average applied tariffs were reduced from 17.1 percent in 1995 to 14.8 percent in 2000), Egypt (which has recently announced further measures to liberalise its trade regime), Fiji (which reduced its maximum tariff to 27 percent in its 1999 Budget and simplied its tariff structure), Honduras (which adopted the lower CACM CET in 1997), Indonesia (which accelerated its tariff reduction programme and which now has an average tariff for agriculture of 5 percent compared to the target of 13.2 percent for 2003 set out in its Pakmei ’95 programme), and Senegal (which lowered its top rate from 65 percent to 20 percent on applying the WAEMU CET).

Table 3. Bound and applied tariffs

Source: Countries marked with an asterisk appeared in the earlier FAO study and, for these countries, the data are from 1999 or the most recent available year before that date. For other countries, the data are drawn from the national case studies commissioned in 2002.

18. While these continued reductions were due in one or two cases to loan conditionality (Indonesia being the main example among the case study countries), in most cases these reductions were deliberately pursued by the countries themselves as part of their chosen economic development strategies. It is important to point out that a reduction in applied tariffs does not necessarily mean that production incentives to domestic farmers have been reduced. In many cases, tariff reductions have been pursued in the context of exchange rate adjustments which have more than compensated farmers for the reduction in tariff protection. Exchange rate devaluation is not an unmitigated blessing even for farmers producing import-competing or export crops as it also increases the cost of imported inputs necessary to take advantage of higher producer prices. But the point is an important one. Macroeconomic factors, such as the level of the exchange rate or of real interest rates, are often a more significant influence on agricultural production incentives than sectoral interventions such as tariff policies.

19. While applied tariffs in many developing countries are often relatively low and falling, the dispersion of tariffs discussed in a number of case studies showed that some countries had difficulty “living with” the ordinary tariff in its simplest form for a number of products, notably basic foods. Tariffs on these products were often higher than average, and were often supplemented by additional measures such as surcharges and variants of price band policies. For example, Costa Rica with an average applied tariff on agricultural products of 14.8 percent in 2000, applied a 57 percent import duty on dairy products and 150 percent on dark chicken meat. Brazil adopted higher bound tariffs for some goods such as wheat, corn, rice, cotton, beef and dairy, although these bound rates are now irrelevant because tariff policy is determined by Mercosur. Egypt retains high tariffs on poultrymeat. Fiji applied a rice tariff of 40 percent compared to the general 10 percent rate until its 1999 tariff reform, after which the new maximum rate of 27 percent was also applied to dairy and meat products. Indonesia applies higher tariffs on rice and sugar which are exempt from its IMF commitment to apply zero tariffs to food items. In Senegal, applied rates are relatively high for fruit and vegetables, cotton and textile fibres, and sugar. Other products deemed sensitive, where a surtax was introduced to compensate for the removal of quantitative restrictions, include rice, bananas, onions, potatoes, millet, sorghum and corn. Price band schemes allow higher tariffs on maize, rice and sorghum in Honduras and milk, maize, sorghum, rice and sugar in Peru. In some countries (Egypt, India, Indonesia) alcoholic beverages continue to be taxed very highly, often with tariffs of 150 percent or more.

20. The list shows that food staples feature prominently among the ‘sensitive’ commodities, but that the list is by no means confined to them. It also includes dairy products, meat particularly poultrymeat, sugar and alcoholic beverages. In fact, a number of countries apply particularly low duties to imported staple foods. These include Egypt (where applied tariffs on wheat, maize and groundnut oil were 1 percent in 2000) and Indonesia (apart from rice). India also had bound rates of zero percent for commodities such as rice, coarse grains and dairy products following commitments in earlier GATT rounds, but these were irrelevant because imports were controlled by quantitative restrictions (QRs) maintained on balance of payments grounds. When India eliminated QRs under pressure from other WTO Members, it re-negotiated higher bound rates for these products (typically in the 40-60% range).

21. The large differences between bound and applied rates imply that most countries could continue to offer concessions on bound rates without this requiring further changes in applied rates. However, the evidence on higher tariffs on ‘sensitive commodities’ implies that, in most countries, there are usually some commodities where applied rates are very close to bound rates. This suggests that developing countries might be interested in a tariff-cutting formula which targets an average cut but allows lower cuts on sensitive commodities. However, if the same formula applies to developed countries, the exception might be used by them to minimise tariff reduction on products of particular export interest to developing countries. Therefore developing countries need to undertake their own analysis whether they are more likely to gain or be damaged by exempting sensitive products from general tariff-cutting formulae. The additional protection gained for certain import-competing sectors must be weighed against the potential for continuing barriers in important export markets against their domestic export-oriented sectors.

22. For those few countries where data were presented, it appears that tariffs tend to escalate by degree of processing, thus providing higher nominal protection to the processing sectors involved. For example, the tariff structure in Costa Rica is 10 percent for raw products, 13 percent for semi-processed and 20 percent for processing products. In Zimbabwe, the current structure of applied tariffs is also characterised by a three-tier structure with tariffs escalating according to the level of processing. Although the case studies do not provide sufficient data to draw a general conclusion, it is not unlikely that tariff escalation is a widespread feature as it is of developed country tariff schedules. Such countries should be aware of the trade-offs (between protecting the import-competing agricultural processing industry and creating additional opportunities for added value in export-competing sectors) implicit in different tariff-cutting formulae which differentiate between low and high tariff rates.

23. The evidence from the case studies suggests that quantitative restrictions (QRs) as a trade measure in developing countries are now a thing of the past. In the past, many developing countries relied on quantitative import restrictions, often implemented through marketing boards (State Trading Enterprises), to provide protection to domestic import-competing sectors. Under the AoA, all such non-tariff barriers had to be phased out, apart from import restrictions maintained on grounds of health, safety or moral conduct. Both Honduras and Peru had eliminated all non-tariff restrictions on imports including quotas and QRs before the implementation of the AoA. In neither case did the country compensate for the elimination of QRs by raising import tariffs; that is: there was no tariffication of these past restrictions. In the Philippines, quantitative restrictions were removed but accompanied by a compensatory increase in tariffs, with the exception of rice. Indonesia has reduced the number of commodities requiring import licenses in response to its AoA commitment, although in practice loan conditionality has been a more important driver. Senegal also introduced surtaxes during the UR in replacement of earlier import quotas. India, which had maintained QRs on balance of payments grounds, phased these out more quickly than it had originally proposed in the light of a dispute panel finding against it, but was able to renegotiate new bound tariffs of up to 80 percent on these commodities, presumably in return for opening TRQs.

24. Relatively limited use is now made of State Trading Enterprises (STEs) with monopoly import powers in the sample countries. In most countries in the sample, such marketing boards were shut down or their monopoly import powers withdrawn as a result of structural adjustment programmes pursued since the mid-1980s. There were only a few remaining instances in the case study countries. Indonesia notified the WTO that both BULOG and BPPC operate as STEs within the meaning of Article XVII of GATT. In Egypt, STEs play a major role in wheat imports and cotton exports. Egypt is one of the largest importers of wheat in the world, and the General Authority of Supply Commodities (GASC), the STE, imports about three-fourths of Egypt’s import requirements. GASC bears the responsibility of purchasing domestic wheat as well as wheat imports both used to produce the 82 percent flour used in baking subsidized “baladi” bread used in Egypt’s food subsidy programme. In the case of cotton, the public sector companies export three-fourths of the total Egyptian cotton exports. Other cotton products, such as yarn, fabrics, and garments are also largely exported or imported by STE’s. In Zimbabwe, maize imports are solely undertaken by the Grain Marketing Board.

25. Commitments on the behaviour of STEs are governed by Article XVII of GATT. The commitment is to ensure that import purchases are non-discriminatory and that the margin between the domestic price and world price falls within the tariff binding for each commodity. The Indonesian case study notes that there may have been a few occasions in the case of sugar where the margin between the domestic price and world price has been close to Indonesia’s bound rate. The other case studies do not suggest that tariff commitments have constrained the operation of domestic STEs in any way.

26. The case study countries have quite diverse experiences in the management of tariff rate quotas. First, relatively few countries in general have opened tariff rate quotas – just five (Brazil, Costa Rica, India, Indonesia and the Philippines) out of the 23 case studies in the sample. The background to these TRQs is diverse; for example, in the case of the Philippines, it was a response to tariffication, while in the case of India, it was part of its deal to provide compensation to be allowed to increase bound tariff rates on certain commodities in post-AoA negotiations. The utilisation experience has also been different. In some countries, such as Brazil and for some TRQs in the Philippines, the TRQs have been redundant because applied MFN tariffs have been lower than the in-quota tariff rates. In the case of the Philippines bovine animal and beef TRQs, their abolition reflected pressure from the domestic cattle industry for lower tariffs on live cattle as well as recognition that the input needs of meat processors cannot be met by the local cattle industry.

27. Where TRQs are binding, as in the case of Costa Rica, India, Indonesia and for some TRQs in the Philippines, utilisation rates have been variable, but in general have been limited. In the Philippines, this led to complaints from the US about the non-filling of domestic pork quotas. The government responded that frozen pork products have a very limited demand in the domestic market where consumers prefer fresh meat. Administration of TRQs has also been diverse. Costa Rica has followed a rigorous MFN allocation principle based on trading TRQ allocation rights on BOLPRO, a commodity exchange. In the case of the Indonesian rice TRQ, imports were the responsibility of BULOG which had an import monopoly in any case. The Indonesian dairy TRQ was administered by adapting a local content scheme in which domestic and imported milk were mixed in a fixed ratio. TRQ quotas are allocated using milk absorption certificates based on the amount of domestically produced milk used in processed products.

28. The case studies demonstrate that tariffs are often the primary, if not the only, trade instrument open to these countries to stabilise domestic markets and to safeguard farmers’ interests in the face of sharp swings in world prices or a surge in imports. This was despite the fact that most case study countries have now implemented trade remedies legislation to address dumping and the use of subsidies by trading partners as well as legislation making possible the use of the emergency safeguards.

29. The use of contingency measures allowed under WTO rules (antidumping and countervailing measures, emergency safeguards etc) are sometimes necessary to deal with the volatility of agricultural trade and unfair competition. Brazil is one country that has implemented anti-dumping legislation and made use of it. In January 1999, the National Confederation of Agriculture (CNA) requested an anti-dumping investigation into imports of milk by Brazil. At the conclusion of the investigation, in February 2001, Brazil imposed an anti-dumping duty on imports of milk powder and whole milk from New Zealand, the European Union and Uruguay. Egypt introduced procedures to be followed regarding the application of safeguard measures as well as anti-dumping and countervailing duties in 1998. In 2001, Egypt initiated a countervailing and safeguard case with regard to powdered milk, where a safeguard margin of 45 percent was imposed on imports of powdered milk. Indonesia has also introduced new anti-dumping and countervailing duty procedures. While about 14 anti-dumping petitions have been filed to date, only one of these involved an agricultural commodity - wheat flour. Although there was a positive finding of dumping and injury in the case of flour, the Government has delayed the imposition of duties pending further investigation of Indonesia’s national interest.

30. Despite this evidence that some of the sample countries are able to successfully implement anti-dumping actions, extensive procedural requirements and conditions make these mechanisms difficult to use. As the Jamaican case study observes, the classic trade remedy instruments of the WTO are not user friendly for small developing countries. The Honduran case study notes that the limited utilization of trade remedies legislation can be due to its complexity; the high cost of the process in relation to the magnitude of the market in dispute; and the limited capacity of the affected organizations or enterprises to provide the required information to justify the opening of an investigation, carrying out the studies, and doing the follow-up the WTO’s rules require. The Philippines case study points out that, despite the availability of such legislation, no action has been taken to counter the dumping of chicken leg quarters from the US in the Philippine market which is detrimental to the domestic broiler industry. Furthermore, some countries, such as Fiji, have yet to introduce anti-dumping legislation in line with the GATT due to a lack of technical capacity and resources.

31. To address the problems inherent in using the classic trade remedies, the AoA introduced the Special Agricultural Safeguard which was available to countries which had used tariffication to eliminate non-tariff border restrictions and which explicitly scheduled it in their Schedule of Commitments. However, a few case study countries had access to this provision because they had used the tariffication option (Botswana, Costa Rica, Indonesia, Morocco, Philippines and Thailand). Costa Rica has the right to apply it in the case of black beans and rice and used it recently to protect against very low priced rice imports. In March 2002, the country placed a temporary import duty of 80 percent on paddy rice imported from the US (when the normal tariff is 35 percent). The issue is currently under debate at WTO where the US has presented a query. Indonesia also has the right to use the special safeguard in connection with dairy products and cloves but, to date, has not done so.

32. In general, other developing countries who have the right to use the SSG have also not used it to date. This may be due partly to the technical conditions required to make use of it. To apply the safeguard, the current nominal price of imports in domestic currency should be lower than the corresponding average price effective during the 1986-88 period, which was a period of very depressed international prices and of strong overvaluation of many developing countries’ currencies. This implies that the trigger price for these countries turns out to be very low in comparison to many current prices.

33. More important, possibly, is the wide margin which many developing countries enjoy between their applied tariffs and their bound tariffs, thus allowing tariffs to be increased without having to appeal to the special safeguard clause. Fiji, for example, raised its tariff on meat products in 2002 from 10 percent to 27 percent. The relatively limited number of cases quoted in the case studies where countries have altered applied tariffs upwards may be because they experienced relatively few import surges of the kind where safeguard action would be appropriate. One country where there is some evidence of import surges that undermined domestic sectors (notably poultry, beef, dairy products and rice) is Jamaica⁶. This issue is further examined in the following section which assesses the evidence on trade flows post-AoA.

34. Of particular interest in this context are the price band schemes used by a number of Latin American countries, including Peru and Honduras among the case study countries. In Peru, application of the price band meant that 29 items were subject to variable specific duties intended as a price stabilization and protection mechanism. The scheme in Peru was one of variable tariffs, that is, a band "without ceiling", until recently. In June 2001, a new price band system was established affecting five product groups: milk, maize, sorghum, rice, and sugar. The price band system in Honduras applies to maize, sorghum and rice. There is a danger that, if applied tariffs are adjusted regularly in response to changes in world market prices, then they will be interpreted as a variable levy whose use is not permitted under the AoA. There is a panel currently looking at price bands in Chile which should clarify the rules in future.

2.2 Domestic support

35. Under the AoA, domestic support is divided into three main categories. The Green Box (GB) covers all outlays which are deemed not to be trade distorting to any significant extent. The GB outlays must meet the two general criteria that the support is provided through a publicly-funded government programme not involving transfers from consumers, and that it does not provide price support to producers. In addition, specific policy criteria must be met for a series of government interventions which are set out in Annex 2 of the AoA. No limits are placed on GB expenditure in either developed or developing countries.

36. The Amber Box contains domestic supports which are deemed to be trade-distorting. Two categories of support measures are included in this box; product-related support (PS) directed to individual commodities and non-product related support (NPS) available in principle to producers of a number of, or any, commodity. The value of these trade-distorting supports is measured by the Aggregate Measurement of Support (AMS) indicator. Developing countries with AMS support committed to bind their AMS at the 1986-88 levels and to reduce these AMS levels by 13.3 percent over the period 1995-2004. Least developed countries committed to bind their AMS levels but were exempted from this reduction commitment. Countries without AMS support in the base period committed not to introduce such support in the future in excess of the de minimis levels.

37. The de minimis levels for developing countries are defined as support up to 10 percent of the value of individual commodity production in the case of PS AMS and up to 10 percent of the value of aggregate agricultural production in the case of NPS AMS. Thus, in theory, even developing countries without an AMS entitlement could provide support up to 20 percent of the value of production of a commodity provided this support is distributed in such a way that no more than 10 percent of the value of production is provided as NPS AMS and that PS AMS does not exceed 10 percent of the value of production of each commodity. Countries which provided AMS support above de minimis levels in the base period have greater flexibility in that they are only required to ensure that all AMS outlays are below their AMS ceiling and they are not limited to the sub-ceilings inherent in the de minimis rules.

38. In addition, developing countries can exclude investment subsidies which are generally available to agricultural producers and agricultural input subsidies generally available to low-income or resource-poor farmers – the Special and Differential Treatment provision - as well as outlays designed to encourage diversification away from the production of narcotic crops. If half of all NPS AMS measures qualified for these exemptions, then the theoretical maximum of AMS support which developing countries can provide to their agricultural sectors would increase to 25 percent of the value of production. It is very unlikely that, in practice, expenditure outlays could be so precisely targeted so as to utilise all the theoretically possible room for manoeuvre. Therefore, the practical ceiling on AMS outlays in developing countries is probably closer to 15-18 percent. Green Box expenditures are in addition to these.

39. Finally, developing countries undertook to notify regularly the WTO of their domestic support expenditures and the categories into which these fell. Of the total of 23 case study countries, only 5 submitted detailed information on support measures, i.e. GB outlays, PS AMS and NPS AMS levels and SDT outlays. Only four of the case study countries have AMS reduction commitments (Brazil, Costa Rica, Morocco and Thailand). The reported AMS outlays in these countries were less than the de minimis levels. Most countries simply reported that their domestic support outlays conformed to the “exempted” categories (GB, SDT or de minimis). In a few cases, countries gave details of their GB measures and in some cases of their SDT measures as well. However, some countries failed to make any notification at all. The subsequent reporting of measures by the case study countries has also been rather patchy.

40. For those countries for which data are available (Table 4), the AMS levels for recent years have been well below the committed or permitted levels, with the “utilisation ratio” on the higher side only for Thailand. It is probably the case that Indonesia’s administered price for rice exceeds the de minimis standard. However, the Government lacks the resources to support domestic prices at the administered level. In other words, the system has been ineffective at providing farmers the full support implied by the administered price. In fact, a number of countries in the sample reported that not only AMS outlays but also GB outlays have been in decline due to budget constraints. Brazil, for example, has eliminated many of the production subsidy programmes in place in the AoA base period and even GB outlays have been declining because of fiscal constraints or change in policies. Uganda, Jamaica and Senegal are other countries where limited resources prevent major support programmes. In the case of other countries, government support measures have been removed in the context of structural adjustment programmes (Zimbabwe). In the Philippines, trade distorting subsidies are reported to be well below the de minimis levels. There has been a conscious effort in that country to phase out input subsidies in favour of more enduring productivity-raising support such as irrigation and market infrastructure.

Table 4. Summary of information on domestic support measures

Source: Countries marked with an asterisk appeared in the earlier FAO study and, for these countries, the data are from 1999 or the most recent available year before that date. For other countries, the data are drawn from the national case studies commissioned in 2002.

41. Peru is an exception, where GB measures tripled between 1995 and 1997 to reach 5 percent of the value of agricultural production. NPS AMS outlays have been fairly constant at around 6 percent of the value of agricultural production. Peru does not appear to have invoked the SDT exemption either because it is not necessary to do so or because most of this support goes to larger and more commercial farms. Its PS AMS level is zero. In the case of India also, unofficial estimates made in the case study suggest there has been a steady increase in the level of NPS AMS once SDT has been accounted for. Its NPS AMS, which includes subsidies on irrigation, fertilisers, electricity, credit and seeds, was about 1.3 percent of the value of agricultural production during the base period after allowing for exemptions which are granted for resource poor farmers in developing countries. In 1995-96, the NPS AMS was roughly 7.5 percent of the value of agricultural production, but in the subsequent notification, this fell markedly to about 1.1 percent for 1996/97 and 1997/98, and by 2000-01 stood at 2.3 percent. The evidence from the case studies suggests that WTO disciplines have not been constraining to the domestic support policies that sample countries wanted to implement. Budgetary constraints and previous commitments under SAPs appear to be much more important in limiting these interventions.

42. Although there were no cases of countries currently breaching the AoA rules and commitments, the case studies identified a number of particular issues that could be of significance for the future.

• In calculating market price support external reference prices remain fixed based on the years 1986 to 1988. The problem of a fixed external reference price in the calculation of AMS assumes special significance because supporting tables in the original submissions that the member countries made to the WTO were expressed in domestic currencies. This has created problems in the calculation of AMS for developing countries where inflation rates are in general high compared to developed countries. The only reference to inflation in the agreement is in the Article 18:4, but there is no clarity as to what it implies.

• There is insufficient clarity in the agreement whether the quantity eligible to receive the administered price is total production, or only the marketed surplus which is actually sold in the market, or the quantity which is actually procured by the government through the price support mechanisms.

• Countries which do not have reduction commitments are eligible to raise their domestic support up to the de minimis level. But Article 13 b (ii) of the AoA states that domestic support that conforms to the provisions of the agreement will be non-actionable only provided that such measures do not grant support to a specific commodity in excess of that decided during the 1992 marketing year. It is therefore not clear whether countries could use the de minimis level of support allowed in the agreement.

• Under Article 6.2 agricultural input subsidies generally available to low-income or resource poor farmers in developing countries are exempt from reduction commitments on domestic support. This also applies to domestic support granted to encourage diversification from growing illicit narcotic crops. The problem with this clause is how to define a low-income or resource poor farmer? No guidance is provided in the agreement.

43. A particular issue is the number of countries which have yet to make any notification of their domestic support outlays to the WTO. In large part, this appears due to the lack of technical capacity to undertake this onerous task. It is important that developing countries are enabled to undertake this, not only to fulfil their WTO notification obligations, but also to facilitate their strategic thinking in the context of the current negotiations. Developing countries must be in a position to develop a view of how the AoA disciplines might constrain in any way the future implementation of agricultural development policy. As part of that exercise, it would be helpful to compute the threshold levels under various forms of support measures.

2.3 Export subsidies

44. Under the AoA, developing countries which had export subsidies in place in the base period committed to binding them and reducing them by 20 percent over the period 1995-2004. Developed and developing countries which did not use export subsidies in the base period committed not to use them in the future. However, an exemption for developing countries allows them to grant subsidies to reduce the cost of domestic marketing and international freight.

45. Export subsidisation was generally not an issue among the case study countries. Most countries did not have the right to grant export subsidies, but this limitation was not generally seen as a problem, as export subsidies are not affordable for most developing countries. Countries which had used export subsidies in the past, such as Peru, Honduras and Senegal, had phased them out. Brazil and Indonesia, which have an entitlement to use export subsidies, have not used them in the post-AoA implementation period. In the case of Brazil, the two programs included in its commitments, tax exemption for selected processed agricultural products and special sales of government stocks for exporters, no longer exist. Costa Rica is one country where the WTO disciplines were effective. To encourage exports, including agricultural exports, the country applied export subsidy certificates from 1994 to 2000 with the objective to motivate non-traditional exports. The program was criticised because it benefited larger firms and also because it was misused to favour fraudulent export operations. The program ended in 2000 as part of the country’s commitments at WTO.

46. However, a potential problem with the non-availability of export subsidies was raised in both the Indian and Indonesian case studies where countries operate stock-holding schemes for price stabilisation purposes. Outlays for such purpose are classified as GB expenditure in the AoA. In India, the inability to sell abroad at less than the domestic price has become a binding constraint in the light of its huge surplus stocks of cereals, which are currently much above the stipulated norms for buffer stocks. Holding such a high level of stocks is expensive, but the options are limited. Releasing these stocks in the open market will lead to a significant fall in prices, which may benefit consumers in the short run, but will have significant implications for future output growth and food security in the long run. Similarly, exporting at such low prices is also not feasible without resorting to direct subsidies which are not permissible.

47. Indonesia faced the same problem in the late 1980s and early 1990s, when there were occasions when BULOG needed to reduce stocks and exported surplus rice at less than the domestic price. In order to permit policymakers as much flexibility as possible in regard to the future disposal of stocks, Indonesia made a commitment on export subsidies but, as noted above, has not subsidized exports of rice since implementation of the AoA.

48. A number of case studies mentioned the importance of continuing the developing country exemption which allows them to grant subsidies to reduce the cost of domestic marketing and international freight. This was seen as potentially important in Senegal, for example, in view of the country's high transport costs, particularly for fruit and vegetables.

49. Most case studies identified that countries had a range of general export assistance schemes in place including, for example, tax rebates, duty drawbacks, tax exemptions, provision of subsidised export finance and export credit guarantees, which were also available to agricultural exporters. These schemes are not specifically referred to in the AoA but are addressed in the Agreement on Subsidies and Countervailing Measures (Annex 1: Illustrative list of export subsidies). As noted in the previous study⁷, it is not entirely clear whether it is legitimate to grant such subsidies on agricultural products (by referring to the Subsidies Agreement) when all forms of agricultural subsidisation (apart from the SDT provision) are prohibited by the AoA for those countries with zero export subsidy commitments. This is also an issue which requires clarification.

2.4 SPS Agreement – domestic obligations

50. The key principle underlying the SPS Agreement is that countries have the right to decide on the measures they deem necessary to protect human, animal or plant life or health. However, to prevent abuse, certain disciplines are applied. Measures should be based on scientific principles, should not be maintained without scientific justification, and should not be applied in an arbitrary or unjustifiable way.

51. Two implementation issues arise for developing countries from this Agreement. One is the extent to which they had in place, or have been able to put in place, standards consistent with SPS principles and the extent to which they have received technical assistance for this purpose. A second issue is whether trade barriers put in place on SPS grounds have been accepted by other WTO Members, or whether other Members have perceived that developing countries have used SPS barriers for protectionist purposes. A third issue, the extent to which SPS barriers in export markets have damaged developing country exports to those markets, is considered later in Section 3 under the heading of export market access conditions.

52. The case studies indicate that many sample countries have either upgraded or reorganised their national SPS systems in response to the introduction of the Agreement. There is greater awareness of the need for vigilance in protecting consumers from potentially unsafe food and of the need to protect plant and animal health. An example was the Egyptian ban in June 1999 on some food products from the EU that might be contaminated with dioxin which affected mainly meat, egg and dairy products.

53. The case studies reported only a very limited number of cases where SPS measures taken by developing countries had been challenged by other WTO Members. The Egyptian case study reported two SPS-related experiences regarding the importation of poultry products. One concerned a decree banning the importation of poultry parts because it was difficult to ascertain whether or not the imported parts came from poultry slaughtered in accordance with Islamic traditions. Indonesia used the same justification in banning the import of chicken legs from the US. The second Egyptian poultry case related to the specification of a maximum moisture content of frozen poultry of 5 percent, which was considered by some WTO members to be well below the average moisture content permitted in many other countries. Egypt’s prohibition of beef imports with a fat content greater than 7 percent was also queried by some WTO members. Fiji, which has had problems with imports of low-grade sheepmeat products from New Zealand which it believes poses a major health problem due to its high fat content, imposed a ban on mutton flap imports in 2000. The WTO legality of this ban has been contested by New Zealand meat exporters. However, the case study evidence suggests that such challenges to SPS measures taken by developing countries are still relatively rare.

2.5 TRIPS Plant Variety Protection

54. The case studies did not provide an in-depth analysis of the experience with implementing Article 27(3) of the TRIPS Agreement but a number of broad conclusions emerged. First, in a number of countries legislative action is still awaited and some case studies noted that the legal expertise and resources to develop and enforce this legislation is lacking. Second, the majority of countries where legislation has been introduced have opted to meet their obligations under this Article by a sui generis system, often deriving extensively from the 1991 UPOV Convention. The rights of farmers and local communities as well as the plant breeders are usually recognised under this legislation. Third, many of the case study countries have expressed an interest in exploring the potential benefits of this legislation to protect traditional knowledge and farmers’ rights. There is a growing awareness of the value of the stock of knowledge concerning traditional medicines and plants, know-how and customs and of the value of obtaining intellectual property protection in commercialising this. Related to this, some countries have worried that overseas companies could use TRIPS protection in their home countries to effectively appropriate into private (and foreign) hands this traditional knowledge. The Indian case study, in particular, documents the way in which the Indian government fought to regain access to particular knowledge which had been patented by overseas companies in Europe and America. Fourth, there is a concern in many countries whether the balance of advantage from TRIPS legislation will be to their benefit or not. While a number of the case studies acknowledge that intellectual property protection is essential for making technological advances in agriculture, there are concerns about the higher cost of bio-engineered seeds, the monopolisation of rents by multinational firms and about differential access of small and large farmers to these seeds leading to increased income inequalities and endangering food security at the household level. It will be important to keep these issues under review as more experience is gained about the operation of plant varieties protection legislation in developing countries in the future.

3. REVIEW OF AGRICULTURAL AND FOOD TRADE

55. The acid test of the AoA is whether it actually impacted on trade flows. Did the market access instruments, the disciplines on export and domestic subsidies and the attempt to achieve greater certainty and transparency in the operation of SPS rules lead to improved export opportunities for developing countries? Did the commitments undertaken by developing countries on tariff bindings and domestic support lead to increased food imports by these countries? Effective trade liberalisation would be expected to lead both to increased exports and increased imports, so it is also important to ask what happened to the overall agricultural trade balance in developing countries as a result of the Agreement. Has this improved or disimproved? If any of these consequences were observed, what were the impacts on food security both at national and household level in these countries?

56. While it is possible to quantify the factual position on agricultural trade flows pre- and post-Agreement, it is much more difficult to identify the Agreement’s impact on these flows. The methodological problem has already been discussed in the Introduction. Some means of isolating the effects of the AoA from the many other developments affecting trade flows which were taking place simultaneously is needed. This would be a complex exercise beyond the scope of the case studies in this volume. Instead, this review and the case studies adopted the simple approach used in the earlier FAO study. The average value of food and agricultural trade during 1995-2000 is compared with that during 1990-94, simply to see whether trade has grown or not. All the case studies reviewed the country experiences at a more disaggregated level, focusing separately on five or six major export and import products, and further identifying the source of the change in terms of volume and price effects. Some of the case studies also tried to relate the observed change in trade flows to particular factors, e.g. access to preferential markets, tariff changes, SPS/TBT measures and domestic productivity growth.

57. In this section, the overall experience with respect to total agricultural exports, total food imports, and the ratio of agricultural exports to food imports is examined. Individual case studies provide more details.

3.1 Agricultural exports

58. Table 5 shows that the value of agricultural exports in 1995-2000 was higher than in 1990-94 for 20 of the 23 countries, the increase ranging from 10 percent to 150 percent. For 7 of these countries (Uganda, Peru, Costa Rica, India, Sri Lanka, Indonesia and Côte d’Ivoire), exports rose by 50 percent or more. Of the three countries where there was no increase, exports remained flat in Bangladesh and the Fiji Islands, while Senegal was unique in experiencing a significant fall in export value (by 19 percent).

59. In the 10-year period prior to the AoA, exports had been increasing in 16 of the 23 sample countries and decreasing in 7 of them. In the post-AoA period, the number of countries where exports were increasing increased to 20 and the number where exports were falling fell to 3. Significantly, exports in the post-AoA period were higher than projected on the basis of the 1985-94 trend in 17 countries and lower in only 6 countries. The latter group includes Fiji Islands –19 percent; Guyana –2 percent; Jamaica –5 percent; Senegal –21 percent; and Thailand –5 percent. Thus, in only two countries was the negative deviation from trend really significant. However, no inferences can be drawn from these figures about the direct impact of the AoA for the reasons given earlier. Also, although the use of five-year averages is necessary in order to summarize a large amount of data, it may conceal important turning points. In some countries, for example Peru and Costa Rica, the upward trend peaked around 1997-98 whether as a result of the Asian crisis or for other reasons.

60. Further light on the export performance of the sample countries is obtained by examining the change in export values in constant prices, a measure of the contribution of volume changes to the change in export values.⁸ Here, a less favourable picture emerges. In 7 of the 23 countries, export volumes declined in the post-AoA period. The implication is that the improvement in export values in current prices in the post-AoA period was due, to some extent, to improved terms of trade rather than increased volume flows. The world market prices of many agricultural commodities were higher in 1995-2000 than in 1990-94. In five cases the terms of trade movement had a negative effect (Bangladesh, Brazil, Guyana, Malawi and Pakistan) although in the cases of Brazil and Malawi the differences are well within the margin of error in the comparison. More detailed accounts of these changes are presented in the individual case studies.⁹

Table 5. Total agricultural exports in 1990-94 and 1995-2000

61. What these aggregate figures conceal, but which emerges clearly from many of the case studies, is the extraordinary diversification in the basket of agricultural export commodities which many countries have achieved in the past decade. In Egypt, the share of cotton, which accounted for 60 percent of agricultural exports in 1985-89, had dropped to 30 percent in the post-AoA period. In Peru, the value of non-traditional exports exceeded traditional exports for the first time in 1996 and their share continues to increase. Senegal has diversified away from groundnuts towards cotton, fruits and vegetables, and hides and skins. In Zimbabwe, the pattern of exports has shifted from cereals to cash crops, particularly tobacco and horticultural products. In Uganda, diversification of new exports benefited from currency devaluation and regional trade liberalisation through COMESA. Fruits and vegetables, floriculture and farmed fish and shellfish exports have been among the most dynamic sectors in this diversification process. In other countries, such as Costa Rica, diversification has taken a vertical form, in the attempt to move downstream along the value-added chain. Attempts at both horizontal and vertical diversification, however, still face many market access barriers in importing countries.

62. Despite the general growth of agricultural exports in most countries, few case studies were able to make a link to improved market access under the AoA. In some countries, the improvement in exports in the post-AoA period was due to improved domestic conditions or world market conditions unrelated to the AoA. Brazil removed its export tax on soybeans in 1996 which had a large effect on exports of soybeans without any significant negative effects on exports of meal or oil, apparently due to significant increases in productivity in the soybean sector. Ugandan exports improved because of the increase in international coffee prices (which accounts for 70 percent of that country’s agricultural exports) and recovery from the ravaging of farms by coffee wilt disease. Some countries benefited from devaluation (Brazil after 1999) while, in others, the reason was improved weather conditions. In India, unilateral reforms which reduced manufacturing protection, introduced rupee convertibility on the trade account and shaped a relatively more open export policy, had a much larger impact on export of agricultural products than the AoA.

63. Where market access had improved, this appeared to occur under regional trade arrangements or as a result of preferential schemes (such as the growth in Zimbabwe’s horticultural exports) rather than resulting from the Uruguay Round. Some studies looked for evidence of export market diversification resulting from new market access opportunities but little hard evidence emerged. On the contrary, the Brazilian case study noted that, while the EU is still the largest market for Brazilian exports (54 percent in 2001), exports were growing more rapidly to other parts of the world, probably reflecting high protection in its traditional markets. One study emphasised the adverse implications of MFN tariff reductions for countries receiving trade preferences under the Generalized System of Preferences (GSP) (Egypt).

64. Continuing market access restrictions and subsidised competition were documented in a number of studies. Export subsidies provided unfair competition to a number of countries (India in the case of its cereals (wheat), dairy products and to a certain extent sugar exports). Further encouragement is given to competing exports by the very sizeable domestic subsidies provided in developed countries. Senegal groundnut oil faces competition from other oils, some of them subsidised. The Brazilian case study emphasises the potential damage arising from the new US farm bill which has increased domestic support particularly for commodities of interest to Brazil such as soybeans, cereals and cotton.

65. The importance of traditional market access barriers was emphasised in the Indian case study, among others. These include continuing problems of tariff peaks and tariff escalation, particularly in the case of non-traditional exports such as dairy products, fruits and vegetables and their preparations, meat and processed foods. It is worth noting, however, that the Indonesian case study, drawing on an earlier FAO study¹⁰ showed that the UR resulted in reduced tariff escalation for nearly all Indonesian agricultural products in the EU, Japanese and US markets, although the reduction was less for these export products of interest to Indonesia than for all agricultural products in general.

66. Problems in accessing tariff rate quotas were also highlighted in a number of the case studies. Unfortunately, very limited information is available on developing country experiences with TRQs opened by others, notably the developed countries. Statistical information on this issue is not readily available as it is traders who make the deals. However, judging by the consensus in the case studies where countries are looking for more transparent and non-discriminatory access to TRQs, there is a strong implication that developing country exporters believe they have difficulty in accessing them and that other exporters have more preferential access. This belief is not surprising given that a number of developed countries converted previous preferential access quotas into TRQs and there was no promise of additional market access.

67. There were relatively few documented cases of anti-dumping measures taken against agricultural exports of the case study countries. Indonesia faced anti-dumping actions on exports of tomato paste in Australia, sorbitol in the EU, and canned mushrooms in the United States. Further actions had been taken against its downstream products processed from leather and natural fibres.

SPS barriers

68. The majority of the 2002 case studies reported that SPS measures taken by importing country markets had adversely affected their exports. There was also evidence in a number of case studies that the number of such measures has steadily increased over time (in the case of Indonesia, for example, from less than 10 holding orders against its processed food exports to Australia in 1993-95 to 40 in 2001). In a number of cases, it was accepted that these measures were justified. Domestic measures were taken and, in some cases, technical assistance was sought, to overcome the problem. In Uganda, for example, the fisheries sub sector faced setbacks in 1999 as a result of fish poisoning and reports of unhygienic conditions at landing sites that led to a ban on fish exports to the European Union that lasted 18 months. Veterinary experts from EU worked with Uganda officials to rectify the handling of fish right from the lakes, fish processing factories and prior to export. Assistance was also extended to the fisheries department to improve fish inspection and surveillance. Thailand had a problem of VRE in frozen chicken exported to the Czech Republic and exports resumed once the problem was resolved. In these instances, problems include failures in domestic supervision, lack of awareness of importing country requirements on the part of exporters and the problem of consistency of domestic SPS standards with the standards of the importing countries. The Costa Rica case study noted the synergy between a well-developed national SPS system and success in penetrating export markets. As a result of a vigorous domestic enforcement, exports of Costa Rican products in foreign markets have a very low rejection rate for phytosanitary reasons.

69. Countries tend to have different rules regarding SPS restrictions such as inspection of imported products, specific treatment or processing of products, fixing of maximum allowable levels of pesticide residues or permitted use of certain specific additives in food. These flexibilities in the SPS agreement leave a lot of room for discretion. In addition, SPS standards are becoming increasingly complex. Sometimes, this leads to products being treated inconsistently in different markets. India faced a ban on the export of marine products to the EU in 1997 after some consignments were found contaminated with Salmonella and Vibrio Cholera bacteria. Yet marine products were being exported to the United States throughout the period when there was a ban imposed by the EU. In some cases, these problems arise because of the lack of mutual recognition of inspections and standards (with several large importing countries often asking for “sameness” in the process rather than “equivalence”). Thus, Philippines bananas which are accepted in Japan under the substantial equivalence principle are denied access to Australia. To persuade Australia to engage in dialogue on banana and pineapple exports, the Philippines suspended the issuance of veterinary quarantine certificates in 1999 for Australia’s live cattle exports. To date, the issue is not yet resolved pending the verification of the risk assessment by the Australian Government regarding banana exports.

70. Foot-dragging on the part of the SPS authorities in the importing countries in approving processing facilities for export appears to be a further problem. The Thai case study noted that poultry exports to the Philippines and Australia had been prevented by delays by the relevant import authorities in making their reports. Fiji lost its horticultural markets when the chemical EDB was no longer accepted as a quarantine treatment in 1990. Fiji acquired high temperature forced air (HTFA) quarantine treatment technology to address this problem. Today, Fiji has a viable industry-operated quarantine treatment facility and a thriving industry in the export of fruit fly host commodities. Unfortunately, Fiji’s initiative in adopting the necessary technology to facilitate exports has not been matched by the regulatory authorities in the importing countries which have yet to grant import permission for Fiji fruit exports.

71. There were also a number of documented examples where SPS measures appeared to be arbitrary and did not appear to be justified. Brazil, for example, faces restrictions on the export of tropical fruits mainly in the US and Japan due to the existence of the fruit fly. Costly procedures are implemented to ensure fruits such as mangoes and papaya meet the standards of these countries. There are cases where the costs are unduly increased due to unreasonable requirements, such as the one that requires a USDA employee to supervise production locally at the expense of domestic producers or traders. Thai food exports to the US experience a greater proportion of detentions than any other supplier to the US relative to their import share with the exception of India. The case study acknowledges that there are problems in determining whether the detentions and complaints against food imports from developing countries reflect real SPS problems or are simply NTB’s under the guise of SPS and TBT measures.

72. In a number of cases, lack of resources and technical expertise in the SPS authorities of the developing country have been a problem in meeting SPS standards in importing country markets. In the majority of case study countries the required technology to do basic testing and certification is not available. In the case of the Fiji horticultural exports, the local industry complains that the quarantine authority has been slow in supplying data required by importing countries, and when it is finally supplied it is not presented in the required format. However, even where measures were accepted as justified, developing countries face problems when regulations change because of the need for investment. Research reported in the Thai case study showed that the tariff equivalent of SPS barriers to Thai exports averaged 29 percent, varying from 4 to 55 percent. A major technical assistance and training program accompanied by financial support is required to ensure the necessary structures are put in place.

73. Despite the importance of continuing trade barriers, in some countries domestic constraints led to a failure to exploit market access opportunities. Among the constraints facing Egyptian exporters attempting to increase sales abroad are low-quality domestic inputs, cumbersome duty–drawback and temporary admission regimes, excessive paperwork, fees and delays for customs and various inspections during export and import; poorly-trained workers; insufficient incentives to export; lack of access to information on foreign markets and product standards. The Malawian case study emphasised the constraints posed by the lack of market and physical infrastructure as well as investment risk and poor access to credit and extension services in explaining the poor trade performance of that country. Export performance in Indonesia has been disappointing, particularly following the huge devaluation of the rupiah after 1997. International commodity price declines and the structural collapse of Indonesia’s trade finance system were partly responsible. In Honduras, the combination of Hurricane Mitch, high real interest rates and low agricultural prices due to low world market prices and also exacerbated by an overvalued exchange rate has produced a considerable financial crisis in the agricultural sector. Investment in Philippines agriculture has been limited in recent years because of the carry-over of the 1997 financial crisis and the political and economic structural adjustments taking place. These experiences from the case studies emphasise the importance of appropriate domestic policy regimes if exporters are to take advantage of new trade opportunities.

74. In summary, almost none of the case studies draw a link between improved export performance and the AoA. Many find explicitly that neither the composition nor volumes of trade in agricultural products have been influenced significantly by the implementation of the AoA. This does not necessarily mean that the commitments obtained by the developing countries from their trading partners in the UR were valueless. There were difficulties in estimating both the significance of the tariff reductions which were made as well as the greater certainty and transparency of market access in assisting the export growth which has been observed. As export volumes grow, the AoA provisions enhance security of access to these markets, particularly for the newer export commodities arising from diversification where competition with producers in developed countries is often greater than in the case of the traditional export commodities of developing countries.

3.2 Import performance

75. Expenditure on food imports increased significantly in nearly all countries in the sample comparing the pre- and post-AoA periods, with the exception of Malawi and Zimbabwe (Table 6). Food imports more than doubled in value from 1990-94 to 1995-2000 in six countries (Costa Rica, Honduras, India, Indonesia, Philippines and Uganda); indeed, the Ugandan figure increased more than three times. As in the case of agricultural exports, it is possible to break down the contribution of increased import volumes and prices to these increases in expenditure. With the exception of four countries (Côte d’Ivoire, Fiji Islands, Malawi and Thailand), where the growth in the volume of imports was greater than the corresponding growth in food import expenditures, countries generally paid more per unit for their food imports in 1995-2000 than they did in 1990-94. Many of the case studies provide further information by disaggregating the sources of the higher food bill by commodity, distinguishing volume and price changes.

76. Food import expenditures were increasing in all countries in the decade 1985-94 with the exception of Bangladesh, Egypt and India. Food import bills in these three countries, measured against their declining trends, increased by even more than shown in Table 6 (by 77 percent for Egypt, 127 percent for Bangladesh and by 218 percent for India). For the majority of countries where import growth in the post-AoA period was a continuation of previous import growth, the question is whether import expenditures in the post-AoA period were above trend or not. Including the three countries mentioned earlier, imports in the post-AoA period were higher than projected on the basis of the 1985-94 trend in 15 countries and lower in 8 countries. As in the case of agricultural exports, it is not possible to draw inferences about the direct impact of the AoA from these figures. As discussed in Section 2, import liberalisation was under way in many countries for other reasons.

77. Inferences from this growth in food imports can be either positive or negative. If food imports add to domestic production and to domestic food availability, the consequences for food security may be positive. The consequences may also be positive even where food imports displace domestic production, provided that those displaced from food production are absorbed into more productive employment opportunities in the nonfarm sector. Negative impacts can be anticipated where food imports displace domestic food production and where those displaced remain unemployed or underemployed. This issue is pursued in the following section where the food security impacts of the AoA are discussed in greater detail.

3.3 Food imports relative to agricultural exports

78. As both agricultural exports and food imports were increasing for most countries in the sample in the post-AoA period, it is important to ask what the agricultural trade experience was on balance for these countries. As in the previous FAO study, this question can be answered by reviewing the trend of total food imports to total agricultural exports. An increase in the ratio indicates that food imports are growing faster than agricultural exports, and vice versa. This ratio is used to summarize the relationship between the two variables most commonly discussed in the context of the AoA. It is not intended to indicate a country’s food import capacity, which is often described in terms of the ratio of food imports to total exports, including services.¹¹

79. Table 7 shows that 14 of the 23 countries have a food ratio less than 1, indicating that agricultural export earnings exceed food import expenditures; in the remaining 9 countries, food import expenditures exceed agricultural export earnings (Peru’s ratio would fall below one if exports of fish products were included). However, the ratio was higher in 1995-2000 for 17 of the 23 countries compared to 1990-1994. The exceptions were Côte d'Ivoire, Guyana, Malawi, Peru, Sri Lanka and Zimbabwe. Many countries show large percentage increases in the ratio but often from a ratio smaller than 1. Bangladesh stands out as a country where food imports were four times the value of agricultural exports in the earlier period, and over eight times their value in the post-AoA period. The ratio for Senegal also increased from two to three-and-a-half times. The rising ratio of food imports to agricultural export earnings in most countries continued a trend evident in the pre-AoA period. The ratio was on a rising trend during the 10-year period 1985-94 in all countries with the exception of India and Jamaica, though in both these countries the ratio began to rise after 1994.

Table 6. Value of food imports in 1990-94 and 1995-2000

Table 7. Ratio of the total value of food imports to the total value of agricultural exports

3.4 Stability

80. Not many case studies commented on experiences with respect to the stability of exports and imports. Many commentators expected that implementation of the AoA, and especially tariffication, would lead to greater stability in world market prices. This might be expected to contribute to greater stability in export revenues and food import expenditures, although changes in quantities also contribute to instability in these flows. Figures for India do show greater stability in exports and, in particular, imports comparing the pre-AoA and post-AoA periods, but there was not a sufficient number of country experiences reported in the case studies to draw any general conclusions on this issue.

4. FOOD SECURITY IMPACTS

81. For all countries in the sample food security is an important objective. Thus, it is important to examine the effects of the AoA on the food security status of developing countries. Has implementation of the Agreement facilitated countries in achieving greater food security, or has it made reaching this objective more difficult? The same methodological problem observed in the case of agricultural trade flows, that it is difficult to isolate the impact of the Agreement from the many other factors at work which influence food security, makes it difficult to reach definitive conclusions. A further factor is that other aspects of the Uruguay Round Agreement, for example, the Agreement on Textiles and Clothing and the overall reduction in protection to manufactured goods, will also have impacted on food security through their overall impact on net trade earnings, employment and livelihoods.

82. One dimension of the food security issue is shown by the percent of the population undernourished in the sample countries (Figure 1). This varied from 46 percent of the total in Kenya to just 4 percent of the total in Egypt in the period 1997-99. In 12 of the 22 countries for which data were available (Fiji was the omission), more than 20 percent of the population were undernourished in the second half of the 1990s.

83. However, the situation has been improving in the majority of countries. In 18 of the 22 countries, the proportion of the population undernourished fell between 1990-92 and 1997-99. Four countries (Botswana, Morocco, Senegal and Uganda) experienced an increase in the proportion undernourished, although in the case of Morocco and Senegal this was by one percentage point. Thus, this dimension of food security has improved in the great majority of the sample countries in the AoA implementation period. It is not possible to isolate the impact of the AoA in contributing to this improvement; indeed, in view of its limited trade effects documented in Section 3, the AoA is unlikely to have been very significant in the period under review. However, the situation has been improving in the majority of countries, even though it is not possible to isolate the impact of the AoA in contributing to this situation.

84. It is possible to make a more direct link between the AoA, trade flows and food security by looking at trends in food availability in the sample countries over time. While the prevalence of undernutrition will be influenced both by average food availability and by the distribution of available food, average food availability plays the crucial role. Figure 2 shows that 90 percent of the differences between the sample countries in the proportion of their population undernourished can be explained by differences in average per caput calorie intake. Figure 3 shows a more demanding test, by comparing the improvement in the prevalence of undernutrition with the improvement in average per caput calorie availability. In this instance, 87 percent of the change in the percentage point prevalence of undernutrition can be explained by changes in per caput food availability. Based on the experience of the sample countries during the 1990s, for every 100 calories improvement in per caput calorie intake, the prevalence of undernutrition fell by 6.5 percentage points.

85. Given the close link between the prevalence of undernutrition and per caput food availability, it is not surprising to find that the pattern with respect to trends in per caput food availability is similar to that discussed above for trends in the prevalence of undernutrition (Table 8). Twenty of the 23 countries experienced an improvement in per caput calorie availability between the immediate pre- and post-AoA periods, the exceptions being Botswana, Senegal and Uganda. Morocco was the only country where per caput food availability increased at the same time as the prevalence of undernutrition. Table 8 also shows the trend in per caput cereal availability. This might be preferred as an indicator of the food security situation of poorer income groups because staple foods are likely to make up a higher proportion of their food consumption. On the other hand, in some countries, roots and tubers may make an important contribution to the diet of the poor but such consumption is not captured by this indicator. Per caput cereal supply increased in all countries between the pre- and post-AoA period except for Bangladesh, Honduras and Senegal.

86. From a food security viewpoint, it may be significant whether the increase in food availability per caput is provided by domestic production or imports. A situation where increased per caput food availability is due to increasing domestic production accompanied by falling imports, or even by increasing imports, may be seen as strengthening food security. On the other hand, a situation where per caput food availability is decreasing because of falling domestic production, or even where it is increasing but only because per caput food imports are growing even more rapidly, may be seen as unsustainable in food security terms in the longer term without a concomitant rise in export earnings.

87. However, it is very difficult to draw a clear interpretation of the consequences of changes in the composition of food supplies between domestic production and imports for food security without more information on the specific country context. For example, domestic production may be growing and displacing imports only because a country is increasing protection to its agricultural sector. On the other hand, a country which experiences falling production and a growing dependence on food imports may simply be shifting resources to non-farm sectors such as tourism or light manufacturing where growth and employment opportunities may be more attractive.

Table 8. Trends in per caput food availability

Source: FAOSTAT; FAO State of Food Insecurity in the World 2001.

88. At least with respect to cereals, some of the case study countries appear to have become more trade dependent in the last part of the decade. This is also evident by examining trends in self-sufficiency ratios (SSRs) for key staple products (Table 10). The cereals SSR declined between 1990-94 and 1995-2000 in 12 out of 23 countries, the vegetable oils SSR in 16 countries, the sugar SSR in 13 countries and the meat SSR in just 7 countries. Individual countries have not necessarily become more dependent on imports for their food supplies in the post-AoA period and there is clearly a wide variety of specific experiences. The individual cases studies explore these differences in more detail.

89. Whether there is evidence of imports displacing domestic production is more difficult to say. There is indeed a small negative correlation between increases in imports and increased domestic production (R² = 0.25), implying that, as imports increase, domestic production falls. However, it is not possible to assert a causal relationship on the basis of this simple correlation. It is equally plausible that the negative correlation is observed because a shortfall in domestic production (due to poor weather, for example) means that a country turns to imports to secure sufficient food supplies. In this case, additional imports are a sign that food security has been strengthened, not diminished.

Table 9. Self-sufficiency ratios for some major food commodities

Source: Computed from FAOSTAT.

Table 10. Indicators of food import capacity

Source: FAOSTAT; World Development Indicators CD-Rom 2001

Dependence on food imports can flag a food security problem at the national level if a country is likely to have difficulty in paying for these imports. A useful indicator of food import capacity is the share of food import expenditure in total exports of goods and services. Because for many developing countries a proportion of foreign exchange earnings is earmarked for debt service and thus not available to import food, the ratio can also be expressed as the share of food import expenditure in total exports of goods and services minus debt service payments.

90. Food imports accounted for 12.4 percent of total exports of goods and services of the sample countries taken together in 1985-89, for 11.5 percent in 1990-94 and for 11.6 percent in 1995-1999. The trend is clearer when debt service payments are taken into account. The share of food imports in the three time periods is then 17.1 percent, 15.6 percent and 15.0 percent. Thus, the capacity to pay for food imports has improved for sample countries taken together over the past decade, despite a gradual hardening of the prices of those foods most often found in the import basket of developing countries (Table 11).

Table 11. Nominal world prices (US$/tonne) for some food commodities

Source: FAO data.

91. Nonetheless, this average statement is not necessarily true for all countries in the sample. For 14 of the 23 countries in the sample, the burden of paying for food imports has risen in the 1995-99 period compared with 1990-94 (Table 10). Countries where the share of food import expenditure in total export earnings (including debt service) has been increasing include Botswana, Brazil, Honduras, Indonesia, Kenya, Senegal, Uganda and Zimbabwe. For the low-income countries in this group, food aid flows can be important in helping to ensure increased food availability at the national level. Food aid flows to the sample countries, however, have been falling. Many of the individual case studies noted that the Marrakesh Decision on Measures Concerning the Possible Negative Effects of the Reform Programme on Least-Developed and Net Food-Importing Developing Countries has not been implemented, and called for more binding commitments to be included in a future Agreement.

92. The most likely groups to benefit from the reduction of trade barriers in foreign markets and the expansion of exports are commercial producers. Small farmers may not be able to participate in growing export-oriented crops and may experience greater competition in accessing resources, including land, marginalizing their position even further. The winners and losers of open trade policies are likely to be different and it is feared that it is often the poor who get hurt most. For example, the Peruvian case study noted that fresh and processed asparagus which emerged as the most dynamic export crop in the 1990s are produced mainly by large farms. Greater export opportunities may also lead to the re-allocation of land and other resources away from domestic food production, with possible adverse consequences for household food security.

93. Whether the position of the poor worsens or not under more open agricultural trade policies, however, depends on much more than what happens in the agricultural sector alone. For example, the availability of non-farm employment in rural areas, the functioning of rural labour markets and the impact of production and trade policies on food prices will all determine the way income distribution, poverty and thus food security is affected. The Peruvian case is also instructive here. In Peru, the incidence of poverty decreased between 1990 and 1994, but increased again in 2000 due to economic recession. However, the proportion of extreme poor, which is deemed a more appropriate indicator of the number of food-insecure people, continued to fall even during the latter period. The case study concludes that any potential negative effects of agricultural trade trends on the rural poor were offset by positive developments in non-agricultural employment and in social welfare interventions. The National Food Program agency is seen as having played an important role in upgrading the social welfare of the poor in both the cities and rural areas of Peru in the 1990s.

94. Trends in the incidence of poverty and the number of households in poverty may be the best indicator to capture the totality of effects on the food security of low-income households. The incidence of poverty has declined in a number of the case study countries, including India, Jamaica and Costa Rica. Poverty indicators also improved during the 1990s in Honduras. In the first half of the decade, the positive impact was greater in rural than in urban areas, consistent with the view that structural reforms were redressing the anti-agricultural bias of previous macroeconomic and sectoral policies. In the second half of the decade, the drop in urban poverty continued but halted in rural areas, while the proportion of extreme poor increased. Hurricane Mitch appears to be mainly responsible for this trend. Evidence also suggests that the proportion of poor households increased in Senegal and Egypt in the second half of the 1990s. In Uganda, poverty has declined although there has been little improvement in income distribution, and income gains in rural areas accrued to relatively few households. There is conflicting evidence from the Philippines, where the head count index of poverty increased in the 1997-2000 period but the family count index decreased albeit in the earlier period 1985-1994. To understand these trends in poverty rates and the reasons for differences between countries would require a much deeper analysis than that undertaken in the case studies.

5. NEGOTIATING ISSUES AND CONCERNS

95. The country case studies contain a concluding section which discusses the negotiating issues and concerns. These reviews draw at times on the official submissions to the WTO Special Session of the Committee on Agriculture by the countries themselves, and in other instances on a more general discussion of the right and obligations of developing countries vis-à-vis the developed countries under the AoA. While in this overview chapter it is neither possible nor appropriate to summarise the detail of the individual country positions, some general points can be made in conclusion.

96. Gaining improved market access, particularly in the developed countries which historically have had the highest levels of trade-distorting support to domestic agricultural production, and removing the unfair competition caused by export subsidies remains a key objective in the current negotiating round. In most developing countries, the import substitution strategies of the past have been replaced by development policies which emphasise the gains from integration into the global economy. However, these gains will not be realised if trade barriers continue to inhibit and restrict the growth of sectors where developing countries have a comparative advantage. The difficulties are clearly seen in the Egyptian case study. With improved market access, Egypt would better adopt a strategy that is based on allocating increasing amount of resources to high-value exportable horticultural products in which Egypt has comparative advantage. However, limited market access improvements will push Egypt to adopt an import substitution strategy that focuses on food production which implies a less efficient use of agricultural resources. Protectionism in the EU [as Egypt’s major market] is not an incentive to modify its position on food self-sufficiency as the means to food security.

97. Under the market access and competition headings, the particular problems raised in the case studies include tariff peaks on the export products of interest to developing countries; tariff escalation which, although reduced in the UR, remains a deterrent to increasing value added food processing activity in developing countries; increasing use of SPS measures and long delays in recognising the equivalence of SPS measures put in place in developing countries; the problem of trade preferences; the need for larger TRQ volumes and more transparent administration of access to them; and the need to dismantle export subsidies.

98. Most of the case studies argue the case for an appropriate safeguard mechanism against low import prices and import surges, in order to allow for further trade liberalisation while guarding against incurring unduly high social costs. Such a mechanism is also justified as a means to protect against the effect of high levels of subsidies and protection to agriculture in developed countries, which depress and destabilise world prices and create unfair competition on both international and domestic markets. There is a widespread perception that the general GATT safeguards are too difficult and cumbersome to be used because injury must be shown. A modified version of the Special Safeguard provision, to be available only to developing countries, is seen as a more effective alternative.

99. The other concern raised particularly by food-importing countries is the fear of higher world prices and price volatility leading to upward price risk for importing countries. The Marrakesh Decision is widely seen as ineffective and some strengthening of its provisions will be essential to gain the support of low-income food-importing countries in the current negotiations.

100. Many of the case studies pointed to the need to ensure that AoA rules did not prevent developing countries from protecting and supporting domestic food production. This concern partly arises from the perception that a high level of food self-sufficiency, particularly in staple foods, was necessary for national food security. It also arises from the fact that the bulk of poverty is found in rural areas and there is thus the need to foster growth and employment opportunities in rural areas and agriculture as part of a food security strategy. Because small producers are often the most vulnerable in the process of trade liberalisation, some kind of targeted support mechanism to increase small farm productivity and to stimulate rural economic diversification may also be required to prevent the marginalization of low income producers. A third argument is one of equity; that the AoA rewarded developed countries for past heavy subsidisation by permitting them to continue this practice, but now withholds from developing countries the same rights, simply because they were too poor to finance such subsidies in the past.

101. The evidence from the case studies suggests that WTO disciplines have not proved constraining to the domestic support policies that developing countries want to implement. Budgetary constraints and previous commitments under SAPs appear to be much more important in limiting these interventions. However, many of the case studies indicate that developing countries are uneasy about the implications of the current disciplines for future domestic support options. Many developing countries want greater flexibility than the current provisions in the AoA allow.

102. Some historical perspective on trends in support to farmers drawn from the experience of the developed countries may be helpful in examining this concern. As countries develop, higher income levels are associated with greater levels of transfers to the domestic farming sector. There are both economic and political reasons why this has been the case historically. Economic growth is associated with a shift in employment opportunities from the farm to the non-farm sectors, and from rural to urban areas. This shift is associated with a growing disparity between average non-farm and farm incomes and, ultimately, with a decline in the absolute size of the farm labour force. Faced with declining relative incomes, farmers have an incentive to organise and to lobby politically for income transfers. At the same time, economic growth means that there is greater scope for the non-farm sector to bear the cost of growing transfers to farmers. Economic growth also means a change towards more meat-based diets with greater indirect consumption of cereals, often leading in the case of net-food-importing countries to a deterioration in the agricultural trade balance. Concerns that declining food self-sufficiency ratios imply diminished food security combine with equity concerns to raise the willingness of the non-farm population to contribute to farm transfers. As the farming sector declines further in size, farmers become more effective in their lobbying efforts while the cost of transfers to the increasingly numerous non-farm population declines in per capita terms. The result is escalating levels of farm support as a country’s per capita income grows. It is likely that the same economic and political forces which have been at work in the developed countries will be at work, particularly in the middle-income developing countries, as their per capita income levels rise.

103. Evidence from the case studies shows that financial constraints in many developing countries, and especially low-income developing countries, mean that these countries will be unlikely to be able to make use of additional flexibility for domestic support outlays. The more likely users of additional flexibility are middle-income developing countries where industrialisation has taken hold and where the absolute numbers engaged in agriculture are falling. Given these likely pressures for increased farm transfers in the future, particularly among middle-income developing countries, the question for developing countries is whether to accommodate these pressures by seeking greater flexibility to provide production-related farm transfers. There are important lessons to be learned from the developed country experience where the scope for protection tends to be captured by producer groups and is not necessarily in the overall national interest.

104. An alternative strategy, more favoured by low income developing countries, is to retain or increase the flexibility to take border measures to protect domestic food production. Higher tariffs to protect domestic food producers do not incur government expenditure, indeed they may contribute to government revenues. There is a permanent tension between raising prices to benefit rural food producers and lowering food prices to enhance the food security of food consumers. The benefits of price support tend to go predominantly to the larger, commercial producers who produce the bulk of the marketed food supply. Many of the rural poor are agricultural workers and net purchasers of food. Indonesia provides a dramatic example where the incidence of poverty, having fallen steadily over the two decades 1976 to 1996, rose sharply as a result of the Asian crisis only to drop back to its pre-crisis level by 2000. Declining food prices, mainly driven by lower rice prices due to trade reform, accounted for a large proportion of that improvement. In some cases, the elimination of over-valued exchange rates would increase incentives for domestic food production (as in Egypt recently) without introducing the distortions associated with a tariff policy, but special measures to help poor consumers following devaluation may also be taken. Future tariff policy also needs to take into account regional integration commitments undertaken by individual countries.

105. A number of the case studies raise specific issues for clarification in the agricultural negotiations other than those mentioned above. Examples of such issues include the status of irrigation subsidies, the methodology used to calculate the market price support element of the AMS and the position of the Least Developed Countries. Some countries would like AoA rules clarified to permit positive PS AMS levels, or positive NPS AMS levels, to be offset by negative PS AMS levels. Such a change would only be relevant to countries which have a bound AMS commitment. For countries whose AMS levels in future must be limited to “exempt” categories, and these are the great majority of developing countries, the offset possibility would not be meaningful because de minimis requirements cap the permitted AMS level of 10 percent of the value of production of each individual commodity, regardless of how high or low AMS levels are for other commodities.

106. As the negotiations proceed and the particular negotiating trade-offs become clearer, developing countries will need to undertake specific and detailed analyses of how they might be affected by individual issues under negotiation. Some countries have a need for technical and financial assistance to improve their capacity for policy analysis. A number of the case studies highlighted the inadequacy of the policy-making machinery in many countries to address AoA issues and the need for improved coordination, including the involvement of civil society organisations in the formulation of national negotiating positions. Also, many developing countries are negotiating regional integration arrangements simultaneously with multilateral trade liberalisation. The need for developing countries to assess and update their AMS levels, not just to comply with the WTO process but as part of the domestic policy analysis process, was stressed in some studies. While some countries had benefited from technical assistance under the Joint Integrated Technical Assistance Programme and the Integrated Framework for Trade Related Technical Assistance to LDCs, there is a continued need for technical and financial assistance to improve the capacity for policy analysis, particularly in the least developed countries.

¹ This paper has been prepared with the assistance of Alan Matthews, Trinity College, Dublin, Ireland. The structure of the chapter follows closely that of the similar overview chapter prepared by the FAO Secretariat in the earlier volume of case studies (FAO, 2000). The assistance of the FAO Secretariat in providing data and other assistance in preparing this chapter is gratefully acknowledged.

² FAO, 2000. Agriculture, Trade and Food Security. Vol. II: Country Case Studies. Rome.

³ Although there will be a temptation to extrapolate the conclusions to developing countries as a whole, the lessons are drawn from these 23 cases and should be treated accordingly.

⁴ FAO is undertaking a parallel project on trade and food security, which will investigate this topic in much greater depth.

⁵ Country models can be constructed in a variety of ways, including synthetic multi-commodity market models, econometric agricultural sector models, computable general equilibrium models and mathematical programming models. The methodologies differ in the maintained assumptions about the behaviour of economic actors and in their data requirements.

⁶ FAO, 2000. Agriculture, Trade and Food Security. Vol. II: Country Case Studies. Rome.

⁷ FAO, 2000. Agriculture, Trade and Food Security. Vol. II: Country Case Studies. Rome.

⁸ Note the changes in export values in constant prices are calculated for the 1995-99 period while the changes in export values in current prices are calculated for the 1995-2000 period.

⁹ Trends are also influenced by the time periods used. For example, Indonesia experienced a deterioration in its export terms of trade between 1994-96 and 1998-2000, even though the experience between 1990-94 and 1995-2000 was positive.

¹⁰ Lindland, J., The Impact of the Uruguay Round on Tariff Escalation in Agricultural Products, FAO,1997.

¹¹ Some evidence on food import capacity ratios drawn from the case studies is presented in the next section.