Forum Syd
GLOBALA STUDIER - NR 5 (ENGLISH VERSION) - NOVEMBER 2000

Peter Einarsson
A Report from Church of Sweden Aid, Forum Syd, The Swedish Society for Nature Conservation and the Programme of Global Studies

© Forum Syd and the author, November 2000
Editor: Johanna Sandahl
Publisher: Forum Syd, Box 15407, 104 65 Stockholm, Sweden
tel: +46-8-506 370 00, fax: +46-8-506 370 99
e-mail: forum.syd@forumsyd.se
website: http://www.forumsyd.se
ISSN 1404-7845

What would agricultural trade policy need to look like if food security and ecological sustainability were the guiding principles for its design, rather than economic growth and comparative advantage?

A number of individuals and organisations in the global NGO community have grappled with that question for the past few years. They have produced a sizeable body of analysis and ideas which is beginning to form a coherent alternative to conventional wisdom, as expressed mainly in the present WTO Agreement on Agriculture.

This paper originated as an internal report for three Swedish development/environment NGOs earlier this year. The task was to review and summarise international NGO thinking about the AoA to date, and to attempt, if possible, some synthesis. This English language version is somewhat expanded and revised, in particular as regards the conclusions, but in the main retains that format.

Thus, it should be emphasised that the text is only to a very limited extent original work. As will be obvious to the reader, it leans heavily on those who have spent more serious time and effort. The list of references, although limited to material directly cited, will give a good indication of who they are. For orientation, there is also a brief guide to the NGO community monitoring the AoA, with contact information.

The reader will need some previous knowledge of WTO and AoA basics, but familiarity with technical terms and details is definitely not presupposed. For those who want to go a little deeper into the details I would recommend Kwa & Bello (1998) and Murphy (1999b), or for those who read French, even more Van Der Steen et al (1999).

It should be emphasised that the scope of this paper is quite strictly limited to the discussion of options for the renegotiation of the Agreement on Agriculture. It does not purport to be a general analysis of agricultural trade or even of WTO influence on agricultural trade. Among the many important things it does not cover are

• other relevant WTO agreements, such as TRIPS and SPS
• the history of AoA negotiations and the politics behind its present content
• the role of regional free trade agreements and IMF structural adjustment programs
• the role of other intergovernmental organisations such as the OECD
• the deterioration of terms of trade between agriculture and industry
• national agricultural policies and politics.

The most questionable consequence of this limitation, pointed out by several of those who reviewed earlier drafts, is that the role of transnational corporations is not directly addressed. The AoA, like all the WTO agreements, works on the assumption that trade takes place between member states, or at least between private companies based in member states. The reality, that global trade increasingly takes place in the form of internal transactions within transnational corporations, is not explicitly acknowledged.

While it can be debated at length to what extent the long term objective of the WTO, completely free trade in all goods and services, would benefit different groups of member states, it is beyond discussion that it would unambiguously benefit transnational corporations. They are in fact the only entities specifically organised to deal with that situation. A borderless world would no doubt further increase their decisionmaking power at the expense of national governments. It is also a public secret that their influence on WTO negotiations, through intimate relations with a number of member states, has been considerable. It may indeed be more accurate to describe the WTO process of trade deregulation as a power struggle between transnational corporations and national governments as a collective, than as a negotiating process between different groupings of national governments.

Yet, for the purpose of this study, I would claim that there is a valid argument for accepting the self-understanding of the WTO and its member states. Although partly fallacious, it does reflect the fact that WTO agreements can formally speaking regulate nothing else than relations between member states. The activities of transnational corporations can be disciplined only to the extent that it is possible by the intermediary of national legislations. Furthermore, it remains true that member states do perceive the negotiation process as one of defending their national self-interest in relation to other states. It is especially true regarding the AoA, because food and agriculture, quite correctly, are seen as crucial national issues by the vast majority of countries. The internal dynamics of the AoA negotiations, which are the focus of this paper, can be understood only in this perspective.

I owe thanks to a number of people from whose insights and critical minds I have benefited in writing this paper. Without listing names, I want to mention in particular the staff of the commissioning organisations, and the group of colleagues and friends who reviewed drafts.

I would also like to clarify that although the general line of argument contained in the paper does reflect the thinking of the three commissioning organisations, the details of the analysis as well as the conclusions and policy proposals are the responsibility of the author alone.

Björkeryd, 13 November 2000
Peter Einarsson
Email: peter.einarsson@ekolantbruk.se
Ph/fx: +46 477 401 60

The basic assumption of this paper will be that food security and ecological sustainability are the two most important policy goals for global agriculture. It will be assumed that feeding the world population, and doing so in a manner which does not compromise the productive capacity of biological systems, are absolute requirements which have priority over all other concerns.

This point will not be further argued here, but some indication will be given of what is understood by each term.

Food security and sustainable agriculture are two concepts with several shared characteristics:

• They enjoy virtually universal support. It would be very difficult to find anyone who would speak up for food insecurity or unsustainable agriculture.
• Definitions consequently vary widely. There are several hundred different ones for each concept, ranging from the inclusive to the devious.[1]
• Yet, each concept has a core which cannot easily be reduced. Differences in definition and interpretation are to a large extent differences in the degree to which that core is taken seriously and allowed to override other concerns.

Sustainable agriculture is a concept closely related to sustainable development. In very general terms thus, it is about satisfying present needs without compromising the possibility for future generations to satisfy theirs.

The core of almost all sustainable agriculture definitions are two requirements which flow from this general principle.

• Preserve productive capacity of natural systems.
• Minimise use of non-renewable resources.

Both requirements, if taken seriously, have very far-reaching consequences for the design of agricultural production systems, and both are routinely disregarded by almost all agriculture today. Any degradation of soils or water, for example, is in contradiction of the first requirement, as is all damage to other ecosystems which provide 'services' to agriculture, such as forests regulating rainfall patterns. The second requirement would virtually exclude any use of fossil fuels, and severely limit all other non-renewable inputs such as mineral fertilisers.

Corresponding to the enormous diversity of agricultural production systems, there has developed a very promising diversity of approaches to the implementation of those principles, which will not be entered into here. One aspect of particular relevance to agricultural trade must however be mentioned.

As a general rule, the two core sustainability requirements are increasingly difficult to fulfil the further physically removed food production and consumption are from each other. The main reason for this is that to maintain high levels of production with minimal non-renewable inputs, agriculture must be tailored to optimise use of locally available resources. In particular, crop diversity and a high level of recirculation of plant nutrients and organic matter are essential. Thus, the regional specialisation which is the central principle of classical free trade theory is difficult if not impossible to reconcile with ecologically sustainable agricultural systems, at least on a large scale and in staple products. In addition, the energy cost of transports adds to the unsustainability.

It should also be emphasised that ecological sustainability is absolutely non-negotiable. Ecosystem degradation leads to reduced production capacity. Once started, this process is extremely difficult, often impossible, to reverse.

It is sometimes argued that developing countries should be allowed certain exemptions from sustainability requirements, having for a long time contributed much less to unsustainable practices. There is very little merit in this argument. Ecosystem degradation, especially in agriculture, is in no way a less serious problem in developing countries. On the contrary, tropical ecosystems are often less resilient to misuse than temperate ones. Furthermore, the food production of developing countries is even more dependent on well-functioning ecosystems than that of developed countries, which are able to partly compensate for degradation by increasing inputs. As regards non-renewable resource use, there may be a limited case for allowing a temporary increase in developing countries, provided it is balanced by a greater decrease in developed countries. But the risk is obvious that the developing world ends up caught in the same unsustainability patterns as the developed world already suffers from.

Many sustainable agriculture definitions also include economic sustainability as a criterion. While the economic viability of new, sustainable practices obviously is a prerequisite for their implementation, it is misleading to include this criterion on the same level as ecological sustainability. What is ecologically sustainable or not is decided by objective factors, by biological and geophysical processes entirely independent of human judgement. Economic sustainability is created by human social organisation and policy choice. The conditions for economic sustainability of ecologically sustainable agriculture can and must be created by political and economic decisions, including notably international trade policy decisions. The opposite is not possible. No matter how economically profitable it is, an ecologically unsustainable agricultural production system cannot be made sustainable by a decision to change the ecological conditions. The notion that the two interact on the same level gives rise to the fallacy that they need to be balanced against each other, and in particular that ecological sustainability requirements may have to be reduced not to threaten economic sustainability.

To avoid any misconception of this kind, this paper will consistently use the term 'ecological sustainability'.

In contrast to sustainable agriculture, food security has a broadly supported official intergovernmental definition, adopted at the 1996 World Food Summit:

Food security exists when all people, at all times, have physical and economic access to sufficient, safe and nutritious food to meet their dietary needs and food preferences for an active and healthy life.

While this definition reasonably well describes the goal, it gives no indication about the way to get there. This reflects the fact that the realisation of food security, like that of ecological sustainability, involves highly sensitive political and economic choices.

Over the past decade, government policies on food security have developed, in parallel with their overall agricultural policies, in an increasingly trade-oriented direction. The declarations of the World Food Summit were highly influenced by this thinking, to the extent that increased trade was highlighted as the main requirement for achieving food security (World Food Summit 1996). The argument is, simply put, that because total global food production is already sufficient to satisfy global demand, what is needed to achieve food security is reduced barriers to trade, allowing existing surpluses to move freely to where there is a demand. The remaining problem, that hungry people do not constitute a demand until they have money to buy for, would have to be addressed by reducing poverty, which incidentally would also primarily involve trade liberalisation as the means to achieve more rapid overall economic growth.

As pointed out by many analysts, this argument is fundamentally flawed by its simplistic quantitative perspective on food availability, disregarding both sustainability and above all the distribution of wealth and power. (See for example South Centre 1997.) In particular for the rural poor in developing countries, who probably constitute the largest group of food insecure, it is stable access to fertile land, clean water and other productive resources for self-provisioning which is the key requirement. To achieve this, political rights and a degree of local self-determination is necessary.

The many NGOs present at the World Food Summit produced, in remarkable unanimity, a statement which challenged governments on precisely those points (World Food Summit NGO Forum 1996). It emphasises the need to

• strengthen the capacity of family farmers and local and regional food systems
• reverse the concentration of wealth and power
• change farming systems towards agro-ecological principles.

In order to achieve this, their common analysis, again in direct opposition to the trade liberalisation agenda of the governments, was that

• primary responsibility for ensuring food security lies with national and local governments, whose capacity must be strengthened and accountability enhanced.

This paper shares this latter understanding of food security, the essence of which is that basic political issues such as social justice cannot be avoided by reference to market mechanisms. Neither food security nor ecological sustainability are compatible with any form of social organisation. A serious consideration of either will have consequences for wider political and economic choices.

This section first gives some basic facts about the size and structure of world trade in agricultural products, then tries to briefly summarise the main effects of the WTO Agreement on Agriculture (AoA).

Global trade in agricultural products is relatively limited. Most people tend to conceive of 'the world market' as something much larger than national/regional markets, but for most agricultural products 'the world market' is not the largest market. The domestic markets of the USA, the EU, China, India and sometimes others are usually larger. Most agricultural products are consumed within the countries of production. Relatively little enters international trade.

The only agricultural products which are produced primarily for export are some of the typical tropical plantation crops, such as coffee (over 80 % of world production exported), cocoa, palm oil, and rubber. But even in this category, export dominance is by no means complete. Several of the major plantation crops are mainly produced for domestic markets, for example bananas, of which only some 20 % of world production are for export, tea (40 %), cotton (30 %), and jute (10 %).[2]

Among the cereals, wheat is the largest export crop with some 17 % of world production traded internationally. Nonetheless, the world market for wheat is just barely larger than the EU domestic market. In coarse grains (maize, barley and other feed grains) the export share is around 11 %. In rice, it is only 6 %, which means that China's domestic market is more than six times larger than the world market.

Not even in soy beans, the crop around which much of the global feed and food industry revolves, is the export share more than about 30 % (in some form: as whole beans, oil or feedstuffs). The figure is roughly the same for sugar.

Looking at animal products, it is only in beef that the world market reaches 10 % of total world production. In comparison, the USA domestic market alone consumes around 25 % of world production. Of pigmeat, only some 3 % is traded internationally, while over 20 % stays within the EU and almost 50 % in China, which is by far the largest producer. Of total milk production, some 6 % is sold on the world market, mainly in the form of powdered milk, butter and cheese.

Agricultural trade is completely dominated by developed countries. They have roughly a 70 % share of both exports and imports.[3] Exports from developing countries are mainly in the tropical plantation crops. Only a handful of developing countries have an export of grains or animal products of any importance.

For the main products it looks like this:

• Wheat: The USA, Australia and Canada between them have 2/3 of total exports. The EU has around 15 %, Argentina around 10 %. Almost 80 % is sold to developing countries, the remainder mostly to Japan, Korea and the former Soviet bloc.
• Coarse grains: The USA dominates completely with well over 50 % of total exports. Australia, Canada and Argentina together have almost 20 %, the EU a little over 10 %. By far the largest individual importers are Japan and Korea with together over 30 %, but almost all the rest goes to developing countries, mainly to the economically stronger among them.
• Rice: Thailand, Vietnam and China are the largest exporters with between them over 50 % of total exports. India, Pakistan and Burma also have considerable exports. The importers are to almost 90 % other developing countries with rice-based diets, in particular Indonesia, the Philippines, Bangladesh, Iran and Brazil, who between them buy 30-50 % (imports fluctuate greatly from year to year depending on how domestic harvests turn out). Developed countries, mainly Japan and the EU, buy about 10 %.
• Soybeans: The USA alone has well over half the exports. Brazil, Argentina and Paraguay have practically all the rest. The EU alone buys almost half the world trade. Japan, Korea and Taiwan between them buy about 15 %, and a whole range of developing countries, with China and Mexico in the front row, share the rest.
• Beef: The USA and Canada together have over 25 % of exports, and Australia, New Zealand, Argentina, Brazil and Uruguay between them over 50 %. The EU has almost 15 %, and India 4 %. However, the USA and Canada import just as much as they export, so they are not net exporters. Other large importing countries are Japan, Korea and Russia with between them almost 40 %. About 15 % goes to developing countries.
• Pigmeat: The EU is the single largest exporter with almost 40 % of the world market. The USA and Canada have a similar share between them. The buyers are Japan, Hongkong, Korea and Russia with 65 % between them, and the USA, which imports almost as much as it exports.
• Lamb: Australia and New Zealand dominate completely with 94 % of total exports. The EU, the USA, Japan and Russia between them buy 2/3 of this.
• Poultry: The USA dominates exports with 40 %, and the EU has some 13 %, but several developing countries also have a net export, in particular Brazil and Thailand with some 18 % between them. Large importers are Japan, Russia and China.
• Dairy products: Australia and New Zealand have roughly half the exports, and the EU almost all of the remainder. The only exceptions are in skimmed milk powder, where the USA has 10 %, and whole milk powder, where Argentina has a similar share. The buyers are completely different categories depending on the product. Of the powdered milk, 85 % goes to developing countries. Of the butter, 1/3 goes to developing countries, 1/3 to Russia, and the rest to the USA and the EU. Of the cheese, a high value product, less than 10 % goes to developing countries, and a similar proportion to Russia. The rest, over 80 %, goes to OECD countries, in particular to Japan and the USA, who between them have 50 % of the imports.

The share of agricultural products in the total trade in goods is in the magnitude of 10 % (in value terms). The share is slightly higher in the trade of developing countries than in that of developed countries (both in exports and imports), but it is not a very marked difference.

It should be clear from the above that agricultural exports are dominated by a small group of countries which can be described as 'natural exporters': the USA, Canada, Argentina, Uruguay, Brazil, Australia and New Zealand. What unites them is favourable climates and soils, sparse population and late colonisation. These three factors have created the preconditions for a large scale, relatively extensive agriculture, operating within physical, social, and economic structures established not much more than 100 years ago. The result is production costs which are very difficult to match for European or Asian farmers. All the 'natural exporters' have, despite extensive forms of production, a productive capacity far beyond their own needs. Historically, they have also built their economies to a large extent on agricultural exports.

In recent years, the EU has also become a large exporter, but of an entirely different character. Western Europe is among the most densely populated areas in the world, and it has limited natural capacity to feed more than its own population. Historically, the EU countries have repeatedly known food shortages and production crises, most recently during the period following the second world war. The EU was a net food importer until the 1970s. Its present role as an exporter is largely artificial, as the exports, mainly grains and animal products, are balanced by roughly comparable imports of feedstuffs (soybean products, maize gluten and other grain by-products, molasses, tapioca, etc.). In addition, present production levels are dependent on a very high use of chemical inputs, fossil fuels, etc.

Of the remaining developed countries several are net importers (Japan, Korea, Switzerland, Norway) but otherwise similar to the EU countries in natural conditions and productive capacity. In the former Soviet Union there are large areas which historically were exporters (in particular the Ukraine), but which cannot under present economic and political conditions use their capacity. This is also true for some of the Central and Eastern European countries in the applicant line for EU membership, while others are more similar to present EU members.

The developing countries can be divided into three rather distinct groups. There are a few net exporters. Brazil, Argentina and Uruguay are, as already mentioned, in a category by themselves, as the only ones who can compete on a large scale with developed countries on the global markets for wheat, feedstuffs and animal products. But a few other developing countries with higher population density and more traditional agricultural structure are also among the net exporters, notably Thailand and Vietnam.

A larger minority of the developing countries are net importers and formally listed by the FAO as Net Food Importing Developing Countries (NFIDC). In other words they are directly dependent on the world market for their basic food supply. Many of these are also among the world's 48 poorest countries, the Least Developed Countries (LDC).

But the large majority of developing countries belong to the middle group which is more or less self-sufficient in food (+/- 10 %). Although they may more often buy than sell food on the world market, they are not dependent on imports on a regular basis. Many of those can also balance a certain import of basic foodstuffs against an export of other agricultural products, typically tropical plantation crops.

To a very large extent, agricultural trade patterns have been shaped by the fact that practically all countries favour their own agriculture in some degree. There are three main methods to do this:

• Border protection against imported products
• Internal support measures for domestic producers
• Export subsidies.

Border protection is the cheapest of the three methods, and consequently the most widespread. Internal support measures normally cost money in the government budget. This means that they are mainly used by developed countries, although most developing countries do have some support programmes, most often in education and technical advice. Export subsidies are exclusively used by developed countries.

The aim of the WTO Agreement on Agriculture (AoA) was to reduce the use of all three methods for favouring domestic production.

• Regarding border protection the AoA introduced a prohibition of all protection measures except fixed tariffs. This means that other measures (for example variable import levies and import quotas) must be converted into fixed tariffs. The process is called tariffication. These tariffs must then be reduced by a certain percentage during the treaty period.
• In addition, the AoA requires all countries to allow a certain minimum market access for each individual agricultural product. What this means in practice is that a country which is self-sufficient in a certain product, and protects its domestic production with a tariff, must offer a reduced tariff for the quantity specified in the AoA, so that import can enter its market. This mechanism is called tariff rate quotas (TRQ).
• The AoA also contains detailed rules for how internal support may be designed. The main principle, often referred to as 'de-coupling', is that support measures cannot be directly related to production volumes.
• For export subsidies there are similar reduction requirements as for tariffs.

Developing countries have been allowed a lower level of reduction requirements (2/3 of those applicable to developed countries) and a longer period for implementation (10 years instead of 6). The percentage for minimum market access is 5 % for developed countries and 4 % for developing countries.

The differently coloured boxes which constantly recur in AoA discussions refer to a popularised categorisation of the various internal support measures, which was introduced during the negotiations. It is based on the traffic light principle.

• The red box thus contained those forms of support which were to be prohibited at the entry into force of the AoA, for example variable import levies. The content of the red box, in other words, is already eliminated.
• The green box contains those support measures which for the time being are accepted without any reduction requirements, for example support to agricultural research, to rural development or production extensification measures, to food security related public stockholding schemes, or to food aid.
• The amber box contains forms of support which violate the basic rules of the AoA and will in principle be prohibited, but are accepted on an interim basis provided they are gradually reduced. The bulk of those support forms are systems for regulation of market prices through guaranteed prices, government intervention etc., but some forms of direct payments are also in this box, as are input or investment subsidies.

In the final stage of the negotiation however, this neat three-way division broke down with the introduction of the blue box, which contains much of what should logically have been in the amber box. Blue box measures thus, like amber box measures, violate AoA principles. But unlike the amber box measures, they are allowed to continue without reduction requirements, provided they have some connection to a production-limiting scheme. The EU direct payments are the largest part of the content in the blue box.

The blue box measures are covered by the so-called peace clause, which means that during a transition period until the end of 2003 they cannot be challenged in the WTO dispute settlement body, even though they are in violation of the basic principles of the agreement.

Overall, the AoA has been considerably less effective than expected. But effectiveness has varied greatly both between its different components and between countries.

The reductions of border protection have come out very unevenly. Most developed countries have minimised the effect of the reduction requirements through a combination of clever manoeuvring during the negotiations and creative use of the many loopholes in the system. Many developing countries lack the expertise needed for such manipulations, and thus have often been harder hit. Those developing countries whose tariffs were low or non-existent have sometimes bound themselves to keep that level, even though the rules would have allowed them to bind the maximum tariff on a higher level to preserve more choice for the future.[4]

The minimum market access requirements have been quite effective, even though the Tariff Rate Quotas have turned out to be a complicated mechanism, prone to manipulation. The percentages are modest, but as they are calculated on the bulk of global consumption, the total trade volumes involved are considerable. Compared to the total size of the world market for most food products (as indicated above), 4-5 % is not a low figure.

The rules governing the design of internal support measures have (together with tariffication) fundamentally changed agricultural policy both in the EU and the USA, and in most other developed countries. Before the AoA, the cornerstone of EU policy was variable import levies, now prohibited. These levies were constantly adjusted to regulate the volumes of import allowed to enter the EU, and thereby guarantee that domestic products were always sold first. It was in preparation for the AoA requirements that the 'McSharry reform' of EU agricultural policy was carried out in 1992, establishing a lower internal price level compensated by direct monetary support to producers. This reform shifted more than enough of the EU agricultural support from the amber to the blue box. The USA, with their FAIR Act in 1996, went one step further and redesigned their support system entirely to the specifications of the green box.

On export subsidies however, the AoA has had very limited effect. One important reason is that the EU, as the largest user by far of direct export subsidies, already during the negotiation managed to ensure that the reduction requirements would have their reference point in 1986-88, the period when EU export subsidies reached their historical maximum. A second reason is that the AoA only regulates direct export subsidies, not other forms of support enabling sales to the world market at prices under actual cost of production.

In sum, it is clear that the AoA has primarily favoured agricultural exporters. The agreement in effect codifies a 'right to export' through the rules about minimum market access, which mean that a country no longer has the right to opt for full self-sufficiency as a strategy in any category of agricultural products. No matter what the reasons might be, as long as there is an exporter anywhere willing to sell at a lower price, the AoA is on the side of that exporter. It should be noted that this goes for developed and developing countries alike, and -- most remarkably -- regardless of whether the lower price is made possible through export subsidies.

Conversely, the AoA in many ways has limited the possibilities to support the development of domestic production. Remaining WTO-legal support options almost invariably require direct payments through the government budget. Obviously, this strikes hardest against developing countries, which have very limited means to offer such support. Developed countries in addition have continued access to many forms of support which are now illegal unless they were already in use at the entry into force of the AoA (amber box measures).

At the same time, the AoA has not led to the expected stabilisation and increase in world market prices. On the contrary, price levels have dropped to historically low levels during the last few years, and fluctuations have increased.

The positions taken by WTO members in preparation for a new round of agricultural trade negotiations flow very logically from their respective roles in present trading patterns and from the size and design of their own agricultural support systems.

Unambiguous enthusiasts for a continued rapid march in the direction established by the present AoA are only those countries which both are 'natural exporters' and have minimal agricultural support systems of their own: the hard core of the Cairns group (Australia, New Zealand, Canada, Argentina, Brazil, Paraguay and Uruguay). The remaining Cairns group members are developing countries,[5] in many cases motivated more by the hope of expanding their exports than by direct defence of existing interests. Some of them are in fact large net importers of food (Indonesia, the Philippines) and partly have objective interests diametrically opposed to the deregulation principles of the Cairns group. This has recently led to an internal discussion about food security among group members.

In its capacity as the leading agricultural export nation, the USA shares many of the basic values of the Cairns group. The important difference is that US exports to a very large extent depend on indirect forms of export support. Under the FAIR Act agricultural policy reform, implemented from 1997, the intention was to eliminate most of those support systems during a 7 year period, and for a short while the USA appeared to be rapidly approaching the Cairns group's positions. But already in 1998 reduced support levels in combination with abruptly falling prices led to an acute economic crisis for US agriculture. The crisis was met by new, 'temporary' support measures. But those measures were followed by additional ones in 1999 and again in 2000, and in practice the USA is now well on its way to abandoning the reform plan altogether. Consequently, its deregulatory profile in the WTO has been considerably softened.

The main opponents to a continued liberalisation are the small group of developed countries which have difficult natural conditions for agriculture, yet strive to maintain as much production as possible at the price of high levels of border protection and internal support: Japan, Korea, Norway and Switzerland.

The EU is fundamentally ambivalent. The odd combination of large exports, high export subsidies, high border protection and high internal support constantly forces the EU to defend incompatible positions. But the emphasis has shifted noticeably since the years immediately following the policy reform in 1992, when EU political rhetoric only revolved around competitiveness and asserting the European role in the world market. Now there is renewed emphasis on the traditional defence of domestic production, and it comes with a new set of arguments centred on the 'multifunctionality' of agriculture. At the Seattle WTO summit the EU proved remarkably serious about this concept, opting to co-ordinate its whole negotiating strategy with Japan, Korea and Switzerland in a coalition informally known as the 'Friends of Multifunctionality'.

Although the largest group of WTO members, the developing countries have so far not played a decisive role in AoA negotiations. Most have tended to lend general support to the positions of the Cairns group, such as the elimination of export subsidies and increased market access, but there has only been joint action of a few major common concerns. One of those, which enjoys virtually unanimous support from developing countries, is the demand to seriously evaluate the effects of the present AoA before entering into new negotiations, with particular emphasis on the effects for food security in LDC/NFIDC.

Post-Seattle, there are however signs that developing countries as a group may become more proactive. Proposals have for example been tabled recently suggesting major changes to the 'box' schemes, taking developing country interests better into account.

This section is an attempt to systematise the key issues raised by the AoA from the perspective of food security and ecological sustainability. There is no ambition to give a complete or objective overview of the agricultural trade negotiations more generally. The discussion is based in large part on work done by NGOs active on agricultural trade issues during the past few years, as should be evident from the list of references. For those unfamiliar with this NGO community, there is a brief guide toward the end of this report.

The most clear-cut issue in the AoA is without doubt export dumping. It is very difficult to find any merit in a systematic underselling of agricultural products. It causes huge costs on the exporting side, and major distortions of agricultural economies everywhere. Export dumping is consequently denounced by the broadest possible range of voices, spanning from the agriculture minister of Australia to Via Campesina. Export dumping is only defended by those who use it. The problem is that the main users are the USA and the EU, the two largest actors on the food world market.

Export dumping of agricultural products occurs in several forms, of which the AoA only deals with one, and rather ineffectively at that.

• Direct export subsidies are now used almost only by the EU. Because internal EU prices are practically always higher than prices on the world market (due to border protection), export can normally take place only at a price below that paid to the primary producer. The EU budget therefore intervenes and compensates the exporter for the difference. This is the mechanism covered by the reduction requirements of the AoA.
• Export credits are widely used by the USA and sometimes by other countries. By offering favourable terms of payment the cost to the customer is reduced (at least in the short term). This is not regulated at all by the AoA. In addition, it is difficult to monitor, because export contracts are usually confidential. The EU refuses to discuss further reductions of their export subsidies unless export credits are also brought into the discussion.
• All forms of direct payments function as a dumping mechanism to the extent that the production supported results in products for export. When border protection is reduced and replaced with direct payments (as required by the AoA), the result is lower prices in protected markets. The gap between the protected internal price level and world market prices is reduced, and the need for export subsidies thus reduced correspondingly (again in conformity with the AoA). But for the importing country, there is no difference. Whether the export price is artificially reduced by export subsidies or by direct payments, the dumping effect is the same.
• Non-monetary forms of dumping, i.e. environmental dumping, social dumping etc., are an even more sensitive subject which practically none of the WTO parties are prepared to discuss. Especially in the 'natural exporter' countries there is very often a correlation between low costs of production and extractive production methods, externalisation of environmental costs, etc. But no countries have a clean conscience in this respect.

Both in the USA and the EU, virtually all agricultural production now profits from some measure of direct payments. This means that practically everything exported from those countries involves some level of dumping. In many of the major commodities, US plus EU exports account for half the world market or more. Thus, a prohibition of dumping which included dumping by direct payments would have enormous repercussions. Both in the EU and the USA it would aggravate present surplus production problems, although in partly different ways.

EU exports would likely shrink to a fraction of their present size, because actual costs of production are so far above world market prices that mainly high value products such as cheese and olive oil could continue to be exported. The EU would have to find some way to avoid producing the surpluses now exported. Supply management measures such as expanded set-aside and production quota programs would be the most obvious solution. But the EU could also relatively easy rebalance its production by reducing intensity, especially in animal production. This would greatly reduce imports of feedstuffs, and more of domestic grain production, now exported, could be used at home.

Increased feed self-sufficiency in Europe would however further complicate the situation for the USA, as it is the main source of EU feedstuff imports. Compared to the EU, the gap between US domestic and world market prices is considerably smaller, so US exports would far from cease without the dumping mechanisms. But they would likely shrink considerably, at least initially. Just like the EU, the USA would be forced to re-evaluate its production strategy and reinstate supply management measures in order to adapt output to actual demand.

In a longer perspective, it is however very plausible that an elimination of export dumping would lead to the permanent increase in world market prices that the present AoA has failed to effect. This would strengthen agriculture everywhere, including in the EU and the USA, although both would have to settle for a less dominant position in agricultural trade.

Considering that systematic underselling is one of the most obvious violations of free trade principles, it is puzzling that the GATT/WTO has never seriously tried to deal with dumping in agricultural trade. There has been a general anti-dumping clause in the GATT text since the very beginning. It might also be expected that the USA, which has the world's strictest national anti-dumping legislation and uses it very actively against any suspicious import, would be interested in establishing similar rules on a global basis.

The general anti-dumping clause of the GATT is however a weak one, primarily because the operative definition of dumping is export sales at a price below the 'normal price' on the domestic market. Whether or not this domestic price correctly reflects actual cost of production is not considered. Only if for some reason a 'normal price' is impossible to establish, does the GATT allow reference to cost of production. In comparison, the US domestic legislation is entirely based on calculations of actual cost of production.

An additional weakness in the GATT provisions is that even when dumping can be shown to have occurred, the country targeted is required to prove actual damage to domestic production before remedial action may be taken (protective tariffs). This is often difficult and in practice severely limits the usefulness of the clause.

Especially after the introduction of compensatory direct payments on a massive scale, this definition of dumping has become very nearly meaningless for agricultural markets. For example, if the EU exports wheat at the internal market price level, this is from a GATT perspective the 'normal price' and not dumping. No direct export subsidies are involved. Yet, this price is well below actual cost of production and would not be possible for the producer to survive on without the additional 15-20 % compensation contributed through direct payments.

Against this background, targeting only the direct export subsidies is pointless, and many NGOs have tried to expand the discussion to export dumping in the broader sense.

A problem with this approach has been that most developing countries react instinctively against any suggestion to strengthen the WTO anti-dumping provisions. The reason is that many of them have bitter experiences with US anti-dumping legislation, which has repeatedly been used to block developing country exports into the USA. In this particular case however, it is difficult to see how developing countries could be hurt by stricter measures against dumping in agricultural trade, because it is exclusively carried out by developed countries.

Mark Ritchie (IATP) has suggested a simple model based on existing PSE calculations, which are already used in the present AoA as basis for the reduction requirements in the amber box (Ritchie 1999). These calculations were developed by the OECD to allow comparison between countries of total compensation to agricultural producers. The PSE value is an estimate of the monetary value of all public support systems to agriculture, and can be expressed in terms of value per unit of each agricultural product.[6] Dumping could, according to Ritchie, be defined as the difference between the export price and a 'normal price' defined as the domestic price plus the PSE value of internal support schemes. In the case of EU wheat exports, for example, the value of direct payments would be covered by the PSE calculation and included in the 'normal price'. This would give a reference point from which the WTO could negotiate a reduction program along the same lines as for tariffs and amber box support.

Daniel Van Der Steen and colleagues (CSA) follows a similar line of thought, but also propose to facilitate the introduction of protective tariffs by partly putting the burden of proof on the dumping country (Van Der Steen et al 1999).

The opening up of markets, primarily by reducing tariffs for goods, is strongly supported by almost all developing countries in trade negotiations, with solid backing from development NGOs (see for example CLONG 1999, ActionAid 1999). The rationale is clear and simple. With lower tariffs, developing countries improve their chances to sell on developed country markets and thereby strengthen general economic development. Together with textiles, agriculture is often singled out as the primary field of export opportunity.

However, there is a number of factors which complicate the matter in agricultural trade, as compared with trade in industrial goods.

• Developing countries most often buyers. Looking at the structure of agricultural trade, it is clear that developing countries are more often importers than exporters of agricultural goods. The overall net effect of increasing minimum market access requirements and continuing general tariff reductions under a new WTO agreement would therefore be to further facilitate developed country exports to developing countries, more than the opposite.
• Developed country products often cheaper. Generally speaking, in trade with goods, the competitive advantage of developing countries is in price (most often because of low labour cost). But in agricultural trade developed country products are often cheaper. Sometimes because of export support of various kinds, but in other cases simply because of large scale, mechanised production. Again, the effect of improved market access would be primarily increased imports into developing countries, not exports from them.
• Developed countries sell food. Developed countries not only dominate agricultural exports in purely quantitative terms (70 %), but their export is overwhelmingly in basic foods and in feedstuffs. This means that increased imports into a developing country directly impacts on its domestic food market and agricultural production.
• Tropical products already relatively favoured. Developing country agricultural exports, in contrast, are to a large extent in tropical products, a category where developed country tariffs are already relatively low. Developed countries protect mainly the sectors where they have production of their own, in other words mostly temperate products. There are some well-known exceptions, such as bananas, sugar (cane sugar competes both with beet sugar and corn starch syrup), and tropical oils (palm and coconut oils compete with temperate oils). But on the whole, the effect of a general tariff reduction would be considerably smaller for developing than for developed country agricultural exports.
• Export sector separate from food production. In developing countries, it is common that the exporting sector is disconnected from food production (other crops, plantation structure, foreign ownership etc.). Even though an increase in exports would mean increased income for the country, it is far from certain that this will benefit agriculture as a whole, in the way it would if export crop production were an integrated part of food producing agriculture. In addition, direct competition between the two sectors is not uncommon (land, water, labour).

Against this background, it is clear that the effects of a further reduction of tariffs on agricultural products would differ widely between individual developing countries. It is certainly not possible to say that it would benefit developing countries as a group. It would benefit countries, or sectors within countries, which have a potential for export of the products impacted (in particular those who compete with traditional developed country exports). On the other hand, it would create increased problems for those dependent on protecting their domestic production.

From a food security and ecological sustainability perspective it is in particular the effects of liberalised imports on domestic production that should be closely watched. Bello (1999) dryly observes that the enthusiasm for tariff reductions of the South East Asian governments serves the interests of plantation crop exporters, while the price is paid by small rice and maize farmers. Madeley (2000) reviews a number of studies covering agricultural trade liberalisation both under the AoA and World Bank/IMF structural adjustment plans in 39 developing countries. His stark conclusion is that liberalisation "has worsened the plight of the poor by sacrificing them to the free play of international market forces", and the influx of cheap imports tops his list of 'chief effects'.

Van Der Steen et al (1999) cite three instructive examples. The first one concerns Mexico, where NAFTA accession in 1994 led to a rapid elimination of border protection. Starting from full self-sufficiency with maize, Mexico's major staple, imports reached 40 % in three years, causing massive displacement of small farmers and radically deteriorating nutrition status in the poorest parts of the population. Maize cultivation has been replaced to a large extent by feed crops and cattle raising, both for export.

The second example is from the Philippines, where the shift from import quotas to fixed tariffs (to meet AoA requirements) have resulted in greatly increased maize imports from the USA, seriously damaging the economy of the maize growing districts in central Philippines. Imported US maize is now about 20 % cheaper than the domestic product, but the gap will increase as tariffs are further reduced under AoA commitments until 2004.

The third example, from the EU, goes several decades back. When the EU Common Agricultural Policy (CAP) was introduced in the early 1960s, an agreement was concluded with the USA in the framework of the GATT negotiations (the Dillon Round). The USA agreed to accept the new border protection mechanisms put in place by the EU for food, in return for a commitment by the EU to allow unlimited import of feedstuffs from the USA (so called Cereal Substitution Products, CSP) at zero tariff. The EU at this point was still a net importer both of food and feedstuffs, so the agreement appeared risk-free from the EU point of view.

No more than 15 years later, however, the EU as we know produced large surpluses of both grains and animal products. This surplus was based to a large extent on the greatly increased imports of feedstuffs, mainly soybeans but also large volumes of maize gluten and other grain derivatives. At first imports originated only from the USA, but over time also from Brazil, Argentina, Thailand (tapioca) and other countries. Without the zero tariff for feedstuffs, the huge surpluses of the 1970s would never have been possible, and export dumping from the EU would have been if not insignificant so at least not a major global problem. In addition, cheap feedstuff imports were without doubt the decisive factor behind the industrialisation of animal production in the EU and its concentration to the vicinity of major ports (Jutland, Holland, Bretagne). By extension, feed imports are the root cause behind a number of serious environmental problems in European agriculture, most of which are related to intensive animal production, and behind the elimination of most of the traditional, integrated animal production in mountain zones and other marginal regions.

Apart from the obvious, these examples also clearly show that the categorisation into developed and developing countries is not always the most relevant one. In the European example, export from several developing countries contributed (and contributes) to the distortions of European production systems, including its ecological unsustainability and the decline of small scale production for the domestic market. Indirectly it also contributes to EU export dumping, which in turn harms food production in other regions of the South. To complicate even more, as some of the soybean production in Brazil (particularly in Rio Grande do Sul) is well integrated into food producing family farms, these in fact have also contributed to destroying the livelihoods of their direct counterparts both in Europe and in all countries affected by EU animal product dumping. Brazil itself in fact is one of those affected countries, so that the exports of Brazilian soybean farmers actually contribute to the dumping pressure on Brazilian dairy farmers.

There are several technical issues concerning market access which can be equally or more important for developing country export opportunities as overall tariff levels. One of these is tariff escalation, the increase of tariff rates with increasing level of processing. Its effect is that the unprocessed raw material may be possible to export, but not the finished product. For example cocoa, but not chocolate bars.

Another is SSG, the Special Safeguard clause, a mechanism allowing a country to rapidly increase tariff levels in response to a sudden increase in imports. This is one of many examples of technical details in the AoA which developing countries did not discover in time. To use the SSG, members had to deliver a notification in connection with their signature. No developing countries did, but almost all developed countries.

Both issues should be possible to solve within the framework of the Special and Differential Treatment (SDT) for developing countries, which in principle should exist throughout the WTO system, but has in practice been very grudgingly implemented. A number of concrete proposals are listed below in the section on SDT.

No part of agricultural policy suffers from so many misunderstandings as the internal support systems. This should come as no surprise given their level of complication, but a few basic facts may be in order before entering into the discussion.

The new EU support system introduced by the 1992 CAP reform offers a good example of the types of changes introduced in response to the AoA. What did the system look like to begin with? A persistent myth is that the EU agricultural budget was enormous and that huge sums were paid to farmers. The opposite was the case.

The original CAP was based on a system of border protection with variable import levies, which were continually (usually weekly) adjusted so that they were always exactly high enough to stop imports from entering at a price below the administratively fixed price on the EU internal market. This was an extremely bureaucratic system, but it was inexpensive to the EU budget and effective in matching imports to demand. The full cost of food was paid directly by the consumer. Both domestic and imported products were sold at the higher, protected price. The EU budget in fact collected considerable income from the import levies, which were paid by importers.

Budget costs occurred only when production increased above the level of domestic consumption. When this happened, the EU first resorted to building up buffer stocks (in the hope that there would be a shortage at a later point). If this was not enough, export subsidies were used to get rid of the product at a reduced price outside the EU. As long as the market was reasonably well in balance, however, the cost of export subsidies was covered by the income from import levies.

It was not until the permanent, structural surpluses started appearing in the mid-1970s (when animal production based on imported feedstuffs had become well established) that costs started to get out of hand, as export subsidies expanded. Toward the mid-1980s EU policy makers were becoming desperate. Some kind of reform was necessary, above all to bring the budget back under control. Export subsidies to agriculture had become the largest single item in the EU budget, and even worse, they fluctuated completely unpredictably as harvests came out larger or smaller and world market prices higher or lower.

At the same time, the USA started to seriously push for having agriculture included in the upcoming Uruguay Round. The model proposed was reduction of both border protection and export subsidies, but with the possibility to compensate agriculture for the lower prices which would result by direct payments straight from agricultural budgets to farmers. That was also the final outcome, first in the EU and then in the AoA.

This solution was considerably more expensive for the EU budget. Some of the money previously paid directly by the consumer to the farmer via prices now had to be covered by the budget. But the advantage was predictability. Now, expenses could be foreseen in detail for several years in advance, as soon as the level of direct payments had been agreed for that period. The new phase of the reform implemented from this year (Agenda 2000) involves a further increase in budget costs after the same pattern (this time it was designed to match the expected demands of the WTO Millennium Round).

In other words, the effect of the AoA for the EU was a considerable increase in internal support and total budget, not the opposite, as often claimed. The increase in support was the cost of decreasing border protection (increasing market access) without drastically reducing the economic viability of EU agriculture.

Most developing countries have the same position regarding internal support as regarding border protection (tariffs). Typically, the stated preferences are total elimination of blue box support, critical evaluation and limitation of green box support, and no prolongation of the peace clause after 2003. Again, most development NGOs agree.

From a purely developing country perspective this position appears almost completely unproblematic. No developing countries use blue box support, and to the extent that they use green box support, their programs will most likely survive an evaluation. A reduction/elimination thus can create no negative effects for domestic production in developing countries in the way that a reduction of border protection can. A further reason for developing countries to oppose continued direct payments is the fact that the whole model in practice is useless for them, because they lack the financial resources to use it.

It is also clear that the internal support payments of developed countries (in contrast to their border protection) directly harms domestic production in many developing countries, because they now are, as we have seen, the main mechanism for export dumping.

The problem with reducing internal support is instead the effects within developed countries, which would in turn have consequences for the global food production system as a whole.

The economic reality of developed country farmers is that despite relatively high compensation levels (prices plus payments) they only barely survive on their farming. Just like in developing countries, farming pays less than practically any other job. The image often conveyed by media of huge 'subsidies' raining over farmers, which could be eliminated with no other consequence than a reduction of excessive profits, is simply false.

Within the EU, the price level for virtually all agricultural products is now considerably below actual cost of production. This is not accidental, but a deliberate consequence of the AoA requirements (reduced border protection). Direct payments are a necessary complement to fill the gap between the price level allowed by the AoA and the real cost of food production. The situation in the USA is very similar, although production costs are lower and the gap to prices therefore smaller. (Besides, support systems are technically quite different.) In sum, an elimination of internal support would mean red figures for most farmers in both the EU and the USA.

But is it not the high level of support which in itself drives cost of production? Would not the whole cost level be adjusted downward if supports were eliminated?

No, probably not significantly. What would go down over time is capital cost, because such a radical drop in compensation would drastically reduce the value of the farms. Present owners would not be able to pay their mortgages and be forced to sell at a loss. The buyer could start over with a much lower debt level. (Alternatively the present owner could go bankrupt and start over himself. Then his creditors would take the loss.)

But mainly the cost level is simply a consequence of running a business in a developed country. Compared to most other branches of business, the difference in cost of production is in fact unusually small between agriculture in developed and developing countries. The difference between the EU and USA/Canada or Australia/New Zealand is in turn easily explained by differences in natural conditions.

A strongly contributing factor to the confusion in the discussion about price levels and support is the established habit of relating all other prices to world market prices.

Nobody will deny that world market prices in general are on a level far below the production cost of the majority of producers. This is because the world market, as we have seen, is relatively small and heavily influenced by various kinds of dumping. Yet all of the AoA requirements take world market prices as their point of departure. There may be legitimate technical reasons for this (some constant must be used as reference), but the unfortunate effect is that world market prices achieve a status of 'normal price'.

The AoA reduction requirements for internal support (amber box) are based on a calculation called Aggregate Measure of Support (AMS). It is based, in turn, on the so-called PSE calculations (Producer Subsidy Equivalents) developed by the OECD during the 1980s. PSE is a method for translating all forms of agricultural support to a monetary equivalent. PSE gives a measure of how much the sum of all public intervention (border protection and other price regulation, direct payments, etc.) are worth to the farmer in terms of dollars/ton of product (or dollars/farm, or dollars/hectare).

The extreme 'subsidy levels' regularly reported in media usually are PSE figures. The reference point is the world market price. If domestic prices are higher, as they are in most countries, this counts as subsidy. If then there is actual public expenditure which benefits farmers (whether it reaches them in cash or in kind) this is added to the sum. The PSE is the total of this calculation, in other words the difference between the world market price and total compensation to farmers. For example, if a 50 % subsidy level is reported, this means that total compensation to the farmer is 50 % above the current world market price.

Because world market prices are very unstable, this 'subsidy level' fluctuates greatly, independently of whether compensation to farmers actually changes. This complicates in particular the discussion of long term trends. An example: The price of a given product in a given country is 125 dollars/ton. In addition, the farmer receives the equivalent of 25 dollars/ton in public support of various kinds, so total compensation is 150 dollars/ton. On the world market, this same product sells at 100 dollars/ton. Total compensation thus is 50 dollars over world market level, and the PSE percentage calculates to 50/100 = 50 %.

Now, if the price level for this product drops 25 dollars both on the world market and on the domestic market of this given country, this means that total compensation to the farmer is now 125 dollar/ton, down 17 %. In relation to the new world market price, 75 dollars/ton, it is still 50 dollars higher. But in terms of PSE, the 'subsidy level' actually increases, as the new, lower total compensation level is a higher percentage of the new, lower world market price. More exactly, the PSE is now 125/75 = 67 %.

Needless to say, the PSE value gives no indication of the relation between prices and actual cost of production. It is quite possible to have a relatively high PSE 'subsidy level' even when total compensation to the farmer is well below actual cost of production.

Another consequence of the model of calculation is that a country which has no internal support payments at all, only tariffs for border protection, still has a 'subsidy level' in PSE terms, simply because the price inside the protected borders is higher than the world market price.

A complete elimination of present internal support systems in the EU and the USA would cause major structural changes in agricultural production. The immediate effect would be that a large proportion of the smaller and less intensive family farms would be forced out of business, in particular in the EU where probably whole farming regions would be closed down. The only adaptation strategy available would be the industrial model: increased specialisation and scale.

The long term effects would likely be a further strengthening of the global agricultural structure we are already moving toward, with the EU countries as increasingly specialised animal producers, their competitive advantages being their location inside the world's second largest market and their skills in utilising feedstuffs from all over the world in a highly intensive production. A few regions in the EU might be able to continue competing in grain production, but the bulk of it would move to the plains of the Americas and Australia. In addition, the EU would certainly continue to defend its turf within some limited niche markets (wine, fruit, olives, cheese, various regional specialities).

A development along these lines would greatly strengthen the grip of transnational corporations over the food sector. They would obviously be in control of an expanding global trade, but they would likely also move further into primary production, following the pattern already established in the USA, where corporations more and more replace traditional family farming even in agriculture proper.

Would this development benefit developing countries? Again, there is not a universally valid answer. It is a reasonable assumption that more agricultural production would be relocalised to developing countries if transnational corporations had freer hands to organise it unencumbered by national borders or legislations. This has happened before in many other industries. There are already tendencies in this direction in animal production (beef, chicken, feedstuffs). From a strictly economic point of view, it is clear that there would be increased opportunities for foreign direct investment, and consequently for employment and general economic development. On the other hand, there is reason to suspect that effects on domestic food production would be negative, as more resources were diverted to the export sector, and as the influence of transnational corporations in national affairs would increase. In sum, whether it would be a positive development to move larger parts of the food production of developed countries to the developing world depends on how one values export income relative to food security.

While food security concerns have played a very marginal role in the negotiation and implementation of the AoA, they have been the main focus for most of the NGOs monitoring the agreement. There are a number of NGO analyses and proposals specifically related to the (negative) consequences of the AoA on food security and possibilities to counteract those. Three levels of discussion can be discerned.

The first level relates to strengthening and expansion of the specific concessions for developing countries already in the agreement. A number of proposals in this line are listed below in the section about Special and differential treatment.

On the second level there are proposals for more substantial changes and additions to the AoA to promote developing country interests. These include both general demands for reduction in border protection etc., as already discussed, and more specific proposals. One recurring idea is to create a set of targeted derogations tailored to developing country needs, a 'food security box' or 'development box', in the same way as developed countries were already granted derogations by the blue and green boxes.

On the third level there are more far-reaching proposals about fundamental changes to AoA basic principles, most of them under the heading of 'food sovereignty'.

Although the idea of a separate 'box' designed to the specifications of developing country agriculture has gained broad support among development NGOs, there exist no detailed proposals for what it would contain or how it would technically be constructed, only brief texts in position papers.

Those who speak about a 'food security box' (see for example UK Food Group 1999, CLONG 1999, ActionAid 1999) are exclusively concerned with derogations that would improve developing country possibilities to protect their domestic food production and markets: exceptions from the tariff reduction and minimum market access requirements, and the right to increase internal support until a higher level of self-sufficiency is achieved. South Centre (1999), which speaks about a 'development box', additionally would give developing countries specific exceptions concerning export development support, and more favourable conditions of access to developed country markets.

A common feature of the 'box' proposals are that they clearly place these measures in the same category as the strongly questioned (not least by developing countries) blue and green box measures. By defining the proposals so clearly as derogations only, they implicitly lend support to the fundamental principle of the agreement, totally deregulated agricultural trade in the long term.

It must also be noted that interest from developing country governments in making food security a major issue in the new negotiation has been limited. A change may however be underway. In June 2000 the 'development box' idea was formally tabled in the WTO Committee on Agriculture by a group of developing countries (Cuba et al 2000). The proposal closely follows the South Centre in combining food security and market access related concerns in one 'box', although the emphasis is on measures to protect and develop domestic food production. Notably, the proposal calls for a very broad derogation allowing all developing countries to exempt whatever part of their agriculture they choose from AoA commitments on a product by product basis. The proposed mechanism is a positive list. Only products declared by a country on its positive list would be subject to AoA disciplines.

Those who speak about 'food sovereignty' clearly state that it is fundamental principles of the AoA which need to be changed, alternatively that the whole agreement should be repealed. On the policy level, the core of the 'food sovereignty' concept is simply "the right of each country to define its own agricultural policy to meet its internal needs" (Via Campesina 1999; for alternative formulations see Food Sovereignty Platform 1999, ERA Consumer 1999, Confédération Paysanne 1999, Van Der Steen et al 1999). But the concept is usually defined to also include the right of food self-determination down to local levels, and closely linked to the defence of rights to land, water and other productive resources.

'Food sovereignty' thus in essence turns the fundamental principle of the AoA upside down. Instead of each nation's equal right to export to others, the starting point is each nation's equal right to choose how to secure its livelihood. It is, as it were, the 'food security box' transformed from exception to general principle, and valid for all countries, not only for the developing world. But 'food sovereignty' also has a wider scope, effectively merging the food security concerns of developing countries with ecological sustainability and other concerns equally valid for the developed world. For example, it would encompass the right to independently define a national level of quality, health or environmental standards for domestic production as well as for imports.

It must be noted, however, that 'food sovereignty' is very far from the current political mainstream. Very few politicians dare face the wind and openly question the primacy of free trade over food security, and even many development NGOs seem to fear being categorised as 'protectionists'. But it is also evident, especially after Seattle and the 'Friends of Multifunctionality' that the strong momentum enjoyed by agricultural free trade proponents is beginning to wane, even though it is not yet comme il faut to say so.

Perhaps the clearest signal of a fundamental change in the nature of the game is the very firm stance of the EU regarding 'multifunctionality', despite extremely fierce reactions from both the USA and the Cairns group.

The argument about the multiple roles played by agriculture in addition to its production of commodities -- in environmental protection, rural development, and cultural, biological and social diversity -- is very familiar for us on the NGO side. It is what we have been preaching for decades, whether in criticising agriculture in the North or discussing development paradigms for the South.

But neither the Agriculture Directorate of the EU Commission or COPA-COGECA, the central organisation of the European agricultural lobby, have usually been allies in those discussions. On the contrary, they have in close symbiosis designed an agricultural policy increasingly geared toward 'unifunctionality', production only. So when those two institutions jointly launched the multi-functionality rhetoric in 1997, reactions were very sceptical.

That this move was a part of the positioning for a new WTO round was obvious. The level of support to EU agriculture was a prime issue and the EU Commission needed a line of defence. Many, including all other major groupings in the WTO, have dismissed the whole concept as nothing but new clothes for traditional protectionism.

During the three years since then, however, the EU has shown a willingness to take a fight over agriculture in partly unexpected contexts, particularly in issues related to gene technology. Substantial examples include the refusal to accept imports of hormone-treated beef despite a negative outcome of the WTO dispute settlement procedure, the de facto moratorium for GMO crops, the total ban against BST hormones for dairy cows, and most lately, the alliance with a majority of developing countries in the final negotiation of the biosafety protocol.

No doubt the outcome in several of these issues has been decided by a strong and active opinion in many member countries, but the fact remains that this opinion would not have had the same leverage had the EU leadership not opened a door. A new dynamic has been created, the strength of which was most likely not foreseen by the EU Commission and COPA/COGECA. European environmental organisations initially were among the sceptics, but have gradually found the multifunctionality argument useful in putting pressure on the EU institutions.

Among developing countries however, the attitude to 'multifunctionality' is still, with very limited exceptions, suspicion. Likewise, more surprising, among many South-based NGOs.[7] In terms of objective interests, there should be a considerable potential for alliances between the 'Friends of Multifunctionality' and the non-exporting group of developing countries, whose 'non-trade concerns' are in fact very similar. Both groups would benefit from a 'food sovereignty' type solution, allowing a higher degree of protection for domestic markets against penetration by 'natural exporters'.

The main, and well founded, reason for developing countries' suspiciousness is the EU's use of the multifunctionality argument to defend continued dumping with the help of 'blue box' payments. Unless this is addressed, there is little reason to expect the potential for alliances to be realised.

An issue which, like export dumping, would seem to be high priority for a free trade organisation, is the proliferation of monopolies and oligopolies in the agricultural sector.

There has been some limited discussion within the WTO about state trading enterprises in agriculture (STEs). The USA has been critical of the public or semi-public export monopolies maintained in for example Canada and Australia. Several developing countries have similar organisations as well, both for export and import trade.

As pointed out in a report from the South Centre (Murphy 1999a), US criticism of STEs can be applied with at least equal justification against the private corporations in control of US agricultural exports. Those corporations, in addition, have strong links to the chemical, seed and food industries, and thus a great measure of vertical control of the whole chain from technology development to consumer product.

This is only one instance of a more general problem with the WTO conceptual framework, which is still largely rooted in the increasingly fictional assumption that trade is something that takes place between countries, or to be exact, between nationally based companies. The fact that transnational corporations by and large are in control of global agricultural trade is not reflected in the present AoA. Their role and presence is not acknowledged, much less disciplined.

The GATT/WTO acknowledges the need for Special and Differential Treatment measures (SDT) for developing countries. In a number of the WTO agreements, there are both specific exceptions and derogations for developing countries and direct support commitments from developed countries. In the AoA however, SDT measures are very limited. Only LDCs have permanent exceptions (from all reduction commitments). All other developing countries have a longer implementation period (10 years instead of 6), and a lower level of reduction commitments (2/3 of the developed country level). For minimum market access the requirement is 4 % for developing countries, as opposed to 5 % for developed countries, plus a longer implementation period (until 2004 instead of 2000).

In addition, there are some minor exceptions for so called S&D support in developing countries, including support to investments and to rural development. Just like green box support, these are not included in AMS calculations. Developing countries also have a higher threshold level (10 % instead of 5 %) for when AMS values trigger reduction requirements (the de minimis rule).

Both developing countries themselves and many development NGOs have called for more generous SDT rules. Proposals include new or extended derogations or support measures, but also technical adjustments to the present agreement, which as implemented has tended to affect developing countries more than developed. One important reason, as already noted, was that actual consequences of the commitments were in many cases not clear to developing countries when the agreement was concluded in 1994.

There is also an unresolved conflict between developing and developed countries about the so called Marrakesh Decision, which promised special compensation for LDCs and NFIDCs if affected by higher food import costs, but has not been implemented.

Many of these issues are technically complicated and will not be further explored here. What follows is a brief listing of a number of proposals which have been formulated. Some of these measures can alternatively or initially be implemented on a bilateral basis.

• Elimination of all tariffs for LDC products in developed countries. The EU has in principle given its support to this proposal, although with some unclarity regarding certain 'sensitive' agricultural products. Can be implemented bilaterally without waiting for agreement in the WTO.
• Allow asymmetric relations in free trade agreements. This issue was central in the recently completed renegotiation of the Lomé agreement. The preferential access for ACP countries to EU markets is not compatible with the GATT principles (non-discrimination), but has been covered by a special derogation. Unless the WTO accepts a similar solution for the renegotiated agreement, it is likely that ACP countries will meet higher tariffs, as the EU is not prepared to extend similar access to all developing countries.
• Create preferential access to developed country markets through other means, for example the GSP.
• Give preferential access to developed country markets for processed goods (exception from tariff escalation).
• Provide resources for capacity development and financing for expanded participation in negotiations. Important especially for LDCs but also for many other developing countries. Can be done either through the WTO or bilaterally.
• Allow developing countries to use (new) border protection measures for 'green box' purposes (or wider). The green box is largely meaningless for developing countries as they cannot afford agricultural policies based on direct support.
• Allow developing countries to change their tariff bindings. Many developing countries, especially among the poorer ones, bound their tariffs on their actual 1994 level (often zero), while most other countries tried to bind them on the highest possible level to retain flexibility.
• Allow developing countries to recalculate their AMS values. Most developing countries reported zero or negative AMS in 1994.
• Allow developing countries to use the Special Safeguard Clause (SSG) even if they did not claim that right in 1994.
• Allow increased flexibility regarding forms of internal support.
• Eliminate all restrictions for internal support in countries lacking basic food security.
• Allow a proportionally higher level of internal support in countries with a large part of the population engaged in agriculture.
• Allow the 'credit' of unused internal support allowances to be saved for later use.
• Increase the de minimis level for developing countries.
• Allow developing countries to subsidise exports by support to marketing, processing, compliance with SPS requirements, etc.

At the conclusion of the Uruguay Round, expectations were that the new AoA would lead to increasing prices for agricultural products on the world market. This would benefit most countries, both the developed and the majority of the developing. The exception are those countries which are dependent on food imports (NFIDCs/LDCs). They would suffer from increasing costs to maintain their supply of basic foodstuffs.

At the concluding Ministerial of the Uruguay Round in Marrakesh 1994 a decision was adopted which acknowledges this problem and establishes the responsibility of developed countries to compensate affected countries. The decision is not formally a part of the AoA, but the agreement refers to the decision and commits developed countries to comply.

Even though the AoA has not resulted in the expected price increase, costs of maintaining the food supply in NFIDCs/LDCs have increased, according to FAO calculations with 20 %. One important factor is that less of the food arrives as development assistance, and more on purely commercial terms. Another contributing factor is increased price volatility on the world market.

So far, nothing substantial has been done from developed countries to assist import-dependent countries. Formally, they can correctly claim that the Marrakesh Decision does not specify either at which levels of price increase measures need to be taken, which these measures should be, or which channels (multilateral or bilateral) should be used. It is also questioned whether the cost increase can be attributed to the AoA at all.

Regardless of formal correctness, the moral responsibility is obvious, as pointed out by practically all NGOs. Concrete proposals for action are most often along the following lines (see for example ActionAid 1999):

• The legal status of the decision must be clarified by incorporating it into the AoA.
• Clear criteria must be laid down specifying at which cost levels support measures shall be triggered, and/or an automatic triggering mechanism be created.
• The WTO must take the responsibility for organising the mechanisms of support, for example in the form of a special fund built up by regular contributions from developed countries.

It is clear from the material presented and discussed in this paper that a serious consideration of food security and ecological sustainability concerns will have fundamental implications for the upcoming renegotiation of the AoA. This section first draws some general conclusions, then gives some concrete policy proposals based on those.

• A 'right to export' is not compatible with food security or ecological sustainability
  The most basic assumption of the AoA, and indeed of the WTO agreements as a whole, is that increased trade is always to the benefit of all parties involved. In practice, this translates into a very strong bias on the side of the exporter. As discussed above, the AoA in essence establishes a 'right to export' by mandating minimum market access and restricting the means available for protection of domestic production, including protection against exports at dumping prices. Although as yet only modestly implemented, the ultimate objective is elimination of all barriers to exports.

In contrast, as discussed in the beginning of this paper, both sustainability and food security concerns argue strongly against further expansion of agricultural trade. Ecological sustainability absolutely requires basing agriculture primarily on local resources and maintaining a high level of nutrient and organic matter recirculation. Food security depends on a variety of factors, but almost invariably it requires stable economic conditions for domestic food production. This is especially true for the food security of the rural poor in developing countries, who as a rule depend mainly on local production with few external inputs.

This is a fundamental conflict which must be addressed. As noted above, little empirical analysis yet exists of the effects of the AoA, but the studies available clearly indicate that liberalisation of agricultural trade has the potential to rapidly undermine local food production. Furthermore, whether or not empirical evidence exists, it is a fact that this effect is precisely the intended one. The whole point of liberalisation is to allow free price competition between agricultural producers across borders, in order to eliminate production where costs are higher and increase it where they are lower. This policy goal, to optimise economic efficiency by eliminating barriers to trade, is given priority over all others.

It is not possible for the AoA to seriously address food security or ecological sustainability, or indeed any other policy goals relevant to food and agriculture, as long as absolute primacy is maintained for the 'right to export'. There must instead be a right to maintain separate national markets for certain or all agricultural products when a country deems this necessary to meet its policy goals.

• Ad hoc derogations for 'non-trade concerns' are not sufficient
  There is no reason to assume that the effects of agricultural trade liberalisation will be substantially different if the process is delayed or moderated by various ad hoc derogations from the AoA requirements. Longer implementation periods or specific time-limited waivers will not make a negative effect less so. If there are fundamental ecological sustainability or food security reasons for limiting agricultural trade liberalisation today, there will likely be in 5-10 years as well. In the main, as we have seen, price differences reflect different cost levels, most of which result from objective factors which will not change with time. Important competitive advantages flow primarily from favourable natural conditions, and to a lesser extent from social, economic and cultural factors such as technical development, access to capital, or the national legislative and regulatory environment. Cost levels can be influenced to some extent by, for example, lowering national environmental standards, introducing new technology or relaxing tax policies. But more important factors such as a lack of fertile land, sunshine or rainfall cannot be changed.

Conversely, the advantages of 'natural exporter' countries will remain. The overall net effect of agricultural trade liberalisation will be to give those countries a larger share of global food and feed markets, at the expense of less advantaged regions. Again, this is entirely intentional and completely in line with the theory behind trade liberalisation. In fact, it is the main argument for trade liberalisation, as it increases overall efficiency (reduces overall cost) in the global food production system.

Needless to say, the basic criteria for ecological sustainability, determined primarily by biological and geological factors, will definitely not change. Nor is it likely that the dependence of food security on domestic and in particular local production will decrease. More likely it will increase, as pressure on natural resources grows over the coming decades.

The problem, in sum, is not the pace of liberalisation. It is that its effects, whether they happen sooner or later, in important respects contradict other important policy goals. Framing policy proposals in terms of derogations, as both developing countries and many development NGOs have done, is therefore insufficient and perhaps even counterproductive, as the very notion of derogations imply acceptance in principle of the measures temporarily derogated from.

What needs to be squarely confronted is the assumption that all perspectives on trade save the most narrow economic one are somehow external, 'non-trade', concerns which only exceptionally and marginally should be allowed to influence the design of trade agreements. This is not only presumptuous, but also, as was evident from the failed Seattle ministerial, unrealistic and improductive. As regional and global trade agreements are rapidly evolving into the primary arena of international political life, and ambitiously demand increased influence over central livelihood areas such as food and agriculture, it follows that they will also have to accept the difficulties that come with a political role. In particular, they must learn to deal with many conflicting policy goals, among which trade liberalisation is only one.

• Agricultural trade cannot be primarily understood as a North/South issue
  As is obvious from this paper, the agricultural policy goals of different countries are very closely related to their objective position in world agricultural trade. The 'natural exporters' strongly push for liberalisation because it is in their economic self-interest. On the other end of the scale, countries like Japan strongly resist liberalisation because they are trying, against all economic odds and to achieve entirely different policy goals, to maintain a domestic agricultural production. In between, most other countries pursue more ambiguous policies, often supporting liberalisation in some respects, but not in others, depending on how they perceive their national self-interest.

This is equally true for developing as for developed countries. There is no common 'developing country interest' in agricultural trade matters, no more than there is a developed country one. Nonetheless, many of the development NGOs have tended to apply a North/South filter to the AoA, more or less indiscriminately supporting all demands of developing countries. This has frequently led to over-simplification and blurring of important distinctions.

In particular, the distinction between food security concerns and agricultural export concerns is crucial to uphold. For food security, the key factor is the ability to provide stable conditions for the development of domestic production, including protection from import pressure as needed. For agricultural exports, the central issue is improved access to foreign markets. These are entirely distinct, and in fact often contradictory, as improved market access for one country equals reduced protection for another. This can involve two developing countries just as well as one developed country exporter and one developing country importer. It needs to be asserted that food security is the priority concern.

From an ecological sustainability perspective, it is also not clear that increased developing country exports to developed countries should always be supported. For example, as related above, it is clear that feedstuff imports from developing countries have contributed considerably to unsustainable production patterns in European animal production, in exactly the same way as imports from developed countries. In this case, sustainability must be the priority concern.

To complicate matters even further, it is not uncommon that export-oriented production in developing countries competes with local food production for various resources, and it also tends to be less ecologically sustainable.

This does not in any way imply that providing increased export opportunities for developing countries is unimportant. On the contrary, it is crucial to their general economic development, and for reasons of global justice this should be reflected in the AoA by expanded preferential access arrangements giving developing country exports, as a rule, priority over comparable developed country exports. But it must also be clear that food security and ecological sustainability are the more fundamental concerns.

• Europe plays the key role in agricultural trade negotiations
  For a number of reasons it is Europe, and in particular the EU, which holds the key to the outcome of the AoA renegotiation.

The EU is one of the two largest agricultural trade actors, a leading exporter as well as one of the largest importers in many products. It is under pressure as the largest provider of agricultural support, much of which directly or indirectly contributes to export dumping.

In a historical perspective, the EU countries in their capacity as colonial powers shaped many of the basic trade patterns still with us today. European colonies, not only present developing countries but also the USA and Australia, had their agricultural production system designed to supply the European market, and to a large extent they still do. The EU history in agricultural exports is very short, and in contrast to all other major exporting nations, it is not really a net exporter in volume terms, mainly because its animal production is massively based on imported feedstuffs. Also largely because of its colonial history, the EU has strong links to developing countries both economically and culturally, including through development co-operation, which enjoys strong public and political support.

For other historical reasons, in particular two recent wars, food self-sufficiency is a deeply entrenched policy goal in Europe, again with solid public support. The new 'multifunctionality' rhetoric builds on genuine popular sentiment in a way that previous 'competitiveness' rhetoric, introduced at the time of the 1992 reform and the conclusion of the Uruguay Round, never could. In this respect, the European attitude to agriculture is much closer to that found in most developing countries than to that of the USA.

All in all, the EU, like the majority of developing countries, does have a choice in agricultural policy, a possibility to opt for a balanced agricultural production system which does not presuppose large agricultural exports. The 'natural exporters' do not have that choice. Unless they can defend or preferably expand their export markets, they will have no outlet for large parts of their production and their economy will suffer. The EU is not dependent on exports in this fundamental way. On the whole, its net production fits reasonably well to its population. If it were only allowed by trade rules to reorganise its agriculture to use domestic feedstuffs instead of imports, there would be no major surplus problem and no need for export dumping.

On this basis, there is a potential for consensus between the EU and most developing countries regarding the AoA. The effects of export dumping is the major grievance developing countries have with the EU. Apart from this, there is considerable overlap between developing country policy demands and the 'Friends of multifunctionality', especially in the insistence on flexibility to protect domestic production for reasons of food security, food safety, environment, rural development and culture.

The other alternative for the EU, like for developing countries, is to accept a continuation of the liberalisation policies, as foreseen in the present AoA, at a slower or faster pace. As noted above, regardless of the pace, the end result will be the same, a restructuring of world agriculture which will concentrate much of world production to 'natural exporter' countries.

But by virtue of its economic and political strength, it is the EU, not developing countries, which holds the power to decide what direction agricultural trade policy will take.

To put the conclusions above in more concrete terms, two major changes in the AoA are called for.

First, and most fundamentally, a rebalancing between the rights of exporters and those of producers for domestic markets. The two key measures are to eliminate all export dumping and to restore freedom of choice in national agricultural policy, including the right to opt for a self-reliance strategy. Such a rebalancing would address most of the food security and ecological sustainability concerns with the present AoA, as well as most other 'non-trade' or 'multifunctionality' concerns.

Second, a systematic strengthening of Special and Differential Treatment (SDT) measures for developing countries, to address remaining food security concerns plus the role of agricultural exports for general development. Key measures here are a food security guarantee for NFIDCs/LDCs, and mandatory preferential treatment for all developing country products in developed country markets.

Export dumping practices are the most extreme manifestation of the pro-export bias in the present AoA, and the major cause of disturbances in developing country markets threatening food security. To eliminate export dumping in the broad sense outlined above would involve at least the following measures.

• Prohibit direct export subsidies
  The prohibition of direct subsidies is relatively straightforward, as their construction is quite transparent. In addition, the EU is now the only large scale user of those subsidies.
• Allow exports from supported markets only if monetary equivalent of all support is added to export price
  Indirect export subsidies are now the main form of dumping, but much more difficult to discipline. Short of totally banning all exports from supported markets, the most workable system is probably to utilise the existing and accepted model for calculation of monetary support equivalents (PSE) in order to arrive at a 'real' price which includes the value of all support systems, monetary as well as non-monetary. Export of a product benefiting from any combination of public support (direct payments, export credits, free public services, or other) would be allowed only if the exporting country applied an export levy equalling the value of that support.

In principle, all kinds of public support should be included in the calculation, whether now classified in the amber, blue or green boxes. But it may be reasonable to exclude some types of green box support, notably those where there is a clear relation between the support and a specific service provided by the farmer, such as maintenance of historical features in the agricultural landscape. There should also be, following normal practice and to avoid excessive administration, a de minimis level of support allowed before a levy has to be applied. (The present box division, it may be noted, would not have any function under this proposal and should be abolished; see also below.)

• Discipline export monopolies and oligopolies, both public and private
  Much of global agricultural trade is controlled by public or private monopolies and oligopolies. Especially the transnational corporations, typically present in a large number of countries world-wide, play a central role in controlling price levels. Despite this, their activities are as yet entirely undisciplined by the WTO. As a minimum, the transparency requirements valid for WTO members must be extended to transnational corporations, in order to allow the WTO to monitor compliance with export dumping rules. There would also be reason to review the activities of the corporations from a competition perspective, in particular relating to the use of dumping prices for penetration of new markets or elimination of smaller competitors.
• Strengthen the rights of the import country to apply countervailing measures
  Given the difficulty for an importing country to prove dumping practices, and the fact that dumping is typically practised by strong actors (developed countries, transnational corporations) against weaker ones (developing countries), there should be a right to apply countervailing measures already on suspicion of dumping, subject to later clarification. In addition, the burden of proof should be shifted to the exporting country.

The right to freely choose the basis of national agricultural policy is effectively abolished by the present AoA, in particular the right to opt for self-reliance as the basic strategy. This severely limits the possibilities to address food security, ecological sustainability, and a number of other relevant policy goals. At least the following is required to restore freedom of choice.

• Remove minimum market access requirements
  The requirements for minimum market access serve only one purpose, to force open new markets to exporters. They amount to a direct prohibition against food self-sufficiency.
• Allow all types of internal support
  Restrictions for internal support were introduced in the AoA to reduce world market distortions resulting from different levels of support between countries, in other words export dumping though indirect export subsidies. If exports from supported markets are regulated as proposed above, no distortions will occur, so there will no longer be a need to regulate internal support schemes. Their size and design can be entirely left to national governments to decide.
• Allow all types of border protection
  Restrictions on the form and level of border protection (tariffication and tariff reduction commitments) were introduced in the AoA to facilitate the process toward elimination of all border protection in a complete free trade regime. As repeatedly noted, the absolute priority accorded to this policy goal is not compatible with a serious consideration of other legitimate policy goals such as food security and ecological sustainability. Flexibility in this respect therefore needs to be restored both for developing and developed countries.

In order for developing countries to have a choice in agricultural policy, they must be allowed to use border protection measures, as internal support is not a realistic option for financial reasons. This requires that the limits resulting from general AoA rules and from country tariff bindings are lifted.

But there is also a strong argument for removing the limits on border protection in developed countries. Border protection is essentially an alternative method for achieving the same goal as internal support, a higher level of compensation to producers on a domestic market. In comparison, border protection has the advantage of being much more transparent, because its effect is directly on the price level of the protected market. Consumers in that market get more correct information about the real cost of production than when part of the bill is picked up by internal support measures paid via taxes. Also, in the context of this proposal, there is much less insecurity and complication involved in establishing the correct export price from a border protected market than in estimating the value of internal support by way of PSE calculations. As border protection thus substantially contributes to market transparency both on domestic and global markets, it seems advisable to encourage the replacement of the patently intransparent internal support measures favoured by the present AoA.

In addition, it must be noted that the attempt to reduce border protection levels in developed countries under the present AoA has been almost entirely unsuccessful, due to advanced technical obstruction. There is obviously a lack of political will to actually reduce protection, although there is a wish to make believe that reduction is occurring. The meaningfulness of continuing this ineffective exercise must be questioned. It should instead be acknowledged that protection of domestic production is a legitimate policy choice, as long as it does not involve export dumping practices harming markets in other member states.

There are also compelling reasons to reverse the tariffication process and again allow a variety of border protection measures. Although administratively more complicated, quota systems or variable import levies have great advantages when the purpose is to flexibly regulate import volumes over time. In particular, they allow countries to maintain a desired level of protection for the domestic market without unnecessarily blocking imports. Tariffication was introduced as part of the strategy to gradually eliminate all border protection on all products, mainly because a unitary system of protection facilitates comparison. If flexibility is reintroduced as to the level of protection, there is no longer a reason to maintain the ban on other systems.

No doubt these policy proposals will immediately be denounced as protectionist by free trade fundamentalists. Measured against the extreme exportism of the present AoA, it may be understandable if they appear so. The basic idea is, however, that agricultural trade policy should provide space both for exports and for protection, without making either an -ism, that is, an end in itself.

An agricultural trade agreement along these lines would in no way stop trade, nor the continued development of free trade relations. What it would do is give countries a choice regarding to what extent and in what products they would participate in agricultural free trade. The basis of trade is mutual economic benefit. Where countries judge that it exists, they will certainly want it. If, as some claim, it always exists, free trade in agricultural products will be universal. If, as this paper assumes, this depends on circumstances, the development of free trade will be more patchy.

What these proposals would stop is only the right of exporters to force their way into domestic agricultural markets against the will of the respective governments. They amount, as it were, to a prior informed consent requirement.

On the other hand, these policies would arguably be much more effective than the present AoA in guaranteeing fairness in trade relations. The requirement that all internal support must be directly reflected in export prices should virtually eliminate all distortive effects on other markets, while the present AoA may in fact have increased them by stimulating the expansion of intransparent blue and green box support systems. There is also reason to believe that these proposals could achieve the increase in world market prices which the AoA so far has failed to deliver. By requiring exports from protected and supported markets to take place at their real price levels, the cost levels of those (large) markets would have more influence on world trade pricing than today.

The basic idea of these proposals is close to that of the 'development box' proposal about a 'positive list' mechanism by which developing countries would be allowed to opt into the AoA commitments product by product. The mechanism proposed here adds flexibility, as there would be no need to opt in completely, but a right to increase (or decrease) participation in free trade gradually. Even from a highly protected or supported market, there would be a possibility to export, as long as some buyer were prepared to pay the full price, as would likely continue to be the case with for example specialty French cheese and wine. Nor would a higher level of protection in one product compromise the possibility to participate fully in free trade with another. An important difference from the 'development box' is however that in this proposal all countries would have this right, while SDT concerns are addressed additionally (see below).

It can be argued that neither food security, ecological sustainability or other agricultural policy goals legitimise total freedom for countries to decide protection levels and internal support measures. Unless there is some international mechanism to ensure a proper relation between such policy goals and measures taken, countries are also free to expand protection for any other reason. There is much to this argument, but under present circumstances, leaving those decisions to individual countries seems the only realistic option. If a specific list of allowable policy goals would be included in the AoA, the decisions would in practice fall to the WTO dispute settlement mechanism, which is an intransparent and undemocratic body, completely unsuitable for this task. In the absence of a global political body with the competence and transparency necessary to credibly make those difficult decisions, they will have to be left to national political decisionmaking.

The role of the WTO would under these proposals be centred on the monitoring and policing of the interface between protected/supported markets and global trade. In particular, it should focus on the implementation of the anti-dumping measures, and on ensuring non-discrimination in a situation where countries have increased freedom to devise border protection measures. Monitoring of internal support schemes would continue to be necessary, as the control of export pricing would require constantly updated information.

In terms of the NGO discussion, these policies would establish a framework for 'food sovereignty' on the national level, providing governments with the freedom to design national policy. It should be emphasised, however, that this implies no guarantee for 'food sovereignty' in the more important sense of self-determination in food security matters on regional or local levels. This will still be dependent on national political processes. The important difference from the present situation is that WTO rules would no longer limit government freedom to react and act on public opinion pressures.

While most food security concerns of developing countries would be addressed by an overall rebalancing to restore freedom of choice in national agricultural policy, the specific and urgent problems of NFDICs and LDCs would not. Indeed, they may be worsened, if such a shift succeeded in reversing the falling trend in world food prices. A food security guarantee needs to be included in the new AoA, but concrete action cannot wait for the possibly distant conclusion of a new negotiation round.

• Include a legally binding food security guarantee in the new AoA, connected to a fund with mandatory contributions
  The food security of net food importing countries is the responsibility of the international community. While it is not self-evident that the WTO is the most suitable intergovernmental body to take this responsibility, governments have already chosen it by adopting the Marrakesh Decision. There should be a general food security guarantee for NFDICs/LDCs in the new AoA. A fund should be created with mandatory contributions in relation to share in global agricultural exports, plus opportunity for additional voluntary contributions. The criteria for use should be developed in co-operation with other relevant intergovernmental bodies, and perhaps the administration should also be shared. The full responsibility for securing adequate funds should however fall to the WTO.
• Interim action is needed to meet urgent needs
  Regardless of the solution eventually negotiated, governments have a responsibility to honour the Marrakesh Decision. The assumption that NFDICs and LDCs have to provide conclusive evidence that increased food import costs are due to AoA effects is unreasonable.

Although developing country exports would likely benefit considerably from a total elimination of export dumping, they should also be provided systematic preferential treatment when exporting to developed countries. At least the following should be implemented:

• Zero border protection for LDC exports to developed countries
  Agricultural exports is one of the few possibilities for LDCs to leverage economic development, and the total volume is small. Even if free imports in a few cases might compete with domestic production in developed countries in a way which conflicts with national policy goals, such concerns should be waivered and if necessary other solutions devised to compensate national producers.
• Mandatory differentiation of border protection in favour of developing country exports
  While developed countries should also have the freedom to regulate imports by any combination of border protection measures, they should be required to differentiate all forms of protection in favour of developing country exporters. The differentiation could be in the form of lower tariff rates, priority access to quotas, etc., as applicable. Differentiation should apply to all products, and be strongly enforced, for example by making all border protection illegal unless the differentiation is upheld.
• Higher de minimis allowance when exporting from a protected market
  Developing countries should be subject to the same rules as developed when exporting products for which they have border protection or internal support. But there should be a more generous de minimis allowance before the requirement to add an export levy is triggered.

This section is a brief guide to the NGOs most actively involved in AoA issues. The list is intentionally short and only contains organisations which have capacity for independent analysis, have published original material and have staff exclusively or mainly working with WTO/AoA issues on an international level. All mentioned here have produced material directly useful in preparing this report. Many others, particularly among development NGOs, are active mainly on a national level, but often stably networked with several of those mentioned below.

A group which for natural reasons is both active and competent is farmers' organisations and other agricultural interest groups. In general, however, they do not share the basic assumptions of this report. The exceptions are the radical farmers' organisations in Via Campesina (listed below), and also the organic agriculture organisations in IFOAM (International Federation of Organic Agriculture Movements), which show a growing interest in trade policy issues but still have some way to go both in capacity and competence.

ActionAid
Development NGO based in the UK but with strong regional offices, of which several actively work on agriculture and trade (for example Brazil, Ethiopia, the Gambia, India and Kenya).

Contact: Laura Kelly (laurak@actionaid.org.uk)

CLONG
(Liaison Committee of Non-Governmental Development Organisations to the EU)
A large umbrella organisation for development NGOs in the EU, of which several are active on agriculture and trade. Maintains a Joint Food Security Group in co-operation with Euronaid.

Contact: Peter With (Danchurchaid; Chair of Joint FSG; pw@dca.dk)
Web: www.oneworld.org/liaison

Collectif Stratégies Alimentaires (CSA)
Small agriculture policy think tank in Brussels. One of few able to span both EU and global policy. Also key member in Belgian NGO coalition Food Sovereignty Platform.

Contact: Daniel Van Der Steen (csa@csa-be.org)

Focus on the Global South
Research institute attached to Chulalongkorn University in Bangkok. Has produced one of the best critical introductions to the AoA (Kwa & Bello 1998).

Contact: Aileen Kwa (a.kwa@focusweb.org)
Web: www.focusweb.org

Food First (Institute for Food and Development Policy)
A small institute also acting as spider in FIAN (Food International Action Network). Based in Oakland, California.

Contact: Anuradha Mittal (amittal@foodfirst.org)
Web: www.foodfirst.org

Institute for Agriculture and Trade Policy (IATP)
Rather large think tank based in Minneapolis. Long track record in US and global agriculture policy. One of the spiders in the NGO networking on agriculture for the Seattle WTO ministerial. Lots of internet-based information, including excellent web archive of WTO materials.

Contact: Sophia Murphy (smurphy@iatp.org)
Web: www.iatp.org or www.wtowatch.org

South Centre
The South Centre is not an NGO but an IGO (intergovernmental organisation) with around 50 developing countries as members. Assists developing countries with policy studies Based in Geneva.

Contact: Rashid Kaukab (kaukab@southcentre.org)
Web: www.southcentre.org

UK Food Group
A broad coalition of UK NGOs.

Contact: Jagdish Patel (ukfg@dial.pipex.com)

Web: www.ukfg.org.uk

Via Campesina
A global coalition of small farmers' organisations. Based in Tegucigalpa, Honduras. Particularly strong in Latin America, but member organisations also in North America, Asia and Europe.

Contact: Rafael Alegría (via@sdnhon.org.hn)

Web: www.sdnhon.org.hn/miembros/via

WWF International
The trade and environment programme of the WWF has a solid history in trade issues. Not specialised in agriculture, but lately increasingly involved.

Contact: Mireille Perrin (mperrin@wwfnet.org)
Web: www.panda.org

Many of these references can be found in electronic formats on www.wtowatch.org and/or the websites of the publishers.

ActionAid. 1999. A recipe for change. Food security -- the key issues for the WTO Ministerial Conference, Seattle, November 1999. Briefing Pack. ActionAid, London.

APRODEV. 1999. Trade and the hungry. How international trade is causing hunger. Ed John Madeley. Association of World Council of Churches Related Development Organisations in Europe (APRODEV), Brussels.

Bello, Walden. 1999. Asia, Asian Farmers, and the WTO. Ag-impact mailing list, IATP, 2 July 1999.

CLONG. 1998. Global Food Security Ignored? Between EU's 1999 reform of the Common Agricultural Policy and the WTO Ministerial in Seattle. Madrid, November 1998. Conference report. Liaison Committee of Development NGOs to the European Union (CLONG), Brussels.

CLONG. 1999. Putting Food Security into the WTO. Fourteen recommendations addressed to the European Commission and the Council of Ministers on the negotiating mandate of the EU for the WTO Conference of Ministers in Seattle to promote world-wide Food Security and Sustainable Development. Liaison Committee of Development NGOs to the European Union (CLONG), Brussels.

Confédération Paysanne. 1999. How to subject WTO to basic human rights. November 1999. Confédération Paysanne, Bagnolet (FR).

COPA-COGECA. 1999. Le modèle européen d'agriculture un modèle d'avenir. Novembre 1999. Comité des organisations Professionelles Agricoles de l'Union Européenne et Comité Général de la Coopération Agricole de l'Union Européenne, Bruxelles.

Cuba et al. 2000. Agreement on Agriculture: Special and Differential Treatment and a Development Box. Proposal to the June 2000 Special Session of the Committee on Agriculture by Cuba, Dominican Republic, Honduras, Pakistan, Haiti, Nicaragua, Kenya, Uganda, Zimbabwe, Sri Lanka and El Salvador. Document G/AG/NG/W/ 13. 23 June 2000. WTO, Geneva.

ERA Consumer. 1999. Food Sovereignty. Malaysian NGO Position Paper on Agriculture. November 1999. Education and Research Association for Consumers, Selangor.

FAO. 1999. Assessment of the Impact of the Uruguay Round on Agricultural Markets and Food Security. Document CCP 99/12 Rev. October 1999. FAO, Rome.

Food Sovereignty Platform. 1999. Placing food sovereignty before commercial interests. Position paper. Food Sovereignty Platform, Brussels.

Jadot, Yannick. 1999. Les enjeux des négociations multilatérales pour les pays d'Afrique, des Caraïbes et du Pacifique. Document réalisé pour le Ministère des Affaires étrangères (FR). Janvier 1999. Solagral, Nogent sur Marne.

Kaukab, Rashid. 1999. TRIPs, TRIMs, Agriculture and Services: a Southern Perspective. Presentation at Wilton Park Conference: Multilateral Trade Negotiations: The Way Forward. 28 June - 2 July 1999.

Kwa, Aileen. 1999. Will food security get trampled as the elephants fight over agriculture? Focus on Trade #36, July 1999. Focus on the Global South, Bangkok.

Kwa, Aileen & Bello, Walden. 1998. Guide to the Agreement on Agriculture. Technicalities and Trade Tricks Explained. November 1998. Focus on the Global South, Bangkok.

Madeley, John. 2000. Trade and hunger. An overview of case studies on the impact of trade liberalisation on food security. October 2000. Forum Syd, Stockholm.

Manitoba Cooperator. 1999. Canada needs a food policy. Editorial. July 1999. Reprinted in The Ram's Horn 173, September 1999.

Murphy, Sophia. 1999a. Market Power in Agricultural Markets: Some Issues for Developing Countries. T.R.A.D.E. Working Paper 6, November 1999. South Centre, Geneva.

Murphy, Sophia. 1999b. Trade and Food Security. An Assessment of the Uruguay Round Agreement on Agriculture. Catholic Institute for International Relations, London.

Oxfam. 2000. Agricultural Trade and the Livelihoods of Small Farmers. Discussion paper for DFID towards development of a White Paper on Globalisation. March 2000.

Oxfam GB Policy Department, London.

Ritchie, Mark. 1999. Eliminating Export Subsidies: One Way Forward. Ag-impact mailing list, IATP, 18 November 1999.

Ritchie, Neil. 1999. Testimony on the World Trade Organization Negotiations. U.S. Agriculture Policy Forum. June 7, 1999. St Paul, MN. Institute for Agriculture and Trade Policy, Minneapolis.

Rosset, Peter. 1999. Food First Trade Principles. Backgrounder. Fall 1999, Vol 5, No 2. Institute for Food and Development Policy, San Francisco.

Solagral. 2000a. Is it worth defending the concept of multifunctionality of agriculture? Contribution to CSD 8. February 2000. Solagral, Paris.

Solagral. 2000b. World Trade, Food Production and the Diverse Roles of Agriculture. Discussion paper for CSD 8. February 2000. Solagral, Paris.

South Centre. 1997. Universal Food Security: Issues for the South. South Centre, Geneva.

South Centre. 1999. Issues of Interest to Developing Countries in the Future Negotiations on Trade in Agriculture. South Centre, Geneva.

UK Food Group. 1999. Agricultural trade rules: are they stacked against the poor? Position paper. November 1999. UK Food Group, London.

Van Der Steen, Daniel; Danau, Alex; Poznanski, Marek. 1999. L'organisation Mondiale du Commerce et l'agriculture. La souveraineté alimentaire menacée par les accords commerciaux. Novembre 1999. Collectif Stratégies Alimentaires, Bruxelles.

Via Campesina. 1999. Via Campesina demands: Take agriculture out of the WTO. Position statement, 1 December 1999. Via Campesina, Tegucigalpa.

World Food Summit. 1996. World Food Summit Plan of Action. FAO, Rome.

World Food Summit NGO Forum. 1996. Profit for few or food for all? Food Sovereignty and Security to Eliminate the Globalisation of Hunger. NGO Forum Statement to the World Food Summit. 13 November 1996. Rome.

WTO. 2000. Agricultural Trade Performance by Developing Countries 1990-98. Background Paper by the Secretariat. Document G/AG/NG/S/6. 23 May 2000. WTO, Geneva.

WWF International. 2000. Recommendations to CSD-8. Sustainability Assessment of Agricultural Trade Liberalisation. Position Paper. February 2000. WWF International, Gland (CH).

1 'Why so many definitions? Because everyone is desperate to find a slot that will include them, and if they don’t find one, they make up their own.' Sustainable Farming, Spring 1990, REAP Canada, Ste-Anne-de-Bellevue, Quebec.

2 All figures in this section should be taken as indicative. They were rather rapidly compiled from a few different sources, mainly the USDA Foreign Agricultural Service (www.fas.usda.gov), FAO, and the World Bank. All numbers refer to trade volume and generally reflect the situation in the mid-90s. Averages of several years were used in most cases, as fluctuations are considerable.

3 There is some confusion about this figure. The WTO reports a considerably lower percentage (WTO 2000). The main reasons for the discrepancy appears to be that WTO figures exclude all intra-EU trade (that is, counts the EU as one country), and include all of Eastern Europe and Russia plus countries like Korea, Turkey and Israel in their definition of developing countries.

4 Many details in Actionaid (1999) and Kwa & Bello (1998).

5 Chile, Colombia, Costa Rica, the Philippines, Fiji, Guatemala, Indonesia, Malaysia, South Africa, Thailand.

6 More on how PSE is calculated below in the section on internal support.

7 '... most if not all of those other issues are really developed country concerns -- environmental protection, cultural landscape etc.' (Kwa 1999).