Introduction
As the study of the World Trade Organisation (WTO) matures, scholarship is becoming more specialised. Consequently, the links between the different areas within the subject are weakening, resulting in more fragmented analysis. For example, it is increasingly possible to be an expert on intellectual property without necessarily having an in-depth knowledge of anti-dumping too. This trend brings obvious benefits for the depth of analysis of the various areas, but there is a negative side which can be seen most clearly in the context of agriculture. Agricultural trade comes to be regarded as arcane and technical and of interest only to experts in the field. This article will pursue the idea that there is some value in exploring the resonance between agriculture as a specific topic and the general notion of international trade.
Where scholars have addressed the big picture of international trade, they tend to rely on a range of abstract ideas including constitutional theory, governance and systems theory, which then inform analysis of the subject matter of the subject agreements.Footnote 2 Although this is a very productive way of exploring the new complexities, these methods do not always fully reflect the dynamics within specific agreements like agriculture. Since, according to this picture, the problems that are inherent in international trade and the WTO elucidated at this high level then drive the analysis of the specific subject matter. I suggest that this is an incomplete picture. It is equally possible to begin by exploring particular problems as a means of illuminating more general concerns. In this more ascendant way of looking at things, certain aspects of agricultural trade regulation have direct resonance for other areas covered by the WTO agreements. The analysis I propose is unrelated to systems theory or ideas of governance, but deserves to be reflected on in its own right. We might see these additional dimensions as addressing a different series of questions: the nature of agreement and the different ways in which agreement can manifest itself. Whereas systems theory helps us determine what the subject matter should be within a WTO agreement, or what should be within the WTO agreements as a whole, ideas of governance enable us to see how agreements may be implemented more effectively. In essence, traditional analyses treat agreement as a generic and homogenous idea, while this article looks at the specific nature of agreement in a more contextual and nuanced way.
The two dimensions to explore here are the nature of international agricultural trade as a polycentric problem and the way accepted concepts and categories influence the way we see problems in international agricultural trade. Polycentrism goes to the nature of the stability of the Agreement on Agriculture, whereas the question of concepts and categories concerns the way in which we represent the problem as coherent to ourselves. These two are important phenomena which are missed or glossed over by traditional approaches. I then briefly indicate how these questions relate to international trade regulation more generally, and what lessons they teach us regarding the true nature of agreement.
Agriculture as a polycentric problem
Agriculture is the problem child of the international trading system.Footnote 3 Seemingly resistant to comprehensive regulation, it lay largely outside the reach of the multilateral trading rules in the General Agreement on Tariffs and Trade (GATT). According to historical analyses of the pre-WTO period, there are two reasons for this exclusion: the agricultural sectors of the two dominant economies, the United States and the European Community (as it then was) were excluded from the rigours of the GATT rules.Footnote 4 The United States’ Agricultural Waiver protected the United States’ use of quantitative restrictions whereby it insulated its domestic markets from adverse GATT panel findings from the 1950s onwards. The GATT was also unable to control effectively the European Community's Common Agricultural Policy (CAP). The CAP operated to protect European Member State markets through a combination of a complex variable import levy (with the potential to operate as a zero quota) and various financial support mechanisms, thus enabling European agricultural products to be more competitive on world markets irrespective of global market conditions. Although the GATT rules on quantitative restrictions and subsidies theoretically applied to the CAP, uncertainty surrounding the scope of the Article XXIV GATT regional trading arrangements exemption meant the CAP was not really successfully challenged in the GATT era.Footnote 5 Inevitably, these exclusions did not engender compliance to the GATT rules by other contracting parties to the GATT. Despite numerous attempts to regulate trade in agricultural products throughout the life of the GATT, success was only achieved through a last-minute compromise between the US and the European Community at Blair House, which resulted in the conclusion of the Uruguay Round of multilateral trade talks and the creation of the WTO. The WTO is an international organisation with separate legal personality presiding over a comprehensive regulatory framework governing many aspects of international trade, including an agreement on international agricultural trade.Footnote 6
The WTO Agreement on Agriculture is regarded as a major breakthrough. For the first time, international trade in agricultural products is regulated through a comprehensive series of rules which aim to reduce and gradually to dismantle barriers to trade over time.Footnote 7 Whether the Agreement applies to a specific agricultural product at all is determined by that product's customs classification code, rather than a generic definition of what an ‘agricultural product’ might be.Footnote 8 Although a product might fall within the Agreement's scope, its rules only apply to a narrow range of trade measures. If a measure is not one to which the Agreement relates, then another WTO agreement will apply to it.Footnote 9 The Agreement on Agriculture regulates market access measures, export subsidies and domestic support measures, or more specifically, domestic subsidies. These measures are each addressed in one of the Agreement's three ‘pillars’: Part III covers market access, Part V addresses export subsidies and Part IV concentrates on domestic support.Footnote 10 Members’ specific reduction commitments on market access measures, domestic support and export subsidies are contained in their Schedule of Concessions which are annexed to the Agreement on Agriculture. Members are not permitted to impose measures in excess of their bound rates, unless this is permitted by the rules. Article 3 makes it clear that the detail in Schedules for domestic support and export subsidies are an integral part of the regulatory framework, although there is nothing in the original Agreement regarding market access.Footnote 11 However, this lacuna was addressed in EC-Bananas III where the WTO Appellate Body made it clear that a Member's Schedule was indeed part of the Agreement on Agriculture and consequently any alleged violation of that Schedule could form part of a complaint under the WTO Understanding on Rules and Procedures Governing the Settlement of Disputes (the Dispute Settlement Understanding or DSU).Footnote 12 The Agreement on Agriculture also permits Members to use trade measures to pursue non-trade objectives like development policies, environmental protection and the preservation of rural ways of living, but only to the extent permitted by the rules.Footnote 13
Despite the new rules, early studies revealed difficulties in implementation. The conversion of non-tariff barriers to tariffs did not open agricultural markets in the way originally envisaged by negotiators and in fact resulted in very high bound agricultural tariffs.Footnote 14 A Reference Finger, Winters, English, Hoekman and Mattoo2002 study by Finger and Winters concluded that significant disquiet about the impact of the rules on the general levels of protection on agricultural markets remained, especially in regard to developed country markets: specifically that, on average, the depth of the tariff cut by developing nations significantly exceeded that of developed nations.Footnote 15 A 2005 study by Anderson and Martin indicated the average tariff imposed by high-income countries (i.e. developed nations against developing nations’ products) was 14 percent, compared with an average tariff of 1 percent on non-agricultural products.Footnote 16 This imbalance in the levels of protection between agricultural and non-agricultural products extended to developing country markets too: according to the Anderson and Martin study, the average tariff on agricultural products on South–South trade was 18 percent.Footnote 17 The study went on to find that many agricultural exporters remained protected by export subsidies and domestic support measures. For the authors, this finding suggested such agricultural producers were still able to sustain existing agricultural production levels because they achieved high incomes irrespective of any decline in world market prices caused by their excess production despite the Agreement on Agriculture's rules which were designed to curtail such activity.Footnote 18 In other words, their production decisions were still largely disconnected from the ‘normal’ operation of the demand for agricultural products: the market for agricultural products was not working without artificial constraints.
These studies apparently indicate that the Agreement on Agriculture is not successful in the way the Uruguay Round negotiators hoped. What is the problem? Why is agriculture so resistant to effective regulation? The traditional response to this question is that comprehensive regulation is relatively new to international agricultural trade.Footnote 19 These ‘new’ rules in the WTO Agreement on Agriculture were designed as only a first step in the long road to liberalisation of the sector and so it is not surprising that they do not adequately reduce restrictive barriers to trade, do not fully address developing countries’ concerns sufficiently or do not truly accommodate non-trade issues like human rights and the environment. In essence, the rules are the problem: the difficulties of international agricultural trade are self-evident, but we have not yet discovered how to modify the rules to accommodate (or not) all the complicated and controversial issues. Academic commentators, civil society representatives and negotiators therefore concentrate firmly on finding the ‘right’ solution based on this apparently shared understanding of what the problem is.Footnote 20
I suggest below that the search for a solution to the problem of international agricultural trade, on which there is universal agreement, is the search for nirvana. Finding a solution suggests it is possible to discover fully the true nature of the problem: all its complexities and its fullest extent are revealed so it appears to be fixed at a moment in time, like a snapshot. It is tidy; there are no loose ends; all the elements neatly fit together in a coherent fashion. The solution is then one which accounts for all these revealed elements and contains them into one scheme which can (but need not) be translated into legal rules.Footnote 21 The reality is very different. The problem of agricultural trade is not tidy, with readily determinable parameters that fit coherently together; it instead resembles a spider's web.Footnote 22 Lon Fuller famously used the analogy of a spider's web to describe situations where there is not one set of disagreements which can be resolved, but many interrelated disagreements which occur simultaneously. There is therefore no final correct description of the problem.Footnote 23 In Fuller's analysis, the various argumentative trajectories represent the different strands of the spider's web. Each point of convergence is a distinct centre for the distribution of tensions across the whole web. Hence, the description of the problem as many-centred, or ‘polycentric’. As with a spider's web, exerting pressure at any of these points is likely to stretch certain strands and break others, so that attempts to resolve any one of these centres of tension will merely increase the tension elsewhere in the web. This means in effect that if a pressure is exerted on any one of these centres, the resulting tensions will be different each time. The problem is therefore in a constant state of flux. Agricultural trade as a polycentric problem is explored fully below.
The studies on the problem of international agricultural trade referred to earlier indicate that the Agreement on Agriculture significantly failed to liberalise international trade in agricultural products: Members continue to restrict access to their domestic markets through high tariffs and further insulate their farmers from the rigours of the market mechanism by the use of domestic and export support measures (where money is available).Footnote 24 Moving forward from this position seemingly requires a reduction in existing barriers to market access and a reduction in export and domestic support measures. However, a decision to liberalise international agricultural trade is not just a decision to reduce the permitted levels of every trade measure: rather the decision to cut one type of measure has implications for every other trade measure Members use. Members may not agree to a deep cut in tariffs if this is not compensated by a lesser reduction in the levels of permitted domestic support; in turn the lesser reduction in domestic support is dependent on the cut in export subsidies. Lesser cuts in tariffs change the way the tensions distribute through the web: an agreement to a smaller cut in tariffs is in turn dependent on deeper cuts in domestic support levels and lesser cuts in export subsidies, or even, deeper cuts in export subsidies and lesser cuts in domestic support as a result. Likewise, if tariff levels are not cut at all, this may mean Members want deeper cuts in domestic support in ways which may or may not translate into differential cuts in export subsidies too. The problem therefore is highly polycentric because it is impossible to determine one cut without deciding on all the others because each decision changes the deliberation of each of the others. Even where one cut is determined, that determination may need to be re-opened as a consequence of subsequent decisions elsewhere: for example, cutting tariffs by 50 percent may not be as satisfactory as it seemed when the original decision was made if later a discussion focuses on whether the Member must cut domestic support by 70 percent.Footnote 25
This analysis isolates just one aspect artificially from a wider picture for the purposes of illustration of the basic idea. I will now discuss other aspects to the problem. Hertel et al. (Reference Hertel, Keeney, Ivanic and Winters2006), following on from a number of key studies on the impact of trade liberalisation post the Agreement on Agriculture, showed that whilst global gains from agricultural liberalisation may accrue, the distributional effects of those gains may not filter down into all sectors, and some segments of society, like farmers in some developing nations, may be adversely affected by liberalisation efforts.Footnote 26 Likewise, in Reference Morrison and Sarris2007, Morrison and Sarris suggested the ‘consensus’, that all countries gain in the same way from fully liberalising their agricultural markets irrespective of their trading partners’ behaviour, is not shared by all WTO Members. In fact many developing and least developed countries are gravely concerned about opening their markets to too much competition too quickly.Footnote 27 Deciding whether or not to take into account the needs of countries at certain stages of development in international agricultural trade regulation adds further layers of disagreement interrelated to those issues discussed above. It is not a different problem to that of reducing the high incidence of trade barriers in agriculture in the way these reports suggest.
The decision to make deep cuts in tariffs is connected to the determination to provide lesser cuts for developing nations over longer implementation periods; this lesser cut is in turn connected to deciding whether other Members win concessions on the level of tariff cut tied to their economic status, for example because they are economies in transition. A deep tariff cut may again only be possible if this is ‘paid for’ by a lesser cut in domestic support levels (i.e. domestic subsidies). However, given that the depth of the tariff cut is influenced heavily by the economic status of the Member and is not an absolute one across all Members, the cut in domestic support may be less to compensate for the lesser reduction in the overall level of tariff barriers. The reduction in domestic support is in turn dependent on agreement on export subsidies. If the cut in domestic support is also significantly modified by concessions predicated on a Member's economic status, this affects the result in export subsidies, which in turn affects outcomes in domestic support. Equally, a smaller cut in tariffs coupled with differential implementation commitments for developing nations, but not for those economies in transition, will change the decision on domestic support, and in turn that on export subsidies.
As in the discussion of reductions of key trade measures, ‘agreement’ on one reduction when development concerns are added into the deliberations is intimately connected to the adjustments on all other measures. The path of tensions created through the web is not miraculously found by merely tracking the different economic exemptions through tariffs to domestic support and export subsidies and making equal compensatory adjustments at each stage for economic status. Rather, the decision on tariffs, including any development concessions, changes the decisions on domestic support (and export subsidies) fundamentally because tariffs remain high on some products and not others in complex ways that adversely (or positively) affect some members more than others. The result is a highly fragmented pattern of tensions which feed into the decision process at each point of convergence.
This picture seems complex enough, but it is evident that other non-trade concerns also feed into the problem, adding yet further dimensions to the problem: for example, preservation of the environment, broadly conceived. The relationship between trade and the environment has occupied trade scholars and policy-makers since the time of the GATT.Footnote 28 A comprehensive discussion of these issues is outside the scope of this article, but there are a number of issues which now potentially feed into the existing regulatory framework for agricultural trade and may influence ideas for reform. A 1999 WTO report on Trade and the Environment made it clear that the WTO was not an environmental organisation, but its existing rules in Articles I, III and XX of GATT would accommodate the use of trade measures for environmental purposes provided these were applied in a non-discriminatory fashion.Footnote 29 Environmental challenges for the WTO should instead focus around the relationship between the WTO and other Multilateral Environmental Agreements, transparency obligations and the nature of restrictions on market access, especially those caused by eco-labelling, standards, and sanitary and phytosanitary (SPS) measures aimed at effectively exporting a specific level of animal husbandry as well as the prevention of disease. Other relevant areas were the use of domestic subsidies to support environmental programmes, promote certain animal husbandry techniques and preserve rural ways of life to undo any damage caused by migration to the cities. These ideas have since evolved to include using trade measures in innovative ways to preserve the environment, especially in agriculture. In 2009, a joint report by the WTO and the United Nations Environment Programme (UNEP) suggested that the inextricable link between trade and climate change could see developing countries in particular using traditional trade measures to ameliorate a drop in major cereal yield of 5–10 percent. The impetus for using trade measures in this way was very important for some African countries, where declines in yield are predicted to be as high 50 percent in 2020 and 90 percent by 2100.Footnote 30 Amongst other practical possibilities, it is suggested that countries could adopt asymmetric tariff regimes so that goods having a lower impact on climate change could enjoy lower tariffs. Domestic support measures could also be offered as incentives for producers to adopt more carbon-neutral production techniques.Footnote 31
Adding this environment dimension to the existing trade measures and development dimensions of the problem of international agricultural trade regulation creates a new pattern of tensions through the web, potentially re-opening decisions that appeared settled. The cut to tariffs now has implications not only for the reductions in domestic support and in turn export subsidies, but also for environmental concerns as well as those on development. The polycentric elements therefore become more diffuse. Members may be less likely to cut tariffs deeply if they also wish to include a differential tariff regime for so-called environmentally friendly products. These decisions may also change the type of concessions for developing nations and even whether some countries should benefit from concessions at all. For example, a member may no longer agree to a lesser tariff cut for a country with an economy in transition if the overall reduction in tariff levels is not as deep as that member originally thought. Further concessions may therefore no longer be appropriate. For this Member there is no longer any special consideration for economies in transition and they instead rank as developed or developing countries (depending purely on those countries’ accession status). Alternatively, another Member might be more likely to agree to a deep cut in tariffs if they believe an environmentally friendly tariff concession is the only pro-environment exception they have to agree to. This decision may itself be revisited, dependent on the outcome of the decision on variable implementation commitments for countries at different stages of economic development.
Similar issues arise for other non-trade concerns like food security, food aid to developing nations, sustainable development and the right to food. Further systemic issues may also be added, including how properly to control adverse Member behaviour within the rules on agriculture; what the relationship is between the Agreement on Agriculture, Members’ domestic agricultural policies (i.e. what is an appropriate degree of state autonomy) and what the relationship should be between international agreements and the WTO rules on agriculture. In each case, the outcome at each point where these issues collide is dependent fully on what is decided at every other point. All these points collide simultaneously, so it is not possible to ‘stop the clock’ whilst one is resolved, leaving the others for later. An outwardly settled point becomes unsettled as Members try to grapple with other interconnected issues and find themselves re-opening the settled decision in order to accommodate the current one.
Ambassador Crawford Falconer's covering letter to the December 2008 Draft Modalities on the Agriculture negotiations states: ‘[c]ertain things are manifestly not yet agreed. And, depending on where we end up on these matters, other areas that seem relatively “stable” may well be revisited.’Footnote 32 This statement indicates that negotiators appreciate the complexity of their negotiating task. However, the statement goes astray because it perceives the problem as linear and capable of resolution, albeit in the long term, whereas recognising international agricultural trade as a polycentric problem is to appreciate that every issue is permanently and intimately connected to every other. In this vision, the outcome at every point is dependent on the outcome at every other point: deciding one issue changes the decision at every other, so it is not possible to determine finally what the problem is, nor determine a final solution. International agricultural trade would be a linear problem if it was possible to make a determination on one point and then move to the next one and the next, assuming the previous ones remains fixed. The history of attempts at regulation throughout the GATT and the Doha Development Round show that this is clearly not the case.
A polycentric problem raises further challenges for multilateral negotiations because not every negotiator understands the problem in the same way. Negotiators are therefore unaware they may be talking past each other. So, negotiations are no longer a question of understanding another negotiator's viewpoint and deciding how far one can compromise with that view before consensus is reached. Instead, the depth of disagreement may be more subtle and reflect fundamental differences in the nature of seemingly settled concepts and classifications, for example whether an issue is trade or non-trade. How each person represents a polycentric problem as coherent to themselves, and particularly how this is subtly affected by their understanding of settled classifications and concepts, is addressed in the next section.
Understanding international agricultural trade as a polycentric problem
The full extent of a polycentric problem cannot be described, as the issues arising at each point are dependent on the outcome at all the other points simultaneously. As a result, the problem is dynamic and shifting rather than homogenous and static, so there are endless variant interpretations of the problem. How each person decides what the problem is, is a constructive process: their understanding is shaped by how they couple together the disparate dimensions to make it coherent, that is, the way the dimensions contribute to particular policy goals and how they fit, or fight, with existing principles.
Each person engaging with the problem makes it coherent in their own way. As a consequence, even slight changes in the way the diffuse dimensions are brought together means that each person sees the problem of international agricultural trade differently. The multiple ways in which these issues can combine means that a single homogenous notion of the problem is unlikely even when negotiators appear to be talking about the same issue. Wittgenstein observed that whilst we might follow pathways of thought in each individual which suggest points of commonality at the point where those thoughts cross, in fact, these crossing points are an illusion and instead reflect only thoughts cutting across one another travelling towards different final conclusions, rather than points where we can truly say agreement on the problem's scope has been established.Footnote 33 This divergence also infuses subtly different meanings into seemingly settled concepts and categories like ‘trade’, ‘non-trade’, ‘development’, ‘environment’ and ‘agriculture’, suggesting that disagreement between negotiators and commentators is a more complex and subtle idea than is acknowledged by the existing literature.Footnote 34
It is tempting to think that making the problem of international agricultural trade coherent involves settling such difficulties as, for example, how best to balance trade and non-trade objectives to produce the most efficient outcome in the light of all the competing political views. In this view, the categories ‘trade’ and ‘non-trade’ are seen as concepts whose content evolves over time to meet the needs and convenience of current generations of legal thinkers. Consequently, it is their content which changes, not the existence of the concepts themselves. However, this view treats concepts like ‘trade’ and ‘non-trade’ as ‘plastic devices whose content changes to accommodate our immediate needs and convenience’.Footnote 35 Working with these categories in order to define the problem of international agricultural trade appears then to be only a question of adding, or subtracting, issues from the trade category and placing them into the non-trade one (or vice versa) to reflect what is convenient at any specific time.Footnote 36 Yet the significance of defining the problem of international agricultural trade using such accepted concepts and categories is more profound than this. The patterns we discern to form these concepts and categories also reflect assumptions of coherence: that is, the way the patterns contribute to particular policy goals, how they combine with existing principles or contradict them. To achieve this, each person engaging with the problem of international agricultural trade must necessarily work within a specific legal theory. This theory should be seen as being a broad account of how goals and principles are implied from the structure and content of the legal order, rather than some higher legal theory. Accepted concepts and categories are therefore not ‘neutral boxes into which the law can be slotted’,Footnote 37 but are instead the product of particular interpretations of the international agricultural trading regime's underlying coherence. The dispute between commentators and negotiators on the problem of international agricultural trade is not about how the notions of trade and non-trade concerns (for example) should be constructed to best serve the needs of the individual negotiators/commentators, but rather how the foundations of the international agricultural trading legal order should be philosophically understood. It is then each commentator's and negotiator's view of this underlying legal order which drives how and what they place within the different concepts and categories. Their idea of the international agricultural trading legal order should not be seen as something which is fixed, but instead as something which evolves and changes over time as circumstances external and internal to the commentators/negotiators shape their understanding and views of it. These ideas are best illustrated with an example.
Existing commentaries examining the problem of the relationship between trade and non-trade concerns in international agricultural trade focus on difficulties arising from the conflict, congruence or disassociation of trade rules with non-trade issues either in the context of the rules in their own right, within a discrete subject area, or as indicative representations of the trade regime's jurisdiction.Footnote 38 The debate's focus is firmly on how to interpret/construe trade rules within the Agreement on Agriculture and the WTO generally, rather than a general philosophical debate about the nature of trade dissociated from any regulatory system. These important contributions offer incisive insights into where the division lies between the legitimate use of a trade measure to achieve a non-trade purpose and a non-legitimate use, and when outside non-trade ideas and influences should be allowed to permeate international agricultural trade jurisprudence. Failure to stay within these clearly defined limits would be regarded as a problem by each of these commentators.
Notwithstanding their erudition, these contributions are based on a specific theoretical approach taken by the commentators, which represents their conception of what the Agreement on Agriculture's underlying aims and objectives are.Footnote 39 In this sense, the commentators’ theoretical approach informs their choice of issues as trade or non-trade, which enables them to conclude whether the issue falls within the agricultural trade rules/regime or not. In this way, each person is able to build up a coherent concept of what they understand as trade and as non-trade. This in turn then drives how that person understands the problem of international agricultural trade. For example, a commentator may believe that the Agreement on Agriculture's rules are designed to maximise Members’ opportunities to generate income from their ability to transfer agricultural products across the world free from trade barriers: the rules are therefore the mechanism for enabling the reduction of those barriers to maximise unrestricted market activity which in turn automatically facilitates other non-trade goals like the preservation of indigenous peoples who derive higher incomes from more access to export markets for their agricultural products. This may be understood as a free-market approach. Another commentator instead sees the Agreement on Agriculture's rules as achieving moral or social goals, labour standards, development and the protection of the environment: the Agreement on Agriculture's rules in this view are designed to facilitate these moral objectives through the reduction of trade barriers. In this second instance, achieving the moral or societal objective is more important, and any market liberalisation is only essential to the degree that it facilitates the moral or societal goal: this might be labelled a moral or societal approach.
In both approaches, welfare maximisation is the ultimate goal, but to the commentators understanding the problem through the lens of the free-market approach, there is a consequential relationship between market liberalisation and the non-trade goal because it is only through the re-establishment of the pure market mechanism that the non-trade goal can be achieved; a measure only fulfils a trade goal to the extent that it liberates the market from its inability to function unimpeded, and obligations in another international agreement will only be part of/connected to the trade regime to the extent that it too facilitates market liberalisation. In contrast, for the commentator who feels the moral or societal interpretation of the Agreement on Agriculture is correct, global welfare is achieved through the attainment of the moral or societal goal, irrespective of whether all distortions to the market are removed. A measure is therefore an appropriate trade measure even if it is not the most economically efficient in terms of market liberalisation, but rather because it is the most effective way of achieving the moral/societal goal in the specific set of circumstances discussed by the commentator. Likewise, rules in another international agreement could be part of, or connected to, the trade regime if they too fulfil the moral and societal goals which the commentator thinks are the fundamental aims/objectives of the Agreement on Agriculture. The distinction between the two views may lead both commentators to the same conclusion on any specific question, but the methodology by which they reach this conclusion lies at the core of their analysis and is not merely an incidental part of it.Footnote 40
There is therefore no definitive determination for how the concepts ‘trade’ and ‘non-trade’ are understood in the problem of international agricultural trade, because the discussion is not the pursuit of a truth where the content of what is trade can be fully known. Instead, it is a concept which each person imbues with their own ideas about the theoretical foundations of the trade regime. The concept of trade should therefore not be seen as fixed, but as a changing construct which evolves as existing approaches are rejected in favour of new ones.Footnote 41 Each person engaging with the problem of international agricultural trade makes the problem coherent in their own way, so there may be slightly different nuanced ideas between them. This means that agreement on an issue may be much more fragile than it first appears, as the meeting of minds is at best only on the surface. Recognising that both the problem of international agricultural trade and the way each person understands it are in constant flux has important consequences for the stability of the existing Agreement on Agriculture and any new rules which are introduced as a consequence of the Doha Development Round of multilateral trade talks. These ideas also have important implications beyond international agricultural trade regulation.
The implications for the regulation of international trade
International trade negotiations involve many negotiators each trying to accommodate their own politically infused domestic policy agendas and the future of the international trading regulatory system's legitimacy into any final text. Reaching a final agreement is a difficult process and inevitably collapses in the trade talks occur. Agriculture is the most extreme example of this phenomenon, but it is clear from the current impasse on trade in services, for example, that agriculture is not the only trade issue to cause difficulties.Footnote 42 This article has argued that two interlinked issues cause difficulties for the stability of existing international agricultural trade rules and the multilateral negotiations on agriculture: agriculture's nature as a polycentric problem and the way each person subtly interprets that problem for themselves. These ideas and their implications can be extrapolated to include all those areas of trade which seem resistant to adequate regulation.
A polycentric problem is made up of a series of overlapping and converging points of disagreement which make it resemble a spider's web. This interconnection means that every issue is inextricably linked with every other, so one issue cannot be resolved unless all others are simultaneously determined. However, an ultimate resolution is impossible because the outcome of one issue is intimately connected with every other, so that what seems to be a settled issue must be re-opened time and again as each new issue is addressed. The problem therefore moves around continually on its axis, always incapable of final resolution. If a regulatory structure is placed on top of a polycentric problem this only represents a static picture of the problem. The structure does not resolve the problem but rather becomes part of it because the snapshot is static, whereas the problem is in a constant state of flux. For example, at the heart of the trade/non-trade concerns debate is the question: When is it legitimate to use a trade measure for a non-trade purpose? This question does not change merely because the Annex II Agreement on Agriculture, the Agreement on Sanitary and Phytosanitary Measures, or the general exemption clause in Article XX of the GATT, etc. have allegedly addressed this,Footnote 43 as can be seen from the litigation of this issue in cases like US-Shrimp, Brazil Tyres and EC-GMO. Footnote 44 For the rules become as much part of the problem as the overall issue the rules tried to resolve. In reality, the regulatory framework only masks the ambiguity and disagreement because the imposition of the regulatory framework is not changing the nature of the problem. It is in fact disguising it.
Moreover, just as the problem is constantly in motion, the starting point from which each person addresses the problem is also in a perpetual state of flux. This makes multilateral trade talks more difficult than people already imagine them to be. This is because any discussion of a problem vis-à-vis other disputants depends upon the identification of common ground as a beginning place from which to argue. But where a problem is characterised by the presence of multiple disputants this common ground is not static across the numerous instances of argument that must take place. Rather, each instance of argument will have an influence over the conceptual starting point taken.
Suggesting that agreement can be achieved suggests an end point, but a polycentric problem has no beginning or end. People therefore appear to agree on a single series of rules, but this may hide the fact that beneath this superficial level their views may be diametrically opposed. When different people operate from different starting points, they can collide in their judgments, or converge in judgment for very different reasons. Where there is collision, typically, the result is a third outcome; but where there is convergence without genuine agreement on the meaning, the effect is that the parties are talking past each other, so no specific resolution is achieved.Footnote 45 Consequently, the problem remains unresolved and may therefore suddenly re-emerge in any context to frustrate the parties’ aims. Much of the frustration felt in multilateral trade negotiations in agriculture and other sectors results from the parties’ belief that they are tantalisingly close to agreement, when in fact this discussion showed that the level of agreement is much lower than it appears.
In a speech to the United Nations Economic and Social Council in 2007, Pascal Lamy, the Director General of the WTO, stated that only small concessions were required to bring the Doha Round of multilateral trade negotiations to a successful conclusion. He argued that the United States need only offer a cut in agricultural subsidies equivalent to ‘a week's worth of transatlantic trade’ and a ‘handful’ of percentage cuts would then be needed from the European Union, Japan, Brazil and India.Footnote 46 This suggests that at that stage agreement was close and involved only a small and arguably insignificant question of adjusting the mathematics. Despite this optimism, the trade talks dragged on into 2008 and eventually collapsed again in July 2008, even though all aspects were allegedly settled apart from the Special Safeguard Clause for developing nations.Footnote 47 This collapse supports the contention made here that the agreement was fragile for more complex reasons than the negotiators and commentators believed: the tentative significant consensus burst apart as if it had never existed. Attempts made in late 2008 and early 2009 to restart the talks show negotiations starting on many issues that were allegedly settled in early 2008.Footnote 48
Agreement occurs where everyone is moving in the same direction on every level. In contrast, disagreement is not a single idea, but involves a multiplicity of directions. Disagreement then entails a much more complex pattern in which there is constant crossing and recrossing of paths by parties whose ideas are structured by different starting points. Arguments are not without consequences. Any attempt to press the argument again, after it has been questioned or rejected, will not simply duplicate the effects on the whole web as previously, because later arguments will in fact have broken strands of the web along which the original argument travelled. Instead, pressing the original argument creates new patterns of tension across the web. This means the problems in international trade are not capable of final resolution, but remain instead permanent problems within the WTO regulatory scheme.
Thus, the appropriate strategy is one of management, rather than final resolution. It is the failure to recognise this step change in addressing the issues which causes frustration. It is therefore critical that we be on constant guard against approaching a polycentric problem with the lawyerly attitude that we should be able to find a final solution. There are two interlinked dangers if we pursue this path: we become disheartened by the complexity of the task when it is viewed in its correct light and, consequently, abandon the task in favour of superficially acceptable solutions which, though they are intended as final resolutions, can only ever be temporary and partial redistributions of the tensions within the web of ongoing disagreement. Incremental management techniques ideally placed to address polycentric problems are already built into the WTO in Article IV of the Marrakesh Agreement Establishing the World Trade Organisation and could be used to ameliorate trade issues as they arise. Small steps must be taken, not giant leaps for mankind.
