22140 * ~~~~~~December 1989 UNDP-WORLD BANK TRADE EXPANSION PROGRAM OCCASIONL PAPER 5 HOW DEVELOPING COUNTRIES MIGHT INFLUENCE THE TALKS ON AGRICULTURE IN THE GATT'S URUGUAY ROUND L. Alan Winters Iti occsonal pape r a pdulct of te joit UNDPMWodd BIdT,Ee ExBpnsn Pngu wtch pves bdnkal and poi ad= b =MtO hkedng I Mfrm f, We rne h s contaed heree tiose of fte autws and do not nesauy fese of te U*d Nalions Dek Powtam or toe Wxd Bark FIllE COP>Y HOW DEVELOPING COUNTRIES MIGHT INFLUENCE THE TALKS ON AGRICULTURE IN THE GATTS URUGUAY ROUND L. Alan Winters December 1989 Trade Policy Division The World Bank Washington, DC HOW DEVELOPING COUNTRIES MIGHT INFLUENCE THE TALKS ON AGRICULTURE IN THE GATTS URUGUAY ROUND L. Alan Winters December 1989 Trade Policy Division The World Bank Washington, DC Table of Contents Page Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Industrial Country Objectives . . . . . . . . . . . . . . . . . . . . . 1 So-Called Noneconomic Objectives . . . . . . . . . . . . . . . 1 The Budget Constraint and Political Acceptance . . . . . . . . 2 Developing Country Interests . . . . . . . . . . . . . . . . . . . . . . 4 Liberalization and World Price Levels . . . . . . . . . . . . . 4 Stability and Risk . . . . . . . . . . . . . . . . . . . . . . 7 The Background to the Geneva Accord . . . . . . . . . . . . . . . . . . 11 The Cairns Group . . . . . . . . . . . . . . . . . . . . . . 11 Preparations and Breakdown in Discussions . . . . . . . . . . 12 The Final Agreement . . . . . . . . . . . . . . . . . . . . . 15 The Next Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Tariffication . . . . . . . . . . . . . . . . . . . . . . . . 19 Liberalization of Other Elements of Agricultural Policy . . . 23 Pitfalls in Negotiations . . . . . . . . . . . . . . . . . . . . . . . 28 Exchange Rates . . . . . . . . . . . . . . . . . . . . . . . 29 Preferential Trading Arrangements . . . . . . . . . . . . . . 31 Special and Differential Treatment . . . . . . . . . . . . . 33 Developing Country Food Imports . . . . . . . . . . . . . . . 34 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 SuaMmnAry This paper considers how developing countries might attempt to influence the next stage of negotiations on agriculture in the Uruguay Round. It sketches the industrial countries' objectives and constraints in agricultural policy and the principal adverse effects of these policies on developing countries. It shows that developing countries have much to gain from the process of world agricultural liberalization and that the threat to their import capacity is less than is often supposed. The paper describes the events surrounding the breakdown of the mid-term talks in Montreal and the emergence of the subsequent Geneva Accord on agriculture. The Geneva Accord leaves a little scope for developing countries to influence the final outcome of the talks; the paper shows how they may exploit this to avoid market-sharing arrangements, which will inevitably be to the developing countries' detriment. The paper also lists a few of the possible pitfalls in negotiation and suggests how developing country negotiators should react to them. These include the problem of exchange rate variability for defining the level of agricultural support; preferential trading arrangements, which offer fewer overall benefits than liberalization; special and differential treatment for developing countries, which disempowers them in negotiations and reduces the concessions offered to them; and the issue of developing country food imports and the importance of establishing transitional aid arrangements. Several conclusions of the paper are particularly relevant for developing country negotiators and should be considered by them in formulating their positions over the coming critical year: -vi- * Developing countries will reap few benefits from a "quantitative solution" to world agricultural problems. Quantity controls would raise prices, offer little scope for increased exports, and exacerbate market instability. * Tariffication will address countries' interests in stabilizing world prices. * Aggregate measures of support will not define liberalization but may help to measure it; among such aggregate measures, the producer subsidy equivalent is preferable to the support measurement unit. * Developing countries should negotiate fully on agriculture as on other issues. Special and differential treatment promises very few direct economic benefits and amounts to little more than nonparticipation in negotiation. S World food prices may rise slightly with liberalization--it is difficult to tell by how much--but the correct response to any resulting hardship is aid and adjustment not the blocking of liberalization. -1- The interests of developing countries as well as industrial countries lies in the liberalization of world agricultural trade. Thus the current arrays of protectionist measures in industrial countries, with their obvious consequences for world agricultural markets, has serious repercussions for developing countries. The Uruguay Round of multilateral trade negotiations within the General Agreement on Tariffs and Trade (GATT) may be able to reverse the tide of protectionism and alleviate some of the harm it has done. But not every ostensibly successful conclusion to the Round will be beneficial in these respects, for how the industrial countries address their own problems will be important for developing countries. This paper identifies ways in which developing countries might influence the outcome of the GATT trade talks in agriculture and move them in the direction of mutual advantage. Industrial Country Objectives So-Called Noneconomic Objectives Industrial countries claim a wide range of objectives for their agricultural policies. The principal one, whether admitted explicitly or not, is simply to raise incomes in the farm sector, but this is buttressed I am grateful to the following people for comments on an earlier draft of this paper: J. Michael Finger, Sir Michael Franklin, Richard Howarth, Alan Matthews, Jim de Melo, Paul Meo, Costas Michalopoulos, Stefan Tangerman, Ken Thomson, and the participants in a seminar at the World Bank. I am also most grateful to Sue James and Maria D. Ameal for typing. More strongly than usual, I must dissociate these people from the paper's views and remaining short-comings. The views expressed are the author's alone and should not be attributed to the World Bank. -2- by a whole series of what I have termed SNOs--so-called noneconomic objectives (Winters 1988).1 These include shifting the distribution of income toward farmers, preserving the environment, maintaining viable and vital rural communities, and guaranteeing secure and stable food supplies. These are legitimate objectives for governments, and they also figure prominently among the aims of developing countries. But these objectives are neither truly noneconomic nor are they greatly enhanced by current agricultural policy. They are important, however, for two reasons. First, they describe, at least to some extent, the genuine aspirations of, and constraints on, industrial country negotiators. Second, they are selectively appealed to by negotiators in objecting to various reform proposals, and thus they define issues for which developing country negotiators should have answers (see Winters 1987b and 1988 for some answers). The Budget Constraint and Political Acceptance In considering industrial countries' policies, however, a genuine constraint must be added to the list of SNOs: the public budget. Budgetary pressure is the most obvious and, to date, most potent source of pressure for agricultural reform. Farm policy nearly bankrupted the European Economic Community (EEC) in the mid-1980s. But as soon as the pressure was relieved in 1988-89 by a combination of small reductions in support and significant increases in world prices, the Commissioner for 1. While I refer to these factors as objectives, some commentators prefer to think of them as constraints. As long as the constraints bind but are not absolutely rigid, the distinction hardly matters for my arguments here. -3- Agriculture started talking again about increased farm aid. Similar pressures have been felt in Japan (over the rice program in the 1970s, for example) and in the United States and the European Free Trade Association (EFTA) countries. The budget is among the most transparent, and hence controversial, aspects of farm support, and for politicians and bureaucrats the trade-off between budgetary savings and economic efficiency is almost always settled in favor of the budget. Although they have eased recently, budgetary pressures are still likely to play an important role in the completion of the Uruguay Round talks. For example, the EEC's Common Agricultural Policy has received very bad press because of its voracious appetite for public money. It not only requires substantial taxation to maintain, but it also effectively blocks EEC initiatives in other dimensions because the EEC authorities are tightly constrained in the total resources made available by member states. Thus nearly every EEC policy statement promises/demands that agriculture's share of expenditure be scaled down in favor of regional and structural funds-- that is, general industrial subsidies. As the process of completing the internal market approaches its climax, other claims on EEC funds will become stronger (for example, for research and development, training, and infrastructure), and a number of political compromises requiring public subsidization will probably be necessary. Taming the agricultural monster will not only win friends directly but will also allow the EEC Commission more latitude elsewhere. The budgetary dimension has also been important in the United States and will become increasingly so. The U.S. government budget deficit is the largest single macroeconomic malaise in the United States (and the -4- world?), and yet the Administration has promised no new taxes while regularly finding, along with Congress, new and deserving causes requiring additional expenditures. As the need for macroeconomic action increases, so will the political pressures to shuffle more of the costs of farm support on to consumers at home and abroad.2 Full liberalization of agricultural trade is not a feasible outcome of the Round, and budgetary considerations will make it more difficult to ensure that the farm policies that remain are economically efficient. An additional tactical objective in industrial countries is to make agricultural support politically acceptable. High implicit taxes on consumers are only occasionally politically embarrassing, but stories of fraud, rotting stockpiles, and huge payments to rich farmers appear to worry the politicians much more. The propensity of the Common Agricultural Policy to generate surpluses is thus an important issue bureaucratically; it figures strongly in the minds of EEC negotiators and pushes them further toward solutions based on managed supply. Developing Country Interests Liberalization and World Price Levels The principal effect of industrial country protectionism is to reduce the world prices of most agricultural commodities. Precise figures 2. That the burden has not yet been fully passed on reflects the fact that until the choice is between taxes rather than between taxing or not taxing, politicians pay regard to the political and economic costs of imposing taxes, even implicit ones. That is, "burden shifting" of this kind is more likely a means of avoiding expenditure cuts than a means of directly financing expenditure increases. -5- are impossible to calculate, and the range of estimates is very large. At one extreme, Tyers and Anderson (1988) suggest that agricultural policy liberalization by Organization for Economic Cooperation and Development (OECD) countries would raise wheat prices by about 25 percent, beef by 50 percent, and dairy products by 100 percent in 1995. At the other extreme are commentators who argue that liberalization could even lower some world prices. The most likely outcome is that liberalization will induce mild price increases. This would clearly have economic implications for developing countries, although not all effects would be in the same direction. Exporters of the temperate products whose prices would be most affected by industrial country liberalization (for example, Argentina and Thailand) have strong and direct interests in the dismantling of protection; their revenues and income would increase significantly. In contrast, chronic food importers (for example, Bangladesh) would undoubtedly suffer. A third group of developing countries falls into an intermediate category: at the current low prices for temperate food they are net importers, but they would be self-sufficient or net exporters at free trade prices.3 Some conmmentators argue that developing countries as a group would gain from OECD liberalization (see, for example, Loo and Tower 1988), while others are more pessimistic (see, for example, Matthews 1985, Parikh et al. 3. It must be admitted that it is even more difficult to predict trade reversals than other effects of liberalization. Reversal depends not only on a developing country's own policies but also on the wedges between import and export prices, which, in turn, depend on trading costs, the discounts necessary to break into new markets, and similar factors. Overall, however, at least some trade reversal looks very likely. -6- 1987, and Tyers and Anderson 1988). The most pessimistic estimate is that of Tyers and Anderson who, allowing time for liberalization to occur and extrapolating current trends until then, estimate that the effect of liberalization for temperate crops would be a net welfare loss of US$14 billion at 1995 quantities and prices for developing countries as a whole. There are several reasons for questioning the basis for this analysis now, and even more for questioning its relevance to the 1990s and 2000s. First, the results are based on historical responses to changes in prices that were estimated on data from an era when industrial country policy was increasingly interventionist and developing country governments penalized and discouraged agriculture. If industrial country liberalization were genuine and bound under the GATT, developing country production responses might be stronger than past evidence suggests. For example, Maddala (forthcoming) has shown that supply elasticities estimated without taking into consideration the disequilibrium effects of price supports or buffer stocks can be seriously biased downward. If, in addition, developing countries liberalized their own agricultural regimes, which have often taxed farmers, supply responses would be even stronger. For several reasons, however, the direct supply response of developing country farmers to liberalization-induced price rises may be exaggerated. Among these are developing countries' general refusal in the past to allow their agricultural policies to be strongly influenced by external factors, the likelihood that price changes deriving from liberalization will be smaller than the secular declines that occurred during the 1980s, and the fact that a number of developing country import- competing sectors are protected rather than taxed (Krueger, Schiff, and -7- Valdes 1988). Thus, while increased agricultural supply may be anticipated from developing countries, it will occur less through the direct channels of price responses than through certain indirect channels, such as research and development. This brings up a second reason for expecting stronger liberalization effects on output than Tyers and Anderson allow. Greater price stability or higher price levels plus a more secure atmosphere and policy environment for developing country agriculture would encourage infrastructural investment and research and development, none of which figure in the supply functions of existing models. A third reason is that Tyers and Anderson deal only with seven temperate product groups. Price rises in these would, at least to some extent, spill over into tropical product markets in which developing countries are already net exporters. Moreover, if agricultural liberalization also included tropical products directly, its effects on developing country export earnings and income could be redoubled, as Valdez and Zeitz (1980) have shown. Stability and Risk A critical dimension of the analysis of the effect of industrial country liberalization on developing countries concerns stability. Most industrial country policies are designed to insulate domestic consumers and producers from changes in world prices. The effect, however, is to increase world price instability. Because fewer countries adjust their supply and demand to any particular quantity shock, the effects of the shock on world prices is increased. Commentators all agree that industrial country liberalization will reduce world price instability. Tyers and Anderson suggest it might do so by a factor of two for wheat and dairy -8- products and four for beef. Model-builders do not, however, allow instability to feed back into production decisions. In part this is justifiable because farmers in both the developed and developing world are insulated from most instability. But insulation shifts the burden of risk onto government, and in developing countries such burdens are large relative to overall public sector resources. For the governments of agricultural exporters and potential exporters, this assumption of increased risk has several repercussions. First, long-term insulation is much less feasible in developing than in industrial countries. Second, agricultural shocks tend to destroy macroeconomic equilibrium when insulation breaks down and are then shifted to other sectors (especially to capital transactions). And third, fearing such shocks, developing country governments tend to undervalue agriculture. This last point represents perhaps the most pernicious effect of industrial country policies on developing countries. Industrial country interventions distort world markets and encourage developing country governments (or in their view, oblige them) to intervene in their economies beyond desirable levels and far beyond their ability to do so effectively. The results are the promotion of industry above agriculture; the consequent distortion of normal patterns of savings, investment, and development; and the extension of bureaucracy, inefficiency and, frequently, corruption. If industrial countries were to stabilize world markets by trading openly, developing countries would have much less to fear in liberalizing their own agricultural sectors. This, in turn, would curtail the pressures to promote industry artificially and, as a result, might reduce some of the competitive pressure on industrial country manufacturers. -9- For food importers, the effect of world price instability is more complex than it is for exporters. The desire for insulation may well lead countries to seek a higher degree of food self-sufficiency than otherwise. If this is achieved by subsidizing agriculture or by taxing it less than other sectors, and if the subsidy or tax rate is maintained in the face of world liberalization, locally produced quantities may increase with industrial country liberalization, but this effect may not be welfare enhancing. If, however, importing governments respond to the improved world price stability by reducing their net support for agriculture, then local output may fall, but welfare is likely to increase. Krueger, Schiff, and Valdes (1988) found a surprising amount of support for import-competing agriculture in developing countries, so developing country liberalization may tend to offset some of the output-enhancing effects of industrial country liberalization. This may, although it need not, increase developing countries' overall import dependence. Another problem for developing countries arising from industrial country agricultural intervention is that it replaces risk by uncertainty. Agricultural supply and demand are inherently variable, and while free world trade cushions and spreads the risks, the risks are nonetheless still there. However, these are risks in Knight's sense: they are amenable to probabilistic quantification and thus are at least partly addressable by contingent decisionmaking and partly ameliorated through the law of large numbers. By contrast, intervention in its current form gives rise essentially to Knightian uncertainty. For example, the trade policies that the U.S. Congress or a budget-stressed Administration may dream up--and their timing--are not predictable. Moreover, policy decisions tend to be -10- long-lived rather than serially uncorrelated like weather shocks. Thus, having politics rather than commerce in the driving seat changes the nature of the environment, making it much more dangerous for developing countries to rely on exports of agricultural products. Related to this issue is the general inability of developing countries to influence the critical political decisions affecting their trade. This is more the case in agriculture than in manufacturing, for agriculture has always lain farther beyond the grasp of the GATT while also being safeguarded by remarkably powerful domestic lobbies in industrial countries. Thus as long as intervention is the rule, developing country producers can expect to do badly as a result of it. The evolution of U.S. sugar quotas or voluntary restraint agreements on EEC imports of feedgrain substitutes is sufficient to prove the point. The conclusion of this analysis is that developing countries have differing interests in the reform of industrial country agricultural policies. While a movement toward free trade and the imposition of commercial discipline in the making of agricultural policy would cause some developing countries to suffer from a terms of trade deterioration, others would benefit. Liberalization will improve the food trade balance for some and worsen it for others, but there is no simple link between the quantities of trade and economic welfare: much depends on how the changes in quantities are brought about. As I argue below, focusing on the amount and direction of trade as criteria of success--as negotiators are prone to do--would be very misleading for the longer run. The true requirement is not for market discipline defined by increases or reductions in trade but for genuine liberalization defined as the reduction of distortions. Indeed, despite possibly differing interests in liberalization, developing -11- countries should nearly all be able to agree to resist a "quantity-based solution" to the agricultural trade problem. The Background to the Geneva Accord The Cairns Group During preparations for the Uruguay Round, agriculture figured low on the agenda of the EEC, EFTA, and Japan. Although the EEC had some minor objectives in the agricultural talks, it would probably have been happy to have seen the sector ignored altogether. When that proved impossible, the EEC sought to scale down expectations about what could be achieved. Thus the EEC maintained--and still maintains--that the basic structure of the Common Agricultural Policy is nonnegotiable. The pressure to include meaningful talks on agriculture was of two sorts. The budgetary strain and popular derision of the Common Agricultural Policy forced the EEC's hand internally, while the United States and the Cairns Group exerted effective external pressure. The formation of the Cairns Group, a coalition of nonsubsidizing exporters,4 has been one of the most interesting and innovative features of the Uruguay Round of negotiations relative to previous rounds. The group has pursued the single issue of agricultural reform in a clear but politically realistic fashion. By building bridges between the two protagonists, it has kept the issue alive and, at least to some extent, has 4. Argentina, Australia, Brazil, Canada, Chile, Colombia, Fiji, Hungary, Indonesia, Malaysia, Philippines, New Zealand, Thailand, and Uruguay. -12- kept on the table the broader economic issue of liberalization rather than market-sharing. Although it is widely recognized that any settlement in agriculture must meet the requirements of the EEC and the United States, other parties are also involved. It is not difficult to envisage some advantage accruing to a third party--the emergence of the Swiss formula in the Tokyo Round showed clearly the possible role of an intermediary. While the presence of Canada and Australia increases the Cairns Group's political power, the group's mere existence provides a useful reminder that, in major set-piece negotiations, developing countries are not altogether impotent. That said, however, developing countries and the Cairns Group are going to have to make concessions if they are to cash in their influence in any concrete way, and they are going to have to maintain a fairly unified stance. Preparations and Breakdown in Discussions The Punta del Este Declaration of December 1986 dealt with agriculture very fulsomely. It talked of "correcting and preventing distortions," "greater liberalization ... of import access and export competition," and "strengthened GATT rules and disciplines." While the declaration promised a far greater degree of commitment than any previous round had made, it was sufficiently imprecise to permit widely varying interpretations of the terms of that commitment and of the solutions to be considered. Much of 1987 and 1988 was taken up with the preparation of basic positions and the analysis of technical issues. Among technical issues, perhaps the most important was the advocacy of producer subsidy equivalents -13- as an aggregate measure of, or even as a target for, the degree of protection. The producer subsidy equivalent attempts to calculate the extent to which the full range of protective policies increases producer incomes. The measure was originated by Josling (1973) and has been calculated by OECD (OECD 1987) and other official bodies and championed as a negotiating tool by Tangermann, Josling, and Pearson (1987). Tangermann, Josling, and Pearson suggest that producer subsidy equivalents could be both negotiated and bound. They argue that bound producer subsidy equivalents would form a basis for an orderly reduction of protection that could not be undermined by new policies and that would not involve explicit trade-offs between the very disparate forms of intervention used by different countries. Negotiations are not yet at the stage of quantitative offers and requests, but it looks increasingly unlikely that producer subsidy equivalents will be anything more than an information measure, if that. Their inability to separate the extent of trade distortion from the degree of income support and their inevitable fluctuation when fixed internal prices are compared with variable world prices have been the principal stumbling blocks. The EEC has been particularly anxious to obtain-"credit" for its recent quantity controls which, while boosting farm incomes, are claimed also to reduce the disruption of world markets.5 For this reason, the EEC invented its own measure, the support measurement unit, which, among other things, allows credit for quantity controls. 5. This mercantilist, or quantity-based, conception of trade problems besets the whole of the EEC's agricultural policy and has colored the Uruguay Round significantly to date. -14- Despite a risk of caricature, one can summarize the three major parties' basic objectives in their initial statements as follows: EEC - a completely regulated system of agricultural trade United States - a complete abolition of import restrictions and subsidies affecting trade (the "zero" option) Cairns Group - a reduction of supports affecting trade, but with remaining discretion over (some) domestic policies.6 The Japanese and smaller European countries were also reluctant to liberalize--ostensibly because of their SNOS--and certain developing countries (including some Cairns Group members) also put forward positions of varying liberality. During 1988, the United States and the EEC showed little sign of compromise and took part in an acrimonious debate, calling each other's proposals "disappointing" and "unrealistic.' Fundamentally, the question was about whether the basic structure of the Common Agriculture Policy was negotiable. The EEC had maintained ever since the Kennedy Round that it was not, while the United States has insisted that it should be and began to argue that without genuine agricultural reform (defined by the United States as the "zero" option), the whole Round would fail. 6. The positions, but not their attribution, are based on de Zeeuw (1989). -15- With such gaps to span, it is hardly surprising that the Montreal mid-term talks of December 1988 foundered over agriculture. However, within days of the breakdown, the signs of a possible reconciliation were evident as the breakdown was presented as a means of increasing the time for consultations. Contingent progress in other areas was recognized as such, and bilateral U.S.-EEC meetings occurred once the two administrations had changed. At a more fundamental level, it became clear that the U.S. Congress was not wholly behind the 'zero" option. This eased the passage of negotiation toward the Geneva Accord, although in the long run the active interest of the U.S. Congress in negotiations is hardly a recipe for agricultural liberalization. The Final Agreement By March 1989, the United States had somewhat moderated its position and was talking of a "ratcheting down' of farm policy rather than abolition, but the EEC was still inflexible. Indeed the EEC argued that far from being expected to offer further reform of the Common Agricultural Policy it should receive credit for modifications made since 1984; that is, other parties should reciprocate over and above any concessions made to win further EEC reforms. The Cairns Group, however, threatened to abandon the Uruguay Round if the United States and EEC could not reach agreement over a text on agriculture. This was a more realistic threat than the earlier U.S. insistence that the zero option was the sine qua non of the Round because agriculture figures more prominently in the Cairn Group's objective function than that of the United States. Thus the threat probably helped to create an atmosphere in which settlement was possible. This influence -16- is another illustration of the role that can be played by an intermediary committed to obtaining a solution. At this point in the negotiations, the Secretary General of the GATT, Arthur Dunkel, presented his first compromise document. While this document was not accepted, it did provide a focus for the negotiations. However, the negotiations mostly had the effect of weakening the document's reform content by such means as the introduction of a studious imprecision about base-years from which to measure reforms. The final agreement emerged a week later with the following principal long-term conditions: * "progressive substantial reductions' in agricultural support, which was defined as all measures bearing directly and indirectly on import access and export competitiveness, and expressly including quantitative restrictions, tariffs, internal and export subsidies, and export restrictions; * special treatment for developing countries, including 'taking into account' any negative effects of reform on food importers; * attention to nontrade objectives (basically SNOs); * establishment of a timetable in which reform plans were to be presented by the end of 1989 and the first step of reform was to be executed in 1991 (plans were to cover the application of aggregate measures of support, the application of GATT rules, and tariffication and decoupling, among other issues); and * multilateral surveillance of the reform program. -17- The short-term dimensions were considered to have been the more contentious to negotiate, but once having been agreed on, they have largely dropped from sight. The short-term conditions include the following: * a freeze on price support levels and import restrictions, but not on any changes already negotiated (price support levels were to be defined in national currency--or European Currency Units [ECU] in the EEC--with a 1989 base; access was to be assessed on the average for 1987 and 1988, with reductions defined either commodity by commodity or on an aggregate measure of support basis); * a presentation to the Agriculture Negotiation Group by October 1989 of plans for reductions in support and improvements in access, with six monthly reports on progress to be presented starting in December 1989; - exemptions for developing countries; and * a waiver from the provisions under "exceptional circumstances." The short-term conditions did not require much adjustment on the part of the major parties. Using the current year as the base from which to measure support "validates' the recent relaxation of set asides by the United States and allows it to claim credit for a future tightening of limits to their former levels. While export subsidies figure in the text of the accord, no mechanism is recorded for measuring or freezing them. With no targets for short-term deescalation, the EEC will receive negotiating credit for existing plans to reduce excess production of -18- cereals through the production stabilizer, while the calculation of support prices in ECU will disguise the protection inherent in the EEC's green- money system.7 The Cairns Group claimed that the short-term agreement represented a down payment on liberalization, but in truth it was not much more than a means of keeping agricultural brushfires under control while the rest of the Round proceeded. The long-term agreements also fall short of the aspirations voiced at the beginning of the Round. For example, the use of an aggregate measure of support--if agreed on--appears likely to sanction EEC 'rebalancing" such as trading increased protection in oilseeds against reduced protection in cereals. Some of the countries of EFTA publicly registered their misgivings about the accord. The significance of this dissent is not yet clear, but EFTA objections on agriculture might turn out to be important to the United States or the EEC if they need EFTA's support on other issues during the Uruguay Round. EFTA's chief objections are also important given current explorations of further EEC and EFTA integration in the context of 1992 and Common European Economic Space. The outcome of the interim talks appears to be very close to the EEC's original objectives. The basic mechanisms of the Common Agricultural Policy were unchallenged, there was an implicit acceptance of supply management (in that it earns credit), and the EEC had to make only a few short-term concessions beyond the reductions in support mandated by its own 7. Effective EEC wheat prices were the same in 1988 as in 1980 in terms of ECU but rose by 25 percent in terms of local currencies. This arises from a complex interaction between MCAs and green exchange rates (Field, Hearn, and Kirby 1989). -19- budgetary situation. The United States, encouraged by Congress and the farm lobby, effectively surrendered its zero option--while little remained of the Cairns Group proposals for deescalation. Nevertheless, the picture is not entirely bleak for further liberalization. The challenge now is to bolster the U.S. and Cairns Group positions sufficiently to allow them to impose some rationality on the Europeans (EEC and EFTA) and on Japan. Cementing the liberal coalition will require considerable political skill and compromise by the developing countries in meeting the demands not only of international market managers but also of the U.S. farm lobby. The Next Steps The Geneva Accord hardly represents the pinnacle of a liberal trading system or even of short-term agricultural deescalation, but it does contain a few grains of hope. Among these are a commitment to progressive, substantial reductions in the support (including internal support) and protection of agriculture; a commitment to the surveillance of agricultural policy under the GATT; and a promise of continuous debate about how to calculate the level of support. All of these leave a small opportunity for developing countries to influence the final outcome. Tariffication One of the most promising routes to trade liberalization in agriculture is tariffication (switching from nontariff trade barriers to tariffs) followed by negotiated tariff reductions. Tariffication has been suggested by several commentators in the past (see, for example, Zeitz and Valdes 1988 and the Dutch Minister of Agriculture as quoted in Agra Europe -20- of November 18, 1988) but was not seriously considered in the negotiations prior to Geneva. It gets only a subsidiary mention in the Geneva Accord, and many commentators considered it to be a dead issue at that stage. In mid-July, however, the United States revived the issue as a practical approach to the relaxation of import-access restrictions. The United States proposed a system in which a base level of imports would be subject only to bound tariffs while imports above the base level would be subject to additional extraordinary tariffs. Over the course of ten years, the base level of imports would expand and the extraordinary tariff would fall until all trade faced only the bound tariff. The United States suggested that actual 1990 imports and nominal protection coefficients would be suitable definitions, respectively, of base imports and the extraordinary tariff, but these measures are negotiable. The tariffication of barriers to agricultural imports has several advantages: transparency (domestic and international), restored sensitivity of supply and demand to market signals, and simplicity of administration. The first two of these advantages will be difficult for the EEC to swallow given its explicit and implicit objectives, but there may be a large enough constituency among members to overcome these objections. Even simplicity of administration may present an obstacle to reform given that all industrial countries have large agricultural bureaucracies to administer their present detailed market regulations. Tariffication would threaten some bureaucratic jobs directly and would serve as a warning on others, because it will be more difficult to defend detailed internal regulation in the face of a market-driven international trade sector. Moreover, tariffication is mistrusted not only by the bureaucracies themselves but also by farm groups, whose lobbying-power is -21- at least partly related to the size and power of their related bureaucracies. Thus if tariffication is to succeed, it must be pursued at the political level, or at least at the level of "general" ministries such as the ministries of finance and foreign affairs. This, in turn, suggests that tariffication will be most likely to be taken seriously during the later stages of the negotiations, when cross-sectoral deals are being struck and governments are assessing their attitudes to the Uruguay Round as a whole. A fourth advantage frequently claimed for tariffication is that precedents and well-tried procedures exist for the negotiation of tariffs. This is true, but there are no guarantees that the trick that unlocked industrial tariffs in the 1950s and 1960s will work as effectively for agriculture in the 1990s. The key to the GATT's early success was its members' political acceptance of the need for liberalization, coupled with the use of reciprocity to hitch the free-traders' wagon to the mercantilists' tractor. The driving force of real-world trade policy is almost invariably mercantilist--the desire to increase sales.8 By providing a means whereby exporters' aspirations could match or overcome import- competing sectors' apprehensions, the GATT enabled governments to put together sufficient coalitions for liberalization. The situation in agriculture is potentially more difficult than this. The desire for liberalization is not entirely certain, and the scope 8. The only serious exceptions to this arise from countries' concerns for guaranteed supplies of "essentials" such as food and fuel, in which case export prohibitions suddenly become acceptable and trading partners are held to have a duty to maintain unrestricted supplies. -22- for reciprocity within agriculture is severely limited. The benefits of agricultural liberalization are diffuse, general equilibrium effects--known to economists but apparently to no one else; there are virtually no producer groups in Japan or Europe who would gain from agricultural liberalization. Thus there is no well-focused opposition to the farm- lobby's resistance to change. Such opposition will emerge only if agriculture is linked to other issues. While this may be possible, such linkage is not generally seen as part of tariffication. Even if tariffication does not lead to immediate liberalization, however, it would still be desirable for stabilizing world markets by allowing market signals to permeate domestic agricultural decision-making. It would be even more desirable if it were coupled with a ban on export subsidies, as Zeitz and Valdes (1988) suggest, for then trade reversal due to policy intervention would be virtually impossible.9 This analysis has several implications for the way in which developing countries approach tariffication. First, since tariffication is so clearly in their interests--especially with respect to stabilizing world markets--they should promote it and should be willing to undertake tariffication themselves. Second, the principle is more important than the numbers involved. It would be unwise to allow negotiations to become stuck on precise numbers before the principle is agreed to. Third, since tariffication is already on the agenda, work should begin on calculating suitable levels for base imports and extraordinary tariff rates and for the 9. If trade in intermediate goods is heavily distorted, trade reversal is possible under a regime in which tariffs are the only policy instruments, but it does not seem particularly likely. -23- rate of relaxation, so that the proposal is not stymied by the apparent practical difficulties of implementation.10 Finally, developing countries must explore linkages between tariffication and nonagricultural issues in order to be able to take advantage of late-stage cross-issue negotiations at senior ministerial levels. Liberalization of Other Elements of Agricultural Policy The United Stages has now also made outline proposals for liberalizing other elements of agricultural policy. It proposes a complete ban on export subsidies and export restrictions and a threefold classification of internal measures. The first group of measures comprises production and price-based policies (including some marketing policies) that should be phased out over ten years. The second group comprises "permitted" policies that include bona fide food aid, income support, environmental policies, and similar measures. The third, intermediate group of policies is defined as a residual and would include policies such as certain uniform input subsidies that have some, but not direct or strong, impacts on international trade. The United States proposes that these policies be subject to "discipline" through the binding and negotiated reduction of an aggregate measure of support. Given that this group excludes the measures that cause the most difficulty in defining an aggregate measure of support, such as quotas, set-asides, and environmental programs, settling on an appropriate measure for this limited purpose should not be too difficult. 10. This point is relevant given the likely bureaucratic resistance to tariffication. -24- In one respect, the U.S. proposal for this intermediate category of measures follows traditional GATT practice. Having defined the sets of measures that are either definitely legal or definitely illegal, the proposal requires that remaining policies be subject to discipline "to prevent their use in ways that would nullify or impair concessions" (the wording is that of Article XXIII of the GATT). In the absence of a satisfactory list of transgressions, the GATT is enforced (1) by the explicit recognition of each contracting party's right to the benefits stemming from implementation of the GATT by other contracting parties and (2) by the right of any party to complain against any act or situation that interferes with the delivery of those benefits, regardless of whether that act or situation figures in GATT regulations. Thus it is the effect of the GATT not the procedures of trade policy that are protected. This formula postpones the full elaboration of GATT responsibilities until concrete cases arise, and in so doing helps the process of negotiation and initial implementation. It probably offers the best hope of alleviating the suspicions surrounding industrial countries' agricultural policy (and its inherent ambiguity) sufficiently to allow an agreement to be struck. Moreover, providing that not too many policies fall into the intermediate category and that negotiations on GATT discipline allow confidence in future rulings, the use of this traditional approach may well represent a major step toward liberalization. The U.S. proposal will clearly not be immediately acceptable to the EEC, but it may be broad enough to allow negotiation. If the EEC does not reject the suggestions out of hand--which would risk exposing itself to opprobrium for sabotaging the Round--it may seek to shift measures from the "phasing-out" to the "discipline" group and from there to the "legitimate" -25- group. Such a strategy has clear political attractions: the United States "wins" the procedural points but concedes enough substantive points to the EEC to allow the EEC to maintain its honor. If the process gets out of hand, however, it has the potential of legitimizing substantial agricultural protection for the next several decades. Thus developing countries should seek to prevent too much redefinition of policies and to ensure that a market-based rather than a quantity-based outcome is achieved. Aggregate Measures of Support and Quantitative Restrictions The EEC foresees a more prominent role for aggregate measures of support than does the United States, proposing that they be used to provide a central means of disciplining agricultural policy. The EEC has developed its own aggregate measure of support, the support measurement unit, through which it seeks to define a measure that (arguably) more closely reflects a policy's impacts on the world market (the discipline motive) and is more amenable to direct policy control (the management-of-markets objective). The EEC's support measurement unit differs from the producer subsidy equivalent in making allowance for quantity-reducing measures such as quotas and for fluctuations in exchange rates and world prices. If aggregate measures of support are to be an important part of the final outcome, as opposed to a measure of residual policy effects as in the U.S. proposal, it is important that producer subsidy equivalents be favored over support measurement units. If the EEC is allowed to apply support measurement units, it will receive credit for existing and future quantity controls, which can only encourage the sort of implicitly market- sharing arrangements that the EEC so earnestly desires. Such market- sharing arrangements will almost inevitably favor the major parties. It -26- will prevent developing countries from expanding their market shares in industrial country markets, for as they come forward to do so individually they will be picked off one by one. At the same time, such individual quotas may well remain subject to unilateral or bilateral (U.S./EEC) adjustment, just as sugar and butter quotas are dealt with at present. Even putting aside their political weakness, developing countries would still suffer from a regime of quantitative restrictions because it would almost inevitably worsen market instability. Along with market sharing would come fixed internal prices in the industrial world, rigid patterns of production, and, most probably, rigidities in world trade patterns. These factors would concentrate the effects of natural supply shocks in residual world markets on which developing countries will find themselves dependent for at least their marginal trade. Finally, quantity constraints are designed to raise prices without allowing the quantities supplied to follow suit. If developing country net exporters ever become part of a "market stabilization" scheme, they will find themselves prevented from increasing their sales while net importers will still face higher prices than presently. Thus developing countries, no matter what their differences, should all be able to agree to resist quantitative market-sharing arrangements. Certain developing countries have expressed reservations about the producer subsidy equivalent measure, fearing both its complexity and its inability to take account of their SNOs (Runge and Stanton 1988). However, such objections would apply to any mechanical aggregate measure. So as a matter of tactics if not of principle, developing countries should throw their weight behind the producer subsidy equivalent in any negotiation of aggregate measures of support. -27- If GATT surveillance of agriculture is to achieve anything, it must have clear summaries of the degree of protection both for confronting the negotiating parties and for alerting public and political opinion to the continuing distortion of trade. Although calculation of producer subsidy equivalents is complex (and even contentious), their message is clear: should we be boosting farmers' incomes by x percent? The importance of this point is illustrated by the EEC's cereals policy during the 1980s. At the same time that the EEC was pursuing a "prudent" price policy (since 1981) and a reform package (1985-86), the producer subsidy equivalent on wheat doubled from 28 percent to 58 percent. If countries can negotiate on the basis of rhetoric alone, international surveillance will be a sham. An attractive goal of the Uruguay Round would have been to bring agriculture back into the (strengthened) GATT by making it subject to Articles XI and XVI banning quantitative restrictions and export subsidies. As long as agriculture maintains its special position, it will be difficult to liberalize. For this reason it is important that negotiations ultimately deal directly with policy instruments rather than just with producer subsidy equivalents or other aggregate measures. Thus while aggregate measures of support may be used to monitor progress and possibly even to measure degrees of reciprocity, in the end commitments will have to be written on specific policies. Moreover, while formulas may be used to define the skeleton of a solution, the final package will require considerable amounts of commodity-specific negotiation. It is important, therefore, that developing countries consider the scope they have for easing the process of measure-specific negotiations. -28- I have previously argued (Winters 1987a) that developing countries may have more influence in bilateral negotiations than they realize. They maintain negotiable quantitative restrictions on a large number of items for which the major industrial countries are principal suppliers. Indeed I identified significant scope for negotiations between industrial and developing countries over quantitative restrictions on a principal-supplier basis. A fair number of developing country quantitative restrictions are in the agriculture and food sectors, which suggests that even within the Agriculture Negotiating Group, developing countries have something to offer.11 Moreover, Finger's (1979) analysis suggests that countries offering reciprocal concessions are in a far stronger position to influence the outcome of negotiations than are those that simply sit by and complain. Pitfalls in Negotiations The Uruguay Round is an immensely complex set of negotiations. Grand concepts such as general liberalization, tariffication, and aggregate measures of support are being discussed at the same time as work is proceeding on detailed matters such as the precise specification of particular policy instruments and the definition of certain classes of 11. Of the 787 nontariff barriers on imports for which the United States, Japan, or-the EEC was the principal supplier, 166 referred to agriculture/food. The barriers were defined at the four-digit Standard Industrial Trade Classification (SITC) level, and the importers covered were Argentina, Brazil, Chile, Hong Kong, Korea, Pakistan, Philippines, and Singapore. -29- goods. At an intermediate level lie several technical economic issues which, while not directly a part of the concept of liberalization, are directly related to its implementation. If inappropriately handled, these issues have the power to subvert and undermine the process of agricultural policy deescalation. This section briefly discusses four such issues and suggests how the pitfalls they entail may be circumvented. Applying the old maxim "forewarned is forearmed," the discussion attempts to alert economists on the negotiating teams of developing countries to the dangers that may await them during negotiations. Exchange Rates Exchange rate variability can pose problems for negotiators in defining the level of support for an agricultural product if a fixed internal price is compared with a variable world price. This difficulty has led the EEC to propose that reference prices be fixed in terms of local currencies for purposes of calculating support measures. That is, the EEC proposes that for each producer country, protection be measured relative to the domestic-currency equivalent of internationally agreed prices, with the conversion from international prices to domestic equivalents made at a single historical exchange rate rather than at the current market rate. Thus 'reference prices" will not reflect subsequent changes in exchange rates, so that, say, as the ECU appreciates, the EEC's internal ECU price of imported agricultural goods would remain fixed while the prices of other EEC tradable goods fell. The resulting distortions between each country's internal prices for agricultural and other tradable goods are obvious and unavoidable, and the scope for political chicanery in the periodic renegotiations to reset reference prices or reference exchange rates is -30- evident. The parallels with the EEC's complex green-currency system are close, and just as that system is manipulated for protective purposes so too could be the proposed international system. Moreover, fixing local currency prices, as the EEC proposal implies, would perpetuate current insulation policies and the world price instability they engender. A different response to exchange rate fluctuations is the argument that it is more important to get the world monetary system right than to worry about trade policy since exchange rate movements can swamp all but the largest sector-specific interventions. A superficial reading of some of Schuh's writing (1988, for example) on the macroeconomic influences of farming might give a similar impression--namely, that until we have macroeconomic stability, farm policy is of second-order importance.12 Such arguments are both distracting and misleading. Trade policy affects the relationships between different tradable goods, and artificial wedges between marginal rates of transformation for different tradables are almost always harmful. By contrast, the exchange rate and macroeconomic variables affect trade-offs between tradable and nontradable goods or between goods and money; while distortions in these links may be harmful, they are largely independent of the trade-policy issue. Moreover, macroeconomic conditions change rapidly, so that a trade distortion designed (or permitted) as an offset to a macroeconomic problem in one year would very probably be inappropriate to macroeconomic conditions a few years later. 12. A proper interpretation of Schuh is probably that for any individual country, macroeconomic stabilization should precede agricultural liberalization, although even this view poses some dangers of the sort alluded to in the text. -31- Thus the reduction or removal of distortionary trade policy will be beneficial in the long run almost regardless of what happens macroeconomically, and any claim to the contrary is probably best seen as obstructionism. Preferential Trading Arrangements Preferential trading arrangements constitute a second potential stumbling block during negotiations. An apparent concession by the United States and the EEC would be able to allow developing country members of the Cairns Group and other developing countries a greater share of their markets while keeping their basic protective structures in place. Such an offer of preferential treatment might seem attractive to the parties concerned, but it would not achieve the market shares that would be generated under a liberal trade system. In addition, establishment of such preferential arrangements would set an extremely dangerous precedent for future trading relations. First, it would reinforce the mercantilist notion of trading efficiency. Even if the initial allocations increased the relative shares of efficient producers and so were welfare-improving, the rigidities introduced would almost guarantee that distortions would increase in later years. Second, because only existing producers would be considered in the negotiations, the outcome would inevitably favor them (including those in the United States and the EEC) over producers who are merely potentially efficient. Third, the rent transferred to developing countries by such means would be limited. Fourth, increased rents would accrue to existing producers in the developing countries while efforts to distribute them in some other way would entail increased government intervention. Rents generated in this way would have quite different -32- development implications from those associated with increased export earnings deriving from increased supply at world prices. Fifth, preferential access for certain producers at given prices has insulation effects and thus increases price variability in the residual unrestricted world market; moreover, unless the United States and the EEC reduced their output in return for increased market access, the residual world price would inevitably be reduced by such a scheme. Finally, the quotas agreed to in 1990 could not be guaranteed for future years. The disadvantages of such preference schemes are manifest in the experience with the Generalized System of Preferences (see, for example, Langhammer and Sapir 1988) and in the history of world sugar trade. Both the United Stages and the EEC grant preferential market access to certain developing country sugar producers. The United Stages has reduced its preferential quotas steadily since the mid-1970s. Although the EEC has maintained its quotas, they are very unevenly distributed and at least to some extent accrue to countries with no natural advantages in sugar production. Much of the revenue-enhancing effect of preferential quotas, even for the few lucky countries holding significant quotas, is undermined by the reductions in the residual world price that the quotas induce by increasing world supply and by the price variability that market rigidities entail. None of the documents pertaining to the Uruguay Round suggests that preferential arrangements are being considered, but that does not mean that they will not figure at a later stage. Preferential market access can be more attractive than free trade to the favored producers, and such schemes have the negotiating advantage of internalizing nearly all the gain. That is, all the "cost" to the EEC or the United States of increasing imports through a preferential quota can be presented as a -33- benefit (increased exports) to particular and identifiable negotiating partners. These partners can then be expected to offer concessions in return--in this case, backing the final mercantilist market-sharing package. This is in direct contrast to genuine market liberalization in which, although it is plain that efficient suppliers as a whole can gain, it is not clear beforehand which among them will. Thus it is much more difficult to extract reciprocal concessions and build coalitions for genuine liberalization than it is for market-sharing arrangements. Put crudely, the negotiators from the major markets can bribe certain producers to connive with them by offering fat pickings from their consumers' pockets. In the face of deadlock, and particularly if progress in the rest of the Uruguay Round comes to depend on agriculture, the political attractions of preferential-access schemes may become very strong. Ultimately, few if any developing countries would gain from preferential trading, but avoiding such agreements will require a high degree of coordination and cooperation. In this regard, groups such as the Cairns Group and the Group of 77 may be important, as will be the GATT Secretariat, who should make it clear that preferential-access agreements will not be acceptable solutions to the issue. Special and Differential Treatment A third distraction is special and differential treatment, which has become something of an icon for developing country negotiators, although it serves their interests very poorly. Not only is the protection they seek to maintain almost always economically harmful, but insisting on maintaining it amounts to nonreciprocity and so disempowers developing -34- countries in negotiation: because they are offering so little, they are offered so little. Finger (1979) has shown the advantages to developing countries of participating fully in the Kennedy Round, while I have shown (Winters 1987) that developing countries have 'concessions" to offer, even in bilateral principal-supplier negotiations and even in agriculture. If developing countries wish to break down the restrictions of U.S., Japanese, and European agricultural policies, they have to be prepared to make concessions in their own policies. Otherwise, reluctant liberalizers, especially within the EEC, will be able to shelter behind the excuse of reciprocity. Developing Country Food Imports A more serious issue that may yet subvert the process of negotiating agricultural reform is the position of the major developing country food importers. According to most analyses, the farm policies of OECD countries have lowered the world prices of temperate food products and, through their substitution effects, those of tropical food products as well. These reductions are clearly beneficial to net food importers. I have argued above that these benefits can easily be exaggerated, especially in the medium to long run, but it is undeniable that some of the world's poorest countries are vulnerable to the effects of liberalization on food prices. In terms of the concerns of this paper, however, the issue is less whether any countries will actually suffer than how to prevent the possibility of such suffering from being used as an excuse for no action. First, the size of the problem must be put in perspective. The OECD model (OECD 1989) suggests that liberalization would reduce grain prices, and Loo and Tower (1988) find developing countries gaining from -35- industrial country liberalization. Even Tyers and Anderson (1988), who overstate the problem, find net benefits from OECD liberalization of US$20 billion for industrial countries and net costs of only US$2 billion for developing countries at 1980-82 prices and quantities; at projected 1995 values, the respective figures are US$51 billion in net benefits and US$14 billion in net costs. Thus, there is plenty of net benefit to go around. This observation amounts to saying that current agricultural policies are an appallingly inefficient way of supporting developing country food imports. The obvious way of addressing the distributional consequences of agricultural reform is through increased aid flows. These should come from the beneficiaries of any reform--consumers in the OECD countries and producers in the potential. exporting countries like those in the Cairns Group. A useful way to ease the path of the negotiations would be for the Cairns Group of countries to pledge a percentage, say 5 percent, of their incremental export earnings to aid very poor food importers.13 For OECD countries, the politics are more complex. Politicians in most of them, especially their agricultural negotiators, already perceive themselves as making major sacrifices by contemplating reform; thus, the attractions of making additional fiscal transfers are not great. Several means are possible for easing the perceived burden of increased aid. First, in many respects, this issue is a matter of presentation: the huge potential benefits of liberalization to consumers must be publicized. Second, during the transition period of any reform, OECD countries will still be taxing consumers quite heavily. A commitment 13. Incremental earnings might be defined relative to, say, 1990 exports. -36- to funnel say 2 percent of this implicit or explicit tax to food importers should be feasible.14 As reform proceeds, the amount of transfer would decline; this approach would be attractive politically and would marginally aid the process of reform. It might also focus attention in recipient countries on the need to address long-term food requirements independently of aid. To the extent that aid is needed for longer periods, this approach would at least allow time for its gradual introduction into the general tax base. Third, the rise in world food prices would increase incentives for research and development applicable to production outside the protected OECD markets. Much of this is conducted inside OECD countries, so a modest increase in support would be a (politically acceptable) internal transfer rather than an external one. Finally, goods and other aid flows are currently distributed in a perverse and inefficient fashion. Large amounts of aid go to relatively well-off middle-income countries (for example, Egypt) while the very poor receive little. Rationalization of these flows would further reduce the burden of any increment required as a result of agricultural reform. A major problem of bilateral aid flows is their use for political rather than economic purposes. This is reflected both in their direction and, frequently, their nature. For example, food aid flows owe more to donors' needs to find outlets for surpluses than to the needs of recipients. For this reason, the problems faced by the very poor food importers should be addressed on a multilateral basis, with revenue commitments to an an aid institution and disbursements made quite 14. For practical purposes, the tax could be defined by the nominal protection coefficients of certain products multiplied by consumption. -37- independently of political objectives. While the process of establishing such an institution would occur through the GATT (in collaboration with development organizations such as the U.N. Conference on Trade and Development [UNCTAD] and the World Bank), its natural location would probably be within UNCTAD. Moreover, it is important to note that it would be a temporary body, established as part of the transitional arrangements of the Uruguay Round, and would cease to exist when those arrangements were complete. To be effective, such transitional aid would need to be directed carefully within recipient countries. Cheap food benefits the urban poor at the expense of the rural poor, so it would be necessary to ensure that any reverses were offset in the right quarter. This difficulty does not appear to be insurmountable, however, for the inhabitants of urban areas are frequently more articulate and more easily reached than are those of rural communities. The problem will be to prevent aid from being captured by the middle-classes. As with many other issues discussed in this paper, however, the simple existence of this problem is not a reason for preventing liberalization: any claim to the contrary should be met by the demonstration that the urban poor can be protected while positive benefits still remain for other sections of society. Conclusions Developing countries are not major players in the Uruguay Round of multilateral agricultural negotiations. They will nonetheless be seriously affected by the outcome of the talks and hence have an interest in exerting as much constructive influence as possible. This paper highlights a number -38- of issues that developing country negotiators should consider in formulating their positions over the critical period of the next year: * Developing countries will reap few benefits from a "quantitative solution" to world agricultural problems. Quantity controls would raise prices, offer little scope for increased exports, and exacerbate market instability. * Tariffication will address countries' interests in stabilizing world prices. * Aggregate measures of support will not define liberalization but may help to measure it; among such aggregate measures, the producer subsidy equivalent is preferable to the support measurement unit. * Developing countries should negotiate fully on agriculture as on other issues. Special and differential treatment promises very few direct economic benefits and amounts to little more than nonparticipation in negotiation. * World food prices may rise slightly with liberalization--it is difficult to tell by how much--but the correct response to any resulting hardship is aid and adjustment not the blocking of liberalization. -39- References de Zeeuw, A. 1989. "GATT and Agriculture." European Affairs 1:13-17. Field, H., S. Hearn, and M.G. Kirby. 1989. The 1988 EC Budget and Production Stabilizers. Discussion Paper No. 89.3. Canberra: Australian Bureau of Agricultural and Resource Economics. Finger, J.M. 1979. 'Trade Liberalization: A Public Choice Perspective." In R.C. Amacher, G. Haberler, and T.C. Willett, eds., Challenges to a Liberal International Economic Order. Washington, D.C.: American Enterprise Institute. Josling, T. 1973. Agricultural Protection: Domestic Policy and International Trade. Rome: Food and Agriculture Organization. Krueger, A.O., M. Schiff, and A. Valdes. 1988. "Agricultural Incentives in Developing Countries: Measuring the Effect of Sectoral and Economywide Policies." World Bank Economic Review 2, no. 3 (September):255-71. Langhammer, R.J., and A. Sapir. 1987. Economic Impact of Generalized Tariff Preferences. Thames Essay No. 49. London: Gower Press for the Trade Policy Research Centre. Loo, T., and E. Tower. 1988. Agricultural Protectionism and the Less Developed Countries: The Relationship Between Agricultural Prices, Debt Servicing Capacities and the Need for Development Aid. Working Paper No. 36. Durham, N.C.: Duke University Program in International Political Economy. Maddala, G.S. Forthcoming. "Estimation of Dynamic Disequilibrium Models with Rational Expectations: The Case of Commodity Markets." In L.A. Winters and D. Sapsford, eds., Primary Commodity Prices: Economic Models and Policy. Cambridge: Cambridge University Press. Matthews, A. 1985. The Common Agricultural Policy and the Less Developed Countries. Dublin: Gill and Macmillan. OECD. 1987. National Policies and Agricultural Trade. Paris: Organization for Economic Cooperation and Development. OECD. 1989. Economy-Wide Effects of Agricultural Policies in OECD Countries: Final Results. WP1 of Economic Policy Committee, March (restricted). Parikh, k.S., et al. 1987. Towards Free Trade in Agriculture. Amsterdam: Nijhoff. -40- Runge, C.F., and G.H. Stanton. 1988. wThe Political Economy of the Uruguay Round Negotiations: A View from Geneva." American Journal of Agricultural Economics 70:1146-52. Schuh, G.E. 1988. "Some Neglected Agricultural Policy Issues in the Uruguay Round of Multilateral Trade Negotiation." Paper presented at Eighth Agricultural Sector Symposium, World Bank, January. Tangermann, S., T. Josling, and S. Pearson. 1987. "Multilateral Negotiations on Farm Support Levels." The World Economy 10:265-82. Tyers, R., and K. Anderson. 1988. Journal of Agricultural Economics. Valdez, A., and J. Zietz. 1980. Agricultural Protection in OECD Countries: Its Cost to Less Developed Countries. Research Report No. 21. Washington, D.C.: International Food Policy Research Institute. Winters, L.A. 1987a. "Negotiating the Abolition of Non-Tariff Barriers." Oxford Economic Papers 39:465-80. Winters, L.A. 1987b. 'The Economic Consequences of Agricultural Support: A Survey." OECD Economic Studies 9 (Autumn):7-54. Winters, L.A. 1988. The So-Called "Non-Economic" Objectives of Agricultural Policy. OECD Working Paper No. 52. Paris: Organization for Economic Cooperation and Development. Zietz, J., and A. Valdez. 1988. Agriculture in the GATT: An Analysis of Alternative Approaches to Reform. Research Report No. 70. Washington, D.C.: International Food Policy Research Institute.