II. Agreement on Agriculture

PURPOSE

The objective of this module is to summarize key issues that need to be taken into account by countries in reviewing their bound tariff levels, given the possibility that they may have to reduce these rates further in the next round of negotiations. Any rules agreed will also apply to other countries and will influence export market access. How developing countries might evaluate proposals to reduce bound tariffs in developed countries from the point of view of improving export market access is examined in Module II.4. The issues outlined in that module should be considered together with the implications of different bound tariff structures for itself as outlined in this module by each country.

CONTENTS

2.1 Introduction

2.2 How were tariffs bound in the Uruguay Round?

2.3 An overview of the post-UR tariff structure of the developing countries

2.4 Possible approaches to reduce bound tariffs in the next round

2.5 Some considerations for setting new bound tariffs

KEY POINTS

• Tariffs were bound in the UR using a tariffication formula set out in the Modalities document accompanying the AoA, although developing countries could opt for ceiling bindings.
• The overwhelming majority of developing countries chose to offer ceiling bindings rather than follow the tariffication route. Bound tariff levels differ across developing countries, with higher levels in sub-Saharan Africa, the Near East and South Asia compared to East Asia and Latin America.
• A variety of approaches to reduce bound tariffs are available for the next round of negotiations; a key difference is the extent to which alternative approaches address the question of tariff peaks in tariff schedules.
• Developing countries should take into account a number of economic and non-economic considerations in setting bound tariff levels in the future.

2.1 INTRODUCTION

Bound vs. actual tariffs

One major achievement of the Uruguay Round (UR) Agreement on Agriculture (AoA) was the prohibition of non-tariff barriers (NTBs) to agricultural trade (except for some specific derogations), requiring that all such trade takes place under a tariff-only regime. By agreeing further to bind nearly all tariffs, WTO Members also capped the levels to which applied rates can be increased at any time.

As the next round of agricultural trade negotiations approaches, one of the challenges WTO members face is preparing themselves for further reduction of the bound tariffs, in the spirit of the long-term objective of the reform process in agriculture. Although what matters as a policy concern on a day-to-day basis is the applied rate, bound rates define the extent of the flexibility to vary the applied rates upward in response to particular economic circumstances. Hence, the setting of bound rates is a matter of strategic importance.

The determination of the appropriate level of bound tariffs is a fairly complex task as this involves judgement on several factors, such as future movements of world market prices and exchange rates, the evolution of domestic competitiveness, availability of contingency measures, revenue considerations, and so on.

This module will cover the following four aspects:

• the guidelines for tariff binding set out in the Modalities document;
• the tariff structure which emerged in developing countries after this process;
• possible formulae for tariff reductions in the next negotiations; and
• issues for developing countries when considering new bound tariff levels.

2.2 HOW WERE TARIFFS BOUND IN THE URUGUAY ROUND?

In the AoA, reforms in the area of border measures were effected in three steps: prohibiting the use of NTBs; binding of tariff rates; and reducing these over time. Article 4 of the AoA, Market Access, covers these subjects, along with some other market access issues such as Tariff Rate Quotas and special agricultural safeguards. Box 1 reproduces Article 4 which, despite its immense importance in the AoA, is one of the shortest.

Box 1: Article 4 of the Agreement on Agriculture of the Uruguay Round

Ceiling bindings were an option for developing countries

As guidelines for countries to prepare their Schedules, including tariff offers and other market access provisions, a negotiating document called Modalities was prepared. These guidelines took into account three factors: i) whether or not a particular tariff line was bound previously; ii) whether or not trade in that product was subject to any NTB; and iii) the development status of a country. In addition, bilateral tariff negotiations also played some role in many cases. The general rules are summarized in Table 1.

Table 1: Guidelines for preparing tariff offers and bindings in agriculture

2.3 AN OVERVIEW OF THE POST-UR TARIFF STRUCTURE OF THE DEVELOPING COUNTRIES

Given the options and possibilities as described in Box 1, an overwhelming majority of developing countries chose to offer ceiling bindings, rather than follow the tariffication route. A majority of developing countries, and virtually all least developed countries, also chose to offer a single rate of binding for all agricultural products. The bindings were mostly done at relatively high rates, e.g. at 100 percent or even more, which was not uncommon in the UR even in the case of many developed countries.

Whatever the arguments in favour of lower tariffs, higher bindings do have one advantage in that they give some scope for bargaining, e.g. to be able to reciprocate with deeper cuts in return for greater access to others' markets. In practice, however, this would depend upon the size of the market, which is typically small in the case of a majority of developing countries individually.

Most SSA countries opted for high ceiling bindings...

In Sub-Saharan Africa (SSA), a majority of countries (19 of the 36 shown in Table 2) bound tariffs at 100 percent or more, 10 countries at between 50 percent and 100 percent, and the remaining 7 at less than 50 percent. These seven include five countries belonging to the Southern African Customs Union (SACU). Most of the countries in the first two groups offered ceiling bindings at a single rate for all agricultural products. Where applicable, they also bound "other duties and charges" as required by the AoA. The SACU countries, on the other hand, followed the tariffication approach. Also, because of their common external tariff structure, there were small differences in tariffs between them, and their bound rates were relatively low.

Applied tariff rates in recent years, e.g. 1995-98, have been generally low, especially when compared with the bindings, giving a considerable margin for discretionary increases in applied rates if and when required. As tariffs were bound at an uniform rate, there is no apparent distinction between "sensitive" and other products for most countries. But in the few cases where uniform bindings were not made, there was a clear pattern of setting higher tariffs on products that compete with domestic production and lower tariffs on non-competing importables or non-tradables.

...as did Near Eastern countries...

In the Near East, a majority of developing countries, and virtually all least-developed countries, chose to offer ceiling bindings rather than follow the tariffication process. In a large number of cases, the bindings were made at high levels, e.g. 100 percent or more (Table 2). Bindings were also rather high for countries that underwent tariffication, e.g. Morocco and Tunisia. In this region, Egypt was an exception in that bindings were done at relatively low levels, e.g. between 10-20 percent for most agricultural commodities, with only some exceptions, e.g. 60 percent for poultry meat and oranges.

In almost all the countries, the bound rates are substantially higher than the rates currently applied (e.g. during 1995-98 period). To the extent that these bound rates are mainly cosmetic, i.e. without corresponding protective effect, their reduction, up to some extent, should have little impact on real sectors. But this is something to be analysed. Egypt is an exception here, with a very narrow gap between the bound and applied rates, which means that the scope for the further reduction of the bound rate, without causing an impact on the real economy, is much less.

...while bound tariffs in South Asia are higher than in East Asia

In Asia, the bound tariffs for agricultural products are relatively high in South Asia compared to East Asia. In South Asia, the simple average of the bound tariffs for over 600 agricultural tariff lines was 122 percent for India, 101 percent for Pakistan and 200 percent for Bangladesh, but much lower at 50 percent for Sri Lanka. In East Asia, the corresponding rates were 40 percent for Fiji, 49 percent for Indonesia, 31 percent for Malaysia, 63 percent for Republic of Korea, 31 percent for the Philippines, and 37 percent for Thailand. While Sri Lanka, Bangladesh and Fiji offered a uniform rate for all products, these were done at different rates by others.

Further tariff reductions would impact these countries differently. For example, while a 50 percent reduction would still mean a binding of 100 percent for Bangladesh, the new rates would be very low, relatively, for Fiji and Sri Lanka. For other countries, there are commodity-specific implications depending upon the currently bound rates. If one were to define "sensitive" commodities as those with high bindings, these would appear to be as follows: several vegetable oils for India; wheat, maize and several others for Pakistan; rice, maize and dairy products for Indonesia; rice, dairy products and poultry meat for Malaysia; several types of meats for Myanmar; sugar, poultry meat, wheat flour and dairy products for the Philippines; and soy oil, skimmed milk, palm oil and some livestock products for Thailand.

Tariffs levels are generally low in Latin America

In Latin America, bound tariffs are generally low and applied tariffs even lower. There are some important exceptions to this general pattern in that some countries have not as yet switched to a full tariff-only regime, implementing rather some variant of a price-band policy, defended with variable tariffs, for a select list of "sensitive" food products. But on the whole border protection is low and so these countries will find it more difficult to further reduce the bound tariffs in the next round. For them, access to a simpler contingency protection measure, such as the agricultural safeguards, could become an important negotiating issue.

2.4 POSSIBLE APPROACHES TO REDUCE BOUND TARIFFS IN THE NEXT ROUND

No method or formula for further reduction of the tariffs has been identified as yet for the next round within the formal WTO process - in fact, this itself would be a subject for negotiations. However, reflecting the importance of this matter, this subject has attracted considerable attention from analysts. What follows is a summary of various ideas, albeit all informal, by which tariffs may be reduced. Given that tariff binding is a matter of strategic concern, it is important for countries to be aware of these possible methods and how these would affect their currently bound tariff rates.

Linear formulae are straight-forward but fail to tackle tariff peaks

Across-the-board linear reduction. A linear reduction formula is simply T_n = (1-r*t)*T_0, where Tn and T₀ are new and original tariff rates respectively, r is agreed reduction rate and t is the time period for reduction. For example, if r = 0.06 (i.e. 6 percent reduction per year) and t = 6 years, a 100 percent tariff is reduced to 64 percent. This method was applied in the Kennedy Round with the "r*t" set at 50 percent. As a result of some exceptions negotiated subsequently, the final reduction was 35 percent. The approach is both simple and transparent. While tariffs could be cut significantly if the reduction rate is high (e.g. 50 percent compared to 36 percent on average in the UR), another linear cut would still leave many tariff peaks in agriculture left by the UR formula.

Linear reduction with conditions on minimum cuts. This was the formula used in the UR AoA (36 percent average reduction with a 15 percent minimum per tariff line). Although tariffs were reduced by an average of 36 percent, the method left many tariff peaks, as countries had the freedom to cut tariffs on "sensitive" products by only the minimum 15 percent while reducing by more for others, in order to reach the (un-weighted) average of 36 percent. This formula could be improved, e.g. by raising the minimum to, say 25 percent, or by seeking a balance in the trade volume between those with higher and lower than average cuts, i.e. trade-weighted tariff reductions.

The UR formula with the same base as in the UR. Rather than using the bound rates reached at the end of the implementation period of the UR as the benchmark for further reduction, a further 36 percent cut in the average level of tariffs from the same base as in the UR would imply a 72 percent cut over the two reform periods, a significant reduction over a dozen years or so. This approach has some other advantages, e.g. giving a sense of the continuity of the process of reform by using the same formula; no controversy over the choice of a new base period; and full "credit" for unilateral reductions during the negotiation period.

Successive linear reductions. Compared with the linear method, here the base tariff rate, T_0, is adjusted every year to its new level. The formula for this, also known as a radial formula, is T_n = (1-r)^t * T₀. With this, if r = 0.06 and t = 6 years, a tariff level of 100 percent is reduced to 69 percent, compared with 64 percent with the linear formula. As the base itself gets reduced every year, the overall reduction at the end of the period is smaller. However, for a smaller reduction rate and a shorter time period, the difference in reduction rates from the two formulae is not much.

Options to tackle tariff peaks

Harmonization of tariff rates - the Swiss Formula. This formula was used in the Tokyo Round to harmonize tariff peaks on industrial products left as a result of the linear formula used in the Kennedy Round. The Swiss formula is Tn = (a_max * T₀)/(a_max + T₀), where a_max is the upper bound on all resulting tariffs. With a_max = 50, an initial tariff of 40 percent would be reduced to 22 percent while a 100 percent tariff would be reduced to 33 percent. On the other hand, with a_max = 25, a 40 percent tariff is reduced to 15 percent and a 100 percent tariff is reduced to 20 percent. The value of a_max then becomes the parameter for negotiations. Figure 1 shows how three of these methods discussed here compare in terms of tariff reductions.

Capping all tariffs at some maximum rate. For example, a maximum rate of 60 percent could be agreed to which all higher tariffs would have to be reduced over an agreed period. This rule may be applied in conjunction with other reduction methods.

Squeezing the water in the tariff

Using actual protection rates for recent years as the benchmark. In this approach, negotiators agree to eliminate the gap, or a good part of it, between the bound and the applied rates, the so-called "discretionary protection" or "water in the tariff", using some recent period to measure the gap, e.g. 1995-97. This approach, while it makes some economic sense, appears problematic due to problems associated with measuring (or agreeing with the measurement of) the protection rate. This was one of the problems that led to inflated tariff equivalents (and thus bound tariffs) on many commodities in the UR, which came to be known as "dirty tariffication". This method is less helpful for developing countries where domestic prices tend to be lower than or similar to world reference prices, resulting in negative or zero bound rates, which would not be acceptable.

Figure 1: Examples of tariff reductions resulting from three tariff cutting formulae

2.5 SOME CONSIDERATIONS FOR SETTING NEW BOUND TARIFFS

Bound tariffs provide flexibility to vary applied tariffs in order to pursue certain goals, e.g. some stabilization of markets. What duties are actually applied at a particular time, however, depends upon economic circumstances prevailing then. The question asked here is simple - what approximate range of bound tariffs is likely to provide the necessary degree of flexibility for pursuing policies in coming years? As bound tariffs have a longer-term connotation in the WTO system, i.e. they remain bound until re-negotiated, approaching this problem involves strategic thinking and considerable analysis, at the level of individual commodities. What follows is a summary of the various economic and non-economic considerations that should ideally be taken into account in such an exercise.

Why bind tariffs at different rates?

Consider offering a differentiated tariff structure. In the UR, most least developed and many developing countries committed a single bound duty for all agricultural products, e.g. 100 percent or 150 percent. But there are some advantages in considering differentiated tariff structure. One is that it allows for higher bindings for selected commodities (e.g. the so-called "sensitive" ones) to reach a given target of overall average reduction, by cutting tariffs more deeply on others, e.g. the "less-sensitive" products. This is especially so if the UR-type formula is used for reductions. This approach was in fact followed by many countries in the UR. Possible drawbacks of this approach should also be considered. A differentiated tariff structure increases the distortions in the price signals which producers face and adds to the overall cost of operating an administered price policy.

Identify a sub-set of sensitive commodities. If a country decides to differentiate its tariff bindings, the next step would obviously be the identification of these "sensitive" products or sub-sectors. For this, there have to be some clearly-defined criteria. For example, food security is often mentioned as one of these considerations; but then one needs to go further to identify food security commodities and sectors, which do not necessarily have to be foodstuffs if other commodities/sub-sectors contribute significantly to incomes and entitlements of the poor and food insecure. Other criteria could be employment, or the protection of infant industries, income distribution, or even revenues. Further analysis would then be required to define an approximate range of flexibility on tariffs that may be required for these commodities or sub-sectors. Expected instability in world commodity prices would also be a factor to reckon here. At the same time, as in any negotiation, one needs to be also prepared to give up something - by identifying commodities/sub-sectors where it would be possible to "live with" low bound tariffs. These could be, for example, commodities where the industry is in a position to absorb some shocks or those where the country is expected to be competitive.

Ensure tariff bindings are consistent

Carefully review the case of related commodities. Where commodities compete or substitute in production and consumption, bindings for them should be made coherently. In Sri Lanka, for example during 1995-98, the government found it desirable to vary duties on rice by almost the full maximum extent allowed by the bound rate in response to price developments on wheat, which competes closely with rice in consumption. In another example from a different country, import duties on four types of vegetable oils were bound at 45 percent while for the rest of the oils the binding was at 300 percent. As vegetable oils substitute highly in consumption, the 45 percent binding (plus or minus some margin due to differences in world market prices) essentially sets the limit to which applied tariffs can be levied on any vegetable oil, unless the goal is to regulate the import of a particular type of oil with higher duty. In this case, the 300 percent binding has little practical significance. These are the types of considerations that need to be taken into account in analysing a new set of tariff bindings for related or "like" products.

Tariffs on products involving processing. In many cases, it would be necessary to take an industry-wide view in setting tariffs. For example, if the objective is to protect (or not to dis-protect) flour mills or the vegetable oil processing industry, it would be necessary to bring into the analysis the concept of effective protection rate (see Module I.4), as this would determine the appropriate relationship between tariffs on inputs (e.g. wheat) and outputs (e.g. flour). Many examples can be found in the UR Schedules where this consideration did not seem to have been taken into account in setting tariff bindings.

Addressing the problem of very low and zero bound rates. Some countries are also confronting the problem caused by very low or even zero bindings in the UR. It is not clear why they chose to bind tariffs at a relatively lower rate, e.g. at rates even lower than the extent to which world market prices are known to fluctuate. An example is the bindings by a country of between 10-20 percent on most cereals. Could it be due to some oversight, i.e. inadequate appreciation of the importance of tariff bindings or was it a conscious step? In some cases, zero bindings, e.g. India's rice, some coarse grains and some dairy products, were done in earlier Rounds when, and until recently, this was not much of a problem as agricultural trade was mainly regulated by non-tariff barriers. Under a tariff-only regime, and perhaps with changed economic circumstances, these could be the sources of difficulties, especially so when these are "sensitive" commodities.

Ensuring consistency between MFN and regional tariff rates. In negotiating new levels of Most-Favoured-Nation (MFN) tariffs in the multilateral forum, countries need also to take into account the bound MFN committed by other members of a Regional Trading Agreement (RTA). This is not an issue in the case of a Customs Union, where there is one set of common external tariffs, but applies to the case of a Free Trade Area where members normally have different MFN duties. Such a review would help maintain consistency in the MFN rates of the members and avoid possible trade problems.

Other issues in setting bound tariffs

Accessibility to contingency safeguard measures. These measures, such as antidumping, countervailing and safeguards, have little value as long as bound tariffs are high. But as the process has been set for reducing the bound rates, the time has come for countries to consider whether they are in a position to resort to these measures. For the developing countries in particular, it is important to review whether their institutional, analytical and legal capability permits them to resort to one or more of the contingency measures as and when required. This aspect would also have to be taken into account while considering the levels of the bound rates for the next round.

Economic efficiency issues and border protection - most of the discussion in this paper focussed on UR rules and the issue of flexibility in border measures inherent in higher tariff bindings. Having a necessary degree of flexibility is useful, but it is also important to recognize the possible socio-economic costs associated with protection, which include possible increased resource costs and inefficiency, misuse by special-interest groups and rent-seekers, and so on. Also, unnecessarily high tariff bindings mean less transparency and predictability of trade, which is likely to hurt private sector investment. Obviously, these economic considerations have to be factored into the process of the analysis discussed above.

Conclusion

In conclusion, it is clear that there is no one general approach or formula for determining appropriate levels of tariff bindings that would apply to all countries. Rather, as said above, this would depend upon particular economic circumstances, both current and prospective, affecting the commodity or the sub-sector in question. Moreover, unlike the case with applied tariffs that are often set in response to short-term developments, bound tariffs remain fixed for a much longer period and so their determination requires some degree of strategic thinking. Such an analysis should be a high priority task for countries preparing for the round of negotiations.

Table 2: Uruguay Round Tariff Commitments on Selected Agricultural Commodities - Sub-Saharan Africa

Source: Uruguay Round Country Schedules.

Table 2 (cont'd): - Sub-Saharan Africa

Source: Uruguay Round Country Schedules.

Table 2 (cont'd): - Sub-Saharan Africa

Source: Uruguay Round Country Schedules.

Table 2 (cont'd): Uruguay Round Tariff Commitments on Selected Agricultural Commodities - selected Near East Countries

¹ Bahrain maintains a bound rate of 200 percent on beer, wine and other fermented beverages, 100 percent on tobacco and 35 percent on the rest of the agricultural products.

Source: Uruguay Round Country Schedules.

Table 2 (cont'd): Uruguay Round Tariff Commitments - selected agricultural commodities and countries - Asia

² Postponed tariffication.

Source: Uruguay Round Country Schedules.