Économie internationale 2001- 3 (no 87)| ISSN 1240-8095 | ISSN numérique : en cours | ISBN : | page 89 à 106 Distribution électronique Cairn pour les éditions La Doc. française. © La Doc. française. Tous droits réservés pour tous pays. Il est interdit, sauf accord préalable et écrit de l’éditeur, de reproduire (notamment par photocopie) partiellement ou totalement le présent article, de le stocker dans une banque de données ou de le communiquer au public sous quelque forme et de quelque manière que ce soit.

Agricultural tariff rate quotas as a development instrument

Alan Matthews
Cathie Laroche Dupraz [1]

1

The use of tariff rate quotas (TRQs) was legitimised and introduced as a market access instrument in the Uruguay Round Agreement on Agriculture (AoA). The motivation behind such an instrument was to guarantee minimum levels of market access and to safeguard current levels of access in the face of the high MFN tariffs resulting from tariffication. 1371 TRQs have been notified by 37 countries to the WTO as a result of the Uruguay Round (G/AG/NG/5/7) [2]. Already, in the scheduling of TRQ commitments, it was alleged that various strategies were used to limit quota volumes (a process which has been called “dirty quotification”, see Bureau and Tangermann, 1999). For example, a group of developing countries complained that domestic consumption calculations had been manipulated in establishing minimum access levels (G/AG/NG/W/37). They also emphasised that the common practice of establishing TRQs based on highly aggregated commodity groups, rather than on a product-by-product basis was contrary to the specifications set out in the Modalities where it was agreed that minimum access levels should be established on a relatively disaggregated product level. Various problems have also been identified in the implementation of TRQs (de Gorter and Sheldon, 2000 ; Skully, 2001). A majority of TRQs are not being filled and thus minimum access commitments are not being met (G/AG/NG/5/7). While there may be market explanations for this lack of demand, there is widespread agreement that quota underfill is in part attributable to the administrative methods employed to implement TRQs. TRQs also generate rents, and the allocation procedures to distribute those rents distort trade and can be subject to political influence (Abbott and Morse, 1999 ; Abbott, 2001).

2

In the agricultural negotiations currently under way in the Special Session on Agriculture in the WTO, a number of countries have made proposals to improve the administration and size of TRQs. Exporting countries are calling for increases in tariff rate quotas to improve market access opportunities. The US, for example, has proposed that these should be substantially increased by annual increments over a fixed period and their functioning improved, including dealing with unfilled quotas. For this purpose, it proposes to base the reduction of in-quota duties on the historical performance of TRQ fill rates : the lower the fill rate, the deeper the duty cut. An automatic trigger mechanism is suggested to reduce in-quota duties in response to falling fill rates (G/AG/NG/W/58). The Cairns Group has also called for substantial increases in all tariff quota volumes (G/AG/NG/W/54). Canada has proposed that, where tariff peaks remain after a further round of MFN tariff reductions, new TRQs could be opened in order to guarantee a minimum level of market access (G/AG/NG/W/12). The EU’s position is noticeably silent on the issue of increasing TRQs. However, it does propose that rules and disciplines should be defined to increase the transparency, the reliability and the security of the management of TRQs, in order to ensure that the concessions already granted are fully realised (G/AG/NG/W/90).

3

Developing countries have also called for a substantial increase in TRQs and for the simplification of their administration. One group of developing countries wants arrangements to ensure that new suppliers from developing countries have equal access to allotments within TRQs and has called for the mandatory filling of quotas, in developed countries, before imports take place at the above-quota level (G/AG/NG/W/37). India has called for the stricter application of the MFN principle in allocating TRQs but with special preference being given to developing countries having less than $1,000 per capita income (G/AG/NG/W/102). Nigeria also calls for tariff quotas to be made global, and that where bilateral quotas exist, global quotas should be in addition to these countries’ quotas and allocated to countries that are not covered by bilateral country quotas (G/AG/NG/W/130). The Small Island Developing States (SIDS) propose that certain percentage increases in minimum access TRQs should be allocated at a zero in-quota tariff to SIDS. They also propose that specific duty-free TRQs outside minimum access quotas should be provided to SIDS (G/AG/NG/W/97). Mauritius, as a significant TRQ beneficiary, is concerned to maintain the value of trade preferences and wishes to maintain a meaningful difference between in-quota and out-of-quota tariffs (G/AG/NG/W/96).

4

There are two important policy ideas contained in these proposals. First, with varying degrees of enthusiasm all participants appear to accept that TRQs will continue to be part of the next agricultural agreement, and most are seeking an expansion in their volume. This is despite the recognition in India’s proposal that TRQs have perpetuated trade distortions by legitimising quantitative restrictions, generating quota rents and denying market access to newcomers. The other idea is that TRQs might be administered preferentially to favour market access by developing countries or a subset of them. Preferential treatment of developing countries under TRQs could take a number of forms. First, in-quota tariffs might be levied at a lower rate or eliminated for developing country exporters. Second, a proportion of increased tariff rate quotas might be reserved for developing country exporters or a sub-set of them. Third, licensing arrangements could be administered to ensure the transfer of quota rents to developing country exporters.

5

Any or all of these options would appear to lead to net gains to developing country exporters. However, a full discussion must recognise that gains in multilateral trade negotiations require the expenditure of negotiating effort and capital. Pursuing one negotiating objective comes at the cost of neglecting others. In particular, efforts to obtain more favourable TRQ access could distract developing countries from seeking MFN tariff reductions. It is thus important to weigh up and evaluate the costs and benefits of alternative approaches.

6

Have developing countries benefited from existing TRQ access ? Which countries would be likely to benefit from expanded TRQ access, and for which commodities ? How would preferential TRQs fit into other schemes of preferential access for developing country exports ? Are the gains from pursuing expanded TRQ access likely to be significant ? This paper makes a limited attempt to answer these questions using the EU market as a case study. First it examines the use developing countries make of the EU’s TRQs in agricultural trade. It examines the legal issues raised by the proposal for preferential TRQ administration. Then it discusses whether developing countries should spend their trade-negotiating capital on improving market access by increasing TRQs instead of seeking further reductions in MFN tariffs. The last section summarises the conclusions of the paper.

Use of EU TRQs by developing countries

7

We wish first to determine whether developing countries benefit from TRQ access. To keep the data task manageable, we focus on the EU market for agricultural and food imports. In 1997, this amounted to 51.7 billion ECU (TABLE 1) [3]. The most important commodity group in terms of imports is processed foods and drinks which accounted for about 16 billion ECU, followed by fruits, vegetables and nuts (about 11 billion ECU), tropical products which include coffee, tea and cocoa (about 8.7 billion ECU) and oils and oilseeds (about 8.1 billion ECU). Almost 60 per cent of these imports are supplied by developing countries. Latin American countries have a 28.0 per cent share of the total, non-least developed ACP countries account for 12.3 per cent, and non-least developed Asian countries account for a further 12.2 per cent. Least developed countries (LDCs) account for only 3 per cent of the EU’s agricultural imports.

Table 1EU agricultural imports, 1997

8

These imports enter the EU on both MFN terms and also under a complex network of preferential access agreements on both a reciprocal and non-reciprocal basis (WTO, 2000). More than 20 countries in Eastern Europe, Africa and Latin America have now signed free trade agreements with the EU. These include the Central and Eastern European countries in the context of European Agreements, and neighbouring countries in the Mediterranean basin under the so-called Euro-Mediterranean Agreements. The EU has also entered free trade arrangements with South Africa and Mexico, and negotiations are under way with Chile and MERCOSUR. Apart from reciprocal free trade arrangements, it has initiated two non-reciprocal trade arrangements : the Generalized Scheme of Preferences (GSP) and the Lomé Convention (now Cotonou Agreement) which extends preferential trade terms to African, Caribbean and Pacific (ACP) countries. More recently, the EU announced an “Everything But Arms” (EBA) initiative directed to all least developed countries.

9

The EU established 85 TRQs in its Schedule resulting from the Uruguay Round. An extra quota for grape juice and grape musts was added in September 1996 following negotiations in the context of EU Northern enlargement. A quota for rum and taffia was added in July 1997 resulting from an agreement between the EC and the USA on spirituous beverages. This rum quota has not been taken into account in the following analysis which is based on 1997 data. In its notifications to the WTO, the EU distinguishesd between current access and minimum access quotas. Current access quotas were designed to safeguard historical quantities imported under special arrangements, whereas minimum access quotas were designed to fulfil the minimum access obligations of the AoA. 44 of the EU TRQs are current access quotas, 36 are minimum access quotas, while the remaining 6 are non tariffied product quotas (these quotas were opened to comply with the AoA for products which did not have to undergo tariffication to convert non-tariff barriers to tariffs).

10

Tariffs under the EU’s current access quotas are much lower than the respective out-of-quota tariff. Bureau and Tangermann (1999) estimate that, on average, for the 50 quotas under current access and for non-tariffied products, the in-quota tariffs show a reduction of 80 cent compared to the out-of-quota tariff at the beginning of the implementation period. Since the in-quota tariffs remain unchanged over the implementation period while the out- of-quota tariffs are reduced by an average of 36 per cent, these authors estimate that in- quota tariffs for current access quotas will be about one-third of the corresponding out-of-quota rate by 2001.

11

For TRQs under minimum access, the EU applied a relatively uniform reduction. Most in- quota tariffs have been set at 32 per cent of the out-of-quota MFN initial (base) tariff. Exceptions include milled rice, durum and quality wheat which are subject to a zero in-quota tariff and for high quality meat where there is a much lower in-quota tariff set compared to the very high MFN tariff. Also, these in-quota tariffs are not scheduled to change over the implementation period, so they will be closer to 40 per cent of the out-of quota tariff by the end of the implementation period (Bureau and Tangermann, 1999).

12

The EU used current access quotas to maintain previously-existing preferential access arrangements. Of the 44 current access TRQs, 14 are allocated to a particular list of countries. Several of these quotas list developing countries as beneficiaries (including China which was not a WTO member when the EU Schedule was submitted). Some of these quotas are allocated to ACP countries as a result of the Lomé Convention (now the Cotonou Agreement). This includes four quotas for sheep, goats and mushrooms as well as the 1.2 million tonne quota of sugar. Others are allocated to Central and Eastern European countries. The EU Schedules mention that, for 18 out of the 36 minimum access quotas, the EU may count against these quotas preferential imports from Central and Eastern European countries under the Europe Agreements. This is the case for pigmeat (5 quotas), poultry (3 quotas), dairy products (7 quotas) and processed eggs (3 quotas). However, neither the quantities admitted under quota nor the eligible countries are specified in the Schedule itself.

13

EU TRQs are usually defined at the HS8 digit level of the Harmonised System tariff classification. The 86 TRQs considered in this paper cover 335 individual HS8 tariff lines. To provide a clearer idea of the products covered by TRQs, TABLE 2 sets out the 86 TRQs grouped according to the HS4 digit categories of the commodities which they cover. Most TRQs include tariff lines from within the same HS4 category, although there may be more than one TRQ opened within a HS4 category. There are 43 HS4 categories in which TRQs have been opened for individual tariff lines, or around 22 per cent of the eligible total. Not all products specified at the 8-digit level within a HS4 category are necessarily covered by a TRQ, so the proportion of eligible HS8 lines covered by a TRQ is lower than this percentage.

Table 2EU TRQs grouped by HS4 digit codes by volume

14

TABLE 2 distinguishes between whether access is provided under current access, minimum access or non-tariffied product quotas. Four products stand out with respect to the importance of their TRQs in terms of the volume of imports covered. In each case, the TRQs concerned are mainly current access quotas. The products are : manioc, arrowroot and sweet potatoes (where three TRQs were introduced to guarantee access to Thailand and Indonesia for manioc products previously exported under a Voluntary Export Restraint (VER), as well as to China for arrowroot and sweet potatoes) ; maize (where the current access quota of 2m tonnes represented compensation to the United States for accession of Spain to the EU) ; bananas (where the TRQs arose from the EU’s attempts to regulate market access following the introduction of the common market organisation for bananas in 1993) ; and sugar (where the TRQ represents commitments to import sugar under the Lomé Convention from ACP countries and India). Other products where sizeable TRQs have been opened include meats, some dairy products and eggs, fruits and vegetables and cereals.

15

The final column of TABLE 2 shows actual imports in each HS4 category compared to the TRQs opened in that category. Actual imports are calculated only for those HS8 tariff lines in the category for which TRQs have been opened and thus do not represent total imports in that category. Where the percentages are less than 100, there is a strong presumption that all imports enter under the TRQs and that the TRQs have created or, in the case of current access quotas, maintained trade flows which would not otherwise occur because of high over-quota tariffs. This would appear to be largely the case for meat (except poultry), dairy products, cereals (except rice) and cane sugar. In the case of HS4 categories where the percentages are substantially greater than 100 and there are significant trade flows outside of the TRQs, then the main function of the TRQ will be to create rents, which may or may not accrue to the exporter, rather than to create additional trade. This appears to be the case for fruit juices, apples, citrus fruits, potatoes and other vegetables.

16

TABLE 3 provides information on the importance of TRQs, in value terms. For presentational purposes the HS4 categories have been aggregated into 8 commodity groups. The is to see how important TRQ trade is with respect to total EU agricultural imports. Three indicators are used. The first is the actual value of imports in the HS8 tariff lines covered by TRQs and their importance relative to total imports (shown in Column 3). The drawback of this measure is that, as shown in TABLE 2, for some of these tariff lines considerable trade takes place outside of the TRQ. Thus the values in Column 2 are generally greater than the value of actual TRQ imports reported in Column 6, with the exception of sugar and sugar confectionery (this exception is discussed below). A second indicator tries to estimate the potential value of TRQ quota trade. This is done by calculating the unit value of the HS8 trade flows for which TRQs are opened and multiplying the eligible or maximum TRQ quantities by these unit values (Column 4). The assumption made is that imports entering the EU market under TRQs are broadly representative, in terms of source and quality, of all imports entering under a HS8 tariff line. The third indicator is similar to the second one, except that notified imports under each TRQ are used instead of the eligible TRQ volumes to measure the actual value of trade (see Column 7). Notified imports are the volume of imports notified by the EU to the WTO and, for some TRQs, correspond to the volume specified in the licenses given to importers, not to the actual quantities imported (G/AG/N/EEC/13). It has been argued that this leads to an over-estimation of actual TRQ imports as licenses may not be turned into imports. The EU holds that this is not the case as importers must provide a deposit when applying for a licence and would be unlikely not to make use of this import permission (Bureau and Tangermann, 1999). In 1997, it appears that the value of notified imports in the sugar and sugar confectionery sector did exceed the value of actual imports in that year. Apart from the licence utilisation issue, this could also be explained by differences in the periods covered by the licences and actual imports [4].

Table 3Relative importance of TRQ products and TRQ imports by main commodity, 1997

17

The total value of TRQ imports in 1997 reached ECU 5.5 billion or just under 11 per cent of EU agricultural and food imports. Expressing TRQs in value terms emphasises the greater importance of TRQs for live animals and meats. These potentially cover a greater value of trade than TRQs for fruits and vegetables (which include manioc, arrowroot, sweet potato and banana TRQs, as mentioned previously) although actual imports are still greater under the TRQs opened for fruits and vegetables. This conclusion would be reinforced if account were taken of the ACP banana quota which is not notified as a TRQ. TRQs in these two commodity groups accounted for 69 per cent of all TRQ imports in 1997. No TRQs have been opened for tropical products or oils and oilseeds where substantial imports take place in any case. On the other hand, half or more of actual imports in the case of sugar, dairy products and live animals and meat entered under TRQs. Potentially, EU agricultural TRQ imports could amount to 14 per cent of actual imports though, in practice, these reach just under 11 per cent. The difference between actual and potential TRQ imports can be defined as the fill rate. [5] The average fill rate expressed in value terms is 76 per cent. Sugar, dairy and fruit and vegetable TRQs have the highest fill rates, while fill rates for cereals and milling products, live animals and meat and other processed foods and drink are below average [6].

18

In the context of this paper, we are interested in the extent to which developing countries have been able to make use of the EU’s TRQs. TABLE 4 shows the ECU 5.5 billion distribution of TRQ imports by region of origin for the six commodity groups where TRQ imports actually took place in 1997 [7]. In absolute terms, TRQs are most important for developing countries in the fruit and vegetable, meat and sugar sectors. Developing countries make little use of TRQs in the dairy, cereal or processed food sectors. Overall, developing countries account for 60 per cent of TRQ imports and developed countries 40 per cent. These shares are very similar to the shares of these groups in total EU imports.

Table 4Usage of tariff rate quotas, by country grouping, 1997

19

This similarity of shares is coincidental, given that TRQs are opened on a commodity-specific basis and the profile of agricultural exports from developing countries is different from developed countries. TRQ exports as a proportion of total exports of the different country and commodity groupings are shown in TABLE 5. This gives an indication of the dependence of developing countries on TRQs for market access in each commodity grouping, and the scope either for increasing market access or quota rents to developing countries if TRQs were increased. The figures are interpreted as showing, for example, that of all imports of live animals and meat from developing countries in 1997, 61 per cent entered under TRQ arrangements.

Table 5Ratio of TRQ trade to total trade by country grouping, 1997

20

Overall, around 10 per cent of developing country agricultural exports to the EU entered under TRQs. This was only a slightly smaller proportion than for developed countries in 1997 although the difference is not a significant one. These figures are based on actual in-quota imports and are thus influenced by differences in fill rates between developing and developed country suppliers for bilateral quotas. There are particularly high shares of TRQs in total trade for meat imports from non-LDC ACP countries and for sugar imports from ACP and Latin American countries. These are current access quotas, opened specifically to benefit the developing country recipients. The remaining developing country exports entered the EU market under MFN tariffs (which might be zero) or benefited from preferences under GSP, Lomé or preferential trade agreements. With just 10 per cent of their exports covered by TRQs, and with some quota levels clearly binding, there is scope to benefit developing countries by a further extension of the volumes and commodities covered by TRQs. Expansion of the sugar and meat TRQs would likely lead to increased export volumes, while expansion of TRQs for fruit and vegetables, cereals and other processed foods could lead to either increased volume or increased rents, depending on whether existing trade is deflected through TRQ channels or not. Thus, the evidence from the EU market suggests that increasing TRQs could be an important means of improving market access for developing country exporters.

Would WTO rules allow preferential TRQs ?

21

One of the reasons why developing country exporters benefit from the EU’s TRQs is that a number of these were opened specifically as current access quotas to maintain pre-existing preferential access commitments. Some developing countries have suggested that reserving quota allocation for developing countries could be further extended in the allocation of increased TRQs. Other ways of providing preferential treatment under TRQs could also be envisaged, including offering preferential in-quota tariffs or requiring licensing arrangements to maximise the rent transfer to developing country exporters. This section examines the legal issues surrounding proposals to introduce a preferential element into TRQ administration.

22

TRQs were indirectly introduced by the WTO Agreement on Agriculture through Article 4 which specifies that “market access concessions contained in Schedules relate to bindings and reductions or tariffs, and to other market access commitments as specified therein” (italics added). These other market access commitments were set out in the Agreement on Modalities for the Establishment of Specific Binding Commitments under the Reform Programme (GATT, n.d.). These Modalities were never formally adopted and thus are not part of the AoA, but they remain as suggestions or guidelines for the determination and management of TRQs (Skully, 2001). They explicitly state that both minimum and current access opportunities should be introduced on a MFN basis. The question is whether there is a legal basis to discriminate in favour of developing countries in the administration of TRQs.

23

The principal reference here is the Decision on Differential and More Favourable Treatment, Reciprocity and Fuller Participation of Developing Countries (the Enabling Clause) introduced in 1979 to allow the granting of more favourable treatment to developing countries on a non-reciprocal basis. Paragraph 2(a) of the Enabling Clause permits preferential tariff treatment to products originating in developing countries in accordance with the Generalized System of Preferences. A footnote to this paragraph describes this scheme as relating to the establishment of generalized, non-reciprocal and non discriminatory preferences beneficial to developing countries. Prima facie, this waiver appears not to apply to preferential in-quota TRQ tariffs unless they were introduced within the context of GSP schemes. However, GSP preferences have well-known drawbacks. They are not negotiated but instead are unilateral. Tariff concessions are invariably subject to a safeguard clause and can be withdrawn at any time. Eligibility is often subject to additional conditions (on labour standards or observance of intellectual property protection, for example). It would be possible to envisage a separate waiver for in-quota TRQ tariffs which might be bound in Members’ Schedules, but the complexities of negotiating this on top of negotiating the concessions themselves would certainly add to the negotiating cost.

24

A further issue is whether in-quota tariff preferences could be confined to particular groups of developing countries, as proposed by the Small Island Developing States. The Enabling Clause appears to require that all developing countries benefit and that there should be nondiscrimination among beneficiaries. The one clear exception is the possibility of granting more generous preferential treatment to least-developed countries, as confirmed in the 1994 Decision on Measures in Favour of Least-Developed Countries and the 1999 Decision on Preferential Tariff Treatment for Least-Developed Countries. Preference donors have disputed this interpretation on the grounds that since the decision to grant preferences is discretionary, there should be no obligation to extend these to all developing countries. This issue may be clarified by a panel which has been initiated by Thailand where the footnote in the Enabling Clause referring to non-discrimination will be examined (WT/DS242).

25

In the case of import restrictions or tariff rate quotas, which inherently require some administrative allocation mechanism, it can be difficult to determine if the MFN principle is being applied. For this reason, Article XIII of the GATT sets out rules to govern the administration of TRQs. The basic principle set out in paragraph 2 is that contracting parties “shall aim at a distribution of trade … approaching as closely as possible the shares which the various contracting parties might be expected to obtain in the absence of such restrictions…”. Countries may operate either a global quota or allocate country-specific shares. Paragraph 2(d) sets out the rules countries should follow if TRQs are allocated to particular supplier countries. In such cases, either the importing country should seek agreement with respect to the allocation of shares in the quota with all other contracting parties having a substantial interest in supplying the product concerned, or the importing country should allocate shares to those contracting parties having a substantial interest based on the proportions supplied by those contracting parties during a previous representative period, due account being taken of any special factors which may have affected or may be affecting trade in the product. Clearly, there is much scope for argument over what special factors might have operated or may be operating in allocating supplier shares based on historical data.

26

Preferential access to increased TRQ volumes would be in breach of Article XIII. However, paragraph 2(b) of the Enabling Clause provides a waiver for differential and more favourable treatment concerning non-tariff measures of this kind governed by the provisions of instruments multilaterally negotiated under the auspices of the GATT. If there was agreement to provide preferential access to increased TRQ volumes to developing countries in a new Agreement on Agriculture, then this would appear to allow it to be covered by the Enabling Clause waiver. Again, the preferential access would seem to be required to be made available on a global basis (apart from special treatment for least-developed countries). Alternatively, if importing countries adopted an allocation method based on country-specific shares, as is allowed under Article XIII, preferential access might be interpreted as confining the countries eligible to receive allocations to developing country suppliers.

27

A third mode of preferential treatment under TRQs would be to institute arrangements designed to ensure that the quota rents created were transferred to developing countries. As pointed out in de Gorter and Sheldon (2001), WTO rules are only concerned with how quota administration influences the volume and distribution of trade and are not directly concerned with the distribution of rents. However, they also point out that rents cannot be arbitrarily distributed in ways which alter competitive conditions between countries. Under the Banana Framework Agreement (BFA), the EU required some but not all countries to issue export certificates for their country-specific TRQs. This regulation was intended to transfer part of the quota rent to the suppliers of bananas from these countries. The panel set up to adjudicate on complaints against the EU’s banana regime did not object to the use of export certificates per se, but ruled that the EU was in violation of GATT Article I which states that “… with respect to all rules and formalities in connection with importation and exportation, … any advantage, favour, privilege or immunity granted by any contracting party … shall be accorded … to the like product originating in … the territories of all other contracting parties”. This was because requiring export licences from some suppliers but not all was inherently discriminatory (WT/DS27/R). The discrimination in the BFA made distinctions between developing country suppliers. However, it might be possible to defend an arrangement which discriminated in favour of all developing countries as a preferential non-tariff measure under the Enabling Clause. Again, however, this would only be possible if the measure was included in a multilaterally-negotiated WTO agreement.

28

This discussion of the legal context in which preferential TRQs might be sought highlights the limited room for manœuvre that exists under the Enabling Clause. While the possibility of negotiating a more extensive waiver or interpretation of the Enabling Clause always exists, it would add further to the negotiating capital which would have to be expended in order to gain worthwhile concessions for developing countries under this heading. Other issues in the cost-benefit calculus are considered in the following section.

Is the expansion of TRQs worthwhile ?

29

In this section we draw a distinction between systemic criticisms of TRQs as a way of managing global trade and objections to TRQs on the grounds that the benefits to developing country recipients may be transitory, non-existent or even negative. For example, there is less benefit in terms of trade liberalisation of the TRQs to limit a TRQ to a specific country’s exports than to open a global quota to any country. Nevertheless, if a recipient country is able to increase its exports in the reserved situation relative to the global quota situation, then from the country’s point of view, the reserved quota system may be the preferred situation. Our concern in this section is whether there are specific reasons to suggest that the economic benefits of TRQ access to developing countries may be limited.

30

The growing literature on TRQ administration emphasises that the apparent benefits of TRQ access can be nullified or reduced in a number of ways (Abbott and Morse, 1999 ; de Gorter and Sheldon, 2000 ; Skully, 2001). These include rent dissipation, business uncertainty, and resource misallocation effects. TRQs have two main potential benefits : (1) to allow some trade to take place in the face of prohibitive over-quota tariffs, thereby generating a welfare surplus for the exporter on this trade ; and (2) to create rents which the exporter may hope to access as a form of trade-linked transfer. However, which agents actually capture the rents created by TRQs depends on institutional factors and market structure. Developing country exporters do capture the rents in some specific TRQs (for example, those allocated to ACP sugar exporters under the Lomé Convention, now the Cotonou Agreement) but in many cases these are captured by distribution agents or consumers in the importing country. Exporting countries may try to put in place institutions to capture these rents, such as state trading enterprises or producer associations with some degree of monopoly power. These institutions, in turn, may generate complex schemes to distribute these rents to domestic agents. As a result, the actual benefit to exporters from this form of tied aid is often much less than initially foreseen (Abbott and Morse, 1999). It is also easier to implement such arrangements where quota allocations are country-specific, and there is no requirement that preferential access to increased TRQ volumes would be administered on a country-specific rather than global quota basis.

31

The expected increase in market access may also be illusionary depending on the nature of the TRQ administrative mechanisms used to allocate quotas shared by more than one country. Under the first-come, first-served system, for example, with no specific import rights allocated to either the exporter or the importer, an exporter runs the risk of shipping the product and then finding that the quota has been filled. In this situation, the cost is borne by the exporter. Under these conditions long-term contracts and business relationships are hard to establish. Even where new markets are successfully established, they may encourage a wasteful resource misallocation in the exporting country if their exploitation depends on the continued protection of MFN tariffs in the importing market. The availability of preferential TRQ access might encourage a country to become an exporter even where it has no long-term comparative advantage in the production of that commodity. Sugar exports from the Philippines are given as an example by Abbott and Morse (1999).

32

Finally, there are arguments based on negotiating strategy. TRQ preferences by definition will be temporary as their value will be eroded by successive rounds of MFN tariff reductions. It does not appear sensible to invest huge resources into gaining what at best would be a temporary concession. There is also the danger of fragmenting the developing country negotiating effort because, inevitably, seeking to improve market access through increased TRQs, and particularly through preferential TRQs, would detract from the efforts to gain further improvements in MFN access. Thus, trade flow data suggest that there would be potential benefits from increased TRQ access for developing countries, whether or not on a preferential basis. However, economic arguments about the real value of these benefits, suggest that greater gains would ensue to developing countries if they pursue MFN tariff liberalisation rather than increasing TRQs, owing principally to the legal uncertainty about whether preferential access to, or preferential treatment under, increased TRQs could be given to developing countries under existing Enabling Clause provisions ; the costs in terms of negotiating capital in pursuing increased TRQ access, suggest that greater gains would ensue to developing countries if they pursue MFN tariff liberalisation rather than increasing TRQs.

Conclusions

33

Developing countries are proposing improved market access under TRQs in the current round of trade negotiations. The majority of proposals from developing countries call for a substantial increase in tariff rate quota volumes, as well as greater transparency in and the simplification of administrative arrangements. A number of countries have suggested that special and differential treatment could apply to TRQs, and that developing countries, or subgroups of them, might be given either preferential in-quota tariffs or preferential volume access. There is also the proposal of requiring administrative arrangements to ensure that, as far as possible, quota rents are transferred back to exporters of developing countries exporters rather than captured by importing country agents.

34

The EU is taken as a case study to examine the potential usefulness of increased TRQ access for developing countries. Over 60 per cent of the EU’s agricultural imports are currently supplied by developing countries. Potentially, TRQ imports could amount to 14 per cent of actual imports though, in practice, just over 10 per cent of imports enter under TRQs. The average fill rate (weighted by value) is 76 per cent. Sugar, dairy and fruit and vegetable TRQs have the highest fill rates, while fill rates for cereals and milling products, live animals and meat and other processed foods and drink are below average. The EU has bound a number of its preferential access commitments to developing countries as TRQ commitments. The most important cover sugar and beef (but not banana) imports from ACP countries, banana imports from Latin American suppliers (though these have been the subject of successive WTO disputes and TRQs will be eliminated when a tariff-only regime for banana imports is introduced not later than 1 January, 2006), manioc imports from Thailand and Indonesia and meat imports from Latin American suppliers. Partly as a result of these specific bilateral commitments, TRQs cover a higher value of developing country exports to the EU than for developed countries (in percentage terms, however, TRQ access is slightly less important for developing country exporters). The evidence from the EU market on the relatively low share of developing country exports accounted for by TRQ trade suggests that increasing TRQs could be an important means of improving their market access. The main beneficiaries would be Asian and Latin American suppliers not eligible for the even more favourable access terms announced for least-developed countries under the EU’s Everything But Arms initiative.

35

Having established an a priori case for the relevance of increasing TRQ access, the paper discussed the legal issues raised by the proposal for preferential TRQs. Preferential for developing countries with respect to tariffs and non-tariff measures is sanctioned by the 1979 Enabling Clause. However, a close reading of this Decision raises doubts whether the waiver from MFN treatment it contains would automatically cover either preferential in-quota tariffs, preferential access to increased quota volume or discriminatory licensing arrangements. Preferential in-quota tariffs could be justified as part of GSP preference schemes, but would suffer from the well-known drawbacks of these preferences. They are unilateral and are not bound and thus are subject to change. Furthermore, they often come associated with eligibility conditions.

36

There are also strong economic arguments as to why developing countries should be wary of putting their efforts behind improving TRQ access rather than maximising the size of MFN tariff reductions, even if the dangers of distortionary trade and rent-seeking are not necessarily those of most concern to potential beneficiaries of the system. Developing countries want better market access for their agricultural exports, and if this could be brought about easily through expanded TRQs, they might not be too concerned about the systemic impacts. What is important is whether there are self-interested arguments advising developing countries against this choice. In our view, three arguments are convincing. First, TRQ preferences by definition will be temporary as their value will be eroded by successive rounds of MFN tariff reductions. It does not appear sensible to invest huge resources into gaining what at best would be a temporary concession. Second, related to this point, there is a danger that building up markets whose value depends in part on continued protection would encourage resources in developing countries into sectors which are not sustainable in the longer-run. Third, there would be a clear danger of fragmenting the developing country negotiating effort because, inevitably, gains on TRQ access would be traded off against further improvements in MFN access. Our conclusion is that developing countries should focus primarily on seeking improved market access through MFN tariff reductions in the ongoing Special Session on Agriculture negotiations.

37

Article received on November 20, 2001

38

Accepted on February 11, 2002

References

• AMAD, Agricultural Market Access Database. Available on-line at http:// www. amad. org/ .

• Abbott, P., and Morse, B.A.., 1999. TRQ Implementation in Developing Countries. Paper presented at the “Conference Agriculture and the New Trade Agenda in the WTO 2000 Negotiations”, October 1999, Geneva, World Bank and World Trade Organisation.

• Abbott, P., 2001. Tariff Rate Quotas : Failed Market Access Instruments ? Paper read to the 77^th EAAE Seminar/NJF Seminar No. 325, August 17-18, Helsinki.

• Bureau, J.-C., and Tangermann, S., 1999. Tariff Rate Quotas in the EU. Agriculture and Resource Economics Review, 29(1).

• De Gorter, H., and Sheldon, I. (Eds.), 2000. Issues in Reforming Tariff-Rate Import Quotas in the Agreement on Agriculture in the WTO. Commissioned Paper No. 13. International Agricultural Trade Research Consortium.

• GATT, n.d.. Agreement on Modalities for the Establishment of Specific Binding Commitments under the Reform Programme, MTN. TNC/W/FA, Geneva.

• Skully, D.W., 2001. The Economics of TRQ Administration. Technical Bulletin n° 1893. Economic Research Service, US Department of Agriculture, Washington D.C.

• WTO, 2000. Trade Policy Review : European Communities, 2 volumes, Geneva, World Trade Organisation.

Notes

[1]

Corresponding author : Alan Matthews, Jean Monnet Professor of European Agricultural Policy in the Department of Economics, Trinity College Dublin, Ireland (Alan. Matthews@ tcd. ie).

Cathie Laroche Dupraz, Economist in the Département d’économie rurale et gestion, Pôle d’enseignement supérieur et de recherche agronomique de Rennes, France. The very helpful comments of Gabrielle Marceau and two anonymous referees on an earlier draft of the paper are gratefully acknowledged, but final responsibility for the arguments made and conclusions drawn remains with the authors.

[2]

WTO documents are referenced by the WTO document code in this paper and are available to download from the WTO website : www. wto. org.

[3]

This value is the sum of HS Chapters 1 to 24 less fish and fish products (Chapter 03). The Agreement on Agriculture in addition covers a small number of additional products in HS Chapters 25 and above which are not included in TABLE 1 (See Annex 1, AoA).

[4]

The problem which arises does not reflect valuation effects of imports since the discrepancy is also evident in volume terms.

[5]

Normally, fill rates are calculated for individual TRQs using data on import volumes and commodity aggregations are based on unweighted averages. The fill rates presented in TABLE 3 are similar except that individual TRQ fill rates are aggregated to the commodity group totals using the relative unit import value of each TRQ as weights.

[6]

The fill rate for cereals and milling products may be underestimated since the available TRQ is reduced by imports of maize gluten feed, brewers’ grains and citrus pulp, which was not taken into account in these indicators.

[7]

Many TRQs cover a number of HS8 tariff lines and it is not possible to state which tariff lines will be used to exploit the TRQ. The assumption made is that imports under TRQs take place in proportion to the total imports of all the HS8 tariff lines covered by each TRQ. Imports under current access TRQs are allocated to the named supplier countries, while for minimum access and non-tariffied product TRQs, it is assumed that each country grouping benefits from the TRQ according to its shares in the relevant HS8 tariff lines. The method, although the best available given the data, is crude and is unreliable for small values of trade.