1. Agriculture

The basic thrust of reform in agriculture was to establish tariff-based protection. The obligations (Box 2) were differentiated according to level of development. ^3/ In many developing countries, the nature of the agreement in agriculture is not very relevant for reducing protection of the sector. This is because it was mostly negotiated between the European Union (EU) and the United States (US) which tended to tailor it to the existing agricultural policies in industrial countries. ^1/ While the agreement clearly addresses the worst distortions in world agriculture, it leaves many agricultural distortions especially in low-income developing countries outside its scope. Many African countries do not subsidize, but tax agriculture either implicitly by giving higher protection to industry, or more explicitly by taxing exports of many commodities, or by maintaining government controlled domestic prices below world prices (Schiff-Valdes 1992). These distortions were not part of the UR agenda, and some of them are not even covered by the GATT (export taxes or domestic pricing policies that “tax” agriculture). Actual liberalization of industrial tariffs in Africa within the UR is also modest (see below) and unlikely to reduce the existing bias against agriculture. African trade policies have also suffered from frequent policy reversals and from the impact of exchange restrictions on trade flows. Some of these policies have been addressed in the countries’ structural adjustment programs over the past decade (World Bank 1994a, Schadler-Rozwadowski-Tiwari-Robinson (1993)).

Tariffication. All African participants that have submitted schedules bound 100 percent of their tariff lines in agriculture. However, the level of the bindings by many countries is at prohibitive levels (100 percent - 300 percent) (Table 1). The only “reasonable” bindings (less than 50 percent) were by Congo (30 percent ceiling binding), Central African Republic (ceiling binding at 46 percent), and by four of the five countries of the Southern African Customs Union (Botswana, Namibia, South Africa, Swaziland). In the latter, bound duties on average will decline by 43 percent from about 70 percent to 40 percent over a six-year period. ^2/ Few of the developing countries that are not in the group of least developed countries offered to make reductions from their ceiling bindings, although this was required by the agreement. ^3/ Only Cote d’Ivoire, Ghana, and Zimbabwe offered minor reductions on a few items.

The bindings consist of a customs duty and other duties and charges (ODC). As mentioned previously, the ODC rates were not listed in a country’s schedule in GATT. Not all countries added ODC rates to their schedules, although the practice seems particularly frequent in Africa. ^l/ In some countries that had old bindings, such as Cote d’Ivoire, reduction commitments in the UR were made on the customs duty (old binding), to which the now “confirmed” ODC rate was added. For example, in Cote d’Ivoire a 7 percent bound duty on fresh milk (HS 040110) will be reduced to 6 percent over a ten-year period, but a 200 percent ODC will be added to this item. The end result will be a substantial increase in bound rates on these items. ^2/

The high levels of bindings resulting from the UR may have a number of unexpected effects on protection of agriculture in Africa and amount to a lost opportunity to liberalize. First, the results indicate that the UR will not liberalize agriculture in Africa in the sense of reducing actual tariff rates. It is likely (and to be hoped) that most countries will not apply the high bound rates. Available data indicate that applied rates are in general lower (Table 1). The policy implications of this are that rationalization of protection in agriculture in SSA will have to rely on continued unilateral liberalization.

Second, the high bindings do not impose an effective constraint on policy reversals. ^3/ The high level of the bindings can by itself undermine their stability objective and the transparency objective of tariffication in general. Despite the relatively uniform rates used in the bindings by many countries, applied rates within these bands can be highly dispersed. In principle, countries can continue to change tariffs within the margin provided by the binding. The likelihood of this depends on whether the recent trend of unilateral liberalization continues. Most of the maximum rates are too high to provide a meaningful cap on rates to improve the security of market access.

Third, the high level of bindings is not likely to serve the objective of removal of all non-tariff barriers in agriculture. Countries can charge duties at varying levels within the margin of the binding, which could be linked to a domestic threshold price. The duty can be set as the difference of the given domestic price and the world price as long as the duty charged does not exceed the binding. ^1/ In practice this can work like a variable levy or a minimum price. However, the levies might be challenged as such under the agreement. A footnote to the agriculture agreement states that variable levies etc. are forbidden and will be converted to tariffs. Furthermore, the Agreement in Customs Valuation forbids the use of minimum prices, except for developing countries that make a special reservation for it. Clarification of the issues may require recourse to the dispute settlement proceedings. The peace clause ^2/ may prevent this for some years to come, and the fact that EU’s partners did not challenge this during the verification and negotiation process suggests that the setting of variable levies within the bound ceilings might be accepted.

Subsidies. Except for South Africa, the commitments on domestic agricultural policies or on export subsidies required by SSA in the UR are likely to have little if no impact on their domestic agricultural policies. Only South Africa made any commitments to reduce either domestic or export subsidies (Table 1). ^3/ The main reason is that the others do not subsidize agriculture or its exports, and that least developed countries were exempt from the commitments. As mentioned many traditional export commodities are taxed, not subsidized, in many SSA countries. In some countries, such as Cote d’Ivoire or Senegal, earlier export subsidy programs covered mostly manufactures and in any case have been terminated for lack of funds. Only Cote d’Ivoire, Mali and Senegal have submitted lists of domestic agricultural programs that they consider permissible, i.e., not subject to cuts, and that may have some subsidy elements. The rules require South Africa to reduce domestic agricultural subsidies by 20 percent over a six year period from a 1986–88 base, which South Africa has already done in its reform program.

Minimum access. Apart from South-Africa, SSA countries have made no minimum access commitments. ^1/ The agreement requires this only from countries that did not make ceiling bindings. ^2/ South Africa has listed amounts designated to minimum access in its schedule, but the agreement leaves much discretion on their implementation. No indication is given on how the quotas would be allocated. ^3/

In summary, in agriculture the commitments made by and required of SSA developing countries in the UR do not provide much evidence that these would seriously limit their present options on domestic policies or that they have undertaken liberalization commitments. This is because (i) the UR rules will not in themselves affect most agricultural distortions in Africa, (ii) commitments made by most of SSA will not result in more liberal policies or a more stable policy framework, and (iii) the general exemptions in the GATT or those for the balance of payments in Article XVIII, albeit under slightly more restrictive conditions, are still available for developing countries to justify existing policies.

2. Industry

In the UR there was no uniform rule for making tariff commitments apart from having to achieve the stated goal of a 33 percent reduction in tariffs over a five- to ten-year period. This made actual cuts dependent on the results of bilateral negotiations, which tended to bias them in favor of cuts on products of interest to main industrial country partners. Developing countries were encouraged to make either ceiling bindings, as many had no previous bindings (Table 1), or reductions in previously bound items. As in agriculture, the ODC rates were to be included in new schedules (see above).

As most of the agreed liberalization commitments in non-tariff measures (NTMs) were designed to reduce the types of policies prevalent in industrial countries, their relevance for developing countries is reduced. The main obligations from the Round in industrial NTMs are that all countries are (i) to remove voluntary export restraints within four years, ^4/ and (ii) to eliminate all GATT-inconsistent restrictions on textiles and clothing imports within a ten-year phase-in period. ^5/ Should developing countries maintain quantitative restrictions in textiles and clothing, they have to be removed within the ten year transition period (unless justified by other GATT 1994 provisions such as Article XVIII). All restrictions have to be notified to the relevant WTO body at the start of the transition period, even if the measure has a GATT 1994 justification. ^1/ Unless developing countries notify existing restrictions (and a timetable for their removal), in principle, they are assumed to maintain no GATT-inconsistent restrictions in the sector.

Existing tariff and non-tariff barriers in many African countries on industrial goods vary substantially and accurate information on existing policies is at times difficult to get. Policies in a sample of SSA countries ^2/ reviewed in the GATT’s Trade Policy Review mechanism indicate that, apart from tariffs and fiscal duties, applied border measures include an array of bans and prohibitions on imports and exports, licensing, exchange rate rationing, quantitative import restraints, and minimum prices. In many countries the structure of the tariff is highly dispersed as high duties coexist with numerous exemptions. A recent World Bank study on African trade policies underlined the importance of exchange controls in restricting imports and exports in many SSA countries (Nash 1993). With overvalued exchange rates foreign exchange was often allocated to imports with licenses amounting to an implicit import subsidy; exports were taxed with various surrender requirements at official rates below market rates. In this environment the impact of border measures such as tariffs on trade flows may be minimal. Over the past decade many of these issues have been addressed by adjustment programs supported by multilateral donor agencies resulting in some improvements (World Bank 1994a, IMF 1994). However, reforms have suffered from credibility problems due to frequent reversals ^3/ and slow implementation.

Tariffs. As many SSA countries did not have previous bindings, they made mostly ceiling bindings, but only on a handful of tariff lines (Table 1). Twelve countries made bindings on a large number of items (50 to 100 percent of tariff lines), while the majority bound less than 2 percent of tariff lines during the UR. Only eleven countries will have bindings on over 90 percent of tariff lines after the Round (Djibouti, Guinea Bissau, Niger, Rwanda, Sierra Leone, Zaire, and the five countries of the Southern African Customs Union). Only four members of the Southern African Customs Union and Cote d’Ivoire offered to reduce previously bound rates. ^1/

In many cases the resulting bound rates are prohibitive at rates over 100 percent. Only in ten countries were the few duties bound at levels below 50 percent. This includes the four of the five countries of the Southern African Customs Union, in which bound duties will decline to an average of 17 percent. ^2/ As applied rates, at least in countries for which data are available, are in general lower, the bound rates may not be applied in practice and will only serve as upper bounds.

Apart from the countries of the Southern African Customs Union, tariff commitments on industrial goods by SSA do not represent meaningful liberalization and are unlikely to address any of the key distortions in SSA trade policies. First, as with agriculture, the high level of bindings can undermine their impact on security of market access, as duties can readily be changed within the maximum bounds. Security is further reduced by the small number of lines bound. But even high bindings can be better than no bindings. Second, credibility of policies is unlikely to improve. SSA countries have not taken advantage of the opportunity offered by the Round to bind domestic reforms to an international anchor to improve credibility of reform. Third, the UR in most cases will not change existing levels of protection as applied rates will not change in most SSA. Therefore, the UR will do little to promote, for example, regional trade.

Non-tariff measures. In general the UR results seem to put few new obligations on developing countries to reduce existing non-tariff measures. A review of submitted schedules indicates that so far in textiles the only commitment in SSA is by the three of the five countries of the Southern African Customs Union, which have agreed to reduce barriers (high specific duties in local currency) in textiles and clothing imports over a ten year period. The only other commitments in NTMs were made by Senegal and Cameroon. Senegal offered to remove prior authorization requirements on a number of tariff lines (paper products and some machinery). Cameroon agreed to eliminate import quotas on a number of products (paper products and miscellaneous items).

The agreements on Balance of Payments (BOP) and Customs Valuation may have some impact on existing non-tariff barriers in developing countries. However, the agreement on BOP mostly reinforces earlier recommendations for greater transparency by public announcements of plans to terminate the measures and for preference for the use of price-based measures over quantitative restrictions. Apart from some tightening, the agreement does not create new obligations to reduce existing measures or the use of BOP restrictions in the future.

The Agreement on Customs Valuation in the WTO (Table 3), in principle, requires countries to eliminate official minimum prices at the border as customs valuation is to be based on the price actually paid or payable for the imported good. ^1/ The single undertaking means that all countries will have to adopt these rules. However, developing countries, that so notify the Director General of the WTO, have five years to comply with this agreement. ^2/ The adoption of the rules can be further postponed by a further request to the WTO. In practice, this means that, although developing countries have undertaken the new obligations, the long transition periods and potential for their extension require little changes in policies in the near future. Implementation of the new obligations therefore depends on the will of trading partners to grant exemptions or on the countries themselves to implement the obligations.

View Full Size

• Download Figure
• Download figure as PowerPoint slide

If applied, the new rules on Customs Valuation may imply changes to existing laws in SSA. At present many of them use the so-called Brussels Definition of Value at the border. ^3/ Compliance may imply more stringent administrative requirements, but the removal of minimum prices could increase efficiency, and the use of an international standard for valuation could imply lower trading costs for importers by reducing discretion in valuation at the border, and hence reducing rent-seeking opportunities.

A main impact of the UR agreements on NTMs in industry for developing countries in SSA is likely to be an increase in the transparency of policies. This will arise from the increase in notification requirements (textile restrictions, subsidies, TRIMS, see below). The conduct of trade policy reviews will also add to the transparency of policies. This is reinforced by the general trend in recent years, for example, to require developing countries to list and justify more carefully all measures maintained under the BOP exceptions. In the past, many developing countries have not notified their non-tariff measures to the GATT as required and their trading partners have not actively challenged the GATT consistency of applied policies. For example, only South Africa and Nigeria of the SSA countries have followed GATT procedures and notified their import restrictions and held consultations in the GATT BOP Committee. The increased pressure for transparency may be less for the very small and poor countries with a minor role in international trade, but may effect the larger traders of SSA.

In summary, the UR results will not liberalize industrial (tariff) protection in most of SSA. Furthermore, in the near future, long transition periods and potential to extend (in many cases indefinitely) the adoption of the new obligations, coupled with continued application of old exemptions in the GATT such as Article XVIII:B, will mean that the UR will set few new constraints on existing border protection (NTMs) in SSA. The case of South Africa and other members of the Southern African Customs Union is different and stands out from the rest of the SSA: the Union has made substantive commitments in terms of increased bindings and cuts in existing bindings, which are close to applied rates. Within the Union, South Africa used the UR to bind domestic unilateral reform which can work as a powerful argument against domestic lobbies resisting change. South Africa also made its commitments according to the rules for industrial countries. For South Africa the UR coincides with a domestic process of unilateral liberalization.

3. Services

All developing countries were required to make market access commitments (liberalization) in services and accept the framework agreement (the rules). The General Agreement on Services (GATS) obliges countries to give most-favored-nation (MFN) treatment to foreign service providers in sectors and modes of supply, ^1/ in which liberalization commitments have been made. A one-time exemption from MFN treatment could be made at the time of making the market access commitments. ^2/ Services was one area, which developing countries initially were reluctant to include in the negotiations. Many felt that they had no comparative advantage in services or that liberalization of this sector would be of little relevance for their development strategies. But as the Round progressed developing countries started to see a new potential in the sector both directly for their exports, and indirectly as liberalization of services may have an important impact in improving the competitiveness of many export industries. Openness and competition are likely to improve efficiency in local provision of services, which often are important inputs to manufacturing or other economic activities (banking, transport).

Very little information is available on the structure of protection in the services sector in SSA. A large part of the sector is likely to be dominated by state enterprises with privileged market positions or government monopolies often raising prices of services. Government-provided services are not covered by GATS. Many international transport routes are also subject to cargo sharing arrangements, which have a tendency to maintain the price of freight at high levels. Provision of professional services is often subject to tied aid to a specific country, which limits competition on potential suppliers. Liberalization in the sector could bring substantial efficiency gains, although in Africa competition in transport or other sectors is also restricted by many natural factors such as small size of the markets.

The nature of the liberalization commitments in the UR vary substantially among the thirty-seven SSA countries that had submitted their schedules by May 1995. The extent of the liberalization in the submissions is difficult to estimate for the complex nature of the market openings in the different modes of delivery. One rough measure can be the number of sectors in which commitments were made. According to this measure the range of sectors opened varied from one in seven countries (Chad, Burkina Faso, Madagascar, Mauritania, Mozambique, Tanzania, Uganda) to 12 in one country (Gambia) of the 12 sectors and 37 countries covered (Table 2). The sectoral breakdown indicates that most commitments were in tourism (34 countries), business services (18 countries) and communication (13 countries). Twelve countries had also made reservations to the MFN obligation mostly in maritime transport. Commitments in financial, business, and communication services, which are important “input sectors” to exports, were made by less than half of the countries. This can reduce the potential for efficiency gains or transfer of technology in many of these countries in sectors that are inputs in export production. Most of the commitments are likely to reflect the status quo with little additional liberalization.