CHAPTER 4. Policy Implications

Sanjeev Gupta, and Yongzheng Yang
Published Date:
September 2005
  • ShareShare
Show Summary Details

The generally disappointing record of African RTAs warrants a reexamination of the underlying assumptions. Is the record disappointing because no preconditions were established to make the RTAs beneficial, dooming them to failure before they even began? Or is this record the result of poor design and/or implementation? What can we learn from the successes of regional trade integration in other parts of the world as well as from Africa’s failures?

The evidence presented in the literature suggests that most African RTAs lacked initial conditions for success because of limited intraregional trade, weak complementarity in resource endowments, and inadequate transport infrastructure and local capacity. The design is generally poor, particularly with respect to external tariffs, nontariff barriers, and trade facilitation—although in some cases (e.g., WAEMU) the early involvement of international financial institutions and some donor governments helped improve the design. For virtually all African RTAs, implementation has been weak, often with delays, which partly results from overlapping membership. Thus, to improve the performance of African RTAs, a broad approach is required to tackle a whole range of design and implementation problems as well as to create necessary preconditions for successful regional trade integration.25 Overall, the following policy messages have emerged from the literature:

  • African countries should reduce their trade barriers against non-RTA members when pursuing RTAs. As Jebuni (1997) points out, MFN liberalization, either unilaterally or multilaterally, is even more important in the presence of RTAs. Lower external trade barriers reduce the risk of trade and investment diversion. Compared with unilateral liberalization, multilateral liberalization can mitigate the potential terms of trade effects on liberalizing countries if they are large economies. Because African economies are small and hence have no market power, unilateral tariff reductions will be welfare-improving as long as supporting policies are put in place to ensure that resources dislocated by liberalization are reemployed. Tariff reductions will also reduce the tariff-induced bias against exports. Subramanian and others (2000) find that, for Africa, a 1 percentage point reduction in trade taxes leads to an increase in trade of between 0.7 and 1.1 percentage points.More generally, MFN liberalization is found to be more conducive for growth and trade than discriminatory liberalization. Using time-series data for 1950–92, Vamvakidis (1998) shows that economies grew faster after broad-based liberalization, in both the short and long run, but slower after participating in an RTA. He also finds that economies had higher investment shares after broad-based liberalization, but lower shares after joining an RTA. A recent World Bank (2004) study finds that MFN liberalization tends to increase intraregional trade most rapidly and efficiently. Nowhere is this more evident than in East Asia, where there have been very few RTAs until recently, and yet intraregional trade has exploded following broad-based, nondiscriminatory trade liberalization in the region (Box 1). With the rapid growth of trade comes a finer division of labor within the region and intraindustry trade, which, in turn, reinforces regional trade ties. There are some encouraging signs in Africa, too. For example, when the WAEMU customs union was established in 2000, the average external tariff in its member countries was reduced substantially, fostering both extra- and intraregional trade (World Bank, 2003a).The success of the European Union, notwithstanding its continued high protection of agriculture, is largely due to the steady reduction of its external trade barriers against non-EU members under successive rounds of GATT/WTO trade liberalization (Box 1). The experience of EU trade integration suggests that MFN liberalization can be effectively pursued by RTA members despite likely resistance to liberalization from vested interests. To overcome such resistance, policymakers in RTA members must show a strong commitment to effective implementation of reforms, as demonstrated by the EU members over the past decades. It is only through sustained efforts and a willingness to be bound by both multilateral and regional commitments that RTAs can foster intraregional as well as extraregional trade.
  • African countries should avoid the pitfall of import substitution in their RTAs. One of the key reasons that many African RTAs have failed to promote trade is their inward orientation, as seen in an import substitution policy.26 RTAs have often been designed to increase incentives to export to regional markets at the expense of exports to the world market (Foroutan, 1993). In addition, since local firms have to pay external duties, they are unable to access the most efficient sources of supplies and production equipment, thereby reducing Africa’s ability to compete internationally.
  • Most African RTAs, even if successfully implemented, would have only limited potential to expand intraregional trade because countries in the region can meet only a very small share of regional import demands (Yeats, 1998; African Development Bank, 2000). For example, machinery and transport equipment account for approximately three-fourths of total (global) African imports, but they account for less than 4 percent of intra-African trade (Yeats, 1998). Similarly, Khandelwal (2004) finds that with low product complementarity among African countries, even three of the largest, most diversified economies in the region—Egypt, Kenya, and South Africa—might not function as growth poles in COMESA and SADC. While low intra-African trade flows point to great growth potential in the long run, it also suggests that the greatest boost to African trade in the short to medium term must come from policies promoting trade with the rest of the world. The African market is too small to sustain high export growth.
  • African countries need to strengthen their domestic supply response to take advantage of unprecedented opportunities to export to world markets. African countries typically face very low protection in industrial countries, either because MFN tariffs (except on agricultural products) are already low (about 2–6 percent on manufactures and 16–49 percent on agricultural products) or because extensive preference is granted to African goods under the GSP, the EBA, AGOA, and Japan and Canada’s zero-rated tariff schemes.27 Any of these markets is much larger than Africa, and, when combined, they provide far greater potential to exploit economies of scale.28 To increase domestic supply response, African countries need to continue to undertake structural reforms as well as MFN trade liberalization. They must also improve infrastructure and upgrade workers’ skills.
  • To exploit economies of scale and enhance domestic competition through RTAs, African countries need to reduce transport costs within the region.29 It is difficult to take advantage of economies of scale in Africa when shipping a car from Japan to Abidjan, Côte d’Ivoire, costs $1,500 while the same car costs $5,000 to ship from Addis Ababa, Ethiopia, to Abidjan (ECA, 2004). The high transport costs and other barriers to intra-African trade are reflected in the fact that such trade is often “cross-border” and that “geographically distant” African countries trade very little with each other despite extensive regional trade arrangements (Yeats, 1998). Hummel (2001) finds that, in general, each additional day a shipment is in transit is equivalent to an extra 0.8 percentage point increase in applied tariffs. For historical reasons, infrastructure in Africa has been oriented toward facilitating trade with former colonial powers in Europe. To facilitate regional integration, Africa needs to devote resources to regional infrastructure. Such investment is also necessary to enhance domestic competition. With a more integrated regional market, MFN liberalization could then effectively substitute for a competition policy, given the small size of individual African markets.
  • African countries should focus on cross-border, sectoral cooperation in areas of common interest. Customs administration, which is weak in most African countries, impedes trade within the region as well as with the rest of the world. Crossing a border in Africa can be equivalent to the cost of more than 1,000 miles of inland transportation, while in Europe the cost is equivalent to 100 miles. According to the World Bank (2004), delays at the main border crossing between South Africa and Zimbabwe (Beit Bridge) amounted to six days in February 2003, leading to an estimated loss of earnings per vehicle of $1,750—equivalent to the costs of a shipment from Durban to the United States. African countries could also cooperate on a range of other areas, such as energy, water resources, research and education, environment management, and regional conflict prevention and resolution. The relative success of SADC is perhaps partly due to its efforts to facilitate regional cooperation in various projects that bring tangible benefits to member countries.30 Unlike the promotion of intraregional trade through preferential trade agreements, cooperation in such projects does not lead to trade diversion.
  • African countries should participate more actively in multilateral trade liberalization. Numerous studies indicate that Africa stands to gain substantially from multilateral trade liberalization. The World Bank (2004), for example, estimates that Africa would gain as much as $24 billion from global merchandise trade reform. Such benefits do not come without any costs, of course. Both Africa and other parts of the world would have to reduce their trade barriers and undertake structural adjustment. The Uruguay Round was a missed opportunity for Africa. Only a little over 2 percent of tariff lines in Africa are bound, often at levels well above the applied rates (Oyejide, 1997). Binding African tariffs at levels close to the applied rates would increase the credibility of Africa’s trade policy. A recent study by Kowalski (2004) finds that binding tariffs closer to the applied rate could significantly increase trade.The current Doha Round of multilateral trade negotiations can bring substantial benefits for African countries. The long-term dynamic gains from a successful Doha Round are likely to be much larger than typical estimates (such as the number cited above) from general equilibrium models. In agriculture, for example, eliminating trade distortions (export and domestic subsidies and high tariffs) around the world could boost world agricultural prices by 12 percent according to the U.S. Department of Agriculture (2001), thus increasing the returns on agricultural investment and productivity growth in Africa.31 Reductions in industrial tariffs in Africa would increase domestic competition and provide much needed incentives for exports in Africa. Liberalization of trade in services in Africa would reduce the production cost for other sectors, as well as increase competition and hence efficiency of the domestic services sector. Many of these and other gains (such as those from improved trade facilitation, which is also part of the Doha Round agenda) are difficult to quantify using standard general equilibrium models. However, if the prosperity of some developing economies (e.g., Hong Kong SAR, Korea, Singapore, and Taiwan Province of China) over the past decades is any indication that multilateral liberalization can help poor countries achieve, Africa has a much higher stake in a successful Doha Round than a mercantilist approach to trade liberalization would suggest.
  • African countries should take early action to bolster domestic tax mobilization. As noted earlier, trade taxes remain an important source of government revenue in most African countries despite tariff reductions in recent years. At an average tariff level of 17 percent at present, any further tariff cut can result in revenue losses. The need for early mobilization of domestic taxes has become more urgent because of the ongoing negotiations on multilateral trade liberalization in the WTO and on the FTAs with the EU. For countries that plan to undertake unilateral liberalization, domestic tax mobilization should be part of the reform program to safeguard government revenue.
  • Preferential trade arrangements between African countries and outside countries are likely to bring greater benefits. Although such arrangements are inferior to MFN liberalization, as noted earlier, they provide several advantages over RTAs among African countries alone and could serve as a second-best policy when MFN liberalization is not feasible. Empirical studies show that trading partners matter. Trade ties with fast-growing partners bring more benefits, and trade with high-income countries is also more beneficial (Arora and Vamvakidis, 2004). A key weakness of RTAs in Africa is that the region as a whole has been growing more slowly than the rest of the world. Preferential trade agreements with industrial countries could help reduce policy reversals by acting as an “agency of restraint” when commitments under the multilateral system do not exist (Collier, 1991). Furthermore, trade agreements with industrial countries are more likely to have broader product and policy coverage. Properly balanced, such coverage—such as trade in services, trade facilitation, investment protection, institution building, and regulatory reforms—can yield more benefits. But there is the risk that trade agreements with industrial countries are skewed toward areas that are more in the interest of industrial countries than of developing countries. For example, financial services are often covered in North-South agreements, whereas temporary movement of labor is not. Also, such agreements often introduce labor and environmental standards that may not be appropriate for Africa, given its current level of economic development.
  • African countries should streamline RTAs with overlapping membership. Most African countries have multiple commitments arising from different agreements, and the commitments are not always consistent. For example, some Eastern African countries (Burundi, Kenya, and Uganda) are involved in two planned customs unions (COMESA and EAC). As long as the CETs are different (or even if their final levels are the same but have different schedules of transition), these countries would face the impossible task of implementing multiple tariff commitments. In the case of FTAs, the implementation is further complicated by the different ROOs that these countries have to administer. In addition to the administrative costs, these rules give customs officials more discretion and are therefore vulnerable to rent seeking by domestic interest groups. Despite the recognition of these problems, little progress has been made in addressing them.32 The current negotiations on FTAs with the EU represent an opportunity to begin streamlining the arrangements. For example, some countries could reduce their multiple memberships to a single one; small and unsuccessful RTAs could be absorbed by the large ones that have been designated to represent groups of African countries in negotiating FTAs with the EU. Given the number of RTAs involved and the rivalries between them, any such process will be a long and difficult one.33

Box 1.Regional Trade Integration With and Without RTAs: The EU and East Asia

Regional trade integration in the EU and East Asia represents two success stories. The two regions have, however, taken different paths to success: EU integration has been driven by formal institutional arrangements, whereas East Asian integration has been a result of “natural” market forces. But both stories highlight the importance of initial economic conditions and sustained MFN reductions on external trade barriers.

The European Union, successor to the European Economic Community (EEC) created in 1957, has consistently pursued not only the elimination of trade barriers between its member countries, but also the reduction of barriers against nonmembers. Before the EEC was established, tariffs in its original members were high, and nontariff barriers were prevalent. Through successive rounds of multilateral trade liberalization under the auspices of the GATT and its successor, the WTO, the average MFN tariff on manufactures has been brought down to about 4 percent, although tariffs on agriculture remain high (22 percent). In 1957, intra-EEC trade was already about 30 percent of its total trade. The formation of the EEC led to sharp increases in intraregional trade, and, by the early 1970s, intra-EEC trade had reached 60 percent of total EEC trade. While intraregional trade has increased, trade with the rest of the world has also increased, albeit less rapidly. Beginning with trade, the EU has successfully moved to deeper economic and political integration and extended its membership over time. Successful economic integration has also contributed to greater regional stability.

Formal regional trade schemes are a recent phenomenon only in East Asia,1 which has pursued trade liberalization largely on an MFN basis. The region’s first integration arrangement, ASEAN, established in 1967, largely for the purpose of regional security, was designed only to facilitate trade. Nevertheless, with the shift from an inward-looking to an outward-looking growth strategy in the late 1950s (Japan began after World War II and China in the late 1970s), the region has consistently opened its markets to the rest of the world. Trade has since expanded rapidly. In 1956, the year for which data are available for all countries except China and Malaysia, total East Asian exports were only 4.6 percent of world exports, of which 2.6 percent was contributed by Japan. By 2003, the region accounted for more than 23 percent of world exports. At the same time, the share of intra–East Asian exports in the region’s total exports rose from 23 percent to 47 percent. Rapid export growth to industrial country markets has generated demand for imports within the region, and income and resource diversity among the countries has enabled them to specialize according to their global comparative advantage. Thus, East Asia’s trade integration has succeeded without much trade diversion.

1 For the discussion here, East Asia includes China, Hong Kong SAR, Indonesia, Japan, Republic of Korea, Malaysia, the Philippines, Singapore, Taiwan Province of China, and Thailand.

The approach proposed here is similar to that advocated by the World Bank in the Regional Assistance Strategy papers—for example, World Bank (2001, 2003b).


If RTAs are designed to provide local industries with a learning-by-doing period before exposing them to full-scale international competition, a clear, finite transition period could be set to avoid prolonged protection.


The preferential market access is subject to ROOs, which reduce African exporters’ ability to take full advantage of the preferential treatment. In addition, technical standards and sanitary and phytosanitary measures in industrial countries may impose substantial compliance costs on African exporters.


GDP in sub-Saharan Africa is about one-tenth of that in Japan. In 2000, the EU, the United States, China, and Japan accounted for more than three-fourths of total African exports.


There is an issue whether African countries should give priority to investment in infrastructure that reduces transport cost for intraregional trade or in infrastructure that helps extraregional trade. On efficiency grounds, the decision should be based on a cost-benefit analysis. However, if the promotion of intraregional trade is the objective, then there is a second-best argument for transport investment in favor of intraregional trade.


Sectoral cooperation in Africa has its own shortcomings, as documented in Kritzinger-van Niekerk and Moreira (2002). Ineffective coordination of sectoral activities and institutions could lead to failures to take advantage of synergies of various sectoral activities, resulting in slow regional integration.


Some net food-importing African countries could, of course, suffer terms of trade losses.


For some countries, multiple memberships reflect their desire to pick and choose various options offered by competing RTAs—in terms of the level of integration and the extent of commitments (e.g., the exchange rate regime).


According to the current plan, four regional EPA negotiating groups will be formed in Africa to negotiate FTAs with the EU (Hinkle and Schiff, 2004). ECOWAS has been designated the negotiating bloc for Western Africa; The CEAMC plus São Tomé and Príncipe grouping has been chosen by the Central African states; 16 countries, including most of the COMESA members, will form an Eastern and Southern Africa group; and SACU members and three neighboring countries are forming a Southern African negotiating group.

    Other Resources Citing This Publication