The discussion so far highlights the need to take a broad perspective in thinking about how to make African RTAs work better. Careful design and sustained implementation are necessary to make any RTA deliver, but the more fundamental determinants of RTA performance seem to be policies and conditions that affect the overall environment for trade. In this context, MFN liberalization, improvements in regional infrastructure (particularly transport), and reductions in trading costs at borders are critical. These are all conventional development issues but nonetheless pose difficult challenges for African policymakers. African countries need to protect their revenue base in undertaking MFN tariff reductions, and, when such tariff cuts are implemented in their trading partners, they need to address the consequence of preference erosion.34 Efforts to improve infrastructure and reduce trading costs will require adequate local capacity of implementation, in addition to financial resources. In all these areas, the IMF can lend its support to African countries, beyond its core expertise area of macroeconomic management, which is also essential to create a conducive environment for trade.
A. Establishing Sound Macroeconomic Frameworks
The IMF plays a central role in helping its African members promote trade by establishing sound macroeconomic frameworks. Macroeconomic stability is a key precondition for trade growth. Low inflation and financial stability create an enabling environment for trade growth. Prudent fiscal policies not only help keep inflation low, but also leave more resources to the private sector to increase their productive capacity. Appropriate exchange rates help maintain external balance and competitiveness. They also help reduce resistance to trade liberalization. African experience has shown that overvalued exchange rates often lead to pressure for policy reversals even if reforms have been undertaken in the first place (O’Connell, 1997). Many African countries have made progress toward macroeconomic stability in recent years. Inflation has moderated and public finance is on a more sustainable path. Many countries have a stronger balance of payments position thanks to more flexible exchange rates and higher foreign reserves. Still, much more needs to be done.
B. Managing Revenue Losses from Trade Reform
Trade liberalization does not necessarily lead to revenue losses. In the initial stages of liberalization, revenues may even increase, especially when certain nontariff barriers are converted into tariffs, high tariffs are lowered, and certain exemptions on the payment of duties are eliminated (Ebrill, Stotsky, and Gropp, 1999).35 In the subsequent stages of liberalization, it is possible to recover revenue losses by strengthening domestic indirect taxes, provided that the country has the administrative capacity to do so (Keen and Ligthart, 2001).36 To some extent, revenues from domestic taxes are likely to increase in the aftermath of trade liberalization because of higher domestic consumption arising from lower prices of tradables and because of higher growth resulting from the improved allocation of resources in the economy. However, the size and timing of such dynamic effects on revenues are highly uncertain. The experience in Africa shows that only half of the countries where trade tax revenue declined from the early 1980s to the late 1990s managed to recover tax revenues from domestic sources. In general, empirical research shows that revenue losses from trade liberalization vary considerably across countries (Box 2).
Empirical research suggests that appropriate supporting policies mitigate trade revenue loss when it occurs. First, trade liberalization will need to be accompanied by a strengthening of indirect taxes. A two-year period has often been found sufficient for introducing an effective VAT in countries where this tax does not exist. However, for countries where tax administration is weak, like many in Africa, it may take up to five years to introduce an effective VAT. Countries that already have a VAT could moderate revenue losses by improving its design. Second, strengthened income taxes on both individuals and companies can aid in capturing revenues from rising incomes. Third, attention will need to be paid to strengthening the administration of the wider tax system—for example, by establishing large-taxpayer units. Finally, the adoption of sound macroeconomic policies promotes growth, expands the tax base, and, ultimately, raises government revenues. In all these aspects, the IMF has provided, and should continue to provide, assistance to countries in need of mobilizing more government revenue.
Trade Liberalization and Revenue Losses
The empirical evidence on the revenue effects of RTAs in sub-Saharan Africa is mixed. The World Bank estimates (cited in ECA, 2004, p. 101) that Uganda could lose 9 percent of total government revenue in 1998 as a result of COMESA liberalization. An IMF study of the free trade area between 11 (out of a total of 19) COMESA member states1 shows that the elimination of intra-COMESA trade flows would not pose significant threat to revenues.2 Only in the case of Rwanda would the direct loss of import duty be about 4 percent of total revenues. For Uganda, Tanzania, and Mauritius, it would be between 1 and 2 percent of total revenues; in all other countries, it would be below 1 percent. Trade diversion effects could magnify these results in Rwanda, Uganda, and Tanzania, whose imports from Kenya could increase somewhat at the expense of imports from non-COMESA countries.
How well countries can manage revenue losses from trade liberalization depends on the macroeconomic policies they follow and their level of economic development. One recent study (Agbeyegbe, Stotsky, and WoldeMariam, 2004) shows that trade liberalization in sub-Saharan Africa may not have a significant effect on overall revenues if it is accompanied by sound monetary and exchange rate policies. Following trade liberalization, a depreciation of the exchange rate often increases overall tax revenue, with the impact varying across different types of taxes. Another study (Baunsgaard and Keen, 2005), however, provides evidence that is more troubling for low-income countries. Using panel data for 125 countries during 1975–2000, it concludes that following the loss of trade revenues, high-income countries recovered them with ease, middle-income countries recovered only 35–55 percent of revenues, and low-income countries recovered none. The authors were unable to find evidence to support the notion that the presence of a VAT makes it easier to offset revenue losses from trade liberalization. Moreover, Keen and Simone (2004) find that international competition in tax incentives (including those offered to investors in free trade zones) has reduced both the corporate tax rate and base, making it more difficult to recover revenue losses from trade liberalization.
1 The countries studied were Kenya, Madagascar, Malawi, Mauritius, Namibia, Rwanda, Swaziland, Tanzania, Uganda, Zambia, and Zimbabwe. 2 These estimates focused only on changes in import duties and therefore ignored the implications of the tariff reduction for excises, VAT, or sales tax. They were derived on the assumption of unchanged behavior.C. Addressing the Impact of Preference Erosion
African countries have expressed concern about preference erosion in the context of multilateral liberalization. As noted earlier, African countries currently enjoy extensive preferential market access in industrial countries and some developing countries. MFN liberalization in Africa’s trading partners would reduce its preference margins in these markets and, hence, reduce the competitiveness of African exports. Empirical research indicates that the impact on exports of preference erosion from multilateral liberalization is generally small and highly concentrated in a small number of countries and commodities (Box 3). Preference erosion can also result from the further proliferation of regional trade arrangements. For instance, the EU-Common Market of the South (Mercosur) FTA that is currently under negotiation could reduce the tariff advantage that African countries enjoy in the EU under the EBA. Similarly, a continued U.S. push for FTAs in the Americas, the Middle East, and Asia could reduce the preference margin for African exports in the U.S. market.
The IMF has taken steps to address the problem of preference erosion. In April this year, the Executive Board of the IMF approved the Trade Integration Mechanism (TIM). It is designed to finance balance of payments shortfalls as a result of textile shocks, increases in the prices of food imports, and preference erosion stemming from multilateral liberalization. It is hoped that the TIM will mitigate the apprehension of many developing countries about trade shocks arising from a successful Doha Round. In July 2004, Bangladesh became the first country to be granted TIM support under the umbrella of the country’s Poverty Reduction and Growth Facility (PRGF)-supported program. Bangladesh expects a decline in its exports of garments as a result of the scheduled removal of textile quotas at the beginning of 2005. Some African countries (e.g., Lesotho, Madagascar, and Mauritius) face similar shocks.
D. Strengthening Surveillance and Trade-Related Technical Assistance
As RTAs in Africa continue to expand and the Doha Round negotiations make progress, the IMF will need to strengthen its surveillance and trade-related technical assistance. As RTAs and multilateral trade liberalization broaden their coverage to services, especially financial services, the Fund has the important task of helping African countries strengthen their financial systems, through surveillance activities and technical assistance. The IMF may also need to examine regional financial issues and be vigilant about potential vulnerabilities arising from more integrated regional financial markets and from any impediments to the resolution of financial crises that arise. In the context of overlapping RTAs in Africa, tariff reforms have become increasingly complicated. Working together with other international financial institutions (e.g., the WTO and the World Bank), the Fund should continue to assist African countries in streamlining their tariff structures to enhance transparency and efficiency, and broaden their domestic tax base to safeguard revenues for the budget. The decision by WTO members to negotiate trade facilitation—the sole area of the four so-called Singapore Issues that remains on the Doha agenda37—will require the IMF to strengthen its technical assistance in customs administration.38 As part of surveillance and program work, the Fund will also need to take a more proactive approach to infrastructure development and reductions in trading costs both at and behind the border. Finally, the Fund should strive to strengthen its analysis and assistance in support of Africa’s regional integration arrangements.
Preference Erosion from Multilateral Liberalization
The impact of preference erosion on African countries is likely to be small and concentrated in a few countries. Subramanian (2003) finds that a 40 percent MFN tariff reduction by the Quad countries (Canada, the EU, Japan, and the United States) would likely lead to a loss of less than 2 percent of exports for least developed countries as a whole. However, the loss would be significant in a few countries, but still small in absolute value and in relation to exports. Only one country (Malawi) may face a loss exceeding 10 percent of exports. Alexandraki and Lankes (2004), focusing their research on the impact on middle-income developing countries, reach a similar conclusion: the preference erosion is heavily concentrated in a subset of preference beneficiaries—primarily small island economies dependent on sugar, bananas, and textiles.
These results must be qualified. First, there is considerable uncertainty about supply elasticities and the elasticities of substitution between imports from preference beneficiaries and those from other countries. Second, although both studies deal only with exports, imports can also be affected by preference erosion (e.g., reduced exports of garments result in fewer imports of textiles). Third, the general equilibrium effects of preference erosion need to be considered, because the exchange rate may well depreciate in response to export losses. More important, multilateral liberalization that leads to preference erosion as a by-product is likely to result in overall welfare gains for current preference beneficiaries, although short-term adjustment costs may occur. And, finally, to the extent that the impact of preference erosion represents a permanent shock that often occurs gradually and can be anticipated, it can be dealt with effectively.
E. Creating an Enabling Global Environment
The IMF should continue to advocate that multilateral trade liberalization be a top policy priority for its membership. The success of the Doha Round trade negotiations remains vital for maintaining confidence in the multilateral trade system, which has been a cornerstone of global economic prosperity since World War II. Regionalism in its current form is not a substitute for multilateral liberalization. Regionalism also runs the risk of marginalizing small and weak countries outside major trade blocs even if they are WTO members. Overall, the outcome of a global trade system in which multilateralism coexists with proliferating regionalism is uncertain, produces fewer widespread benefits, and may not serve the IMF’s objective of facilitating the expansion and balanced growth of international trade, as set out in its Articles of Agreement.
The framework agreements reached at the WTO in August 2004 are welcome and place the Doha Round back on track. What is required now is determination and commitment by all countries to deliver on the goals of the Doha Development Agenda. Both rich and poor countries carry responsibilities for promoting the fuller integration of developing countries into the global trading system. Rich countries should eliminate their agricultural export subsidies as soon as practical and substantially reduce trade-distorting domestic subsidies. They also need to reduce tariff peaks and escalation as well as barriers to manufactured exports (e.g., textiles) that are of special interest to poor countries. At the same time, trade barriers in many developing countries remain high. Poor countries, including those in Africa, should firmly reject the idea of a “round for free” (i.e., no reciprocal liberalization by certain developing countries) and undertake to reduce their own trade barriers while insisting that rich countries make strong commitments to open their markets further.
The IMF should continue to advocate coherence between trade and development assistance. To enable developing countries to reap the benefits of trade liberalization, rich countries have a responsibility to deliver the promised technical assistance to help developing countries build their capacity in trade areas. More generally, industrial countries need to boost their overseas development assistance. Current aid flows are not only insufficient, but are also unpredictable and often uncoordinated among donors. Better aid coordination and multiyear commitments are key steps in making development assistance more effective.
If trade preferences in favor of African countries have not worked in the first place, then there is no risk of preference erosion from MFN liberalization. The ability of African countries to utilize trade preferences seems to vary. For example, in the case of textile and garment exports to the United States under the AGOA, several African countries—namely, Kenya, Lesotho, Madagascar, Mauritius, Namibia, and Swaziland—have been able to increase their exports substantially, while many others have not demonstrated such a capacity.
Insofar as multilateral tariff reductions focus on bound rates rather than applied rates, the revenue impact of trade liberalization is muted because bound rates in Africa tend to be well above applied rates (Baunsgaard and others, 2003).
It is argued that countries that have already a well-functioning value-added tax (VAT) are best placed to recoup revenues lost through trade reform.
The other three issues that have been dropped from the Doha Round agenda are investment rules, competition policy, and government procurement.
Other international organizations, such as the International Organization for Standardization, the World Customs Organization, the WTO, and the World Bank will need to help African countries overcome technical barriers to trade and barriers arising from sanitary and phytosanitary measures in importing countries. These barriers are becoming increasingly important as tariff levels fall.