V The External Environment and Trade Developments

Arvind Subramanian
Published Date:
September 2000
  • ShareShare
Show Summary Details

External Environment

ESA countries faced changes in their external environment in the 1990s, caused by the increasing trend toward globalization and worldwide trade liberalization, most notably as a result of the conclusion of the Uruguay Round. These changes had four implications for ESA countries’ trading opportunities. Two of these relate to the preferential market access enjoyed by ESA countries and the other two to non-preferential access.

  • First, the effect of the tariff reductions implemented in the Uruguay Round was to reduce the margin of preferences enjoyed by ESA countries in industrial countries’ markets.20 The magnitude of the losses for ESA countries has been estimated, however, to be relatively small, varying between 0.1 percent and 0.5 percent of their exports (Appendix II, Table A12).21

  • Second, perhaps a more important source of disruption for ESA countries is the reduction in preferential access on items such as beef, sugar, and textiles, which these countries enjoy in the form of guaranteed quotas, particularly in the EU market. In the case of beef and sugar, there are reduced rents accruing to ESA countries’ exporters. In the case of textiles, the loss could be even more severe if ESA countries fail to adapt to the competitive conditions that would arise from the elimination of quotas on textiles and clothing products. For example, Mauritius, which has sizable exports of these products, could well face severe competition from lower cost suppliers in Asia, threatening its current market niche.

  • Third, tariff reductions by industrial countries on tropical commodities will benefit ESA countries, but the effect is expected to be small as these commodities were already subject to very low tariffs prior to the Uruguay Round (Harrold, 1995).

  • Lastly, as a result of the tariff cuts under the Uruguay Round, the tariff escalation faced by ESA countries has been reduced to some extent, although it will endure for a number of important products (wood, textiles and clothing, fish, and leather).22 Tariff escalation has the effect of encouraging exports of raw materials from ESA countries rather than products higher up the value-added chain, thereby militating against diversification and industrialization in those countries (Harrold, 1995).


During the 1990s, there was an appreciable increase in the openness of ESA countries as measured by the ratios of exports and imports of goods and services to GDP (Figure 5.1). Between 1990 and 1999, the average trade-GDP ratio increased by 9 percentage points, although wide variations existed between countries in the region, and eight countries actually experienced a decline in their trade-to-GDP ratios (Table 5.1). These developments reflected the extent of trade liberalization in the region and resulted in higher merchandise trade, rather than trade in services. On average, the share of trade in nonfactor services in GDP remained broadly unchanged at about 30 percent.

Figure 5.1.Eastern and Southern Africa: Trade Liberalization and Economic Performance

Table 5.1Eastern and Southern Africa: Openness Indicators, 1990 and 1999(Percent of GDP)
Exports of Goods

and Services
Imports of Goods

and Services
Total Trade
South Africa242523234748
ESA (excl. South Africa)1303543477382
Sub-Saharan Africa1283027325561
Source: IMF staff estimates.

A statistical analysis of the determinants of the openness of ESA countries reveals that trade liberalization was one of the factors behind the observed increase in openness during the 1990s (Box 5.1). On average, a 1 percent reduction in trade taxes led to an increase in the trade-to-GDP ratio of about 0.9 percentage points, and the corresponding coefficients were statistically significant.

Export Performance

Export performance of ESA countries lagged behind that of the rest of the world, particularly other developing countries. The average annual growth in export volume for ESA countries was 5.4 percent, compared with 6.4 percent for the world as a whole. However, there was a substantial pick-up in exports in the second half of the 1990s, with export growth of 6.4 percent, compared with 4.5 percent between 1990 and 1994. Within ESA, a few countries—Ethiopia, Lesotho, Madagascar, Mozambique, Seychelles, and Uganda—registered impressive export performance, comparable to that of other fast growing developing countries (Table 5.2). For ESA countries (as for sub-Saharan Africa in general), the growth of exports of manufactures was relatively strong, with the share in total exports increasing from 28 percent in 1985 to about 37 percent in 1996.23 Particularly striking was the performance of Madagascar, Uganda, and the SACU countries (Appendix II, Table A13). The share of traditional exports—encompassing agricultural materials, fuels, ores, and metals—all decreased relative to the late 1980s. Within manufacturing, the sector exhibiting the most dynamic growth was clothing (Appendix II, Table A14). However, the export performance of countries in the ESA region has been negatively affected by anticompetitive arrangements in the transportation and insurance sectors, as reflected in the relatively high level of freight costs that exceed the level of tariff barriers by a wide margin.24

Table 5.2Eastern and Southern Africa: Growth in Export Volume
Average Annual Rate of Growth
Congo, Democratic
Republic of-14.73.6-5.6
South Africa3.64.94.2
All countries4.56.45.4
All countries, exc.
South Africa4.66.55.5
Sub-Saharan Africa3.45.54.3
Source: Country authorities and IMF staff estimates

Box 5.1Trade Protection and Trade Outcomes in Africa

Is trade in Africa determined exclusively by exogenous factors, such as geography, or does trade policy play an important role? To examine this question, and consistent with the approach in Rodrik (1999), a set of regressions was run on a sample of 37 sub-Saharan African countries for the period 1984–98. To provide a suitably large data set for the analysis, the sample was split into two time periods, 1984–91 and 1992–98. The dependent variable was the ratio of exports and imports to GDP or the ratio of exports to GDP (the trade outcome variable), and the independent variables included the log of initial per capita income (logincome) and population (logpop), a measure of the proportion of a country’s land area in the tropics (tropics), trade taxes (tradetax), import taxes (importax), and export taxes (exportax). The tax variables were measured in terms of collections: for example, import taxes were measured as customs duty collections divided by the value of imports. Timedum is a dummy variable aimed at capturing any structural shifts across the two time periods. The results are presented in the table below.

The regressions show that trade and export outcomes are determined by income levels, geographical variables, and trade policy. The fit of the regressions, particularly for this type of cross-section analysis, is quite good. The coefficients on the size variables (income and population) are usually significant at the 5 percent level and are correctly signed. They were also quite robust to alternative specifications. More important, the trade policy variables also turned out to be significant, robust, and correctly signed. The results suggest that a 1 percentage point reduction in trade taxes leads to a trade increase of between 0.7 and 1.1 percentage points. Interestingly, and consistent with Lerner’s theoretical proposition that import taxes are equivalent to export taxes in terms of their relative price and resource allocation effects, the coefficients on the two tax terms turn out to be broadly similar.

Logpop-6.89**-2.63 *-1.40
Tropics-67.06 ***-14.06**-8.12
Tradetax-0.88 ***
Importax-0.71 ***-0.41 **
Note: ***, **, and * denote significance at the 10 percent, 5 percent, and 1 percent levels, respectively. X: value of exports. M: value of imports.

This aggregate manufacturing export performance was accompanied by a small increase in the diversity of products exported by ESA countries. Four of the eight ESA countries studied showed an increase in the number of items exported, among which Zambia stands out as a strong performer (Appendix II. Table A15). On related measures of diversification—the single-and three-product concentration ratios25—ESA countries’ performance was positive, with the concentration ratios falling in six of the eight countries. The performance of ESA countries contrasts sharply with that of six other sub-Saharan African countries investigated, which showed a marked decrease in the level of diversification measured on all three counts (the number of products exported and the one and three-product concentration ratios).

One of the key aspects of globalization has been the recent trend toward increased specialization at the level of the firm, the plant, and product lines—as reflected in increasing intra-industry trade. This specialization has also led to the phenomenon of global production sharing, whereby different components of a good are produced in various geographical locations. With the exception of South Africa, however, most African countries have insignificant levels of intra-industry trade, reflecting their low level of industrialization and the limited degree of regional vertical integration (Appendix II, Table A16). Even South Africa ranks only in the middle of emerging market countries—Brazil, Taiwan, Province of China, and Korea have intra-industry trade ratios that are nearly twice as large. Similarly, apart from South Africa, there has been no discernible dynamism in trade in parts and components in the other ESA countries.

    Other Resources Citing This Publication