76975 APRIL 2013 • Number 114 Can Trade Reduce Poverty in Africa? Maëlan Le Goff and Raju Jan Singh While most economists accept that, in the long run, open economies fare better in aggregate than closed ones, many fear that trade could harm the poor. African countries, for example, have realized significant improvements in trade liberal- ization in recent decades, yet Africa remains the poorest continent in the world. It seems that the large gains expected from opening up to international economic forces have been limited in Africa, especially for poor people. Drawing on the findings of a recently published working paper (Le Goff and Singh 2013), this note argues that the benefits of trade are not automatic, but rather depend on accompanying policies aimed at developing the financial sector, promoting primary education, and improving governance. This accompanying policy agenda allows people to take advantage of the opportu- nities offered by freer trade, by reallocating resources away from less productive activities to more promising ones. Trade liberalization therefore should not be implemented on its own, but with the necessary complementing policies. Trade and Poverty in Africa mon Market for Eastern and Southern Africa (COMESA), the East African Community (EAC), and the Southern Afri- Trade liberalization is being actively promoted as a key com- can Development Community (SADC)—are testimony to Af- ponent of development strategies. Through greater efficiency in resource allocation, specialization in production, dissemi- rican leaders’ willingness to ease trade restrictions. nation of knowledge and technological progress, and competi- Although Africa’s participation in global trade remains tion, trade liberalization should favor economic development limited, trade to and from Africa has significantly expanded. and growth. How is this reflected in poverty? In theory, a The continent is still one of the least open regions of the more liberalized trade regime will make greater use of the fac- world. In 2011, African trade accounted only for 3.5 percent tors of production that are most abundant, thus increasing of global exports and imports (down from 5.3 percent in their relative price. If poverty and relative low income stem 1980), compared to about 6 percent for developing Latin from abundance of labor, greater trade openness should lead American countries and 32 percent for Asia (figure 1). Ex- to higher prices for labor and a decrease in poverty. ports from African countries have nevertheless accelerated, Hence, the removal of tariff and nontariff barriers (ex- expanding annually, on average, by 2.6 percent in the 1980s, port and import bans, restrictive rules of origin), which re- by 8 percent in the 1990s, and by 15 percent in the 2000s. main particularly high in Africa, and improvements in trade Over the last decade, this rate of increase for Africa outpaced facilitation have been recognized as a means for supporting the world average of 9.7 percent. During the same decade, the Africa’s economic development. Recent attempts to consoli- average yearly growth rate of African imports reached about date trade integration within Africa—for example, the Com- 17 percent. This increase in both exports and imports has re- 1 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 1. Share in Global Trade by Region, 1960–2011 Figure 2. Trade Openness by Region, 1980–2011 45 50 percent of total trade 40 45 35 40 percent of GDP 30 35 25 30 20 25 15 10 20 5 15 0 10 60 65 70 75 80 85 90 95 00 05 10 80 85 90 95 00 05 10 19 19 19 19 19 19 19 19 20 20 20 19 19 19 19 20 20 20 developing countries African developing countries developing countries African developing countries Latin American Asian developing countries Latin American Asian developing countries developing countries developing countries Source: UNCTAD statistics. Source: UNCTAD statistics. sulted in greater trade openness in Africa, measured as the from trade. Krueger (1983) has used this insight to argue that sum of exports and imports as a share of gross domestic prod- trade reforms in developing countries should be pro-poor, uct (GDP; figure 2). since these countries are most likely to have a comparative ad- Yet, with almost 50 percent of the population living be- vantage in producing goods requiring unskilled labor. low US$1.25 a day, sub-Saharan Africa remains the poorest From a dynamic perspective, economic growth is critical continent in the world. The large gains expected from open- to sustained poverty alleviation, and trade liberalization is ar- ing up to international economic forces have, to date, not gued to require increases in productivity to sustain growth. been realized in many African countries, especially for poor Freer trade provides greater incentives for investment, the people. It seems that countries are not equally able to make benefits of scale and competition, limitation on rent-seeking use of the opportunities arising out of increased access to in- activities favored by trade restrictions, and openness to new ternational markets. ideas and innovations (Berg and Krueger 2003; Grossman So, can trade actually reduce poverty in Africa? How and Helpman 1991; Lucas 1988). could poor people in Africa benefit more from the economic However, for comparative advantage to increase the in- opportunities offered by greater trade openness? Are there comes of unskilled workers, workers need to be able to move some complementary policies that could allow the benefits of out of shrinking sectors and into expanding ones. Davis and trade to be more widely shared among the population? Mishra (2006), Goh and Javorcik (2006), and Topalova (2006) suggest that labor in the real world may not be as What Does Trade Theory Tell Us? mobile: there are too many barriers to entry and exit for Both theoretically and empirically, the impact of trade open- firms, and too many barriers to labor mobility for workers. ness on poverty is ambiguous. Most economists accept that, In this case, the expected benefits from freer trade may not in the long run, open economies fare better in aggregate than materialize. do closed ones, and that relatively open policies contribute Trade openness may also increase the size of the informal significantly to development. Many analysts fear, however, sector (Goldberg and Pavenik 2003). Being more exposed to that in the shorter run, trade may harm the poorer actors in foreign competition, firms may be incited to reduce their the economy, and that even in the longer run, successful open costs by hiring temporary workers instead of permanent ones, trade regimes may leave some people behind in poverty. or even to lay off workers, who may in turn obtain informal Bhagwati and Srinivasan (2002) distinguish between jobs. Depending on the wage differences between sectors, this two broad argument strands when discussing the effects of could lead to an increase in poverty. freer trade on poverty: static and dynamic. This note exam- In addition, if the poor are mostly completely unskilled, ines, in the former case, how freer trade effects poverty, taking and it is semiskilled labor that is in increased demand, pov- resources and technology as given, and in the latter case, erty will be unaffected—or possibly, worsened. Trade liberal- growth effects and the evolution of poverty over time.1 ization may even be accompanied by skill-biased technical Following the static approach, the Stolper-Samuelson change, which can mean that skilled labor may benefit rela- theorem, in its simplest form, suggests that the abundant fac- tive to unskilled labor. Lower prices for capital goods or in- tor should see an increase in its real income when a country creased competition following trade liberalization could opens up to trade. If the abundant factor in developing coun- encourage firms to import machines and increase their de- tries is unskilled labor, then this framework suggests that the mand for skilled labor (Acemoglu 2003). Furthermore, poor (unskilled) in developing countries have the most to gain many developing countries are rich in natural resources: 2 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise trade would stimulate this sector rather than the labor-in- benefits of trade do not accrue automatically, and that poli- tensive sectors. cies that complement trade opening are needed. To reap the Krueger (1983) shows through case studies that develop- potential wage, employment, and income gains associated ing countries’ manufactured exports were, indeed, labor in- with trade, trade reforms should be accompanied by policies tensive, but that the employment effects of freer trade poli- that ensure (i) macroeconomic stability and a sound invest- cies were generally rather limited. A number of cross-country ment climate; (ii) protection for workers; (iii) maintenance of studies on poverty, while not dealing with trade explicitly, in- high-quality working conditions; and (iv) facilitation of labor corporate trade openness as a control variable and showed transitions. similar results: at best the benefits of greater trade openness What does this literature entail for Africa? This note, seem to have bypassed the poor (Beck, Demirgüç-Kunt, and based on Le Goff and Singh (2013), argues that the effect of Levine 2007; Dollar and Kraay 2001; Guillaumont-Jean- trade openness on poverty in Africa depends on complemen- neney and Kpodar 2011; Kpodar and Singh 2011; Singh and tary reforms that help a country take advantage of new op- Huang 2011). portunities. Working with pooled cross-country and time se- ries data for 30 African countries averaged over five-year Toward a Policy Agenda periods from 1981 to 2010, Le Goff and Singh (2013) pro- This lack of any clear correlation between openness measures vide new evidence on the links between trade liberalization and poverty indicators in aggregate may suggest that trade lib- and poverty reduction in Africa. Their results uncover an in- eralization requires combination with other policies to have a teresting pattern of reform complementarity: trade openness significant impact on poverty. The sort of policies envisaged tends to reduce poverty in countries as their financial sectors would be those that encourage investment, allow effective grow deeper, their education levels higher, and their gover- conflict resolution, and promote human capital accumula- nance stronger. tion (Winters, McCulloch, and McKay 2004). These results correspond to a certain logic. Three dimen- For example, Sindzingre (2005) suggests that institu- sions (finance, education, and governance) capture an econo- tions could help explain the heterogeneity in the globaliza- my’s ability to reallocate resources away from the less produc- tion-poverty relationship. She argues that domestic political tive sectors to the more productive ones. This, in turn, allows structures and institutions (such as oligarchic or predatory countries to take better advantage of the opportunities of- regimes) may prevent the poor from benefiting from global- fered by trade. ization. Similarly, Bolaky and Freund (2008) show that trade A more developed financial sector, as measured by the reforms actually lead to income losses in highly regulated private sector credit-to-GDP ratio, allows banks and investors economies. Excessive regulations restrict growth because re- to quickly identify new and promising sectors and to redirect sources are prevented from moving into the most productive credit. A more educated population, as measured by primary sectors and to the most efficient firms within sectors. completion rates, would be more able to acquire the new skills More recently, Haltiwanger (2011) and McMillan and sought by growing sectors and adjust more rapidly to the new Verduzco (2011) argue that benefits of trade depend to a conditions of the labor market. Finally, better governance, as large extent on national institutional settings. The process of measured by the rule of law, would facilitate the entering into trade-induced growth entails a continual reallocation of re- and dissolution of contracts and make conflict resolution sources away from less productive activities to more produc- easier. tive ones. Many things can go wrong in this reallocative pro- It’s easy to imagine how any of these factors could help cess if economies are distorted, for instance, if transportation pull people out of poverty. An easy-to-get business loan could or communication infrastructure are not sufficiently devel- help a new grocery store open and create new jobs, for exam- oped, if ineffective (or nonexistent) competition policy can- ple. A literate worker would be better able to transition from not prevent large firms from abusing their market power, or if low-paying agriculture work to a job in a new factory. And for- financial markets are not sufficiently developed to fund new eign investors, observing that a country’s government enforc- and expanding businesses. In such distorted economic envi- es contracts, might be more willing to invest in a new plant ronments, there is little chance that the benefits of greater and hire local workers. If these conditions are not met, how- trade openness will materialize and—in extreme cases—a “de- ever, greater openness to trade could be associated with higher coupling� may take place, that is, where policy reforms induce levels of poverty. some firms to downsize or exit the market, but do not lead to Eyeballing the data seems to confirm this intuition. Fig- the expansion of other firms. ure 3 compares the trade-poverty relationship in the top and Similarly, reviewing the new wave of research under the bottom country groups in terms of financial development, International Collaborative Initiative on Trade and Employ- education, and governance.2 These charts suggest that for ment, Newfarmer and Sztajerowska (2012) conclude that the each conditional variable considered, the relationship be- 3 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Figure 3. Poverty and Trade Openness for Top and Bottom A more rigorous examination of the data confirms these Reformers observations. Following the same approach as Chang, Kaltani, a. Trade and poverty depending on �nancial development and Loayza (2009), analysis starts with a linear regression 5 specification and then extends the regression to account for 4 interaction terms between an openness measure and proxies log of poverty headcount for various country characteristics (financial depth, educa- 3 tion, and governance). While, on average, trade does not seem 2 to be associated with lower poverty in Africa, this observation hides important nonlinear relationships. First, the poor start 1 benefiting from trade when the development of domestic pri- 0 vate credit reaches a threshold of 17.7 percent of GDP, which 2 3 4 5 6 is far below the sample average. Second, trade openness starts -1 being favorable to the poor when the share of the population -2 log of trade openness over age 15 with completed primary education exceeds 46.7 bottom group linear (bottom group) percent. Third, trade openness could be favorable to the poor top group linear (top group) when institutional quality (measured by the law and order variable) reaches 3.3 (on a 0–6 scale, 6 indicating high-quality b. Trade and poverty depending on the quality of institutions institutions). 5 The good news is that these results suggest that many Af- 4 rican countries meet these conditions and are well positioned log of poverty headcount 3 to take advantage of the opportunities offered by trade. On average, however, while the financial sector is deep enough 2 and the level of education high enough for countries to bene- 1 fit from trade, institutions are generally too weak. Countries below these thresholds will need to enact a wide range of poli- 0 2.5 3 3.5 4 4.5 5 5.5 6 cies designed to harness the power of trade for economic de- -1 velopment. Inadequate policies and institutions, weak hu- log of trade openness man capital, and limited financial development are not only -2 bottom group linear (bottom group) bad for a country’s welfare, they also hold back the poor, deny- top group linear (top group) ing low-income residents of developing countries the oppor- c.Trade and poverty depending on education level tunity to benefit from freer trade. 5 About the Authors log of poverty headcount 4 Maëlan Le Goff is an Economist at the CEPII (French Research 3 Center in International Economics), and Raju Jan Singh is a Lead 2 Economist in the Poverty Reduction and Economic Management (PREM) Network of the World Bank’s Africa Region. 1 Notes 0 2 3 4 5 6 -1 1. See Winters, McCulloch, and McKay (2004) for a detailed discussion on the possible channels for freer trade to affect -2 log of trade openness poverty. bottom group linear (bottom group) 2. Countries are classified in the top (bottom) group if they top group linear (top group) belong to the top one-fourth (bottom three-fourths) of a rank Source: Authors’ calculations and the International Country Risk Guide for institutions; UNESCO for primary education; and IMF for financial development. distribution given by each conditional variable (financial de- velopment, education level, and quality of institutions). tween trade openness and poverty appears negative and more References pronounced in the top group than in the bottom group. This would suggest that trade does tend to reduce poverty, but Acemoglu, D. 2003. “Patterns of Skill Premia.� Review of Economic Studies 70 (2): 199–230. only when the conditions are right: in countries where finan- Attanasio, O., P. Goldberg, and N. Pavenik. 2004. “Trade Reforms cial sectors are deep, education levels high, and governance and Wage Inequality in Colombia.� Journal of Development strong. Economics 74 (2): 331–66. 4 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise Beck, T., A. Demirgüç-Kunt, and R. Levine. 2007. “Finance, In- Kpodar, K., and R. Singh. 2011. “Does Financial Structure Matter equality and the Poor.� Journal of Economic Growth 12: 27–49. for Poverty? Evidence from Developing Countries.� World Bank Berg, A., and A. Krueger. 2003. “Trade, Growth, and Poverty: A Se- Policy Research Working Paper WPS5915, Washington, DC. lective Survey.� IMF Working Paper WP/03/30, Washington, DC. Krueger, A. 1983. Trade and Employment in Developing Countries: Bhagwati, J., and T. N. Srinivasan. 2002. “Trade and Poverty in the Volume 3: Synthesis and Conclusions. New York: NBER. Poor Countries.� AEA Papers and Proceedings 92 (2): 180–83. Le Goff, M., and R. J. Singh. 2013. “Does Trade Reduce Poverty? A Bolaky, B., and C. Freund. 2008. “Trade, Regulations, and Income.� View from Africa.� World Bank Policy Research Working Paper Journal of Development Economics 87 (2): 309–21. 6327, Washington, DC. Chang, R., L. Kaltani, and N. V. Loayza. 2009. “Openness Can Be Lucas, R. E. 1988. “The Mechanics of Economic Development.� Good for Growth: The Role of Policy Complementarities.� Journal of Monetary Economics 22 (1): 3–42. Journal of Development Economics 90 (1): 33–49. McMillan, M., and I. Verduzco. 2011. “New Evidence on Trade Davis, D., and P. Mishra. 2006. “Stopler-Samuelson Is Dead and and Employment: An Overview.� In Trade and Employment: Other Crimes of Both Theory and Data.� In Globalization and From Myths to Facts, ed. Marion Jansen, Ralf Peters, José Manuel Poverty, ed. A. Harrison. Chicago: University of Chicago Press Salazar-Xirinachs, 23–60. Geneva: ILO. for NBER. Newfarmer, R., and M. Sztajerowska. 2012. “Trade and Employ- Dollar, D., and A. Kraay. 2001. “Trade, Growth and Poverty.� ment in a Fast-Changing World.� In Policy Priorities for Interna- Mimeo, Development Research Group, World Bank, Washing- tional Trade and Jobs, ed. Douglas Lippoldt, 7–73. Paris: OECD. ton, DC. Sindzingre, A. 2005. “Explaining Threshold Effects of Globaliza- Goh, C., and B. S. Javorcik. 2006. “Trade Protection and Industry tion on Poverty: An Institutional Perspective.� Research Paper, Wage Structure in Poland.� In Globalization and Poverty, ed. A. UNU-WIDER, United Nations University, No. 2005/53. Harrison. Chicago: University of Chicago Press for NBER. Singh, R., and Y. Huang. 2011. “Financial Deepening, Property Goldberg, P., and N. Pavenik. 2003. “The Response of the Informal Rights, and Poverty: Evidence from Sub-Saharan Africa.� IMF Sector to Trade Liberalization.� Journal of Development Econom- Working Paper, WP/11/196, Washington, DC. ics 72 (2): 463–96. Topalova, P. 2006. “Trade Liberalization, Poverty and Inequality: Grossman, G. M., and E. Helpman. 1991. Innovation and Growth in Evidence from Indian Districts.� In Globalization and Poverty, the Global Economy. Cambridge, MA, and London: MIT Press. ed. A Harrison. Chicago: University of Chicago Press for NBER. Guillaumont-Jeanneney, S. and K. Kpodar. 2011. “Financial Devel- Winters, A., N. McCulloch, and A. McKay. 2004. “Trade Liberal- opment and Poverty Reduction: Can There Be a Benefit With- ization and Poverty: The Evidence So Far.� Journal of Economic out a Cost?� Journal of Development Studies 47 (1): 143–63. Literature XLII: 72–115. Haltiwanger, J. 2011. “Globalization and Economic Volatility.� In Making Globalization Socially Sustainable, ed. M. Bacchetta and M. Jansen, 119–46. Geneva: ILO and WTO. The Economic Premise note series is intended to summarize good practices and key policy findings on topics related to economic policy. They are produced by the Poverty Reduction and Economic Management (PREM) Network Vice-Presidency of the World Bank. The views expressed here are those of the authors and do not necessarily reflect those of the World Bank. The notes are available at: www.worldbank.org/economicpremise. 5 POVERTY REDUCTION AND ECONOMIC MANAGEMENT (PREM) NETWORK    www.worldbank.org/economicpremise