In recent years, sub-Saharan African countries have grown remarkably. According to data from the Penn World Table 7.0 (Heston et al. 2011), average annual real GDP per capita growth from 2005-9 has been over 2.5% (3.5% when excluding 2008 and 2009). This recent growth performance is remarkable given that, for over four decades since 1960, real GDP per capita growth in sub-Saharan Africa was dismal, averaging less than 0.5% per annum. Sub-Saharan African countries in recent years have also made significant progress in terms of poverty reduction (Chen and Ravallion 2010).
By casual empiricism, it is interesting to note that the average sub-Saharan African country is today over 30% more open to international trade than in1960 (as measured by the ratio of exports plus imports over GDP). The big question is, of course, whether this increase in trade openness is a cause or a consequence of the increase in economic growth. Answering this question is important for economic policy. Indeed, the World Bank (2012) published a report on the virtues of regional integration in Africa, emphasising a policy agenda that goes well beyond trade policy reforms, touching on numerous policy issues that affect international trade costs within the region. What would Africa get in return? It is unfortunate, but the truth is that most proposed policies will not be properly evaluated. In some cases, the policies are too broad to be subject to impact evaluations in the tradition of randomised control trials. We must look elsewhere for estimates of the potential effects of increasing trade. International comparisons are to some extent unavoidable.
Identifying the causal effect of trade openness on growth in sub-Saharan Africa
In Brueckner and Lederman (2012), we tackle causality issues by using panel data and novel instrumental-variable estimations. The use of panel data allows us to exploit within-country variations in countries' trade openness and GDP per capita, controlling for any time-invariant country characteristics that affect both international trade and economic growth. The empirical literature that has explored the effects of international trade openness on economic growth in cross-sections of countries has been plagued with omitted variables bias, stemming from cross-country differences in history, geography, and ethnic composition.1
The first strategy uses rainfall as an instrument for GDP per capita to estimate the response of trade openness to within-country variations in GDP per capita. This strategy builds on prior literature that has established a robust effect of rainfall on African countries' GDP per capita.2 The strategy enables us to obtain useful information on the extent to which international trade is itself a function of GDP per capita. Importantly, we can compute a residual trade-openness variable that, by construction, is exogenous to within-country variations in GDP per capita. We use this residual trade-openness variable as an instrument to estimate the within-country growth effect of openness to international trade.
The second and complementary strategy uses the GDP growth rates of OECD countries as an instrument for trade openness in sub-Saharan African economies. Higher real GDP growth of OECD countries can increase trade openness in Africa through two main channels – the supply channel (higher OECD GDP growth increases OECD countries’ exports of goods and services) and the demand channel (higher OECD GDP growth leads to an increase in the consumption of goods and services produced by sub-Saharan African countries). The fact that countries' GDP is only a tiny fraction of OECD countries' GDP ensures that variations in OECD countries' GDP growth are plausibly exogenous to within-country variations of sub-Saharan African countries' openness to international trade.
We find that openness to international trade increases economic growth in sub-Saharan Africa. The instrumental-variable estimates suggest that, on average, a one percentage point increase in trade openness is associated with a short-run increase in GDP per capita growth of about 0.5% per year. The long-run effect is larger, reaching about 0.8% after ten years. Importantly, these results are robust to controlling for year effects and other growth correlates related to political institutions and intra-national conflict. They are quantitatively in line with the cross-sectional growth estimates reported in, for example, the seminal paper by Frankel and Romer (1999) and more recently by Feyrer (2009).
The panel regressions also allow us to explore how the growth effects of openness to international trade vary across countries. In the context of sub-Saharan Africa, we are particularly interested in the role that ethnic divisions play in shaping the impact of international trade openness on economic growth. This is motivated by the theoretical literature on the “voracity effect” (Lane and Tornell 1998, Tornell and Lane 1999). The voracity literature predicts that trade windfalls can have adverse effects on economic growth in polarised countries with weak legal-politico institutions. Consistent with the theoretical literature, we find that the positive effect of trade openness on economic growth significantly declines with ethnic polarisation.3 Hence, while for sub-Saharan Africa as a whole increases in international trade openness were, on average, good for growth, our findings call for some caution in expecting large growth benefits associated with international trade openness in countries that are characterised by strong ethnic divisions. In that regard, our results echo Easterly and Levin (1997) who document that strong ethnic divisions are associated with growth-prohibiting policies.
Much has been written on the causes of Africa's so-called "growth tragedy". Amongst the most popular explanations are deep-rooted ethnic divisions, colonial heritage, and the slave trade (e.g. Easterly and Levine 1997, Acemoglu et al. 2001, Nunn 2008, Nunn and Watchekon 2011, and Michalopoulos and Papaioannou 2011). While these explanations are popular, in particular, for differences in cross-country averages of GDP per capita growth over many decades, it is more difficult to see how these variables could explain the recent increase in GDP growth (or more generally, any within-country change in growth), although deep rooted country characteristics such as ethnic divisions could possibly interact with time-varying variables that determine economic growth.
Acemoglu, D., S. Johnson, J. Robinson (2001), "The Colonial Origins of Comparative Development: An Empirical Investigation," American Economic Review 91: 1369-1401.
Barrios, S., L. Bertinelli and E. Strobl (2010), ‘Trends in Rainfall and Economic Growth in Africa: A Neglected Cause of the African Growth Tragedy’, Review of Economics and Statistics 92(2): 350-66, May.
Brueckner, M. and A. Ciccone (2011), ‘Rainfall and the Democratic Window of Opportunity’, Econometrica 79(3): 923-47, May.
Brueckner, M. and D. Lederman (2012), 'Trade Causes Growth in Sub-Saharan Africa', World Bank Policy Research Working Paper 6007.
Chen, S. and M. Ravallion (2010). "The Developing World is Poorer than we Thought, But No Less Successful in the Fight Against Poverty. Quarterly Journal of Economics 125: 1577-1625.
Easterly, W. and R. Levine (1997), 'Africa's Growth Tragedy: Policies and Ethnic Divisions,' Quarterly Journal of Economics 112(4): 1203-50, November.
Feyrer, James. 2009. “Trade and Income – Exploiting Time Series in Geography.” Mimeo, Dartmouth University,
Frankel, J.A. and D. Romer (1999), ‘Does Trade Cause Growth?’ American Economic Review 89(3): 379-99, June.
Heston, A., Summers, R., and B. Aten (2011), Penn World Table Version 7.0. Center for International Comparisons of Production, Income and Prices, University of Pennsylvania, May 2011.
Lane, P.R. and A. Tornell (1998). "Are Windfalls a Curse? A Non-Representative Agent Model of the Current Account." Journal of International Economics 83-112.
Levine, R. and D. Renelt (1992). "A Sensitivity Analysis of Cross-Country Growth Regressions." American Economic Review 82: 942-963.
Michalopoulos, S. and E. Papaioannou (2011). "The Long Run Effects of the Scramble for Africa." CEPR DP 8676.
Miguel, E., S. Satyanath and E. Sergenti (2004), ‘Economic Shocks and Civil Conflict: An Instrumental Variables Approach’, Journal of Political Economy 112(4): 725-53, August.
Nunn, N. (2008). The Long-Term Effect of Africa's Slave Trades.” Quarterly Journal of Economics 139-176.
Nunn, N. and L. Wantchekon (2011). “The Slave Trade and the Origins of Mistrust in Africa.” American Economic Review (forthcoming).
Rodrik, D., Arvind Subramanian, and Francesco Trebbi (2004), "Institutions Rule: The Primacy of Institutions Over Geography and Integration in Economic Development," Journal of Economic Growth 9: 131-165.
Tornell, A. and P. Lane (1999), "The Voracity Effect," American Economic Review 89: 22-46.
WDI (2011), World Development Indicators, Online Database.
World Bank (2012), De-Fragmenting Africa: Deepening Regional Trade Integration in Goods and Services, Washington, D.C.
1 In a seminal paper, Frankel and Romer (1999) showed that exogenous cross-country variations in international trade were positively correlated with GDP per capita across countries. Their identification strategy relied on the gravity model of trade, whereby bilateral geographic distances between trading partners was used as an exogenous instrument for each country’s volume of trade. However, subsequent research argued that these geographic instruments were closely correlated with countries' historical experiences during colonial times, which in turn help explain the international differences in governance and institutions (Acemoglu et al. 2001). Rodrik et al. (2004) thus argued, and showed empirically, that “institutions rule”. Once the institutional channel is controlled for, international trade appears to have no impact on GDP per capita across countries.
2 See Miguel et al. (2004), Barrios et al. (2010), and Bruckner and Ciccone (2011).
3 While there is a significant interaction effect between trade and ethnic polarisation, there is no significant interaction effect between trade and ethnic fractionalisation. This asymmetry in results with the two ethnic diversity measures is consistent with the Lane and Tornell voracity model. The voracity model predicts that the voracity effect is strongest when there are two powerful groups, and that the effect diminishes as the number of groups increase. This non-linearity is captured by the polarisation index, but not by the fractionalisation index.