This paper empirically identifies the main driving forces behind the recent development in economic growth across Sub-Saharan Africa based on a two-step procedure. Given the role of convergence in explaining the level of economic development, the first step employs the new extension of the sigma convergence developed by Phillip and Sul (2007) to test and endogenously identify the formation of different steady state paths across a sample of 34 countries selected based on available data over the period 1996-2010. Empirical results vindicate the existence of three main convergence clubs and a divergent group of 8 countries; suggesting that Sub-Sahara African countries do not form a homogenous club. The second step implements a Bayesian Averaging of Classical Estimates (BACE) method on the only convergent groups in order to explicitly account for the assumed conditional convergence in cross-sectional growth regressions. Estimation results prove support that 8 out of 18 selected explanatory variables documented in the literature are significantly and strongly associated with the long term economic growth. Particularly, investment and the relative price of exports are found to be favourable to the recent regional economic performance while public consumption and remittances appear to be of less contribution. Other important variables include scientific research, trade taxes, land availability and population growth which are unexpectedly found to be negatively associated with economic growth. Although their sign certainty probabilities are reportedly insignificant, these results raise a number of policy challenges including poor quality of institutions, the exposure to world shocks given the dependence to international trade taxes, the poor quality of human capital and more importantly a threat of skilled labour immigration.
What explains the recent growth performance in Sub-Saharan Africa? Results from a Bayesian Averaging of Classical Estimates (BACE) Approach
Working paper 578
Subscribe to new publication notifications (quarterly)