Fractionalization, polarization and banking stability in Africa

Abstract I examine the impact of diversity (ethnic and religious fractionalization and polarization) on banking stability in Sub-Saharan Africa (SSA). Using data from 1996 to 2014, I employ the system Generalized Method of Moments (sys-GMM) approach to examine this relationship. I find that countries in SSA are more polarized religiously than they are ethnically. The region is, however, more ethnically fractionalized than they are religiously. Further, I conjecture that banks in more heterogeneous societies will experience poor asset quality and lower stability. I however postulate that banks offset the risks from diversity at certain levels of net interest margin (NIM). I provide empirical evidence to support these conjectures. I find varying threshold NIM values for each diversity indicator depending on the stability measure used. Opening up the banking system to foreign entry can help offset the negative impact of diversity on banking stability. Policy implications are discussed.