African polygamy: Past and present

Motivated by a simple model, I use DHS data to test nine hypotheses about the prevalence and decline of African polygamy. First, greater female involvement in agriculture does not increase polygamy. Second, past inequality better predicts polygamy today than does current inequality. Third, the slave trade only predicts polygamy across broad regions. Fourth, modern female education does not reduce polygamy. Colonial schooling does. Fifth, economic growth has eroded polygamy. Sixth and seventh, rainfall shocks and war increase polygamy, though their effects are small. Eighth, polygamy varies smoothly over borders, national bans notwithstanding. Finally, falling child mortality has reduced polygamy.