Is there a nonlinear relationship between financial development and poverty in Africa?

Purpose The paper examines whether there is a threshold between financial development and poverty in African economies. Design/methodology/approach The study adopts the innovative dynamic panel threshold model of Seo and Shin (2016) made practicable by Seo et al. (2019)–the model estimates threshold relationship even in the presence of endogeneity. Also, following the recommendations of Cihak et al. (2013) and Sahay et al. (2015), we also adopt a robust measure of financial development based on the four pillars of financial deepening, stability, efficiency and access derived from the principal component analysis (PCA). Findings The empirical results show that there exists a threshold level of financial development necessary for poverty reduction in Africa. Research limitations/implications Our result is important for policy formulations. First, individual African country must discover the level of financial development necessary for spurring poverty reduction. Second, policymakers, especially in lower-income countries, must keep improving their financial sector development to achieve the threshold level necessary for achieving poverty reduction even though financial development might seem less relevant at its present level. Practical implications The policymakers in Africa should note that there exists a threshold level of financial development that reduces poverty. Hence, the present level of financial development might have not yielded a considerate effect on poverty. Still, the policymakers must keep pushing on until the threshold is achieved. Social implications Financial development reduces poverty level but it must reach a certain threshold level before it does so. So, we advise African policymakers to continue to develop their financial sector to achieve this threshold. Originality/value This seems to be the first work to document the threshold relationship using the dynamic panel threshold. Besides, the study specifically concentrates on Africa dividing the continent into different income levels. Moreover, we adopt a robust measure of financial development unlike extant studies on Africa.