Panel logit regression analysis of the effects of corruption on inflation pattern in the Economic Community of West African states

Abstract

It is widely believed that increase in money supply increases inflation rates in the long run. Dissenting evidences have later emerged to show that the inflation-money growth link has weakened over time. This study provides a political interpretation of inflation rates dynamics by investigating corruption as a potential risk factor associated with high inflation in the Economic Community of West African States (ECOWAS) using panel logit model. All 15 ECOWAS members were selected and studied between 1995 and 2019. Estimates revealed that, in the long run, ECOWAS inflation was significantly determined by past inflation shocks, level of corruption, real GDP and exchange rates; the link between inflation and money supply growth had indeed weakened over time. In particular, results showed that, other things given, a unit increase in the level of corruption increased the risk of high inflation by 82.6%. Results also indicated that seigniorage was not enough to explain the link between inflation and corruption, suggesting that some other channels exist through which corruption affects inflation. In addition, taking 0.5 as the benchmark for forecast probabilities, the logit model reconstructed all the high inflation episodes experienced in ECOWAS in spite of the outlying observations of Liberia and Guinea. The study therefore concludes that other things given, the more corrupt ECOWAS members are, the higher the likelihood to experience high inflation.