Financial development and economic growth in Africa: an examination of causation and efficiency

Abstract

This thesis assesses the significance of real bank credit in stimulating real output paying particular attention to the factors that prompt financial intermediation within the economy. The thesis contributes to the existing literature on finance and growth by providing fresh empirical evidence in the case of the Nigerian economy and Africa as a whole. In the context of Nigeria, credit Granger causes output, but the reverse is not true. In testing the factors that mobilise credit, I find that exports are negatively related to credit. Moreover, since credit usually fund non-oil exports, I also find that oil exports is negatively related to credit, whereas non-oil exports is positively related to credit. The latter also explains why capital inflows and imports are positively related to credit in my study. Extending the analysis to Africa as a whole, I find that causality is bi-directional. In examining the factors which mobilise credit (based on three measures of output); I find that output consistently exerts a positive influence on credit, whereas inflation and exports exert the opposite effect. However, the impact of government expenditure on credit is ambiguous. These results are re-confirmed when I use an alternative estimator for robustness. In line with the variables used in the Nigerian case, both capital inflow and imports positively influence credit while the impact of exports is negative for the whole of Africa. When examining the drivers of output in the African context, I find that credit and exports positively influence output whereas inflation exerts the opposite effect. The role of government expenditure is equally ambiguous. A further robustness test again confirms these results. The relationship between exports and credit in the literature is positive hence, it is important to investigate why the opposite holds in the Nigerian and African context. As such, I examine the efficiency of the banking system using three different measures, which includes loans, other earnings and other operating income since this may explain the counter intuitive result: export sales in Africa are largely intermediated by multi-national firms who prefer to obtain financing from credit markets that are more efficient than the African banking system. Across Africa, efficiency of the banking system is 74%, 76% and 92% when loans, other earnings and other operating income are respectively used as the output variables. This implies that 26% of credit is allocated in an unproductive way while 24% and 8% of expenditure could be better managed. When dividing the sample into medium and low-income countries, I find the respective levels of efficiency for each of the measures to be 94% and 11%; 83% and 0%; 90% and 0% for loans, other earnings and other operating income as the output variables respectively. This result supports bank loans as the best output variable, which I use further in the estimation. Further clues as to why there should be such differences in efficiency are obtained when the sample is split by regions, since there are regional variations in the use of credit. The Central African region is the least efficient. In these economies, resources are typically held and allocated by a few individuals.