The main findings are as follows. First, regulatory reforms in Kenya have enabled agency banking to positively affect access to formal financial services (specifically services for transactions and savings usage, but not credit access in general or investment product usage). Arising from this finding, we argue that improved financial access brought about by an expanded services supply has not been used for purposes that can enhance welfare gains broadly in the near- or intermediate-term. State regulatory and policy agents need to look into making loans more accessible through additional supply points and encouraging households to use financial services for investment purposes.
Second, we establish a strong relationship between the set of regulations known as “know-your-customers,” and access to financial services in Kenya. Know-your-customers regulations require financial institutions to have a complete record of customers’ identification documents which should be stored for several years after the first encounter with the client. Incidentally, about 14% of Kenya’s adult population does not own any identification documents. We recommend that Kenya takes the route followed by other developing countries, notably India, to relax the identification requirement in specific well-defined instances when its imposition is likely to result in financial exclusion.
Third, by analyzing the relationship between two of the more prominent macro-prudential regulations – capital adequacy and liquidity – and credit to highly risky small-scale agriculture, we find that tightening prudential regulations could negatively affect provision of credit to small businesses and therefore conflict with the country’s financial inclusion objectives. In particular, the capital adequacy requirement significantly diminishes banks’ ability to provide credit to small-business clients and may result in credit rationing. However, the adverse effect of capital regulations on credit provision can be mitigated with proper macroeconomic stabilization policies. That is, when we control for macroeconomic and financial market developments, the sign of our coefficient estimate remains negative, but their magnitudes decline.