The Outreach Measurement Debate in Microfinance: Does Average Loan Size Relate to Client Poverty?

Depth of outreach in microfinance is conceptualised as the degree to which a microfinance institution serves (very) poor clients. However, this concept is commonly approximated in academic research by 'average loan size'. This rough proxy has previously also been used by policy makers such as USAID but they have moved away from it recently due to reliability concerns. Academics have voiced a numerous conceptual critiques of this proxy but, to date, the lack of reliable client poverty data for a sufficient number of MFIs prevented researchers from an empirical investigation of the merits of this proxy. Fortunately, we have access to the Microfinanza dataset, which contains client poverty and average loan size data for a reasonable number of MFIs. Our empirical results largely confirm the critiques, as we cannot identify more than a weak relationship between average loan size or one of its variants and measures of client poverty. Regressing MFIs financial performance indicators in a controlled research design on various depth of outreach measures, we crucially find that the relationship changes substantially when an average loan size proxy is replaced by an actual client poverty measure. Our results imply that researchers aiming to understand the relationship between depth of outreach and MFI financial performance might want to substitute crude average loan size proxies by actual depth of outreach measures.