ASSESSING THE SPATIAL ACCESSIBILITY OF MICROFINANCE IN NORTHERN BANGLADESH: A GIS ANALYSIS

This paper attempts to understand and operationalize the notion of spatial accessibility (SA) in the context of microfinance. Using geographic information system (GIS) data from northern Bangladesh, we have generated a kernel‐smoothed map and found remarkable spatial variation in access to microcredit. Results suggest that areas isolated from physical infrastructure, administrative establishments, and prone to ecological shocks, exhibit lower degree of SA. Moreover, using an instrumental variable framework, we found that SA has a significant positive impact on household's decision to borrow and on the number of loans: one standard deviation higher SA is associated with a rise in participation probability and average number of microloans by, at least, 3.5 percentage points and 16 percent, respectively.