The impact of banking services on poverty: Evidence from sub-district level for Bangladesh

The extent to which banking services can reduce poverty is under question as recent experimental evidence has suggested that there is no impact. Our findings, based on survey and administrative data, are to the contrary. We compile a unique dataset of banking measures and poverty indicators at the level of Bangladesh’s 544 administrative sub-districts. We find a relationship between banking services and poverty reduction, and show that the relationship works through the deposit channel rather than the credit channel. We exploit variation in branch placement by sub-district between 2010 and 2015 to implement a difference-in-difference estimation approach. We also estimate ordinary least squares and fixed effects models to explore the role of factors other than banking services in poverty incidence. Broad findings supporting a role for the deposit channel, but not the credit channel, are confirmed. These results have significant policy implications as governments in developing countries are actively involved in promoting financial inclusion through the banking sector.