Microfinance on the margin: why recent impact studies may understate average treatment effects.

A series of recent randomised trials estimate the impact of microfinance on incomes, consumption, and other key measures of welfare. This comment demonstrates why impact estimates obtained from experimental designs focusing on marginal microfinance borrowers are likely to understate the impacts yet realised by inframarginal borrowers, those having taken microfinance loans prior to implementation of an experiment, when field experiments are implemented in areas broadly served by microfinance. [ABSTRACT FROM PUBLISHER]