Measuring Microfinance: Cognitive and Experimental Bias with New Evidence from Nepal

This study examines reasons for the discrepancy between the claims of large microfinance impact by development practitioners and the smaller impacts found in recent experimental studies. It uses a unique data set to study the impact of microfinance in an area of Nepal in which microfinance did not previously exist, and where 51% of the target population took microfinance loans when microcredit was introduced.The study demonstrates that the discrepancy is the result of two sources of bias, one upwards by practitioners and one downwards by researchers. Practitioners observations overstate the true causal impact of microfinance, while downward bias likely exists in the local average treatment effects estimated by recent experimental and quasi-experimental studies of microfinance. Findings include:

• Small enterprises display large gains in business investment, home improvement, and durables consumption after taking microloans;
• Causal impact of microfinance based on the availability of microfinance is more modest, limited to impacts on small enterprise capital investment, and increases in livestock herds;
• About 68.3% of apparent microfinance impact observed by practitioners is an illusion driven by correlated unobservable factors.