Overcoming Selection Bias in Microcredit Impact Assessments: A Case Study in Peru
利用秘鲁小额信贷机构的面板数据，研究发现借款者收入本就更高，控制选择偏差后，借款仍显著提升企业利润，而忽视偏差的模型会高估影响。
There are several potential sources of bias in microcredit impact assessments. This paper uses a panel data set from a Peruvian MFI to test for impact of credit on microenterprise profits, while controlling for these biases. We find that those who will eventually become borrowers have significantly higher incomes than those who will not become borrowers, implying that selection into the lending programme is a substantial problem. After controlling for this selection, we find that an average microentrepreneur who borrows earns significantly higher enterprise profits than one who does not borrow, and that na�ve models, which do not control for selection, overestimate this impact. Fixed effects estimates give roughly the same results as the quasi-experimental cross-section analysis.