Essays on Microfinance

Brian C Hunter, Purdue University


This dissertation examines three questions on microfinance borrowers, institutions, and markets. In chapter 1, we examine the association between varying loan officer busyness levels, in terms of number of clients serviced by the average loan officer, and microlender loan portfolio performance. Loan officers issue new loans, collect loan payments, and minimize repayment issues. If an MFI increases the workload of their loan officers, and/or relaxes borrower selection standards due to competitive pressures, it may reduce the capability of staff to screen and monitor borrowers which would, in turn, lead to lower loan repayment rates. We use MIX Market panel data from 2007 to 2014 and find that MFIs with busier loan officers also tend to have higher loan default rates on average, but that the relationship depends on the lending methodology used by the MFI. MFIs lending to individuals exhibit a positive association between loan officer busyness and loan portfolio performance. For lenders using the solidarity group methodology, the relationship is non-linear, indicating that loan portfolio performance increases as loan officers work toward managing 433 clients on average, then decreases as busyness increases beyond 433 clients. A better understanding of a lender’s capacity to increase the productivity of loan officers, without threatening loan portfolio performance, has implications for MFIs seeking to increase sustainability and impact. In chapter 2, we search for evidence of adverse selection in a Mexican microinsurance market. Using 2011 administrative data from the Mexican MFI Compartamos, we run a natural experiment which exploits a business administration irregularity that provided a shock to the insurance coverage decision-making process. We run four additional tests to determine if deceased borrowers purchased coverage adversely or by chance. With the exception of a test examining unused observables, we find no evidence for adverse selection. This study has implications for microinsurers serving poor households in developing countries, who have limited underwriting capabilities but still desire to deal in simplified low-cost insurance policies with uniform premiums. In chapter 3, we conclude by analyzing whether borrowers use their voluntary purchase of additional life insurance coverage as a signal, by which they indicate to their peers and to their lender that they are responsible, reliable, and capable of repaying debts (i.e., safe borrowers). Again using 2011 administrative data from the Mexican MFI Compartamos, we find that purchasing insurance serves as a signal of borrower quality, as borrowers who signal default less often and receive a larger loan amount from the lender, on average. This was the case for both new and experienced borrowers.




Laschever, Purdue University.

Subject Area


Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server