Lending to Poor Borrowers and Incentives for Micro-credit Agents

This paper focuses on the effect of positive correlation between wealth and reimbursement on a credit institution's schemes to select borrowers and the contracts it offers to screening agents. The paper uses case studies in Uganda, Peru and Guatemala to confirm its theoretical results. Using screening agents to specifically select poor borrowers is difficult when they have a lower probability of success than rich borrowers. The design of an adequate incentive scheme includes the choice between one agent performing all tasks, and several specialized agents, where group incentives can be used. The case studies indicate that agents are offered bonuses when they reach some threshold in repayment rate and number of borrowers. Branch managers also randomly select and assess borrowers' wealth to compare it with the agents' reports. MFIs try to hire altruistic agents who value helping the poor out of poverty. It is, however, not always easy since the prospect of bribes from asset-rich borrowers can also attract non-altruistic agents. Since the agent has a direct incentive to report rich borrowers as poor, MFIs must design his wage to give him correct incentives.