In microfinance and similar relational lending situations, circumstances often make the external enforcement of contracts either unavailable or too costly. Instead, in maintaining loan repayment incentives, lenders depend on reputation. With competition, maintaining this reputation-based incentive structure becomes increasingly difficult. As more lenders enter the market, the incentive to maintain a sound reputation with any single lender declines. Borrowers increasingly engage in strategic behavior and renege on loan contracts. The dissertation explores how the introduction of credit bureaus and similar information sharing mechanisms constrain opportunistic behavior.
Using experimental methods, two information regimes are compared: private reputation, as observed in relational lending, and public reputation, as exemplified by credit bureaus. Game theory is used to model lender behavior, given asymmetric information about borrower types (conditionally honest and strategic) and to derive predictions about market breakdown and the extent of credit rationing. The central determinants of willingness to lend are the beliefs about the expected proportion of honest borrowers in the pool and the variance of these expectations. Experiments are conducted to observe actual behavior.
Results show a connection between information regime, lending frequency, loan size, and repayment. Public reputation leads to less credit rationing (larger loans), lower frequency of (exploratory) lending, and a smaller lender share in surpluses generated. Stronger relationships, measured by multiple repeated loans with the same lender, also lead to a higher likelihood of repayment.