Why Income Enters the Story

This chapter extends the credit-history mechanism from observable borrowing and repayment to private effort. The question is how income can become evidence about a hidden borrower characteristic when the work that produced it cannot be seen directly. That extension matters because reputation may then affect not only the price of credit, but also the incentive to earn income—and income can feed back into the reputation.

Income becomes evidence when effort is hidden

In the earlier credit-market model, the borrower’s type was hidden while saving, borrowing, and default were observable. The moral-hazard extension adds another private choice: at the beginning of life, a person can exert costly effort. Effort raises the probability of receiving a high income in the next period, but the lender does not observe whether the person chose to work harder. The cost is paid now; the possible income arrives later.

This is a different information problem from ordinary adverse selection. Adverse selection means that people know something about themselves—here, their patience—that the lender does not know. Moral hazard means that a person takes an action after entering the situation and the lender cannot observe that action. The two problems interact. A high-patience person values future income more, so they are more willing to pay the current cost of effort. A low-patience person places less weight on the later payoff and is less likely to make that sacrifice.

The income process therefore becomes type-dependent even though type remains hidden. High-patience people are more likely to exert effort, and effort makes high income more likely. When a high income realization is observed, the lender updates the type assessment not only from the earlier asset choice but also from the income outcome. The score now contains information from two channels: what the borrower chose to do with credit, and what income eventually appeared.

The update is conditional rather than automatic. A high-income outcome raises the assessment only to the extent that high types are more likely to produce it than low types after accounting for the earlier asset choice. A low-income outcome can lower the assessment, but it does not prove that the person chose low effort. The lender is comparing likelihoods, not discovering the private action. This is why an observed outcome can be informative without being a direct record of the decision that produced it.

The crucial qualification is that income is an outcome, not a direct observation of effort. A high-income realization may be consistent with effort, but it can also arise from the income process without the person choosing effort. The model uses probabilities rather than certainty for precisely this reason. Income can raise the assessment of high type without proving that the borrower worked hard, just as a low income can lower the assessment without proving that the borrower chose not to work.

In the paper’s model, the income realization also has an ordinary economic value.