Working Paper: CEPR ID: DP10622
Authors: Sebastian Findeisen; Dominik Sachs
Abstract: We study the optimal design of integrated education finance and tax systems. The distribution of wages is endogenously determined by the costly education decisions of heterogeneous individuals before labor market entry. Consistent with empirical evidence, this human capital investment decision is risky. We find that an integrated education and tax system in which the government provides education loans to young individuals coupled with income-contingent repayment can always be designed in a Pareto optimal way. We present a simple empirically driven application of the framework to US data in which individuals make a college entry decision. We find the optimal repayment schemes for college loans can be well approximated by a schedule that is linearly increasing in income up to a threshold and constant afterwards. So although the full optimum could lead to complicated non-linear schedules in theory, very simple instruments can replicate it fairly well. The welfare gains from income-contingent repayment are significant.
Keywords: education; education finance; optimal taxation; student loans
JEL Codes: H21; H23; I21
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Integrated system providing education loans with income-contingent repayments (H81) | Achieve Pareto optimality (D61) |
| Integrated system providing education loans with income-contingent repayments (H81) | Manage tax distortions across different education groups (H31) |
| Optimal repayment scheme for college loans (G51) | Approximated by a linearly increasing schedule up to a certain income threshold (H31) |
| Transitioning from non-contingent loan repayment system to income-contingent system (H81) | Significant welfare gains (D69) |