We estimate the plausibly causal effect of fee-for-service (FFS) contracts relative to less retrospective contracts between health insurers and healthcare providers on treatment decisions for deliveries. We find that covering c-sections under FFS increases the c-section rate by 27% while covering vaginal deliveries under FFS has no impact on procedure choice. Effects are explained mainly by payment retrospectiveness rather than payment amount. Impacts of FFS on treatment decisions do not vary by pregnancy risk, but also have no impact on maternal or infant health outcomes. Results constitute evidence of substantial provider moral hazard generated by retrospective contracts.

Draft

We quantify the impact of federal subsidies for graduate medical education on primary care physician (PCP) supply by examining the impact of Section 5503 of the Affordable Care Act, which increased the number of residents that teaching hospitals in rural and high-need areas could receive subsidies for training. Instrumenting for selection into the program using its eligibility criteria, we find that the provision increased both recruitment of residents into primary care and time spent at teaching hospitals in high-need areas, resulting in an increase in PCP supply in treated counties of 6 percent.

Draft