Exploring the nexus between certainty in injury compensation and treatment selection

We study the effect of reduced medical liability due to the implementation of scheduled damages on the overuse of cesarean sections. Using data from inpatient discharge records on deliveries in Italy, we exploit the fact that hospitals are distributed across court districts and that only some courts introduced schedules during the period of observation. This allows us to identify the effect of a decrease in liability using a difference-in-difference approach while minimizing the heterogeneities between treated and control hospitals. We show that decreased medical liability increases the incidence of unnecessary cesarean sections by 7 percentage points, which corresponds to a 20% increase at the mean of cesarean sections. The magnitude of the response is higher for hospitals with lower quality and that are far from consumer association headquarters. Lower schedules and higher levels of reimbursements per delivery also increase the overuse of cesarean section. The analysis of the response times, combining the difference-in-difference approach with a regression discontinuity design, shows that the response to decreased liability is already detectable in the short run. Our findings are robust to several sets of robustness checks and are not driven by anticipatory effects or a change in the composition of the treated patients.