The life-saving effect of hospital proximity.

We provide a new assessment of the effect of hospital proximity in an emergency situation-road-traffic accidents-exploiting the exogenous variation in the proximity to cities that are legally allowed to have a hospital on the basis of their population size. Our instrumental variable results show that a one-standard-deviation increase in the distance to the nearest hospital (5 km) raises the fatality rate by 13.84% at the sample average. This figure is equal to 0.92 additional deaths per 100 accidents. We show that both ordinary least squares and difference-in-differences estimates, common approaches in the literature, provide a downward-biased measure of the true effect of hospital proximity because they do not fully solve spatial sorting problems. Proximity is more important when the level of road safety is low, when emergency services are less responsive, and when the nearest hospital has relatively low quality standards.