This paper investigates the effect of public spending on health and welfare on total, male and female suicide rates using panel data for U.S. states over the time period 1982-1997. We show that the share of health and welfare in total public spending are strong predictors of suicide rates and yield estimated coefficients that are both statistically significant and economically meaningful. In parsimonious specifications, we also find that suicide rates are systematically higher in states with higher divorce rates, but average income level, income inequality and unemployment rates do not have a robust impact on suicide. The model provides a better fit to male than to female suicide data. Our main results hold up to a series of specification tests.
展开▼
