We use data on a panel of states over a 30 year sample to estimate the response of unemployment to military procurement spending. The state panel provides greater variation in both variables and permits us to examine whether responses to procurement spending shocks vary across states. Our main...