The impact of college graduation on geographic mobility : identifying education using multiple components of Vietnam draft risk

College-educated workers are twice as likely as high school graduates to make lasting long-distance moves, but little is known about the role of college itself in determining geographic mobility. Unobservable characteristics related to selection into college might also drive the relationship between college education and geographic mobility. We explore this question using a number of methods to analyze both the 1980 Census and longitudinal sources. We conclude that the causal impact of college completion on subsequent mobility is large. We introduce new instrumental variables that allow us to identify educational attainment and veteran status separately in a sample of men whose college decisions were exogenously influenced by their draft risk during the Vietnam War. Our preferred IV estimates imply that graduation increases the probability that a man resides outside his birth state by approximately 35 percentage points, a magnitude nearly twice as large as the OLS migration differential between college and high school graduates. IV estimates of graduation's impact on total distance moved are even larger, with IV estimates that exceed OLS considerably. We provide evidence from the National Longitudinal Survey of Youth (NLSY) 1979 that our large IV estimates are plausible and likely explained by heterogeneous treatment effects. Finally, we provide some suggestive evidence on the mechanisms driving the relationship between college completion and mobility. -- Geographic mobility ; education ; college graduates ; internal migration ; instrumental variables