Data and Code for: Do Immigration Raids Deter Head Start Enrollment?

We study whether raids lead to a decrease in Hispanic Head Start enrollment. Specifically, using a nationwide dataset of immigration raids from 2006 through 2008, a period of intensified interior enforcement in the U.S., we provide the first large-scale evidence of the causal impact of local raids on a community. After applying a flexible matching strategy across counties with differential exposure to raids to create analysis samples, we employ a triple-difference strategy to disentangle enrollment changes from leaving a community (mobility effect) versus disengagement from services (deterrence effect). We find robust evidence that immigration raids decrease Hispanic enrollment in Head Start by 10 percent. Further, by comparing this effect to Hispanic students enrolled in first grade, we find evidence that the decrease in Head Start enrollment is driven by a deterrence effect. In other words, after a raid occurs locally, our findings suggest that families are staying in their communities but not enrolling their children in Head Start.