The Electoral Effects of U.S. Prison Construction

From 1980 to 2000, the United States opened over 400 prisons, a rapid expansion of carceral capacity mirroring the punitive turn in public policy and public opinion. Yet few studies have examined the impact of prison construction on nearby residents and their political behavior. We develop a theoretical model tracing how prison openings affect economic, attitudinal, and electoral outcomes in surrounding communities. Using modern difference-in-difference methods and a novel county-level dataset covering forty years of prison openings, we find that prison construction decreases county-level vote share for Republican candidates, especially if Republicans represented the county during prison construction. Drawing on analyses of economic outcomes and prison bond measure support, we argue that voters sanction issue-owning Republican candidates for negative economic externalities and turn against funding for prisons. These findings challenge conventional wisdom about the political economy of prisons and have implications for how we understand contemporary expansions of carceral capacity.