Abstract:
This paper documents the emergence of a race gap in incarceration after the abolition of slavery in the U.S. Counties that relied more on slave labor incarcerated more African Americans, with no comparable effects for whites. An increase of slave reliance by 10% increases black incarceration rates by 1.8-per-1,000. This effect is associated with an increased use of prison labor. Consistent with this, I show that arrests increase before cotton harvesting and incarceration declines after exogenous shocks that decrease the demand for labor. I find no evidence for supply-side mechanisms, according to which former slaves commit more crimes