Unemployment and Imprisonment

Abstract

In recent congressional testimony, Professor M. Harvey Brenner referred to a number of studies which attempted to gauge the relationship between prison admissions and the economy.' Specifically, between 1967 and 1974, Cox and Carr observed that the Georgia prison population fluctuated with changes in the unemployment rate.2 Similarly, Brenner and Jankovic found that admissions to state prisons correlated positively with the unemployment rate.3 Indeed, Brenner testified that for every 1 percent increase in the unemployment rate, state prison admissions rose by 3,340 inmates, even after controlling for the effects of time, the percent of males in the population, per capita income, and inflation. However, Brenner and other researchers have failed to control for practices which could explain the variation in prison admissions. This note examines whether practices, rather than unemployment, explain these fluctuations in prison population. In an effort to test the sentencing thesis, this note replicated an internal staff study by Colin Frank, a former employee of the Bureau of Prisons.4 Frank's study indicated that the quarterly unemployment rate of males age twenty and over explained 59 percent of the variation in the federal prison population sentenced between 1952 and 1974.5 However, Frank's study failed to introduce variables into the analysis. Data from the Administrative Office of the