The determinants of jail use across large US cities: an assessment of racial, ethnic, and economic threat explanations

While empirical research on jails is not as neglected as it once was, there is no study that systematically tests theoretical assumptions on the use of this type of punishment. The present study fills this gap by examining theoretically based determinants of jail admission rates across 157 US cities. The consensus perspective assumes that the legal order reflects a widely held consensus about social harms that require punishment and that the use of formal crime-control mechanisms is a natural response to infractions of this order. Alternatively, the conflict perspective assumes that the legal order reflects the interests of the powerful and that the size of the state punishment apparatus is a response to perceived threats to these interests. Two prominent threat hypotheses are used to assess this approach. Racial and ethnic threat theories suggest that incarceration is more likely in areas with the most blacks or Hispanics. Economic threat theories claim that the rate of incarceration will be greatest where economic differences are the most pronounced. This study uses classical regression to test these perspectives with data from the 1983 Census of Local Jails. With the crime rate, measures of disorder, the size of the police force, the presence of young males, jail capacity, and regional effects held constant, the results provide strong support for the conflict perspective. Income differences between African-Americans and whites, and the size of both the African-American and Hispanic populations are found to explain variations in jail use across large US cities. The findings supporting the threat hypotheses are strengthened in supplemental analyses that restrict the sample of cities to those with a sizable minority population and another that uses updated data from the 1999 Jail Census. Additionally, some models show a quadratic association between residential segregation and the rate of jail admissions providing some support for the contact hypothesis.