Abortion as Crime Control

Abstract

Donohue and Levitt have applied the selective incapacitation hypothesis to account for what they contend is a latent function of the Supreme Court's 1973 decision in Roe v. Wade to legalize abortion—the decline in the crime rate during the 1990s. They contend that abortion, insofar as it removes recidivists from the population before they are born, has the compositional effect of lowering the overall rate of crime within the country. In this article, their thesis is indirectly evaluated by estimating interrupted time series models of the impact of the Roe decision on several nativity time series for two categories of women Donohue and Levitt identify as being disproportionately at risk for giving birth to high-rate offenders. None of the ARIMA transfer function models lends any credence to Donohue and Levitt's perspective.