NCJ Number
241393
Date Published
June 2009
Length
67 pages
Annotation
This review describes the conceptual foundation of the Donohue and Levitt hypothesis which proposed that legalized abortion played a major role in the dramatic decline in crime during the 1990s.
Abstract
Results indicate that that the most credible tests of abortion and crime in both the United States and abroad do not support the Donohue and Levitt hypothesis. In the end, the simple time-series plots of age-specific arrest and homicide rates tell the story: the crime rates of cohorts born just before abortion was legalized follow the same time path as the crime rates of those born just afterwards. There is no discontinuity in crime rates associated with the early legalization of abortion in New York or California, nor is there a break in crime rates in the rest of the States after Roe v. Wade. This review is structured as follows: Donohue and Levitt's presentation of innumerable regression estimates is analyzed. Then, the evidence that links unwanted childbearing and adverse outcomes is assessed. Next presented are Donohue and Levitt's empirical model and results. Then provided are the descriptions of the published criticisms of Donohue and Levitt's work and responses. The final section examines the association between abortion and crime in other countries. Table, figures, references, and appendix