NCJ Number
205180
Journal
Analyses of Social Issues and Public Policy Volume: 3 Issue: 1 Dated: 2003 Pages: 45-64
Date Published
2003
Length
20 pages
Annotation
Using data and a research design that aims to avoid many of the complications faced by similar previous studies, this study examined the evidence for or against any empirical relationship between the legalization of abortion under the 1973 Roe v. Wade U.S. Supreme Court decision and subsequent crime.
Abstract
The study used a quasi-experimental perspective and an interrupted time series design as an alternative to observation studies analyzed with regression-based causal modeling. The unit of analysis was the United States as a whole. The analysis used annual data on the number of homicides beginning in 1970 and ending in 1998, allowing for a roughly equal number of years before and after the time when a change in homicide could be expected. The years before the intervention's possible effect served as the control group, and the years after served as the experimental group. The number of homicides each year was the response variable. Homicide was selected as the crime for analysis because it is the most accurately reported crime and the crime that State penal codes view as the most serious. U.S. mortality data were obtained from the National Center for Health Statistics. Included in the analysis were all deaths classified as homicides according to the International Classification of Diseases. It was possible to obtain homicide breakdowns by age, gender, and race/ethnicity. The study found that the legalization of abortion was associated over a decade later with a gradual reduction in the homicides of White and non-White young men. The effect on the homicide of young women was minimal. The study concludes that the 1990's decline in the homicide of young men was statistically associated with the legalization of abortion. The findings were not consistent with several alternative explanations, such as changes in the crack cocaine drug market. Still, the authors acknowledge that there could be alternative explanations that fit the data, given the myriad of factors that can affect homicide rates. The authors further advise that the purpose of this and similar research studies is not to justify legalized abortion as a means to control crime; however, policymakers should still be aware of all of the implications and consequences of policies pursued. 3 tables, 2 figures, and 26 references