U.S. flag

An official website of the United States government, Department of Justice.

Incapacitative Effect of Imprisonment - A Critical Review of the Literature

NCJ Number
J Cohen
Date Published
Research on the incapacitative effect of imprisonment is reviewed.
A recurrent theme in the review is the inadequacy of current estimates of the individual crime rate. The various papers discussed typically require only an average rate, yet either the crime rate was underestimated or the authors avoided empirical estimates altogether. Aside from the empirical weaknesses of present estimates, there are also some conceptual deficiencies. The discussion of a possible constraint on the prison population highlights the need to go beyond a concern for population averages. In particular, the ability to discriminate between different individuals' crime rates and a characterization by crime type is important for possible allocation decisions. The best estimates of the individual crime rate should account for variations in crime rate by crime type, variations in crime rate across the criminal population, and variations in crime rate during an individual career. Such estimates of crime rate will require better data on criminal careers than are presently available. The Shinnar et al. model represents the best approach to estimating the incapacitative effect available to date. Not only does it incorporate most of the assumptions of the other authors, but is has the added advantage of providing a means of handling the discounting of crime rates by time served and the possibility of dropout from the criminal career. The model could be improved, however, if it would (1) incorporate heterogeneity in crime rates and the probabilities of arrest, conviction, and imprisonment; (2) explore alternative distributions for the different variables and attempt to validate them empirically; (3) consider the implications of various dependent relationships among different aspects of a career; (4) introduce nonstationarity into the process, particularly variations in crime rates over time and changes in the size of the criminal population; and (5) explicitly incorporate possible constraints on individual criminal justice system parameters or the resulting prison population. References and tabular and graphic data are provided.