A review of the scientific literature on the relationship between alcohol and violence and that between drugs and violence is presented. A review and analysis of three major theoretical approaches to understanding these relationships are also presented. A number of conclusions are reached on the basis of these efforts. First, despite a number of published statements to the contrary, we find no significant evidence suggesting that drug use is associated with violence. Second, there is substantial evidence to suggest that alcohol use is significantly associated with violence of all kinds. Third, recent theoretical efforts reviewed here have, despite shortcomings, led to significant new understanding of how and why alcohol and drugs are related to violence. Fourth, these theoretical models and a growing number of empirical studies demonstrate the importance of social context for understanding violence and the ways in which alcohol and drugs are related to violence. Fifth, the shortcomings of these theoretical models and the lack of definitive empirical tests of these perspectives point to the major directions where future research on the relationship between alcohol and violence, and between drugs and violence, is needed.

Analyses of disparity in sentencing most frequently use general felony court samples. Because of this, little is known about the specific patterns of sentencing disparity among defendants convicted of criminal homicide. Using a sample of defendants adjudicated guilty of homicide in Philadelphia, Pennsylvania, over the period 1995–2000, this research seeks to describe and explain patterns of sentencing with reference to legal and extralegal defendant and case characteristics. The findings indicate that it is primarily legally relevant variables that influence sentencing outcomes in homicide cases; however, some evidence is found for case processing effects as well as for the operation of criminal stereotypes in the generation of sentence disparities. Implications of these findings for research designs and theory development are discussed.

Recent innovations in sentencing policy across the United States reveal a renewed interest in the idea of selective incapacitation of criminal offenders. This is perhaps most evident in the proliferation of “Three Strikes and You're Out” habitual‐offender statutes across the nation. Although the term was first introduced by David Greenberg in 1975, Peter Greenwood and Allan Abrahamse's eponymous 1982 Rand report represents the most fully articulated plan for implementing such a strategy. The report's release stimulated much discussion, because of the AUTHOR'S claims that selective incapacitation could simultaneously reduce crime rates and prison populations. Ethical problems inherent in such proposals as well as methodological inconsistencies in the original research warrant a reexamination of the proposal and of the empirical basis for the conclusions offered therein. Greenwood and Abrahamse's original research is replicated with a representative sample of California state prison inmates (N = 2, 188) in light of these limitations, with specific focus on the methodological issues concerning the construction of the predictive scale. The selective incapacitation scheme advocated by Greenwood and Abrahamse performs extremely poorly in terms of both reliability and validity, thus precluding the implementation of such schemes. The article contains a discussion of other, more ethically acceptable uses of an instrument that identifies “high‐rate” or “dangerous” offenders. In conclusion, some observations on the limitations of incarceration‐based strategies of crime control are offered.

Most prior research on intimate partner (IP) homicide sanctioning is limited by research designs that analyze outcomes only for defendants charged or convicted of IP homicide, or that focus on female offenders. These studies preclude comparisons of criminal justice responses to intimate and non-intimate homicide cases, as well as comparisons of male and female offenders. Using an original data set of homicide cases adjudicated guilty in Philadelphia during the period 1995 to 2000 ( N = 1137), intimate and non-intimate homicide cases are examined in bivariate and multivariate contingency table analyses. Contrary to earlier research suggesting harsher treatment for women convicted of IP homicide, the results indicate that male defendants convicted of IP homicide are treated more severely than female defendants at all examined stages of the criminal justice process. Additionally, male defendants are sanctioned more harshly for IP than for non-IP homicides. Explanations for the divergence of these findings from those of earlier studies are discussed.