26 Mar Case Study Arrest and Domestic Violence

In 1983, preliminary results were released from a study on the deterrent effects of arrest in cases of domestic violence. The study reported that male abusers who were arrested were less likely to commit future assaults than offenders who were not arrested. Conducted by researchers from the Police Foundation, the study used rigorous experimental methods adapted from the natural sciences. Criminal justice scholars generally agreed that the research was well designed and executed. Public officials were quick to embrace the study’s findings that arresting domestic violence offenders deterred them from future violence.

Here, at last, was empirical evidence to support an effective policy in combating domestic assaults. Results of the Minneapolis Domestic Violence Experiment were widely disseminated, in part due to aggressive efforts by the researchers to publicize their findings (Sherman and Cohn, 1989). The attorney general of the United States recommended that police departments make arrests in all cases of misdemeanor domestic violence. Within five years, more than 80 percent of law enforcement agencies in U.S. cities adopted arrest as the preferred way of responding to domestic assaults (Sherman, 1992:2).

Several things contributed to the rapid adoption of arrest policies to deter domestic violence. First, the experimental study was conducted carefully by highly respected researchers. Second, results were widely publicized in newspapers, in professional journals, and on television programs. Third, officials could understand the study, and most believed that its findings made sense. Finally, mandating arrest in less serious cases of domestic violence was a straightforward and politically attractive approach to a growing problem.

Sherman and Berk (1984), however, urged caution in uncritically embracing the results of their study. Others advised that similar research be conducted in other cities to check the Minneapolis findings (Lempert, 1984). Recognizing this, the U.S. National Institute of Justice sponsored more experiments—known as replications—in six other cities. Not everyone was happy about the new studies. For example, an advocacy group in Milwaukee opposed the replication in that city because it believed that the effectiveness of arrest had already been proved (Sherman and Cohn, 1989:138).

Results from the replication studies brought into question the effectiveness of arrest policies. In three cities, no deterrent effect was found in police records of domestic violence. In other cities, there was no evidence of deterrence for longer periods (6–12 months), and in three cities, researchers found that violence actually escalated when offenders were arrested (Sherman, 1992:30). For example, Sherman and associates (1992:167) report that in Milwaukee “the initial deterrent effects observed for up to thirty days quickly disappear. By one year later [arrests] produce an escalation effect.” Arrest works in some cases but not in others. As in many other cases, in responding to domestic assaults, it’s important to carefully consider the characteristics of offenders and the nature of the relationship between offender and victim.

After police departments throughout the country embraced arrest policies following the Minneapolis study, researchers were faced with the difficult task of explaining why initial results must be qualified. Arrest seemed to make sense; officials and the general public believed what they read in the papers and saw on television. Changing their minds by reporting complex findings was more difficult but continues to be important. Long-term follow-up studies have found that arrested offenders were more likely to be victims of homicide (Sherman and Harris, 2013). Even more sobering, domestic violence victims of arrested offenders were more likely than victims of nonarrested offenders to have died within 23 years of the experiment (Sherman and Harris, 2015).

Critical Thinking

1. What was your first reaction to reading this case study? Did you know about the contradictory findings regarding the use of arrest in incidences of domestic violence?

2. Can you think of other topics or claims that you think criminal justice researchers should revisit? What other common practices that are seen as absolutely true should researchers reevaluate?