Domestic violence arrests may be counterproductive

Women more likely to die in years following partners’ jailings

9:51am, March 6, 2014
Sponsor Message

After a minor assault, a woman who reports domestic violence to the police may do worse in the long run if her partner is arrested rather than warned. Women whose partners get taken into custody even briefly stand a much greater chance of dying early from heart disease and other ailments compared with women whose partners get a warning from police officers, a new study finds.

That surprising trend applies most strongly to employed black women, say criminologists Lawrence Sherman of the University of Cambridge and Heather Harris of the University of Maryland in College Park.

“Black women are dying at a much higher rate than white women from a policy that was intended to protect all victims of domestic violence,” Sherman says.

Although explanations for this fatal trend remain tentative, it’s time to reevaluate mandatory arrest laws for domestic violence, the researchers conclude in an upcoming Journal of Experimental Criminology. A substantial number of U.S. states and multiple jurisdictions across the United Kingdom have enacted such laws.

Working with Sherman in 1987 and 1988, police in Milwaukee randomly assigned 1,125 suspects, most of them men, who had been accused of misdemeanor domestic abuse to receive a warning or an arrest. Those arrested were released between several hours and around a half-day after being booked.

None of the accused people's partners suffered serious physical harm from the violence, which often consisted of shoving and slapping, and 13 percent sought medical care for injuries. Most of the partners were black women in their 20s and 30s.

By 2011, Sherman and Harris found, 91 of the women who had reported domestic violence had died: 70 with partners who had been arrested and 21 whose partners had been warned. Cancer, heart disease and other internal ailments accounted for 60 deaths, 49 of them in the arrest group. Two women in the arrest group and one in the warning group had been murdered.

Of 192 employed black people who had reported violence to the police, about 11 percent died in the years following partner arrests. No employed black person died after partner warnings. Overall, black people’s death rate after partners’ arrests was 98 percent higher than their death rate after partners’ warnings.

White people who were employed at the time of a partner’s arrest showed no heightened risk of dying during the follow-up. Overall, white people’s death rate following partners’ arrests exceeded their death rate following partners’ warnings by 9 percent.

Factors such as obesity and cigarette smoking could not explain higher death rates following partners’ arrests. Sherman and Harris estimate that, had no partner arrests occurred, 40 percent of black people's deaths and 6 percent of white people's deaths would have been avoided.

Sherman suspects that arrests for domestic violence undermined women’s social status and respectability. This may have proven especially true for employed black women, 84 percent of whom were sole breadwinners living in poor neighborhoods. Chronic worry and other stress resulting from a sudden social demotion could have contributed to deaths from diseases, he proposes.

Pockets of severe poverty in Milwaukee’s black areas deserve close scrutiny as possible influences on domestic violence victims’ later health, remarks Harvard sociologist Robert Sampson. Neighborhood characteristics can influence individual outcomes such as kids’ verbal skills (SN: 12/22/07, p. 388) and adults’ imprisonment rates (SN: 9/11/10, p. 9).

Only long-term studies such as the Milwaukee project can identify delayed, unintended consequences of well-intentioned legal policies, Sampson says. Consider another of Sherman and Harris’ findings: Unemployed Milwaukee men arrested — but not those warned — for misdemeanor domestic violence became progressively more apt to reoffend beginning six years after their initial arrests. 

More Humans & Society articles