What Can We Do About Domestic Murders?

       In recent domestic murders in Ontario, it seems that victims knew of impending danger. Several years ago, when Randy Iles killed his estranged girlfriend in the Collingwood area, there was public outcry. The police and courts knew Iles had a record of assault but they could not protect Aileen May, despite her pleas for help. A coroner's inquest made recommendations about offender risk assessment, but four years later, the inquest into the murder-suicide of Gillian and Ralph Hadley restated these concerns. Recognizing that some progress had now been made into risk assessment, the Hadley inquest urged continued development and use of these tools.

       This fall, Amanda Stamp was rescued after being assaulted and held captive by her violent ex-boyfriend. But the shooting death of young Vanessa Bol and the fatal stabbing of Natalie Bobeita in front of her son in November brought home the fact that we still cannot prevent these tragedies.

       In 2003, the Research Department and Ontario Provincial Police developed an evidence-based risk assessment for wife assault. We identified the best predictors and calculated the minimum information needed to predict repeated wife assault. Thirteen questions about the accused, his violence, and the victim's circumstances were turned into a checklist called the Ontario Domestic Assault Risk Assessment (ODARA). Using the ODARA, police and women's counselors can assess the accused's likelihood of another assault. We tested the ODARA against the OPP's current assessment (the DVSR), against another spousal assault risk assessment (the SARA), and against an assessment designed to assess the victim's risk of domestic murder (the Danger Assessment). The ODARA was the most accurate assessment. Surprisingly, the Danger Assessment was the least accurate. Based only on police information, men who later murdered their partners scored up to 2 out of 15 on the DAS, but up to 8 out of 13 on the ODARA, which is in the worst 7 percent of known wife assaulters.

       In another project with the OPP, we studied a large group of Ontario men who murdered their partners. Many of these men had no previously recorded contact with the police. This finding raises two problems. First, if we have very little information about a man prior to his committing a murder, it is hard to study what was different about him. We can't really learn what predicts domestic murder without this information. Some researchers get around this by comparing women who were murdered with other abused women who are still alive.

       These studies show things that seem to put a woman at risk for murder: having children from a prior relationship, and leaving the partner for a new relationship. The second problem with domestic murders being committed by men with no police record is that there would have been no opportunity for the police to assess risk in these cases. If domestic murder happens out of the blue, how can we predict and prevent it?

       Our studies from the 1990's show that violence in relationships is very common - two thirds of grade11 girls in Simcoe County reported being physically assaulted by a boy. But violence in a relationship changes from year to year, and other studies show that a man who hits his wife early in marriage will not necessarily become a repeated wife assaulter. One in ten men identified as wife assaulters in the OPP records had no previous violence, and no risk factors for being violent again.

       So how can we distinguish a future murderer from the many men who assault their partners but do not kill them? Victims themselves are often aware of the problem long before it comes to the attention of the police. This was certainly true of Gillian Hadley and Arlene May. May was described as having told family and friends that Iles was going to kill her. Our research shows that women's fear of future assaults actually predicts repeated assault quite well.

       We are extending this finding in a study with assaulted women. In partnership with hospitals and shelters in our community, we are recruiting women to complete the ODARA and to tell us their predictions of future assaults. We might modify the ODARA clinical interview if it turns out that women's predictions can improve it. Many of the women in this study have not had contact with the police.

       In sum, the police and courts have an important role in assessing risk for repeated violence by known wife assaulters. Domestic murderers who were known to the police would have scored high on the ODARA before they killed their partners. But some murderers have no record, so partnering with battered women is also an essential part of the effort to prevent domestic murder. Our Research team is working with police and women's counsellors in Ontario to improve our ability to predict and prevent domestic violence.