Monday, October 06, 2014

Why Domestic Violence Prevention Programs Don't Work. My Comments on Tony Dokoupil's Article With That Title.

NBC News website adds another opinion piece to the NFL domestic abuse debate.  This one is headlined:  Why Domestic Violence Prevention Programs Don't Work.

Now take a breather and imagine some other headlines in the same family:  Why Twelve Step Programs Don't Work for Alcoholism.  Why Most Diets Don't Work For Permanent Weight Loss.

And sending people to prison doesn't work terribly well in stopping recidivism.

I probably shouldn't focus on that headline, because those who write the headlines are not those who write the articles.  The writer of the article puts the clothes on the topic, the headline writer puts on the clickbait hat, and often the hat is screamingly inappropriate for the clothing.

It's somewhat inappropriate in this case, because Tony Dokoupil's article is about two different programs:  First, those which aim to change the behavior of someone who has been sentenced for domestic abuse, and, second, general attempts to prevent domestic abuse in the next generation of men.

Dokoupil argues that neither one of those works.  Indeed, he is skeptical about the idea that misogyny or sexism (the view that abuse is based on feelings of entitlement about women and women's obedience) has that much to do with heterosexual men's abuse of their intimate partners:

But although ending sexism is a welcome goal, there’s little evidence to suggest that it’s the root cause of domestic violence, or that combating it with slogans and applause lines will put abusive men on a gentler path.
In 2009 Oxford University Press addressed this evidence-hole in “Intimate Partner Violence,” a 572-page compendium of what we know and how. In the section on big picture theories, three prominent scholars, including one from the Centers for Disease Control and Prevention, called the lack of support for the sexism theory and other top-down notions of change “perhaps the biggest gap” in the field.
Even more troubling is what seems to happen when sexism is used as the basis for treatment. The short answer is nothing, according to Washington State’s Institute for Public Policy. The team of taxpayer-funded researchers recently reviewed all the literature on the sexism model. The impact on an abusers' likelihood of re-offending: “not significantly different from zero.”

Now enter my world.  If I were a perfect blogger I'd go and study all those references Dokoupil gives, then take a few years to come up to speed on the whole field of abuse prevention.

I can't do that.  The alternative I'm left with is to wonder what "sexism" means in this context:  Does Dokoupil mean that talking about sexism to sentenced offenders won't change their behavior?  Does he mean that sexism in itself has nothing to do with the predominant patterns of intimate partner abuse we see?

If the former, I wouldn't be surprised.  What I've spotted online is that people who get accused of sexism tend to become more entrenched in their attitudes, more convinced of the evil plots of feminazis and the general perfidy of womankind, more -- well, sexist.

The latter is a much wider question and much, much harder to study, because what might be meant with the term "sexism" can vary tremendously.

Are we speaking about the traditional social norms which expect that the man in marriage is the leader and the woman the follower?  Are we speaking about the popular culture views about what sex means for men and for women?  The sado-masochistic gendered messages of some pornography?  The religious rules of the Abrahamic religions which, literally taken, appear to allow the physical disciplining of women or at least demand silent subjection from women?  The way "masculinity" is defined as aggressive and dominant, the way teenagers in some cultural sub-groups clearly "other" women (or "slutty" women) and regard them as prey?

The whole article confuses me, as does so much else in the lightning-speed way information and misinformation is now available to us.  Do you want to share in some of that confusion?

First, take this snippet:

The programs—a blend of treatment for abusers and education for men and boys—are new by the standards of social science and psychology. They appeared in the 1970s, multiplied in the 1990s, and now form what amounts to a second front in the public health fight against domestic abuse.
The first front focused on women. It meant shelters and hotlines, all of it supported by the Violence Against Women Act of 1994. A link has never been authoritatively made, but something in this round seemed to work: rates of domestic violence fell by more than 50 percent between 1995 and 2004, according to Justice Department figures.

In other words, the article headlined "Why Domestic Violence Prevention Programs Don't Work" then seems to point out a tremendous drop in domestic violence!  Perhaps that drop wasn't caused by the shelters and hotlines (though they certainly made continued domestic violence more difficult and allowed the victims some ways out) or the Violence Against Women Act (which offered resources for attending to the kind of crimes which fall under the rubric domestic violence).  Perhaps the drop was because of rising awareness about the problem and changing cultural mores?

Second, this closer analysis of one study is commendable.  We should analyze all the studies which make waves, to make sure that they withstand proper criticisms:

Still, some researchers insist that treatment can work. In the 1990s, for example, Edward Gondolf, now a professor emeritus at the Indiana University of Pennsylvania, followed hundreds of abusers as they entered intervention, a four-month to one-year course of weekly group counseling. Four years later, he and colleagues declared success, citing rates of re-assault as low as 10 percent in a given year.
“There is hope!” Michigan State psychologist Cris Sullivan wrote in an email, citing Gondolf's work. But in recent years, the Gondolf study has been excluded from major policy reports, discounted because of methodological flaws and a rosy interpretation of the data. His final figures did not include drop-outs, for example, about 50 percent of the starting population. And when all assaults were factored in, across all four years, the recidivism rate jumps to near 50 percent.
Though I'd like to ask what the recidivism rates are without any kind of treatment.  If they are much higher than 50 percent, the treatment could still be of value.

Third, I agree that the anger management approach, criticized by Dokoupil, is at most a very partial approach, because if the problem was with just anger management, domestic abusers would also be workplace abusers and get into physical fights with their friends all the time.  This doesn't appear to be the general pattern of behavior which suggests that the perpetrators can manage their anger quite well in many contexts.

Fourth, the terms used in the article are confusing.  "Domestic violence" is not the same thing as "intimate violence in heterosexual relationships."  The former includes elder abuse if several generations live together, child abuse by adults in the family, including by the mothers of the children, abuse of non-elderly parents by adult children, sibling abuse and so on.  It includes intimate partner abuse when the partners are in the same household unit, but intimate partner abuse can reach across family units when the partners don't live as a couple.

The terms matter, because sexism probably isn't the reason why fathers abuse their sons or mothers their daughters or why anyone would abuse a grandparent.  And as I'm not sure how Dokoupil's links use the terms I remain confused.

Finally, consider the global prevalence of domestic abuse of several types and its long history as something which was assumed to be a family's private business unless it resulted in deaths.  Parents were allowed to "correct" their children physically, husbands were allowed to demand obedience from their wives and so on.  If all that history and all those beliefs are viewed as a mountain, then the various treatment and prevention programs are like sending a dozen ants to take down the mountain.  Even if the programs worked we'd see very little change in the size of the mountain.

That's not to argue that the programs are efficient, but to point out the scale of what we are combating and the possibility that visible change might well be of a generational nature.