Why Men Right's Activists Prefer Data From Before 1990

In the comments to an earlier post, Brad Benjaminson (who doesn’t identify as a men’s rights activist, but tends to cite writings by MRAs) cited several articles he thought of interest. I read the title of one – “Wives Also Kill Husbands-Quite Often” – and before I even saw the date the article was written (1994), I knew the article would use data from before 1990.

How did I know? Because I’ve read a lot of men’s rights articles about “intimate partner homicide” (that’s murdering a spouse, a girlfriend or a boyfriend), and nearly all of them use pre-1990 data. For instance, a quick search of two MRA (men’s rights activist) websites – Men’s Network.org and MenWeb – found seven articles arguing that women are about as likely as men to commit intimate murder. All of them used data from before 1990 to make their case. In fact, almost all of them used the same data set – a Bureau of Justice Statistics study of intimate homicide in 33 of the 75 largest-population (i.e., urban) counties, which was published in 1994 but used data gathered in 1988. The BJS has published more recent work – so why do the MRAs return to this one source over and over? (Or, if not this source, sources that also used urban data from before 1990?)

Because they want to prove – despite clear data, like these recent FBI figures, showing men are far more likely to murder wives and girlfriends than vice-versa – that men are “equal victims.” (This relates, I believe, to a larger project of trying to show that patriarchy doesn’t exist, women have nothing to complain about, etc.)

So what’s special about Urban data from before 1988? Check out these charts (source), both featuring more recent homicide data than the data the MRAs highlight:

This graph shows the reality: although there have always been more women murdered by intimates than vice-versa, the numbers used to be closer. In particular, there’s been a huge decline in male victims – which, unsurprisingly, isn’t something that MRAs with an ideology of male victimhood want to admit.

So that’s why MRAs avoid recent homicide data. Why do they prefer urban data, rather than countrywide data?

As you can see, before 1988 or so black husbands were more likely to be murdered by wives than vice-versa. The BJS data set the MRAs like to use, contains data from spousal-murder cases in 33 urban counties in 1988. In that data set, “Blacks comprised 55% of the 540 defendants, and whites comprised 43%. Among husband defendants 51% were black and 45% were white. Among wife defendants 61% were black and 39% were white.”

So using out-of-date urban data enables MRAs to use a historic anomaly – the high rate of husband-murder among blacks before 1988 – and pretend it represents the norm.

* * *

So why have husband-murder rates been dropping faster? Obviously, there is no one simple answer: but part of the answer is that abused women now have more resources. “Studies of homicides between intimates show that they are often preceded by a history of physical abuse directed at the women, and several studies have documented that a high proportion of women imprisoned for killing a husband had been physically abused by their spouses… the weight of the available evidence shows that often wives kill their husbands in the context of a history of wife abuse.” (Mercy, J.A. & Saltzman, L.E. “Fatal violence among spouses in the United States, 1976-85” American Journal of Public Health 79(5): 595-9 May 1989)

Many of these studies have found that wives who kill their husbands often felt “hopelessly trapped” in an abusive relationship. Therefore, it seems possible that the growth of resources for abused women since 1970 has made a significant number of such wives feel less “trapped,” hence reducing the murder rate of men. To test this possibility, Browne & Williams looked at state-by-state spousal murder rates compared to a “Resources for Abused Women Index,” (availability of shelters, hot lines, support networks, etc, in each state), after controlling for demographic variables (such as the higher general murder rate in many southern states). (Browne, A. & Williams, K. R. “Exploring the effect of resource availability and the likelihood of female-perpetrated homicides.” Law and Society Review, 23, 75-94, 1989.)

The study found that “the Resources for Abused Women Index, although negatively correlated with rates of both types of partner homicide, is more strongly correlated with female-perpetrated than with male-perpetrated homicide…. Moreover, such resources were associated with a decline in the rate of female-perpetrated partner homicide in 1980-1984 compared to 1976-1979.”

So it seems that, thanks to feminism, abusive men may now be less likely to be murdered by their wives.

It’s also possible, that if battered black women (on average) had less access to resources to get themselves out of abusive relationships, that could explain the unusually high rate of black husbands murdered before battered women’s shelters became (relatively) common.

Another question: Why has homicide of white wives declined while homicide of white girlfriends hasn’t? I’m not sure what explains the racial difference, but one factor contributing to the girlfriend/wife difference is the emergence of no-fault divorce. According to a paper (.pdf link) by Betsey Stevensen of Harvard and Justin Wolfers of Stanford, no-fault divorce signficantly helps women in bad marraiges. From an article written by Wolfers:

The findings reveal that under no-fault laws a wife can threaten to leave an abusive husband, and this becomes a credible threat. Under the old regime, this was not so. Our theory is that the fear of divorce creates a strong incentive for abusive partners to behave.

More generally, easy access to divorce redistributes marital power from the party interested in preserving the marriage to the partner who wants out. In most instances, this resulted in an increase in marital power for women, and a decrease in power for men.

Our analysis of US data revealed the legislative change had caused female suicide to decline by about a fifth, domestic violence to decline by about a third, and intimate femicide – the husband’s murder of his wife – to decline by about a tenth.

Unfortunately, as “marriage movement” and men’s right activists have become more influential in recent years, there has been a movement to defund battered women’s shelters and to repeal no-fault divorce laws. Either of these changes would be incredibly harmful to the interests of battered women.

* * *

(Below the fold are links to the seven MRA articles I looked at, with the relevant bits quoted).

Here are links to the seven articles I looked at, all of which used pre-1990 data to make their points, and most of which used data drawn from urban areas. These articles make many additional claims, which I don’t cover in this blog post; many of them, however, are discussed in this earlier post about “husband-battering.”

1. From “Husband Battering” by David Goss: In 1958, an investigation of spousal homicide between 1948 and 1952 found that 7.8% of murder victims were husbands murdered by wives, and 8% were wives murdered by husbands (Wolfgang 1958). More recently, in a study of spousal homicide in the period from 1976 to 1985, it was found that there was an overall ratio of 1.3:1.0 of murdered wives to murdered husbands, and that “Black husbands were at greater risk of spouse homicide victimization than Black wives or White spouses of either sex” (Mercy & Saltzman 1989).
2. From “Domestic Violence and the Demonising of Men”: If we consider the most extreme form of physical violence – murder of one spouse by another – it is apparent that women are almost as likely to kill as a man. Of urban spouses convicted of murdering a spouse, 41 per cent are wives (Bureau of Justice Statistics, 1994.) [Note: the 1994 BJS report cited uses data from 33 urban counties collected in 1988. –Amp]
3. From “Family Violence” (also available here): Men and children may not report when they are injured by a woman, however, the dead bodies of the men and children who are the victims of violent women are usually reported. Murder statistics are far more reliable than reported abuse statistics. The Bureau of Justice Statistics released a report of family homicides in 33 urban counties [using data collected in 1988 – Amp]. Some gender activists claim that violent women are acting in self- defense. These quoted statistics represent convictions for murder.
   1. “In spouse murders, women represented 41 percent of killers.”
   2. “In murders of their offspring, women predominated, accounting for 55 percent of killers.”
   3. “Among black marital partners, wives were just about as likely to murder their husbands as husbands were to murder their wives: (47/53)
4. From “Assaultive Girlfriends”: In July 1994 the Bureau of Justice Statistics of the U.S. Department of Justice released a Special Report detailing the results of a survey of family homicides in 33 urban U.S. counties. [What a surprise – the same data source, with data collected in 1988, shows up again. –Amp]
5. From – ironically – “Latest Research Findings”: Some of the best data on serious assault by intimates appears to be homicide data, which suggest that four out of ten intimate homicides are of men. (Mercy 1989, Langen & Dawson, 1995, FBI Uniform Crime Reports). [That 1995 article, it turns out, uses BJS stats collected in 1988 from 33 urban counties – just like the previous three references. And although he cites FBI data, current FBI numbers show that about two out of ten intimate homicides are of men. -Amp]
6. From “Domestic Violence: A Two-Way Street”: Nor do husbands murder their wives significantly more than wives murder their husbands. A 1994 Department of Justice study [Yup! The same one! –Amp] analyzed 10,000 cases and found that women make up over 40 percent of those charged in familial murders.
7. Finally, this long list of citations include three that discuss spousal homicide. Of the three, the one with the most recent data is a report “on homicide rates in St. Louis from 1968-1992.”