Murder mystery

by Doctor Science

I was working on a post (about school and other shootings) and wanted to check one fact. Result: a couple days of work later, I have found out something really odd about murder in America. I'm nothing of an expert in the field, so I'm hoping that someone who *does* know about it can come along and say what the experts think is going on -- or if I've noticed something that the experts haven't.

I'm talking about two things, possibly related, concerning intimate partner homicide (IPH):

1. 30-40 years ago (and possibly earlier), it was almost as common for American men to be killed by their partners as for women. This is extremely unusual: in other countries and cultures, the rate of IPH for women is two to four times that for men.
a. Since then, the US pattern has changed, so that nowadays many more women are victims of IPH than men are.

Note on the charts: All data in this post are taken from Homicide Trends in the U.S. from the Bureau of Justice Statistics, especially the section on Intimate homicide. Charts with a gray background and yellow margin are taken directly from the BJS report, while those with a white background were created by me based on BJS data. Let me know if I need to show my work in GoogleDocs.

The fall in the number of Intimate Partner Homicides, by gender of victim.

b. This is almost entirely due to a change in IPH among black Americans. The high American rate of husband-and-boyfriend killing was largely a black phenomenon. Since the 70s, the death rate for black men due to IPH has plummeted, faster than the drop in IPH for black women.

Rate of Intimate Partner Homicide, by race and relationship type. Earlier, I had summed the spouse + x-friend rates for each gender, having overlooked that the rates were calculated using different base populations.

Cut for length and images --

2. The overall murder rate has gone down greatly since peaking around 1980 and again in the early 90s. This has been across the board, for all races, both sexes, and most categories of homicide.

Total number of homicides and intimate partner homicides.

The exception is IPH of white women, which has dropped only slightly in the past decades. White women are being killed by their partners at about the same rate as they were 40 years ago, even though the rate of most other types of homicide has fallen. Whatever has caused the decline in murder -- an industrialized-world-wide phenomenon -- it does not protect white women from IPH.

Numbers of Intimate Partner Homicides, by race. Stacking graph so it's easy to see how the IPH number for white women has stayed steady while the numbers for other categories, and for IPH as a whole, have fallen.

As I dug around looking for more info about IPH, one point was emphasized again and again: statistically speaking, men and women kill their partners for different reasons:

Women tend to commit homicide against a partner as a last resort in a long history of abuse ... Men tend to commit homicide based on perceived threats to their power and control

What the researchers I'm quoting and others have found says that women's shelters save men's lives. The availability of shelters, abuse hotlines, and other networks of resources and support means that many fewer women are driven to the last resort of murder, and so many fewer men are killed. It's pretty clear that the scholars think most of these men need killin', but that doesn't mean the women in question deserve to have to become murderers.

So the drop in IPH of males seems to be explicable. What no-one (that I can find with my limited knowledge of the field) seems to have noticed is that the IPH rate for white women has barely budged.

I'm going to stop this post now, so y'all can discuss. Tomorrow (or so) I'll put up my further thoughts about the whys and wherefores.

The Daughters of Danaus, by Bryce Cameron Liston. The Danaïdes were the second most famous husband-killers of Classical mythology (after Clytemnestra), but there are curiously few images of them in the artistic canon, and none that I can find showing them murdering their husbands on their collective wedding-night. It's a curious lacuna, given how fond Western art is of the combination of sex and violence.

Comments

[Jason Kuznicki]

You'll want to check out Betsey Stevenson, "Bargaining in the Shadow of the Law: Divorce Laws and Family Distress."

You're looking at what appears to be yet another hidden benefit of no-fault divorce.

http://bpp.wharton.upenn.edu/betseys/papers/DivorceQJE.pdf

[wj]

It really is a pity that the DoJ report reports on IPH only with absolute numbers. The rest of the report seems to routinely look at numbers per 100,000 population. But here, no normalization for changes in population size.

the other big hole in the report is anything on the race of the killer vs the race of the victim. That is, for example, is a black man more likely to get killed
by a black IP, or by a white IP? That could be a factor, given the rise in interracial relationships over the course of the study. A change which the study appears to cheerfully ignore.

As for why the rate for white females has held steady, I have no clue. But I would not be surprised if something has changed for other groups during the period studied, which had changed for white women previous to the period studied.

The other interesting thing I see in you chart on rates is that, at the beginning of the study, black men were getting killed at twice the rate of black women. Where as by the end, they were substantially lower -- which is where white men are relative to white women.

[Turbulence]

Interesting post Dr Sci. I have a question about the second graph, the (only) one showing rates: what is the population used to normalize the counts? Is it the total population? The population of that gender/ethnic combination? Or the population of that gender/ethnic combo currently in a long term relationship?

I'm wondering if some of the white/black discrepancy might be caused by the shrinking number of stable partners available to black women as mass imprisonment of black men takes its toll.

[geographylady]

Are Hispanics counted as white females? just looking at graph #4 and the white female stagnant numbers.

[Doctor Science]

Turb:

arrgh, you caught me! Thanks. It turns out that I was summing things I shouldn't have been for that second chart, so I put the less-understandable version up instead. Oddly, I can't find *any* IPH rate-by-gender-and-race figures.

[Turbulence]

Ah, thanks for explaining Doc.

I haven't had a chance to look at it in detail, but this study seems to have some interesting information albeit for a smaller time scale and limited to only California.

[Doctor Science]

geographylady:

Hispanics are sometimes counted as White, sometimes as Black. There have been a few state-level studies that include info on Hispanic origin, but nothing national (that I can find).

[geographylady]

yes, there seems to be a lack of data at that level, at least on homicides. Violence is recorded by ethnicity, but not homicides.

[Brett Bellmore]

With regard to that last chart, the http://en.wikipedia.org/wiki/White_American>'white' population of this country increased by about 20% during the relevant period. That means, for the number of IPHs among white women to stay roughly constant, the rate had to decline about 20%

During the same time period, the http://www.blackdemographics.com/population.html>black population increased by roughly 40%, meaning, again, that the rate of IPHs among black women has, in fact, declined precipitously. Just not as much as for black males.

As for what explains this, is it coincidental that the rate of incarceration in this country increased dramatically over the same period? And mostly of men.

Incarceration of women increased during the same period, but not as dramatically.

It's probably hard to get murdered by one's mate, if you're locked up in a cell. Apparently we're jailing exactly the sort of people who would tend to get murdered by their intimates.

[Brett Bellmore]

It occurs to me that, on that theory, IPH really says more about the victim than the perpetrator. Though what it says about the victim may be that they have a really dangerous taste in intimate partners...

[Brett Bellmore]

"What the researchers I'm quoting and others have found says that women's shelters save men's lives. The availability of shelters, abuse hotlines, and other networks of resources and support means that many fewer women are driven to the last resort of murder, and so many fewer men are killed. It's pretty clear that the scholars think most of these men need killin', but that doesn't mean the women in question deserve to have to become murderers."

So, maybe we need some men's shelters, too? To save the women who really need killing?

[Slartibartfast]

So, maybe we need some men's shelters, too? To save the women who really need killing?

If there was a point hidden in that statement somewhere, it's not volunteering to make itself. A little unpacking is in order.

Or not, as you choose.

One thing that I tell myself I've learned over my years of clumsy blog-commenting is that the only person that truly appreciates the cleverness of my comments is me. Don't be (the hopefully former) me, Brett.

[Brett Bellmore]

You've apparently fixated so on the mild snark, that you can't see the point: If women's shelters save the lives of men, would creating men's shelters save the lives of women?

If it helps, we can call them "cooling down centers", instead of "shelters", to preserve the illusion that the guys are always in the wrong.

But, seriously, if we gave men in abusive relationships somewhere to go, would it help the situation any?

[jack lecou]

If it helps, we can call them "cooling down centers", instead of "shelters", to preserve the illusion that the guys are always in the wrong.

Did you actually read the part about the typical dynamics of IPH, Brett?

Shelters work to prevent the death of the male partner by providing an abused woman with an alternative. Support and perhaps an escape route. Flight rather than fight.

But women who are victims of IPH are killed because "men tend to commit homicide based on perceived threats to their power and control."

Can you explain to me why a man who feels murderously upset about threats to his power and control would go to a shelter?

[wj]

Dr. S, did you notice one other oddity in the report: the murder weapons listed (gun, knife, etc.) do not include poison. Which is, at least in popular myth, considered a woman's murder weapon of choice. I wonder why the omission....

[Brett Bellmore]

Geeze, where's an eye rolling emoitcon when you really need one? Did YOU actually read the part about "illusion"?

[Brett Bellmore]

Anyway, setting aside the issue of feminist dogma about all domestic abuse being by men...

Anybody think there might be something to the idea that the huge increase in incarceration helped reduce the rate of IPHs?

[jack lecou]

Yes.

But I'm not sure why you think that helps. I mean, what exactly are you calling an "illusion"?

The vast majority of IPH cases really do result from approximately the same underlying situation: an abusive relationship in which the male partner is 'in the wrong', in your words. Either the man escalates the abuse to the point of killing the woman, or the man is killed because he's backed the woman into a corner from which she feels she has no other escape.

To avert either of those outcomes, again, the vast majority of potential IPH cases, it's going to be a shelter for women that would seem to be in order. A shelter for abusers makes very little sense, whatever name you use. (Not least because male abusers already have plenty of perfectly safe places to go if they'd like to "cool off" rather than escalate.)

Now, there are undoubtedly tens of thousands of cases in which that dynamic is reversed - high-risk-of-IPH situations where women are the primary abuser. And in a sort of theoretical sense, I guess I'd concede that a "men's shelter" might be of some use in those cases (the heterosexual ones, anyway).

But emphasis on theoretical, because the analogy just doesn't really hold. Problem scale aside, the social dynamics of female-on-male partner abuse are completely different, so shelters aren't necessarily on the short list list of remedies.

And, while men are certainly abused by women, and we should certainly be doing things about that, it's no use pretending, "what about the menz"-style, that the problem is approaching even the same order of magnitude as male-on-female abuse. Particularly, I'd guess, when you narrow the comparison to situations where homicide results. The differing nature and scale of the problems, and the vastly different rates of homicide which result, are no "illusion".

[jack lecou]

Or, shorter:

If you implemented measures (men's shelters, whatever) and completely eliminated IPH resulting from female-on-male abusive relationships, would it show up as anything more than a tiny blip on the charts above?

As such, is giving unequal debate time to the abused men issue vs. the abused women issue really "feminist dogma", or just "sane allocation of resources"?

[Brett Bellmore]

"...would it show up as anything more than a tiny blip on the charts above?"

Honestly, I don't know. But I view it as an empirical question, not something to be dismissed out of hand.

It is my suspicion that, in as much as female abuse of males tends to be more psychological than physical, it's incidence is probably underestimated. It being harder to overlook contused faces, than contused psyches.

However, think about that mechanism via which women's shelters are supposed to save male lives: Man abuses woman, who, lacking escape, finally kills the abuser.

So, woman abuses man, who already starts out with a higher proclivity towards violence. A plausible mechansism. The question is whether shelters are the appropriate response, or something else.

[Doctor Science]

wj:

While poisoning is definitely something of a woman's crime -- women are 36% of poisoners, though only 11% of murderers overall -- the number of (known) poisonings in the US is *extremely* low -- a total of 523 cases for the years 1999-2005. Most poisoning victims were either very old or very young, as well.

[Doctor Science]

Brett appears to be describing a hypothetical scenario in which a man responds to psychological abuse by killing his female intimate partner. There is no evidence that this is a statistically significant occurence. If Brett (or someone else) can find an actual report or paper indicating that this pattern is important, link away.

Otherwise, I have to dismiss this argument as a distracting fantasy.

In reality, most men who kill their inimate partners:

- already had a record of abusing them to the point of serious physical injury
- by strangling or choking, in particular
- also abused drugs and /or alcohol
- showed extreme jealousy -- this is the hallmark of the "out of the blue" partner-killer

If you're going to contend some other dynamic is important, you'll have to find some evidence.

[McKinneyTexas]

In reality, most men who kill their inimate partners:

- already had a record of abusing them to the point of serious physical injury
- by strangling or choking, in particular
- also abused drugs and /or alcohol
- showed extreme jealousy -- this is the hallmark of the "out of the blue" partner-killer

One would suspect these markers or something very much like it in men who kill the women in their lives.

Am I reading these charts right? The total number of IPH's is 1500 annually out of 16,000 homicides?

It's pretty clear that the scholars think most of these men need killin', but that doesn't mean the women in question deserve to have to become murderers.

Probably they do need killin'. And it's probably self-defense, so it's not murder. It's justifiable homicide.

[jack lecou]

Honestly, I don't know. But I view it as an empirical question, not something to be dismissed out of hand.

I had a much longer reply, but on second though, I think "dismiss this argument as a distracting fantasy" just about sums it up.

The null hypothesis here is NOT that those killings up above stem as often as not from female-on-male abuse as vice versa. Certainly the utter lack of evidence that a lot of men are killing their partners to defend themselves from psychological abuse doesn't do much to back up that "feminist dogma" comment. The "under-reporting" only gets you so far.

(And not to continue the sideshow, but I'd also note that the whole point of women's shelters, and the reason they mitigate violence women might be otherwise driven to commit, is that they give women options. The kind of options which men, whether abuser or abused, typically retain as a default consequence of enjoying a generally higher level of physical, social, legal and economic power. Abused men might sometimes kill their partners, but the notion that they do so, as abused women may do, primarily because they literally see no other option is not plausible.)

[Countme-In]

I expect this guy could use a shelter:

http://thinkprogress.org/economy/2012/04/09/460917/wisconsin-state-senator-money-less-important-wome/

Probably though, his bill mandates a shelter for all Wisconsin women as part of the male protection racket renascent in that State.

[liberal japonicus]

Anyway, setting aside the issue of feminist dogma about all domestic abuse being by men...

Brett, Dr. Science made no mention of feminism and discussed her points with references and data. While I'm sure the good doctor would say she is a feminist, you don't get to invoke "feminist dogma" to try and undercut her argument as a glib handwaving gesture. If you have something in mind, state precisely what you think it is. If you can't, or are unwilling to, you are engaging in mind-reading without a license.

[Doctor Science]

McKinneyTexas:

And it's probably self-defense, so it's not murder. It's justifiable homicide.

You might well think so -- but AFAICT the data here do *not* include justifiable homicides, which are quiet rare. Very, *very* few homicides ruled justifiable are by women. If you have a better reference or source of data, link please!

[McKinneyTexas]

If you have a better reference or source of data, link please!

I don't. I assumed from the context of the quoted portion that the circumstances were known and justified killing the BF or H, as the case may be, which, seems by definition, to be justifiable homicide.

I guess I'd need to know how it is a BF or an H needs killing but not in self defense.

I checked your link, and those numbers seemed low, an aggregate of 600 or so a year. I thought we had that many in Houston. I checked another set of stats. They came out slightly higher, but not in any significant way, maybe a total of 700 plus.

[Chris J]

Slightly off-topic, but anyone interested in the long historical context of this topic should take a look at Roger Lane's Murder in America. Lane is emeritus professor of history at Haverford College and one of the leading authorities on historical trends of violence (including suicide) in America. (He is also one of the best teachers I ever had.)

[Hartmut]

Concerning simple homicide (manslaughter, justifiable or not) vs. murder: The fact that women tend to go for 'subtle' methods that need preparation is to their disadvantage once it gets to court. A killing in the heat of the moment (the way male murderers tend to paint their action afterwards) often is not treated as murder because of the absence of 'base' motives. If on the other hand the killing was a prepared deliberate act, it is almost automatically classified as murder and if the method was 'subtle' (like posion), perfidiousness gets added to the charges.
Btw, I have heard forensic scientists that believe that the dark figure in poisoning cases is 90% or more, i.e. 9 out of 10 poisonings are not recognized as such even today, if they are correct with their estimate.

[Brett Bellmore]

"If on the other hand the killing was a prepared deliberate act, it is almost automatically classified as murder"

And rightly so; If you can make preparations to kill somebody, you can make preparations to leave. Unless you're chained to a post, self defense and planning to kill somebody ARE mutually exclusive.

[Brett Bellmore]

Anyway, I take it nobody is interested in the surge in incarceration possibly having been responsible for the reduction in IDHs, by putting the sort of people who would be murdered by their partners in jail? The timing is right...

[russell]

The timing is right...

Correlation and causation.

Although, in the example I cite here, most folks see the causation relationship going in the other direction.

[Rob in CT]

Brett,

I'd say that it's plausible that putting more people in jail might have helped lower the number of "IDH" murders. I'm not sure how one teases the answer out from the data.

Speaking of timing, isn't this also a time period during which a major push was made to highlight the evils of domestic violence? I sure remember growing up (born in '76) consistent messages about it. Is it possible that our society has simply done a better job shaming people into not becoming abusers, or mitigating the extent of the abuse? I'm not advancing that theory, but it seems possible...

[Rob in CT]

Of course in addition to the possible "shaming potential abusers" there is also the obvious (and already discussed) impact of women's shelters, abuse hotlines and no-fault divorce.

Incarceration rates may have played a part, but I think the other possible reasons are just as plausible, if not moreso (how does going to jail for possession of pot prevent domestic violence? I don't really see it).

[Slartibartfast]

IIRC global warming causes and is in turn caused by atmospheric CO2.

So it may very well be that we could remedy global warming by adding more pirates.

[Brett Bellmore]

As XKCD famously said of correlation, http://m.xkcd.com/552/>"Correlation doesn't imply causation, but it does waggle its eyebrows suggestively and gesture furtively while mouthing 'look over there'."

[Sebastian]

Decreasing amounts of lead in the environment.

It is my go-to explanation re declining violence when I can't think of anything else. It also has the weird virtue of maybe even being largely right.

[hairshirthedonist]

I think it's plastic bottles reducing testosterone levels.

[wj]

If on the other hand the killing was a prepared deliberate act, it is almost automatically classified as murder.

Doesn't that depend on the kind of perparation?

Yes, if you spend time setting a trap, that's premeditation and preparation. But it would seem that if you merely go out and buy a weapon, you were engaged in preparation, too. Even if you don't immediately use it, but wait until the next instance of abuse.

"Preparation" may mean something different when used by lawyers, of course -- it wouldn't be the first word to do so. But anything that you do to improve your prospects if bad things happen to you would seem to be preparation the way I routinely use the term.

[Phil]

If you can make preparations to kill somebody, you can make preparations to leave.

Eh, this sort of handwaves away that one of the hallmarks of abusive husbands is that they very deliberately shut their partners off from support systems and outside contacts, making them wholly dependent on the man for nearly everything and making leaving incredibly difficult.

Decreasing amounts of lead in the environment.

It is my go-to explanation re declining violence when I can't think of anything else. It also has the weird virtue of maybe even being largely right.

A triumph for the regulatory state!

[McKinneyTexas]

one of the hallmarks of abusive husbands is that they very deliberately shut their partners off from support systems and outside contacts, making them wholly dependent on the man for nearly everything and making leaving incredibly difficult.

Whether it's a hallmark or a common characteristic, a woman in an abusive, controlled relationship has (1) less options and (2) a less than optimum 'thinking things out' environment.

Also, there is the whole pent up anger thing working, which can and does inform the thought process.

"He beat me", standing alone, probably shouldn't be a clean defense to some kind of reduced level of homicide charges, but the jury should be given the option to exonerate if "the totality of the circumstances" justified the act.

[GoodOleBoy]

Turbulence

'I'm wondering if some of the white/black discrepancy might be caused by the shrinking number of stable partners available to black women as mass imprisonment of black men takes its toll.'

Do you truly think this process is reducing the available population of 'stable' partners? I agree that there are far too many men in prison, but are you suggesting we are incarcerating stable partners and less so unstable partners?

[russell]

I bet if we all put our heads together we could imagine reasons for the decline in black woman on black man murder rate that weren't due to some form of social dysfunction or another.

Who knows, they might even be true.

[Hartmut]

Isn't the effect of chronic lead poisoning apathy? Would one not expect less violence in a fully leaded environment?

[Slartibartfast]

Is lead really the issue? Tungsten or depleted uranium (or even sapphire) bullets of pure love would be equally deadly, I'd imagine.

[Turbulence]

Isn't the effect of chronic lead poisoning apathy? Would one not expect less violence in a fully leaded environment?

There's a fair bit of evidence for the lead hypothesis actually. See here for instance.

[hairshirthedonist]

Boy, I hate to even bring up the abortion-rate theory (but I just did).