Monday, September 14, 2009

Does Violent Crime Disproportionately Affect Women?

"a violent world that disproportionately targets women" - Weer'd Beard said this on a comment on another blogger's website and I had to ask myself: "Is violence disproportionately skewed towards women?" Seems like a rather easy question to answer so I went searching.

The Bureau of Justice Statistics compiles this type of information from their crime victimization surveys. On the most basic level, one could say no based on this graph:

Violent crime (murder, rape, robbery, assault) has been committed against males at a higher rate since at least 1973. However, looking at the graph and examining the underlying data, one can see that males have experienced a sharper decline (63%) than females (52%). Reviewing data from the 2006 and 2007 Crime Victimization surveys one finds that the rate between males (22.5) and females (18.9) has narrowed further. So, in the next 5 years we might see that violent crimes are committed against females at a greater rate than males.

So initially, I'll say that I disagree with Weerd, with some stipulations. 1) I'm looking at data from the US. 2) The trend indicates that Weerd will be right in a few years (so he may just be ahead of his time). If I stopped there, this would be an uncharacteristically short blog. So I won't. Perhaps my initial disagreement is wrong.

Based on the violent crime rates, about 2.7 million are committed against men and 2.4 million are committed against women. While not identical, it should be close enough to find any disproportionate amounts. The two impacts that I want to look at are deaths and injuries. So, let me change my original question to "Does violent crime cause a disproportionate number of deaths and injuries to females?" I'll use WISQARS as my source data for the years 2001 - 2006 (reason being that they have deaths recorded through 2006 and their comparable injury information only goes back to 2001). It should be noted that the death information will be more accurate since it is gathered by counting the total number of deaths. Injuries is done using a survey method and extrapolating to the entire population. As usual, my conclusions will come at the end (hence the word conclusion), and I'll be more than happy to tell you if I am wrong. So, in summary to start, based on the BJS chart, for the present time, I think Weer'd is wrong. Let's see who is right?

Violent crime as defined by the government consists of murder, rape, robbery and assault. WISQARS has a dataset for fatal injuries that has a subcategory of homicide. As part of the non-injury dataset, it has subcategories for sexual assault and other assault. I'll use the sexual assault as a rough approximation of rape injuries. The other assault I will use for the robbery and assault category. These categorizations are not perfect, but we rarely get to do a lab experiment when it comes to crime (those pesky criminals don't always cooperate and it is hard to get volunteers to be the victims). I'll then see how the comparative WISQARS rates match up to the BJS rates. A couple of things that I won't be looking at is the psychological cost of these crimes and the monetary cost of these crimes.

To begin I used the homicide and assault categories to gather information on cut/pierce, drowning, fall, fire/burn, firearm, poisoning, struck by/against, suffocation, and motor vehicle. Then I calculated the comparative rate between the two as a percentage (female rate/male rate * 100). If the percentage is greater than 100% than it disproportionately affects females.

To begin here is the graph of homicides:

The first thing that caught my eye was that only suffocation and drowning had any values above 100%. From a logical standpoint this makes sense. Murder by suffocation and drowning are brute force methods. There are no force multipliers. A rope may be used around someone's neck but this does not necessarily multiply the force. Given the fact that the average male has more upper body strength than the average female and that predators (criminals) tend to target the weak, it makes sense that brute force methods would disproportionately affect women. However, as can be seen from the total, homicide statistics are driven by firearm homicides which disproportionately affect men (five times the rate as women). Looking at the BJS rate comparison from 2005 the comparative rate is 26%. This is very similar to what I found from WISQARS.
Now for the assaults (which would be robbery and assault):
Again, the overall shape is similar to the homicide chart (minus drownings, it is hard to get injured from drowning). Vehicular injuries has a couple of years where women are about equally affected (100%) which is curious to see since, there wasn't the a similar comparative rate for homicides. The total is driven by the Struck By/Against category since they account for nearly 10 times the injuries as the next closest (Cut/Pierce). What these two graphs combined tell me is that Weer'd intention with his comment may be the force disparity that women have in unarmed combat. Suffocation is an indicator of this. Going up to my two assumptions in the preceding paragraph, an average male preying on an average female is going to have a strength advantage. A female preying on another female is going to select her prey so that she has a strength advantage. So if this was Weer'd intention, I would have to agree, however the way he worded it originally doesn't easily lend itself to that conclusion. Looking at the BJS comparative value for 2005 it is 62% when combining robbery and assault numbers. Once again this is similar to the numbers I calculated from WISQARS.

And the last graph Sexual Assualt or Rapes:
The first thing that should be obvious is the scale. Rape and sexual assault disproportionately affect women in a big way. That isn't news to anyone I hope. Further, injuries primarily occur in the struck by/against category. This also shouldn't be a suprise. The BJS found that only about 7% of rapists use weapons during their crime. The vast majority are just predators seeking out easy prey that they can use their strength disparity. Looking at the comparative rate for the BJS it is 1400%. This is higher than the values I calculated, but then again, they are surveying all rapes and sexual assaults while WISQARS is only looking at injuries. Still, as far as the order of magnitude goes, they are very similar.
Overall, violent crime rates are driven by assaults. There are 10 times as many assaults as robberies, 4 times as many robberies as rapes and 20 times as many rapes as murders or 800 times as many assaults as murders. Of the four categories of violent crime, women are disproportionately affected by rape. However, the trend of violent crime victims by gender appears to indicate that we may be seeing more of the violent crime targeted to women in the near future. So in conclusion, I still think Weer'd is wrong, for now.
That all being said, I don't think women (or men for that matter) should sit back and pretend crime doesn't happen. On the reverse side, don't paralyze your life because of what is going on in New Orleans or Chicago. Looking at the Crime Victimization reports can tell you a lot of things.
1) Women are 50% more likely than men to be victims to someone they know (family or friend). Lesson: Know your friends and family.
2) 25-34 year olds are half as likely to be victims of violent crime as 12-24 year olds. The rate drops an additional 25% for 35-49 year olds, an additional 33% for 50-64 year olds and finally an additional 80% for those over 65. In other words, a 65 year old has only 5% of the risk of being victim of a violent crime as a 20 year old. Lesson: Relax as you get older.
3) Never married and divorced people are 3 times as likely to be victims of violent crime as married or widowed people. Lesson: Get married and stay married.
4) Urban residents are 50% more likely to experience violent crime than suburban or rural. Lesson: If you can, get out of the city.
Bottom line. Choose first to not be a victim. Take the necessary steps to not make yourself an easy prey.


  1. The only problem I have with your conclusion is that it doesn't control for various sociological factors.

    The most obvious is the fact that criminals tend to be disproportionately men. Most violent crime (as much as 80 percent by some accounts) is criminal on criminal.

    It also doesn't consider the number of "assaults" that were the result of mutual combat between men. Men, being more aggressive and competitive, tend to be involved in mutual confrontations that can result in criminal charges much more often than women...even if women are often the cause of those confrontations.

    This tells me that there's a good chance that the larger impact of violent crimes on men would be significantly reduced, if not completely eliminated by controlling for these factors.

    Granted, Weerd didn't stipulate, but I think his implication is that violence against innocent victims is skewed toward women.

    I still can't say whether that assertion is accurate or not, but I'd say it's much more probable than the raw numbers indicate.

  2. Sailorcurt,

    The sociological factors were purposely not looked at. I was trying to answer the question of whether women were disproportionately affected.

    I wouldn't jump to the conclusion that violence against innocent victims is skewed toward women. Mainly because of the higher percentage of violent crime that is perpetrated by friends and family of the victim (64% for females, 43% for males). I would hypothesize that criminal on criminal crime is more likely to also be classified as the victim familiar with the offender as opposed to being a stranger.

    The mutual combat issue is interesting. Basically what you have is two arrests for assault as opposed to one arrest. If men were more often involved in these, I believe that would actually show that there are less incidents of assault (since you could count the two as one incident) than previously thought, which would mean that more women are assaulted than men. However, the assault rate for women is 33% lower than for men so it would have to be a significant number of assaults that fell into this category for this to change the numbers.

    Innocent victims also brings up some good questions. When a pimp making a drug deal goes south and he and his prostitute are beat up, does the prostitute count as an innocent victim? What about if she was making the drug/money pass?

    From a logic stand point, I would agree with the assertion that violence against innocent victims is skewed toward women. I'll have to look and see if I can find any stats to back that up. If this is what Weer'd meant, I'll let him speak for himself as I know he reads this regularly.

  3. Wow, I bow *again* to your numbers, Reputo!

    And I indeed had discounted Criminal-on-criminal violence when I made that unqualified statement.

    As a general rule this statement is one I make to women who are anti-gun or afraid of guns (or in the linked case against the Anti-Gun position of most feminist groups) certainly as a man who follows violent crime news, in Mass most of our dead bodies show up in our big cities, mostly Boston (and worse yet the minority districts of Boston) and of these the departed is a gang member with a long criminal history, and if apprehended (there's only a 30% chance of that in Bean-Town) is another gang member.

    Of course both are male, as I am unaware of any gang that recruits women as "Soldiers".

    That being said, I am not in a gang, nor is anybody I associate with. Also gang affiliation is a very easy variable for a person to control. (Ie: if you don't want to be killed or assaulted DON'T JOIN A FUCKING GANG!)

    Of course in no way does simply not being in a gang mean you are safe from violence.

    Of course there is another factor easily controlled, and likely something that will mess with the numbers: Domestic Violence. Easily selected for, by simply not associating with people who feel the need to communicate with their fists. Likely to screw with the numbers because how many assaults are actually reported in this grouping?

    I also don't associate with domestic pugilists, and good thing, as if the police were somehow unable to supply a reasonable amount of safety, I might have to step in, which while moral might be quasi-legal. Thankfully all my female friends who had abusive partners have heeded my advice and got the hell out of dodge and taken legal and physical means of personal defense.

    Again, these two factors don't create safety.

    Again there is the above "crime against innocent victims" as stated. Things like mugging, rape, kidnapping, ect. The type of crime that created those rules of thumb like "never walk alone at night", "avoid dark alleys and unlit streets", "always wear good running shoes when traveling in potentially dangerous areas", "Always be in at LEAST condition Yellow", and "Always carry at least one gun".

    Of course you have exposed that I was simply running off of dogma and haven't looked directed at the data, still the logical mindset is that predatory criminals stalk their victim and select for vulnerability. This would lead me to believe that tall dudes like myself who have been lifting weights to git rid of this soggy gut of mine, and who makes eye contact with every person I cross would not be a good choice, while an elderly man, or a small young woman would be much better odds of minimal physical resistance.

    Would love to see if you can validate that, or disprove it.

    Thanks for the info!

  4. Reputo,

    On the mutual combat, usually only one person gets arrested or charged, unless the other person continues to be combative toward the cops.

    Person A may have been mouthing off, but only Person B gets charged for escalating the situation into a fight. Person A may get ticketed for drunk in public or intox, but rarely are people charged.

    However, I agree that mutual charges are more likely to impact men then women.

    Much like a family dispute, it doesn't matter that both were yelling, only the one using physical force gets the ride.

    From the Bureau of Justice Statistics
    Victim/offender relationship

    Males were more likely to be violently victimized by a stranger than a nonstranger, and females were more likely to be victimized by a friend, an acquaintance, or an intimate.

    During 2005 --

    * About seven in ten female rape or sexual assault victims stated the offender was an intimate, other relative, a friend or an acquaintance.

    * Seventy-four percent of males and 48% of females stated the individual(s) who robbed them was a stranger.

    Intimates were identified by the victims of workplace violence as the perpetrator in about 1% of all workplace violent crime. About 40% of the victims of nonfatal violence in the workplace reported that they knew their offender.

    For murder victims, 43% were related to or acquainted with their assailants; 14% of victims were murdered by strangers, while 43% of victims had an unknown relationship to their murderer in 2002.

    Two thirds of murders of children under the age of 5 were committed by a parent or other family member.

    Intimate violence

    * About 1 in 320 households were affected by intimate partner violence.

    * Female victims are more likely to be victimized by intimates than male victims. In 2005, of offenders victimizing females, 18% were described as intimates and 34% as strangers. By contrast, of offenders victimizing males, 3% were described as intimates and 54% as strangers.

    * The rate of nonfatal intimate violence against females declined by nearly half between 1993 and 2001.

    * Between 1976 and 2002, about 11% of murder victims were determined to have been killed by an intimate.

    * The sharpest decrease in number of intimate murders has been for black male victims. An 81% percent decrease in the number of black men murdered by intimates occurred between 1976 and 2002.