Tuesday, October 21, 2014

More Men Are Raped in US Than Women?


You may have come across that assertion before if you have the habit of visiting MRA sites on the net.  I've seen it, but never found any sources for it, except for fuzzy hints that it's because of prison rape of men is not included in the general rape statistics.

But the Gamergate led me to the tweets of one of its avid supporters, and the  result is that one of this gentleman's earlier tweets provided me with a source for the view that men are raped more often than women in this country.

The link he so kindly provided is from October 2013 Daily MailThe crucial bit is at the very beginning of the story:

More men are raped in the U.S. than woman, according to figures that include sexual abuse in prisons.
In 2008, it was estimated 216,000 inmates were sexually assaulted while serving time, according to the Department of Justice figures. 

That is compared to 90,479 rape cases outside of prison.

Let's look at those figures.  First, a caveat:  It's possible that there are other sources which support the view that more men than women are raped in the United States (though I doubt that). All I try to do in this post is to analyze one particular source, the Daily Mail one.   Second, it's crucial to note that prison rape is horrible, that men, indeed, can be the victims of rape and that all rape victims deserve our concern and support. But we can do that with actual data, right?

To see what drives that paragraph above, note the following three criticisms:

1.  It compares "sexual assault" with "rape" by giving us numbers about the former for prison and jail inmates while quoting some undefined source on the latter for the general non-institutional population.  Sexual assault in the prison study the Daily Mail article refers to includes not only rape and attempted rape but any kind of unwanted sexual touching. 
The concept of rape does not include that third category.  Whether it includes attempted rape depends on where that 90,479 figure comes from.  Thus, the above paragraph compares apples to oranges in the sense of the acts included under the categories of rape and sexual assault.  The latter is a much wider category.

2.  It's difficult to determine from that paragraph the source of the rape counts for the non-institutional population.  Sexual assault statistics tend to come from two sources:  one consists of reports to the police or other authorities, the other consists of self-reporting by random individuals in the community.  The former gives much lower rates of rape and other forms of sexual violence, partly because many forms of sexual violence are never reported to anybody.

The prison and jail data in this case comes from the latter kind of study:  self-assessment by the inmates.  We should compare that data to similarly created data from the general population, not to reported rates of rape. 
I searched for roughly comparable figures for 2008 and found the following:

The total number of sexual assault victims in the non-institutionalized population was 203,850.  The share of female victims was 164,240 and the share of male victims 39,590.

The figure given in the Daily Mail article for rape (90,479)  is not the correct measure for the comparisons the article attempts to make.  For women that would be 164,240 if we compare sexual assault figures*.  Note that you should not now run to compare that figure with the prison-and-jail victimization figure of 216,000.  That's because of the next criticism:
3.  The study the Daily Mail article  relies on  covers BOTH male and female inmates in US prisons and jails.  It's wrong to assume that all the 216,000 inmates reporting sexual assault in 2008 were men.  The quoted 90,479 figure for general rape rates may also include male victims.  But in any case the comparisons become muddled when female victims in prisons and jails are put into the male victim category.  This is how:

The 2008 prison-and-jail sexual victimization survey tells us that 4.4% of all prison and jail inmates reported being the victim of sexual assault that year.

Unfortunately, I couldn't find such overall figures separately for female and male inmates.  The study does give us data by the gender of the inmate but it's split into categories by the identity of the offender:  either other inmate/inmates or prison or jail staff.  Some inmates reported having been assaulted by both other inmates and the staff.
 If we assume that the last category is the same, percentage-wise, for men and women victims, then my rough calculations suggest that 4.3% of adult men in prisons and 3.0% of adult men in jails were the victims of assault that year.  Compare that with 6.3% of adult women in prisons and 4.2% of adult women in jails who reported that they were the victims of assault that year.

Here's where I began truly questioning the original statement that more men than women are raped in the US.  Now, the original data in the Daily Mail article doesn't compare rape with rape, in any case, but looking at the last set of figures it's  clear that the overall victimization rates are higher for female inmates than for male inmates.

In 2008, the general percentages of sexual assault per 1000 non-institutionalized persons over the age of 12 in the US were 0.3% for men and 1.3% for women.  However hard I try, I cannot make the magnitude ranking of those percentages flip by adding that prison-and-jail data with proper population weights. Because the female victimization rate inside prisons and jails is still higher than the male victimization rates.

Looking at the population inside prisons and jails does decrease the overall difference between female and male sexual assault victimization rates. That's because the male population of prisons and jails is much higher than the female population and because sexual assault is much more common in those institutional settings than in the general population.

But it doesn't flip the percentages.  The Daily Mail article is wrong: Self-reported rates of sexual assault are still considerably higher for women than for men.

The point of this post?  It's always good to know what the data actually tells us.  With the general warning that we should be careful about comparing disparate data sets, collected in different ways, it's pretty safe to believe that more women than men are raped in the United States. 

------
*It's impossible to know if the general population survey questions are exactly comparable to the survey questions the inmates answered or if the possible response biases are the same in the two cases.  All the comparisons should be treated with care for that reason.

To give you an example, one of the sources of the Daily Mail article also addresses the sexual victimization in juvenile facilities in 2011-2012.  The rates of reported sexual assault by other inmates in those facilities were 5.4% for female inmates and 2.2% for male inmates.  The rates of reported sexual assault by staff in those facilities were 8.2% for male inmates and 2.8% for female inmates.

These figures define sexual assault or sexual misconduct by staff as including contact initiated by the inmate or contact without any coercion.  The data on female and male inmates together tells us that 3.5% of the respondents reported forced or coerced sexual contact with staff, 4.7% reported sexual contact with staff without any coercion, force or threat.

The latter type of sexual contact amounts to the abuse of authority by the staff.  But sexual contact of this type is not counted in the sexual assault statistics of the general non-institutionalized population.  If it were, any apparently voluntary sexual contact between the young and authority figures should be included.

Even if we decided to include non-coerced sexual contact with the staff of juvenile facilities in the "prison rape" category, the overall statistics on sexual assault by gender would not flip in magnitude rankings, because the number of juveniles in these facilities is so small.

---
Added later:  Jill Filipovich wrote about similar issues in 2012