This is likely the most difficult post I’ve written for this space if not, in fact, in any space. The subject at hand is one that in my estimation we’re being asked not notice, but rest assured our opposition notices and it’s better that we address it up front rather than stand accused.
The news this week of the first murder of a trans person in 2018 is, to be very clear, tragic. Every death by violence is too much death and it is well and good that we call their names and magnify their lives and honor their memories. Christa Leigh Steele-Knudslien was killed on January 5, police allege by her husband. Her life was one remembered as one of service to her community, having helped organize the first New England Transgender Pride Event and also beauty pageants for Trans contestants called Miss Trans America.
Naturally the reporting, while noting her as the first (known) trans murder victim cited the 28 (known) such victims across the U.S. in 2017. In cases like this, the overarching theme – if not the explicit statement – of activists organizations implies that there’s an epidemic or crisis of violent murders of trans people. For example this claim in the above linked article:
The deaths of transgender people, particularly trans women, are disproportionately high…with trans women four times as likely to be homicide victims as their cisgender counterparts, according to Mother Jones.
But is this true? A closer look is necessary.
To be clear, it is certainly more dangerous to be outwardly known as trans. Everything from the elevated risk for those for whom discloser results in homelessness, to the difficulty in obtaining quality health care, to the potential for violent assault is elevated. But the difficulty here is that we’re speaking specifically of murder. Which is a very high profile, attention getting subject and naturally attractive to those who’d like to highlight the real and significant problems tat the trans community has to overcome. But the implication that the number of known murders supports the assertion of elevated risk is…problematic.
The best estimate for the U.S. population in 2016 is 323 million. According to data gathered by fivethirtyeight.com, in this century murder has fluctuated between 4 and 6 cases per 100,000 in this century. That works out to around 16,000 such deaths if we take the middle and assume 5/100,000. But the gross number doesn’t tell you as much as the rate.
For well over a decade the most commonly referenced estimate for the number of trans people in this country was .3%, however better work in recent years has postulated a figure as much as three times that. So a fair representation based on current research would suggest a range of 970,000 to 3,230,000 trans people (personally i lean towards the high end but I’m trying to be objective here). So, if we apply the rates identified for the general population of 5 murders per 100,000 individuals to these estimates, then we should expect 48.5 to 159.5 murders annually. I’ll remind you, organizations which fight for trans equality have noted 28 known murder victims. Even if you assume at least one unknown one for every known one (an EASY assumption given how survivors and officials are often given to post-death erasure) you have a figure that is statistically equivilant to the rate of the general population on the low end. You’d need over four unknown victims for every known one on the high end, that’s that just to get to the “normal” rate. Again, I would not dispute such an assumption. I think it’s completely logical to assume the rate among trans people is comparable to the cis population. But you’d need a rate at least twice that if not more before invoking crisis language.
So first takeaway is, be careful not to say or imply that trans people in general are more likely to be murdered. I don’t see statistical evidence for this in what is commonly available.
However, the Mother Jones article mentioned in the quote above makes specific reference to women. 21.3% of 2016 murder victims were female. so out of those 5/100,000 that would be 1.06/100,000. To circle back and apply these to our range fr number of trans people, that would give you a range of 10-34. The MJ article also says 80% of trans victims identify as female, which would mean we’re using 22 as a base of comparison. So if you take the lowest credible estimate you could claim that a transwoman is twice as likely to be killed as a ciswoman (the HRC report that MJ cites says “four times as likely” and I don’t see how they can claim that unless they are using a much lower estimate for the number of trans people) and at the high end of the range you’d find no elevated risk at all.
The only caveat I can see here is that perhaps HRC is using some estimate of OUT transwomen (i.e. the reality that if – for example – 1% of the population IS trans, it’s certain far less than 1% is publicly identified and visible as trans. It’s almost certainly true over half of us remain closeted and thus, we’d have no way of knowing a trans woman was murdered if said closeted person were killed). IF one operates from the assumption that half of trans people are closeted and take the low end of the range you could get to “four times as likely” but at this point you’re getting into some serious weeds.
In any case, if one is going to be publicizing the idea that a transwoman is far more likely to be murdered than a ciswoman, be very sure you say “out transwoman” or your math won’t work.
Further, and here’s where valid claims are most easily found, the HRC report says that 84% of trans victims are people of color. I wrote up an explanatory paragraph on how one would attempt to calculate this comparison and it’s just too wonky and loaded with assumptions (for example, about even distribution) to be readable. However, it appears to be a very safe claim to assert that trans women of color are at a much higher risk of homicide than cis women of color. When you see that claim made thenthey are on solid ground, however the downside of that is that it makes it that much LESS likely that trans white women are being murdered at a rate higher than the cis population because the deaths among WOC are skewing the math. In all frankness, I do not think that it is easy to statistically support the claim about elevated murder rates without these caveats. It does not appear to me, unless researchers are privy to data not easily publicly available, that white trans women are being murdered more frequently than white cis women.
Moreover, as the case of Steele-Knudslien illustrates, not all of such murders are motivated by anti-trans bias which is the sort of violence to which we’re attempting to draw attention. Again I want to emphasize, all life is precious, all death by violence in particular is tragic, none of those lost should be minimized or forgotten or dismissed. But if we go out into the public square and try to advance the narrative that transwomen in general are more likely than ciswomen to be murdered we’re not being completely genuine. Trans women of color are the ones who are at a greatly elevated risk and we need to be careful about our assertions lest the haters undermine our efforts by nitpicking. I can’t take time and space in this column to get into speculation about why trans WOC are at such elevated risk, it’s certainly a complex topic that needs a great deal of study. But you have to first identify the issue before you can address it. Even if that means surrendering a convenient narrative.
Image by Mark Jensen