Articles / Bad stats

Bad stats, advocacy and HIV in trans communities

In April I posted about a study by Stefan Baral and his colleagues at Johns Hopkins that purported to show that “Transgender Women 49 Times More Likely to Have HIV.” Baral acknowledged in a comment (I have no reason to doubt that it’s really him) that “these are pretty bad stats!” but went on to say that he felt his publication would help trans communities in the long run, and pasted the limitations section that he had ignored in his statements to the Huffington Post.

Claudia says that the police in Jackson Heights are much more respectful than in Hunts Point, focusing on keeping the overall area safe. “The men get rowdy, get drunk, and behave awfully. I don’t care if these women are men, nobody has a right to treat them like garbage.” See more photos by Chris Arnade, and read some of the sex workers' own stories.
Claudia says that the police in Jackson Heights are much more respectful than in Hunts Point, focusing on keeping the overall area safe. “The men get rowdy, get drunk, and behave awfully. I don’t care if these women are men, nobody has a right to treat them like garbage.” See more photos by Chris Arnade in his Flickr set, and read some of the sex workers’ own stories.

In theory I agree that it’s worth disseminating potentially inaccurate information if you believe that it’s going to help people regardless. In practice it’s never easy to predict what effect your actions will have. In this particular case, I think that reports like these do more harm than good for a very specific reason: they obscure important differences.

Baral’s study was actually a “meta-study” that combined studies done in cities all over the world. All of these studies investigated non-random samples of transgender people, usually focusing on subgroups that are at particular risk for violence or disease. They all include disclaimers saying “NON-REPRESENTATIVE SAMPLE NO NO DO NOT GENERALIZE RESULTS!!!”

One of the source studies turned out to be the study that I participated in here in New York, led by Larry Nuttbrock of the National Development and Research Institutes (NDRI). In November I criticized the study for overgeneralizing, but as I mentioned in May, the study does reveal some important things, which may very well be found in other communities.

In particular, the NDRI study reveals the deep divide here in New York between the lives of middle-class white trans people and lower-class black and Hispanic trans people. The difference in HIV infection rates (3.5% vs. 48-50%) between the two groups of participants in the study is huge. In another paper, the NDRI researchers indicate that HIV infection tends to correlate with unprotected anal sex, which in turn tends to correlate with gender abuse and symptoms of depression, as well as with attraction to men and nonwhite ethnicity.

What that means, in turn, is that middle-class white trans people like me are not at significant risk for HIV infection, or gender abuse, or depression. To use a fancy social science term, it’s intersectional: if you’re black or Latina, if you’re feminine, if you’re attracted to men, if you’re poor, if you’re a prostitute, it all adds to your risk. And as Hwahng and Nuttbrock observed, it’s all about power: the power to say no to unprotected anal sex.

That’s why you won’t see me using this study to get anything for myself. These high HIV rates don’t apply to me. I don’t deserve any of the money that governments and donors want to spend on it. I don’t know which donors Baral is trying to convince, or what he wants them to spend the money on, but it shouldn’t be middle-class white people.

This is not a problem of “trans women” or even of transgender people in general. It is a problem of disempowered black and Latina transvestites, and it can only be solved by re-empowering them. As Erica so eloquently put it, nihil de nobis sine nobis.

Of course I’m concerned about these high HIV rates, because I care about my fellow transvestites, and my neighbors. That’s why I’m prepared to act in support and solidarity, and I hope you are too. But don’t talk to mainstream transgender organizations that aren’t doing anything about this issue. Talk to organizations that empower sex workers of all genders, like the Red Umbrella Project.

It's only fair to share...Share on Facebook
Share on Google+
Tweet about this on Twitter
Share on Tumblr


  1. Hi Andrea–indeed, the comment was mine! A few things: 1) I have no control over what Huffington Post writes or does not write. We chatted for nearly 45 minutes and the journalist quoted a few lines. That is what journalists do. 2) I still agree with you about all of your concerns (not about the more harm than good thing :)) in terms of the quality of the inputs. This is indeed part of the tyranny of the mean and can only say that we were very deliberate in providing a thorough description of the limitations as I work in a department of epi (at johns Hopkins), teach epi, etc. 3) This was not a study focused on the US–it was intended to provide a global perspective of how little research has addressed the needs and vulnerabilities of trans communities resulting in these communities often being under-represented or not represented at all in programs and policies addressing HIV. And since HIV funding can often be a means to deliver rights-affirming health programming, there is value in engaging. The funders are PEPFAR, the Global Fund, EU, AUSAID among others. I don’t think this paper will have much sway at the city or state level and hope that more locally relevant data will inform these decisions. The PACHA did pass an exciting resolution this year to include more activities for trans population as part of the national HIV/AIDS strategy–but this was independent of this paper though some people from the PACHA have contacted me to hear more. 4) When we started, we had planned to focus only on low and middle income countries, but there was such limited data that we also included US data. But in the paper, we provide clear references to the data used including links–and people who are interested should go and investigate what data informed these analyses as you have done. 5) This article wasn’t intended to “talk to” mainstream transgender organizations–it was intended to be a tool for those who find it useful! In general, I don’t try to talk “to” those who I want to serve as much as I hope to talk “with” them. And indeed, it has served as a tool for some trans organizations around the world. 6) This was my attempt at a product to serve–and about the only thing that I disagree with you about (or at least we have complementary perspectives :)) is that this product has served some…and yes, still believe that it has done more good than harm.

    I have given my perspective on the benefits of this piece of work and would really like to better understand how this article harmed anyone.


  2. Thanks for writing again, Stefan! I agree that people should read the studies that you reference – your compilation is itself a valuable resource!

    I don’t know if your article has harmed anyone, but I think it does muddy the waters too much. Because the samples that these studies are drawn from are so biased towards sex workers, I would like to see something comparing HIV infection rates of trans sex workers with non-trans ones (male bottom and female). How big a factor is transgender expression, actually, compared with ethnicity, income and not having the status of “man”?

    I focused on the NDRI study because I know New York better than anyplace else. It raises the question: if there’s this big a population of not-at-risk middle-class trans people here, why wouldn’t there be in other cities? I don’t believe that there’s an elite or a bourgeoisie anywhere that doesn’t contain a significant number of people with transgender feelings. When necessary, we may build closets to hide in, so that we can keep our day jobs and say no to unprotected anal sex.

    Yet many of the studies you draw on don’t mention any trans people who aren’t marginalized prostitutes. It makes sense because those are the people that they want to help, but then it doesn’t make sense to say “trans women are…”

    All the recommendations in the President’s Council seem reasonable, but I’m not convinced they will do much to solve the problem of HIV infection in trans communities. I’m also not sure what funding from these agencies will accomplish either.

    I appreciate you not taking this personally! I can tell that you and your colleagues are committed to fighting the spread of HIV/AIDS. This problem of bad stats is widespread, and I didn’t intend to single you out. I’m writing multiple posts about your article mostly because you draw on the NDRI study that I know well as a participant and a reader.

    My position on transgender social science research is that we should be trying to get good census data of people who are out of the closet or willing to trust researchers with their anonymity. Until we have that, we should focus on qualitative work. There are plenty of questions that deserve to be explored through ethnographic means. It may not make quite as splashy headlines as your odds ratio, but then again it may if there’s a compelling story. Thanks again.

Comments are closed.