The Trouble With Counting The Gays

Last week, Carolyn reported on a new study from UCLA that claims that 3.5% is the new 10%:

Nine million people in the U.S. identify as lesbian, gay, bisexual, or trans, according to a UCLA study released Thursday. This translates to 3.5% of the population as lesbian, gay, or bi and 0.3% as trans. An additional 8.2% have participated in same-sex sexual activity, and 11% acknowledge some same-sex attraction but do not identify as anything.

She pointed out some of the study’s flaws and highlighted some of the negative attention it’s drawing from both LGBT groups and all-around party people like the American Family Association, sharing delightful views like, “Homosexuals represent just 1.7% of populace. Time to stop getting pushed around by such a tiny minority.”

As commenter Gregory McDaniel pointed out, “Since the percentage of the population identifying as gay is so small, the right wing should just relax and say, “giving rights and respect to this small number of people can’t hurt anything”

Since lots of you had criticisms to add to Carolyn’s,  I thought it might be worthwhile to look at what’s going on here from a sociological* standpoint (as this is apparently my thing).

First off, surveys are an inherently authoritarian method of inquiry. By giving people a set of answers to choose from, they say , “I already know something, I just want you to tell me that I know what I know.” Maybe you’ve taken a sexual health survey and been asked if you use condoms and thought, “I don’t know how to answer this.” Because maybe you’ve never had sex before. Or you’re a women who has sex with women. Or you’re in a monogamous relationship. Or you’re trying to have a baby. They’re using condom use as an indicator of safe sex practices when your reality is entirely different.

In her rundown of problems, Carolyn asked if people are “identified by their self-identity, by past sexual behaviour, by sexual attraction, or by some combination of these things,” pointing out that “Sometimes these things are related and sometimes they aren’t and making generalizations is counter-productive. Labeling someone else causes everyone problems.”

This is one of the big issues with studying identities. For every genderqueer warrior, there’s a women who deeply believes that she was created to bake the perfect roasted ham for her eight kids. Social science operates on the idea that no one experience is more valid than another. There’s plenty of room in good social science for a person to say that they weren’t born gay. Or that being a man who has sex with men doesn’t make them anything other than straight.

The problem is that [shockingly] academia and the rest of the world haven’t always run hand & hand. The rest of the world consumes academic research in a myriad of ways, from a Yahoo! News headline to The Rachel Maddow Show to political propaganda and deceptive PSAs.

For example — perhaps you’ve heard of the “culture of poverty” and recognize it as a racist, classist myth that was spread to cut back on government assistance programs. The thing is that Oscar Lewis, the author of the study, was horrified to see what was done with his work. He published the research with the intent that it would be used to help fight poverty but instead it was snapped up by politicians who used it to point fingers at poor people and say, “See! You’re doing this to yourselves!” This is the ugly side of research. You never know who’s going to appropriate your data and use it for their own ugly ends.

So in this case — acknowledging that there’s more than one way to be queer has the potential to open up a Pandora’s Box of problem, though. It leaves room for fluidity that proponents of conversion therapy and other psychos might exploit.

But let’s give people the benefit of the doubt. Let’s focus on the idea that we’re all humans with brains and we shouldn’t always have to worry about appealing to the lowest common denominator. Especially when they’re going around making up numbers for themselves.

Good, reliable data is a useful thing to have in your corner when you’re looking to make change. Dr. Gates, the researcher who published the new study, said he “would like to see questions about sexuality become standard on big government surveys such as the census so that the information can be linked to health, income and other factors, just as demographers can currently study race and gender.”

And he’s got an excellent point. Whereas the gay population number is causing us some issues, tracking these things opens up a new avenue to evaluate intersectionality. By doing research that shows correlations between sexuality and earnings or health, patterns of inequality begin to emerge and are harder to write off. While all this goes back to the problem of self-identification, ask any sociologist and they’ll tell you that gender and race are just as much social constructs as sexuality and we ask about them all the time. So let people call themselves what they want to. There are too many people in the closet to get a reliable tally but maybe by doing research, proposing legislation, and changing the way people feel about what it means to be gay, we’ll get gradually closer to a real number. Just remember what commenter Chandra said, “It doesn’t matter if it’s 10% or 0.001%. Human rights aren’t dependent on how you do your math.”

*Caveat: Oh hello, I’m about a month away from graduating with my bachelor’s degree but that really doesn’t mean anything since there are gaps big enough for an elephant to fall through in in my school’s curriculum. My entire class just found out last week that functionalism is not a thing. Apparently it was laughed out of the discipline 40 years ago? Anyway.

Before you go! Autostraddle runs on the reader support of our AF+ Members. If this article meant something to you today — if it informed you or made you smile or feel seen, will you consider joining AF and supporting the people who make this queer media site possible?

Join AF+!


Laura is a tiny girl who wishes she were a superhero. She likes talking to her grandma on the phone and making things with her hands. Strengths include an impressive knowledge of Harry Potter, the ability to apply sociology to everything under the sun, and a knack for haggling for groceries in Spanish. Weaknesses: Chick-fil-a, her triceps, girls in glasses, and the subjunctive mood. Follow the vagabond adventures of Laura and her bike on twitter [@laurrrrita].

Laura has written 308 articles for us.


  1. Homosexuals should be added to the endangered species list. :D protected and treasured. haha.

    • If it means that non-profit groups will send me free return lables with my name and address and pictures of cute lesbians on them as part of their fundraising efforts, I am wholly in support of this!

      • Maybe that could become another way for Autostraddle to raise money…return labels with cute lesbian pictures and/or cute kitten pictures. I would also support this.

        • Along with laws against hunting and destroying our native habitats. And efforts towards expanding our habitats. So that could mean funding towards AS and amazing programming :D also people can’t hit us with cars anymore.

          there ya go AS, I’ve developed the purrfect plan to save you and all of our lovely kittens <3

          • Don’t forget that you get your own *free* stuffed Homosexual toy and your sponsored Homosexual will even write you.

  2. Congratulations on your up and coming graduation, Laura!! May God bless this new time in your life. :)

  3. Thank you for this. I am in an pretty much never ending snit over the massive FAIL that is the way the news media and politicians interpret scientific studies. They can’t even seem to grasp the simple fact that correlation is not causation, never mind the more complicated issues behind the way some studies were conducted in the first place or the actual implications of various findings.

  4. Am I the only immature one who saw the picture of the Count and automatically thought of “The Count Who Counts Censored” youtube video?

    Congrats on graduating soon! <3

  5. I just finished reading this paper and I can see how easily someone can misinterpret the authors findings. The author CLEARLY states multiple times how difficult it is to capture information that is representative of the LGBT community. His findings are based on a collaboration of multiple surveys to pin point an estimate. It is an attempt, if anything. I think that’s the point people fail to see. Hopefully, this will encourage other researchers to pursue better methods in creating and reaching out to the LGBT community.

  6. The problem with when you go based on how people identify themselves is that you are always going to get less gays than there are in the general population. There are a lot of people who are in the closet or just haven’t figured it out yet. I think it’s particularly the case with bisexuals and with trans people, that there’s a huge gap between the actual numbers and just the numbers of who is out. (Trans because less people understand it and because trans people are even more ostracized than the LGB community is, bisexuals because not many people understand it and because I’ve noticed, both with myself and others, that bisexuals tend to take longer to figure it out and subsequently come out.)

    The other thing is that even that small percentage is still a lot more people than they think. That means that in a 100-kid strong elementary school class, as mine was, about 3-4 kids are LGBT. In a high school class of 500, it increases to 15-20. Still not a lot, but enough that they can’t just shrug us off as people they’re never going to have to deal with. Just about everybody, and especially these days in the age of Facebook, knows an LGBT person, whether they realize it or not.

    • In the study, 1.7% identifies as gay or lesbian and 1.8% identifies as bisexual. This isn’t the first study I’ve seen that cites more people identifying as bisexual than gay or lesbian.

      The feeling of invisibility has a strong effect on perceptions. Bisexuals have never appeared invisible to me and I wouldn’t be surprised by a study that reported 3-4 times more bisexuals than homosexuals.

      • Another thing is that I feel that even a lot of people who start out homophobic and self-loathing who are gay, eventually feel forced to come to terms with their sexuality and come out because they are just are not satisfied with the opposite sex. On the contrary, most bisexuals are satisfied in opposite-sex relationships, and therefore don’t feel like they pay as huge of a penalty for staying in the closet. (I don’t say all because for people who are 4-5 on the Kinsey Scale, this probably wouldn’t be true. Even I’m around a 3, and I feel like while I could eventually settle down with either a man or a woman, never getting to experience one or the other sex at least once in my life would leave me unsatisfied.)

        I also think that particularly for women, homosexuality is often defined more in terms of lack of interest in guys, rather than interest in girls. It’s normal for even straight girls to pay attention to other girls as a way of competing, so I just didn’t think anything of the way I looked at girls – not realizing it was any different. And I liked boys, so I just assumed that meant I was straight. And the new “bi until graduation” and “girl crush” trend among straight girls makes it even more complicated.

        So in general, I feel like homosexuality is a lot more straightforward for some people than bisexuality is, and it makes me wonder if there are particularly a lot of bi people who are closeted or not even aware of it who we’re not seeing in these statistics. I mean, I had my first crush on a girl at age 13 (boy at age 11), I went to a queer-friendly high school where a lot of my friends and classmates were gay or bi and in same-sex relationships, and I STILL didn’t manage to completely figure it out until I was a college sophomore.

      • Also, I was just wondering, what do you mean about bisexuals never appearing invisible to you? Do you mean that you just seem to be good at detecting bi people – kind of like gaydar?

  7. Laura, I liked you before, but now I know you’re a fellow sociology geek it’s x1000.

    Ok, yes functionalism is pretty much no longer in use ’cause if the tautology of functionalism were a parking lot you could park the moon there. It survives a little bit in some modified Durkheimian ways in criminology. It was for several years very important and influential. Really no one in sociology even attempts grand theory anymore, it’s much more focused on micro or midrange theorizing. Actually, of the functionalist stuff that makes occasional appearances it is usually Merton’s mid range theories.

    Sorry for the dork out, guys. If anyone is interested in my sociological response to that article, it’s on the original posting.

    • <3

      Where else could I write a nerdy article like this and get more than 2 comments? You guys are amazing. And thanks for all the nice thoughts about graduation!

  8. One thing that always gets to me when I hear about surveys like this is that I was not a part of them. How can they make a statement about whatever percentage of the population is this-or-that, when I am part of the population and I never took their survey? I know it probably has to do with sample sizes and math and statistics, which I was never very good at anyway, but it still bugs.

    • Ok, so I hope I don’t fuck this up. Any statisticians in the audience, please correct me, but yes, it has to do with statistics and sample size. It is not feasible or realistic to try to seek out an entire population of people. For example, if you wanted to say something about African American’s in their 20’s you couldn’t very well try to find every single one of them to ask your question to. But using a carefully targeted random sampling you can get enough people who will be able to represent the characteristics of the broader population to a satisfactory degree. So, continuing our example, hopefully our sample of 20-something African Americans would have a similar percentage of people living in rural areas versus urban areas as the actual population percentages. Using various statistical techniques the researcher can actually estimate how likely it is that their sample is a good fit for the wider population. Maybe a clearer example would be using gender. So we know that women make up about 50% of the population, so in our sample of 20-something African Americans we would want to see close to 50% women in our sample before we make any conclusions about 20-something African American women.

      In fact, sometimes a good sample can be even better than trying to reach everyone in a population because some people are easier to reach than others, so if you are trying to reach everyone you will get far more ‘easy to reach’ people than ‘harder to reach’ people and that will skew your results.

      • Interesting. Please don’t take this as me jumping on your explanation (which is very handy – the more nerds, the better the world, methinks), but I’d like to point something out for more clarity:

        “So we know that women make up about 50% of the population, so in our sample of 20-something African Americans we would want to see close to 50% women in our sample before we make any conclusions about 20-something African American women.”

        How then do we ‘know’ when it’s valid to say we’ve gotten a large enough sample size to calculate an LGBT population if that ‘guesstimate’ number isn’t really known?

        ((p.s. As if guesstimate isn’t being autocorrected lol))

        • You know I thought about addressing that after I had already submitted my response.

          It’s kinda a two prong problem and exactly why there is so much contention over the numbers. The first problem is what you’ve identified: we don’t have a good population estimate and keep getting wildly different numbers. It would be one thing if we kept doing these surveys and most of them resulted in a similar percentage of LGBT compared to the rest of the sample, then we could more safely estimate the population numbers from our consistent samples. My understanding is that most population estimates on basic demographic stuff are based in part off the Census, which tries to do exactly what I suggested is kind of unwieldy. The Census tries to actually track down every person(or householder) in the population. This is why some people have been pushing to include questions about LGBT people on the Census. And they are not the first group to do so for visibility reasons. For example, I’m pretty sure the Census didn’t start including Hispanic on the form until the 80s.

          The second problem is that LGBT people are ‘harder to reach’ people on top of that. Usually ‘harder to reach’ people are transient or don’t have phones (in the case of phone surveys), but in this case its that people are unwilling or unable to identify themselves as LGBT. Sometimes these surveys are filled out by one member of the household, so maybe your husband doesn’t realize you still identify as bisexual or as Laura pointed out people can be engaging in same sex sexual relations, but they do not identify as LGBT.

          So I guess the short answer to your question is when all the surveys come back with similar percentages, which will probably be when people start more consistently identifying as LGBT.

          • That makes sense, and I completely agree with the difficulties in finding/addressing the right groups to get a good number. Even with LGBT boxes to fill out on surveys and censuses (I want to say “censi” lol), there will always be a lower number than truth.

            Once the actual numbers are pegged, it would be fascinating to see the numbers grow and change after a period of time as people understand their identities more. For now, I guess it’d be a safe bet to go up to the Magic Kinsey 10% value to get a good estimate since that’s the ‘upper-limit’ – better to have a larger sample than a small one.

  9. thanks for the commentary – insightful and based in theory and application – I was just teaching intersectionality in my class yesterday – brilliant thinking and you’re going far I can tell. congrats!

  10. I get so annoyed reading the functionalist sections of my intro to sociology textbook! Last weekend I threw it across the room. Ugh. Anyways, thanks for this lovely article ondate and the social sciences, yay having a brain!

  11. I have a strong suspicion that there exists a genderqueer warrior who also deeply believes that she was created to bake the perfect roasted ham for her eight kids. That’s just plain yummy.

Comments are closed.