Numbers Racket

A report last week from the Department of Health & Human Services/Centers for Disease Control finds that only 1.6% of Americans identify as gay or lesbian and 0.7% identify as bisexual, meaning just 2.3% of the population identifies as LGB (T’s were not included). The findings are based on data from the 2013 National Health Interview Survey.

In past years, such low numbers would result in considerable pushback from LGBT activists, who have long bandied about the figure of 10% for gay America, citing Kinsey’s research in the 1940s, to the consternation of social conservatives. Others have put the number at around 5%, and indeed exit polling typically places the self-identified LGB vote around that figure.

Maybe because the new stats come from the Obama administration, the response has been…crickets. Or maybe with victories coming so quickly on marriage equality and given Obama’s new executive order on nondiscrimination among federal contractors, the numbers game just isn’t that important.

But 1.6% seems way out of whack with everyday experience, even with the expectation that gay people gravitate to larger cities in big numbers. Are people lying to government survey takers? Or to themselves despite their sexual behavior? And do we in fact go to the polls in numbers far out of proportion with Americans overall? Perhaps better analysis will be forthcoming.

More. The Washington Post quotes a few spokesfolk at second-tier and regional LGBT groups who take issue with the survey’s low LGB count, while the Human Rights Campaign says the number isn’t important.

Furthermore. I just came across these lessons about sex from big data at Time online:

3. “Most men lead lives of quiet desperation and go to the grave with the song still in them.”
Like any good data scientist, Rudder lets literature—in this case, Thoreau—explain the human condition. Rudder cites a Google engineer who found that searches for “depictions of gay men” (by which the engineer meant gay porn) occur at the rate of 5% across every state, roughly the proportion of the world’s population that social scientists have estimated to be gay. So if a poll shows you that, for instance, 1% of a state’s population is gay, the other 4% is probably still out there.

4. Searches for “Is my husband gay?” occur in states where gay marriage is least accepted.
Here’s a Big Data nugget you can see for yourself: Type “Is my husband” in Google, and look at your first result. Rudder notes that this search is most common in South Carolina and Louisiana, two states with some of the lowest same-sex marriage approval rates.

21 Comments for “Numbers Racket”

  1. posted by Tom Scharbach on

    The recent studies estimating the percentage of gays and lesbians among various populations vary widely, ranging from right around 1% to as high as 8%, but seem to cluster in the 1.5-3.5% range.

    The Williams Institute estimates that about 3.5% are gay/lesbian/bisexual, about evenly split (1.8% gay or lesbian, 1.7% bisexual). The report takes a look at various studies on Page 3 and subsequent, and the numbers are worth a look.

    The CDC estimate is not wildly out of range with the recent studies cited in the Williams report and elsewhere, so my guess is that the lack of reaction is well, a lack of reaction to old news. It doesn’t merit full-blown political paranoia in my view.

    The lower ranges seem to be relatively consistent when estimating consistent settled attraction/behavior by adults; the numbers go up when estimating any attraction/behavior by adults. Include the number of teenagers who have experimented, and the number goes up a lot, but what happens behind the barn stays behind the barn.

    The one thing that reputable researchers (as opposed to those with a political agenda) agree upon is that coming up with an accurate estimate is a difficult task, strongly influenced by variations in methodology.

  2. posted by clayton on

    I have to think the number is under-reported.

  3. posted by Jorge on

    And do we in fact go to the polls in numbers far out of proportion with Americans overall?

    Can’t rule that one out 😐

    Maybe God suddenly changed time? Or he introduced a flaw in the research method? Or even error?

    But what is the racial/ethnic breakdown? This we do not have. The races have different self-report rates.

  4. posted by Houndentenor on

    I don’t know where you listen, Stephen, because I’ve read lots of skepticism on gay blogs about this survey.

    Besides, what does it actually matter? Are we more or less entitled to equal rights if we are 3% of the population vs 2%?

  5. posted by Lori Heine on

    Given the way gay people–especially gay men–have been stigmatized, by the healthcare industry, in this country, I can understand why the numbers may be underreported.

    And now none of our personal information is private anymore. As the corporate and government worlds become increasingly intertwined, abuse of our personal information is likely to reach Kafkaesque proportions.

    I’m out to absolutely everybody in the world. But even I might be careful how I answer questions administered by a smiling droid in a white smock.

    • posted by Houndentenor on

      One of the things I’ve seen most criticized in this study is the even smaller number of self-identified bi-sexuals. I find that hard to believe, although it does seem there are a lot of people who have often had sex with same-sex partners who do not identify as either gay or bi. Self-identification is one way to measure this but it’s certainly not an objective standard by any means.

  6. posted by Matthew on

    Inflation of the numbers of LGBT people in the population is a long-held strategy of the LGBT movement. The reasons are clear: the more that it is supposed there are, the greater their political clout, and the more “normal” such perversion appears to be. Meanwhile, the sexual abuse of children, a trait strongly associated with the LGB community in a number of scientific studies, escalates.

    • posted by Houndentenor on

      We so rarely get such ill-informed bigots here. It’s hard to know where to begin to dismantle such obvious lies, but considering the readership it hardly seems necessary.

      • posted by Matthew on

        Claiming bigotry is the fallback answer of those who deny the truth.

        • posted by Houndentenor on

          What truth is that? There’s nothing in your little rant that is remotely true. It’s just boiler plate propoganda from the religious right.

          • posted by Jorge on

            *Yawn.* The only strong association between sexual abuse of children and LGB people is that the number of peer-reviewed, scientific studies DENY the association.

            I studied social psychology and statistics, and I read Supreme Court cases for fun. Now that I have disproved the sex abuse link, you no longer think LGBT rights refers to perversion. I have laid hands upon you across the internet, and now you are saved.

            Your arrogance will avail you nothing.

            I’ve set the timer. Now watch him self-destruct.

    • posted by Jim Michaud on

      Hey Matthew, go to Gay Patriot. You’ll fit right in there.

  7. posted by Mike in Houston on

    I suspect that there is a bias towards under-reporting because this was a HHS/CDC survey and people who are reluctant to talk to their doctors about their sexual orientation aren’t likely to open up to a survey taker (no matter how anonymous).

    CDC surveys on HIV — particularly in people of color populations — always show a huge under-reporting of LGBT identification, while at the same time what would be a large “MSM” or men who have sex with men that don’t identify as gay.

    Couple that with recent workplace surveys that show that more than 1/2 of LGBT employees are not out at work — even at companies that have non-discrimination policies — and you have a situation that is ripe for underreporting.

    Census data, which tracks same-sex households and same-sex households with children provides a more balanced picture and number.

    All-in-all, though, this is one survey and one data point.

    Census data

    • posted by Tom Scharbach on

      All-in-all, though, this is one survey and one data point.

      Exactly. One data point among many, with results varying depending on the questions, the sample, and the methodology.

      As Houdentenor and others have pointed out, the HHS/CDC survey is subject to criticism, but its results, although probably low, are not anything to get excited about one way or the other.

      That’s my way of looking at it, anyway.

  8. posted by Tom Scharbach on

    And do we in fact go to the polls in numbers far out of proportion with Americans overall?

    I have no way of knowing, but I wouldn’t be surprised if gays and lesbians vote with higher frequency than the general population. We’d have to be hopelessly lethargic not to have become politicized over the last 30-40 years given the plethora of ballot initiatives, laws singling us out for special discrimination, the AIDS crisis, and the amendment fights.

  9. posted by JohnInCA on

    Their methodology was person-to-person interviews, either over the phone or in-person. That’s a bias against getting good numbers right there.

    Second, if you look at the numbers by age (you can find the full data at the CDC) the younger crowd significantly out-performed the seniors. Which matches other age-breakdowns I’ve seen where the 18-29 crowd significantly identifies as gay more then their older cohorts. Unfortunately the CDC didn’t give a very detailed age breakdown, but from what I could see the pattern was there as well.

    Third, their response pool was somewhere around 35000 if I recall correctly. Meaning they got about 700 admitted LGB. If everyone were honest that would be good numbers, but (again) this is one of those questions where people are notorious liars.

    • posted by Houndentenor on

      I’m probably atypical (I often am) in this regard but I probably wouldn’t stop and give personal information to a stranger either in person or on the phone. Actually I don’t answer my phone unless I know who’s calling. (The rest are almost always telemarketers. Anyone else will leave a message.) Besides I’m usually either practicing or in rehearsal or doing something else that can’t or shouldn’t be interrupted for an annoying survey.

  10. posted by Lori Heine on

    If I think the first question is dishonestly worded, I won’t answer it or any of the others. Usually they are. Whoever’s conducting the survey almost always has a particular viewpoint, and you can tell right away what results they want.

    There’s an ad that keeps appearing on IGF that’s like that. “Do you believe people should be allowed to pray in the public schools?” or some such twaddle. (I could look to see exactly what it says, but I’ve seen too much of it already.) The premise being that if you can’t force everyone else in the universe to pray–and in the manner in which you believe–the government is thereby forcibly prohibiting you from praying.

    It’s on a topic off-subject from this post, put up by a group that probably wishes we were all dead, but I suppose the bloggers can’t control the ads they take.

    It must be official robo-call season, because I get about 2 or 3 of them every day. They always call when I’m working, so I just hang up.

    • posted by Houndentenor on

      I feel even better now about my policy about not answering my phone unless I know who is calling AND feel like talking to them. Friends and relatives know to contact me in other ways (except my mom who is allergic to email and social media but she comes up on caller ID). If my phone doesn’t recognize the number, there’s about a 99% chance that it’s a telemarketer or other annoyance.

      I have always been wary of polling concerning people’s sex lives. It’s the one thing pretty much everyone lies about (we all have that one TMI friend!) either to make themselves seem more experienced or conversely to make themselves seem less slutty. For that reason I don’t think any polling on that topic can be trusted and it’s not possible to know which way the numbers are skewed. And besides, if this study accurate reflects the number of bisexuals then every last bi guy in America contacted me on aol between 1996 and 1998. (LOL)

  11. posted by tom Jefferson 3rd on

    It cannot be easy getting accurate polling data on anything to do with human sexuality. How and who asks the question matters, and, yeah, even then people may feel the need to “show off” or give “conventional” answers.

    4 – 5 percent is probably a low, rough estimate. Exit polling date has found similar results, but not every one votes and not every vote is going to give a polling person the “gospel truth”.

    “10 percent is gay” is not quite what the famous sex research said in the 1940s – 1950s. Alas, complex and nuanced scientific answers don’t always play well in politics.

    It’s always interesting how people rarely wonder how many people are really straight or wonder if kids become straight because of some silly popular culture psychology.

Comments are closed.