Is Ten Percent of the Population Homosexual?

It's always nice to know that you can meet up with your high school friends with whom you have kept in touch all these years and still partake in political polemics as if nothing has changed.  Granted, I became libertarian since then, and thus don't agree with the Right like I used to, but it's still pleasant to have that sense of nostalgia with good friends.

I had such a rendez-vous with some Left-leaning high school friends yesterday. Although a good number of topics came up in discussion, the one that lasted the longest and was the most heated was whether a tenth of the population was homosexual.  Whether it was personal experience, intuition, or previous, brief encounter with a small handful of studies, I didn't agree with the "ten percent gospel" that has its origins in the well-renowned Kinsey study.  As I should well know by now, if you question something that someone else deems "unquestionable," the response is "attack mode."  As long as it's not personal, I don't mind the heated debate since the important thing at the end is ascertaining the truth.

I was told by one of my friends that I should look at the Kinsey study, as well as the Time Magazine article from 2004.  A lot more has been published on the subject since Kinsey, and the article from Time didn't answer the question of "what percent of the population is gay?"  Regardless, I was going to look at the data to see if my friends were right.

It's not only a matter of the fact that homosexuality is a hot-button issue in America.  Both sides have a stake in the debate, which is why I am just as skeptical of the Left's 10% claim as I am of the Right's claim of a considerably small homosexual population. 

A paramount question is how one goes about defining "homosexual."  Is a homosexual who has only one sexual encounter with someone of the same sex? Can we define a homosexual as someone who exclusively or predominantly has sexual relations with someone of the same-sex?  Are people who self-identify as homosexual an accurate measurement of the homosexual population?  As we go through the prominent studies on the matter, let's keep these questions in mind.  

A look at the Kinsey study itself: The Kinsey study was under heavy criticism for selection bias.  But let's consider that the Kinsey study isn't methodologically flawed, because no study ever is, right?  Under the Kinsey study, 37% of males and 13% of females have had at least one overt homosexual experience.  A good majority of us are not going to consider translating one homosexual interaction into being homosexual.  The colloquialisms for such an experience are "experimenting" or "being bi-curious."

Analyzing more long-lasting findings, the study shows that 10% of males and 2-6% of females were exclusively homosexual.  80% of the aforementioned males (i.e., 8% of the statistic sample) were exclusively homosexual for at least three years.  However, 4% of males and 1-3% of females were exclusively or predominantly homosexual from the onset of adolescence.  How long does someone have to have same-sex attraction to be considered homosexual?  Your answer is going to vary based on your definition.  In spite of that, the revered study hardly claims that 10% of the population is homosexual.

The Census Bureau findings are commonly used by those on the Right because the percentage is around 2-4%.  The Census Bureau has a few flaws.  First is that there is no explicit question asking about sexual orientation.  The person filling out the form has to report how the other person is related to them, which causes ambiguities in the answer.  There is still the social stigma of being homosexual, which means people are hesitant to put their sexuality on the form.  Even if that stigma didn't exist, there will still be people out there who feel that it's none of the government's business.  

The Center for Disease Control (CDC) recently did a study on sexual behavior. The study measured sexuality on two primary levels.  The first is the sexual act itself (Table 10).  Does having any sexual contact count?  Does it have to be a specific act?  And again, does partaking in a single homosexual act render you homosexual?  Then there's the matter of sexual identity.  How the question is framed brings about different answers.  Being framed in terms of "Are you mostly [or partly] homosexual/heterosexual, or just bisexual (Table 11)" has different results than "Do you consider yourself heterosexual, homosexual, bisexual, or other (Table 12)."  You can finagle over 10% from Table 11 if you lump in people who consider themselves mostly heterosexual.  You can't pull it off with the findings from Table 12.

The Williams Institute, which essentially is a think tank for issues regarding advocacy for sexual orientation, did a study about same-sex couples and individuals in the United States.  Looking at Table 2, the state with the highest estimation of gay men, lesbians, and bisexuals was Washington DC at 8.1%.  Since the sample maximum is below 10%, there is no way the national mean is 10%, which means that according to this pro-gay rights study, the LGB (Lesbian, Gay, and Bisexual) population is not 10%.

Even a 2010 study, which grabbed the Huffington Post's attention because it was "the largest survey on the topic published since the 1990s," does not match up to the 10%.  Looking at Table 1 (p. 258), 92.2% of men and 93.1% of women identified as heterosexual.  You would have to include bisexuals to have a median population parameter of 5.65%, which is a fancy way of saying that it doesn't make it to 10%.  Even if you decide to use sexual acts within the past year as an indicator (which is better than sexual acts over a lifetime, but still is limited in terms of meaning), none of the homosexual acts reach the 10% mark.

Postscript: There are other studies on the issue, but the major, peer-reviewed ones have been covered in this blog entry.  In order to make the "ten percent" argument, one would have to make the definition of "homosexual" so broad that it included bisexuals, individuals with very transient same-sex attraction [much like in the Kinsey study], and those who have had at least one lifetime homosexual act, all of which undermine the claim itself.  This is truly a stretch of statistical data to bolster numbers.  Looking at the data puts the percentage closer to five or six percent.

Does this mean that Americans should stop focusing on making sure that homosexuals are provided their rights?  Absolutely not!  Having a smaller percentage of homosexuals in society does not mean we abandon the endeavor of making sure that the government protects the natural rights for homosexuals, an endeavor that will most likely be the civil rights issue of our time.  However, the continued usage of the "ten percent" statistic only debilitates credibility.  Those who argue for gay rights don't need to use exaggerated numbers when they already have the advantage of a sound argument.

6-5-2016 Addendum: Last Wednesday, the Archives of Sexual Behavior published a study showing the prevalence of same-sex behavior. The number of men having sex with at least one man increased from 4.5 percent in the early 1990s to 8.2 percent in the late 2010s. For women having sexual relations with women, that increased from 3.6 to 8.7 percent within the same time period. Given the increased societal acceptance of homosexuals and homosexuality, these findings could simply mean that more people are comfortable reporting their true sexuality and sexual practices. Assuming that a) these figures are correct and b) we want to use the definition provided in the study, they are still slightly below the 10 percent mark.

5-22-2018 Addendum: Gallup estimates that 4.5 percent of the U.S. population is homosexual.