In the late 1940s and 1950s, sex researcher Alfred Kinsey estimated that about 10% of the population was something other than straight (and then, as now, a much larger number have same sex experiences or attraction). Today scholars believe that about 3.5% of the U.S. population identify as gay, lesbian or bisexual, a considerably lower number. Yet, a telling poll by Gallup shows that Americans wildly — wildly — overestimate the number of people who identify as non-heterosexual:
The table shows that more than a third of Americans believe that more than one out of every four people identifies as gay or lesbian. Only 4% of Americans answered “less than 5%,” the correct answer.
Estimates varied by demographics and political leaning. Liberals were more likely to overestimate, as were younger people, women, Southerners, and people with less education and income:
Interestingly, these numbers are higher than in 2008, when Gallup asked a similar questions. In that poll, only a quarter of the respondents choose “more than 25%” and more than twice as many said that they had “no opinion.”
Gallup concludes: “…it is clear that America’s gay population — no matter the size — is becoming a larger part of America’s mainstream consciousness.”
Lisa Wade, PhD is an Associate Professor at Tulane University. She is the author of American Hookup, a book about college sexual culture; a textbook about gender; and a forthcoming introductory text: Terrible Magnificent Sociology. You can follow her on Twitter and Instagram.
Comments 60
Quinn — June 11, 2012
Depending on how respondents interpret the question (or, more generally, who is "gay or lesbian"), <5% or 5-10% or 10-15% could all be considered the "correct" answer. Some people might really be answering based on same-sex behavior or same-sex attraction, rather than gay or lesbian identity. I suspect that a lot of respondents probably wouldn't distinguish very carefully between the three. Even so, there is still clearly an issue with overestimation compared with current data.
I'd also be really interested in seeing estimates from self-identified LGBT people, though obviously they'd be a small number of respondents within the Gallup survey (assuming they even asked the question). I'd be particularly interested in this because there could be a social network effect (among other effects)--you report based on the people you know. But there also seems to be an interesting effect among people who don't want "gay/lesbian relations" to be legal, which could be a symbolic threat issue. There's also probably some standard numeracy issues here, as well.
This is definitely worth looking into further.
Adfs — June 11, 2012
Did they ask mainly people in cities? Dan Savage pointed out that gay and lesbian people tend to congregate in big cities after high school, because even if their hometown isn't homophobic, its hard to get much of a dating scene going in a small town. So if you confine your population to big cities maybe the estimates are more accurate?
WG — June 11, 2012
"Only 4% of Americans answered “less than 5%,” the correct answer."
At this time there is really no way to be absolute in knowing the "correct" answer. There are way, way too many variables unaccounted for, yet.
hopeless shade — June 11, 2012
It's interesting to find that the "I have lots of LGBTQA friends, they're probably about 20% of the population, right?" and the "these sinful homosexuals are taking over America, they've probably spread their agenda to like 20% of the population, right?" camps both picked the same incorrect number.
I have always gone with 10%, because it sounds usefully significant but realistically low.
Andrew s — June 11, 2012
I think it's going higher because more people are engaging, openly, in bisexual behaviors or simply acknowledging their bisexual tendencies, without engaging in the behaviors. This openness may confuse people.
oddboyout — June 11, 2012
Is there a similar survey on assumed demographics of race or ethnicity in the US?
EDIT: Found an article on JSTOR published in 2001 showing African American population is estimated at 30% (it's 12.4%). http://www.jstor.org/stable/3078787
myblackfriendsays — June 11, 2012
I would think the difference between people who identify and people who actually are (whatever that means,) is statistically significant.
mimimur — June 11, 2012
Actually, from what I've seen, different surveys have results lying at about 5-10%. And then again, as long as there is a stigma towards it, we'll never know how many respondents would have chosen a queer category everything else being equal. So really, I'm a bit sceptical to pointing at these groups as being wrong.
mimimur — June 11, 2012
Actually, from what I've seen, different surveys have results lying at about 5-10%. And then again, as long as there is a stigma towards it, we'll never know how many respondents would have chosen a queer category everything else being equal. So really, I'm a bit sceptical to pointing at these groups as being wrong.
kimadactyl — June 11, 2012
"Yet, a telling poll by Gallup shows that Americans wildly — wildly — overestimate the number of people who identify as non-heterosexual:"
These things are not antonyms. "non-heterosexual" covers a lot more than "lesbian" and "gay".
AMERICANS WAY (LIKE, WAY) OVERESTIMATE THE NUMBER OF GAYS AND LESBIANS « Welcome to the Doctor's Office — June 11, 2012
[...] from SocImages [...]
Barbara — June 11, 2012
I am having a very hard time figuring out who is responding to whom here, and it seems to me that those who complain about others being snarky themselves then become snarky. And please, people, don't get nasty when someone has a typo. It's just a typo, not an indication of stupidity.
According to my reading of it, WG's legitimately sarcastic comment was NOT discrediting sociology nor social statistics; neither was he de-railing the topic. He was making the very valid point that sexuality and identity are at least somewhat linked. Right? This seems like a smart comment to me, even given the fact that there are real distinctions between attraction, identity, behavior, what we might think of as some ontologically "true sexuality," although how on earth we would know this about a person is beyond me. All we CAN ask people is about their behaviors, their experiences of attraction, and their current identity. Right? There is no way of knowing what someone's sexuality "really" is, whether we are professional sociologists or not. This is especially true because an individual's definition of themselves can change over time (this has been found to be more common for women than for men), at least these days, with a person saying that they "are" heterosexual, then lesbian, then bisexual, then claiming that they were only "experimenting," or "making a mistake," and after all, they "were" heterosexual all along. As Oddboyout points out, Americans have been overestimating the Black population for decades. This overestimation is variously explained as being the result of "threat" i.e. "oh dear, there are so many of them out there," or of cultural presence ("wow, there are 'so many' (sic) of them in the movies and on tv that there must be a lot of them in real life"), OR the result of living in a place where the local Black population IS in fact relatively high, and mistakenly ascribing that proportion to everywhere, USA ( I have a friend from Raleigh, NC who was downright shocked that there are New England towns where NO Black people live at all. It never occurred to her that this could be true). These hypothesized explanations provide some possible ones for the current data: threat (as in the social conservatives, as pointed out by Hopeless Shade), cultural presence ("there are so many of them on tv") and living in high-concentration areas or social circles (cities and among liberals, as again pointed out astutely by Hopeless Shade).
Jamie Riehl — June 12, 2012
"Are" and "identify as" are not the same question.
Quickies: 06/13/2012 - Queereka — June 13, 2012
[...] I wish more than 25% of Americans identified as [...]
Liz Bloodbath — June 13, 2012
I don't think this article should use the phrase non-heterosexual and gay/lesbian interchangeably. The Gallup poll specifically used the phrasing 'gay or homosexual men and lesbian or homosexual women'. They didn't include bisexuals or people who otherwise selfidentify as queer. Just a nickpick.
Although the exclusion of bisexuals makes the numbers seem even more usual. I don't know anyone who thinks 1/4th of the people they meet are gay, and gay advice columnist Dan Savage regularly brings up the gay dating pool being very small. Not arguing with the data, it's just surprising.
Survey Respondents Give Awful Estimates, part n of N « A (Budding) Sociologist's Commonplace Book — June 17, 2012
[...] of social phenomena by survey respondents. The first example is from a recent SocImages post: Americans Way (Like, Way) Overestimate the Number of Gays and Lesbians. The title of the post gives away the main finding pretty well: while the best estimates of GLB [...]
Curtis Lam — June 20, 2012
Have to say I'm not too surprised by this and I personally always thought it had to be less than 10%. LGBT-related issues are very much in the public eye these days, and people are being more open about their orientations than ever before.
NoBigGovDuh — August 9, 2012
I think they may be thinking gay/bisexual?
Gay Marriage Opponents Think Most Americans Dislike Equality As Much As They Do - iVoter.com | iVoter.com — March 5, 2014
[…] inaccurate. Heck, one in ten Americans think HTML is a sexually-transmitted disease. And 35 percent think that at least a quarter of people are homosexual. Interestingly, even same-sex marriage supporters seem to perceive less support for it than […]
Gay Marriage Opponents Think Most Americans Dislike Equality As Much As They Do » RickMick | RickMick — March 5, 2014
[…] inaccurate. Heck, one in ten Americans think HTML is a sexually-transmitted disease. And 35 percent think that at least a quarter of people are homosexual. Interestingly, even same-sex marriage supporters seem to perceive less support for it than […]
MalcolmY — December 9, 2014
I can certainly understand why so many people think that the number of Gays and Lesbians is higher than it is (or may actually be). Gays are an extremely vocal minority. You see a gay character on practically every (slight exaggeration perhaps) sitcom and TV drama. You see them marching in their own parades. You see them attempting to infiltrate other parades, you walk past their bars, and you are inundated with rainbow flags and bumper stickers. Then we are constantly hammered by the media coverage of same sex marriage, and special laws for workplace protection. I'm a live and let live type, and I am very surprised that the number in this poll is estimated at 3.5% Gays themselves have always claimed 10% as far as I knew.