Homosexual Demographics and Rick Warren
So, I was wandering the web, and I came across this site, and one of the comments was regarding Rick Warren's estimate of the Gay population at 2%. The comment pointed to an opinion poll, where people believed that the true percentage was closer to 20-25%. The commenter seemed to believe that opinions create truth - that if the public believes that the percentage is 25%, then that is what the percentage actually is.
Which is a really interesting point of view, because then if a majority percentage believe that homosexuality is wrong... it is, isn't it? And, in California... well, let's not go there. Suffice it to say that opinions do not create reality.
But the question interested me, so I went hunting. What is the actual percentage, based on real (or semi-real; people lie to statisticians) data?
Kinsey and crew claimed the percentage at 13-37.
Demographics of Sexual Orientation
Kinsey concluded that all but a small percentage of the population were to one degree or another bisexual (falling on the scale from 1 to 5). He also reported that 37% of men in the U.S. had achieved orgasm through contact with another male after adolescence and 13% of women had achieved orgasm through contact with another woman.[2]
His results, however, have been disputed, especially in 1954 by a team consisting of John Tukey, Frederick Mosteller and William G. Cochran, who stated much of Kinsey's work was based on convenience samples rather than random samples, and thus would have been vulnerable to bias.
But, as stated, Kinsey's work was subject to bias. To get a good survey, you don't post your survey-taker in the women's bathroom if you want to find the population of men in the company. Might be convenient - if you want a result that says 100% of the company is women.
In 1979, Gebhard (with Alan B. Johnson) concluded that none of Kinsey's original estimates were significantly affected by the perceived bias, finding that 36.4% of men had engaged in both heterosexual and homosexual activities, as opposed to Kinsey's 37%.
The problem, of course, is sampling. Where was the sampling done, who did the sampling, and what methods did they use? In the case of Kinsey, his sample included 1500 convicted sex offenders, as well as 600 male and 600 female prostitutes. I think the sample size was about 10k. The 2700 indicated might be a bit skewing to the data. Thus, Kinsey is not a reliable source for an estimate.
More, how does one define "homosexual"?
(same source)
Australia:
...
The study found that 97.4% of men identified as heterosexual, 1.6% as gay and 0.9% as bisexual. For women 97.7% identified as heterosexual, 0.8% as lesbian and 1.4% as bisexual. Nevertheless, 8.6% of men and 15.1% of women reported either feelings of attraction to the same sex or some sexual experience with the same sex. Half the men and two thirds of the women who had same-sex sexual experience regarded themselves as heterosexual rather than homosexual.
How, indeed. Is homosexuality defined by feelings of attraction? Or is it defined by sexual experience? Or is it defined by consistent sexual experience?
Canada:
1988: A study of 5,514 college and university students under the age of 25 found 1% who were homosexual and 1% who were bisexual. [5]
1998: A stratified random sample of 750 males aged 18 to 27 in Calgary, Canada included questions on sexual activity and orientation. 15.3% of men "reported being homosexual to some degree" on the basis of three (often overlapping) measures of homosexuality: (1) voluntary, same-gender sexual contact from age 12 to 27: 14.0%; (2) overlapping homosexual (5.9%) and/or bisexual (6.1%) self-identification: 11.1%; and (3) exclusive (4.3%) and non-exclusive (4.9%) same-gender sexual relationships in past 6 months: 9.2%.[6]
2003: A survey of 135,000 Canadians found that 1.0% of the respondents identified themselves as homosexual and 0.7% identified themselves as bisexual. About 1.3% of men considered themselves homosexual, almost twice the proportion of 0.7% among women. However, 0.9% of women reported being bisexual, slightly higher than the proportion of 0.6% among men. 2.0% of those in the 18-35 age bracket considered themselves to be either homosexual or bisexual, but the number decreased to 1.9% among 35-44 year olds, and further still to 1.2% in the population aged 45-59. Quebec and British Columbia had higher percentages than the national average at 2.3% and 1.9%, respectively.[7]
Hmm. You can read the rest of Wikipedia's article yourself; it's actually fairly well-referenced.
The numbers vary. Many people who do not consider themselves "homosexual" report voluntary same-gender sexual contact. That's not necessarily a contradiction, although it first seems so.
The numbers vary by country, by area of the country, by definitions, by survey methods, and by who is taking the survey. The survey average for the U.S. says probably about 5-6% by definition of "ever having had any same-gender sexual contact". And 2-3% by self-identification as gay or lesbian.
And, only 776,000 same-sex couples (estimated) existed at the time of survey, which was about 0.5% of couples.
So, when Rick Warren uses a figure of 2% - he's not necessarily lying, and he's not relying on a single survey. At least he's not using Guttmacher's survey, which said 1.1%.
Some more references:
Warning. This one is biased:
Family Research, 10% gay discussion
Also biased:
Traditional values: homosexual urban legends
They cite a CDC study, "National Survey of Family Growth", as well as quoting a lawsuit citing another study, "National Health and Social Life Survey".
Also,
According to Census 2000 figures, there are a total of 106,741,426 households in the U.S., only 00.42% of those are unmarried same-sex households.
The Census methodologies changed between 1990 and 2000, so any claims about growth or not, based on those statistics, are unverifiable (if I change my questions in a survey, even slightly, the outcome will change based on the question).
Google Answers
citing http://www.adherents.com/adh_dem.html
I discovered the 2000 census results with the data broken down into
percentages of the total US population based on cnesus data or other
accepted population percentages For gay men the the percentage was
listed at 0.70%- roughyl 2,000,000 gay men for the 200 census. next
to this category was a reference that I have copied below. i hope
this answers you question.
He continues to cite, many of the references being given by one of the above sites also.
He does raise an objection to the survey methodology: that sexual orientation is not counted until the respondent is 18. That is, a 17-year-old is considered non-oriented in terms of sexual preference. So when the Census measures 0.7% as gay men, they're ignoring those under 18. There's a reason for that; when one is under 18, one is generally not asked Census questions!
This is a valid objection to the mathematics of some surveys. One should not apply the findings from one age group (18-35, for example) to the entire population. We do see this in some of the surveys.
This a biased source citing a reversely-biased source:
Dakota Voice, citing Human Rights Campaign, a pro-homosexual advocacy organization
Despite having been told by homosexual activists and their friends in the "mainstream" media for years that homosexuals make up 10% of the population, this study confirmed what accurate but seldom cited studies have been finding for years: homosexuals make up less than 3% of the population.
I need not caution you, but I will anyway. Be aware of definitions and survey methodologies. The survey itself is here.
If you look at Note 5, referencing the 2.9% figure, you'll note that the accuracy of that number is +/- 0.5% points.
Here's another site:
Gaydata.org:
The website was created for individuals and organizations:
1) Who want to find information about lesbians, gays and bisexuals grounded in scientific knowledge,
2) Who want to find datasets to analyze that include sexual orientation data, and
3) Who want to learn how to collect sexual orientation data.
I haven't really looked at the data, although he does have a demographics section that shows a number of survey results, which seem to be many of the same ones examined by the above sites. However, he quotes the surveys directly.
The problem for someone seeking "the actual number" is that one survey is not enough; it might have been terribly biased (like Kinsey's). So we look at multiple surveys, some biased one way, and some biased the other. We throw out the obviously-flawed ones. As an example, the Barna Research Group conducted a survey some time back about divorce. They forgot to include a question: "have you ever been married?" So, of course, when single people responded to the question "have you ever been divorced?" the answer was no, but when your survey is supposed to find how many married people have divorced... well, the survey was flawed. The whole thing was worthless. Surveys are not omnipotent; they suffer from all kinds of problems. Sampling errors (the biggest deal with Kinsey's surveys), definition issues, question phrasing -- all of this stuff can affect the results of the survey.
Cautions: the data are all over the place from the various surveys, indicating methodology, question phrasing, and definition conflicts between the surveys. Probable error is mostly dependent on sample size, and is usually calculated based on a 95% confidence interval.
Nonetheless, the survey given by the Hunter College (sample size of over 26,000) seems rigorously conducted, and sponsored by an organization friendly to LGBT's. Its figure of 2.9% seems reasonable to me, but uses self-identification. This survey excluded those unwilling to participate in a random Internet survey, but would that skew the results? Even granting the error to be on the positive side, the number rises to 3.4% - which still sounds reasonable to me.
Conclusion: while Rick Warren is probably understating the actual percentage, he is nonetheless closer than Kinsey's 37%. The actual number of Lesbian, Gay, or Bisexual persons is likely around 2-3%, for persons aged 18-54. As a rough "guesstimate", 2% is not "off" by much, and is even over the number if you use Guttmacher's survey (estimating 1.1%). Census data shows less than 2%. AND, if we use the Census figure of 0.5% of all couples being same-sex couples, Warren's estimate is high by a factor of 4. So... Warren's estimate is not unreasonable, though he should probably up it to 3%.
Now, his stance on Prop 8, quoting from which is whence this 2% figure comes (essentially, "it only affects 2% of the population; why worry?"), is another matter. THAT is clearly enough an anti-gay bias.
Which is a really interesting point of view, because then if a majority percentage believe that homosexuality is wrong... it is, isn't it? And, in California... well, let's not go there. Suffice it to say that opinions do not create reality.
But the question interested me, so I went hunting. What is the actual percentage, based on real (or semi-real; people lie to statisticians) data?
Kinsey and crew claimed the percentage at 13-37.
Demographics of Sexual Orientation
Kinsey concluded that all but a small percentage of the population were to one degree or another bisexual (falling on the scale from 1 to 5). He also reported that 37% of men in the U.S. had achieved orgasm through contact with another male after adolescence and 13% of women had achieved orgasm through contact with another woman.[2]
His results, however, have been disputed, especially in 1954 by a team consisting of John Tukey, Frederick Mosteller and William G. Cochran, who stated much of Kinsey's work was based on convenience samples rather than random samples, and thus would have been vulnerable to bias.
But, as stated, Kinsey's work was subject to bias. To get a good survey, you don't post your survey-taker in the women's bathroom if you want to find the population of men in the company. Might be convenient - if you want a result that says 100% of the company is women.
In 1979, Gebhard (with Alan B. Johnson) concluded that none of Kinsey's original estimates were significantly affected by the perceived bias, finding that 36.4% of men had engaged in both heterosexual and homosexual activities, as opposed to Kinsey's 37%.
The problem, of course, is sampling. Where was the sampling done, who did the sampling, and what methods did they use? In the case of Kinsey, his sample included 1500 convicted sex offenders, as well as 600 male and 600 female prostitutes. I think the sample size was about 10k. The 2700 indicated might be a bit skewing to the data. Thus, Kinsey is not a reliable source for an estimate.
More, how does one define "homosexual"?
(same source)
Australia:
...
The study found that 97.4% of men identified as heterosexual, 1.6% as gay and 0.9% as bisexual. For women 97.7% identified as heterosexual, 0.8% as lesbian and 1.4% as bisexual. Nevertheless, 8.6% of men and 15.1% of women reported either feelings of attraction to the same sex or some sexual experience with the same sex. Half the men and two thirds of the women who had same-sex sexual experience regarded themselves as heterosexual rather than homosexual.
How, indeed. Is homosexuality defined by feelings of attraction? Or is it defined by sexual experience? Or is it defined by consistent sexual experience?
Canada:
1988: A study of 5,514 college and university students under the age of 25 found 1% who were homosexual and 1% who were bisexual. [5]
1998: A stratified random sample of 750 males aged 18 to 27 in Calgary, Canada included questions on sexual activity and orientation. 15.3% of men "reported being homosexual to some degree" on the basis of three (often overlapping) measures of homosexuality: (1) voluntary, same-gender sexual contact from age 12 to 27: 14.0%; (2) overlapping homosexual (5.9%) and/or bisexual (6.1%) self-identification: 11.1%; and (3) exclusive (4.3%) and non-exclusive (4.9%) same-gender sexual relationships in past 6 months: 9.2%.[6]
2003: A survey of 135,000 Canadians found that 1.0% of the respondents identified themselves as homosexual and 0.7% identified themselves as bisexual. About 1.3% of men considered themselves homosexual, almost twice the proportion of 0.7% among women. However, 0.9% of women reported being bisexual, slightly higher than the proportion of 0.6% among men. 2.0% of those in the 18-35 age bracket considered themselves to be either homosexual or bisexual, but the number decreased to 1.9% among 35-44 year olds, and further still to 1.2% in the population aged 45-59. Quebec and British Columbia had higher percentages than the national average at 2.3% and 1.9%, respectively.[7]
Hmm. You can read the rest of Wikipedia's article yourself; it's actually fairly well-referenced.
The numbers vary. Many people who do not consider themselves "homosexual" report voluntary same-gender sexual contact. That's not necessarily a contradiction, although it first seems so.
The numbers vary by country, by area of the country, by definitions, by survey methods, and by who is taking the survey. The survey average for the U.S. says probably about 5-6% by definition of "ever having had any same-gender sexual contact". And 2-3% by self-identification as gay or lesbian.
And, only 776,000 same-sex couples (estimated) existed at the time of survey, which was about 0.5% of couples.
So, when Rick Warren uses a figure of 2% - he's not necessarily lying, and he's not relying on a single survey. At least he's not using Guttmacher's survey, which said 1.1%.
Some more references:
Warning. This one is biased:
Family Research, 10% gay discussion
Also biased:
Traditional values: homosexual urban legends
They cite a CDC study, "National Survey of Family Growth", as well as quoting a lawsuit citing another study, "National Health and Social Life Survey".
Also,
According to Census 2000 figures, there are a total of 106,741,426 households in the U.S., only 00.42% of those are unmarried same-sex households.
The Census methodologies changed between 1990 and 2000, so any claims about growth or not, based on those statistics, are unverifiable (if I change my questions in a survey, even slightly, the outcome will change based on the question).
Google Answers
citing http://www.adherents.com/adh_dem.html
I discovered the 2000 census results with the data broken down into
percentages of the total US population based on cnesus data or other
accepted population percentages For gay men the the percentage was
listed at 0.70%- roughyl 2,000,000 gay men for the 200 census. next
to this category was a reference that I have copied below. i hope
this answers you question.
He continues to cite, many of the references being given by one of the above sites also.
He does raise an objection to the survey methodology: that sexual orientation is not counted until the respondent is 18. That is, a 17-year-old is considered non-oriented in terms of sexual preference. So when the Census measures 0.7% as gay men, they're ignoring those under 18. There's a reason for that; when one is under 18, one is generally not asked Census questions!
This is a valid objection to the mathematics of some surveys. One should not apply the findings from one age group (18-35, for example) to the entire population. We do see this in some of the surveys.
This a biased source citing a reversely-biased source:
Dakota Voice, citing Human Rights Campaign, a pro-homosexual advocacy organization
Despite having been told by homosexual activists and their friends in the "mainstream" media for years that homosexuals make up 10% of the population, this study confirmed what accurate but seldom cited studies have been finding for years: homosexuals make up less than 3% of the population.
I need not caution you, but I will anyway. Be aware of definitions and survey methodologies. The survey itself is here.
If you look at Note 5, referencing the 2.9% figure, you'll note that the accuracy of that number is +/- 0.5% points.
Here's another site:
Gaydata.org:
The website was created for individuals and organizations:
1) Who want to find information about lesbians, gays and bisexuals grounded in scientific knowledge,
2) Who want to find datasets to analyze that include sexual orientation data, and
3) Who want to learn how to collect sexual orientation data.
I haven't really looked at the data, although he does have a demographics section that shows a number of survey results, which seem to be many of the same ones examined by the above sites. However, he quotes the surveys directly.
The problem for someone seeking "the actual number" is that one survey is not enough; it might have been terribly biased (like Kinsey's). So we look at multiple surveys, some biased one way, and some biased the other. We throw out the obviously-flawed ones. As an example, the Barna Research Group conducted a survey some time back about divorce. They forgot to include a question: "have you ever been married?" So, of course, when single people responded to the question "have you ever been divorced?" the answer was no, but when your survey is supposed to find how many married people have divorced... well, the survey was flawed. The whole thing was worthless. Surveys are not omnipotent; they suffer from all kinds of problems. Sampling errors (the biggest deal with Kinsey's surveys), definition issues, question phrasing -- all of this stuff can affect the results of the survey.
Cautions: the data are all over the place from the various surveys, indicating methodology, question phrasing, and definition conflicts between the surveys. Probable error is mostly dependent on sample size, and is usually calculated based on a 95% confidence interval.
Nonetheless, the survey given by the Hunter College (sample size of over 26,000) seems rigorously conducted, and sponsored by an organization friendly to LGBT's. Its figure of 2.9% seems reasonable to me, but uses self-identification. This survey excluded those unwilling to participate in a random Internet survey, but would that skew the results? Even granting the error to be on the positive side, the number rises to 3.4% - which still sounds reasonable to me.
Conclusion: while Rick Warren is probably understating the actual percentage, he is nonetheless closer than Kinsey's 37%. The actual number of Lesbian, Gay, or Bisexual persons is likely around 2-3%, for persons aged 18-54. As a rough "guesstimate", 2% is not "off" by much, and is even over the number if you use Guttmacher's survey (estimating 1.1%). Census data shows less than 2%. AND, if we use the Census figure of 0.5% of all couples being same-sex couples, Warren's estimate is high by a factor of 4. So... Warren's estimate is not unreasonable, though he should probably up it to 3%.
Now, his stance on Prop 8, quoting from which is whence this 2% figure comes (essentially, "it only affects 2% of the population; why worry?"), is another matter. THAT is clearly enough an anti-gay bias.