Emilio Guerra
gaydemographics.org

One thing the 2000 Census does not do is indicate how many gay, lesbian, bisexual and transgendered people there are. It would be very unwise to extrapolate from these figures the total number of homosexuals.

As you can see from the table below, the official numbers from the U.S. Census Bureau change wildly, specially when sampling is made in the smaller units. Thus, even though California's same-sex couples vary 13% depending on the count, North Dakota's have a 220% statistical variance.

The official figures only represent same-sex couples that decided to be counted. Despite the possible exaggeration and overstatement of the undercount, I still believe that there were a significant amount of same-sex couples that were not counted,or rather, did not count themselves, whether by design or simply by ignorance.

When the information was being released on a state by state basis, the Institute for Gay and Lesbian Strategic Studies announced its doubts that all same-sex couples were indeed counted in the 2000 Census.

Why do I believe that not every same sex-couple was counted?

• Even though the Census information is labeled confidential, the information will be released to Block group level. If you want true confidentiality about your household arrangements, you'll think twice about that. In addition, small communities are more identifiable. If you are a same-sex couple in, say, Athelstan, Iowa (population 18), you were counted. But now, your neighbors may be wondering who is the gay couple in their midst. You may or may not care, but if you submitted it in the Census form, you have no choice.
• Secondly, the word partner. The interesting study, Demographics of the Gay and Lesbian Population in the U.S.: Evidence From Available Systematic Data Sources (more about it below), concludes that the misinterpretation rate of the word "partner" was approximately 0.40% in the 1990 Census. That is, "a respondent may consider his two roommates "partners" if household resources are shared," for example. But also, many people in relationships, committed or otherwise, may feel that the denomination was either too formal or informal, and changed it. Or simply, as the IGLSS contends, they were not aware of the possibility.

Definitions

One stumbling block for demographers is establishing who is "gay". In other words, labels. From the heady days of Kinsey, the "10 per cent is homosexual" label has been toned down lately. Even though The Social Organization of Sex: Sexual Practices in the United States (Laumann, Gagnon, Michael, and Michaels, 1994) study cites that 7.7 percent of men and 7.5 of women have strong same-sex attractions, ony 2.8 percent of men and 1.4 of women consider themselves homosexuals.

The IGLSS again puts its two cents here, citing factors that make many gay people reluctant to identify themselves as gay, even though they would be considered as such by the "mainstream" G/L/B/T Community.

So, for the total number of homosexuals, whether self-proclaimed or otherwise, you don't need the Census. Just multiply the Population by the percentage you like the most (10, 7.6 or 2.1), and voilá, you get a figure.

Now, of course, we have hard data about how many same-sex couples were aware of this option in the Census and decided to be counted as such. Alas, no studies have been done yet.

1990 Census and the University of Syracuse Study

The Center for Policy Research of the Maxwell School of Citizenship and Public Affairs at Syracuse University in New York issued a working paper in 1999, titled Demographics Of The Gay And Lesbian Population In The United States: Evidence From Available Systematic Data Sources.

In it, the CPR states "we provide an overview of standard social science data sources that now allow some systematic study of the gay and lesbian population in the United States. We consider how sexual orientation can be defined in each data source, and we note the potential sample sizes of gays and lesbians from each data source."

You can find the working paper here, in PDF ( Adobe Acrobat) format.

The study basically analyzes figures from three sources: The General Social Survey (GSS), the National Health and Social Life Survey (NHSLS) and the 1990 U.S. Census.

Naturally, it needs to be updated or a new working paper needs to be issued on all the data of the 2000 Census. But it is a start. Here's their conclusion:

Although I may or may not agree with its findings, and must contend that the 2000 Census will probably increase the figures, it is perhaps as scholarly as it gets, for now.

Other resources

• Gay.com has a wonderful section called "60 years counting queers: Gay population studies since Kinsey". In it, most of the important surveys about gay population are included.
• The Institute for Gay and Lesbian Strategic Studies (IGLSS) has several policy papers about this and other subjects, including the closely related subject of the purchasing power of gays and lesbians.
• From the opposite end of the spectrum, the Robertson School of Government of Regent University, had the paper Data Integrity and Homosexual Politics. I included it here because it tries to make a point, from a very politicized and non-too-friendly view, but nonetheless a point was attempted. It was removed from their site, though.
• Gaycensus.com is a web-based undertaking to count the world's gay, lesbian, bisexual, and transgendered population, through an internet-based survey. It is run by the Citizens for Census Equality in California.
• The General Social Survey.