There’s no formal organization or outfit behind gaydemographics.org, it is the creation of Emilio Guerra, a Spanish-born a journalist who lives in North Miami Beach, FL.

I began to post this information on my personal website, www.emilioguerra.com, back in the spring of 2001, when I couldn’t find detailed information anywhere else about same-sex couples. Since the tables and the whole thing really got out of hand, I moved the information to www.gaydemographics.org in February of 2002.

www.gaydemographics.org is first and foremost a community resource, built with the plain truth and facts about the homosexual population in the world, and same-sex couples in particular. It is open for cooperation with any individual or organization that would like to analyze or expand on the data.

Every single statistic in a table is properly attributed to the demographic source, usually the United States Census.

U.S. Census methodology Summary File 1

All number were obtained from the American Factfinder FTP for Summary File 1.

The following tables were obtained from the SF1 FTP server:

P001001: Total Population
P018007: Married couples
PCT014002 Unmarried partner households
PCT014003 Male householder and male partner
PCT014005 Female householder and female partner

Once the tables were downloaded the following format was used for every table that pertains to the 2000 U.S. Census in the website:

NAME (name of the geographic unit)
Population (P001001)
Married Couples (P018007)
Unmarried couples (PCT014002)
Same-sex couples (PCT014003 + PCT014005)
% Same-sex couples ((PCT014003 + PCT014005)/(P018007+ PCT014002))
Male couples (PCT014003)
% Male couples ((PCT014003)/(P018007+ PCT014002))
Female couples (PCT014005)
% Female couples (PCT014005) /(P018007+ PCT014002))

There is a slight variation for the Summary File 2 tables, the ones pertaining to race. For the 1990 Census (http://www.gaydemographics.org/USA/1990_Census.htm), the analysis was obtained from the household (proportion of same-sex couples divided by total households), which is how I did everything originally.

PUMS Methodology

1. Downloaded all 52 5% 5-Percent Public Use Microdata Sample (PUMS) Files.
2. For each file, flagged all households that included an "Unmarried partner" as a person who was living in the household, under the relationship field.
3. Extracted all households with unmarried partners, and compared the sex of the Householder with the sex of the unmarried partner.
4. Another filtered list was made when the Householder and the unmarried partner were of the same sex.
5. State list was filtered with those households that met criteria in item #4.
6. All person records and all household records (2 files) were made separately.
7. Data was added to Microsoft Access database.
8. A total of 104 files were added (one for each state plus the District of Columbia and Puerto Rico, with both the person records and the household records).
9. Prior to adding to database, codes were changed for each of the 104 files. The orginal PUMS files contain no text whatsoever, only numeric codes. Each file has 174 columns.
10. The ensuing database was almost 1 GB in size. Its findings are reflected here.

All in all, it took me about a month to prepare the data, not bad considering that I had no experience with database work and have a rather slow PC.

About Emilio Guerra

I am not a professional statistician, nor do I play one on the web. As to my credentials, I have none as a demographer or statistician. I am a journalist, but have always loved facts, and demographics in particular. Some people have called me a statistics geek. I studied American History at Florida International University, here in Miami. You can find out more about me and my partner, Josh Hyatt, on my personal website.