Census Bureau on same-sex couple households: Read the data, see the movie
The Census Bureau has issued a correction in the numbers of same-sex spouses and same-sex unmarried partners, revising the estimated numbers in the two categories down from 349,377 to 131,729, and from 552,620 to 514,735, respectively. Williams Institute Distinguished Scholar Gary Gates, together with other demographers, concurred in the revision. (Full statement after the jump)
That makes a revised total of 646,464 gay couples, up from the 2000 Census number of 594,391. (Because the Census counts the population by households, there is no figure for the number of unpartnered lgbt people.)
What's really amazing is that the Bureau produced this videotape to explain how the initial overcounting errors occurred.
The Census Bureau press release:
The U.S. Census Bureau released today new statistics on same-sex married couple and unmarried partner households. According to revised estimates from the 2010 Census, there were 131,729 same-sex married couple households and 514,735 same-sex unmarried partner households in the United States.
The results of the 2010 Census revised estimates are closer to the results of the 2010 American Community Survey (ACS) for same-sex married and unmarried partners. The 2010 ACS estimated same-sex married couples at 152,335 and same-sex unmarried partners at 440,989.
The new, preferred figures revise earlier estimates of same-sex unmarried partners released this summer from the 2010 Census Summary File 1 because Census Bureau staff discovered an inconsistency in the responses in the 2010 Census summary file statistics that artificially inflated the number of same-sex couples. In addition, a breakdown of couples who reported as same-sex spouses is now available. The summary file counts originally showed that there were 349,377 married couple households and 552,620 same-sex unmarried partner households.
Statistics on same-sex couple households are derived from two questions on the census and ACS questionnaire: relationship to householder and the sex of each person. When data were captured for these two questions on the 2010 Census door-to-door form, the wrong box may have been checked for the sex of a small percentage of opposite-sex spouses and unmarried partners. Because the population of opposite-sex married couples is large and the population of same-sex married couples in particular is small, an error of this type artificially inflates the number of same-sex married partners.
After discovering the inconsistency, Census Bureau staff developed another set of estimates to provide a more accurate way to measure same-sex couple households. The revised figures were developed by using an index of names to re-estimate the number of same-sex married and unmarried partners by the sex commonly associated with the person's first name.
"We understand how important it is for all groups to have accurate statistics that reflect who we are as a nation," said Census Bureau Director Robert Groves. "As scientists, we noticed the inconsistency and developed the revised estimates to provide a more accurate portrait of the number of same-sex couples. We're providing all three - the revised, original and ACS estimates - together to provide users with the full, transparent picture of our current measurement of same-sex couples."
The 2010 Census preferred estimates have been peer-reviewed by Gary Gates, a demographer with the UCLA School of Law's Williams Institute on Sexual Orientation Law and Public Policy, by Philip Cohen, a professor of sociology at the University of North Carolina at Chapel Hill, and by Megan Sweeney, professor of sociology at UCLA. These experts concluded the methodology behind these revised estimates was sound.
All three sets of estimates are available at both the national and state levels and provide estimates of the presence of the couple's own children. The 2010 Census revised estimates provide a 10-year benchmark, while the ACS estimates are useful for looking at a yearly time series.