Same-Sex Unmarried Partner Couples in the American Community Survey: The Role of Misreporting, Miscoding and Misallocation

Michael D. Steinberger, Pomona College
Gary J. Gates, University of California, Los Angeles

The past quarter century has seen a large amount of research on the demographic characteristics and economic decisions of gay and lesbian Americans. The majority of the nationally representative research on same-sex couples in the United States uses the U.S. Census and American Community Surveys as their data source. We utilize differences in data collection methods in the American Community Survey (ACS), 2005-2007 to explore the role of misallocation in the identified same-sex unmarried partner sample. By comparing demographic and economic characteristics over the entire distribution of responses, we show a significant portion of identified same-sex households are likely incorrectly allocated opposite-sex married couples. Based on our analysis, we provide empirical guidance to researchers interested in obtaining accurate demographic and economic characteristics of same-sex households from the U.S. Census and ACS. For completeness, we also test the robustness of our findings using the PUMS 2000 Census data.