A significant amount of the research on same-sex couples in the United States uses the Decennial Census and the American Community Survey (ACS) as primary data sources. However, research shows that a potentially large fraction of identified same-sex couples may actually be misidentified different-sex spouses where one partner miscoded their sex. Beginning in 2008, the Census Bureau changed its survey instrument and editing procedures for the ACS in a manner which may have significantly reduced the potential for misidentification in the identified same-sex couples'sample. This proposal will help determine if new procedures used in the 2008 and 2009 ACS have improved the reliability and accuracy of Census data on same-sex couples. Findings will allow researchers to understand and account for the misidentification issue when using these data. The urgency of assessing the reliability of same-sex-couple data is further compounded given that the Census Bureau has begun formally releasing population estimates of same-sex spouses. Analyses will focus on two research aims. The first research aim will compare data from the 2007, 2008, and 2009 ACS to help assess if the 2008/2009 ACS data collection and editing procedures yield greater accuracy of responses and improve the reliability of same-sex couple data. Analyses will use confidential ACS data that includes original unedited responses to the household roster to assess the quality of the same-sex couples'data. The second research aim will assess if the complexity of legal relationship recognition for same- sex couples creates discordance between unedited responses to the relationship to householder and the marital status questions. This discordance in their legal status may manifest itself in discordance in how same- sex couples jointly interpret the relationship to householder question (husband/wife vs. unmarried partner) and the marital status question (currently married vs. other responses). The consideration of this discordance will offer insights into how same-sex couples process these questions in the face of a complex legal relationship environment. These analyses will be useful in understanding how to craft questions that accommodate the challenges faced by same-sex couples when they are asked to define the social and legal nature of their relationships.
The US Decennial Census and the American Community Survey (ACS) are perhaps the most widely used data sources to study the geographic, economic, and demographic characteristics of same-sex couples. Beginning in 2008, the Census Bureau changed its survey instrument and editing procedures for the ACS in a manner which may have significantly improved the accuracy and validity of the same-sex couples'sample. This research will help to determine if new procedures used in the 2008 ACS have improved the reliability and accuracy of Census Bureau data on same-sex couples.