Examining Same-Sex Couples Using Dyadic Data Methods

Abstract

The increasing availability of population-level data on same-sex couples has given family demographers greater insight into the characteristics of same-sex couples in the United States as well as the implications of gender and sexuality for individual well-being across multiple outcomes. Whereas most nationally representative research on same-sex couples focuses on individuals, surveys such as the National Health Interview Survey (NHIS) can be used to construct couple-level data files, enabling researchers to consider how couple-level indicators and variables for both partners affect individual well-being. In this chapter, we detail the benefits of dyadic data methods for research on couples and demonstrate a dyadic data analysis of 1262 same-sex couples and 113,642 different-sex couples derived from the 2012 to 2016 NHIS person-level files. Specifically, we examine the effects of respondent gender × partner gender on: reported health status; partner education and reported health; and union status and reported health among those in same-sex versus different-sex couples. Our results suggest that analysis of same-sex couples using dyadic data methods yields more nuanced results than analyses of individuals. For instance, there is a significant positive association between partner college degree and reported health status that is greater in magnitude for women than men—but only if women are in different-sex relationships. In addition, cohabitation and marriage are similarly associated with health for men in same-sex relationships, but cohabitation is associated with poorer health for men and women in different-sex relationships. We conclude with suggestions for future research using dyadic data methods for research on couples.