Measuring Couples' Gender Composition: Evaluating Partner Proxy Reporting on a Two-Step Gender Identity Measure

Many household surveys rely on proxy reports, in which one respondent answers for other household members. Two-step gender identity measures are well-tested for self-identification. Far less research has tested these measures for partner-proxy reports of couples' gender composition. Using dyadic data from the National Couples' Health and Time Study (NCHAT), we evaluated proxy reporting among partnered adults aged 20–60. Nonresponse to proxy reports of sex assigned at birth and current gender identity was comparable to self-reports, and similar to or lower than proxy nonresponse for other demographic characteristics. Proxy and self-reports on these measures showed high agreement, meeting or exceeding levels observed for other demographic measures. Within the two-step measure, congruence was lower for self-identified transgender or gender-diverse partners than for self-identified cisgender partners. Among discordant reports, proxies were also more likely to report a self-identified transgender or gender-diverse partner as cisgender than to report a self-identified cisgender partner as transgender or gender-diverse. Overall, our findings provide new insights into partner-proxy reporting of gender identity and reveal a directional pattern of discordance that warrants further attention. This study is one step toward better measuring gender diversity within couples and underscores the need to intentionally include couples with an LGBTQ+ partner in data collection.