LAUSR

Purpose Quantitative research on the segregation of same-sex partners in the USA is new, and limited by challenges related to the accurate measurement of segregation and data errors. The purpose of this paper is to provide a novel approach to re-examine residential segregation between same-sex partners and different-sex partners in the USA. Design/methodology/approach Two versions of the dissimilarity index and corrected same-sex partners data from the 2010 decennial census were used. Effects of different geographic scales were examined. Findings Results reveal that the levels of segregation of both male and female same-sex partners were higher at metropolitan- vs state-levels; the levels of segregation was lower when measured using the unbiased as compared to the conventional version of the D-index; and male same-sex partnered households were more segregated from different-sex partnered households than were female same-sex partnered households. Research limitations/implications Future studies should be attuned to geographic scale effects and should not ignore the bias of the D-index. Originality/value This study provides a better test of the differences between the two versions of the D-index and contributes to the literature by examining the segregation of both male same-sex partners and female same-sex partners across different geographic scales.