A number of studies emphasizing immediate risk factors for lesbian, gay, and bisexual (LGB) adolescents and young adults suggest that LGB youth are at-risk for lower levels of educational achievement (Robinson & Espelage, 2013; Russell, Seif, & Truong, 2001; Kosciew,et .al, 2011).  Yet limited cross-sectional research on educational attainment and sexual orientation shows that same-sex, partnered men may have higher levels of education and as same-sex, partnered women have lower levels of education compared to opposite-sex partnered individuals (Black et al., 2000; Baumie et. al, 2009; Turner et. al, 2005).  Furthermore, there may be greater variation in educational attainment depending upon how lesbian, gays, and bisexuals are categorized (e.g. attraction, same-sex partnered households, sexual behaviors) and between sexual minority groups  (Black, Sanders, & Taylor, 2007; Pew Research Center, 2013).

The broad goal of this paper is to determine if differences in sexual orientation predict educational attainment in the National Longitudinal Study of Adolescent Health (AddHealth) sample and, if so, whether demographic, parent, or school characteristics explain the disparity.  Using data from Waves 1 and 4, we compare educational outcomes for adolescents and young adults who identify as lesbian, gay, or bisexual with their heterosexual–identified counterparts (N=6753). We also compare education by consistently heterosexual (Wave 1 opposite sex-attraction and Wave 4 heterosexual identification) and consistently homosexual (Wave 1 same-sex attraction and Wave 4 homosexual identification) affiliation with inconsistent-reporting participants. We use Wave 4 self-reports of educational attainment and demographics as well as parent-reported education and school-reported geographical information from Wave 1.  

This study attempts to resolve conflicting information in earlier literature regarding the relationship between sexual orientation and educational attainment. We use ordinary least squares multivariate regression analyses to examine differences in education by sexual orientation. We display three models for educational attainment using both the Wave 4 and consistent categorization of sexual orientation. We then control for age, race, parent education, and school geography in a stepwise hierarchical analysis. 

Preliminary results for males extends beyond previous research on educational attainment for LGB populations, suggesting sustained differences in education for 100% homosexual-identified adults (β_homosexual=0.950, p=0.000) and 100% heterosexual adults, even when controlling for parent education and school geography (β_homosexual=1.422, p=0.000).  While sexual orientation does not account for a large amount of the overall variation in educational outcomes, our findings suggest the difference between AddHealth respondents identifying as 100% heterosexual and non-heterosexual respondents is robust.  The difference between 100% heterosexual respondents and 100% homosexual respondents increases when we add controls for age, race, income, parents’ education, and school geography.

This educational advantage is even greater for AddHealth participants consistently reporting sexual identity (β_heterosexual=.723, p=0.000; β_homosexual=1.568, p=0.000).  Despite the small number of participants in the consistently homosexual category, we find robust results after controlling for demographics, geography and parent education (β_heterosexual=.399, p<0.10; β_homosexual=1.637, p=0.000). Future analysis will examine female participants as well as additional school/family characteristics that may influence educational attainment.  Results may contribute to a better understanding of the unique links between sexual orientation and educational attainment in the context of diverse schools.