The Role of Selection Effects in the Contact Hypothesis: Results from a U.S. National Survey on Sexual Prejudice

Abstract

Empirical research has documented that contact with lesbians and gays is associated with more positive feelings toward and greater support for legal rights for them, but we know less about whether these effects extend to informal aspects of same-sex relationships, such as reactions to public displays of affection. Furthermore, many studies have assumed that contact influences levels of sexual prejudice; however, the possibility of selection effects, in which less sexually prejudiced people have contact, and more sexually prejudiced people do not, raises some doubts about this assumption. We used original data from a nationally representative sample of heterosexuals to determine whether those reporting contact with a lesbian, gay, bisexual, or transgender friend or relative exhibited less sexual prejudice toward lesbian and gay couples than those without contact. This study examined the effect of contact on attitudes toward formal rights and a relatively unexplored dimension, informal privileges. We estimated the effect of having contact using traditional (ordinary least squares regression) methods before accounting for selection effects using propensity score matching. After accounting for selection effects, we found no significant differences between the attitudes of those who had contact and those who did not, for either formal or informal measures. Thus, selection effects appeared to play a pivotal role in confounding the link between contact and sexual prejudice, and future studies should exercise caution in interpreting results that do not account for such selection effects.