Objective. This article uses social movement theory to explain variation in local support for proposed constitutional amendments to ban same-sex marriage in 22 states during 2004 and 2006.
Methods. The analysis uses OLS regression with county-level data to explain variation in local support for the amendments.
Results. Support for the amendments in both years was positively related to the proportion of a county that was evangelical or Republican, but negatively related to its level of education and proportion of Catholics. Amendment support was positively related in only one year to the percentage of a county's population that was professional, young, black, in female-headed households, and Mormon. Amendment support was negatively related to the concentration of gay organizations in 2006, but positively related to the presence of mega-churches in 2006.
Conclusions. Community characteristics have a substantial effect on levels of support for same-sex marriage bans, thus reinforcing the utility of cultural explanations of policy decisions.