Direct Democracy, Agendas, and Presidential Vote: Gay Marriage and the 2004 Election

This study draws from agenda-setting and issue priming theories to develop hypotheses about how state-level ballot propositions may affect presidential elections. We propose that ballot measures prime voters to evaluate candidates in terms of policy questions placed on their state ballot. We test if concerns about gay marriage in 2004 had greater salience for voters in states where proposals to ban gay marriage were on the ballot, and test the impact of gay marriage on presidential vote choice. We find that gay marriage was more likely to be cited as an issue used to evaluate candidates by some residents of states with marriage ban propositions. We also find respondents who were concerned about gay marriage who resided in states with the issue on the ballot were more likely to vote for Bush. The research is among the first to show that state ballot propositions may have subtle but important effects in presidential elections.