Using Battleground States as a Natural Experiment to Test Theories of Voting
35 Pages Posted: 13 Aug 2009 Last revised: 31 Aug 2009
Date Written: 2009
In this paper, we employ a longer time series (1980-2008) than previous research to gauge the effect of status as a battleground (or, pivotal) state on state-level turnout. Our model of state-level turnout also includes (1) midterm elections, which allow us to directly compare the effect of battleground status on turnout to the across the board increase associated with the presidential election and (2) state fixed effects, which capture persistent state-level factors related to turnout. We find that the effect of the presidential election dwarfs the effect of being a battleground state (by a ratio of about 8:1); for all the attention battleground states receive, turnout is only about two percentage points higher than in non battleground states. The large increase in turnout in presidential elections in all states, battleground and spectator states alike, suggests turnout levels are primarily linked to factors that affect the entire electorate, such as the importance of the presidency, rather than factors that influence just a portion of the country, such as intensive campaigning and mobilization efforts or greater chance of casting a decisive vote.
Suggested Citation: Suggested Citation