"The Persuasion Region: A Theory of Electoral Change."
-
Link to current version of the paper.
-
Link to earlier version of the paper
discussed on
The Monkey Cage.
-
Earlier version presented at the APSA Annual Meeting And Exhibition, September 5, 2009, Toronto.
-
Do candidates win elections by mobilizing new voters or by
persuading existing voters? Despite its simplicity, this question is
not straightforward to answer. I begin with a simple theory of electoral
change, derived from the spatial model of voting, from which I
introduce the \emph{persuasion region}, a segment of the electorate
``up for grabs'' by both parties between one election and the next.
To illustrate the theory's explanatory power, I use a new Bayesian
model to leverage the information contained in precinct-level
election returns by merging these data with 36.6 million
individual-level voter registration and turnout records across four
years, three elections, and two states. I show that in election
pairs with large persuasion regions, both-election voters who switch
between the parties are a larger influence on election outcomes. In
elections with smaller persuasion regions, new voters are a greater
influence on the outcome. I then demonstrate the power of the
persuasion region, and the mechanism at work, using individual-level
panel data from the American National Election Studies. Fundamental
characteristics of competing candidates structure the relative
contribution of persuasion and mobilization to election outcomes.