Most research on voting behaviour and elections has focused on cross-sectional analysis of particular elections. This technique is very useful in telling us about the influences on individual voter decision making. Less commonly research has focused on short-term electoral change, and has tried to explain specific election results with respect to one another, or in relation to long-term trends.
* College of William and Mary, Williamsburg. I would like to acknowledge the suggestions of Alan Abramowitz in the preparation of this article and the helpful comments of the referees and Editor on an earlier version.