Conventional wisdom says that voter participation in closely contested elections is higher because of the inherent competitiveness. The logic is that people feel that their vote could be decisive in a close election and hence more turn out to vote. But is that really true? Ron Shachar has crunched the numbers from three Presidential elections using some advanced statistical analysis and says that the answer is a bit more complicated than that.     
While the analysis is too complex to detail here, the basic idea is quite straightforward. Does the competitiveness of an election (defined as the closeness of the race as measured by vote percentage of the two major candidates) always predict turnout, or does its effect disappear when marketing variables are included in the model? Previous research either had not considered this approach or had insufficient marketing variables to use in the model, leading to the conclusion that closeness predicts turnout. 
In this research data on marketing variables such as advertising and grass roots campaigning were available and included in the analysis. It turns out that when you do that, the impact of the closeness on turnout disappears. What does this mean? It means that closeness of the election drives turnout only indirectly through marketing expenditure. In fact, it is even possible in such a model to calculate the precise   impact of the marketing variables.
For example, increasing the proportion of the population contacted by both parties by 10% (i.e. increasing the grassroots efforts) leads to an increase of 2% points in participation, which of course is a huge improvement. Overall Shachar finds that if all marketing activity for the 2004 election had been cancelled, the number of voters would have decreased by 15 million!
What this research really shows is the precise and dramatic impact of marketing variables on turnout. It is not just a question of belief on the political consultants. Spending money appropriately does turn out the vote.
This research was conducted by Ron Shachar who is a Professor of Marketing at Tel Aviv University and is currently a Visiting Associate Professor of Marketing at Duke University. It was published in the Journal of Marketing Research.