In a presidential election, not all votes are of equal value. Because the ultimate decision rests with the Electoral College, votes carry more weight in some states than in others. As a result, campaigns face powerful incentives to concentrate their resources geographically. By 2000, only a minority of states comprising only a minority of the electorate were worth campaigning in at all. But these states were something of a microcosm of the country as a whole, and this made the campaign into a natural experiment: One set of voters got the ad and visit “treatment,” the other did not.

This chapter starts by examining the allocation of resources by campaigns – of candidates' time and of money spent on advertising. The first task is to identify the primary strategic incentive governing allocation of effort across states – which is more important about a state, its size or the competitiveness of its presidential race? Campaigns also must decide how to distribute ads over time – should they spend resources early or should they store them up for the end? The study of resources culminates with the bottom line, the impact of visits and ads on vote intention. In 2000, ads made a difference at the margin, and the Bush campaign was the beneficiary when it counted most, at the end. This is one piece of evidence about the natural experiment.

Advertising is not the only media component of a campaign. Candidates also made use of the “free” media.