The Effects of Campaign Spending in House Elections: New Evidence for Old Arguments

Abstract

The question of how campaign spending affects election results remains open because the simultaneity problem has proven so intractable. Green and Krasno's (1988) recent attempt to solve the problem-and to demonstrate that marginal returns on spending are as great for incumbents as for challengers-again comes up short. A panel feature of the ABC NewslWashington Post Congressional District Poll conducted during the 1986 elections offers a fresh perspective on the question. Analysis of changes in voting intentions during the final six weeks of the campaign shows that, as the tainted OLS studies have invariably suggested, the amount spent by the challenger is far more important in accounting for voter's decisions than is the amount spent by the incumbent.