THE MISMEASURE OF CAMPAIGN SPENDING - EVIDENCE FROM THE 1990 UNITED-STATES HOUSE ELECTIONS

Research on the effects of campaign spending on elections has used total campaign expenditures to measure the amount of money actually spent on campaigns. This article examines a unique data set on the detailed expenditures of House candidates in 1990. We find that the mismeasure of campaign spending seriously affects the magnitude of conventional estimates of the effects of money on votes but does not account for the peculiar result that incumbent money matters less than challenger money. In fact, when correctly measured incumbent expenditures have a significant negative coefficient, evidence that standard estimates suffer from still greater problems of statistical specification.