by Alan Abramowitz
The new issue of Barron’s Magazine, always a model of objective journalism, has a cover story by Jim McTague that argues that reports of a coming Democratic victory in the 2006 midterm elections are greatly exaggerated.
A lot of McTague’s “analysis” appears to consist of little more than wishful thinking. For example, he predicts that Rick Santorum, who has been trailing Bob Casey, Jr. in every poll in the last six months, will win reelection in Pennsylvania thanks to a late surge in support from the western part of the state and that Mark Kennedy will defy polls showing him trailing by double-digits to defeat Amy Klobuchar in Minnesota.
Beyond wishful thinking, McTague’s argument that Republicans will keep control of the House and Senate rests almost exclusively on the fact that most endangered Republican incumbents have raised more money than their Democratic challengers and, in both 2002 and 2004, the candidate who spent the most money in a House and Senate race almost always won.
But there is a fundamental flaw in this argument: 2002 and 2004 were not wave elections–elections in which there is a strong national tide. In wave elections lots of incumbents lose even though they outspend their challengers. This is what happened in 1974, 1980, 1982, and 1994. In 1994, for example, 26 of the 34 Democratic incumbents who lost their seats outspent their Republican challengers. On average, losing Democratic incumbents outspent their Republican challengers by a margin of $969,000 to $663,000. Republicans also won 14 of 25 open seat races in which the Republican candidate spent less than the Democratic candidate.
Using the the relative size of the candidates’ campaign warchests to predict election results in a wave election can yield highly misleading result. If a strong Democratic wave hits the House and Senate on November 7th, as now appears likely, many Republican incumbents will lose despite outspending their Democratic challengers.