According to conventional wisdom, the current scarcity of competitive races is the fault of redistricting, because state legislatures redrew boundaries with an eye toward protecting incumbents. But the findings by Abramowitz and Alexander challenge that notion.
The two political scientists point out that plenty of House districts are still competitive for presidential candidates even though they aren't for congressional candidates. After the redistricting following the 2000 census, 204 districts were "safe" at the presidential level for one of the parties -- an increase of five. (The professors defined districts as "safe" if a presidential candidate had won the district by at least 20 percentage points.) In contrast, after redistricting, just 114 seats looked competitive at the House-race level, a decrease of four. (A district was considered "competitive" if the incumbent had won by less than 10 points.)
Abramowitz and Alexander suggest that incumbency may simply be of greater value to House candidates than it was in the past. Many districts that once hosted pitched congressional battles are now peaceful. One example is my home district, Louisiana's 4th, now represented by Republican Jim McCrery. A rising star on the Ways and Means Committee, McCrery won a tough race in 1988 to get to Washington, but he hasn't had one since and is widely considered to be unbeatable.
Yet when McCrery publicly mused about retiring, he practically gave Republican Party officials heart attacks. They knew that without McCrery the district would be very competitive. And there are dozens of Jim McCrerys, incumbents in districts that, by their makeup, ought to be competitive -- and likely will be once the incumbent departs.
But occasionally, a once "unbeatable" incumbent loses. For years, Maryland's Rep. Connie Morella, a popular liberal Republican, faced only token opposition in an overwhelmingly Democratic district. Then in 1996, a little-known Democratic challenger ran a spirited, underfunded race, losing 39 percent to 61 percent, but alerting Democrats that Morella was potentially vulnerable. Two years later, veteran civil-rights activist Ralph Neas mounted a well-funded challenge but lost to Morella 40 percent to 60 percent. In 2000, a well-connected lobbyist ran a better-organized campaign and drew 46 percent of the vote.
In 2002, the Democratic-controlled Maryland Legislature added thousands of Democrats to the district. And Democrat Chris Van Hollen beat Morella, 52 percent to 47 percent. Now, Van Hollen's seat is extraordinarily safe.
For many years, political scientists put great weight on "surge and decline" as a way of explaining the outcome of congressional elections. This theory holds that when a presidential candidate wins, a considerable number of his party's congressional candidates also surge into office. But, when those freshmen later have to run on their own, many lose.
Abramowitz and Alexander note that the surge-and-decline pattern is less evident today than it was in the past. In the first half of the 20th century, the average number of seats gained by the winning presidential candidate's party was 30 in the House and four in the Senate; since 1976, the averages have dropped to only five House seats and one Senate seat. The subsequent average midterm election losses have also dropped, from 42 House and four Senate seats in the first half of the 20th century, to just 13 House seats and two Senate seats in the past 25 years.
Abramowitz and Alexander make a number of assertions, including, "The advantage of incumbency is largely responsible for lack of competition in individual House and Senate contests." The 99 percent re-election rate for House incumbents in 2002 was a record. Almost 90 percent of the senators who ran that year were re-elected. And 80 House members and three senators faced no major-party challenger. Abramowitz and Alexander note that even with record amounts of money being spent, few races were competitive in 2002. Perhaps money and technology have worked to enhance the strength of incumbency.