Redistricting and Incumbent Protection in 2001-2002
Elections to the U.S. House of Representatives in 2002 were dramatically less competitive than elections after the last round of redistricting in 1991-1992. In 2002 fewer than one in ten races were won by less than 10% and fewer than one five races were won by less than 20% -- less than half of the number of races won by those margins in 1992. Only four incumbents were defeated by non-incumbent challengers, the fewest number in history, and the average victory margin was nearly 40%. Our 2003 Dubious Democracy report [] has additional data quantifying this alarming state of affairs.

By forcing most incumbents to run in changed districts, redistricting typically results in the two elections immediately following redistricting (such as 1992 and 1994) being the most competitive elections of a given decade. But by most measures the 2002 U.S. House elections were the least competitive elections since 1988. This lack of competition in a post-redistricting election makes it quite possible that House elections toward the end of the decade will be less competitive than any in history.

Decrease in electoral competition is due to a combination of factors, perhaps the most important of which is an increase in consistent voting patterns in federal races that makes it much harder for candidates of one major party to win seats in territory controlled by the other major party. But redistricting clearly played an important role.

For example, some might ask whether 2002 was a “stay the course” year (as indeed was the case in 1988, when that very phrase was the mantra of George Bush in his successful presidential bid). But voters showed great restlessness in other elections. For example, half of the 36 gubernatorial races in 2002 resulted in a party change. Of 70 statewide races for governor and U.S. Senate, the winner had 55% or more of the vote in fewer than half (47%), and only 24% of these races were won by 60% or more. In contrast, the winner had 55% or more of the    vote in 91% (396) of the 435 U.S. House races, and fully 81% of House races were won by 60% or more.

The Center has quantified just exactly how incumbents were protected. Here are three examples:

    * Of incumbents who had close relatively close races in 2000, more than three out of four in 2002 ran in a district that was drawn to be more favorable for their party
    * Among all the races with incumbents, 20 districts were moved from being competitive/swing districts to generally safe districts for one party
    * In California and Texas, there were 14 districts in 2000 that were very competitive, with a partisanship measure of 47% to 53%. Using that same measure, there are only two such districts in 2002

In this short report, we address two questions:

1. What was the effect of redistricting for the incumbents who had the most competitive U.S. House races in 2000?

2. How did the distribution of safe and competitive seats in 2000 compare with the distribution in 2002 after the 2001-2 redistricting?

Question One - Incumbents in Close Races in 2000

To answer the first question, we looked at the 59 House races in which the 2000 winner had less than 55%. We then excluded three categories of races 1) states with only one district and Maine, since redistricting did not occur in those states; 2) districts in which the incumbent did not run for re
-election (we will note that some of those retirements were induced by redistricting); and 3) districts that in October 2002 lacked 2002 partisanship data (the 60 districts in the following states: Alabama, Hawaii, Kansas, Kentucky, Massachusetts, Mississippi, Nebraska, New Hampshire, Ohio, South Carolina). Those exceptions left 46 incumbents running for re-election in new districts in 2002 who had received less than 55% of the vote in 2000.

Of those 46 incumbents, 37 of them had their districts made safer in redistricting in that the presidential candidate of the incumbent’s party won a higher percentage of the vote in their 2002 district than the 2000 district; only nine districts were made less safe for these potentially vulnerable incumbents. Here is the breakdown of number of seats by party:

                          Safer                        Less Safe                    Total
Democrat            17                                  6                            23
Republican          20                                  3                            23
Total                   37                                  9                            46

The nine incumbents whose districts became less safe are listed. Note that hardly any had to run in significantly worse districts – meaning ones where the district was tilted more than 1.5% in the direction of the other party. The two exceptions were Republican Connie Morella and Democrat Bill Luther – and they were two of the four incumbents to lose to non-incumbent challengers in 2002. Note also that Washington state and Minnesota account for five of the nine incumbents made less safe; Washington uses a redistricting commission, and Minnesota was impacted by independent candidates Jesse Ventura's opposition to incumbent protection in redistricting.

Democrats (6)
Arkansas-4          Mike Ross (1.1% more Republican)
Minnesota-2        Bill Luther (2.3% more Republican – running in new CD #)
Minnesota-4        Betty McCollum (0.4% more Republican)
Texas-11             Chet Edwards (1.2% more Republican)
Wash.-1              Jay Inslee (0.8% more Democratic)
Wash.-2              Rick Larsen (0.4% more Democratic)

Republicans (3)
Indiana-8             John Hostettler (0.7% more Democratic)
Maryland-8         Connie Morella (5.0% more Democratic)
Minnesota-6        Mark Kennedy (1.4% more Democratic – running in new CD# )

In contrast to very limited damage done to threatened incumbents, there are numerous examples of pronounced moves to protect threatened incumbents. For example, Florida Republicans Clay Shaw and Ric Keller were the only candidates to win by close margins in Florida in 2000. Their districts were made on average 5% more Republican, meaning a margin swing of 10% in the Republican-drawn plan. They both won easily in 2002. In a further sign of partisanship in Florida, the five clearly safe Florida Democratic incumbents had their districts made even safer, but the two Democratic incumbents in potentially competitive seats had their districts made on average 4% more Republican – one of these incumbents, Karen Thurman, was one of the four incumbents to lose to a non-incumbent challenger.

Question 2 - Redistricting and number of competitive districts

To answer the second question, we looked at all the races in which a 2000 incumbent is running for re
-election and excluded the states with one district and Maine (which had not yet redistricted) and states for which in October 2002 we lacked partisanship data. We then classified each district as likely to be competitive or not in open seats based on whether the district partisanship (as measured by the relative vote in the presidential race) was between 45% to 55% or outside that range.

Out of 323 incumbents meeting our criteria above, we found that a total of 20 fewer incumbents (from 132 down to 112) now represent districts that are in our competitive range. The number of seats in each category is as follows:

             GOP Incumbents    Democratic Incumbents.   Total Incumbents
Year      Safe Competitive        Safe Competitive          Safe Competitive 2000      87          74               104          58                 191       132
2002      99          62               112          50                 211       112

Going a bit farther in this approach to incumbent protection, we looked at two states that have been highlighted as ones with incumbent-protection plans: California and Texas. In those two states, there were 14 districts (5 out of 30 in Texas, 9 out of 52 in California) in the 2000 election that we would categorize as very competitive -- partisanship measures of 47% to 53%. In the 2002 districts, however, there were only two out of 85 -- one in each state. No incumbents were defeated in those two states in 2002, and only four of those state’s 85 seats were decided by margins of less than 10%.

California is particularly pronounced in its gerrymander. In 2000, for example our model of making projections outlined in Monopoly Politics 2002 did not make projections in nine of the state's 52 House races. In 2002, the model projected winners in fully 52 of 53 races -- only not making a projection in the race to fill former Congressman Gary Condit's seat, which still had enough of a Democratic lean to contribute to a relatively comfortable Democratic win. On average, the six Democratic incumbents who had been in competitive districts had their district made 4.8% more Democratic, while the nine Republican incumbents in competitive districts had their district made on average 3.5% more Republican.

As a result, our model now projects all but one of current incumbents in California as comfortable winners going into 2004 – meaning projected to win by at least 10%. Note that these incumbents are quite safe in 2004 and their seats are likely safe for their party for the entire decade unless more seats become open. Our model projected 1,091 winners by at least 10% in 1996
-2002; none of these projected winners lost, and only 21 (less than 2%) did not win by 10%.