The House Is Very Much In Play

The House Is Very Much In Play

By Sean Trende - January 19, 2010

With Scott Brown rushing toward the finish line with Martha Coakley in D+12 Massachusetts, apparently (if the polls are to be believed) ahead by a nose, you’d think that the political establishment would slowly wake up to the fact that we're in a political environment that is at least as bad as 1994. But amazingly, the establishment’s slumber continues. The consensus among the political class remains that the House not only isn’t in play, but that it is crazy talk to suggest that control could flip.

These pundits are absolutely wrong. While I think the most likely scenario right now is that the Republicans will pick up somewhere in the neighborhood of 35 seats, Democratic losses could easily go higher. In fact, I’d say the GOP currently has about a one-in-three chance of getting the 40 seats they need to take back the House, with a bias toward higher gains. Pundits have discussed the economy, health care, and last November’s elections ad nauseam, so I won’t rehash this. I think the real reasons that the House is in play are as follows:

Open Seats

The most prominent reason you will hear from the makers of conventional wisdom as to why the GOP can’t take the House is that there are not enough open seats. The GOP only knocked off 34 incumbents in 1994, while it needed (and presently needs) 40 seats to take the majority. 22 of the 52 seats that it picked up that year came from vulnerable open seats (it also lost four open seats of its own). If you look at Stu Rothenberg’s latest numbers, there are only eight open Democratic House seats rated “tossup” or better for Republicans. That’s obviously a lot less than 22, so Republican chances for big gains are arguably limited.

As a threshold matter, it’s probably worth noting that if the GOP knocks off 34 Democratic incumbents again, and picks up the eight open seats rated as tossups or leaning its way, it would just barely take back the House (even losing the Castle and Cao seats).

But overall, the relentless focus on open seats is misguided in a wave election, which 2010 is shaping up to be. This focus is justified in a “normal” year, where a President is cruising at an approval rating in the mid-50s, the economy is doing reasonably well and a party is not overextended into the opposing Party’s territory. In those elections, few incumbents are threatened, and most of the action is in open seats.

Wave elections are different. In wave elections, many incumbents suddenly find themselves in deep trouble, and choose to retire. In other words, many of the incumbents who retire do so because they were likely to lose anyway. Think of Vic Snyder or Byron Dorgan this year, incumbents who were very much in danger of losing their seats before they decided to retire, and whose polling almost certainly played a role in their decisions to quit.

Look at the actual open seats in 1994. In 1994, one of the open seats Republicans picked up was Mike Synar’s in eastern Oklahoma (now mostly in Dan Boren’s district). Synar’s district was open because he lost the primary to Virgil Cooper, a retired high school principal who spent around $20,000 on business cards that he slipped under the windshield wipers of people’s cars. Had Synar made it to the general election, it is highly unlikely that he’d have been successful against the much better-funded Tom Coburn.  Or consider Marilyn Lloyd. In 1992, a fairly good Democratic year, she beat Zach Wamp by a little more than a point. In the toxic environment of 1994, it is difficult to imagine that she would have won.

Or, to turn the partisan tables, think of 2006. Bob Beauprez of Colorado and Jim Nussle of Iowa’s districts both went to the Democrats when they ran for Governor, but it is hard to imagine either of them winning had they stayed in the race – Jim Leach lost in a politically similar district neighboring Nussle’s, while Beauprez’s district was exactly the type of suburban district the GOP had trouble holding that year (think about what happened around Philly).

These aren’t isolated examples. A few months ago, I constructed a regression analysis that examined how critical votes, opponent fundraising, and the incumbent’s 1992 vote share affected the performances of Democrats running for re-election in red districts in 1994. We can use the formula from this regression equation to get a back-of-the-envelope look at how the Democrats who retired would have done in 1994 had they run. Using this, I estimate that eleven of them would have lost anyway had they run for re-election, while another four would be within a point of losing. In other words, had the Democrats not had any open seats in 1994, the GOP nevertheless would have had a good shot at taking back the House (though without the other ten seats, they may not have had working control of the House).

More open seats would make the GOP’s job easier, but they aren’t necessary for the GOP to take back the House.

Vulnerable Incumbents

Moreover, the 34 incumbents that lost in 1994 do not represent a magical cap for the Democrats’ potential losses. The granddaddy of all midterm elections for defeating incumbents in the past hundred years was 1922, when the Democrats beat 63 Republican incumbents (this can’t be blamed on redistricting, since no reapportionment occurred in the 1920s). In 1974, the Democrats defeated 35 Republicans, and won only a handful of open seats. In 1966, the Republicans unseated 38 Democratic incumbents. In 1946, the Republicans defeated 45 Democrats, in 1942 they defeated 40, in 1938 they beat 60, while in 1914 they knocked off 48. This covers most of the great midterm backlashes of the past Century; the others are 1930 (41 Republicans lost) 1958 (34 Republicans lost) and 2006 (22 Republicans lost). The party out of power routinely knocks off more than 34 incumbents in wave elections.

Now, I’m fully aware of the problems with going back that far – certainly before World War II. Congressional races have changed a lot since then and incumbency advantages have grown. But there’s an important point to be made here: The absolute worst incumbent elections on this list – 1914, 1922, 1938, 1942, 1946, 1966 – were elections held immediately after the party in power had made major gains in the previous few elections.

This is immensely important. Although the Democrats are about as overextended today as they were in 1994 in terms of Democrats in red-leaning districts, this is a very young Democratic Congress. Of the seventy-three Democrats in Republican-leaning districts, forty are serving their first or second terms. This is almost twice as many freshman/sophomore Democrats in such seats as there were in 1994.

Moreover, this interacts with the first point above: With a younger caucus, there are going to be fewer retirements. A Congressman like Vic Snyder or a Senator like Byron Dorgan looks at this year and says “it’s been a good career, I should go out a winner.” But for someone like Steve Driehaus, there’s not much downside to giving this a go, and quite a bit of upside.  But this doesn't mean that there are going to be fewer defeated incumbents, since these Freshmen and Sophomores typically represent the most vulnerable type of incumbents.


The polling so far for red district Democrats has been uniformly atrocious, and thoroughly supports the argument that the GOP can defeat 35 incumbents or so. Polls have shown Vic Snyder (AR-02), Frank Kratovil (MD-01) (a partisan poll, but also one that was never answered by the Kratovil campaign), Steve Driehaus (OH-01) and Dina Titus polling around 40%. Those are numbers from which incumbents rarely come back.

Democrats who have avoided voting the party line like Stephanie Herseth Sandlin (SD-AL) and Larry Kissell (NC-08) are polling around 50%, which is only barely outside the danger zone for incumbents. The list of Democrats from red districts who voted “yes” on the stimulus, health care bill, and cap and trade legislation is lengthy, and given this polling, most of these Democrats are likely at least somewhat vulnerable, even against relatively unknown opponents. The fact that Martha Coakley is in bad shape in Massachusetts, and that Barbara Boxer is below 50% against relatively unknown opponents in California can’t be comforting to Democrats whose districts more closely resemble Texas than California.

In the absence of more polling of individual House races, we can look to the generic House ballot. It is ultimately of limited utility, because ultimately voters will choose between two candidates, not “Republican” or “Democrat.” But nevertheless, the national vote gives a pretty good gauge as to which party controls the House; you’d have to go back to 1942 to find a party that won the national vote by an appreciable amount, yet failed to take back the House (depending how you count unopposed candidates, Democrats could arguably have narrowly won the national popular vote in 1996 without taking back the House).

If you look at the RCP average, you see a sea of red markings showing Republicans leading the generic ballot test. It isn’t just Rasmussen Reports: CNN, Battleground, Bloomberg, and Gallup have all measured the GOP ahead in the generic ballot in the past few months.

This is highly unusual. The GOP typically overperforms generic ballot models. In 2004 – a pretty good House year for the GOP – the GOP led in 9 ballot tests all year. In the mini-tsunami of 2002 the polls looked pretty good for Republicans, but probably still not as good as they did today.


Obviously, a lot could change between now and November. It is still early, and there’s a pretty high error margin on any projection at this point.

But the conditions for the Democrats are not good. When you look at the polling, the conditions on the ground, and the datapoints we have from the races in the past few months, it is hard to avoid the conclusion that the GOP can win the House back. This election presently combines the removal of Presidential coattails that drove 1946 in part with the bad economy that drove 1958 with the overambitious agendas that drove 1966 and 1994. If forecasting elections at this point is like forecasting where a hurricane will run aground – the range becomes narrower as the time of impact draws closer – right now a perfect storm is brewing and a GOP takeover of the House is well within it potentially projected path. The storm may dissipate before it runs aground. But it may also intensify. This could get very ugly for the party in power.

Sean Trende is senior elections analyst for RealClearPolitics. He is a co-author of the 2014 Almanac of American Politics and author of The Lost Majority. He can be reached at Follow him on Twitter @SeanTrende.

Sean Trende

Author Archive

Follow Real Clear Politics

Latest On Twitter