Jump to Any Race
House Overview|By David Wasserman, December 12, 2013

If the Murray-Ryan budget deal amounts to a small lurch forward, the last few months' politics have sure felt like a roller-coaster ride. And, we aren't just talking about GOP Reps. Trey Radel (FL-19) or Steve Stockman (TX-36).

For the first sixteen days of October, the government shutdown generated conditions that felt a lot like 2006, with independent voters ready to vent their anger at Republicans. Soon thereafter, the roll-out of the Affordable Care Act generated an atmosphere that felt similar to 2010, when President Obama and Democrats got routed. As Charlie Cook pointed out, gyrations of this magnitude in short time spans are exceedingly rare.

At the moment, the political environment appears to have come back down to earth. And, with the 2014 election back to looking more like a referendum on President Obama than House Republicans, we have updated our outlook to a GOP gain of zero to ten House seats.

The candidate most symbolic of the times is Democratic Omaha Councilman Pete Festersen, who entered the race against weak GOP Rep. Lee Terry (NE-02) in the midst of the shutdown but dropped out this week. This shouldn't come as a shock: every cycle has a "gut check" time when candidates reevaluate the climate or their own capabilities, and many candidates who come storming out of the gate in off-years find they can't sustain their momentum.

But for Democrats to have really built on their October progress, they would have needed 1) the promise of more Republican intransigence on continuing resolutions and debt ceilings, 2) more Republican retirements from marginal or semi-marginal districts, and 3) a raft of five to ten more "grade A" candidates in GOP-held districts. In the aftermath of the ACA's launch, none of the three have materialized.

Towards the very end of the government shutdown, the HuffPost Pollster average of the congressional generic ballot showed Democrats peaking at roughly a 45 percent to 39 percent lead, approaching the point at which the House GOP majority might be in danger. Post-ACA roll-out, the latest average shows Republicans leading 40 percent to 37 percent, with many more undecided voters - a result that points towards small GOP gains.

Many wondered whether Democrats' shutdown momentum could last an entire year, and it turns out it didn't even last a month. The slow-drip nature of the ACA roll-out means it is a more durable problem for Democrats, but odds are it won't be the electoral catastrophe it felt like last month.

There are signs that President Obama's approval ratings have begun to stabilize, and that Democrats have hit their own nadir (much as the GOP did in October). One smart Democratic strategist takes solace and says "This is the low point. It is hard to imagine another scenario as bad for Democrats as what we just went through." The same strategist calls 2014 a "Net Zero" election, one in which voters don't want to reward either party.

Fundamentally, midterm elections tend to be referendums on presidents, not Congress - unless Congress does something (like impeach a popular president, in 1998, or the recent shutdown) to steal the spotlight in a negative way. And if the House GOP were to pass the Murray-Ryan deal, it would show that House Speaker John Boehner and his lieutenants have gotten the "monkey off their back" and are on track to avert a similar spotlight in 2014.

The problem for Democrats is, even a "net zero" or "neutral" election - one in which there is no real wind at either party's back or voters feel reluctant to vote for either party - is likely to produce small Republican gains in the House. We have introduced a new tool below to quantify the partisan "tilts" of past election cycles to demonstrate why this is the case.

Our newest House ratings count 43 competitive races in 25 Democratic-held and 18 GOP-held seats. But of the 25 Democratic seats, 10 are in the Toss Up column. Of the 18 GOP seats, just three are in the Toss Up column. If the election were held today, Democrats would need to win 42 of these 43 competitive races to win the House, a virtually impossible task.

Hot off the press today, Democracy Corps's Battleground Survey, a terrific partnership between Democratic pollster Stan Greenberg and Democratic strategist James Carville, corroborates our ratings. In the most competitive GOP-held seats, named Republican incumbents led generic Democratic challengers 47 percent to 42 percent. But in vulnerable Democratic seats, Democratic incumbents led on the ballot by just 43 percent to 42 percent.

A Tale of Two Midterms: 2006 and 2010

The recent volatility in the political mood has generated a lot of questions in our inbox along the lines of "If Democrats have a great night in November 2014, what's the upper limit of seats they could gain?" Or, "If Democrats unravel and everything is going right for Republicans, what's the maximum number of seats they could pick up?"

In the last several decades, the most successful midterm election for Democrats was the 2006 wave, and the most successful GOP midterm was the 2010 wave. Most would agree that a 2006-like year would be an absolute best-case scenario for Democrats in 2014, a 2010-like year would be their absolute worst-case scenario, and that next November is almost certain to be somewhere in between those two extremes.

However, redistricting took place after these two elections. So is it possible to simulate would happen if a 2006 or 2010 environment were applied to today's district lines? Actually, we think so.

Introducing the Cook Partisan "Tilt" Index

One statistic others have used to quantify the size of waves is the National House Popular Vote. For example, in 2012, Democrats won 59.6 million votes for House, and Republicans won 58.2 million. The imperfection with this metric, however, is that there are many uncontested seats where there are either no votes counted for one party or no votes are counted at all.

So we have devised a new way to measure the partisan "tilt" of any given election year. First, narrow the universe of House races down to those that featured candidates from both major parties (in the past six cycles, this number has ranged from 353 to 406). Then, measure how much better or worse the average Democratic candidate performed versus the "normal" partisan lean of his or her district, using our Partisan Voter Index of districts.

For example, in a D+1 district, a "generic" Democratic candidate might expect to receive 51 percent of the two-party vote in a "neutral" year. But if the 2006 political environment "tilted" to Democrats by four points nationally, that Democrat might have received 55 percent of the vote. And, if the 2010 political environment "tilted" to Republicans by two points nationally, that candidate might only have taken 49 percent that year.

We calculated the average partisan "tilt" of every contested House election for each cycle between 2002 and 2012, and this is what we found:

"Tilt" by Year

Average Two-Party Vote Share vs. Cook PVI in Contested* House Races, 2002-2012
Election YearContested Seats*Average Party Share vs. PVIDems Elected**GOP Elected
20023531.06% More Republican205230
20043700.45% More Democratic203232
20063803.94% More Democratic233202
20083775.21% More Democratic257178
20104061.95% More Republican193242
20123911.51% More Democratic201234

*Contested races are races in which both a Democrat and a Republican appeared on the ballot.

**For 2002 and 2004, Vermont Rep. Bernard Sanders (I) is counted as a Democrat.

Two findings stand out to us: first, Democrats have performed, on average, significantly better in presidential elections than midterms, which supports the idea of the GOP's midterm turnout advantage. Second, even though 2012 featured a stronger Democratic "tilt" than both 2002 and 2004, Democrats won fewer seats under the new lines in 2012 than they did in either of those years.

In fact, the closest thing to a "neutral" political environment in House races we have experienced in the last ten years was 2004, when Democrats performed less than half a point better than their districts' PVI on average. That year, they won 203 seats. But in 2012, Democratic candidates outperformed their PVI by 1.5 points, and Democrats won just 201 seats, illustrating the power of both Democratic clustering in cities and Republican gerrymandering.

So, what would happen if the 2014 environment approximated that of 2006? Would Democrats gain the House? According to our "tilt" index, a 2006-style election would represent a swing of 2.43 points further in Democrats' direction from 2012 (1.51 to 3.94). So, assuming a uniform swing in House races from 2012, any Republicans who won with less than 52.43 percent of the vote in 2012 might therefore be "underwater" in 2014.

Using this method, even 2006-like conditions would only net Democrats 12 seats, short of the 17 they need. To be sure, there is never a uniform swing in the House from election to election. But, there are about as many vulnerable Republicans who aren't on this list (such as Reps. Gary Miller and Steve Southerland) as there are Republicans who won close races last year and are no longer vulnerable (like Rep. Michele Bachmann).

House Republicans Who Won with Less Than 52.4% of the Two-Party Vote in 2012

DistrictIncumbent2012 Two-Party ShareCurrent Rating
IL-13Rodney Davis50.18%Lean Republican
MI-01Dan Benishek50.28%Lean Republican
MN-06Michele Bachmann50.61%Solid Republican
IN-02Jackie Walorski50.74%Lean Republican
NY-27Chris Collins50.79%Solid Republican
NE-02Lee Terry50.80%Likely Republican
CO-06Mike Coffman51.09%Toss Up
PA-12Keith Rothfus51.74%Solid Republican
FL-10Daniel Webster51.74%Likely Republican
NY-23Thomas Reed51.91%Lean Republican
KY-06Andy Barr52.00%Likely Republican
OH-16Jim Renacci52.05%Solid Republican

Now, let's swing the pendulum in the other direction, and pretend 2014 is a lot like 2010. According to our "tilt" index, a 2010-style election would represent a swing of 3.46 points in Republicans' direction from 2012. There are 21 Democrats who won with less than 53.5 percent of the vote in 2012, so a "best-case" scenario for Republicans might produce a pickup of around 21 seats:

House Democrats Who Won with Less Than 53.5% of the Two-Party Vote in 2012

DistrictIncumbent2012 Two-Party ShareCurrent Rating
NC-07Mike McIntyre50.10%Lean Democratic
UT-04Jim Matheson50.16%Lean Democratic
FL-18Patrick Murphy50.29%Toss Up
AZ-02Ron Barber50.42%Toss Up
MA-06John Tierney50.61%Lean Democratic
IL-10Brad Schneider50.63%Toss Up
NY-21Bill Owens51.00%Lean Democratic
CA-52Scott Peters51.18%Toss Up
CT-05Elizabeth Esty51.31%Likely Democratic
CA-07Ami Bera51.68%Toss Up
AZ-01Ann Kirkpatrick51.94%Toss Up
NY-18Sean Patrick Maloney51.95%Lean Democratic
NH-01Carol Shea-Porter51.96%Toss Up
AZ-09Kyrsten Sinema52.20%Lean Democratic
TX-23Pete Gallego52.48%Lean Democratic
NY-01Timothy Bishop52.49%Lean Democratic
NH-02Ann Kuster52.53%Lean Democratic
CA-26Julia Brownley52.69%Lean Democratic
CA-36Raul Ruiz52.94%Toss Up
NY-25Daniel Maffei52.96%Likely Democratic
IL-17Cheri Bustos53.28%Lean Democratic

It's possible to simulate each party's seat share in the House under today's lines for every election since 2002. As it turns out, Democrats would need a monster wave like 2008 (a pro-Democratic turnout we have not seen in a recent midterm) to win a narrow majority. And, even if 2014 were a "neutral" election like 2004, Republicans would pick up about seven House seats:

If Election 2014 Featured the Same Conditions of Election ________...

Election YearDem EstimateGOP EstimateSwing from 2012
2002184251Republicans +17
2004194241Republicans +7
2006213222Democrats +12
2008221214Democrats +20
2010180255Republicans +21
2012201234No Net Change

The Bottom Line

As long as the 2014 cycle doesn't take a drastic turn, the "best case scenario" for House Democrats next fall is probably a gain of 12 seats, and the "worst case scenario" is a loss of 21 seats. However, given the current map, even a return to a "neutral" political environment is likely to produce a small Republican gain. As such, we have updated our House outlook to reflect a GOP gain of between zero and ten seats.

We are also updating our ratings in four districts, all benefiting Republicans:

Ratings Changes

DistrictIncumbentRatings Change
NE-02Lee Terry (R)Lean R to Likely R
NH-02Ann McLane Kuster (D)Likely D to Lean D
OH-14David Joyce (R)Lean R to Likely R
PA-07Pat Meehan (R)Likely R to Solid R

Updated Bottom Lines

NE-02: Lee Terry (R) - East: Omaha and suburbs
Likely Republican. Democrats jumped for joy when Omaha Councilman Pete Festersen launched his campaign at the height of antipathy towards the government shutdown. But after Festersen dropped out, Democrats are back to the drawing board. Terry is a relatively weak GOP incumbent, but is in much better shape than he was even a month ago.

NH-02: Ann McLane Kuster (D) - West: Nashua, Concord, Keene
Lean Democratic. Republicans are jumping all over a video of a November town hall meeting in which Kuster appeared at a loss to address the Benghazi controversy, but our real reason for changing our rating of this race is the entry of two credible Republicans: former state Sen. Gary Lambert, an Iraq veteran, and state Rep. Marilinda Garcia, regarded as a rising star in the party.

OH-14: David Joyce (R) - Northeast: Lake County, Ashtabula
Likely Republican. Joyce, a former Geauga County prosecutor, was a blank slate when he replaced retiring Rep. Steve LaTourette on the ballot in 2012 and got away without a Democratic challenge. But he has mostly voted in the same moderate mold as LaTourette, and Democrats in DC don't appear all that enthusiastic about attorney Michael Wager's progress in the race.

PA-07: Pat Meehan (R) - Southeast: most of Delaware County
Solid Republican. Meehan has emerged as one of the most middle-of-the-road voices in the 2010 freshman class in the House, becoming one of just 20 Republicans to call for a clean continuing resolution during the government shutdown. A former prosecutor with a non-political reputation and a district redrawn to protect him, it's easy to see why Democrats are having such a tough time recruiting here.