Most pundits give Democrats a pretty slim chance of retaking the House. But a new model out of Princeton says they not only have a shot — they're the odds-on favorites.
Sam Wang, a neuroscientist at Princeton who works on election modeling as a hobby, has found than in recent elections, national polling that asks voters which party they plan on supporting in House elections has done an excellent job of predicting what percent of the popular vote each party will get in House elections in November. For example, in 2010 Republicans had a 7 point lead in national generic ballot polls, and won the national House popular vote by 6.7 points.
Right now, however, Democrats are far ahead Republicans in national House polling at the moment. Here are the Huffington Post averages since early summer:
The median poll has a Democratic lead of 4 points, with a standard of error of plus or minus 2 points. By Wang's estimates, that translates to an average Democrats victory of 24 seats. Even if you adjust to take into account the advantage Republicans get due to being incumbents, Democrats only need to win by 1.3 percentage points to retake the House. And they're set to win by significantly more than that.
Of course, these polls could change in the next six weeks. Wang figures that the Democrats' victory margin could move up or down 4 points before the election. Given that possible variation, he estimates that Democrats have a 74 percent chance of retaking the House.
Combined with Nate Silver's estimates that Democrats have a 79 percent chance of holding the Senate, and the fact that most models give Obama a better chance of winning than Romney, if Wang's prediction holds true it 2013-2014 will be 2009-10 redux, with Democrats will have all branches of government, but the filibuster will still make life difficult.
It's worth noting, however, that other models give different results. John Sides and the team at the Monkey Cage have a model that uses GDP, the president's party and approval rating, incumbency, and district-level presidential vote, rather than House polling. Their model gets the seat margin wrong by 2.61 seats, on average, much lower than Wang's error. It gives Republicans a three out of four chance of keeping the House.
So Wang's model isn't definitive. But it does mean one should keep an eye on Congressional ballot polls going forward.
(Link to Sam Wang's study via Zack Beauchamp)