Such a takeover remains unlikely, however. Our House forecasting model currently gives the Democrats about a one percent chance of winning a House majority in 2014. To understand why, it is worth exploring the nature of the “midterm penalty” for the president’s party — what it is, why it exists, and what it means in 2014. Ironically, one of the biggest obstacles to a Democratic takeover is their success at holding the White House in 2012.
One of the perennials of American politics is that the president’s party loses seats in midterm elections, especially in the House of Representatives. Only three midterm elections in the last century featured a seat gain for the president’s party in the House: 1934, 1998, and 2002. Of those, the largest was 9 seats in 2002.
That two of those three anomalous midterms occurred recently might suggest that something has changed. But the 1998 election took place as Republicans geared up for impeachment proceedings against President Clinton — a rare event if ever there was one. (The 1934 election, of course, occurred in the depths of the Great Depression.) And the 2002 election was fairly well-predicted by our model, suggesting that the outcome, however anomalous historically, was not out of line with the underlying fundamentals. So it seems fair to expect that the penalty will continue, as it did in 2006 and 2010.
There is no one established explanation for the midterm penalty, although several theories have been developed and tested. (An excellent summary of this research is in this paper (gated) by Olle Folke and James Snyder.)
One theory is that higher turnout in presidential years consists disproportionately of people who support the winning candidate, and many of them stay home in midterms.
Another is that, in some presidential elections, one party has a really good year — winning seats it probably shouldn’t hold and thus leaving it open to losses the next election.
A final theory is that some voters react so negatively to the president’s policy agenda that they turn against him in the midterm. In essence, these voters help elect members of the opposite party to “balance” the president’s influence. If this theory is true, then the conservative turn in public opinion during the Obama presidency may underlie the “shellacking” the Democrats received in 2010.
The midterm penalty is one key reason why our forecasting model suggests the Democrats are in such a tough spot this year. In the model, the president’s party receives just over two percent less of the House vote in midterm elections than in presidential elections, even after accounting for some of the other reasons why the president’s party might be unpopular.
The graph below shows what our House forecast would be with and without the midterm penalty while varying Obama’s approval rating from its current level to near its historic high.
The challenge the Democrats are facing becomes clear. Assuming an average midterm penalty, there is no reasonable level of Obama approval that is likely to give Democrats a House majority in a midterm election.
But if the penalty were taken away, prospects would look better. For example, if Obama had Bill Clinton’s 1996 approval (55 percent) and it were a presidential election year, the Democrats would have about a 20 percent chance of claiming the majority instead of the 5 percent chance we estimate they would have if Obama’s approval hits 55 percent this year.
In general, large midterm gains almost always appear to be a reaction against the sitting president. This is perhaps the one disadvantage of Obama’s reelection for Democrats. If there were currently a Republican president, and that person were as unpopular as Obama is now, the midterm penalty would work to the Democrats’ advantage: Democrats would have a 67 percent chance of taking back the House.
That is not enough reason for Democrats to yearn for President Romney, of course. Nevertheless, the combination of a Democratic president and the midterm penalty is helping to keep them from controlling both ends of Pennsylvania Avenue.