President Barack Obama's popularity is improving, but not enough. Photographer: Jewel Samad/AFP/Getty Images
President Barack Obama's popularity is improving, but not enough. Photographer: Jewel Samad/AFP/Getty Images

My good friends John Sides, Ben Highton and Eric McGhee updated their Senate prediction model this week. They explain their thinking over at the Monkey Cage. The headline number is very good for Republicans, who are now according to the model solid favorites to win a Senate majority. I highly recommend the three posts.

The most important takeaway is that the fundamentals are lined up against Democrats in this cycle. They are facing a midterm election, which plays badly for the incumbent party. In addition, the elections take place during the president's second term and follow a big first-term win that dragged several unlikely Democratic winners into the Senate. Also in the Republican's favor is that there are plenty of Republican states with Senate elections this cycle. Retirements, too, have played out well for Republicans, though that is partially a function of the political context and partially just luck of the draw.

That’s the lay of the land, regardless of what anyone is doing. For a comparison, look at 1986, when Democrats won a Senate majority in similar circumstances despite Ronald Reagan's popularity and a booming economy. The outlook for Democrats this year may be even bleaker given that Barack Obama isn’t very popular (he's doing better than Richard Nixon in 1974 before he resigned and somewhat better than George W. Bush, but far worse than Reagan, Bill Clinton, and Dwight D. Eisenhower in their second midterms).

That doesn’t mean the Democratic majority is doomed, at least not as definitely as the party's chances of winning the House. The way I read the Sides, Highton and McGhee forecast is that if the midterms are held under current conditions, and if their candidate-quality indicators really work, then Republicans would have “more than an 80 percent chance” of winding up with 51 or more Senate seats.

So two things could be off. Two key predictors are the economy and Obama’s popularity. As I’ve said before, a major improvement in the latter is unlikely (at this stage, sharp downward spikes are more likely than sharp upward spikes), but modest improvements are certainly possible. And economic performance could shift in either direction.

The other variable is that the indicator they’re using for candidate quality might not work well. In House elections, it’s relatively easy to use counts of “quality” candidates (generally, challengers and open-seat candidates with a record of previous electoral success) to get an idea of how well each party has done with recruitment, because with 435 elections, flukes will be evened out.1 That’s less likely to be the case in the Senate, with fewer than one-tenth the number of elections. For example, the model counts candidates coming from the House as strong politicians, which is overall true. But in Georgia this year, Democrats are happy with first-time candidate Michelle Nunn and eager to run against her Republican opponent, Paul Broun, a member of the House. Moreover, in Georgia and several other states, the model can't know who the nominee will be, and individual candidates and their campaigns matter in Senate general elections.

Many of the ways the model could be wrong are symmetrical, with two exceptions: it’s more likely that Obama’s approval rating will be sharply down than sharply up, and it’s more likely that Republicans will pick terrible candidates in their primaries, given the current state of recruitment, than that Democrats will self-sabotage.2 One of the reasons is that Republicans simply have more chances for something to go wrong because they have to win more seats to have a majority. Nonetheless, it should be a good year for Republicans.

1 It's possible that some tiny-town mayor with no clue about how to run a campaign might be coded as a quality candidate, and her opponent, with no previous electoral experience, might be a local celebrity with excellent name recognition, strong party ties and terrific political skills. For the most part, however, there’s plenty of evidence that the measures political scientists use for quality candidates are effective.

2 I’m not predicting a big fall for Obama. I’d guess that his current very slow rise in approval to around 44 percent now will continue and perhaps slightly accelerate. But it’s hard to imagine a sharp spike up.

To contact the writer of this article: Jonathan Bernstein at Jbernstein62@bloomberg.net.

To contact the editor responsible for this article: Max Berley at mberley@bloomberg.net.