There has been a growing sense in recent weeks that the odds of Republicans picking up a Senate majority in November are not only growing, may well have tipped over to better than 50-50.
The numbers, geography, and timing for Senate Democrats have been challenging from the beginning of this election cycle. They have greater exposure, defending 21 seats compared with only 15 for the GOP. Even worse, the exposure comes in tough places for Democrats, who have four seats up in states that Mitt Romney carried by 15 percentage points or more, two in states that he won by 14 points, and another in a state Romney took by 2 points.
The timing is particularly bad in that the party’s exposure comes during a midterm election, when the electorate is usually older, whiter, and more conservative than during presidential election years, when turnout is more diverse. Finally, the political environment for Democrats is bad; the party currently has a president with a national job-approval numbers averaging in the low forties, and considerably worse in at least half the Senate battleground states. Plus, the Affordable Care Act, his signature legislative accomplishment, is distinctly unpopular.
All in all, it’s not a good situation for Democrats.
Republicans have helped themselves with a strong recruiting year. The GOP expanded the playing field in recent weeks with former Sen. Scott Brown’s decision to challenge incumbent Democrat Jeanne Shaheen in New Hampshire. The party has also traded up candidates in Colorado, replacing problematic 2010 Senate nominee Ken Buck for Rep. Cory Gardner.
If you had to bet today on the outcome, the odds would strongly favor Republicans getting halfway to their goal of a net gain of six seats in Democratic open seats: GOP candidates are favorites in South Dakota, West Virginia, and, to a slightly lesser extent, Montana. Four Democratic incumbents are embroiled in very tough races: Mark Begich in Alaska, Mark Pryor in Arkansas, Mary Landrieu in Louisiana, and Kay Hagan in North Carolina. All are running roughly even, slightly ahead, or even behind their GOP rivals. The races, in our view, are absolutely in the Toss-Up column.
Conventional wisdom has labeled Pryor as the walking dead, even though multiple private Democratic polls (by different pollsters) have never showed him down as much as a single point. The one high-quality public poll where all the details are available — conducted by the Democratic polling firm of Hickman Analytics for an energy-industry group — had Pryor ahead of Rep. Tom Cotton by 3 points among all likely voters, and 2 points behind among definite voters; both are margin-of-error variances. This is an example how the perception of a race often can be driven by sketchy polling.
After those four Democratic Senate incumbents (in Alaska, Arkansas, Louisiana, and North Carolina), we still have an open seat in Michigan, where two little-known candidates are battling in a very close race. Yes, the Iowa open seat is worth watching, specifically because the odds of the convoluted GOP nominating process picking an exotic and potentially problematic candidate for the general election are good. Democrats dispute our Toss-Up designation of the race in Michigan, but current polling suggests that is indeed where things stand. In the two “new races,” Colorado and New Hampshire, one or both could end up in the Toss-Up category, though not enough numbers have been released to justify that in the former, and numbers in the latter currently show Shaheen with a lead well beyond the margin.
Then there is the matter of the two vulnerable GOP seats. The conventional wisdom in Kentucky continues to discount the magnitude of Senate Minority Leader Mitch McConnell’s peril. His poor favorability ratings in the state should disabuse anyone of that notion, but apparently they haven’t. The perception of his tenacity is given greater credence than that the data indicate.
My good friend and competitor Stu Rothenberg puts the broad range of potential outcomes at four to eight seats gained by the GOP, numbers that make sense to me. Narrowing it down a bit to a five-to-seven-seat gain, while riskier, is probably an equally logical conclusion. Nate Silver’s terrific website FiveThirtyEight puts the broad range of GOP victory from plus one for Republicans to plus 11, with a net gain of six seats the most likely. While I can quibble with some of the odds that Nate puts on individual races, just as Stu and I disagree here and there, we are all in the same ballpark. The disagreements with FiveThirtyEight are in some cases the difference between looking at things purely quantitatively, as Nate does, or a bit more subjectively as Rothenberg and I do. Larry Sabato’s Crystal Ball is a little less explicit in its weighting of qualitative versus quantitative analysis, but overall looks to be in about the same ballpark as well.
Some people ask if there is room for a Charlie, Stu, or Larry in a world with Nate’s quantitative approach. It is a legitimate question, and I confess to being a big fan of Silver’s, even if we sometimes disagree on the details. But, as the terrific book and movie Moneyball suggests, while there is not a Major League Baseball team that does not employ statisticians using sabermetrics, neither is there one that has fired all of its scouts. Smart teams employ both.