As Sam Wang has been predicting, once there was a more robust set of polls, the other polling models would begin to converge with his own, improving the Democrats’ odds of retaining control of the Senate. The reason? More dependence of the actual average of polls and less speculation about “fundamentals.”
…[A]s the election approaches, other sites are decreasing the bias that they add by using fundamentals. This will inevitably make them approach the PEC snapshot, day by day. If everything converges on the PEC Election Day prediction, I would score that as an argument in favor of using polls only – or at least letting readers see the difference added by the use of fundamentals.
Nate Silver speculates that the real reason for the convergence is the Democrats’ money advantage, but that should already be baked in the fundamentals cake. It’s at least as important to Kay Hagan’s reelection prospects that she has more experience than Thom Tillis as it is that she has a lot more money, but both factors should be considered if you’re going to just begin speculating about factors other than the actual polling.
Another problem with speculating is that there’s no way to really measure the superiority of one get-out-the-vote effort versus another one. People assume that the Democrats have a more robust GOTV strategy, but how do you account for it? Not every close contest is getting the same amount of resources.
Here are a few other observations on the polls.
1. Polls out of Georgia disagree about whether Carter and Nunn are ahead or behind, and the differences are explained entirely by different assumptions about turnout. This is also true with recent polls out of New Hampshire of the Shaheen/Brown race.
2. Sen. Pat Roberts is looking something close to doomed. He has a 29% approval rating, only gets 34% in a four-way race that includes Democrat Chad Taylor skimming 6% of the vote even though he has dropped out of the contest. If Taylor succeeds in getting himself taken off the ballot, Greg Orman’s road to victory looks assured, but he’s beating Roberts by 7% as it is. If they knew for certain that Orman would caucus with the Democrats, the election modelers would all be predicting a Democratic hold of the upper chamber.
3. There’s a new poll out of Arkansas that shows Sen. Mark Pryor up by four and Mike Ross (of Blue Dog fame) tied with Asa Hutchinson 44%-44%. Why are the Republicans doing so poorly in Arkansas? What happens when the Clintons descend on the state to make the case for the Democratic candidates?
4. Great news that a new poll is out showing Senator Mark Begich with a five point lead in Alaska. Unfortunately, if history is any guide, a five point lead in the polls in Alaska is not good enough for a Democrat to actually win.
Finally, as things stand, I think Begich, Mark Udall, Kay Hagan, Jeanne Shaheen, and Mark Pryor should all be favored to win. Our candidates in Iowa and Michigan are moving in the right direction and should be narrowly favored. Michelle Nunn and Mary Landrieu look strong enough to force run-off elections at the very worst. Pat Roberts will lose, giving Orman the option to caucus with the Democrats. And Mitch McConnell is holding strong, at least for the moment.
Assuming things stay as they are and election day goes as I expect it to, the Dems will lose three seats (Montana, South Dakota, and West Virgina) putting us at 52 seats. Orman will caucus with the majority, bringing us to 53 seats. And there will be runoffs in Louisiana and Georgia that could lead to either a 51, 53, or 55 seat majority.
If I’m right, and remember things can and will change, the runoffs will not determine the outcome. Even if we lost both runoffs and Orman chose to caucus with the Republicans, causing a 50-50 tie, Joe Biden would cast the tie-breaking vote.
I am worried about Alaska however, as polling there has consistently overestimated the strength of Democratic candidates.
