RT @fivethirtyeight: When you cut through the clutter, the bottom line is that tonight should make Dems more optimistic about November.
RT @fivethirtyeight: When you cut through the clutter, the bottom line is that tonight should make Dems more optimistic about November.
Someone is going to have to explain 538’s logic to me. This was yet another anti-incumbent day through-and-through, and last time I checked, the Dems have a lot more incumbents than the Reeps. Blanche Lincoln needed to hit 50% for last night to mean anything, and she didn’t, so now she has a runoff where she may very well lose, and even if she survives, there’s the election in November. The Democrats in Connecticut went from having a 3% chance of losing Dodd’s seat to having a 30% chance, thanks to the pick of a George O’Leary Democrat as their favored nominee. The loss of Specter in Pennsylvania virtually hands that seat to Toomey. The lone bright spot is that the Dems held on to Murtha’s seat, but Murtha’s seat is about as gerrymandered as they come to cobble together sufficient Blue Dog votes to keep a Democrat competitive, and turnout was quite low. And what you got in Critz was a Blue Dog Democrat who is pro-life, pro-gun, and ran against Obamacare. Winning that seat would’ve been nice for the GOP, but at best the result is neutral for Obama and the liberal agenda.
Briefly:
• Based upon the polling and whatever other nifty analytical tools he uses, Nate at 538 regards Halter as a slightly better general-election candidate than Lincoln. It’s a virtual tossup, but in any event, the prospect of Halter winning the nomination doesn’t scare him. (Though let’s be honest: the Dems are going to lose Arkansas. It’s almost as much of a foregone conclusion as North Dakota. So really, whatever you think of Lincoln vs. Halter, who cares? If the Dems manage to keep that seat with either candidate, it means the election is going FAR better nationally than expected. If we assume the November election is close enough that majorities are in play, then last night’s Arkansas result is simply a wash. Doesn’t matter. Doesn’t affect the math.)
• Likewise, Sestak is at least as good a candidate as Specter, and probably a better one. Indeed, 538 had a post the other day with a compelling critique asking why on earth the White House went to bat for Specter. In any case, how can you possibly square your “anti-incumbent” talk — with which I agree, BTW — with the notion that “the loss of Specter in Pennsylvania virtually hands that seat to Toomey”? Last I checked, Specter is an incumbent; Sestak is not. Toomey’s campaign narrative just got a lot more complicated, because he is no longer the only “change” candidate in the race; either candidate represents change. Now, obviously, Toomey is still running against the incumbent party, and I’m not downplaying that. Toomey is still the favorite in this environment. But he was going to be the favorite either way, and all things being equal, if I’m the Dems, I’d rather have a “fresh face” than a tired old opportunist hack as my candidate in an anti-incumbent year. Like Nate, I’m not sure why Obama felt differently. But in any event, I don’t see how Specter losing is some sort of catastrophic blow to the Dems’ November chances. At the very worst, it is, like Arkansas, a wash.
• The Dems got the best possible matchup in Kentucky, in terms of general-election chances. That doesn’t mean they’re likely to win, but they’ve got a better outside shot than in any other matchup, at least according to the way Silver looks at the polls and crunches the numbers.
• Most importantly, they held onto Murtha’s seat, which, if they’d lost it, would have led to a cascading avalanche of CW that “OMG THEY’RE GOING TO LOSE THE HOUSE!!!1!” They still might well lose the House — and they may well lose that seat in November en route to doing so — but for now, they’ve dodged a bullet. And, CW or no CW, winning a hotly contested House seat is obviously unambiguously good news, with the only debate being over how good.
Really, the only bad news from last night is that the most electable GOP candidate won the primary in Arkansas. But again: the Dems are going to lose Arkansas regardless, barring a huge change in the national environment that would mean they solidly keep their majorities. So winning a House seat, getting a non-incumbent candidate in Pennsylvania, and getting a favorable matchup in Kentucky, far outweigh any damage in Arkansas.
P.S. Here are the numbers, per 538’s model:
ARKANSAS
Lincoln: 4% chance vs. Boozman (polling average -15.9)
Halter: 8% chance vs. Boozman (polling average -13.0)
(Their chances ranged from 9%-21%, in Lincoln’s case, and 25%-45%, in Halter’s case, against the other GOP candidates. So perhaps Boozman’s victory is a bigger blow than I thought. But the odds of anyone other than Boozman winning the nomination always seemed low — roughly 20%, by Silver’s estimation — so it’s also an expected blow. It would have been a shock if anything else had happened.)
PENNSYLVANIA
Specter: 29% chance vs. Toomey (polling average -5.0)
Sestak: 29% chance vs. Toomey (polling average -5.1)
(So this was a virtual tossup, per the model. I had mis-remembered that. But I still stand by my belief that a non-incumbent candidate is better.)
KENTUCKY
Conway: 23% chance vs Paul (polling average -6.9)
Conway: 11% chance vs. Grayson (polling average -11.0)
Mongiardo: 17% chance vs. Paul (polling average -8.8)
Mongiardo: 10% chance vs. Grayson (polling average -11.4)
(With Rand Paul as the GOP nominee, the Dems have a substantially better chance of picking up Kentucky than of holding Arkansas. Would I bet on it happening? No. But they’ve got a fighting chance now.)
The strength of 538 is that they are looking at gobs of polling data. The weakness of 538 is that they’re looking at gobs of polling data.
Lincoln would do better in the general election because she’s more moderate and could raise more money. I suspect that Halter polls better than Lincoln largely on the fact that he’s not the incumbent and he’s less well known (i.e. the voters prefer the devil they don’t know), but as we get closer to November, Arkansas Dems are going to sorely miss having the money advantage as Boozeman successfully drives up Halter’s negatives.
Sestak vs. Specter may very well have been a wash polling wise, but as with Arkansas, Sestak’s numbers mask the fact that he is much less well known than Specter. Specter’s advantage is that he’s considered a moderate, he can raise money, many people in Pennsylvania have pulled the lever for him in the past, and he’s already successfully beaten Toomey before. In a wave of “throw the bums out”, Sestak is a great example of “meet the new boss — same as the old boss.” Sestak is more reliably liberal and he will not attract moderates very successfully.
With the Kentucky seat, I truly and honestly believe the numbers don’t matter at this point. Those numbers are reflective of a primary season with multiple relatively unknown candidates. Yesterday’s winners now have five months to pivot and shape their appeal. The surface horse-race numbers aren’t what is important at this stage — what matters are the internals (e.g. strongly like / strongly dislike, strengths and weaknesses on issues, etc.), and 538 doesn’t delve much into that data. But coming from a campaign manager perspective, those are the only numbers that I care to see because that is what will shape my message, my attacks on my opponent, and my media campaign.
Regarding Murtha’s seat, again, a victory there for the Republicans would’ve been great from a CW perspective, but it really portends nothing about November because it’s a highly unique seat and Critz is a very blue-dog Democrat. You can extrapolate from a Republican victory there, sure, but you can’t really extrapolate anything from a Blud Dog victory there one way or the other, so there’s simply no reason to be optimistic from a Dem perspective other than “the worst didn’t happen”.
Bottom line: Yesterday didn’t change anything, and it didn’t portend anything, it was a blip on the radar screen, and partisans of both sides are foolish to try to read much into the results.