The bad news is that on Wednesday, Fox News laid down quite the smackdown, with a salvo of five polls, none of which had any good news for Democrats (with 4-point GOP leads in the Senate races in Alaska and Kentucky, a 6-point GOP lead in Colorado, a 7-point GOP lead in Kentucky, and even a 5-point GOP lead in Kansas, where Greg Orman hadn't trailed a poll since the Chad Taylor dropout). The good news, though, is that these Fox News polls barely changed the overall trend in the Election Outlook model, dropping Democratic odds of retaining
control of the Senate down from 33 percent on Monday to 31 percent now.
The bad news behind that good news, though, is that the Fox News polling wasn't that much worse than the averages in some of those races, so there wasn't much further damage they could do; in Alaska, for instance, Mark Begich's odds of winning stayed flat at 21 percent, and Mark Pryor's odds in Arkansas fell only from 21 percent to 20 percent. Colorado was the only big mover; Mark Udall had moved back above water at 51 percent after leading by 3 in this weekend's YouGov poll, but fell back to 35 percent with the addition of the Fox News poll.
You might think the state that would have the biggest impact from the Fox polls was Kansas, since Greg Orman has led consistently for the last month, and now all of a sudden two polls had him down (a CNN poll on Wednesday also found him trailing, though only by 1). However, a SurveyUSA poll released on Tuesday found Orman leading by 5, consistent with previous averages, so the cumulative effect wasn't great. And more importantly, while Orman's odds of winning the race fell significantly (from a 90 percent chance of winning on Monday, to a 67 percent chance now), our model is more interested in the question of whether Orman caucuses with the Democrats.
Since we're taking him at his word that he'll caucus with whichever party has a majority, Orman caucuses with the Dems in only a minority of all our simulations, since the Democrats win a Senate majority without him only about one-third of the time. So the odds of a "Democratic" victory in Kansas fell only from 33 percent to 29 percent, mostly because there are now more simulations where Pat Roberts wins (though that's still only in a minority of all cases).
And Democratic odds in Kentucky actually improved significantly in the last half week; that Fox News poll was basically in line with previous polling, but the difference-maker in Kentucky was the SurveyUSA poll released on Monday giving Allison Lundergan Grimes a 2-point lead over Mitch McConnell. Whether that polls turns out to be an outlier or the beginning of a new trend, it singlehandedly boosted Grimes's odds in Kentucky from 19 percent to 27 percent. That means she now has better odds than the any of the three red state incumbents that Democrats are trying to defend (Begich, Pryor, and Mary Landrieu in Louisiana).
We'll discuss gubernatorial races over the fold:
Smoothed trendlines for SD-Sen
Before getting to the gubernatorial races, there's one other Senate race you might be wondering about. Probably the most attention-grabbing poll of the week wasn't the Kentucky poll, but the SurveyUSA poll of the South Dakota race on Tuesday that found Republican frontrunner Mike Rounds barely clinging to a lead. The poll put Rounds at 35, independent candidate (and former Republican senator) Larry Pressler at 32, and Democratic candidate Rick Weiland at 28. The disparity between Rounds and Weiland is still enough that the Democratic odds in South Dakota didn't move much, from 1 percent to 2 percent (although you could spin it by saying Weiland's odds doubled).
That's mostly because previous polls, where Rounds had a significantly bigger lead, still shape much of the trendline. However, the trend, as you can see above, is unmistakable, with Rounds heading down and Pressler heading up. If the trend continues, we may need to start separately calculating odds for a Pressler victory — and unlike Orman, Pressler hasn't even laid out any ground rules suggesting how he might caucus, in fact going so far as to say that Senate rules don't require him to caucus with anybody (which is true, so long as he doesn't mind not serving on any committees!).
From there, we could really start going down the rabbit hole, in ways that a model simply can't predict or even comprehend. If Orman and Pressler both win and ally with each other (and potentially pull in Angus King, who puts the "I" and "me" in (I-ME)), then there's an actual critical mass of flaky centrist prima donnas who can call the shots, holding out and throwing their lot in with whichever party makes them a better offer. In other words, the U.S. Senate could suddenly start looking a lot more like the New York and Washington state Senates!
Finally, let's turn to the gubernatorial races, where the Democrats' hopes seem a lot better than they do in the Senate. Although they're still on track for a median of 22 seats (a gain of one over where they are now), and the overall odds of gaining seats have stayed flat (going from 58 percent on Monday to 56 percent now), they're made some significant gains in important blue-state seats.
For instance, things keep improving in Illinois, where Pat Quinn, looking dead all cycle, seems to have finally Romnified his wealthy Republican opponent Bruce Rauner. A We Ask America poll from Tuesday put Quinn up by 4 (effectively worth 7, since they're a Republican pollster), and that boosted his odds from 49 percent to an all-time high 72 percent. Similarly, Dan Malloy, the incumbent in Connecticut also facing a vulture capitalist opponent in Tom Foley, seems to be peaking at the right time. Malloy led by 8 in a PPP poll, his first lead of that size all cycle, and also fought to a tie with Foley in a new Quinnipiac poll after trailing by 6 in their previous poll. Those two polls combined boosted Malloy's odds from 26 percent to 60 percent.
The main place those gains are offset is in Alaska, where the independent Bill Walker (whom we're treating as a de facto Dem, since his Dem opponent dropped out and became his Lt. Gov. ticket-mate) also trailed by 5 in the Fox News poll of Alaska, the first poll in several months where he didn't lead. That pushed Walker's odds down from 64 percent to 39 percent.
Also offsetting the blue-state gains was a decline for Democratic challenger Paul Davis in Kansas, also thanks to the gubernatorial portion of the Fox News poll there (where he trailed by 6, the first poll in nearly a month where he wasn't winning); Davis's odds fell from 83 percent to 62 percent. Martha Coakley in Massachusetts also continues to, well, pull another Coakley; her odds fell from 63 percent to 54 percent, thanks to a 5-point deficit in a poll from Emerson College.