Kavanaugh effect? Indiana poll shows Republican Mike Braun leading Joe Donnelly by four

Good news, but four points happen to be the margin of error of this poll. That means the race is effectively tied, which would be in line with other recent polls showing the incumbent, Donnelly, with a lead of 2-4 points or less.

As for the Kavanaugh effect, eh. There *is* evidence in this particular case to support it but there’s a bunch of evidence cutting the other way right now nationally.

The poll, conducted by Mason Strategies with a margin of 3.9 percent, showed Braun leading Donnelly 47-43, Libertarian Lucy Brenton had three percent of the vote and seven percent were undecided…

The poll also indicated that Donnelly’s vote against Brett Kavanaugh to the Supreme Court might be having some impact, albeit not as much as Republicans might have hoped. Thirty percent of likely voters said they more likely to vote for Donnelly because of his vote against Kavanaugh while only 35 percent said they were less likely. And 32 percent of likely voters said Donnelly’s vote did not matter.

A 30/35 split on the Kavanaugh question isn’t dramatic, although the pollster notes elsewhere that even moderate Republicans and Democrats tended to react along partisan lines when asked about Donnelly’s no vote on Kavanaugh. Liberal Republicans disapproved of it, conservative Democrats supported it. The result is that each party may be more unified in its choice of Senate candidate than it was before — which is good news for Republicans in a red state like Indiana. On the other hand, the share of “very enthusiastic” voters who say they’re much more likely to vote for Donnelly now exceeds the “very enthusiastic” share who are much less likely to vote for him by eight points. The “Kavanaugh effect” in polls is nice but if it doesn’t translate into actual net votes then it’s worth zippo.

As for that national evidence cutting in the opposite direction, new from Montana:

In the state’s high-profile U.S. Senate race, Democratic incumbent Jon Tester holds a narrow lead over Republican challenger Matt Rosendale, 46 percent to 43 percent, according to a Montana Television Network-Montana State University poll.

The mail-ballot poll of more than 2,000 registered voters, conducted during a three-week period in late September and early October, has an error margin of plus-or-minus 2 percent.

Given that error margin and the poll results, the race between Tester and Rosendale is statistically a toss-up going into these final weeks of the campaign.

A toss-up isn’t the worst-case scenario against a twice-elected Democratic incumbent but remember that Kevin Cramer is stomping Heidi Heitkamp across the border in North Dakota right now. Not only that, but the last two polls of Montana taken before this one came in mid-September, shortly before the Ford/Kavanaugh hearing. In those two Tester led by two and four points, respectively, averaging a three-point lead — precisely the number he hit in this latest poll taken a month later. If all of the data is accurate, there’s no “Kavanaugh effect” in Montana. Unless you believe Tester was poised to pull away and win easily over Rosendale and the Kavanaugh fiasco re-tightened the race, which seems hard to believe.

Meanwhile, the NYT is talking to GOP and Democratic consultants and relaying the gossip:

There is one state on which both sides agree: Tennessee, despite some polls that showed Republicans enjoying a post-Kavanaugh bounce, remains a real race for the open Senate seat.

Tennessee? Marsha Blackburn led there by eight and 14(!) points in two post-Kavanaugh polls. It looked to all the world like she was pulling away a la Cramer in North Dakota. Then, last week, a poll arrived showing Democrat Phil Bredesen up by a single point. Was that poll the outlier or were the blowout polls for Blackburn outliers? (Or could the race really be so volatile that Blackburn bounced out to an actual 14-point lead, only to lose it within a few days)? If the Times’s anecdotal evidence about Tennessee is accurate than the Kavanaugh effect is overhyped there as well. And as I noted in last night’s post, Democrat Bill Nelson’s lead in Florida has actually *increased* over Rick Scott lately despite the Kavanaugh backlash on the right.

None of this is to say that the GOP’s in danger of losing the Senate. With Cramer on his way to victory and Cruz looking solid against Beto! the Democrats’ odds of a majority hang by a thread. They’d need two out of three in Tennessee, Nevada, and Arizona plus victory in all of the toss-up races. That means Nelson in Florida, Donnelly in Indiana, Tester in Montana, and McCaskill in Missouri. And things aren’t looking great in Missouri for McCaskill, if you believe the experts. Although the public polling shows her effectively tied with Josh Hawley, sources told the NYT yesterday that Republicans think it’s “virtually certain” they’ll pick up Missouri. Why they think that, I have no idea. Imagine, though, if Bredesen really did pull off a shocker in Tennessee — only to have Republicans retain a Senate majority because Claire McCaskill got blown out in Mizzou. Strange times.