We now have every indication that Donald Trump's campaign is in something like a free fall. But it's a free fall down to his core of support which seems to be in the very high 30s. The current margin between the two candidates is roughly comparable to what it was between Barack Obama and John McCain in the final week of the 2008 election. Remember, 2012 was within a single percentage point on election day, though Obama won by about 3.5 percentage points.
With that kind of margin, the Democrats should be able to easily take control of the Senate. But something odd happened over the three week period over which Trump cratered: Senate Republicans managed a mild resurgence. The big stand out is Ron Johnson in Wisconsin. Johnson has been the most likely Republican incumbent to lose his seat for months. Yet over the period in which Trump cratered he's come from way back to pull almost even with Russ Feingold. I have yet to see any good explanation of why that is. Nothing in the race itself jumps out and it comes over a period in which Clinton was consolidating her lead in Wisconsin.
There's nothing as dramatic in other states. But Kelly Ayotte managed to reassert a very small lead in New Hampshire and Pat Toomey has held on in a virtual tie even as Clinton has consolidated a solid lead. Across the country we have a total of four senate seats in the toss-up category - which means within 2 points in the PollTracker Average: Missouri, Wisconsin, Pennsylvania and New Hampshire. Meanwhile you have anther very close race in North Carolina where Richard Burr has a 2.5 percentage point lead.
I don't think we know why GOP senate candidates haven't been hurt yet by Trump's steep drop in support. There are a number of potential explanations. But there are too many different possible factors to state with any confidence which is most likely. It could simply be that events have been moving so quickly with Trump that Democrats haven't had enough time yet to hammer Toomey, Ayotte, Johnson and the rest with him. Perhaps it simply hasn't sunk in yet. Or maybe voters are deliberating shifting toward supporting a Republican senate as a Clinton presidency seems more certain.
In any case, assuming that the polls stay more or less where they are now, there's one pretty big advantage Democrats likely have in their favor. In three of the four races currently in the toss-up category, Clinton has mid-to-high single digit leads. That creates a very steep downdraft for an incumbent senator who's in a near tie race. That might even end up being a factor in North Carolina where Richard Burr currently has a 2.5 percentage point lead but where Clinton appears to be edging into a lead at 3.2 percentage points.
History may not be the best guide this year. But ticket splitting tends to create cognitive dissonance for most people. Plenty of people do it. But if a senate race is tied and there's an 8 or 9 point spread for the presidential race, the candidate of the opposite party to the presidential leader is going to have a very hard go of it. Lots of people will split their vote. But people who are undecided or only lightly committed will tend to break for the presidential winner.
There may be an additional factor as well. Presidential campaigns, national parties and individual candidates each have overlapping ground operations. But a big, big part of that mix is driven by the presidential campaign. We're accustomed to presidential races where the campaigns have at least broad parity. On any given Sunday the worst team in the NFL might beat the best. They're broadly comparable. But the Trump campaign's field operation might be more like a pro football team squaring off against a high school squad or no team at all. We just don't have any track record for a competition that mismatched. What's more, most of the key senate races are in battleground states - the places where the Clinton campaign will have invested the most resources. It's not implausible that this difference could count for a point or two or conceivably even more in the final result.
Enough states are in the undecided that a number of very different scenarios are possible. But at least 50 senates for the Democrats (which would constitute a majority with a Democratic vice president) seems pretty likely and one or two more senate are possible. The big question is whether over the next three weeks we start to see signs in the polls of a Democratic wave which solidifies Democratic advantages in New Hampshire, Pennsylvania, Wisconsin and perhaps North Carolina.