I cast my ballot several days ago, although not immediately after my most recent post as I had suggested. On that Friday afternoon I had only a 30-minute window in which I would have time to vote, but after 30 minutes in line I was still 10-15 minutes from being through, so I had to leave without voting. I came back on Monday morning, and that time it only took about 20-25 minutes to vote.

Early voting turnout has indeed been robust, particularly in states like Georgia and North Carolina that had not embraced alternative voting mechanisms in 2020 as much as many other states had. It seems unwise to attempt to draw any conclusions from this year’s early voting patterns, though, as voter behavior appears to be changing more rapidly than we have data points from which to extrapolate.

A week ago tonight Trump had his rally at Madison Square Garden, which will be best remembered for one of his warm-up acts having told a (supposed) joke in which he referred to Puerto Rico as a “floating island of garbage.” A NYTimes headline referred to the rally as “a closing carnival of grievances, misogyny and racism.” Two days later, Harris gave her last major speech of the campaign before a reported crowd of 75,000 at the Ellipse in D.C. More recently, last night she made a surprise appearance in the cold open of Saturday Night Live, appearing with former cast member Maya Rudolph, who has been providing the show’s Harris impression this fall.

None of these last-minute activities appear to have changed the underlying dynamics of the race, which remains tight tight tight. Possibly too tight; Nate Silver and other armchair observers have recently noted that the spread of reported polls in swing states is actually far too narrow than what one would expect to see from polling a race that was a dead heat, raising the possibility that pollsters are, consciously or unconsciously, massaging poll results so that they do not deviate much from the conventional wisdom.

Not every poll of late has been boring, mind you. The Des Moines Register’s polling operation has one of the best reputations in the business, and in particular did fairly well this spring with its final poll of the four-way race in the Republican Iowa caucus. Yesterday they published their final presidential poll of Iowa, with the startling result that Harris was ahead, 47-44. For what it’s worth, the same outfit’s final poll 4 years ago was Trump 48-41, and the final result in Iowa was Trump 53-45. Back in August Iowa was on the list of states that I categorized as “expected Republican” (meaning an 80% to 95% chance of Trump victory in Silver’s model), the same category as Ohio and Texas – a set of three former and perhaps future swing states that, in the politics of the early 2020s, look staunchly red. As such Iowa has been getting next to no attention in this campaign, although I imagine ads from the Blue Dot have been filtering into the western counties.

But if a first-rate pollster is showing Harris with a lead in Iowa… well, just maybe we’re in a for a very different election night than we thought we were. Of course, never forget that in late October 2020 there was a poll from ABC and the Washington Post showing Biden ahead 57-40 in Wisconsin, a state he would ultimately win by only 0.63%.

The Senate race continues to look more rosy for the Republicans than the Democrats. The Democrats have a tough parlay to get to 50 Senate seats. First, there are three swing-state incumbents that need to win re-election: Rosen in Nevada, Casey in Pennsylvania, and Baldwin in Wisconsin. Next, the Democrats have to defend two open swing-state seats, namely Simena’s old seat in Arizona and Stabenow’s seat in Michigan, as well as an open blue-state seat in which the Republicans have run a compelling candidate, namely Cardin’s seat in Maryland. If all that happens, the Democrats should be at 48, so two of the following five things would also need to happen: Brown holds Ohio; Tester holds Montana; Allred upsets Cruz in Texas; Mucarsel-Powell upsets Scott in Florida; and independent candidate Osborn prevails over Fischer in Nebraska. (Moreover, it is not clear that Osborn would caucus with the Democrats if he were to win.) All of this seems like a parlay that certainly could happen on a very good night for the Democrats, but feels less likely than not to actually happen. Senate Republicans are set to vote on November 13th between Cornyn, Thune, and (assuming he wins) Florida’s Scott as the replacement for McConnell as Senate GOP head and, perhaps, the new Majority Leader.

And then there’s the House, which is too chaotic to really model with confidence. There will be no shortage of interesting seats to watch on election night and beyond. In addition to warily watching my own seat (Craig’s MN-02), I will be very interested in the neighboring seat of WI-03, where an appealing young woman Democrat named Cooke is challenging the incumbent freshman Republican, Van Orden. My instinct is that Hakeem Jeffries is a slight favorite to become Speaker, but we shall see.