In the last two days, the revelation that Donald Trump had contracted COVID-19, and has been hospitalized as a result, has roiled the presidential race, but at this point, there seems to be no reason to expect that it will make a material difference. It is clear that the trajectory of the race has shifted back to Joe Biden. Readers may recall that in my last update, 30 days ago, I struck an ominous tone about the trajectory of the presidential race for Biden. Democrats will find my new update more heartening.

It is clear, from an examination of the polling trends in the key states, that Biden’s course has stabilized and has actually moved back in his direction. He maintains a lead in most of the key states and is, in fact, in better condition in some unexpected states than I expected he would be at this time.

A trend is emerging in which Biden’s position, relative to the position of Hillary Clinton four years ago, is improving among white voters in Midwestern states, but not improving significantly among their counterparts in southern states. It seems clear that white voters in the Midwestern and northeastern battlegrounds are much more “elastic” than white voters in southern states, who remain much more difficult to budge out of the Republican column.

The trends now clearly point to the first major shakeup in my view of the state of the race, as I am cautiously shifting Ohio to “Tilts Democratic.” Polling data clearly points to a race in the Buckeye State that has shifted in Biden’s direction, with Biden leading 10 of the last 12 polls since July. That said, his leads are very small—but consistent.

I am also moving Iowa at this time to “Tilts Democratic,” even though Trump’s polling position has been better in Iowa than in Ohio. It appears the trend is moving against him, with Biden recently moving into a small polling lead, and Biden’s improving position among Midwestern white voters and older voters augurs well for him in Iowa.

It is also clear that Michigan and Wisconsin can be moved back to “Leans Democratic,” although Michigan polling still appears softer for Biden than I expected it would. That bears watching. I also am moving the race in Nebraska’s District 2 back to “Leans Democratic,” and Biden appears to be in a very strong position to win that single electoral vote in Nebraska.

I remain convinced, as I have for months, that Pennsylvania is a harder lift for Biden than either Michigan or Wisconsin, and polling data is now bearing that out. I am keeping Pennsylvania at “Tilts Democratic,” as it is clear that Trump’s team is putting a lot of marbles in Pennsylvania. However, some recent polling in Pennsylvania has been very favorable for Biden, and the state appears likelier to move toward “Leans Democratic” than “Tilts Republican.”

I am also keeping Florida at “Tilts Republican,” despite a narrow polling lead for Biden. All signs I am seeing point to Democrats completely misplaying their hand with the state’s non-Cuban Latino communities. I have been in contact recently with a well-placed source in south Florida Democratic politics who indicates to me that the Democratic failure to engage these communities is even worse than I imagined. Make no mistake: with Biden’s improved position among white voters and seniors, he should be well ahead in Florida right now, by at least two or three points on average. The fact that he isn’t should be concerning to Florida Democrats.

I am also keeping North Carolina at “Tilts Republican,” because I find it hard to imagine that North Carolina will flip blue while Florida stays red.

But I am moving South Carolina to “Leans Republican,” based on a continued trend toward stronger-than-expected showings by Biden in that state’s polling. That, coupled with a very strong race by Democratic Senate candidate Jaime Harrison, convinces me that something unexpected is going on in the Palmetto State, and it bears watching.

At this point, it is clear that Biden’s position in Minnesota and Maine is strong, and I am moving Minnesota back to “Likely Democratic,” and three of Maine’s four electoral votes to “Safe Democratic.” I am also shifting the electoral vote in Maine’s Congressional District 2 to “Tilts Democratic.” Polling in Maine is now very strong for Biden, and it is hard to imagine him winning the entire state by double digits while losing the 2nd District. The same trends that are benefitting Biden in Ohio and Iowa—improved performance among white voters and older voters—will also benefit him in northern Maine.

The ratings changes now show states that fall into the safe, likely or leaning Democratic categories have a total 270 electoral votes, exactly the number required for Biden to win. If all the states listed as “Tilts Democratic” also go to Biden, he would win the electoral vote by a count of 315-223.

U.S. Senate 

One of the dumbest things a person can do when his theories are cast into doubt is to double down and ignore all evidence to the contrary.

It has long been my contention that it is unlikely that any given state will vote for one party’s presidential candidate and another party’s candidate for Senate this fall. I still consider it unlikely. However, weeks of consistent polling data have made it clear that Democratic Senate candidates in IowaKansasMontana, North Carolina and South Carolina are clearly running ahead of Joe Biden in those states. While I still expect Republicans will win most or all of those races, it would be stupid to rule out the possibility of a split result in any of them.

In the last 20 years, the results of Senate and presidential races, when both are contested in the same year, have shown almost an 85 percent correlation. The correlation is even more stark when an incumbent senator is on the ballot and a presidential candidate of his or her party wins the state. There has been only one exception since 2000, and that exception occurred with the incumbent senator (Ted Stevens, R-Alaska) under indictment.

This is why I have been hesitant to give much of a chance to Democrats Steve Bullock of Montana or Jaime Harrison of South Carolina. Both are running against incumbent Republican senators in states that Donald Trump is likely to win. In the end, unless Trump faces surprisingly close races in those states, I still expect both Democrats will lose. But it would be foolish to deny that they are competitive and may have a realistic chance to win.

I have also been hesitant to give much of a chance to Democrat Barbara Bollier in Kansas. Although Bollier is not running against an incumbent, she is running in one of the most safely Republican states in America—one which hasn’t elected a Democrat to the United States Senate since 1932, and one which has not voted Democratic in a statewide federal race since 1964. In the end, I also expect she will likely lose, but again, it would be foolish to deny that she is close enough to have a puncher’s chance at an upset. That said, I’ve been fooled by Kansas before. In 2014, I bought into the polling that showed independent Greg Orman beating incumbent Republican Senator Pat Roberts. In the end, Roberts won comfortably, which has rendered me exceedingly cautious about overstating Democratic prospects in Kansas.

I am moving all three of these races to “Leans Republican,” but at this time, I still expect Republicans to win them all.

I still remain convinced that Democrats, in the end, will probably lose both Senate races in Georgia, with both likely headed for runoffs in December. Given a historical pattern of Democratic turnout drop-off for post-election runoffs, the dynamics favor the Republicans in both races.

Right now, the likeliest state for a split result split appears to be North Carolina, where Democratic Senate candidate Cal Cunningham is clearly running ahead of Biden in his race against Republican Senator Thom Tillis. That said, unlike in Iowa, Biden also maintains a razor-thin polling lead in North Carolina, so it may happen that both Democrats prevail in the Tar Heel State. That said, the recent revelations of romantically charged texts between Cunningham, who is married, and a married California political consultant, may change the trajectory of this race. At this time, out of an abundance of caution, I am moving this race to “Tilts Democratic.”

The next likeliest state for a split result appears to be Iowa, where Democratic Senate candidate Theresa Greenfield has established a small, but consistent, lead over Republican Senator Joni Ernst, but Biden is in a very tight race with Trump. I am moving this race to “Leans Democratic,” but I also have Iowa now going narrowly to Biden as well, so there may not be a split result here after all.

A split is also possible in Alaska, where independent candidate Al Gross won the Democratic primary and is polling very strongly against incumbent Republican Senator Dan Sullivan. Biden is also polling surprisingly well in Alaska; as in South Carolina, there may be something interesting happening there worth watching as well. I am now moving this race to “Tilts Republican.” Keep an eye on this one.

In addition to characterizing Iowa and North Carolina as “Tilts Democratic,” I continue to rate Colorado as “Safe Democratic,” while setting Senate races in Arizona and Maine as “Leans Democratic.” If Democrats prevail in these five races, while losing their seat in Alabama, as expected, they will finish with a 51-49 majority in the Senate.

U.S. House

I will not get into an exhaustive race-by-race breakdown here. There are always some House races on both sides that go an unexpected direction. Most people were taken completely off guard by Kendra Horn winning a seat in Oklahoma in 2018. That seat was on my list of potential surprises, given the fact that it was the kind of urban/suburban district that had been trending blue since Trump won in 2016. It is the same kind of seat, in locales across the country, where I expect Democrats will add to their House majority in 2020.

After 2018, my assessment of the results was that there were about 20 Republican-held seats in suburban districts that Democrats narrowly missed picking up in their 40-seat wave. Two of those seats have gone back to the Republicans as the result of a special election loss (California 25) and a party switch (New Jersey 2), but I expect them both to flip again, back to the Democrats, in 30 days.

This year, I think Democrats have a top-out number of +21. I don’t think they will top out; it is probably more realistic to expect that they will net about half of what they left on the table in 2016, roughly 10 seats. Using those two numbers as a range, my best estimation at this time is that Democrats will make a net gain of 16 U.S. House seats, which would give them a 249-186 advantage. This figure would approach the crushing 257-178 majority Democrats held in the House after the 2008 elections. However, with no filibuster available in the House, as there is in the Senate, and a continued trend toward partisan polarization, the margin in the House does not matter nearly as much as it once did.

However, these expected Democratic gains would not change the balance of power in state delegations, where Republicans would continue to control the delegations of at least 26 states. In the event that disputed results in one or more states send the presidential election to the House of Representatives, Trump would still likely have the edge, despite Democrats controlling more seats, because a House vote for president provides that each state, not each representative, would have one vote.

Final Update and Election Night Guide

I will issue a final update on the election on the morning of Monday, November 2, and I plan to publish a comprehensive election night guide on my website at that time. The quadrennial Cliston Brown Election Night Guide will provide a rundown on the presidential race and key House and Senate races in every state, listed by poll closing times, so that viewers can have a sense of what to be watching for as the returns come in.