I. The Horse Race.
Last week I wrote one day early because of the first debate. It occurred to me that I will do my last forecast on election eve, so I will stick with Mondays for these posts over the last four weeks of this blessed election season.
We have received enough post-debate polling to confirm what we all saw with our own eyes-Joe Biden won the debate. More importantly and decisively, Donald Trump lost the debate, spectacularly. He needed a win and he couldn't even muster a tie. Joe Biden leads the RCP average by 8.3 points, up from 6.8 points one week ago. Biden leads the 538.com average by 8.2 points, up 1.3 points from one week ago.
State-level polling has also moved in Biden's direction. Biden now has stable leads in enough states to win the presidency. He now leads every swing state that Real Clear Politics tracks closely. He leads in each of the critical states of MI/PA/WI by five or six points. He leads in Florida by two and Arizona by three. Polling has also been close in the states that Biden doesn't need to win. Iowa, Ohio and Georgia are all competitive and the Democrats are making a play for Texas.
Some of this polling was conducted after President Trump tested positive for COVID-19. But most of it was taken before we learned the full extent of how reckless trump was in the days between the first debate and the public disclosure of his diagnosis. There was some concern that people might rally to the president because he is ill but his cavalier treatment of the COVID crisis works against that. And, as can be depended upon, his erratic and reckless behavior only underlines the fact that he is not up for this job.
It has been difficult to get accurate information about the President's health. His doctors and political staff have both obfuscated the details. Some doctors on social media have interpreted the information about his treatment as indicating a severe case. But it is equally plausible that the President is steam rolling his doctors into prescribing whatever treatments are available, without regard to side-effects or medical efficacy.
In the past 24 hours he seems to be doing better. The most encouraging sign so far is that he woke up this morning and released a flurry of nonsensical all caps Tweets from his iPhone. Nature is healing, and perhaps too the president. But he's not out of the woods yet, even though he talked his way out of the hospital a few hours ago. He has been given oxygen at least twice. He is old and he is obese. If he really is loading up on experimental drugs, then he might experience severe side effects. He could take a turn for the worse and leaving the hospital early has all the earmarks of being premature and stupid.
His health will dominate the news, at least until the Veep debate. Now that he is discharged, he will want to recklessly throw himself back into campaigning. Maybe he will wear a mask more often. Maybe he will avoid large crowds. And maybe, just maybe, he will keep being the same brazen jackass that we elected four years ago. (I am purposely avoiding any medical predictions here. That is not my beat.)
But time is not on his side, no matter what his prognosis might be. Trump needs to change the narrative. He probably thinks "I beat COVID" will rally his people to him. But I think this story reminds a lot more people that we are so very bad off because he has been reckless with everyone's life and health, not just his own.
II. Veep Debate Prep.
The Vice-Presidential Debate is not an important event most cycles. Bob Dole screwed up really badly in the 1976 debate, Admiral Stockdale became a punchline in the 1992 debate and the 2008 demonstrated how low the media would set the bar for debate performances by a certified moron when they acted like Sarah Palin deserved praise for not literally shitting on the carpet.
This year will be hyped as a bigger deal because both candidates are elderly and at least one of them has suffered a serious medical condition in the closing weeks of the campaign.
Senator Harris has the cadence and style of a prosecutor. She knows how to build a case. But this doesn't necessarily help in a forum where she can't ask follow-up questions. She was not as effective defending herself in the primary debates and I expect Pence to have a few lines of attack prepared. But her main objective is to look like a moderate, dependable person should the worst happen to her boss in the next four years.
Mike Pence has one incredible gift for these debates. He can can gaslight better than anyone in politics. At some point on Wednesday night, he will say "Of course the president does not believe "X" with an a mildly indignant tone. It will not matter, even if we have all seen vide of the president saying exactly "X" on multiple occasions.
Pence has two goals for this event. First, present a more mature face on Trumpism. It is possible that he will be the candidate by November 3rd and he will almost certainly be a candidate in four years. This is a chance to build up his brand.
III. The Forecasts.
I want to spend a little more time with these this week, because the movement has been significant.
Biden's national lead is significant. He has clear and steady leads in enough states to win the election. He also had stable, though smaller leads in Florida and Arizona. Ohio, Iowa and Texas all look to be competitive. But each forecaster has a slightly different take on where the race stands.
1. FiveThirtyEight.
Nate Silver started this season late and with a conservative estimate of Biden's chances. He opened the forecast giving Biden 71% chance of victory. Since then, our weekly tracking had him at 73-70-72-76-77-78%. So the trend has been good for awhile. As of tonight, Nate gives Joe Biden an 81 percent chance of winning.
Nate's model has a built-in hourglass. Trump is behind. As time and events pass, he loses opportunities to turn that around. Most significantly, Nate had Bide narrowly winning Ohio for most of this week. Right now, it is a true toss up: 49.5% to 49.5%. The leaves his score at 335 Biden, 185 Trump and 18 Toss-up.
2. The Economist.
This continues to be Biden's best forecast. Last week they had him at 85% and tonight they have moved him all the way to 90%. Their EV projection is officially 341 to 197, up from 332-206 last week..
The Economist is one of 2 forecasters who advertise a "mode" result but also publish their state by state probabilities. I prefer the latter method. There Biden remains at 334-204, with no changes from last week.
3. Sabato Crystal Ball
Prof. Sabato just projects the states as individual results, so I don't know what percentage he would assign to a Biden victory. But he currently has Biden winning 279 electoral votes and Trump just 179 with 80 tossups.
Sabato moved both IA and OH from Leans-Trump to Toss-up. And he moved WI from toss-up to lean Biden. He now forecasts that Biden will the Hillary states plus WI/MI/PA and NE-2nd.
He considers FL, IA, OH, NC AZ and ME-2nd to all to be toss-ups. That gives Biden a lot of outs.
4. 270 to Win.
Similar methodology and temperament to Sabato. But he made those changes a week ago, so his ratings didn't move this week. He gives Biden a 278 to 169 lead with, 91 Toss-up votes. (OH, ME-2, NE-2, AZ, NC, GA and FL)
5. JHK Forecasts.
Biden did well here nationally this week, moving from 79.7% to 83.9%. But there were no flips in the individual races. He has Biden getting 335 votes to 203 for Trump. But he thinks Biden has a good shot at winning Georgia. (47%).
This forecast is probably most similar to my own "hunch cast."
6. Plural Vote.
Plural Vote includes internet search data as a component of their forecast. If more people Google Donald Trump, then Trump's numbers go up. I assume they adjust for the fact that Trump is the incumbent but this week, a lot of people Googled to find out more about Trump's medical condition. That affected the model. They currently give Biden a 73.2% chance of winning, with 329 electoral votes as their most common outcome. (This is up slightly from last week when they gave Biden 70.7% and 322 EVs.)
This forecast diverges the most between mode and aggregate mean result. If you award each state to the candidate who wins each contest most often, Biden is all the way up to 352 votes. (They project Biden to win GA about 56% of the time. That seems bullish to me.)
7. Betting Markets.
There was real movement here too. One week ago Biden had a 57.3% chance of winning the election according to the bettors on Predictit.com. Tonight Biden has a 61.5% chance there. This number bounces a little bit as Trumps health waxes and wanes but the movement is clearly in Biden's direction.
8. Yours Truly.
I think this week was a very bad one for the president. He is obviously ill but the whole week unfolded with a distinctly Keystone Cops feel. A number of close associates, including his wife and most of his debate preparation team have now tested positive, along with three friendly senators. His arrogance has been on full display on how he has treated the disease and even his own well-being. I don't really see how he can overcome this.
I am moving Biden up from 86% to 91%. Trump needs to change the trajectory of this course. To do that, he needs to recover quickly, to do well in the remaining debates and to hope that Biden has something bad happen to him. We might look back on the first weekend of October and realize that this was where all of his bullshit caught up with him. The polling data is encouraging. The narrative could not be worse for him. There are 29 days to go and his campaign is in very bad shape. Almost as bad as the candidate himself is in.
IV. Taxonomy, Again.
Standard reminder of my taxonomy of states.
A. It will seem close if he doesn't win Florida. He probably only gets to 290 maximum without Florida.
B. It will seem comfortable if he wins NC and FL That probably gets him to 334 electoral votes. Getting over 350 will look like a nice win, especially if he can win a state like Georgia, which Obama never carried and which ought to be trending blue over time.
C. It will seem like a blowout if he adds Iowa and Ohio. This means he recreated the 2008 coalition, with Georgia subbing in for Indiana.
D. The cherry on top would be Texas. This probably only happens if Trump craters in some way, but the psychological value of flipping Texas will be enormous.
E. A true blow out is possible. If Trump does not recover physically and/or his behavior descends further into the realm of laughing stock, we have to look for some long shot states that might shock the world. I am partial to Montana, a state which has a track record of not liking incumbent presidents. I also think Alaska is in play during a true blowout. Beyond that, South Carolina is the only state I can even entertain as a possibility now. (Nate Silver gives Biden a 14% of winning, which is a bit better than the average National League pitcher getting a hit in any given at bat.)
No comments:
Post a Comment