Once upon a time, Labor Day was the official kickoff of the General Election campaign. Candidates would rest up and prep after their conventions, followed by a two month sprint. That sort of calendar is long in the past. Joe Biden won’t be addressing a large Labor Day crowd in Detroit at Cadillac Square, as JFK and many other Democratic nominees did in the past.
It’s likely Biden won’t address a single large crowd at any time this fall. Even Donald Trump has made huge compromises in his preferred methods of campaigning. After the Tulsa embarrassment, he’s not going to try to fill a large arena real soon.
With the virtual conventions drawing less attention than the historical norm (admittedly, conventions already weren’t what they used to be), a limited amount of truly undecided voters, and COVID messing with any and every norm, there are times when it seems the campaign hasn’t started yet.
It’s an odd thought in an era where candidates regularly announce their intentions to run an entire year ahead of the first primary or caucus. Maybe the traditional Labor Day starting gun will fire, and we’ll wake up Tuesday morning feeling like an election is on. Given that early voting is beginning this month in many states, it’s gotta sink in eventually.
So where are we? How likely is Trump to win? Are the polls moving? Is Biden fading? Can we even trust the polls? Will Democrats take Texas? Is Biden going to fall short in his native Pennsylvania?
Some of these questions are very answerable. Biden is ahead right now. He’s led Trump in the Real Clear Politics polling average for the past year. He and Bernie Sanders led Trump in most national surveys from 2017 though 2019 also. Pick whatever polling average you want, from just about whenever you want, and you’ll find Biden ahead of Trump. It’s a struggle to find individual national surveys that put the president ahead. Impossible if they were taken since COVID got out of control.
Contrary to popular belief, the polls were largely accurate in 2016. Hillary Clinton underperformed her final national numbers by a percent, maybe a percent and a half. Even in the crucial states of Michigan, Wisconsin, and Pennsylvania, surveys were only off by a couple points. All of this was well within the margin of error.
Polls in the 2018 midterms were solid. Most primary and caucus surveys for Democrats in February/early March of this year were accurate enough. I’m not sure how pollsters are managing this with so many Millennial and Gen Z voters refusing to pick up the phone, but whatever the obstacles, polls are as reliable now as they’ve ever been, particularly if you’re averaging instead of picking your favorite.
So Biden is leading. And given the small amount of up for grabs voters, I’m confident in asserting he’s going to win the popular vote. At least 50.1% of the country is going to vote against Trump. But we know the popular vote doesn’t decide the presidency. Not only did Trump win 304 electoral votes while losing by 3 million votes, but he likely had enough margin to have won even if the gap was 4, maybe 4.5 million, perhaps even five.
Nate Silver has done his usual mathematical modeling, and reached the following conclusion:
0 to 1% : 6% chance of Biden Electoral College win
1 to 2% : 22%
2 to 3% : 46%
3 to 4% : 74%
4 to 5% : 89%
5 to 6% : 98%
6 to 7% : 99%
Biden is up by roughly 7% right now, yet Silver is estimating he’s got a 70% chance of winning the election, not 99%. Why?
Two things. Time and polling error. If the election were today, he’s figuring Biden would be at least 90% to win. You can see this on the chart. Deduct a couple points for error, and he still looks pretty good. But, should Biden lose a couple of points, plus a small polling error, and now we’ve got a toss up.
Whether it’s good third quarter economic numbers (at least relative to expectations), a last minute vaccine, multiple bad Biden debate performances, Trump voters getting to the polls/mail in higher percentages, it doesn’t take a flying logical leap to see how the gap could get cut in half, at which point Trump has a legit chance.
It’s just as possible that Biden wins the popular vote by 10 points. No vaccine in time, bad economic numbers, Trump stumbles badly in debates, continued controversy over comments about soldiers being “suckers.” The difference is, while I’m sure Biden would want as large a victory as possible for reasons of legitimacy and leverage, his Electoral College odds are already 99% plus at a 7 point margin. It would get him declared the winner a little closer to Election Night. But that’s it.
If Trump picks up a couple points, we have a real puzzle. Surveys are indicating Biden is doing better than Hillary with older voters and white voters, while Trump is running ahead of his 2016 pace with Latinos and African American men. You might find this surprising, but the data is consistent. While Trump’s Electoral College advantage is similar to last time, it may manifest itself in different states.
To this end, I’m planning on going state by state and taking a look at which way it’s recently leaned, if there’s a difference this year, how it got how it is (or seems to be), and how likely it is to flip from 2016 this time. If all goes according to plan, we’ll begin tomorrow in Michigan.
Unless, or until Biden has a clear double digit national lead, within a week of Election Day, this contest is far from over. Let the exploring begin.