One mistake pollsters have more than corrected for since 2016 is in continuing to poll right up through the end in order to detect any late movement in the electorate. And although the race appears to have tightened by maybe a point or so in the final days, Trump hasn’t closed the gap nearly enough to have significantly improved his chances of winning—particularly not in a race where more than 96 million votes representing nearly 70% of the total 2016 vote count have already been cast.
Nationally, The New York Times aggregate of polls has Biden up by eight points, a one-point decrease from where the race has stood for the past couple weeks.
Another place to look for late movement was a one-day spot poll from NBC News/Wall Street Journal of 800 registered voters conducted Sunday in the 12 battleground states of Arizona, Florida, Georgia, Iowa, Maine, Michigan, Minnesota, Nevada, New Hampshire, North Carolina, Pennsylvania, and Wisconsin. The poll found Biden running ahead of Trump by five points in those battlegrounds, 51-46%, while the outlets’ final national survey gave Biden a 10-point advantage, 52-42%. Not only is that tighter window in the battlegrounds to be expected, it also didn’t budge more than a point from the battleground poll the organizations conducted from Oct. 29-31, which found Biden up six points, 51-45%. In other words, the poll found almost no late movement, not even in the battlegrounds. And a stable race at this point is a good race for Biden and Democrats.
A few states to keep in mind as we head into election night coverage are North Carolina, Florida, and Pennsylvania, but for different reasons. North Carolina and Florida both allow the counting of their mail-in ballots to start before Election Day, which suggests they will likely report results much earlier than the Midwestern Rust Belt states where mail balloting counts will proceed at a much slower pace. North Carolina, where The New York Times aggregate shows Biden up by two points, could be an early indicator of whether Democrats are moving in a good direction right out of the gate or whether it could take longer to sort out the night’s results. Losing North Carolina wouldn’t necessarily be a death blow to Trump, but it makes his path to 270 much more narrow while giving Biden an early edge in the Electoral College.
Florida, on the other hand, would basically ruin Trump if Biden won the state. Biden is also ahead by two points there, but Florida is always on a razor’s edge and it’s, well, Florida—a scrap heap of mangled Democratic dreams from elections past. So …
Pennsylvania, where Biden is up six points in the Times aggregate but the vote count could take days, is where the election is most likely to be decided. The last couple days of high-quality surveys from noted pollsters have looked solid for Biden. Here’s a sampling:
|Monmouth (Oct. 28-Nov. 1, LV)||51-45%||+6|
|Ipsos (Oct. 27-Nov. 1, LV)||52-46%||+6|
|NYT/Siena College (Oct. 26-31, LV)||49-43%||+6|
|ABC/Washington Post (Oct. 24-29, LV)||51-44%||+7|
|Muhlenberg College (Oct. 23-28, LV)||49-44%||+5|
That’s a very consistent data set in the final days of the campaign in what could prove to be the most consequential state of the race. Biden has multiple paths to 270, but winning Pennsylvania is the surest way for him to seal the deal, flipping it and both Michigan and Wisconsin back to the blue column while holding all the other states Hillary Clinton won in 2016. North Carolina could be a stand in for Pennsylvania if Biden also takes Michigan/Wisconsin, which is why it could be an early indicator of a good night for Biden.
Nate Cohn has generally provided some very good day to day analysis of the polls, and I think his frame is about right: Trump would need a much bigger polling error this year to pull off the upset he did in 2016. Of course, that doesn’t account for Trump’s efforts to disrupt the vote count, sow disinformation, and potentially declare an early/baseless victory—that’s a different post.
Cohn has regularly written that how big a lead Biden actually has in this race really comes down to how big a lead Biden has in the Keystone State. If he’s only up by four points there, that’s enough within the margin of error that it amounts to a slim Biden lead. If it’s an eight-point lead, that’s enough outside the MOE that it’s a significant lead. Where things appear to stand is somewhere in between—solid but not exactly decisive.
So if you need to take the edge off in the next 24 hours, definitely sign up for phone banking or text banking to turn out every last Biden voter in Pennsylvania.