One hundred days out from Election Day 2020, a lot of indicators point toward a Joe Biden victory over Donald Trump. Yet consistent majorities of Americans, including a sizable number of Democrats, tell pollsters they think Trump will win. At an anecdotal level, I can tell you that every time I write a piece documenting Biden’s lead, all sorts of people — mostly people who plan to vote for Biden — tell me they don’t believe the numbers or don’t trust them. Invariably it’s because of what happened in 2016. Just this week, Michigan congresswoman Debbie Dingell (who calls herself a “gut politician”) was interviewed by The Atlantic and said this:
I don’t trust polling. I don’t believe that Biden is 16 points up in Michigan; that’s a bullshit poll, and it’s the same people who said Hillary had it in the bag. I worry about polling suppressing votes. I don’t want anybody to think their vote doesn’t matter. I’m seeing lots of Trump signs start to pop up. There are some very complicated issues that Trump is playing to divide this country. He is energizing his base, and we have to energize ours.
There’s a lot going on in this quote, but it bespeaks a refusal to even look at objective evidence of how the presidential race is going, and an almost superstitious belief in Trump’s ability to translate hate into votes at the most critical moments.
Now, I can’t dispel the evil images from the 2016 Election Night from the minds of those still traumatized by them.
But it’s worth looking at the many reasons this election cycle is different from the last one, and is not somehow preordained by it.
1. The polls weren’t off that much in 2016, and it’s less likely they are this time around
The final RealClearPolitics national polling averages in 2016 showed Hillary Clinton with an advantage of 3.2 percent in the national popular vote. She won by 2.1 percent. As Nate Silver has explained, the national polls didn’t miss the final result by much, and did so by roughly typical margins.
Yes, the less frequent state polls were also less accurate, particularly in the three states (Michigan, Pennsylvania, and Wisconsin) that hadn’t gone Republican in years and that Trump narrowly won to pull off his inside straight of an Electoral College majority. But even those polls weren’t off by that much. In Pennsylvania, for example, eight of the last nine polls showed a Clinton lead within each poll’s margin of error. That was also true of seven of the last ten polls in Michigan, and five of the last eight in Wisconsin. The problem with both national and state polls wasn’t so much polling error as it was misinterpretation of polls. Reading them more carefully this time around is appropriate, as opposed to refusing to read them at all.
To the extent that there was a source of real polling error in 2016, it was the tendency of some surveys to inadequately poll non-college-educated white voters, which turned out to be a crucial element of Trump’s base. Going into this year there’s been a significant increase in the share of pollsters that weight their results by education level to ensure a reasonably accurate sample of white working-class voters. So poll accuracy, on average at least, ought to improve accordingly.
2. Biden’s polling leads are larger and more stable than Clinton’s were
At present Joe Biden leads Trump by 8.7 percent in the RCP polling averages. Four years ago today Trump led Clinton by 0.9 percent. Biden’s lead at RCP has been at 7 percent or higher since the end of May, and hasn’t dropped below 4 percent the entire cycle. After Trump won the GOP nomination in 2016, Clinton’s lead over Trump topped 7 percent for just two days in August and two days in October.
Biden is also doing consistently better in battleground states than Clinton did once the general election campaign was fully underway. For example, Biden hasn’t trailed in a Michigan poll all year, and is generally holding leads outside the margin of error. And he’s currently ahead in the RCP averages in Arizona, Florida (by seven points), and North Carolina. Georgia and Texas are in play as well.
3. The minor-party vote is likely to be down a lot
It’s not just remarkable that Trump won in 2016 while losing the national popular vote by more than 2 percentage points. He also won with a mere 46.1 percent of the vote, less than any major-party winner or loser since 1996, when Ross Perot’s second run cut into both Bill Clinton’s and Bob Dole’s vote share. Why? Mostly because the minor-party/independent vote was nearly 6 percent, as compared to less than 2 percent in 2012.
There were several reasons for high non-major-party voting in 2016, including better-known-than-usual Libertarian (Gary Johnson) and Green (Jill Stein) Party nominees who were repeat candidates; terrible favorability ratings for both Trump and Clinton; and above all the belief that Clinton had the election in hand from the get-go, which made “protest votes” (particularly from former Bernie Sanders supporters) seem less risky.
This time around the Libertarians and Greens have new and less-well-known candidates; have struggled with ballot access and getting state campaigns underway thanks to the COVID-19 pandemic; and as I noted at the top of this post, Joe Biden’s polling lead hasn’t convinced much of anybody the election is “in the bag.” In particular, Democratic voter determination to get rid of Trump has significantly increased the intensity of progressive voter support for Biden, while creating a lot of negative pressure on would-be “protest voters.” So far, despite generally ineffective Trump campaign efforts to demonize “Sleepy Joe” as a puppet of the Radical Left, the Democrat’s favorability ratios are far better than Clinton’s were.
The odds are very good the minor-party vote will be closer to 2012’s numbers than 2016’s, which means Trump needs to do something he’s never been good at: reach beyond his base.
4. Trump is the incumbent in 2020 in a bad year for incumbency
In 2016 Trump was able to run as the “change candidate,” palpably hostile to both major-party Establishments, while Clinton was the candidate of the party that had held the White House for eight years characterized by sluggish economic conditions and partisan gridlock in Washington. That probably had a lot to do with Trump’s relatively strong performance among voters who didn’t like either candidate.
In 2020 Trump is inescapably the incumbent. His independence from things voters don’t like about Republicans is gone after his intensely partisan policies and conduct while in office.
Given the pandemic and the economic fallout from it, it’s not surprising that Americans are very sour about the direction of the country, as reflected in right track/wrong track polling (19/72, according to the most recent NBC News/Wall Street Journal survey), which has typically foretold the fate of incumbent presidents. While perhaps another president could have treated this year’s calamities as Acts of God for which no blame can be assigned, Trump’s steadily deteriorating public assessments for his handling of COVID-19 make that difficult. The pandemic has also drawn attention to associated policy areas (e.g., health care and racial justice) where Trump and his party are not trusted at all.
It’s never easy for an incumbent president to turn reelection from a “referendum” into a “choice,” but the uphill battle Trump faces is illustrated by the fact that this time around he’s losing badly with voters who don’t much like either candidate. Right now he is poorly positioned to win either a “referendum” or a “choice” election.
5. Trump’s act has grown old
While Trump was a familiar public figure in 2016 thanks to his relentless self-promotion and his television career, he was the ultimate untested quantity as an elected official, much less a president. Now, after he has dominated public discourse to a degree probably unmatched by any president since FDR, his act is universally well-known, and is generally either adored or deplored in the polarized atmosphere to which he has contributed so much. You could probably count on one hand the number of occasions on which the 45th president has tried seriously to transcend his abrasive and divisive persona with any sort of unifying message, and even in 2020, when he desperately needs to expand his electoral base, he can’t seem even to try. Instead he’s doubling down on racism, authoritarianism, and mendacity.
Unsurprisingly, a large segment of the electorate has all but ruled out voting for him no matter what happens between now and November. Yes, his favorability ratio was poor in 2016. But what’s remarkable now is the number of Americans who give him a very unfavorable rating — fully half the electorate in a recent Monmouth survey — which suggests he may need to win virtually all persuadable voters to win reelection. Meanwhile, the excitement Democrats express about voting Trump out of the White House is clearly offsetting any misgivings or mixed feelings they have about Biden.
There is, however, at least one difference between 2016 and 2020 that doesn’t necessarily benefit Joe Biden:
6. We probably won’t know who won the morning after Election Day
The razor-thin margins by which Trump put together his Electoral College majority and the unexpected nature of the outcome shouldn’t make us forget the fact that his victory was called by the Associated Press at 2:29 a.m. EST. Hillary Clinton called Trump to concede just a few minutes later.
That’s not going to happen this November unless one candidate or the other has won by a landslide.
COVID-19 fears about voting in person will almost certainly produce record levels of voting by mail, particularly among Democrats (Republicans may be inhibited from following suit by Trump’s relentless attacks on voting by mail). Mail ballots are often counted after Election Day as a matter of law. Even where they can be counted quickly they must undergo initial processing to verify deadline compliance (most states require that mail ballots be received by Election Day, though a few, like California, accept ballots postmarked by Election Day) and signature verification. The bottom line is that we are likely to see a slow, slow count in November in which (relatively speaking) strong early Republican performances based on in-person voting are eroded or reversed within a day or a week by Democratic-leaning mail ballots as they are finally tabulated.
It’s hardly a stretch to envision Trump claiming victory on Election Night and then crying “fraud” as his leads in key states vanish. The country could easily enter a nightmare landscape of litigation, skullduggery, and threats — perhaps accompanied by violence in the streets. It’s unclear how far Trump would go to hang onto the presidency after losing it when all the votes have been counted. But it could get ugly and dangerous.
7. Nothing will surprise us
Suffice it to say with Donald Trump in the White House and the electoral system in chaos, things could happen that are far more surprising — and alarming — than Trump’s 2016 upset win. But that could indeed be Joe Biden’s hidden weapon: zero complacency in his campaign or in the ranks of his supporters. It may take heroic turnout to produce a Biden win big enough to avoid a contested election. But that’s certainly on the table. Nobody’s going to coast toward November 3, ready to party the moment polls close. Until Donald Trump leaves the White House for good, those Democrats who gazed in horror as his victory was proclaimed in 2016 are not going to rest easily, or perhaps at all.