Why the polls were wrong about Trump (again)

Andrew Romano

As of Wed afternoon, nosotros all the same don't know for certain who the next U.Southward. president will be. That'due south because — as expected — cardinal states such as Michigan, Pennsylvania, Nevada, Arizona and Georgia haven't finished counting their votes.

But one matter is articulate, even at present: Pollsters systematically underestimated President Trump'south support — over again. This fourth dimension, they missed by an even bigger margin than in 2016. And Yahoo News was no exception; our concluding YouGov poll gave Democrat Joe Biden a 10-indicate pb.

And then even if Trump doesn't beat Biden, he has largely beaten his pre-election polls.

The question is why almost everyone was off.

The 2020 polling error "matches the blueprint of the 2016 mistake really well, so in that location really does seem to be something wrong here," explained G. Elliott Morris, a data journalist who runs the Economist's election forecast, during a Wednesday postmortem on the "Scientific discipline of Politics" podcast. "It's non just ii random polling errors."

U.S. President Donald Trump arrives an election night party in the East Room of the White House in Washington, D.C., U.S., on Wednesday, Nov. 4, 2020. (Al Drago/Bloomberg via Getty Images)
President Trump at an ballot night political party in the Eastward Room of the White House. (Al Drago/Bloomberg via Getty Images)

Beyond the Rust Belt in particular, the polls and the results weren't fifty-fifty close:

  • In Michigan, this yr'southward final FiveThirtyEight polling boilerplate (the most comprehensive and careful in the field) showed Biden alee by 7.9 percentage points. The race is at present on track to be decided by less than 2 points. Four years ago, Hillary Clinton led by an average of nearly 4 points in Michigan heading into Election Solar day. She lost by one-quarter of a point. That means this year's Michigan polls were off by 6 or 7 points — and less authentic than 2016'southward past something like 2 or 3.

  • In Wisconsin, this year'southward final polling average showed Biden ahead by viii.iv per centum points. He is now on track to win past less than 1 bespeak. 4 years ago, Clinton led by an average of nigh 5 points in Wisconsin heading into Election Day. She lost to Trump by less than a point. That means this year'south Wisconsin polls were off by as much as viii points — and less authentic than 2016'south by about 2.

  • In Ohio, this yr'southward final polling average showed Trump ahead by 0.8 per centum points. He is now on runway to win by viii. Iv years ago, Trump led by an average of 2 points in Ohio heading into Election Mean solar day. He won by 8. That means this year's Ohio polls were off by about 7 points — and less authentic than 2016'due south past almost 1.

  • And in Iowa, this twelvemonth's final FiveThirtyEight polling average showed Trump alee by 1.3 points. He is now on track to win by about vii. Four years ago, Trump led by an average of 3.4 points in Iowa heading into Election Mean solar day. He won past 9. That ways this year's Iowa polls were off by nearly 6 points — and just every bit inaccurate every bit 2016's.

The list goes on.

Residents cast their votes on November 3, 2020, at Eisenhower Elementary School in Flint, Michigan. (Seth Herald/AFP via Getty Images)
Residents cast their votes in Flint, Mich., on Tuesday. (Seth Herald/AFP via Getty Images)

  • In Florida, Biden led in the final polling boilerplate by 2.5 points; information technology looks like he is going to lose by almost 3.v, for a vi-point polling miss.

  • In Nevada, Biden led in the concluding polling average by 5.three points; it looks like the results will be extremely close, for a roughly 5-bespeak polling miss.

  • In Texas, Trump led by 1.ane points in the final polling average; it looks like he is going to win by 6, for another 5-bespeak polling miss.

  • In Pennsylvania, Biden led in the final polling average by four.seven points; that will probably prove to be off by a few points once all the votes are tallied.

Fifty-fifty the national polls appear to take missed the mark this time around. In 2016, Clinton led by 3 to 4 points in the final polling average, depending on which average you consulted. She wound upwards winning the national popular vote by two.i points, and so the polls back and then weren't that far off.

In 2020, however, the final FiveThirtyEight national polling average showed Biden ahead by a lot more than Clinton: eight.4 points. Like his Democratic predecessor, the quondam vice president volition win the popular vote once big bluish states such as California, New York, New Jersey, Maryland, Massachusetts and Illinois terminate counting more than than 10 million outstanding ballots. One gauge has Biden, who is currently leading by three million votes, winning by roughly 5 million to 7 million — virtually twice Clinton'due south 2016 margin.

Democratic Presidential candidate Joe Biden speaks at the Chase Center in Wilmington, Delaware on November 4, 2020. (Jim Watson/AFP via Getty Images)
Joe Biden speaks in Wilmington, Del., on Wednesday. (Jim Watson/AFP via Getty Images)

But in calorie-free of this yr's historic turnout on both sides, that would equal a gap of just iii to 5 percentage points. In other words, the national polls may have missed this twelvemonth by nearly twice as much as in 2016, if the estimates concord.

Yahoo News was amongst the media organizations whose national numbers probable overstated Biden's advantage. We partnered with YouGov early this wheel and asked it to conduct regular national opinion surveys about politics, coronavirus and the news of the calendar week on our behalf. In those polls, the smallest lead YouGov e'er found for Biden was 5 points. YouGov's final Yahoo News survey showed Biden leading Trump 53 percent to 43 percent.

At the time, we wrote that a 10-indicate Biden victory "is hardly bodacious. But given the size and stability of Biden's national atomic number 82, the scope of his competitiveness in key states and the staggering number of Americans who have already voted … the president volition secure a second term only if polls are underestimating his support by far more than in 2016, when they missed by about 1 signal nationally and about 4 points in Michigan, Wisconsin and Pennsylvania."

That turned out to be exactly what happened on Election Day — aside from the part most Trump securing a 2nd term, for which he faces an uphill battle in the Electoral College. As a result of similar errors, Democrats are also coming up short in more Senate and House races than expected.

Regardless of who wins the White House, talk will turn in the days alee to why 2020'southward polling was even more than wrong than 2016'due south. Right now, the answer is "nosotros don't know."

Polling errors are commonplace; they happen in about elections. But they tend to be random and to change directions from i cycle to another. Certain, in 2016 the swing-country polls underestimated Trump by iv points or and so. But four years earlier, in 2012, they underestimated Barack Obama by about the same amount.

An attendee at a watch party for Republicans waves a Trump 2020 flag on election day, November 3, 2020 in Austin, Texas. (Sergio Flores/AFP via Getty Images)
An attendee at an election nighttime watch political party for Republicans in Austin, Texas. (Sergio Flores/AFP via Getty Images)

One reason this happens is that pollsters are always adjusting their methodologies to avoid making the same error twice. From 2016 to 2020, this was supposed to ensure that white voters without college degrees — a hard-to-reach but unduly pro-Trump demographic — would exist better represented in their surveys. The shortfall was supposed to be fixed.

Only the troubling thing about 2020 is not just that the polls missed once more. It's that they missed again in the verbal same places they missed in 2016 — and they missed past fifty-fifty more than. This suggests that any went amiss in 2016 has only worsened over the final four years, and that "weighting for education" or other methodological tweaks can't correct for information technology.

Instead, the bug causing pollsters to increasingly understate Trump's back up are probably more profound than that. The polls were very authentic in 2018, when Trump wasn't on the ballot, and certainly in states, such as Arizona, it seems that pollsters are really underestimating Democrats.

But those are the exceptions that evidence the dominion. For two elections in a row now, Trump has drawn on unanticipated wells of support to perform improve on Election Twenty-four hours than the polls predicted — and despite the best efforts of the best pollsters in the business to become meliorate at measuring his vote, they seem to take gotten less accurate, not more than.

It'due south going to take time to identify the cause — or more probable, causes — of this stumble. Early theories abound. During Wednesday's "Science of Politics" postmortem, Morris and host Matt Grossman, a political scientist at Michigan Country, considered several possibilities.

In s Florida, pollsters failed to catch a huge Trump-ward shift amid Cuban Americans that largely accounted for the president'due south statewide victory; Hispanic districts in Southward Texas as well swung heavily toward Trump.

In the Northern battlegrounds, meanwhile, surveys seem to accept correctly pegged Biden'south improvement with white, college-educated suburban voters — but greatly overestimated the Democrat's inroads among the white, non-college voters who surprised pollsters in 2016.

Voters lined up to cast their ballots at the Bishop Leo E. ONeil Youth Center Ward 9 polling place in Manchester, New Hampshire, one of the city's busiest, as Americans rushed to vote in the presidential election on November 3, 2020. (Jodi Hilton/NurPhoto via Getty Images)
Voters at a polling identify in Manchester, N.H., on Tuesday. (Jodi Hilton/NurPhoto via Getty Images)

Researchers accept largely ruled out the idea of "shy Trump voters" who prevarication to pollsters and say they're undecided or that they favor someone else when they really favor Trump. Simply it'southward possible, Grossman and Morris speculated, that pro-Trump, non-college-educated whites are merely less inclined to choice up the phone or participate in polls than pro-Biden, non-college-educated whites.

Why? Because the pro-Trump cohort also tends to have less "social trust" — i.due east., less "trust in other people or institutions," as Morris put it. Spurred past Trump's "fake news" mantra, participating in polls may have itself become politicized. When overall response rates are as low equally 4 per centum, this could skew the results against Trump in places similar the Rust Chugalug or Texas.

A similar dynamic may have also made it seem like more Republicans were flipping from Trump to Biden than ultimately did — over again, because pro-Trump Republicans may be less inclined than pro-Biden Republicans to answer a pollster'southward call or participate in an online survey.

Other potential reasons for 2020'southward big miss may take been beyond anyone'due south control. Information technology's unlikely that belatedly-breaking voters who decided within the last week made the departure, fifty-fifty though they told exit pollsters they favored Trump over Biden past 14 pct points. There simply weren't enough of them this yr — just four to 5 percentage of the overall electorate vs. most fourteen percent in 2016 — to explain Trump's overperformance on Ballot 24-hour interval.

A more plausible scenario, Morris said, is that a meaning number of pandemic-induced mail ballots are either arriving late, or being rejected, or not existence returned at all. If tons of people tell pollsters they've voted past mail and then, for any reason, some those "probable votes" don't actually count on Election Solar day, it could widen the gap between the polls and the results.

Political scientists and pollsters will debate these bug for years to come up, and they'll probably devise new approaches to bargain with them. But after existence told that Joe Biden could win in a landslide — and and then watching as Trump beat his polls by even more 2016 in state after state — the broader public might be more than inclined to dismiss political surveys in the hereafter.

There are "systematic issues that they oasis't solved since 2016, and in fact seem to be worse this time," Morris said. "That'due south pretty troubling if you're a pollster — especially if you lot've spent the final four years trying to reckon with the fact that polls were missing Trump supporters. Then they have a big reckoning ahead of them."

_____

Read more from Yahoo News:

  • How to watch election night 2020: The definitive hour-by-hr guide

  • From legal battles to voter intimidation, a short guide to what could go wrong on election nighttime

  • Black lawmakers explain what Kamala Harris means to them

  • Don't freak out over early leads for either Biden or Trump on ballot nighttime

  • Fauci's balancing act: Navigating Trump and the pandemic