Dem pollsters acknowledge ‘major errors’ in 2020 pollingApril 13, 2021
A group of top Democratic Party pollsters are set to release a public statement Tuesday acknowledging “major errors” in their 2020 polling — errors that left party officials stunned by election results that failed to come close to expectations in November.
In an unusual move, five of the party’s biggest polling firms have spent the past few months working together to explore what went wrong last year and how it can be fixed. It’s part of an effort to understand why — despite data showing Joe Biden well ahead of former President Donald Trump, and Democrats poised to increase their House majority — the party won the presidency, the Senate and House by extremely narrow margins.
“Twenty-twenty was an ‘Oh, s---' moment for all of us,” said one pollster involved in the effort, who was granted anonymity to discuss the process candidly. “And I think that we all kinda quickly came to the point that we need to set our egos aside. We need to get this right."
That’s about where the answers end. The collaboration’s first public statement acknowledges that their industry “saw major errors and failed to live up to our own expectations.” But the memo also underscores the limits of the polling autopsy, noting that “no consensus on a solution has emerged.”
According to Democrats involved with the internal review, Tuesday’s statement marks the beginning of a years-long process to examine why, since 2012, most major elections have tilted against the party, despite favorable polling data before the vote. Up and down the ballot, Democrats have been, more often than not, shell-shocked by defeats in races they thought to be competitive, or narrower-than-expected, victories in contests they thought they led comfortably.
Democrats are not alone in reviewing what went wrong last year. The polling industry is engaged in multiple reviews of its 2020 performance, including a forthcoming report from the American Association for Public Opinion Research’s task force that is expected to address the overestimation of Democrats’ performance, from the presidential race down to races for Congress and state offices.
The previously undisclosed Democratic polling review is not being replicated by Republicans, who ultimately lost the presidency and the Senate, and won fewer House seats than Democrats. While some in the GOP were also surprised by the party’s competitiveness last November and are studying their methods, there is no similar, organized effort moving into the 2022 midterms and the 2024 presidential election.
There’s no simple answer for why the polls have missed the mark in recent elections. But one likely culprit for some of the errors is the deteriorating public trust in institutions, like government and the news media — and the correlation between that wariness and voting for Trump. Between his public statements and Twitter account, the former president cast doubt on polling specifically, which the Democratic consultants suggested led to his supporters refusing to participate in surveys.
“Trump went after the polls,” said another Democratic pollster involved in the partnership. “He was really pretty overt to those that were listening about some of his distrust of polls or media.”
The 2020 election shattered turnout records — and since November, pollsters have been eagerly awaiting official information from the states about who voted, and who didn’t. That data is now almost entirely available, and there are clues hidden within.
The Democratic pollsters, who typically compete against each other for business, acknowledge that Trump was able to activate large numbers of voters who had turned out less reliably in the past. Looking at one state where the polls were off — Iowa, where Trump beat Biden handily and what had been seen as a toss-up Senate race went decisively for incumbent GOP Sen. Joni Ernst — Republicans classified as “low-propensity voters” turned out at four times the rate of Democrats in that category, according to the Democratic memo.
“This turnout error was clearly one factor in polling being off across the board, but especially in deeply Republican areas,” the memo reads. “It also meant, at least in some places, we again underestimated relative turnout among rural and white non-college voters, who are overrepresented among low propensity Republicans.”
But sky-high turnout for Trump among irregular voters only explains a small slice of the problem, the pollsters concluded. Even if the polls conducted last year were properly adjusted for future turnout, they still would have been biased toward Democrats.
The memo floats at least three possible causes: late movement toward Trump and Republican candidates that polls conducted in the run-up to the election failed to catch, the Covid pandemic causing people who stayed home to answer the phone at a greater rate than those who did not follow restrictions, and the decline of social trust and faith in institutions.
But there’s little clarity about how significant each of those hypotheses was.
“While there is evidence some of these theories played a part, no consensus on a solution has emerged. What we have settled on is the idea there is something systematically different about the people we reached, and the people we did not,” the memo reads. “This problem appears to have been amplified when Trump was on the ballot, and it is these particular voters who Trump activated that did not participate in polls.”
Some Democrats believe these errors are a direct Trump effect — that he is a singular force in politics, engendering extreme opinions on both sides — and it will fade if he’s no longer a candidate.
“I don’t think we know what it is. I think we still have a lot of work to do to figure it out,” one pollster said. “I’m marginally optimistic that if Trump is on the ballot in ’24 that we can fix it. I don’t know. If he’s not, I do think a lot of it could resolve itself.”
The Democratic effort stands in contrast to the last major review of a party’s polling practices. Following the 2012 election, the Republican National Committee’s so-called “autopsy” — best known for its recommendation that party leaders moderate their views on immigration and other social issues — included a list of best practices for pollsters, who were also summoned to party headquarters for discussions.
The five Democratic firms that signed onto the memo are ALG Research, Garin-Hart-Yang Research Group, GBAO Strategies, Global Strategy Group and Normington Petts. Together, they are five of the top six polling firms working for the Democratic Party apparatus, along with Greenberg Quinlan Rosner Research, according to financial disclosure reports. ALG Research was Biden's lead pollster in last year's election.
Participants in Democrats’ review said the process — which, unlike the GOP’s 2012 effort, was not dictated from party officials — was collegial, despite the fact that the five firms compete against each other for business.
“One should feel comfortable talking to your competitors because, ultimately, we all want the same thing,” said a third participant. “We all want a useful way to help give guidance to Democratic candidates and progressive causes.”