HUFFPOLLSTER: Here's What We Know About Why Polling Missed The Mark

A lot of work is going into figuring it out.

The American Association for Public Opinion Research has a committee to investigate what went wrong with pre-election polling. There apparently was a hidden Trump vote that polls weren’t capturing. And low voter turnout helps explain how Hillary Clinton won the popular vote but lost the Electoral College. This is HuffPollster for Thursday, November 10, 2016.

POLLING ASSOCIATION: ‘POLLS CLEARLY GOT IT WRONG’ - The American Association for Public Opinion Research, in an emailed statement: “Election years present particularly high profile moments for public opinion and survey research. This is a time when polls dominate the media and the accuracy of polls can be confirmed or refuted by the actual poll vote outcome. The polls clearly got it wrong this time and Donald J. Trump is the projected winner in the Electoral College.  Although Clinton may actually win the popular vote, her margin is much lower than the 3 to 4 percent lead the polls indicated.  And many of the state polls overestimated the level of support for Clinton. There is much speculation today about what led to these errors and already the chorus of concerns about a ‘crisis in polling’ have emerged as headlines on news and social media sites. As final results continue to be tabulated it would be inappropriate for us to participate in conjecture. Pre-election polling is critical to the industry. Such polling can support the democratic process and it offers a very public opportunity to showcase the benefits, and weaknesses, of survey research. Therefore, understanding and being able to articulate the overall outcomes of election polling, the changing methodologies being used, and the potential for variation in the accuracy of polls is vital for the industry.”

Task force will investigate results - More from AAPOR: “As it has done in the last several elections, AAPOR has already convened a panel of survey research and election polling experts to conduct a post-hoc analysis of the 2016 polls. The goal of this committee is to prepare a report that summarizes the accuracy of 2016 pre-election polling (for both primaries and the general election), reviews variation by different methodologies, and identifies differences from prior election years. The committee was convened in April 2016 and is chaired by Courtney Kennedy (of the Pew Research Center) and includes Scott Clement (Washington Post), Kristen Olson (University of Nebraska-Lincoln), Claire Durand (University of Montreal), Lee Miringoff (Marist College), Doug Rivers (YouGov), Josh Clinton (Vanderbilt University),  Mark Blumenthal (SurveyMonkey), Chris Wlezien (University of Texas), Kyley McGeeney (Pew Research Center), Evans Witt (PSRAI and President of NCPP), Charles Franklin (Pollster.com and University of Wisconsin), and Lydia Saad (Gallup). The committee should have completed its work by May of 2017.”

POLLSTERS OFFER PRELIMINARY THOUGHTS ON WHAT WENT WRONG - Carl Bialik and Harry Enten: “It will take a while to figure out exactly why polls missed. Reviews by pollsters and their professional organizations can take months. ‘The polls were largely bad, including mine,’ Patrick Murray, director of the Monmouth University Polling Institute, wrote us in an email. ‘But if anyone thinks they have the answer right now, they are just guessing.’... Individual polls were wrong. Aggregated, they missed in individual states, including in many swing states. National polls were off in the same direction: Polls overstated Clinton’s lead over Trump. And her true lead wasn’t enough to overcome her weak position in the Electoral College….While the errors were nationwide, they were spread unevenly. The more whites without college degrees were in a state, the more Trump outperformed….Pollsters also cited lower-than-expected turnout, particularly in the Midwest.” [538]

FORECASTS MISSED TOO - HuffPollster: “I didn’t think there was any way Republican Donald Trump would win. But my presidential and Senate forecasts for The Huffington Post badly differed from what played out on Tuesday night and Wednesday morning…. The model structure wasn’t the problem. The problem was that the data going into the model turned out to be wrong in several key places….  [T]here are some key lessons regarding the forecast model that I can address now. Polls alone probably don’t make a reliable forecast model…. Political science forecasts had a better night than poll-based ones. As John Sides pointed out, several very early forecast models constructed by political scientists pointed toward a Trump win, or at least a very close race…. We may have excluded polls that mattered from our model…. Our decision to not include all-landline automated polls might have been more consequential... Steve Shepard... pointed out that many of the most accurate final polls were from all-landline automated pollsters. The landline polls most likely just got lucky in that their biases toward older people in rural areas who still have landlines trend conservative, and the electorate was more conservative than anticipated. We’re going to look into whether that affected our averages substantially.” [HuffPost]


-Amy Walter: “[A]t its core the Obama coalition lost and the ‘missing white voter’ won.” [Cook Political Report]

-Sean Trende: “It is far too early to say for certain, but what happened Tuesday certainly looks consistent with the missing-whites thesis.” [RCP]

-Nate Cohn: “Donald J. Trump won the presidency by riding an enormous wave of support among white working-class voters….the polls were wrong about one big thing: They missed Mrs. Clinton’s margin in the Midwestern and Rust Belt states, like Wisconsin, Michigan and Pennsylvania….In the end, many of the factors that made Mrs. Clinton appear favored to win in these states simply weren’t there.” [NYT]

-Larry Sabato, Kyle Kondik and Geoffrey Skelley: We heard for months from many of you, saying that we were underestimating the size of a potential hidden Trump vote and his ability to win. We didn’t believe it, and we were wrong.” [Sabato’s Crystal Ball]

-John Sides: “Early political science forecasts were, on average, correct….Party loyalty is still very potent….Candidates and campaign activity seemed to matter less than we thought…. Identity politics can help either party…. Politics is cyclical.” [WashPost]

-Andrew Mercer, Claudia Deane and Kyley McGeeney: “Across the board, polls underestimated Trump’s level of support....The fact that so many forecasts were off-target was particularly notable given the increasingly wide variety of methodologies being tested and reported via the mainstream media and other channels.” [Pew Research]

-Jon Cohen and Mark Blumenthal: “Whatever the final explanation, the bottom line was a polling error, on internet polls as well as live-interviewer and automated telephone polls, big enough to lead to the wrong conclusion about who would win….One thing is already clear: how pollsters determine who is actually going to vote is broken, regardless of the various approaches, whether by public pollsters or the campaigns themselves.” [SurveyMonkey]

-David Rothschild: “Polling does two things: estimate the voter population and the support for each candidate from the voter population; relying on other people’s public polling it is very hard for a forecaster to estimate where they will go wrong.” [PredictWise]

-Sam Wang: “In addition to the enormous polling error, I did not correctly estimate the size of the correlated error – by a factor of five….We all estimated the Clinton win at being probable, but I was most extreme. It goes to show that even if the estimation problem is reduced to one parameter, it’s still essential to do a good job with that one parameter. Polls failed, and I amplified that failure.” [PEC]

-Andrew Gelman: “Trump outperformed the polls in several key swing states and also in lots of states that were already solidly Republican….During the campaign we talked about the idea that swings in the polls mostly didn’t correspond to real changes in vote preferences but rather came from changes in nonresponse patterns: when there was good news for Trump, his supporters responded more to polls, and when there was good news for Clinton, her supporters were more likely to respond. But this left the question of where was the zero point.” [Andrew Gelman]

HOW CLINTON LOST THE ELECTORAL COLLEGE, BUT WON THE POPULAR VOTE - Michael McAuliff, Janie Velencia and Nick Baumann: “With votes still trickling in, Clinton held about a 200,000-vote edge in the popular contest, even as Trump appeared headed for a stronger Electoral College win than George W. Bush secured in his first election ― when he lost the popular vote by 500,000 ballots. How could that happen? The short answer is key Democratic voters didn’t bother to go vote, while white Republican voters in rural areas of swing states did. In 2000, turnout was low, with a 55.3 percent turnout by eligible voters helping to hand Bush the White House after the infamous contest in Florida was decided by the Supreme Court. Based on initial vote tallies, turnout for the 2016 election will be the lowest since then, when Al Gore won the popular contest, but lost the Electoral College, 271 to 266. Trump is doing better. And just 55.7 percent of people eligible to vote ― about 128 million people ― showed up at the polls, according to early estimates.” [HuffPost]

HUFFPOLLSTER VIA EMAIL! - You can receive this daily update every weekday morning via email! Just click here, enter your email address, and click “sign up.” That’s all there is to it (and you can unsubscribe anytime).

THURSDAY’S ‘OUTLIERS’ - Links to the best of news at the intersection of polling, politics and political data:

-The Trump campaign’s analytics team didn’t even think he would win. [Yahoo] 

-The economy and security were the top issues for voters according to an election day poll. [Morning Consult]

-The New York Times visualizes Trump’s reshaping of the electoral map. [NYT]

-Michael Tesler finds that Hillary Clinton reaped little benefit from group solidarity among women. [WashPost]

-Clare Malone digs into Clinton’s failure to win over white women. [538]

-Daniel Cox examines white Christian support for Trump. [PRRI]

-Larry Bartels sees 2016 as an “ordinary election,” not a realignment. [WashPost]

-The pollster behind USC/LA Times’ experimental tracking poll reflects on the outcome. [LAT]

-Steve Lohr and Natasha Singer write about the limits of data science. [NYT]