The Hidden Group that Won the Election for Trump: Exit Poll Analysis from Edison Research

By: Larry Rosin

“I don’t think there’s ever been two more unlikeable candidates,” said Michael Che during the Weekend Update sketch on Saturday Night Live this week.  “Not one time in this election have I heard anyone say: ‘You know what? I like them both.'”

The data from the Exit Polls conducted by Edison Research for the National Election Pool show Mr. Che to be correct – an extremely small portion of the voting public (only 2%) told our exit pollsters they had a favorable view of both.  While most voters did have a favorable view of one of the two major candidates – an astonishing 18% of the electorate told us they had an unfavorable opinion of both Hillary Clinton and Donald Trump.  And this is the group that won the election for Trump.

2016-exit-poll-data-favorable-unfavorable

The fact that nearly one-in-five voters who didn’t like either major candidate still came out to vote is pretty remarkable.  This number is double what we saw four years ago (9% were unfavorable to both in 2012) and nearly four times what we saw in the Bush-Kerry match-up of 2004 (favorability ratings were not asked in the 2008 exit polls).

As you might expect, if you had a favorable impression of one candidate and not the other, in virtually every case you voted for that one candidate.  So had those with a negative view of both candidates split evenly, Clinton would have won rather easily.  However, as the graph below shows, this “Neithers” group broke strongly to Trump 47% to 30%.

The story gets even more pronounced when we look at the states that swung the election to Trump.  In each of the cases in the table below, the votes gained by people who said: “I don’t like Trump but I’m going to vote for him anyhow” is greater than his total margin in these states.  In other words – it was the “Neithers” who pushed Trump over the top in these states and ultimately won him the election.

State % “Neithers” Trump Clinton
Wisconsin 22% 60% 23%
Pennsylvania 17% 56% 31%
Michigan 20% 50% 29%
Florida 14% 61% 24%
North Carolina 16% 62% 26%

The “Neithers” are more likely to be men (61%) and are more likely to be age 30-44 than in the younger or older age groups.  They are 78% white, as compared to the total electorate which is 70%.

One of the most intriguing aspects of the “Neithers” is that a significant portion of those who were unfavorable to both Clinton and Trump were favorable to President Obama.  Nearly half of those who didn’t like either of this year’s two major candidates do have a favorable impression of President Obama – and a significant portion of this group voted for Trump.

The 2016 election was unique in so many ways.  One distinguishing characteristic is just how many people had an unfavorable impression of both of the major party candidates.  To be sure, some of these people decided not to vote for either – Gary Johnson and Jill Stein combined for 19% of the vote among the “Neithers.”  However in the end, far more people who liked neither candidate chose Donald Trump and that provided him with his margin of victory in the battleground states.

Behind the Numbers: The 2016 National Election Exit Poll

By Gage - 2012 Electoral College map, CC BY-SA 4.0, Link

By Gage – 2016 Electoral College map, CC BY-SA 4.0, Link

On November 8, 2016, Edison Research engaged in the largest single-day research project in the world: Exit Polling the nation on behalf of the National Election Pool (NEP). A staff of over 3,000 exit poll interviewers, precinct vote return reporters, call center workers, and analysts all across the country helped us provide the sole record of who voted, and why. We collected, processed, and analyzed over 100,000 interviews in a 17-hour period to not only create that record, but also to provide the NEP with the guidance to make the right projections for their viewers and readers.

Whether you saw our name or not, over 71 million Americans watched the fruits of our labor on ABC, CBS, CNN, FOX, and NBC–and millions more have seen visualizations of our data in newspapers like the New York Times, the Washington Post, the Wall Street Journal, and USAToday. Indeed, the entire world was watching: Edison’s exit poll clients also included major international news networks.

This election was extraordinary in both the level of interest and the level of scrutiny on the changing nature of the pre-election polls. Many pre-election forecasts showed a high probability of a Clinton win, only to reverse those projections as the night went on. Because of the exit polls, our clients knew something was up – something possibly very significant – the moment they looked at their screens, and were able to plan their coverage accordingly.

Any kind of survey research involves making estimates of a population–the exit polls are no different–and reconciling those estimates with historical data, statistical models, and of course the “received wisdom” of the professional pollster. Throughout the day, we were able to analyze millions of fields of data in real time to spot potential “surprises” and provide insights to our network clients that allowed them to make accurate projections all night long. For instance, we were able to recognize early in the day that many states that were not presumed to be close would in fact be nail-biters until the end.

While the Presidential Election naturally received the most attention, Edison’s exit polls covered a total of 103 races–everthing from hotly contested Senate races, to consequential gubernatorial battles, to the legalization of marijuana in several states. Today you can see the results of our work all over the Internet: where there is credible data on who voted for these races, how they voted, and why, it came from Edison.

The humble profession of “pollster” has undergone a lot of criticism over the past several weeks. Some pundits argue that polls are no longer effective. We at Edison are proud of the fact that our work covered scores of races and initiatives on Election Day, and our clients did not make one mistake, issue one retraction, or make one incorrect projection. Our exit polls have proven once again to be the single most reliable source of information on who voted and why those voters made their decisions.

In fact, since we began serving the NEP in this role in 2003, Edison and its network clients have never made an incorrect projection based upon exit poll data. We are very proud of this, and of our role in our great democratic process.

Links to our work:

ABC: http://abcnews.go.com/Politics/election-2016-national-exit-poll-results-analysis/story?id=43368675

CBS: http://www.cbsnews.com/news/cbs-news-exit-polls-how-donald-trump-won-the-us-presidency/

CNN: http://www.cnn.com/election/results/exit-polls

NBC: http://www.nbcnews.com/storyline/2016-election-day/election-polls-nbc-news-analysis-2016-votes-voters-n680466

FOX News: http://www.foxnews.com/politics/elections/2016/exit-polls

New York Times:  http://www.nytimes.com/interactive/2016/11/08/us/politics/election-exit-polls.html

Washington Post: https://www.washingtonpost.com/news/post-politics/wp/2016/11/10/exit-polling-popular-vote-have-democrats-challenging-the-trump-mandate/

Election Polling - man speaking to people waiting in line

What Makes an Exit Poll Legitimate?

We here at Edison are proud to be the sole providers of exit polling data for the National Election Pool this Election Day. It’s a responsibility that we take very, very seriously indeed, because we know that our work will be studied and scrutinized for years to come. On election night, it will be our exit polling data that you will see when you watch the election night returns on television, and that work is the product of decades of expertise, care, and the unfailing commitment to excellence of my Edison colleagues.

For this election, we have been seeing a number of reports on social media and other outlets that various entities are considering doing their own exit polls on November 8th. While we certainly believe in fair and open elections, we also think it’s important to be able to ascertain the motives and methods of any exit polling effort.

If you vote on Election Day, you might be asked to take an exit poll. If it’s from us, you’ll see our logo, as well as the logos of all of the major U.S. news networks clearly identified on the materials being used by our pollsters. But if you don’t see this, that doesn’t necessarily mean that the exit poll isn’t legitimate. In Utah, for instance, the Utah Colleges exit poll has been administered by student volunteers from BYU and other universities since 1982, and has been a valuable resource both for academic study and for understanding the Utah electorate.

Other efforts, however, may not be quite so transparent. With that in mind, I asked Joe Lenski, Edison’s guru of all things exit polling, to come up with this list of how to tell if an exit poll is legit.

  • First and foremost, a legitimate exit poll should use proper sampling techniques – both to identify sample precincts and to identify sample voters within each precinct.
  • The exit poll interviewers should be rigorously trained and non-partisan.
  • The questionnaire should not have any biased or leading questions.
  • A hallmark of a legitimate exit poll is a guarantee of the respondents’ privacy, confidentiality and anonymity–an exit poll should be every bit as secret a ballot as the actual vote, and the pollster should not know the voters’ responses. No exit poll should require the respondent to publicly state their choices.
  • A properly executed exit poll accounts for absentee and early voters who will not be voting at election day voting locations, a practice that is becoming increasingly common and requires a great deal of care on the pollster’s part to properly model.
  • The methodology should also adjust for non-response bias by demographics that can be observed by the interviewer, such as age, race and gender. If a voter sampled for interviewing is missed for any reason (they refuse, for example) their demographic information is still captured observationally and these data are factored into the final weighting of the results.
  • Exit Poll interviewing should take place during almost the entire time the polls are open, in order to account for any time of day effects. Certain demographic groups might vote before or after work, for instance, while others might vote in the middle of the day–and it is important that any biases observed during those times be mitigated by coverage throughout the day.
  • Finally, any legitimate exit polling effort should be fully transparent about its methodology, questionnaire instrument and who is sponsoring or paying for the exit poll.

Now, chances are you won’t be asked to take an exit poll on Election Day. But if you are–keep the points above in mind as you participate in the voting process. And whether or not you take an exit poll, make sure you take the most important poll of all: vote on November 8th.

Infinite Dial 2016 - Political Following Via Social Media

One Quarter Of Americans 12+ Follow Politics On Facebook

Infinite Dial 2016 - Political Following Via Social Media

March is an exceptionally busy month for Edison:  not only are we hard at work on our upcoming Infinite Dial 2016 study with our partners at Triton Digital, but we are also the sole providers of Exit Polling to all of the major news networks. That’s right–for those of you who stayed up watching the results of things like last night’s Michigan primary, we were behind the scenes all night, providing valuable information on who voted, why they voted, and helping our network clients make the most accurate projections possible.

So, we thought it might be interesting to “cross the streams” a little, and mix a little politics into the Infinite Dial. Here’s one small “tease” of some of the information we’ll be rolling out as the year goes on: what social networks do people use to follow political news and politicians?

Well, as you can see from the graph above, the top answer with the general population should be no surprise: it’s Facebook, with 25% of Americans 12+ saying they use the popular social network to get information about politics. Twitter is a solid second for political information, which is significant in that Twitter isn’t the second most popular social networking service (more on that in The Infinite Dial.)

This graph is taken from The Infinite Dial 2016, and is based upon a representative sample of 2001 Americans 12+, conducted via mobile phone and landline, to the most meticulous standards in research. These data, like all Infinite Dial data, are projectable to the U.S. population.

For more on podcasting, online radio, social media and more, sign up for The Infinite Dial Webinar, Thursday, March 10th, at 2PM. Attendance is strictly limited to 1,000, and it will fill up–so sign up now, and get there early!

Register today for The Infinite Dial 2016.

Edison Research Successfully Conducts Entrance Polls For The 2016 Iowa Caucus Amidst Record Turnout

On Monday, Feb 1 2016, Edison Research successfully conducted entrance polling for the Iowa Caucuses on behalf of the National Election Pool (NEP). Edison interviewed 3454 voters at randomly-selected Republican and Democratic caucus sites. The Republican turnout was approximately 187,000 people, a record for the Iowa Caucus, and the Democratic turnout was 171,000. Senator Ted Cruz won the Republican contest, while Hilary Clinton and Bernie Sanders continue to be in a statistically close race for the Democrats.

Edison’s entrance polls were used to determine key information from voters, including demographic data, important issues and the “electability” of the various candidates. Edison’s election team captured, processed and analyzed thousands of data points within the short duration of the caucuses and enabled our member clients and subscribers real-time access to in-depth analysis of the Iowa results.

Full coverage of our entrance polls can be found here:

NBC – http://www.nbcnews.com/politics/2016-election/primaries/IA

CBS – Dem – http://www.cbsnews.com/elections/2016/primaries/democrat/iowa/exit/
Rep – http://www.cbsnews.com/elections/2016/primaries/republican/iowa/exit/

CNN – Rep – http://www.cnn.com/election/primaries/polls/IA/Rep
Dem – http://www.cnn.com/election/primaries/polls/ia/Dem

New York Times

Dem – http://www.nytimes.com/interactive/2016/02/01/us/elections/iowa-democrat-poll.html?action=click&pgtype=Homepage&clickSource=story-heading&module=span-abc-region&region=span-abc-region&WT.nav=span-abc-region
Rep – http://www.nytimes.com/interactive/2016/02/01/us/elections/iowa-republican-poll.html?action=click&pgtype=Homepage&clickSource=story-heading&module=span-abc-region&region=span-abc-region&WT.nav=span-abc-region

Washington Post – https://www.washingtonpost.com/graphics/politics/2016-election/primaries/iowa-entrance-poll/