Think about Sunday, October 23, 2016. Like many of you, it was inconceivable to me that Donald Trump could become President-elect in just two weeks. That afternoon, CNN blared this headline: New poll shows Clinton over Trump by double-digits. By election day, the Real Clear polling average showed Clinton leading Trump by 3.2% The final results: Clinton: 48.2% - Trump: 46.1%, a 2.1% win. That’s within the margin of error, but remembering that CNN headline, many Americans thought it was an enormous miss.
Electoral College Result: Trump 306 - Clinton 232
Let’s look at 2020. This time from Pew Research, a late-cycle “poll of polls” showed Joe Biden at 53% of the vote and Donald Trump at 41% of the vote. The final results four years ago: Biden 52.3% - Trump 46.9%, a victory of four-and-a-half points - not within the margin of error. Yet we think of the polls in 2020 as closer, since they at least got the winner correct.
Electoral College Result: Biden 306 - Trump 232
What Happened in ‘16?
If a poll was taken of likely voters who were asked, “What was the polling error in 2016?” The number one answer would be that Donald Trump outperformed the polls. Countless articles and books have been written on the data and questions surrounding that opinion. While the result in 2016 was within the margin of error nationally, there is plenty of evidence to indicate that many voters were reluctant to tell a stranger on the telephone that they planned to vote for Donald Trump. To add to the surprise, even after the “Access Hollywood” tape, late-deciding voters most definitely broke for Trump and against Hillary Clinton.
Against Hillary Clinton - that phrase deserves elaboration. As strongly as some voters admired Secretary Clinton’s intelligence, preparation, and policy positions, there were millions of others who, rightly or wrongly, resented her and held low opinions of her. Those voters were looking for a reason to vote against her and it didn’t have to be a very good one. The argument that her negatives were media-driven rather than fact-driven doesn’t change the fact that a lot of people simply disliked her. The final result can be explained between that and the undercount of Trump voters.
What Happened in ‘20?
Pew Research analyzed the difference between the twelve-point lead anticipated in the days before the election and the final result. Did we once again undercount Trump voters? They give an example to illustrate the challenge of accurate polling in Presidential elections. See the charts below from Pew:
Every polling organization makes decisions about which 1000, or 1200, or 5000 polling respondents to include in any given poll. They make assumptions based on the reported age, party affiliation, and demographics that they think will match the electorate of that year. Miss by 38 voters of a thousand (3.8%) and you get the 2020 polling error.
Yes, there is some truth to the continued, “Trump undercount.” There is also truth to the notion that pollsters wrongfully predicted that the 2020 electorate would duplicate that of 2016. One occupational hazard of the polling business is trying to measure the wrong electorate - last time’s rather than this time’s.
Another factor in predicting which voters will successfully cast a ballot is the level of enthusiasm a voter has for a candidate, issue, or the election in general. Every voter talking to a pollster is asked about their anticipated likelihood of actually showing up. Is it possible that Trump voters in 2020 were more motivated than those planning to vote for Biden to actually cast a ballot? Many people think so.
Which Explanation Will Explain the Inevitable 2024 Polling Error?
I have four potential errors in mind for this cycle, plus one that I think could happen in reverse. Any one of these, were it to occur, could flip the results from the last-minute poll results.
Trump Voter Undercount (again)
We’ve seen a version of this twice. Couldn’t this be the case again this year? Maybe, but I doubt it. Here’s why: We know that some voters were reluctant in 2016 to admit that they planned to vote Trump. We know that, in 2020, core Trump voters were hell-bent to show up at the polls. Although a strong MAGA voter is probably well-motivated to vote again this year, there have been cracks in the Trump base. We discussed that in my article titled, “A Cracking Cult”, a few weeks ago. While there are some new members, some traditional Republicans are peeling away. Here is my theory: For every Liz Cheney who speaks out publicly about her break-up with Donald Trump, there are five more Republicans who have made the same decision; but kept it private — even to their families and friends.
What does that mean? This year, there is a real possibility for a Trump OVER-count as formerly loyal Trumpists quietly decide not to vote for him again. How many voters match this? We’ll find out. What we do know is that some former Trump voters have strong incentive to keep their changed-minds secret, away from the pressure of the MAGA faithful.
We Aren’t Ready for a Woman President
We have seen a movie about this once already. At the same time, social values have evolved in eight years. Most new voters compared with 2016 are younger, and even more likely to support the idea of a female President. Yet there is another factor to consider. Listen to Notre Dame political science professor Christina Wolbrecht:
“Most Americans are well-trained,” Wolbrecht explains. “They know to say that they would vote for a woman president. One of the challenges that scholars have is determining exactly how true that is.”
There is only one direction that this particular polling error can fall, and that direction favors Donald Trump.
The Dobbs Effect
Ever since the Supreme Court decided Roe v. Wade in January 1973, the politics of abortion have been established and predictable. Republicans wanted to ban abortion and went to the polls over and over to vote that way. Democrats who supported a woman’s right to choose were sometimes reluctant to campaign on the issue. But the American opinion on the matter was not set in stone. Poll after poll showed that a solid and growing majority of Americans believed in the rights established in Roe. Even Bill Clinton once famously proclaimed that abortion should be, “Safe, legal, and rare.”When POTUS overturned Roe, the script was flipped. Now, Democrats and young voters are strongly motivated. Put more succinctly, millions of voters were pissed. In state after state, when ballot measures appeared asking voters if, in their state, abortion rights should be codified, the answer has always been yes.
At the time of the 2022 midterms, this energy was palpable and led to Democratic wins in a wide range of races. From Governorships that weren’t close (think Gretchen Whitmer) to several house races, candidates who loudly supported women’s rights performed better than polling suggested.Is 2024 going to be another example of how motivated a wide swath of Americans are about the stripping away of a 50-year right? Will abortion-rights voters again decide to support Democrats up and down the ballot as a signal that they believe the GOP is extreme on this issue? I believe the answer is a resounding yes. Even though Americans have short political memories, this issue is different. Women are dying as Trump abortion bans have swept through state legislatures, creating chaos and confusion. This is a huge advantage for the Harris-Walz ticket. Remember, an abortion rights voter may well have stayed home in both 2016 and 2020 (before Dobbs) and yet be strongly motivated now.
New VotersYoung conservatives have indeed registered in large numbers since 2020. They may be children of MAGA parents or have some other driving ideological source, but they are there. Gen Z has already shown itself to be a vigorous group of socially active citizens. This year, over 41 million will be eligible to vote including 8.8 million more than in 2022. Add to those sheer numbers the fact that nearly half of this group are voters of color. What do we know about the combination of the young and voters of color? We know that they are socially liberal as a group. We know that they aren’t afraid to let their voices be heard.
What we don’t know … what pollsters don’t know, is how many will turn out. Young voters are notoriously easy to motivate at a rally and hard to get registered and to do what’s necessary to cast an actual ballot. What clues do we have about these challenges in 2024?
Vote.org, one of the premier voter registration organizations in the country, reported that, on National Voter Registration Day in September, voters under 35 made up 81% of all registrations. 11% on that day were exactly 18 years old, up 53% from the same day four years ago. That little endorsement from Taylor Swift resulted in almost 350,000 visits to voter registration sites from one social media post alone. Yes, young voters get very fired-up at rallies and always have. My bet is that, in this election, more than expected are going to actually vote. And Vice President Harris is going to overwhelmingly win their votes.
Bottom Line
This election is going to be close. Probably. Not necessarily as close as it looks now. There is going to be a “Reverse Trump undercount”. We just don’t know how big. There is going to be an “Unready for a woman,” polling error, but how big? The Dobbs Effect is still potent enough to make a difference. And the ever-unpredictable younger vote will surely be sufficient to flip a state or two.
If you’re nervous, that’s a good sign. We all should be nervous. And motivated. And doing whatever we can, starting with making an honest-to-God plan for our own vote. Find out the details of your early voting options. Bring someone with you. Do what you can. We’ll sleep when we’re dead. As they say at Vote Save America, LFG!
Share
