Table of Contents
It’s still early in the primary season, but a whiff of a possible polling error is already in the air.
That’s because Donald J. Trump has underperformed the polls in each of the first three contests.
In Iowa, the final 538 polling average showed Mr. Trump leading Nikki Haley by 34 points with a 53 percent share. He ultimately beat her by 32 points with 51 percent. (Ron DeSantis took second.)
In New Hampshire, he led by 18 points with 54 percent. In the end, he won by 11 points with 54 percent.
In South Carolina, Mr. Trump led by 28 points with 62 percent. He ultimately won by 20 points with 60 percent.
In the scheme of primary polls, these aren’t especially large misses. In fact, they’re more accurate than average.
But with Mr. Trump faring well in early general election polls against President Biden, even a modest Trump underperformance in the polls is worth some attention.
So what’s going on? We can’t say anything definitive based on the data at our disposal, but three theories are worth considering.
One of them, described at the bottom, seems especially plausible and consistent with something we’ve written about before: Anti-Trump voters are highly motivated to turn out this cycle. It wouldn’t mean the polls will be wrong in November, but it would be good news for Democrats nonetheless.
Theory No. 1: Undecided voters
One simple explanation is that undecided voters ultimately backed Ms. Haley, the former South Carolina governor.
This is plausible. Mr. Trump is a well-known candidate — even a de facto incumbent. If you’re a Republican who at this point doesn’t know if you support Mr. Trump, you’re probably just not especially inclined toward the former president. It’s easy to see how you might end up supporting his challenger.
It’s also a theory with some support in the polling patterns. Other than Mr. DeSantis dropping out of the race, which led that voting group to shift toward Mr. Trump, Mr. Trump’s support in the early states was flat over the month or so before these elections. Over the same period, Ms. Haley tended to make gains — gains most easily attributed to undecided voters coalescing behind her.
That was even true in South Carolina, where she closed the gap somewhat in the final round of polls.
But while this theory could easily be part of the story, it’s not a complete explanation. Besides his lower margin of victory compared with pre-election polling, Mr. Trump tended to run behind his pre-election vote share, which can’t simply be attributed to undecided voters coalescing behind Ms. Haley.
Theory No. 2: The electorate
Another possibility is that the polls simply got the makeup of the electorate wrong. In this theory, pollsters did a good job of measuring the people they intended to measure, but they were measuring the wrong electorate. In particular, they did not include enough of the Democratic-leaning voters who turned out to support Ms. Haley.
It’s impossible to prove, but I think this is probably a major factor. It’s always relatively hard to predict the makeup of the electorate in a presidential primary, but the large number of Democratic-leaning voters motivated to defeat Mr. Trump is a particularly great challenge this cycle. For the first time since 2012, there’s no competitive Democratic presidential primary to draw Democratic-leaning independents away, and the Republican runner-up is a relative moderate who may be palatable to many Democratic-leaning voters.
We don’t yet have turnout data on how many Democratic-leaning voters actually participated in these primaries, but there’s good reason to believe this is part of what’s going on.
For many pollsters, the problem is baked in from the start: They don’t even interview prior Democratic primary voters. Take, for instance, the methodology of a Monmouth/Washington Post poll — one of the few polls to disclose their methodology in sufficient detail for close analysis:
The Monmouth University-Washington Post Poll was conducted from Jan. 26 to 30, 2024, among a probability-based sample of 1,045 South Carolina voters who have voted in at least one Republican primary election since 2016 or have newly registered since the 2020 election and not voted in a primary.
The decision to survey prior Republican primary voters is understandable — it makes the poll much cheaper and homes in on the respondents likeliest to vote — but it will obviously miss any previous Democratic voters who hadn’t voted in a Republican primary and now choose to do so.
How much of a problem for pollsters is this? It could be a big one. The pre-election turnout estimates we used for our election night live model — you may know it simply as the Needle — supposed that 8 percent of the Republican primary electorate would be composed of former Democratic primary voters who hadn’t previously voted in a Republican primary, those who wouldn’t be eligible for the Monmouth/Washington Post poll. That group seems likely to have backed Ms. Haley.
That might seem like a lot of Democrats, but the final results suggest it might have actually been too low. In fact, these same pre-election turnout estimates unequivocally underestimated the turnout in Democratic-leaning areas relative to Republican-leaning areas, suggesting that the turnout from Democratic-leaning voters was even more vigorous than projected.
The story was the same with our turnout projections in New Hampshire last month: The turnout in Democratic areas was quite a bit better than we expected. And realistically, the same challenge could keep dogging pollsters as long as the primary stays competitive, at least in open and semi-open primary states like South Carolina and New Hampshire.
Warning: The Michigan primary on Tuesday is also an open primary, though the campaign to vote “uncommitted” to protest the war in Gaza may give Democrats a good reason to vote in their own primary.
There’s not a lot pollsters can do about this turnout problem. Many pollsters don’t have the money to survey the whole electorate for a low-turnout primary. Even if they do survey everyone, they still need to conclude that these Democrats are likely to vote in a Republican primary, and I’m not sure that’s so easy to determine. If asked by a pollster, how many of these voters will really say something like “I’m almost certain to vote in the Republican primary”?
This is an unusual decision for Democratic-leaning voters to make, but many appear to be making it.
If you’re a Democrat hoping that the polls are underestimating Mr. Biden in the general election, your best-case scenario is the polls are wrong because there’s a Hidden Biden vote, or at least a Hidden Anti-Trump vote.
In this theory, the polls did well in modeling the electorate while undecided voters split between the candidates, but anti-Trump voters simply weren’t as likely to take surveys as pro-Trump voters. If this theory were true, then the general election polls might be underestimating Mr. Biden by just as much as they’ve underestimated Ms. Haley.
There’s no great way to prove (or disprove) this theory. Usually, nonresponse bias theories gain credibility through a diagnosis of exclusion: Once other explanations are ruled out, then we’re left with the possibility that there’s an unobserved bias in the data. That’s mostly because nonresponse theories usually don’t have any clear evidence on their behalf, which is the case here as well.
The absence of evidence for nonresponse bias doesn’t disprove it. Far from it. But in this case, the turnout and undecided voter theories are credible enough that there isn’t reason to assume any nonresponse bias either.
And realistically, neither the undecided voter nor the turnout theories would have much bearing on general election polling. There’s no reason to expect that voters undecided between Mr. Trump and Mr. Biden will break toward Mr. Biden, at least not for the reason that undecided Republicans might break toward the newcomer Ms. Haley. The unusual turnout challenge for pollsters posed by Democratic-leaning voters in open and semi-open Republican primaries doesn’t have any analogy to the general election either.
There is one reason the anti-Trump turnout might have relevance for general election polling: It’s consistent with other data showing Mr. Biden with the edge among the most highly engaged voters. This could yield a slight turnout advantage, even in a general election. It may also mean that the current polls of all registered voters slightly underestimate Mr. Biden compared with the narrower group of actual voters.
This wouldn’t mean the polls today are vastly underestimating Mr. Biden, but it could make the difference in a close election.
