Three theories as to why Trump’s key results don’t match expectations

It’s still early in the primary season, but there’s already a whiff of potential election error in the air.

That’s because Donald J. Trump has underperformed in the polls in each of the first three elections.

• In Iowa, the last FiveThirtyEight poll average showed Trump leading Nikki Haley by 34 points with a 53 percent share. He ultimately defeated her by 32 points and 51 percent. (Ron DeSantis came in 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 and 62 percent. He eventually won by 20 points with 60 percent.

In the scheme of the primaries, these are not particularly big misses. In fact, they are more accurate than average.

But with Trump doing well in early general election polls against President Biden, even a modest underperformance by Trump in the polls is worth some attention.

So what’s going on? We can’t say anything definitive based on the data we have, but three theories are worth considering.

One of these, described at the bottom, seems particularly plausible and aligns with something we’ve written about before: anti-Trump voters are highly motivated to break 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

A simple explanation is that undecided voters ultimately supported Mrs. Haley, the former governor of South Carolina.

This is plausible. Mr. Trump is a well-known candidate — even, in fact, a sitting president. If you’re a Republican who isn’t currently sure whether you support Mr. Trump, you’re probably not particularly inclined toward the former president. It’s easy to see how you might ultimately support his challenger.

It is also a theory that finds some support in voting patterns. Aside from Mr. DeSantis withdrawing from the race, which caused the voting pool to shift toward Mr. Trump, Mr. Trump’s support in early states was flat for the month or so before this election. During the same period, Ms. Haley tended to make gains — gains that were most easily attributed to undecided voters who rallied behind her.

That was even true in South Carolina, where she closed the gap somewhat in the last round of voting.

But while this theory could easily be part of the story, it is not a complete explanation. In addition to his lower margin of victory compared to pre-election polls, Trump was favored as a candidate behind his pre-election vote share, which cannot simply be attributed to undecided voters rallying behind Ms. Haley.

Theory No. 2: The Electorate

Another possibility is that the polls simply misrepresented the composition of the electorate. According to this theory, pollsters did a good job of measuring the people they wanted to measure, but they measured the wrong electorate. In particular, they did not include enough Democratic-leaning voters who turned out to support Ms. Haley.

It’s impossible to prove, but I think this is probably an important factor. It is always relatively difficult to predict the composition of the electorate in a presidential primary, but the large number of Democratic-leaning voters motivated to defeat Mr. Trump poses a particularly significant challenge this cycle. For the first time since 2012, there is no competitive Democratic presidential primary to draw away Democratic-leaning independents, 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 are good reasons to believe this is part of what’s going on.

For many pollsters, the problem has been there from the start: They don’t even interview previous Democratic primaries. Take, for example, the methodology of a Monmouth/Washington Post poll – one of the few polls that discloses their methodology in enough detail for accurate analysis:

The Monmouth University-Washington Post survey was conducted from January 26 to 30, 2024 among a probability-based sample of 1,045 South Carolina voters who have voted in at least one Republican primary since 2016 or have reregistered since the 2020 election and have not voted in a primary.

The decision to survey past Republican primaries is understandable — it makes the poll much cheaper and focuses on respondents most likely to vote — but it will obviously miss any past Democratic voters who choose to participate in a Republican primary .

To what extent is this a problem for pollsters? It could be a big one. The pre-election turnout estimates we used for our election night model — you might just know it as the Needle — assumed that 8 percent of the Republican primary electorate would be made up of former Democratic primary voters, those who would not qualify come for the Monmouth/Washington Post poll. It appears that this group has supported Ms. Haley.

That may seem like a lot to Democrats, but the final results suggest it may have been too low. In fact, these same pre-election turnout estimates unequivocally underestimated turnout in Democratic-leaning areas relative to Republican-leaning areas, indicating that turnout among Democratic-leaning voters was even more robust than expected.

The story was the same with our turnout projections in New Hampshire last month: turnout in Democratic areas was a lot better than we expected. And realistically, the same challenge could continue to haunt pollsters as long as primaries remain competitive, at least in open and semi-open primary states like South Carolina and New Hampshire.

Warning: Tuesday’s Michigan primary is also an open primary, although the campaign to vote “unfettered” in protest of the war in Gaza may give Democrats a good reason to vote in their own primaries.

There’s not much pollsters can do about this turnout problem. Many pollsters don’t have the money to survey the entire electorate for a low-turnout primary. Even if they survey everyone, they still have to conclude that these Democrats are likely to vote in a Republican primary, and I’m not sure that’s that easy to determine. If a pollster asks this, how many of these voters will actually say something like, “I’m pretty sure I’m going to vote in the Republican primary”?

This is an unusual decision for Democratic-leaning voters, but many appear to be making this decision.

Theory No. 3: A hidden Biden vote?

If you are a Democrat and are hoping that the polls underestimate Mr. Biden in the general election, then your best case scenario is that the polls are wrong because there is a Hidden Biden Vote, or at least a Hidden Anti-Trump Vote.

In this theory, the polls did well at modeling the electorate as undecided voters split between the candidates, but anti-Trump voters were simply not as likely to take surveys as pro-Trump voters. If this theory were true, the polls in the general election might underestimate Mr. Biden as much as they underestimated Ms. Haley.

There is no good way to prove (or disprove) this theory. Normally, non-response bias theories gain credibility through a diagnosis of exclusion: once other explanations have been ruled out, there remains the possibility that there is an undetected bias in the data. This is mainly because non-response theories usually lack clear evidence, which is also the case here.

The lack of evidence for non-response bias does not refute this. Far from it. But in this case, the turnout and undecided voter theories are so credible that there is no reason to believe there is non-response either.

And realistically, neither the undecided voter nor turnout theories would have much impact on the general election polls. There is 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 to pollsters posed by Democratic-leaning voters in open and semi-open Republican primaries also has no analogy to the general election.

There’s one reason the anti-Trump turnout could be relevant to the general election: It lines up with other data showing Biden has the edge among the most engaged voters. This could provide a small turnout advantage even in a general election. It could also mean that current polls among all registered voters slightly underestimate Mr. Biden compared to the smaller group of actual voters.

This would not mean that today’s polls vastly underestimate Mr. Biden, but it could make the difference in a close election.