Gov. Ron DeSantis of Florida and President Trump in July 2020, when they were working together. Credit…Al Drago for The New York Times
Is Donald J. Trump the clear favorite and front-runner to win the Republican nomination? Or is he badly weakened and even an underdog against Ron DeSantis?
At the onset of the Republican campaign, the polls are exceptionally divided on Mr. Trump’s support among Republican primary voters.
In national surveys since last November’s midterm election, different pollsters have shown him with anywhere between 25 percent and 55 percent of the vote in a multicandidate field.
That’s right: a mere 30-point gap.
In just the last two weeks, an Emerson College poll found Mr. Trump leading Mr. DeSantis by 26 points, 55 percent to 29 percent, in a multicandidate field, while a Bulwark/North Star/Dynata poll over a similar period found Mr. DeSantis leading by 11 points, 39 percent to 28 percent.
This is not normal. It’s also a recent development. In the three months before the midterm election, 10 polling firms showed a much more typical 12-point spread in Mr. Trump’s share of support, between 45 percent and 57 percent.
Whether Mr. Trump is at 25 percent or 55 percent is no small matter. Believe it or not, early polling is fairly predictive of the eventual outcome in presidential primaries. It also has real-world consequences. It affects the decision-making of potential candidates, operatives and activists, many of whom have adopted a wait-and-see approach in part because there are so many conflicting signs of Mr. Trump’s strength.
And the existence of such a wide split betrays that the survey research industry may be in far worse shape than one might have otherwise guessed. While the exact reason for the vast spread in survey results is hard to ascertain, the likeliest explanation is that many well-known pollsters are collecting profoundly unrepresentative data.
Although there’s not a clear picture, a rough pattern in the data might hint at the actual state of the race. Higher-quality surveys havetended to show far less support for Mr. Trump.
If that’s confirmed by additional higher-quality polling as primary season gets underway, Mr. Trump is no front-runner.
Why such a large gap? (Wonkiness rating: 6/10)
There are many reasons polls can disagree, but most of the usual explanations don’t add up to a huge 30-point gap:
It’s not about telephone versus online polling. Almost all of the polls have been conducted online, so the difference can’t be attributed to a so-called mode effect — like the possibility that Mr. Trump’s supporters won’t divulge their preference to a live interviewer.
The way pollsters define the Republican primary electorate (say: self-identified Republicans versus people who say they’ll vote in a Republican primary) doesn’t explain what’s happening, either. An analysis of New York Times/Siena polling last fall suggests that these choices do have effects, but that they are fairly modest on the scale of the 30-point gap in question. And multiple pollsters with similar definitions of the Republican primary electorate nonetheless show fundamentally different races.
The gap persists regardless of the number of Republican candidates listed by the pollster — including in head-to-head polling between Mr. DeSantis and Mr. Trump.
It’s probably not weighting, the statistical adjustments made by pollsters to ensure a representative sample. Most of the polls are weighted by roughly the same set of demographic characteristics, including by self-reported education.
It’s probably not the timing, but I’ll hedge a little bit on this one. Most of Mr. Trump’s worst polls were conducted in November and December, after the midterm election. That’s potentially relevant because the disappointing Republican showing in the midterms is the likeliest explanation for the apparent decline in Mr. Trump’s support. On the other hand, none of those pollsters have returned to the fray, and most of the polls that have conducted multiple surveys since the midterms have shown no change or no loss of ground for Mr. Trump in this period.
If it’s not the mode, the population, the timing, the question or the weighting, there’s really one explanation left: the sample itself. For some reason, some pollsters are getting a vastly more Trump-friendly group of Republican respondents than others.
Or, to be more blunt about it: Someone’s data could be extraordinarily and unacceptably inaccurate — inaccurate to a degree we would have never guessed until pollsters started asking about a new race.
So which pollsters are right?
It’s really hard to tell which of these polls might be “right” or “wrong.” There are countless ways to collect survey data online and, in general, there’s very little transparency about the process. Even when there is transparency, there aren’t well-established best practices that make it easy to evaluate whether a given approach is a sound one.
But there are two reasons to err toward the polls that are showing Trump weakness.
First, the so-called probability polls have uniformly showed relative weakness for Mr. Trump.
Probability sampling is where the respondents are more or less recruited at random, such as by calling random telephone numbers or by sending a mail invitation to random addresses to participate in an online poll. It’s traditionally considered the gold standard in survey research. A nonprobability sample, by contrast, isn’t selected at random. It might instead be recruited from banner ads on certain websites.
The five probability samples — from Ipsos, Suffolk, Monmouth, Quinnipiac and Marquette Law (fielded by SSRS) — tend to give Mr. Trump relatively bad news. Ipsos and Monmouth found him trailing Mr. DeSantis with just 25 percent and 26 percent of support in a multicandidate field. Suffolk University and Marquette Law/SSRS found Mr. Trump at just 36 percent and 33 percent in a one-on-one matchup (and did not ask a multicandidate question).
The fifth probability poll, Quinnipiac, found Mr. Trump faring somewhat better, with 43 percent in a one-on-one matchup with Mr. DeSantis. But this 43 percent figure is only among self-identified Republicans, who are more supportive of Mr. Trump than the broader group of Republican primary voters in Times/Siena data and other polling. Mr. Trump would probably fall into the upper 30s if the result also included Republican-leaning independents, like most other polls.
The nonprobability polls, on the other hand, are all over the place. The two wildly divergent polls cited in our introduction, from Emerson and Bulwark, are both nonprobability surveys.
This basic pattern — relative agreement among the probability surveys and variance among nonprobability polls — is consistent with the possibility that the problem is on the sampling side.
Nonprobability polls have produced solid results under many circumstances, but that kind of polling is very hard. Recruiting a representative sample from unrepresentative sources is a self-evident challenge. And then there are recurring data quality concerns, from panel conditioning to bots. Anecdotally, many pollsters believe these issues are getting worse with time, but it’s hard to evaluate from the outside. And it’s just as hard to evaluate which firms do a good or bad job of handling these issues.
Whatever the explanation for the variance among the nonprobability shops, the consensus among probability polls suggests that the polls showing a relatively weak Mr. Trump are closer to the truth.
Second, the state polling is almost entirely consistent with a weak or relatively weak Trump.
Nearly every nonpartisan state poll shows him running a few points worse than his performance in the same state in the 2016 primary. Most of those state primaries were held at a time when Mr. Trump’s national support was in the mid-30s, suggesting he sits in that range or a little lower today.
If we exclude the unusual cases of Florida and Utah (Florida because Mr. DeSantis is its governor; Utah because Mr. Trump had such vanishingly low support there in the 2016 caucus), Mr. Trump is underperforming his 2016 vote share by an average of four points in polls with a one-on-one matchup with Mr. DeSantis and by 10 points in multicandidate state polls.
That would imply that Mr. Trump has around 30 percent of the vote in a multicandidate field and 35 percent in a one-on-one race with Mr. DeSantis.
The usual caveat applies that it’s early, and news events and campaigns could reshape things. And to be clear, this is a very rough inference. There are only a half-dozen nonpartisan state polls at this point, and it is always hard to divine national estimates from state polls. But no matter how you look at these data points, they’re on the bad side of the ledger for Mr. Trump.
And several of these state polls — Siena, Vanderbilt/SSRS, University of New Hampshire — are well-regarded surveys with solid track records and a reasonable methodological foundation. When you combine them with the national probability polls, that’s a sizable if not quite dispositive amount of higher-quality data on the side of a relatively weak Mr. Trump.
We probably won’t be sure until there’s a large enough wave of rigorous national polls — think ABC/Post, Times/Siena (sorry, won’t be soon) — to calibrate our expectations for the race. In the interim, the assessment that Mr. Trump is stuck in the low-to-mid 30s nationwide represents a best guess.