As is now his habit, Donald Trump took advantage of the press waiting for his arrival at his criminal trial in Manhattan on Monday to talk about the things he wanted to talk about. To wit: new polling from Siena College showing him leading in several battleground states — and maybe a new indicator that a blue state might be shifting purple.
End of carousel“I think we’re probably leading in New Jersey,” Trump told reporters. “We had a rally, over 100,000 people this weekend. A lot of the mainstream media didn’t want to say how many people. They didn’t want to cover it.”
Trump is referring to his rally in the beach town of Wildwood in the southern part of New Jersey on Saturday. You can read stories about it and the turnout in the mainstream media here, here, here, here and here, among other places.
But the important point here is the first part of his comment — that, because he got tens of thousands of people to show up at a rally, he is “leading” in New Jersey polling. In a state that he lost by 14 points in 2016 and 16 points in 2020. Because he got all those people (probably fewer than 100,000) to a city that’s no more than two hours from at least four other states, he has taken the lead. Who needs polls when you have people showing up at a rally six months before anyone can vote?
Advertisement
This is an argument that Trump has made for years. Over and over, we’ve heard him talk about the size of crowds at his rallies and inaugurations, insisting that these were useful proxies for his popularity. During the 2016 cycle, his allies repeatedly pointed to rally attendance to make a point similar to the one Trump offered about New Jersey. In 2020, Trump expressed amazement that he drew huge crowds while Joe Biden stayed home in his basement to avoid the pandemic. Biden didn’t actually stay home in his basement, but this rhetoric was later used to bolster the idea that somehow the election had been stolen.
When assessing Trump’s claims this time, though, we have an advantage over 2016: We know what ended up happening. We can look at how many people attended his rallies and where they were to see whether there was any demonstrable link.
In fairness, this is a high bar to set. “Demonstrable link” means being able to pick out that his rally crowds were correlated to how he fared. But picking out single factors in state-level election results is next to impossible. States — particularly battleground states — are complicated things with lots of voters making lots of decisions in lots of places for lots of reasons. That Trump held a rally cannot be easily disentangled from everything else.
Trump held more than 100 rallies during the 2016 general election. He won about 55 percent of the counties where those rallies were held. In the final two weeks, he held 36 rallies and won the host counties in 22 of them.
When we compare the results in counties where Trump held rallies with those in the surrounding counties — operating under his assumption that rally attendance represents popularity in that place — we see that he fared worse. During the general election, Trump did worse in most counties where he held rallies than he did in the surrounding counties. That’s true of the final two weeks, too; he did worse than in the surrounding counties for 19 of his last 36 rallies.
This isn’t proof that his rallies hurt, certainly. Just that it doesn’t tell us much.
Advertisement
We considered the scale of the rallies, too. The places where his turnout was largest relative to the county population were places where he did better in 2016 and where he outperformed surrounding counties. But this is mostly because they were in small, rural counties — places where relatively large crowds are easier to attain and where larger vote margins are easier to land.
There is other evidence that Trump didn’t do himself many favors. Analysis of the 2018 midterms found that Trump’s endorsement often hurt Republican candidates in the general election, with his unpopularity triggering more support for the Democratic candidate. When he stumped for Republican candidates in Louisiana and Kentucky in 2019, there was evidence of a similar effect. Other research found that his endorsements in 2018, 2020 and 2022 had either no effect in the general or probably hurt his endorsed candidates.
Comparing support for Trump to that for his endorsed candidates isn’t fair, of course. The whole point of Trump showing up at a rally, even for another candidate, is that he is personally popular. It’s just that being so popular you can get huge numbers of people to show up to a rally — let’s grant him the 100,000 in New Jersey, for example — doesn’t mean much when the state’s presidential election turnout is 45 times as large (as New Jersey’s was four years ago).
Advertisement
Some of the chatter about Trump’s crowd sizes is just chatter, meant to make a point by any means necessary. Longtime Trump adviser Roger Stone, for example, shared a photo of a large crowd by the ocean, taunting Biden by asking whether he could draw a similar crowd.
The answer, as people quickly pointed out in response, was that Trump probably couldn’t, either. The photo was from a 1994 Rod Stewart concert in Brazil.
The photo does raise an important question, though: Could Stewart have been elected presidente do Brasil three decades ago? Donald Trump seems to think so.
ncG1vNJzZmivp6x7uK3SoaCnn6Sku7G70q1lnKedZL2wuMitoJyrX2d9c4COaWxoaWNkwbPBzKlknKqfrLFuv8iznGaqlajCrcDSaA%3D%3D