Sen. Bernie Sanders (Vt.) talks about the need for a political revolution in the Democratic Party. He may bring it about. But another presidential candidate looks much better poised to bring about a political revolution. If Donald Trump wins the Republican nomination, he will fundamentally change the character of the party.
Recent polls suggest a Trump win is a real possibility. An average of polls in Iowa has Trump ahead of his nearest competitor by about six percentage points, and he is dominating in New Hampshire, South Carolina and nationally.
But traditional Republican voters are not the strongest drivers of this support. Trump has assembled a new and different coalition of supporters. They will either disband in failure — or usher in a new brand of Republicanism.
In some very interesting analysis, the Upshot’s Nate Cohn writes that Trump “is strongest among Republicans who are less affluent, less educated and less likely to turn out to vote. His very best voters are self-identified Republicans who nonetheless are registered as Democrats.”
So Trump appeals to new Republican voters, and he can bring them into the party if he can get them to the polls. But he also alienates traditional Republicans who reliably vote. These voters either support other Republican presidential candidates or remain undecided. In Iowa, many have coalesced behind Sen. Ted Cruz (Tex.). In other states and nationally, polls show they are deeply divided as to who should be the alternative to Trump.
So will these Republicans support a Trump candidacy should he win the nomination?
We conducted a new experiment to gauge whether, and how willingly, Republicans who do not support Trump would work for him. We did the same for other candidates, including Cruz, Trump’s stiffest current competition in Iowa, and Sen. Marco Rubio (Fla.), a Republican who has shown some recent momentum.
We didn’t just ask folks whether they would work for particular candidates. We actually tested their willingness to do it. Here’s how:
We invited participants to play a game. In each round, participants were paired with candidates for the Republican nomination. For each pairing, participants were asked to complete three relatively simple timed math problems. For each problem they completed correctly, they were told that $1 would be contributed to the presidential candidate with whom they were paired. Participants could skip any round without penalty. The order of candidate pairings was randomized for each participant.
Our design is based on methods developed to study the realignment of voters during the primary process. The data were collected online using a national survey of 1,872 Republican respondents drawn from Survey Sampling International (SSI) over a period of four days from Jan. 8 to 11.
With this method, it’s much harder for participants to deceive survey researchers, or themselves, about their intentions.
Who actually worked for Trump?
Trump supporters are nearly unanimous in their willingness to work for him; 97.9 percent did so. But only 55.58 percent of those who supported other Republican candidates agreed to complete math problems in aid of Trump. Uncommitted voters were even less likely to help Trump, with only 49.06 percent working on his behalf.
Who worked for Cruz?
Cruz supporters, like Trump supporters, worked for their candidate nearly unanimously, at 95.77 percent. This is where comparisons with Trump end. Participants who did not support Cruz were quite willing — or at least, 67.31 percent of them were — to work on Cruz’s behalf. That’s 11.73 percentage points more support than Trump attracted from other candidates’ supporters. Similarly, uncommitted voters were 14.46 percentage points more likely to work in aid of Cruz than in aid of Trump. Trump supporters were the least likely to work for Cruz, at 61.07 percent.
Who worked for Rubio?
As with the first two candidates, nearly all of Rubio’s supporters are willing to work to support him; 96 percent completed the math problems. Rubio sets himself apart from Trump and looks much like Cruz, with 71.35 percent of those who support other candidates and 61.64 percent of the uncommitted finishing the math problems for him. Once again, Trump supporters are the least likely to work to support Rubio, with only 57.63 percent doing so.
There is something different about Trump supporters, and about the Trump candidacy. If he wins, he is likely to bring in new Republican voters and shed old ones.
Trump may well win — but not if only the most reliable Republican voters caucus or vote in the primaries.