Not So Fast on the New Poll That Says Bernie Can Beat Trump

“Socialist” is still a problem.

Bernie Sanders laughing while holding a mic
Sen. Bernie Sanders during a campaign stop in Iowa City, Iowa, on Saturday. Mike Segar/Reuters

Would people who don’t like socialism vote for Sen. Bernie Sanders against President Donald Trump? That’s a big question for Iowa Democrats as they prepare to caucus on Monday night. Socialism is consistently unpopular in polls. Even many Democrats say they wouldn’t vote for a socialist. If that makes Sanders unelectable, Iowans would be foolish to propel him toward the nomination.

A new poll taken by Data for Progress, a left-leaning organization, casts doubt on this criticism of Sanders. DFP hasn’t posted the survey or its methodology, but it released some of the data to Matthew Yglesias, my former Slate colleague. In a writeup in Vox, Yglesias presents the numbers with care and circumspection, reading them as evidence that Sanders could overcome the socialist label. I read them differently. The DFP poll tells us more about Trump than about Sanders. And it fits the theory that Democrats are better off nominating a different candidate.

The survey, taken between Jan. 9 and Jan. 19, posed three versions of a standard ballot test. One question asked how you’d vote “if the candidates were Bernie Sanders and Donald Trump.” The second version added partisan cues, asking how you’d vote “if the candidates were Democrat Bernie Sanders and Republican Donald Trump.” The third version asked how you’d vote if the candidates were “Democrat Bernie Sanders, who wants to tax the billionaire class to help the working class,” and “Republican Donald Trump, who says Sanders is a socialist who supports a government takeover of healthcare and open borders.”

The three versions yielded similar results. In the first scenario, respondents chose Sanders over Trump, 47 percent to 41 percent. (Vox doesn’t say whether the respondents were adults or registered voters, but DFP has generally polled registered voters.) In the second scenario, they chose Sanders by a narrower margin, 45 percent to 43 percent. It’s possible that partisan cues helped Trump or hurt Sanders. It’s also possible that the difference was just random variation.

On the third version of the question—the one that mentioned socialism—respondents still chose Sanders, 47 percent to 42 percent. That’s almost identical to the first version. From this, Yglesias infers that “affixing that label to Sanders doesn’t really shift polling at all.” In its headline, Vox declares, “Bernie Sanders leads Donald Trump in polls, even when you remind people he’s a socialist.”

That’s a comforting conclusion. But it doesn’t quite match the survey. The question didn’t ask whether people would vote for Sanders when you remind them that he’s a socialist. It asked whether they’d vote for Sanders if Trump called Sanders a socialist. In the latest Fox News poll, 58 percent of voters said Trump wasn’t honest or trustworthy. Only 38 percent said he was. So when you tell respondents that Trump has called somebody a socialist, most of them won’t assume that person is a socialist.

Yglesias has a theory to explain why the “socialist” version of DFP’s question produced no effect. “Republicans have been characterizing Democratic Party support for higher taxes and a more generous welfare state as ‘socialism’ for a long time,” he notes. This “may have somewhat deadened the argument.” In a subheadline, Vox puts this point in colloquial terms: Republicans are “the political party that cried wolf.”

That’s a good theory. It implies that Trump would have trouble persuading voters that a self-styled progressive pragmatist, such as former Vice President Joe Biden, Sen. Amy Klobuchar, or former Mayor Pete Buttigieg, is a socialist. It would be Trump’s word—which most people don’t trust—against the word of the accused candidate. But if Sanders were the nominee, the claim would be conceded. Sanders calls himself a socialist. When the wolf says he’s a wolf, you don’t need to trust the shepherd boy.

In this way, the DFP survey fits the larger case against Sanders. If voters discount Trump’s allegations of socialism, it’s possible—but unlikely, given the findings of other polls—that they don’t care about socialism. It’s also possible that they dislike socialism but prefer Sanders anyway. But the simplest explanation is that they don’t believe Trump. They’re not going to accept that the Democratic nominee is a socialist unless he says so himself.

Sanders doesn’t just embrace the label. He backs it up with a long history of Marxism. And he’s the only major candidate who insists on abolishing private health insurance and replacing it with a government monopoly. (Sen. Elizabeth Warren, who initially shared his position, has since deferred the idea of taking over the health insurance market.) If you want to test how Sanders would fare in a general election, you can’t just tell voters that Trump calls Sanders a socialist. You have to tell them what Sanders has said and done.

Yglesias believes that Sanders’ performance as a candidate in Vermont supports an optimistic reading of the DFP survey. “Sanders consistently does a bit better in elections for his Senate seat than you would expect from the state’s baseline party lean,” he observes. But winning in Vermont is very different from winning a presidential election. According to Pew data, Vermont is 35 percent liberal, 32 percent moderate, and only 30 percent conservative. The United States, as a whole, leans the other way: 36 percent conservative, 33 percent moderate, and only 24 percent liberal. Wisconsin is 41 percent conservative, 36 percent moderate, and only 19 percent liberal. A minor liability in Vermont can be fatal in Wisconsin.

The good news for Sanders is that DFP’s survey dispels the worst-case scenario. The poll could have found that if Trump were to call Sanders a socialist, most people would vote against Sanders. It didn’t. Trump’s accusation, by itself, isn’t decisive. The question for Democrats is whether to nominate a candidate who will concede that accusation—or one who will fight it.