Last week a survey from Washington-based Mercury Analytics surfaced with a most curious finding: Roughly 20 percent of Democrats said they would defect from their party to vote for Donald Trump, with a surprising number of them declaring with 100 percent certainty they would be comfortable doing so. A smaller chunk of Republicans (14 percent) said they would switch to vote for Hillary Clinton. 

The survey ignited a firestorm of intrigue. Are Democrats really so taken with Trump? Given Trump’s extremist boosters (most recently an openly white-supremacist super PAC), the ready crossover seemed especially confusing. How could the right-wing candidate with some of the most extreme views in the race attract voters who ostensibly lean toward the center or left?

As The Washington Post’s Philip Bump points out, there’s probably not as much in the Mercury Analytics poll to get excited about as there initially seemed. But the prospect of crossover voting between the parties still deserves attention, especially in the unique cases of Trump and Bernie Sanders, whose mutual appeal to many of the same voters has been well noted. “Voters who were on the fence between the seemingly polar opposite candidates said both communicated well with working-class people and made strong cases for how they would improve the economy,” The New York Times reported from Vermont last week, observing the odd phenomenon of voters who consider the two candidates quite comparable. In the words of one voter: “Bernie is my No. 1 choice, and Trump is No. 2. They’re not that different.”

That voter’s view is not ill represented in the media. Despite the fact that Sanders and Trump are worlds apart politically, they have suffered endless comparisons, with some of the more ambitious parallels suggesting the two candidates are political twins. If they were similar in terms of policy, that could account for why they sometimes capture the interest of the same voters. And yet they aren’t, as I recently argued, in agreement on their major policies: The conclusion that they are can only be reached by cherry-picking issues they agree on along with several more mainstream candidates, or by abstracting a policy issue to the point that important distinctions are buried under general categories. 

So if matching policies aren’t attracting voters to the Sanders-or-Trump binary, what is? Why don’t the vast policy differences between Sanders and Trump result in totally distinct bases of interest? 

One explanation is that Americans aren’t necessarily well-informed when it comes to policy particulars. A 2010 Pew Forum survey on policy and the public found that only 14 percent of Americans knew the current inflation rate; less than half knew which party held the majority in the U.S. House of Representatives; and only 16 percent knew that more than half of loans made under TARP had been paid back. Meanwhile, a 2012 Pew Forum poll on last cycle’s presidential campaigns found that just under half of voters knew what a super PAC was, and only 39 percent could identify John Roberts as the current Chief Justice of the U.S. Supreme Court. In other words, the technical policy differences that distinguish candidates from one another on paper might not always be entirely clear to voters, meaning their interest in various candidates shouldn’t automatically suggest that they view candidates’ policies as similar. 

At the same time, voters are pretty well-versed in the other side of the equation: their own problems. These, after all, are what savvy politicians aim to address. And this is where Sanders and Trump converge, and where they could compete fiercely for crossover votes. Rather than offering the same solutions, Trump and Sanders tend to speak strongly to the same problems, which captures voters who feel their grievances are rarely addressed by mainstream candidates in either party.

Both speak in terms of decline, especially the decline of the American middle class, of workers and everyday sorts who have seen their status and stability dip over the past decades. “Here is the reality of the American economy,” Sanders wrote in a June 2015 Boston Globe op-ed, “Despite an explosion in technology and a huge increase in worker productivity, the middle class continues its 40-year decline. ... Meanwhile, the wealthiest people and the largest corporations are doing phenomenally well.” And here’s Trump, for comparison, at an August rally: “The hedge-fund guys are getting away with murder. ... They’re making a tremendous amount of money. They have to pay tax. I want to lower the rates for the middle class. ... They’re getting absolutely destroyed.” 

Both Sanders and Trump complain about American resources being squandered abroad, while many Americans do without at home. They mourn the outsourcing of jobs to workers overseas, and promise to return jobs to American shores. These are threads of their noted populism (a label that Trump, of course, denies). They’re also, I suspect, the tones and attitudes that attract the same voters despite the two candidates’ utterly different methods for addressing the problems they identify.

Sanders himself has suggested it’s the anxieties he and Trump both address that results in the crossover interest—and that what Trump offers to assuage those anxieties is scapegoating, not problem-solving. “This country has experienced racism for hundreds of years,” the Vermont senator said at a September roundtable event. “Clearly Trump is scapegoating the Hispanic community. Immigrants are not responsible for the disappearing American middle class, the Wall Street collapse brought on by huge financial institutions’ greed and illegal behavior, the war in Iraq, income inequality or climate change.” 

Voters disgruntled with declining economic opportunities and falling status can blame their plight on immigrants or refugees, as Trump has invited them to. Or they can be equally convinced that the culprit of American middle-class decay is a rather vicious financial industry, its incessant political lobbying, and the collapse of old supports like unions—Sanders’s message. The solutions differ, but the problem is the same. 

This is why Sanders could have some luck with Trump fans, and to great effect: If the senator can convince them that their impulses against immigrants, refugees, and minorities at large arise from their own disenfranchisement, then the same anxieties that could transform (vis-a-vis Trump) into mobbish rage could be channeled into energy to advance progressive policies. If Sanders is reaching even a few Trump supporters, then, it’s not so much evidence of the two candidates’ sameness as a suggestion that even worrying fears can find constructive expression, an outcome we should all hope for.