It’s usually best to ignore Jonathan Chait, a man whose trouble predicting electoral outcomes is surpassed only by his contempt for the Left and his condescension toward working-class Americans.

These qualities come out in Chait’s latest polemic, “Running Bernie Sanders Against Trump Would Be an Act of Insanity.” Despite poll after poll showing Bernie Sanders beating Donald Trump by wide margins nationally and in must-win battleground states, Chait insists that “Sanders is an extremely, perhaps uniquely, risky nominee.”

And although the Vermont senator has a proven record of defeating Republicans, receives the highest polling numbers among independents, and has already won twelve electoral races, Chait nevertheless would have us believe that Sanders is somehow virtually unelectable in 2020.

The Sanders campaign and numerous analysts have repeatedly refuted these charges and, as evidenced by Sanders’s surging popularity, the establishment’s tired attacks no longer seem to be sticking. Nor is it likely that many voters will be swayed by the electoral wisdom of a journalist who penned the 2016 classic, “Why Liberals Should Support a Trump Republican Nomination.”

But buried within Chait’s article is an argument that Sanders’s supporters do need to be better prepared to respond to. According to Chait, the results of the 2018 midterms demonstrate the futility of trying to defeat Republicans by running left-wing candidates. He notes, correctly, that none of the congressional candidates backed by Our Revolution, Justice Democrats, and Brand New Congress were able to flip Republican seats. In contrast, centrist Democrats supported by the party establishment flipped forty seats. The Democratic Party, he concludes, ran “a field experiment between two factions, and the moderate faction prevailed.”

Chait’s facts on this question are sound, but his political conclusions are faulty. The reason for this is simple: presidential contests and midterms are very different beasts.

Take the question of turnout. Sanders hinges his electoral strategy on motivating nontraditional voters, who are disproportionately young, poor, and nonwhite, and who tend to favor redistributing wealth as well as other reforms. Yet one of the long-standing features of midterms is that they bring out a significantly lower percentage of voters than presidential elections.

In 2018, turnout was only 51 percent — a drop of more than 7 points from the 2008 presidential contest that Obama won by promising progressive change and building a multiracial, multigenerational coalition to win it.

Midterm voters are also significantly older than in presidential elections: in 2018, 69 percent of those over sixty-five voted, compared to only 22 percent of eighteen to twenty-four-year-olds. As one detailed analysis of the midterms concludes, “young voters certainly appear to have the most room to grow” in 2020 — a fact that bodes well for the prospects of Sanders’s youthful movement.

There is also a deep difference between the political geography of congressional and presidential races. Political scientist Jonathan Rodden demonstrates in his recent book Why Cities Lose that because the working class is concentrated in urban areas, left-wing parties generally do best in statewide and national elections, where a single urban vote matters just as much as a rural vote. Yet in legislative elections for discrete geographic districts, the Left almost always fares significantly worse, since conservative-leaning rural regions have a disproportionately high number of seats.

And though 2018 saw a significant increase from past midterms, this only underscores just how few people normally vote — turnout rates in the United States trail virtually every other industrialized country in the world. Moreover, the most dramatic voter increase voter in 2018’s midterms came from suburban white college graduates. As Matt Karp documents, mainstream Democrats were generally only able to flip wealthier districts:

Among the nearly forty House districts where Democrats took control in 2018, about thirty are rated “prosperous” or “comfortable” by the Economic Innovation Group, a bipartisan think tank. Of the forty-three “distressed” districts held by Republicans, Democrats flipped just two (NJ-2 and NM-2). In other words, the midterms confirmed that the Democrats have become — perhaps more than ever before in their two-hundred-year history — a party of the prosperous.

Beating Trump won’t be easy — despite his unpopularity, the president has held his core support intact. Sticking with the Democratic establishment’s orientation to affluent moderates will spell disaster in 2020, just like it did in 2016 when 4.4 million Obama voters stayed home.

An anti-establishment candidate like Sanders is our best bet to bring out young and working-class voters to the polls, particularly in the crucial battleground states of Ohio, Michigan, Pennsylvania, and Wisconsin. And as a recently leaked audio recording demonstrates, Trump is well aware that Sanders is uniquely poised to do just that.

Much of Sanders’s appeal has to do with political stances shared by other insurgent candidates — Trump, for instance, cited Sanders’s pro-worker stance on trade as one of the central reasons he feared his candidacy. Nevertheless, there are also features unique to Sanders that underscore why he’ll certainly outperform the 2018 anti-establishment cohort.

After forty years of political consistency and a nearly victorious 2016 presidential run, Sanders has an unparalleled level of name recognition, favorability, trustworthiness, and an unmatched appeal to independents and Republicans, urban and rural alike. Most of 2018’s leftist congressional candidates, in contrast, were relative unknowns to voters, making it harder to generate the depth of popular mobilization necessary to overcome well-funded opponents.

To win elections, resources matter — and none of the 2018 insurgents had anything resembling the massive media and canvassing apparatus that Sanders has built up through millions of working-class donations.

The 2018 losses point to a real political problem — but it’s not the one Chait has cynically concocted. Though most Americans support progressive policies, we don’t yet have strong working-class political organizations to build the class consciousness, popular enthusiasm, and day-to-day presence in workplaces and communities necessary to consistently win down-ballot races across the country. But for Sanders to effectively push his agenda through Congress, he’ll need many more allies in the halls of power like Alexandria Ocasio-Cortez, Ilhan Omar, and Rashida Tlaib.

Retreating to a mythical political “center” won’t resolve this dilemma. There’s only one solution: building independent working-class organizations to spread political revolution to all corners of the United States. And a Sanders presidential nomination would provide a once-in-a-lifetime opportunity to make this possibility a reality by unleashing an unprecedented groundswell of volunteer energy and enthusiasm, a wave that’ll make the current upsurge pale in comparison, one deep enough to lift insurgent candidates to office all the way from the local school board to the White House.

“To nominate Sanders,” writes Chait, “would be insane.” He’s wrong. The definition of insanity, as democratic socialist Albert Einstein long ago pointed out, is doing the same thing over and over again and expecting different results.

By imposing a corporate Democrat nominee in 2016, the Democratic establishment and its apologists like Chait delivered us a four-year-long Trump nightmare. We can’t afford to let this history be repeated.