Normally the political party that loses an election goes through a period of soul-searching and vigorous internal debate, while the winning party embraces a smug certainty about its own inevitable multigenerational dominance. In 2021, though, the roles are reversed.

The widespread belief that Donald Trump was, in some sense, the real winner of an election that he lost has succeeded in preempting a Republican debate about why the Democrats captured the White House last year. Meanwhile, the Democrats, despite their control of the Congress and the presidency, are increasingly the ones arguing as though they’re already in the wilderness.

The Democrats’ angst strikes me as a healthy development for liberalism. One problem with the emergency thinking that Trump inspires in his opponents — and one reason to resist it — is that it occludes real understanding of the political conditions that put him in power, and that might do so again. This is what you saw happen to the Democrats after 2016: The sense of being lightning-struck sent the center-left wandering into a maze of conspiracies, a haunted wood where villains like Vladimir Putin and Mark Zuckerberg loomed larger than the swing voters they had lost and savior figures like Robert Mueller were supposed to unmake Trump’s power for them.

Only the party’s left, its Bernie Sanders wing, fully developed a more normal theory of the 2016 defeat, trying to understand Obama-Trump voters in the context of globalization and deindustrialization as well as racism, fascism and Putinist dirty tricks. But this created a fundamental imbalance in the party’s conversation: With the Sanders faction trying to pull the party toward social democracy and the establishment acting as if its major challenges were Russian bots and nefarious Facebook memes, there was hardly anyone left to point out the ways that Democrats might be in danger of moving too far left — and the writers who did so were generally dismissed as dinosaurs.

So it was up to Democratic voters to exert a rightward tug on their party — first by saving the party from the likely disaster of nominating the intelligentsia’s candidate, Elizabeth Warren, and ultimately by putting up a nominee, Joe Biden, whose long career as a moderate gave him some distance from the “Great Awokening” that swept liberal institutions in 2020.

Now, though, with the increasing awareness that Bidenism is probably not a long-term strategy, we’re finally getting the fuller argument that should have broken out after 2016 — over what the Democrats can do, and whether they can do anything, to win over the working-class and rural voters alienated by the party’s increasingly rigorous progressive litmus tests.

A key player in this argument is the pollster and analyst David Shor, whom my colleague Ezra Klein interviewed for a long essay last week, and who has emerged — after a temporary 2020 cancellation — as the leading spokesman for the pragmatic liberal critique of progressive zeal.

This critique starts with a diagnosis: Democrats misread the meaning of Barack Obama’s 2012 victory, imagining that it proved that their multiracial coalition could win without downscale and rural white voters, when in fact Obama had beaten Mitt Romney precisely because of his relatively resilient support from those demographics, especially across the industrial Midwest. And this misreading was particularly disastrous because these voters have outsize influence in Senate races and the Electoral College, so losing them — and then beginning to lose culturally conservative minority voters as well — has left the Democrats with a structural disadvantage that will cost them dearly across the next decade absent some kind of clear strategic adjustment.

From this diagnosis comes the prescription, so-called popularism, glossed by Klein as follows: “Democrats should do a lot of polling to figure out which of their views are popular and which are not popular, and then they should talk about the popular stuff and shut up about the unpopular stuff.”

You will note that this banal-seeming wisdom is not an ideological litmus test: Where left-wing ideas are popular, Shor Thought would have Democrats talk about them more. But where they are unpopular, especially with the kind of voters who hold the key to contested Senate races, Democrats need a way to defuse them or hold them at a distance.

Thus a “popularist” candidate might be a thoroughgoing centrist in some cases, and in others a candidate running the way Sanders did in 2016, stressing the most popular ideas in the social-democratic tool kit. But in both cases such candidates would do everything in their power not to be associated with ideas like, say, police abolition or the suspension of immigration enforcement. Instead they would imitate the way Obama himself, in his first term, tried to finesse issues like immigration and same-sex marriage, sometimes using objectively conservative rhetoric and never getting way out ahead of public opinion.

Which is easier said than done. For one thing, the Democratic Party’s activists have a different scale of power in the world of 2021 than the world of 2011, and the hypothetical “popularist” politician can’t make their influence and expectations just go away. For another, as my colleague Nate Cohn points out, Obama in 2011 was trying to keep white working-class voters in the Democratic fold, while the popularist politician in 2022 or 2024 would be trying to win them back from the GOP — a much harder thing to achieve just by soft-pedaling vexatious issues.

At the very least a Democratic strategy along these lines would probably need to go further along two dimensions. First, it would need to overtly attack the new progressivism — not on every front but on certain points where the language and ideas of the progressive clerisy are particularly alienated from ordinary life.

For instance, popularist Democrats would not merely avoid a term like “Latinx,” which is ubiquitous in official progressive discourse and alien to most U.S. Hispanics; they would need to attack and even mock its use.

Likewise, a popularist candidate — ideally a female candidate — on the stump might say something like: I want this to be a party for normal people, and normal people say mother, not “birthing person.”

But then along with this rhetorical fire directed leftward, popularists would also need go further in addressing the actual policy concerns surrounding the issues they’re trying to defuse. Immigration is a major political problem for Democrats right now, for instance, not just because their activists have taken extreme positions on the issue, but because the border is a major policy problem: The effects of globalized travel and communication make it ever-easier for sudden migrant surges to overwhelm the system, and liberalism’s shift away from tough enforcement creates extra incentives for those surges to happen under Democratic presidents.

So in the long run — especially given climate change’s likely effect on mass migration — there is no way for Democrats to have a stable policy that’s pro-immigration under the law without first having a strategy to make the American border much more secure than it’s been under the Biden administration to date. How to do that humanely is a policy challenge, but if you really want to court voters for whom the issue matters, you have to take the challenge seriously.

Even a strictly defensive strategy, one that just prevents more Hispanic voters from shifting to the Republicans and holds on to some of Biden’s modest Rust Belt gains, would buy crucial time for Democrats — time for a generational turnover that still favors them, and time to seize the opportunities that are always offered, in ways no data scientist can foretell, by unexpected events.

Douthat is a columnist for The New York Times.