Adam Bonica and Jake Grumbach join a high-stakes debate: Which way forward for the Democrats? We find much to agree with in their argument—no amount of moderation is going to quell Americans’ “deep-seated cynicism” with the status quo. But on several points we think their analysis obscures more than it clarifies, resulting in a prescription that is hollower than it should be.

First, the core issue they discuss—do moderates outperform progressives?—hinges on how we define overperformance and moderation. A lot of ink has been spilled on the former question, but the latter remains unclear.

Consider how the term has been used. Lakshya Jain of Split Ticket argues that moderation means “being willing to cut against your own party.” The New York Times editorial board bizarrely praises Jain only to criticize a measure of moderation based on how frequently members of Congress vote against their party; moderation, we are told, should be understood as a “new centrism” that “promises sweeping change while criticizing the two parties as out of touch” rather than “locating itself midway between the two parties.” For Ruy Teixeira, moderation means being attentive “to the issues voters are most concerned about, rather than just singing from the progressive hymnal.” Data journalist G. Elliot Morris describes it as pairing “an extreme left position with an extreme right position.” For their part, Bonica and Grumbach interpret it as chasing the median voter by “shifting dramatically to the ideological center.”

Good empirical studies of all these things matter. But the findings aren’t particularly meaningful if we don’t agree on what we’re talking about.

Second, most of the analysis we are aware of—including Bonica and Grumbach’s—relies on measures of moderation that make unrealistic assumptions. One assumption is that individual policy positions can be objectively arrayed along a left-right axis. Another is that we can accurately represent a particular candidate’s or voter’s total package of views somewhere on that same axis. Both are bad assumptions. Is a ban on new fracking more or less left-wing than raising capital gains taxes? And is someone who supports ICE and fracking bans more or less left-wing than someone who opposes ICE and higher taxes? Although Bonica and Grumbach acknowledge that voters typically hold a mix of views, they rely on measures that reduce ideology to a single numerical metric.

We doubt such heroic abstraction can be of much use to political strategists. The reality is that it’s very difficult to estimate the electoral effect of any concrete instance of a change in policy or messaging. And even if we agree on what moderation means, moderating on some issues could matter very little, while doing so on highly salient issues could do a lot. If voters seem to care a great deal about illegal immigration and very little about industrial policy, moderating on immigration may compensate for adopting stronger positions on state intervention in the economy—an asymmetry that standard measures of moderation are unable to detect.

Third, leaders may make history, but they don’t make it as they please. Bonica and Grumbach are right that public opinion isn’t fixed, even on highly partisan issues—plummeting approval of ICE is a case in point. They are also right that effective political leadership can move public opinion. But neither of these facts implies that voter attitudes are a tabula rasa or that they should play no role in strategic calculations. Indeed, Bonica himself has conceded that “running moderates in the most difficult swing districts is a sound tactic.” It’s hard to see how doubling down on salient issues that poll very badly wouldn’t hurt Democrats in competitive races—those where they must persuade some independents or Republican voters in order to win.

The upshot is not that Democrats have no choice but to parrot Stephen Miller or Matthew Yglesias. Rather, there’s a huge issue—the issue that voters actually care about the most—that Democrats should be raising the salience of but aren’t: the economy. Increases in partisan polarization over the past few decades have been driven by social issues. In other words, voters who agree with Democrats on social issues already vote for them. But a large chunk of the electorate holds left-wing attitudes on economics and yet does not vote for Democrats. Reclaiming credible leadership on the economy—and realigning the electorate around issues of economic fairness—is the obvious way to make inroads in the competitive districts that Democrats need to win in order to take back the federal government. Just consider Dan Osborn, the independent economic populist Senate candidate from Nebraska, who overperformed virtually all congressional Democratic candidates in 2024 by focusing on kitchen-table economic issues and shifting power away from plutocrats and toward working Americans.

For the reasons we gave above, focusing more squarely on economic populism could help inoculate Democrats from attacks on other issues. For example, the Trump campaign ran an ad telling voters that Harris was for “they/them,” not for “you.” If you believe arguments that it struck a chord with some voters who weren’t already going to vote for Trump, it likely did so not only because of the soundbites it used—clips of Harris expressing her support for taxpayer-funded gender-affirming care for inmates—but because voters didn’t trust that Harris really was for them. Such attacks likely wouldn’t have landed as effectively against an economic populist candidate like Bernie Sanders—who, incidentally, also supports transgender health care, indeed health care for all.

In other words, we agree with Bonica and Grumbach that Democrats have to “reframe the battlefield . . . by building unlikely coalitions around a broadly resonant grievance.” But we think their “anti-corruption” agenda isn’t going to cut it. Billionaires aren’t just “buying Supreme Court justices”—they are wrecking the economy for working people. Procedural reforms like “rejecting corporate PAC money” and “banning congressional stock trading” aren’t bad ideas, especially as a means to economically populist ends, but they won’t be enough to earn voters’ trust. The missing ingredient in this vision is class war—which means, at a minimum, a convincing campaign to tax the wealthy, crack down on corporate power and abuse, create good, stable jobs, and ensure affordable health care and housing for all.

As Bonica and Grumbach suggest, this can’t be the task of individual candidates. The Democratic Party’s best performers, like Sherrod Brown, now outrun the top of the ticket by only a few points. Instead, constructing a convincing economic populist alternative has to be a project of the national party—a challenge made all the more difficult by the toxicity of the party’s brand. Democratic America has become increasingly wealthy, well-educated, and professional class. In the rest of the country, Democrats are seen as out of touch and elitist; overcoming that perception will take more than populist rhetoric. The party has to engage in sharp, visible, and polarizing policy fights—fights that convince skeptical voters it is actually doing something different and working to make material improvements to the lives of ordinary Americans.

It was precisely the demise of sustained, confrontational, dare we say “extreme” policy commitments in defense of working people over the past four decades that led to the rise of Trump. If anything is going to convince working-class Americans to place their trust in the Democratic Party again, it will require picking risky fights with the corporations and politicians who have rigged the economy and advancing policies that claw back power from the top and put it in the hands of everyday Americans.