A guy walks into a doctor’s office and says, “Doctor, it hurts when I do this.”

The doctor says, “Then don’t do that!”

After Barack Obama’s reelection in 2012 and in the wake of the Sandy Hook massacre, the Democratic Party re-engaged with the argument for gun control—a topic it had largely set aside as electorally inexpedient during the 2006, 2008, and 2012 cycles. Under pressure from climate activists, the Obama administration abandoned the “all of the above” energy policy that the president had run and won on twice, killing the Keystone XL pipeline. In the wake of Michael Brown’s death and large-scale protests in Ferguson, Democrats took up the cause of criminal justice reform. After the failure of a legislative push for comprehensive immigration reform, activists pushed Democrats—successfully—to embrace reduced enforcement on a party-line basis as an alternative to the quest for bipartisanship.

None of these decisions was undertaken capriciously or for no reason. But all of them involved running the risk of alienating more culturally moderate working-class voters whose primary attraction to the Democratic Party was economic issues.

What were they thinking? Well, to quote the headline of a 2013 Ron Brownstein article, the idea was that “With New Support Base, Obama Doesn’t Need Right-Leaning Whites Anymore.” We learned in 2016 that this was not true and that the views of older non-college whites in the Midwest continue to be quite decisive in American politics. And we learned in 2020 and 2024 that Hispanic and Asian Americans are broadly susceptible to many of the same kinds of arguments that worked for Midwestern whites in 2016. Both groups’ vote share for Democrats has declined significantly.

One logical reaction to this set of facts would simply be to shrug. Life is complicated. Politics is complicated. Election outcomes are heavily shaped by macroeconomic circumstances, military casualties, and other things that exist outside the four walls of strategy and tactics. A Democratic Party that hews tightly to causes beloved by donors and activists but not by swing voters will lose more races more often, but it’s not like they will never win. Given a bad enough recession, an ideologically dogmatic party may sweep into office and shoot the moon on public policy.

But, of course, this attitude runs the risk of empowering fascists, threatening the foundations of American electoral democracy, costing millions of people their health insurance, subjecting the country to a terrifying new regime of internal immigration enforcement, making less-than-zero progress on climate change, and depriving millions of women of their basic rights. To me, that makes “shoot the moon” a bad bet—Democrats have been trying a version of shoot-the-moon since Obama’s reelection, it has hurt, and the solution is to stop doing it. But it would be an intellectually stimulating debate. Highly ideological leftists are aware, I think, that the mood in the Democratic Party is very alarmed by Trump and Trumpism and that if we had square argument about the benefits and risks of shooting the moon, their side would lose.

So instead we have been greeted with a series of academic obfuscations and circular arguments designed to occlude the obvious fact that there are political costs to pushing the envelope on policy views, especially when your views are unpopular.

Adam Bonica and Jake Grumbach claim to find only “small” electoral benefits of moderation, on the order of one to two percentage points. Their subtext is, who could possibly care about an effect of this size? In fact, campaign professionals emphasize to me that this is an enormous effect compared to what you get from spending on campaign ads or mounting field programs to mobilize and turn out voters. A swing of one to two percentage points would have been enough for Hillary Clinton to win in 2016 or for Kamala Harris to win in 2024. Two is a small number, but I don’t think that’s small change for the world. It’s a little bit hard for me to understand what this debate is even supposed to be about given that the skeptics of moderation are willing to concede that both of Trump’s election wins were within the bounds of what they say can be plausibly achieved by moderating. To underscore the point: since 2012 there have been seventeen Senate races decided by less than the Bonica-Grumbach margin of moderation.

Similarly, Bonica and Grumbach’s discussion of turnout hinges on a false dichotomy between persuasion and mobilization. If taking extreme positions generated meaningful turnout benefits, you would expect Justice Democrats endorsees to run stronger than moderates. In fact, they run weaker. Nor do people seeking to emulate them do any better. In the 2020 House cycle, Ryan Grim of The Intercept reported that “Kara Eastman and Dana Balter are giving the left a chance to show that it can win swing districts.” They failed, running on positions further to the left than the national party—both lost key races to incumbents despite a favorable national political climate with widespread anti-Trump sentiment. (Biden even took Eastman’s district, Nebraska’s 2nd, that year, flipping back from Trump in 2016. She faced a three-way race in 2020, yes, but even if she had won all of libertarian Tyler Schaeffer’s votes, she still would have lost.)

A better way to think about turnout is to view nonparticipation as a kind of halfway house between voting for Democrats and voting for Republicans. You get nonvoters to vote for you, opposition voters to not vote, and hardcore voters to vote for you rather than the opposition all with roughly the same set of persuasive techniques.

Last but not least, Bonica and Grumbach veer between descriptive analysis of strategic choices made by elite political actors and counterfactual causal claims. They emphasize, correctly, that contemporary politics is highly polarized and that contemporary elections are highly nationalized. But polarization is not exogenous, some fact of nature that parties have no control over. If Democrats had not shifted left after 2012 (or if Republicans had not repeatedly nominated a maniac), politics would be less polarized. In Obama’s first term, Sherrod Brown positioned himself to Obama’s right on climate and Jon Tester voted against the DREAM Act. By Biden’s term, both men had come closer to the national party on these issues even as the national party itself moved left. The “nationalization” of politics isn’t just something that happened to them; it’s something that they helped do.

The other aspect of nationalization is that congressional politics has become more parliamentary, so the role of individual members matters less than their role as proxies for national congressional leadership. This shows that Bonica and Grumbach’s one- to two-point estimate for the electoral benefit of any individual candidate moderating must be a massive underestimate of the benefit of the national party moderating. Susan Collins’s voters care that she is more moderate than the national GOP, but they also care that she empowers John Thune. If the whole Republican caucus were more like Collins, that would help her win.

David Broockman’s paper on moderation is fascinating and important, but it simply does not support Bonica and Grumbach’s position. Broockman shows that most moderates have a mix of left and right views rather than being centrist across the board. For example, both Social Security cuts and laxity on immigration are unpopular—a typical “moderate” ten years ago would have been liberal on Social Security and conservative on immigration. Trump brought Republicans into closer alignment with these voters. Democrats have engaged in some course-correction on immigration since 2024, but they spent Biden’s presidency validating the concerns of voters who worried about their approach to border security.

Bonica and Grumbach note, rightly, that pro-Harris paid media largely featured popularist themes, but it is a fallacy to suggest that since she lost, the campaign had no positive effects. If you look at the map in detail, the 2020–2024 swing against Democrats was much smaller in the swing states where the campaign ads aired than it was on a national basis. The Harris campaign’s themes were good themes! It’s just that campaign ads are limited in their impact. In 2019 Harris supported a fracking ban; given the party’s posture and track record on this subject, one-off statements in interviews and the presidential debate weren’t going to make voters believe her when she said, in the 2024, that she wouldn’t. The people who crafted her ads did not tell Harris to refuse to distance herself from any Biden administration policies. The moral is that, to the extent that popularist ideas were put in place, they were effective—but Harris failed to go all the way and convincingly portray herself as a person who seriously disagreed with progressive positions that are deeply unpopular in critical districts. Conversely, much of Trump’s campaign message highlighted Harris’s earlier progressive stances. Do Bonica and Grumbach think this was a huge mistake on his part, and Republicans should have tried to obscure this from voters in order to depress turnout?

But here comes a critical point of agreement. Bonica and Grumbach write that “in a nationalized media environment with an authoritarian movement capturing one major party, electoral politics has become existential conflict.”

This is correct. It is why attempting to shoot the moon is irresponsible. It’s also why the party establishment’s dream of addressing the problematic Senate map purely through recruiting is unlikely to work. If Democrats want a Senate majority, it will run through states like Ohio, Iowa, Texas, Florida, Georgia, and Alaska. Given the nationalized media environment, it’s not good enough to nominate solid candidates in those states. You need a national party message that appeals to voters there—voters who dislike illegal immigration and want a secure border, who care more about jobs and growth than carbon emissions, who favor tough-on-crime policies, who have somewhat traditionalist views about gender, and who are skeptical of left-wing racial politics.

The bottom line is that transforming the Democratic Party into a more ideologically coherent, more progressive force in a country where conservatives outnumber liberals and where the electoral map is skewed against Democrats hurts it. The solution is to stop doing that.