Democratic leaders who want a winning message against Donald Trump’s military occupation of Washington, D.C., should accuse the president of orchestrating “a stunt to distract from the pain his tariffs are causing families” and of “creating fear to distract” from his Republican Party’s cuts to Medicaid. What Democrats should not do: accuse Trump of “manufacturing a D.C. crime crisis” or of committing a “historic assault on D.C. home rule and is more evidence of the urgent need to pass a D.C. statehood bill.”
That’s the takeaway of a memo from Blue Rose Research, the outfit led by Democratic data guru David Schor, that’s making the rounds on liberal-left social media. The five-page document summarizes a poll, conducted August 12-13, that tested various messages about the federal takeover of the District. The survey found that messages accusing Trump of a “stunt” or “distraction,” and then pivoting to the damage caused by tariffs or Medicaid cuts, were the biggest drag on Trump’s approval; in fact, of the 16 Democratic messages tested, only those two tested above average, “and even still were barely above average,” the memo states.
This memo, which my colleague Perry Bacon reports “is being sent to Democratic leaders/elites,” could help explain why Democrats in the House and Senate have repeatedly said this exact thing in interviews and on social media. “As Donald Trump attempts to create chaos that distracts from his problems, we’ll call it out for what it is,” Illinois Governor JB Pritzker wrote on Bluesky on Friday. “Trump and Republicans are trying to distract from the pain they’re causing—from tariffs raising the prices of goods to stripping away healthcare and food from millions.”
The memo’s timing is impeccable: It has surfaced at the very moment several influential writers on the left have been debating the merits and uses of polling. But the issue is less with polling itself than with how the information is used—particularly by the Democratic Party. Many on the left have criticized the Democrats’ messaging as far too tame and predictable amid Trump’s increasingly authoritarian second term, accusing party leaders of parroting poll-tested messages rather than speaking passionately from the heart. The party’s real problem in the Trump era, though, isn’t simply that they’re letting public opinion—as represented by polls—shape their messages, though of course they’re doing that. The problem is that they’re scared of public opinion, in particular the opinions of the swing voters who elected Trump.
Earlier this month, John Ganz, a Substack blogger, author, and Nation columnist, posted what he later acknowledged was a quickly written “take.” “Supposedly, the way you make a successful political campaign is that you go out and you ask people what they want, and then you make your message based on that. Except that’s bullshit. It doesn’t work,” he wrote. “Politely put, the data-based approach to politics is based on a fallacious understanding of the world. Not so politely put, it’s a racket for political consultants so they can scam hapless hacks and wealthy donors.”
This caused a stir—and a flurry of further takes, which I will only briefly summarize here and invite to read at your own pleasure (or misery, as the case may be). Vox’s Eric Levitz rebutted Ganz’s piece, arguing that polling—while flawed, and hardly the only tool politicians and their strategists should use—is a necessary bulwark against progressives being carried away by wishful thinking. (As the essay’s subheadline put it, “Progressives can’t afford to trust their guts.”) Ganz responded to Levitz with a piece lambasting “vulgar positivism,” a reference to a centrist political theory of late that says that the Democrats should say things that are popular (and whether these things are popular is, of course, defined by polls). Matt Yglesias tried to split the difference in this brawl in a question from a reader that asked Yglesias about the debate. He and Levitz approached the issue as if there were only two choices: paying attention to polling data or navigating blindly by intuition. A new center-left, pro-Abundance publication, The Argument, also weighed in with a defense of polling—as a way of promoting their own new data project.
As a reporter who values fact-finding, empiricism, and evaluating evidence—and who, before joining TNR, worked at the famously nerdy political website FiveThirtyEight—I’m sympathetic to any arguments about the appropriate use of data. But I side primarily with Ganz, and that’s because I have spent most of the past three years talking to voters and listening to what they say. The issue isn’t that polling is “90% bullshit,” as Ganz exaggerated in his first piece (and I don’t think he believes that either). I understand how frustration with current political discourse would inspire such hyperbole. The Democratic Party has handcuffed itself with data rather than use it to their advantage.
A perfect example is the issue of immigration, which Ganz highlights in his first post. Democrats have generally moved right on the issue since Trump won in 2016, and especially so since his victory last year. They’re doing so because they’re trying to align with the imagined median voter, as gleaned from polling. But public opinion on immigration shifts over time, sometimes drastically, and for reasons that are not particularly mysterious.
Before Trump first became president, even Republicans did not see building a wall on the southern border with Mexico as a priority. But throughout Trump’s first term, as he claimed the U.S. was being “invaded” by foreign criminals, Republican support for building more border walls rose, while Democratic support fell. When Trump was out of office, Republican faith in an expanded border wall softened a bit, with 72 percent of Republicans in 2024 saying that it would improve the border situation. But in the 2024 election, which came after a multi-year spike in migration to the U.S. under Biden, immigration was a bigger driver of support for Trump than in 2016.
Please don’t yell at me data nerds: I know that correlation does not equal causation. But I am also a voter who lives in Trump country, and has for most of the past decade. I hear the way my fellow voters talk about immigration. They’ve begun to adopt the kind of rhetoric Trump has used. At the same time, real-world events have conspired to bring unprecedented numbers of migrants seeking refuge in the U.S., and a compliant right-wing press has helped spread Trump’s anti-immigrant propaganda. Much of the rest of the Republican Party has jumped on board. Trump wouldn’t have been able to do it alone, but he has helped shape his party’s views on immigration, period.
You don’t need to dig too far into the data to determine that these views aren’t very deeply held. While many voters were persuaded by Trump’s rhetoric on immigration and said they supported his policies, they were less keen on his ideas when those issues were framed in a different way, highlighting the economic and social destructiveness of mass deportations. In fairness, Vice President Kamala Harris did warn that these things would happen, as Levitz pointed out, but she also stressed her support for a bipartisan bill—which the GOP blocked—that would have secured the border. “And let me be clear, after decades in law enforcement, I know the importance of safety and security, especially at our border,” she said in her convention speech. “Last year, Joe [Biden] and I brought together Democrats and conservative Republicans to write the strongest border bill in decades. The border patrol endorsed it. But Donald Trump believes a border deal would hurt his campaign, so he ordered his allies in Congress to kill the deal.”
Which is to say, Democrats let Republicans frame the debate on immigration by ceding ground on the issue of “border security,” and they did so because polling showed that it’s popular. Of course it’s popular! “Border security” sounds like a good thing! But how much do Americans think about border security in their daily lives? Unless they actually live on the border, probably very little. And is the issue of “border security” more important than reforming the U.S. immigration system, which Democrats have been trying to do in earnest since 2013, only to be thwarted by unserious Republicans?
That is the trouble with issue polling in general. It is useful to know what the American public feels about a given issue, but these are usually quick questions that provide a snapshot in time. For instance, how voters define “border security,” what part of the immigration situation bothers them or pleases them, and what solutions they support are questions that are not always asked. Levitz accused progressives of being vulnerable to motivated reasoning without data, but moderates are, too—because data is not that deep. It doesn’t really explore beyond a surface-level understanding, and shouldn’t trump morals, values, and the ability to take on a fight when one arrives at your doorstep, which is what’s happening now.
Democrats are always overlearning the lessons from the last election and ignoring the new realities unfolding every day. Trump’s approval rating continues to fall, even among those who voted for him in 2024. They disapprove of his actions on immigration, the economy, and on other issues. Yes, I’m making these assertions based on polling, but the evidence is becoming undeniable. And this is precisely the kind of evidence Democrats usually find persuasive, so I would hope they now recognize a clear political opportunity to talk about how absolutely hideous and destructive this presidency is, from ICE’s gestapo tactics to the tariffs tanking the economy, while also pitching their solutions to it. But they need to start building that case now, cementing it in voters’ minds—not wait until the next presidential election rolls around in the belief that whatever message they concoct in 2028 will be persuasive and perfectly timed.
It’s important to remember—since Democratic strategists often seem to forget—that swing voters are not the only voters. There are also Democratic voters who want to have faith in their party again! They’re mad as hell about what’s happening and want their representatives to feel the same way. There are also nonvoters, lots of them young people, who might be energized by a little righteous anger and moral crusading.
Look at state Assemblymember Zohran Mamdani in New York City. Amid all the handwringing about socialism and whether it will play in the heartland (as if it needs to), too many strategists have missed the fact that the Democrats’ mayoral nominee is talking like a normal person about the issues that normal people care about, like how impossibly expensive the city has become. He has lots of ideas, too, and some are unorthodox or even pie in the sky. But that’s actually smart politics. Voters are less concerned about whether Mamdani will able to, say, create city-owned grocery stores than the fact that he’s promising to do things. He’s also engaging voters where they are—not only on social media, but at their businesses and on the streets. And he’s not always so serious; he can be funny, and fun. His rivals dismissed his scavenger hunt last weekend as a silly game, but Mamdani is showing that politics isn’t just about the most dire things that are happening, but also reminding people what they are fighting for: a vibrant city that is not the violent hellhole portrayed by Trump.
You won’t find much levity among national Democrats these days, and perhaps that’s understandable. But why can’t they just sound normal? When they dismiss Trump’s takeover of D.C. as a “distraction” and pivot immediately to tariffs or Medicaid, it seems so practiced because it probably is; maybe some of them have even read the Blue Rose memo. They’re quickly moving past the deepest concern here—a fascist president using the capital as a dry run for a national police state—to get to “kitchen table” issues, as if fearful they might say the wrong thing about crime. Democrats are so worried about alienating a small slice of the electorate that they’re hesitating to condemn the most abusive act yet by a historically unpopular president from the opposing party. I’m no political strategist, but boy does that sound like a losing political strategy.
