The very worst thing Joe Biden could do if he hopes to defeat Donald Trump is forgive student loans for the millions. The second-worst thing is talk up the tax credit for electric vehicles his administration finalized earlier this year, the one that will make it more affordable for millions of people to afford something other than a gas guzzler. The third-worst thing he could do is fail to talk about his administration’s efforts to slow the crush of migrants crossing the southern border.
If you spend any time in the discourse of Democratic Party politics, these findings may seem at minimum like bad political strategy, even morally repugnant. Why avoid talking about all the great things you have done for young people and the environment? And yet they are what is necessary for Biden to have the best chance of winning in November according to Blueprint, a new Democratic data and polling whose findings have upended much of the conventional political wisdom this election year.
Among them: Biden doesn’t need a specific plan to win back parts of the Democratic coalition like young voters and voters of color who have defected, because their concerns are no different than other voters, namely inflation; that instead of talking up his efforts to combat climate change, he should talk about how under his administration the country has seen record energy production; that instead of talking about the vast new spending to rebuild the country’s infrastructure, the president should talk about how he is lowering the national debt; and that even Latino voters favor stricter immigration controls.
“There is a lot of polling out there, but what we felt was missing from all of it was polling that is just victory-minded,” says Evan Roth Smith, Blueprint’s lead pollster, and the founding partner of the political-consulting firm Slingshot Strategies. “The Democratic Party needs polling that just says, ‘We have to win this election, and so here is where the electorate is, here is what the Democratic Party has done and can credibly run on.’ Let’s see what works and just tell everybody what we find.”
He was sitting at Blueprint’s headquarters in a faded brick building that until recently was occupied by a fashion-photography firm in an ultrahip corner of Williamsburg, Brooklyn. Thick fashion magazines still sat on the bookshelves, while nearby nearly a dozen data scientists and political strategists pecked away on laptops. Three days earlier a new slate of polls had been released by The New York Times that showed Trump ahead in several swing states, and another wave of panic was washing over the Democratic ranks.
“One of the reasons why we suspect that the polling numbers are where they are is that voters don’t think Joe Biden and the Democratic Party are sufficiently focused on the things they’re most concerned about,” Roth Smith says. “And so our job is to highlight the things they are most concerned about and explain what the Biden administration has done.”
The EV tax credit and college-debt relief are both losers, according to Blueprint’s data, because they are both coded as preoccupations of the elite; instead, what Biden should be focusing on is his efforts to bring down the cost of pharmaceuticals, take on big corporations, tax the rich, and lower prices in the face of rising inflation. Roth Smith believes Democrats talk too much about Trump’s odious character, his legal liabilities, and the threat he poses to democracy instead of his economic record that includes a massive tax cut for the rich, which Blueprint found to be staggeringly unpopular.
These unorthodox findings have helped make Blueprint the buzzy polling outfit of 2024, its findings pinging around the internet and making their way into articles where reporters and pundits try to make sense of an election cycle that so far has avoided familiar narratives. Blueprint’s findings have regularly been written up in the Times and the Washington Post, chewed over on the A-block of primetime cable-news shows, and become the topic of the day on politics Twitter. More significantly, the firm’s findings are making their way to the upper reaches of the Biden campaign and those of Democratic congressional-campaign offices, and they seem to be having a result: After a Blueprint poll found that voters were far more concerned about prices than they were about jobs, Biden began talking more about his administration’s efforts to curb inflation and less about their jobs record. After Blueprint found that even Latino voters favored strict enforcement at the border, the White House announced an executive order that did just that.
“There is such an overload of horse-race numbers. If you are a practitioner, they give you stuff you can really use,” said A.J. Lenar, a Democratic strategist working on several congressional races. “Blueprint is a good name for what they do, because it gives you a good blueprint of where people are, and how to talk about an issue this way it will persuade voters and if you don’t it won’t.”
But Blueprint has also become a target of attacks from the left wing of the Democratic Party, who accuse the group of trying to nudge Biden into the center.
“We asked these swing voters, ‘Do you think Biden has done too much or too little?’ and the majority say that Biden has actually done too little,” said Danielle Deiseroth, the executive director of Data for Progress, a left-wing polling firm. “That goes against a lot of the punditry that says that somehow a radically left Biden economics agenda is turning people off of Biden.”
A lot of the skepticism of Blueprint comes from the fact that its biggest donor is Reid Hoffman, the co-founder of LinkedIn and someone who has given hundreds of millions of dollars to Democratic causes and who also believes that the party has drifted too far to the left. He’s been known to engage in the dark arts of politics by, for example, helping fund an online disinformation campaign to discredit far-right Republican Senate candidate Roy Moore (Hoffman later apologized, and said he didn’t know where the money was going) and by donating $250,000 to Nikki Haley’s presidential campaign (for the purpose of slowing down Trump).
“The idea that a billionaire is going to be less self-interested than an advocacy organization, or that a billionaire has better ideas about the approach the Democratic Party should be taking instead of a labor union is a little hard to swallow,” said Max Berger, a longtime progressive operative. “A lot of donors pretend they are neutral about policy and just want what is best for the Democrats, and that is somehow just a coincidence that the things they think Democrats should run on are also the things that would protect their material interests. The Democratic Party becoming friendly to corporate interests is what got us into this trouble in the first place.”
Some other operatives and pollsters from the center-left wing of the party said they distrust Blueprint after The American Prospect reported that the group were engaging the services of Sean McElwee, the founder of Data for Progress and lightning rod for many on the left after he was forced out of that position for betting on elections and his close ties to Sam Bankman-Fried. Blueprint’s leadership disputed that report that said that McElwee was part of the project.
As Dmitri Melhorn, Hoffman’s top political adviser, puts it, the problem with much of the current polling apparatus was that it did not focus on the narrow slice of voters who could be persuaded — and that much of it had been captured by interest groups foisting their narrow agendas onto the party. “The polling industry, prior to Blueprint, at least, was dominated by people whose revenue streams came largely from small progressive issue-advocacy groups, who obviously have strong market preference for pollsters who would tell them that that kind of issue they are advocating for is exactly what would win the next election,” Melhorn tells me. “We came to realize that a lot of decisions were being made in the center of the Democratic Party that we disagreed with, and we came to find out it was because of polling that was mismodeling the actual presidential election.”
“What is driving overwhelmingly the Democratic messaging decisions is bad polling that has made us too soft on crime, too soft on the border, not strong enough on fossil fuels, too soft on campus protests,” he adds. “There’s just this huge ocean of winnable voters in the center that we’re not reaching because our polling is broken.”
All polling involves choices about how questions are worded and in what order to present them, and outside pollsters who have worked with Blueprint said they found that the group is genuinely trying to get answers to questions, not to skew the results. But still, when a Blueprint poll found support for measures to harden the southern border, officials at Immigration Hub, which supports a more liberal approach to the issue, released a sharply worded response that said, “Blueprint fails to survey voters on the specific policies currently being considered by negotiators. The polling fails to communicate that families will be separated, asylum seekers denied protection, and detention will increase exponentially under the proposal,” and that pointed to other polling that showed a lack of support for Trump-era policies.
“If you go against what the activists want it can be tough for you. The social-media reaction isn’t great, the incoming you get from activist groups is going to be unpleasant,” said Lakshya Jain, a data scientist at Split Ticket, a political-modeling organization. “I think someone like Evan may have a different view of some of these issues than what the public thinks, but he doesn’t let that get in the way of the data.”
