They called it the “Big Send.” Democrats gathered in living rooms, libraries, and coffee shops across the country to write letters to millions of potential voters in swing states and competitive congressional districts, urging them to vote in November. During the 2020 pandemic election, the novel but decidedly 20th-century tactic had cut through the glut of digital messages that inundated Americans’ cellphones and inboxes, and organizers hoped it would similarly boost turnout for Democrats in 2024.
It did not.
In a study set to be released later today, the group behind the letter-writing effort, the nonpartisan Vote Forward, found that personal messages sent to more than 5 million occasional voters deemed at risk of staying home last fall had no effect on turnout. (The group’s campaign produced a modest increase in turnout among a second, slightly smaller set of low-propensity voters, but it still fell short of previous Vote Forward programs.) What’s unusual is not Vote Forward’s lackluster findings, but that the group is ready to tell the world about them. Every election, a constellation of progressive organizations sells donors and volunteers on the promise that their data-driven turnout programs will deliver victory at the polls. These mobilization efforts have taken on ever-greater importance in an era of tight elections, where the presidency and majorities in Congress can hinge on just a few thousand votes.
Progressive groups are only too happy to brag about their wins; they’re much less likely to divulge details about their campaigns that flopped. Driving this reticence is a fear that donations will dry up—or go to other organizations in a highly competitive campaign industry—if funders find out their money made little difference on the ground. In several instances, researchers told me, Democratic firms have either pushed them to suppress the results of studies that didn’t produce desired findings or cherry-picked data to make the numbers look better. “We have a people-pleasing problem in our party,” Max Wood, a progressive data scientist, told me.
Yasmin Radjy, the executive director of Vote Forward and its progressive campaign arm, Swing Left, is trying to change that culture. Just as Democrats are now debating, sometimes fiercely, why their party’s message failed last year, Radjy believes that to emerge from “the political wilderness,” they need to have candid conversations about their organizing and turnout efforts. Radjy has been frustrated by what she describes as Democrats’ lack of introspection and transparency. For months, she’s been asking party organizers and consultants what they learned in 2024, and what they’re going to do differently going forward. “We’ve got to actually be honest about both what works and what doesn’t work,” she told me. In the next election, “if we are serving volunteers, donors, and voters reheated leftovers from 2024, we are doing it wrong.”
The risks of a bad field operation are greater than people might think. The goal of any persuasion or get-out-the-vote program is to boost support for your party’s candidate. Many make only a small difference in turnout, or none at all—especially in presidential elections, for which most people already know plenty about the candidates. The worst of these efforts, however, can backfire entirely.
In 2008, then-candidate Barack Obama built the largest field operation in history, relying on both data-driven targeting and community-organizing tactics in a way that revolutionized presidential campaigning. But a study involving more than 56,000 targeted voters in Wisconsin found that a visit from a volunteer supporting Obama appears to have turned some potential voters away from Obama’s candidacy—in a state the Democrat won handily that year. The researchers suggested that people who rarely engaged in elections found the visits bothersome.
During the Obama era, Democrats relied on support from infrequent voters to capture the presidency, although they struggled in low-turnout, off-year elections. They poured millions of dollars into research and organizing programs to identify and mobilize those voters. But since then, the parties’ bases have shifted, and many of these hard-to-reach voters became Donald Trump supporters—especially working-class white voters and, in 2024, a large number of young and nonwhite people.
Some Democrats worry that their party’s vaunted turnout operation has, in recent years, produced a significant number of votes for Trump, reducing, if not negating, the benefits for their own candidates. Early last year, a top progressive data scientist warned donors in a memo that if Democratic mobilization groups “were to blindly register nonvoters,” they could be “distinctly aiding Trump’s quest for a personal dictatorship,” The Washington Post reported.
Radjy acknowledged that had been a concern, but she said Vote Forward’s postelection study found no evidence that its letter-writing campaign helped Trump or Republicans. “If we found that, it would hurt, but we would also share it transparently,” she told me.
It’s not clear that everyone else would. The biggest spenders in Democratic politics frequently test their turnout operations, in many cases through randomized controlled trials in which one group of people receives a particular form of engagement—a door knock, phone call, or text message, for example—while another gets nothing. (This is what Vote Forward did to test its letter-writing success.) After the election, organizers can check to see which group voted at a higher rate. These findings have shown that in presidential-election years, traditional canvassing methods have become less effective as voters get bombarded with campaign ads and reminders to vote. “In a saturated environment, it’s getting harder and harder for individual pieces of campaign communication to break through,” David Broockman, a political scientist at UC Berkeley who studies voting behavior, told me. “I expect the effects of everything are just going to keep on going down.”
Occasionally, the studies that groups conduct are widely shared, but some political organizations suffer from a phenomenon known as the “file-drawer problem”: “A lot of bad results never see the light of day,” Joshua Kalla, a political scientist at Yale University who studies voter persuasion, told me.
Wood, the data scientist, learned that firsthand. He told me he’s worked with Democrats who have urged him not to publish studies with unfavorable findings: “Basically the attitude is, There’s a lot of hype and a lot of willingness to fund this work. And if you put this out, all the funders are going to clam up and point to this as a reason not to do it.” In other cases, he said, clients have misused data to make tactics seem more effective than they really are.
Another researcher, who spoke on the condition of anonymity to avoid alienating allies in the party, told me about working on a study that found a campaign tactic had produced no boost in turnout. When the researcher later saw a published version of the report with their name attached, however, the findings made it seem as if the experiment had been successful. “The big problem,” the researcher told me, is that in addition to using research to improve campaigning, Democratic groups “also use it as effective marketing or to try to get clients. People’s incentives are misaligned.”
Democrats have become much more sophisticated over the past decade in understanding how to assess the effectiveness of campaigns, said Yoni Landau, the CEO of Movement Labs, an anti-Trump operation that ran dozens of large-scale experiments last year. “The challenge now is about political will,” he told me, “whether the people making the decisions—the funders and the organization leaders—want to know whether it worked.” To incentivize rigorous studies, which can help address the file-drawer problem, Landau said Movement Labs is launching a program it’s calling the Prove It Prize, which will encourage groups to test campaign tactics by offering money for experiments that produce positive results. For now, he said, many of the largest investments aren’t tested, and the reluctance to share poor results remains “very prevalent.”
When I called around to some of the largest progressive campaign organizations, most of them told me they had done extensive studies on their field programs in 2024, or were in the process of conducting them. Hardly any would share details of what they learned. Jenny Lawson, the executive director of Planned Parenthood Votes, told me the group would not risk sharing “trade secrets with political entities that exist to end Planned Parenthood.” An official with another major group plainly acknowledged, on the condition of anonymity, that it feared a loss of donations and was unlikely to publish a study showing poor results. A spokesperson for the Democratic National Committee told me it is conducting its own extensive postelection audit, incorporating “insights from inside the DNC and from external partners in the ecosystem” that the committee will make public in the coming months.
Many progressive groups, including Planned Parenthood, do submit their findings to the Analyst Institute, an organization founded in 2007 that both runs and collects experiments on voter-contact programs. The institute serves as a database for Democratic-aligned groups to share research on campaign tactics—successes as well as failures. But some people told me the party’s file-drawer problem extended there too. Christina Coloroso, the Analyst Institute’s executive director, told me its officials coach Democratic organizations to not expect huge positive results in presidential-campaign years. She acknowledged that groups can be reluctant to share data even within the Democratic community “when the results don’t look great,” but she said the institute allows its members to submit research anonymously to allay fears. “It’s true that we may not see every single test that exists across the ecosystem, but all the work that we do is to try to get to a critical mass of studies,” Coloroso said.
The search for the decisive edge in political campaigns has always been a hunt for novelty. Any new tactic that works doesn’t work that well for long. Everybody starts doing it. Voters get tired of—and sometimes quite annoyed at—the calls, the texts, the emails. “The first time that people got direct mail, it was like printing money,” recalled Michael Podhorzer, a former political director of the AFL-CIO who has been working on campaigns since the 1970s. “Oh my God. I just got this letter from George McGovern or from Ronald Reagan. I’m going to read it, and I’m going to send a check here.”
A generation ago, MoveOn.org helped pioneer the use of email to raise money and drive engagement, Podhorzer said. “Then it’s quickly like, Who opens an email?” More recently, the new thing was text messages, which took off in 2020, when Democrats in particular relied more on digital communications—and old-fashioned letter writing. “You just keep finding some way that people aren’t expecting to hear about politics, and so they are actually open to it and listen to you. But then it gets completely swamped,” Podhorzer said.
Conventional turnout methods—door knocking and phone calls, for example—can still have a big impact in low-turnout races, such as primaries, special elections, and campaigns for local office. But with the parties now spending more than $1 billion on the presidential campaign every four years, they’ve seen diminishing returns on each individual mobilization tactic. Vote Forward emerged out of a letter-writing experiment conducted during the 2017 special Senate election in Alabama, a deep-red state where the Democrat Doug Jones narrowly defeated Roy Moore, a former judge who had been accused of sexual assault or misconduct by several women. The turnout rate for people who received handwritten messages was three points higher than for those who did not. “That was the holy cow,” Radjy said. “This is a tactic that can really, really move the needle.”
The impact of the group’s letter-writing program has decreased over time, Radjy told me. Vote Forward found that its letters had no effect on the initial group of “surge voters,” people who had participated in at least one major election since 2016. But the organization was able to expand its program to additional groups, mainly newly registered voters. Among these groups, the campaign boosted turnout by 0.16 percentage points, enough for Radjy to consider that part of the effort a success, because it was similar to the average effect for all previous measured presidential-election turnout programs.
Vote Forward estimates that it drove an additional 9,000 voters to the polls nationwide. As paltry as that number might seem, it’s larger than the total margin of victory in the battle for control of the House during each of the past two elections. The letter-writing program is also relatively inexpensive, costing about $175,000. The group has concluded that although it will still use the tactic in small campaigns, it likely will not do so in the same way in 2028.
Democrats can take some solace in the fact that the nation’s rightward shift last year was much smaller in the states where they campaigned most aggressively. That suggests that the hundreds of millions of dollars they poured into advertising and voter-turnout efforts did make a difference. And even the best ground game cannot overcome a flawed candidate or message.
But the party’s defeat is accelerating a broader questioning of its organizing and ability to connect with the millions of voters who are up for grabs in presidential-election years. “Democrats have much bigger problems on their hands than what they’re doing on the doors at the end of the election,” said Billy Wimsatt, the founder of the progressive Movement Voter Project, a clearinghouse for donors to Democratic groups. He said the party needs to learn from the success of the well-funded MAGA movement, which he calls a “vertically integrated meta church” that, “feels like one big purpose-driven team,” even with all its faults. “Their billionaires are savvier than our billionaires,” Wimsatt told me, “and they’re more interested in winning.”
Wimsatt is one of many Democrats who believe that the party needs to invest in much deeper engagement with voters—outreach that must start long before an election. So does Radjy: “We need to be talking to people earlier,” she said. “We need to be talking to people in a more curious and reciprocal way.” But first comes honesty about what went wrong in 2024. Democrats will appreciate it. They might even demand it. “Even candor that is not rosy,” Radjy told me, “is more appealing than rosy bullshit.”
The post The Real Problem With the Democrats’ Ground Game appeared first on The Atlantic.
