It was Memorial Day weekend, and Massachusetts Sen. Elizabeth Warren was telling a crowd in Ottumwa, Iowa, that the Democratic Party had a choice.
She said the party could nominate a Democratic candidate who would “scoop up all the money in the world and hold it there to run a bunch of television ads right at the end.”
“But that candidate sure isn’t going to help the team,” she said, referring to the Democratic Party.
There was another option, though: “You could spend the time right now building the grassroots operation. You can spend it on the face-to-face, person-to-person, neighbor-to-neighbor reaching out to people.”
She said such an operation would make a campaign resilient, able to weather nearly any storm the media or Republicans could throw at it.
“Every time the crazy news comes in, every time there’s some big dust-up that’s not true, we’ve got our grassroots operation already in place,” she told a crowd of 200 in the ballroom of the Hotel Ottumwa.
More than any other time, on that day in Ottumwa, Warren got explicit about her plan to win the Democratic nomination and the presidency by running what her aides often called a campaign they could be proud of.
That meant shunning traditional media consultants and pollsters, whose high fees had often drawn scorn from the leftists who doubted their effectiveness. It meant forgoing traditional fundraisers and big-money politics in favor of collecting millions of dollars from small donors online. It meant building a massive, unionized organizing operation to knock on doors and make personal connections with voters. It meant winning earned media by focusing on her policy plans, not on campaign tactics or the up-and-down of the polls.
Parts of this plan were successful, even wildly so: Warren’s core promise to combat corruption connected with voters, and she steadily climbed in the polls throughout the summer, becoming one of the leading candidates for the Democratic nomination in the early fall and peaking in October. And her decision to stop attending big-money fundraisers didn’t harm her campaign’s ability to rake in cash. She raised more than any candidate except two billionaire self-funders and Vermont Sen. Bernie Sanders, who also forgoes high-dollar fundraising.
Some elements of her plan, however, ended up being more fashionable than effective, or simply ill-suited for a fast-moving, chaotic primary in which undecided voters were jumping from candidate to candidate, often at the last minute. The high costs of Warren’s field operation ate into the budget for paid media. Her campaign’s desire for discipline sometimes meant missing a chance to win a news cycle. And a lack of polling meant the campaign seemed to see-saw from message to message during the final portion of the campaign in search of something that would stick, with messages that sometimes drifted away from Warren’s core image as a fighter.
And, as a candidate who could have been designed in a lab to set off alarm bells about electability — she was a woman like Hillary Clinton, a liberal like George McGovern, an intellectual like Adlai Stevenson, from Massachusetts like Michael Dukakis — Warren and her campaign struggled to pivot from talking about her ideas to proving she was Democrats’ best hope to tame and defeat Trump.
“I think that both [the] campaign and [Warren] put together a plan and a strategy that in almost any other election cycle would’ve been a winning strategy,” said Doug Rubin, a top official on billionaire Tom Steyer’s presidential campaign who helped lead Warren’s Senate bid in 2012. “The X factor in all of this is Donald Trump. What they didn’t take into account is the intense, overwhelming desire of Democrats, above anything else, to beat Donald Trump in 2020.”
Warren’s campaign confronted many variables far outside its control. Both a former vice president and the runner-up for the last presidential nomination joined a crowded field, instantly laying claim to large chunks of the vote. The electability obsession created even more barriers for a female candidate than usual, and Warren’s campaign rarely felt comfortable letting her go on the attack, fearing a sexist backlash. Her focus on plans meant she had to supply a level of policy detail the other candidates did not. A slump in the latter part of 2019 hurt her fundraising at a crucial point, and an impeachment trial froze the race and limited possibilities for a comeback.
This cycle, more than 20 campaigns tried to lift someone to the Democratic presidential nomination. All but one of them will lose, and Warren’s campaign was one of a handful that gave its candidate a true chance at victory. Interviews with more than a dozen Democratic operatives and elected officials nationally and in the early states, along with Warren allies and staffers, reveal how the senator’s most important plan came excruciatingly close to working, then heart-breakingly fell apart.
Warren was the first major candidate to enter the Democratic race, sending out a video to supporters and announcing an exploratory committee on New Year’s Eve 2018. For the first few months of the race, her campaign teetered. Her decision to release a DNA test aiming to prove she had Native American heritage went poorly, with her poll numbers sinking as Democrats wondered if she could bat away Trump’s “Pocahontas” attacks.
She quickly began building a large field staff in both Iowa and New Hampshire, hiring dozens of operatives and organizers in the first two states to vote. The cost of the campaign quickly grew, but her online fundraising did not, at least initially. Operatives chattered that she would run out of money and be forced to drop out in the summer.
Warren raised just $500,000 on the first day of her campaign, one-third of what California Sen. Kamala Harris brought in on the day of her campaign’s launch, and one-twelfth of what former Texas Rep. Beto O’Rourke raised. She raised just $6 million in the first quarter of the year while spending over $5 million.
Warren changed this by talking about policy, incessantly. Her plans ― to wipe out most (but not all) student debt, to impose a wealth tax on fortunes over $50 million, to break up major tech companies, to limit the influence of lobbyists ― drove the news cycle for months, leaving candidates with less of a policy focus grasping for ways to butt in.
Her team emphasized the plans were all based on a single idea. “America’s middle class is under attack,” Warren said in her launch video ― and those attacks were coming from corporations and the lobbyists who guaranteed their power in Washington.
The strategy worked. Voters, especially the college-educated, slowly moved to Warren, providing her with both their support in polling and their small-dollar donations. In the second quarter of the year, she banked $19 million, enough to expand her already expansive campaign. Slowly and steadily, she surpassed Harris, O’Rourke and even Sanders in polling, prompting her supporters to put turtle emojis into their Twitter handles. Warren drew jaw-dropping crowds in urban hubs ― 20,000 people in New York City, 12,000 in Minneapolis, and 5,000 in Austin.
“For the first 10 months, they ran the most disciplined communications operation I’ve ever seen,” said Rebecca Katz, a progressive Democratic strategist.
Warren’s position peaked with a late October Quinnipiac University survey showing her with the type of coalition necessary to win the nomination. She had 28% of the vote, seven percentage points ahead of former Vice President Joe Biden and 13 points ahead of Vermont Sen. Bernie Sanders, her fellow progressive. She won a near-majority of very liberal voters, was even with Sanders with voters under the age of 35 and had double the support of any other candidate among white voters with a college degree, and even led among white voters without one. (One weakness, which would later severely restrict her hopes of a comeback: She was in second place among Black voters, 22 points behind Biden.)
There is considerable debate about exactly how and precisely when things turned sour for Warren. But most of her allies and outside observers point to some combination of two people and one newspaper: Rep. Alexandria Ocasio-Cortez, Pete Buttigieg and The New York Times.
Buttigieg, the former mayor of South Bend, Indiana, was competing with Warren for white college-educated voters, and smartly took advantage of Warren’s weaknesses. While she had released plans on relatively obscure topics like limiting the influence of military contractors on the Pentagon and revitalizing the diplomatic corps, she had been vague on the most important issue for both Democratic primary voters and general election voters: health care.
Warren had tried to split the baby on the issue, saying she preferred Sanders’ “Medicare for All” plan but was open to other ideas on health care. “I support a lot of plans,” she said after the debate in September, a position that directly contrasted the specificity of her ideas on seemingly every other topic.
“Your signature is to have a plan for everything, except this,” Buttigieg said on the debate stage in Westerville, Ohio, an attack echoed by Biden and Minnesota Sen. Amy Klobuchar.
Warren’s team had actually been working on a detailed health care plan since early in the summer but had been unable to reach a consensus on the best ideas. Warren adjusted and then readjusted in the weeks after, first releasing a detailed plan on how to pay for Medicare for All, then releasing a plan to focus on more popular health reforms first and push for Medicare for All later in her first term.
But Buttigieg and Biden, buoyed by extensive health insurance industry advertising in Iowa, managed to convince moderate voters a candidate who backed a Medicare for All plan that banned private insurance could not win. (After the attacks, public polling showed voters continued to trust Warren on health care issues but thought she was less likely to beat Trump.)
The same night, The Washington Post reported that Rep. Alexandria Ocasio-Cortez, an icon for younger progressives, was backing Sanders. Warren’s team had hoped to sway the 30-year-old Ocasio-Cortez, who had volunteered for Sanders in 2016, or at least persuade her to keep neutral. But Ocasio-Cortez, inspired by Sanders’ recovery from a heart attack, endorsed him at a massive rally in New York City the next weekend.
Around the same time, Buttigieg and Sanders both began airing television ads in Iowa, which helped inflate their poll numbers even further, quickly creating a narrative that they were ascendent.
The Warren team, with much of its money tied up in a large staff and skeptical of the final impact of television ads months out from the election date, stuck to its plan and didn’t respond with ads of its own.
The combined effect was to eat away at the two key portions of Warren’s base: Buttigieg began pulling away white college-educated voters, while Sanders started consolidating the party’s left wing, especially among young voters.
A survey of college students conducted by the textbook rental company Chegg showed Warren’s support peaked Oct. 15 with 30% of the vote, just ahead of Sanders. By the time of the Iowa caucuses, Sanders was winning a majority.
A few weeks later, The Times released a poll showing Warren performing worse than Sanders or Biden in the three Midwestern states key to an Electoral College win over Trump. It didn’t matter that the results in Wisconsin or Pennsylvania were within the margin of error or that the result in Michigan was an outlier. The survey — and questions about Warren’s electability — dominated discussion on cable news and Twitter, two sources frequented by the highly informed, highly educated voters Warren did well with.
That meant the conversation was shifting to one of Warren’s least favorite topics: polls.
Warren and her team avoided poll conversations like a single mention of crosstabs would give them coronavirus. This was meant to insulate them from having to discuss the up-and-down, day-to-day swings of a primary. They didn’t talk about polling when they were winning; they didn’t talk about polling when they were losing.
One of the few times Warren did talk about polls was to swear them off. On that Memorial Day weekend in Ottumwa, Warren proudly told the crowd she hadn’t conducted a single national survey since starting her presidential run.
“I don’t do polls. You want to know how many polls I did before I ran for president?” Warren, who had used a polling firm during her two runs for Senate, asked rhetorically. “You want to know how many polls I’ve done since I started running for president? I know what I believe in, what I’m going to fight for.”
Warren’s position was an odd one, especially for an academic who used data to determine what was wrong in the American economy and how to fix it. Politicians in both parties rarely actually use polls to determine their stances on issues. Instead, they use surveys to figure out the best way to pitch their policy preferences to the electorate, how to persuade voters, and where to invest their campaign’s time and money.
“She was saying the right stuff, and if you ran a poll, the polling would confirm that,” said Matt Canter, a pollster at Global Strategy Group who worked at the Democratic Senatorial Campaign Committee when Warren first ran in 2012. “But polling would’ve helped answer a lot of the questions around her campaign, including why voters didn’t close ranks around her when she was driving the conversation.”
Instead of an outside polling firm, Warren had an eight-person in-house data and analytics team, some of whom had previously worked for prominent polling companies. And the campaign did conduct phone surveys ― just not through a polling firm. The surveys measured Warren’s favorability, standing in the race and the popularity of her various plans ― but only after those plans were released.
Still, the campaign sometimes seemed slow to react to bad political news. Even though Buttigieg’s health care onslaught against her began in October, it took her until December to actually fight back. She called on Buttigieg to release his bundlers, immediately putting him on the defensive about his donors.
Warren’s team quickly recognized concerns about her ability to beat Trump and tried to address them early on, releasing a memo highlighting her triumph over GOP Sen. Scott Brown in May and working to weave arguments for her ability to beat Trump into her day-to-day message. But her campaign didn’t address the concerns in a broad way with paid media until a week before the Iowa caucuses, when they released a television ad featuring talking heads on MSNBC and CNN discussing the Trump campaign’s fear of running against Warren.
“He’s done everything he can for the wealthy and well-connected,” Warren said of Trump at the end of the ad. “I’m Elizabeth Warren and I approved this message because I’m going to beat him.”
That ad aired just 942 times on broadcast television in Iowa, according to a Democrat tracking media buys. Another ad, narrated by the author Roxanne Gay and contrasting Warren and Trump’s life stories, aired just 517 times.
Meanwhile, an ad about how Warren would refuse to hand ambassadorships over to wealthy donors aired more than 2,000 times. The ad, meant to highlight how EU Ambassador Gordon Sondland had enabled Trump’s corruption in the Ukraine scandal, baffled the other campaigns, which believed it raised issues far removed from voters’ daily lives. The Warren campaign, however, believed it was the only ad in the race that spoke to a news cycle dominated by talk of impeachment.
Many Democratic operatives think Warren needed a heavier focus on her biography, including her upbringing in Oklahoma and how anger at the treatment of American families drove her switch from academic to politician.
“She was one of the only candidates who hadn’t been running for president since they were in kindergarten, and she clearly came to this decision much more recently,” Canter said. “I think illuminating that decision, and what led up to it, would’ve been very compelling to voters.”
In general, Warren spent a far smaller portion of her budget on paid media than the other campaigns. Of her television spending, the campaign says more than half went toward highlighting her biography, and a fifth went toward pushing her corruption message. Just 1% went toward arguing she could defeat Trump.
And though her team was open about its skepticism toward television ads, it didn’t make up the gap with digital ads.
In the two-month run-up to Iowa, Warren spent $185,000 on digital ads in the state, according to a Bully Pulpit Interactive tracker, compared to $200,000 from Buttigieg and $430,000 from Sanders. In New Hampshire, she spent even less, just $80,000 against Buttigieg’s $115,000 and Sanders’ $151,000.
Though the Warren campaign insists it did not have to deviate from its Iowa plan, it blamed a six-week fundraising downturn following the October debate for restricting plans to spend more in the remaining early states and Super Tuesday. Before Warren’s donors responded to a last-second plea for funds before the New Year, her top aides were fearful she would raise just $17 million in the fourth quarter. (She ended up raising $21 million.)
And team Warren stuck to its approach of exclusively discussing the candidate’s plan and message, which contrasted sharply with the Biden and Sanders campaigns. Biden talked at nearly every event about how well-positioned he was in Michigan, Pennsylvania and Wisconsin, and about how polls showed him defeating Trump. The Sanders campaign seemed obsessed with metrics ― number of donors, number of video views, polling ― proving how popular the Vermonter was.
While Warren’s team eventually highlighted the Trump campaign’s worries, it did little to promote the idea that the president feared Warren’s populism when the first reports on the topic popped up in July.
And when the chair of the Democratic Party in Michigan’s Macomb County ― arguably the quintessential swing county in American politics ― endorsed Warren less than two weeks before the caucuses, the campaign never even sent out a release about it.
The Warren campaign had planned to address the electability question once and for all with a rally in Madison, Wisconsin, in the run-up to the caucuses. Other campaigns had used similar visits to swing states to show they could generate the excitement needed to beat Trump, but Warren had mostly ignored the Blue Wall states in favor of visiting key Super Tuesday states like Texas and California.
But the impeachment trial intervened, consuming a majority of Warren’s time and forcing the campaign to abandon plans for the rally.
The campaign’s electability pitches would get more explicit — by the end, Warren would open her rallies by pledging to defeat Trump — but it was clear in the run-up to Iowa that she had not completed the sale.
“That’s the big elephant in the room,” said Patrick Peacock, a Davenport city council member who is Black and endorsed Warren a week before the caucuses. “I’m going to be honest with you: The question is whether or not white men will vote for her. Will white men cross over and vote for a white woman?”
The final weeks before Iowa saw Warren try a mishmash of messages in order to regain the footing she had lost. At one point in December, she delivered three different major campaign speeches in as many weeks, eventually settling on a pitch as a “unity candidate” who could bridge the divide between Sanders and Biden backers.
An endorsement from former candidate Julián Castro helped drive home the message, and Warren began playing up the fact that she had adopted ideas and hired staffers from the defeated campaigns of Harris, New York Sen. Kirsten Gillibrand and others.
Warren aides were confident they would eventually secure the endorsement of Washington state Gov. Jay Inslee to give the pitch additional heft. But Inslee, wary of looking like he had his eyes on a gig in the other Washington as he ran for reelection as governor, ultimately never offered it.
A portion of Warren allies disliked the “unity candidate” message, which they felt downplayed Warren’s history as a fighter who was willing to hold even her own party accountable.
And soon after, the campaign would be faced with a fight it swears it did not want. CNN reported that Sanders had told Warren in a private meeting late in 2018 that a woman could not win the presidency in 2020. The next night on the debate stage, Sanders denied saying it, and Warren confronted him afterward.
“I think you just called me a liar on national TV,” Warren told her ideological ally in a curt exchange picked up by the network’s microphones.
While operatives in both campaigns and unaligned progressives pushed to downplay the fight, tensions had been escalating for months before. Warren staffers eventually assembled a more than 40-page research document chronicling what they saw as unfair attacks on the senator from Sanders staffers and surrogates.
“I was told at the beginning of this whole undertaking that there are two lanes: a progressive lane that Bernie Sanders is the incumbent for and a moderate lane that Joe Biden is the incumbent for and there is no room for anyone else in this,” Warren said after the end of the campaign. “I thought it was possible that that wasn’t the case, that there was more room, and more room to run another kind of campaign. But evidently that wasn’t the case.”
The showdown between Sanders and Warren was one of the last stories to break through before the three-week impeachment trial began dominating the news cycle, leaving little chance for any campaign to change the narrative before voting began in Iowa.
In late January, the Warren campaign’s earliest decisions were finally put to the test. It had hired more than 1,000 staffers across 31 states, some of which aren’t scheduled to vote until mid-March, with 80% of the staffers dedicated to field organizing. Warren spent as much on payroll every five days as Klobuchar spent every month, according to FEC records. The campaign paid out biweekly, and with three paydays in January, needed to take out a line of credit at the end of the month.
The biggest operation was in Iowa, where polls showed Warren with a chance at victory.
But the plan had a problem. The point of field organizations is to identify committed supporters and bring them out to polls and caucus sites on Election Day. But a large portion of voters, especially the educated voters who had made up Warren’s base, were still undecided.
The entrance polls showed 36% of Iowa voters made up their minds in the final days of the election, compared to 16% in the 2016 race and 20% in the 2008 race.
“In a world where 50% of people are waffling on their decision until the day before an election, field organizations are less effective,” said Addisu Demissie, a Democratic operative who built a similarly large Iowa field operation as the campaign manager for New Jersey Sen. Cory Booker’s presidential bid.
The trend continued in New Hampshire, where 51% of primary voters decided in the final days.
Warren won 19% of the raw vote and finished third in Iowa, but struggled mightily in New Hampshire, turning in a middling debate performance and failing to net delegates.
Those two states were critical for Warren ― her campaign operations there were considered top-notch, and her brand of progressive politics was thought to be a good fit for both. And her struggles with voters of color meant she was likely to underperform in both Nevada and South Carolina, the next two states to vote.
Warren, a senator for a state that is more than 80% white, tried to reach out to voters of color by centering racial inequality in her plans and engaging activists, many of whom gushed about her to reporters with regularity. But it never translated to votes.
When Warren was at her peak, her campaign began discussing spending more time in South Carolina, wooing Black voters. But her standing in Iowa and New Hampshire grew worse, and the pivot never happened. It’s not clear it would’ve helped, considering Biden’s decades worth of ties to South Carolina.
“One campaign’s worth of outreach is not going to be able to overcome 40 years’ worth of relationships,” Demissie said. “Booker and Harris dealt with the same thing, even as African Americans.”
A brief ray of hope appeared in Las Vegas, where Warren essentially dismantled Michael Bloomberg and his $600 million campaign over the course of a two-hour debate. It also enabled her to aim more attacks at other candidates, with Warren strategists believing voters wouldn’t mind body blows to Klobuchar and Buttigieg as long as Bloomberg had a black eye at the end of the night. It energized Warren’s campaign and delighted her donors, helping her raise nearly $30 million in February.
“That was the final puzzle piece in terms of proving that she could take on Trump, and allowing people to really see what it would look like,” said Meredith Kelly, a Democratic strategist who had worked as communications director for Gillibrand’s campaign. “But it came too late.”
And so, after a disappointing Super Tuesday, Warren appeared before a swarm of cameras outside her home in Cambridge, Massachusetts, and ended her bid. Having ignored clarion calls to run for president in 2016 and insisted throughout the 2020 campaign she didn’t need to be president, she betrayed no regrets about her campaign’s tactics or strategy.
“I’m not sorry for one nickel I spent on field,” she told MSNBC’s Rachel Maddow on Thursday night, mere hours after she left the race. “If anyone does the woulda, shoulda, coulda, I [would] probably spent less money on television advertising and put more money into field because field is the person-to-person. It is the part about rebuilding our democracy.”
Her field offices, Warren explained, are where volunteers would come in, could eventually learn and get entry-level staff jobs, or start talking to each other about local issues like education funding or the city council.
“How about we get together and we repair this little bit of democracy?” Warren said, imagining a conversation between two volunteers.
It seemed like the kind of thing she could be proud of.
It was Memorial Day weekend, and Massachusetts Sen. Elizabeth Warren was telling a crowd in Ottumwa, Iowa, that the Democratic Party had a choice.