Thursday’s Democratic debate in Los Angeles may be the last chance for Elizabeth Warren to reverse her slide. After peaking a few months ago, she hasn’t fallen as far or as fast as Kamala Harris, who exited the race earlier this month. But she’s losing educated white professionals to Pete Buttigieg while Bernie Sanders reclaims progressives disillusioned by Warren’s hedging on Medicare for All. 

In fairness, the four leading candidates are all within the margin of error, and the race is still in flux. Iowa and New Hampshire are make-or-break for Warren. If she loses both, it’s hard to see a pathway back. 

Until recently, Warren did everything right. She has a substantive and well-organized campaign. She is an energetic and focused campaigner. And she is slowly shifting her image from Harvard intellectual to Oklahoma public-schooler who for a time was a single mom trying to make ends meet.

She has a plan for every problem, a pledge that defines her campaign along with the four-hour selfie lines she does with voters. One of those plans, Medicare for All, and her inability to tamp down criticism from the left and the right, is contributing to her slide in the polls.

“The factor that propelled her success, her policy specificity, became a double-edged sword,” says Bill Galston, a senior fellow in the governance program at the Brookings Institution.

It began when she was pressed by the other candidates, and by the media, to say how she would pay for Medicare for All without raising taxes on the middle class. Warren put out a policy paper with numbers critics derided as “fuzzy math.” That only fueled more skepticism about the feasibility of a proposal that she said would cost an eye-popping $20.5 trillion over a decade. 

Another candidate could have taken the punch, Galston told The Daily Beast. “But not when policy specificity is your stock in trade.”

By contrast, Sanders has said that taxes would rise on the middle class to pay for Medicare for all. He gets credit for being forthcoming without releasing much additional detail.

At the same time, voters leaving Warren for Buttigieg cite fears about her electability while skating over several obvious flaws that he has in winning a national election. (He won the mayor’s race in South Bend with 8,000 votes, not by 8,000 votes; he has minimal traction in the black community, and he would also test cultural mores as a gay married man.) 

Voters are still shopping around in the early states, and perhaps not surprisingly, Sanders’ supporters are the most likely to say, by a ratio of almost 2-to-1, that they are committed to Sanders and not likely to change their mind. Warren’s numbers are in the high 20s to low 30s when it comes to that kind of iron-clad commitment, says Galston. “To some extent, it was easy come, easy go,” for Warren, he says. “Now the question is what’s the next act?”

Warren gained steady support because she had plans to address real problems that people care about, and she put a lot of emphasis on her ground-level organization, rather than relying on television ads. The popular assumption was that Sanders would eventually pass the baton to her, but her slippage in the polls suggests that it could be the other way around.

The results in last week’s British election had their repercussions here as well. Third Way, a moderate Democratic group, put out a statement without mentioning Warren or Sanders by name, warning of the risks of “mimicking” the British Labour Party’s “theory that left-wing populism can win.” Working-class districts that hadn’t voted Conservative in decades broke the logjam over Brexit and gave Boris Johnson, a Trump lookalike and ally, a resounding victory. 

Matt Bennett, a co-founder of Third Way, told the Daily Beast that the British Brexit referendum vote was “clearly a harbinger in 2016 (foreshadowing Donald Trump’s election) and could be again. There’s no doubt British voters rejected the maximalist approach and left-wing populism and chose a right-wing populism, which has enormous implications for us,” says Bennett. “There’s still time to do things differently,” he added, noting that Warren has begun talking about choice in her Medicare for All plan.

Bennett cites Warren’s stumble on health care, the No. 1 issue for voters, as the reason she’s lost ground. “It’s the Godzilla of issues, and she got crossways with it,” he says. 

If the debate goes ahead as planned Thursday, and isn’t canceled over an ongoing labor dispute, we will see how Warren adapts to the speed bumps in her campaign. “The attack mode is not her best mode,” says Galston. “She simply has to try to reinforce what she is, and if people like it, fine, and if they don’t, well....” That’s good advice for Warren, who has soldiered through earlier critiques that wrote off her candidacy only to rebound and become one of the top contenders.