These Working-Class Populists Are Struggling to Win Over the Working Class

So far, the 2026 midterm elections are shaping up to be a Democratic wave. That’s obviously great for the party. But an election in which anyone with a “D” next to their name stands a good chance of winning doesn’t tell us all that much about which kinds of campaigns are best at reaching the working-class voters Democrats need to secure durable national majorities, rather than more fleeting ones that, like a wave, can recede as quickly as they appear.

Over the past decade, the party has steadily lost ground with working-class voters, and a growing number of commentators have argued that the way to win them back is to run working-class candidates on an economic populist message. We at the Center for Working-Class Politics, in partnership with The Nation and Nira Data, have been trying to test this question on the 2026 campaign trail. That’s why, last month, we launched the Economic Populist Tracker, in which we follow four working-class populist candidates running this year in places Donald Trump won in 2024—Dan Osborn, an independent running for Senate in Nebraska, and three House candidates, Brian Poindexter in Ohio’s Seventh Congressional District, Bob Brooks in Pennsylvania’s seventh, and Rebecca Cooke in Wisconsin’s third.

While three of the four candidates we’re tracking hold slim leads over their Republican opponents, in a wave year victories alone may not tell us much about whether the candidates’ working-class, economic populist appeals made any difference. Instead, we have to look at whether these candidates do better than a generic Democrat would in their districts.

Two months into our tracking, the answer is mixed. Cooke and Osborn (who, though he is technically rejecting the Democratic Party label, is running with the Nebraska Democratic Party’s support) are running a few points ahead of the Democratic Party, Poindexter is even, and Brooks is slightly behind. The two who are running ahead of the party label are doing it mainly thanks to their support among independents and Democrats—plus, in Osborn’s case, a small number of Republicans. Ironically, though, their edge over the party does not come from working-class voters: None of the four is yet doing measurably better than the party among voters without a college degree in households earning less than $100,000.

Challenges Ahead?

In our September survey, Osborn trailed GOP incumbent Pete Ricketts among likely voters by eight and a half points, compared with 10 points in August. Poindexter led Max Miller by five points (down from six in August), Brooks led Ryan Mackenzie by two points (up from one in August), and Cooke held a slight edge over Derrick Van Orden by two points (down from six in August). None of these figures changed by more than the margin of error.

But steady overall numbers can mask changes that may matter in the home stretch. Here is a brief look at how each race has developed.

Ohio’s seventh district. Poindexter still leads, but Republican voters are rallying to Miller, who was initially damaged by abuse accusations from his ex-wife (accusations he denies). Miller responded to his own scandal by dredging up a 1998 incident in which Poindexter, then 18, was arrested but never charged, and he seems to have won back some of his base. In September, more of the district’s Trump voters called Poindexter “too extreme” compared with other Democrats (27 percent, up from 13), and Miller’s support among Trump voters rose from 44 to 54 percent, mostly as undecided Trump voters came off the fence. Poindexter’s own support among Trump voters held steady at about 10 percent, and his overall lead remains essentially unchanged. But 44 percent of the seventh’s undecided likely voters are Republicans, while 23 percent are Democrats and 13 percent are independents. As the district’s Republican-leaning undecideds make up their minds, Poindexter’s lead could narrow to well within the margin of error.

Nebraska. Ricketts spent September portraying Osborn as a Democrat in disguise, calling him “Democrat Dan” about 40 times at their September 1 debate. The attacks, along with leaked recordings in which Osborn appeared to contradict his stated position on immigration and mocked a small-town Trump supporter, appear to have hurt Osborn’s standing with Republicans.

In a state Trump won by 20 points, Osborn cannot win without a sizable share of Republican votes. He won about one in eight Trump voters in his 2024 loss and likely needs closer to one in five; right now he has about 7 percent.

Among Nebraska’s Trump voters, the share calling Osborn “too extreme” in our September tracking rose from 27 to 42 percent, and among those with an opinion of him, the share with a favorable view fell from 21 to 13 percent. Undecided voters won’t close that gap on their own. Nebraska’s undecided likely voters are 44 percent Republican, 24 percent Democratic, and 14 percent independents. Even if undecided Democrats and independents break his way, Osborn would still trail unless he won over most of the undecided Republicans. His other option is Nebraskans who aren’t sure they will vote, a group about a fifth the size of the likely electorate. They are far less Republican than likely voters (29 percent, compared with 55 percent), so bringing them to the polls could help him.

Pennsylvania’s seventh district. Brooks is picking up steam among Democrats, and undecided voters lean in his favor. In September, his campaign put his working-class biography at the center of its messaging, including an ad featuring Brooks decked out in firefighter gear promising to “smoke out the billionaires and their lobbyists,” as well as the creation of a Latino outreach coalition, “Bombero Bob.” Brooks’s support among likely-voting Democrats rose from 63 to 72 percent as more Democrats became familiar with him, though small losses elsewhere kept his overall lead about the same. More working-class voters are also learning about Brooks’s working-class background: among voters with an opinion of the candidate, the share who say he did working-class work before politics rose from 46 to 60 percent—mostly among voters without a college degree, although that hasn’t yet produced an increase in support among non-college-educated voters in the district. Crucially, however, Pennsylvania’s seventh is the only one of the four races where undecided Democrats outnumber undecided Republicans, 36 to 27 percent, and another 18 percent of undecideds are independents. If each of these groups breaks the way voters in the same group have who’ve already made up their mind, Brooks’s lead would grow to about five points.

Wisconsin’s third district. This is the one race we’re tracking where the challenger appears to be losing ground with the Democratic base. It also shows that a working-class image can help Republicans as well as Democrats. Van Orden spent the month of September attacking Cooke’s working-class credentials, calling her a “phony” who “occasionally waits a table.” In August, Wisconsin voters were about equally likely to say that Cooke and Van Orden had done working-class work before politics; by September, 37 percent said so of Van Orden, a retired Navy SEAL, and 30 percent of Cooke. Over the same month, Cooke’s overall lead narrowed from six points to two. Wisconsin has the fewest undecided voters of the four races, and they are 31 percent Republican, 24 percent Democratic, and 16 percent independents who lean toward neither party. If they break the way decided voters in their party have, Cooke would still eke out a narrow lead.

Are the Populists Doing Better Than a Generic Democrat Would?

Even if all four candidates win, however, it won’t be clear that their particular blend of working-class economic populism was the driving force. In a wave year, any other Democrat might have done just as well. To assess this, we compare each candidate’s support with which party the same voters say they will support in the 2028 presidential election. Brooks, for instance, trails the Democratic Party by about three points on this measure but still leads Mackenzie. What the comparison tells us is how much support each candidate is getting beyond the partisan lean of their districts.

Here the results are surprising. These candidates were recruited on the theory that working-class backgrounds and economic populist messages would help them pick up more working-class voters. Two of the four are indeed running ahead of a generic Democrat, but that additional support is coming from independents, Democrats and, in Osborn’s case, affluent college graduates, not from working-class voters.

Here again, little has changed between the two months. Cooke was about four points ahead of the party in both August and September, and Osborn about three points ahead in August and two in September. Poindexter matched the party in both months, and Brooks sat about three points behind. Because this picture has stayed roughly stable since August, below we combine our data for August and September to look more closely at which groups of voters are driving these differences.

Independents, who make up about one in 10 voters in each district, are the clearest source of the populists’ edge. Among them, Osborn outpaces the Democratic Party by about 21 points, Cooke by about 20 and Poindexter by about nine, while Brooks does no better. Put simply, if Cooke and Osborn did only as well as the party among these independents, much of their advantage would disappear: Cooke’s four-point edge over the party would shrink to about two points, and Osborn’s three-point edge to nearly zero.

The candidates differ most in their support among Democrats. Cooke does about eight points better than the party among 2024 Harris voters. Brooks lags the party by about nine points among Democrats. Because Democrats make up only about a third of his district’s voters, that shortfall works out to about three points overall, and since he is about even with the party among everyone else, it accounts for virtually his entire deficit relative to the party brand. Poindexter falls slightly short of the Democratic Party among Democrats, which offsets his gains among independents and leaves him even with the party overall.

Osborn is about six points behind the party among Democrats, but he makes that up elsewhere: He is the only one of the four who outperforms the Democratic brand among Trump voters, by about three points, and among people who voted for a third-party candidate in 2024, by about 13. The other three do about as well among Republicans as the Democratic Party.

Our findings on working-class voters are the most sobering for the economic populist theory of victory. Combining both months gives us enough respondents to split each district’s voters by education and household income as well as gender. None of the four candidates is measurably ahead of the party among voters without a college degree in households earning less than $100,000. Among this group, Cooke is about three points ahead, Osborn is even, and Poindexter and Brooks trail by about three to four points, all within the margin of error.

The candidates’ performance among working-class voters splits sharply by gender. Cooke does better with working-class women than with working-class men: Among women without a college degree in households earning less than $100,000, she is nearly eight points ahead of the party, while among men in the same group she is even. Brooks and Osborn show the reverse pattern: Brooks trails the party by about six points among women without a degree in lower-income households, and by about 10 points among lower-income women with a degree, while among working-class men he is about even with it. Osborn sits about four points below the party among women without a degree in lower-income households. His advantage comes instead from affluent voters: He outperforms the party by about 10 points among college graduates in households earning $100,000 or more, and by about 14 points among the men in that group. Poindexter’s differences among all these groups are small, generally within about three points of the party label.

Putting these pieces together, the populists who are running ahead of the Democratic Party are doing it mainly by winning independents who don’t lean toward either party, and by how effectively they inspire Democratic voters. But they are not outperforming the party with working-class voters: Among voters without a college degree in lower-income households, all four are doing about as well as a generic Democrat would, with the exception of Cooke’s strength among working-class women. Republicans make a difference only in Osborn’s race, where he is running a few points better than the party among Trump voters.

The Road to November

The populists outperforming their party are doing it mostly with independents who lean toward neither party. But winning in November depends on who is still undecided, and in every race most undecided voters are Democrats or Republicans, as independents who lean toward neither party make up only 13 to 18 percent of undecided voters. That makes undecided Democrats the most important target for the three House candidates. Brooks, whose undecided voters include more Democrats than Republicans, can grow his lead simply by becoming better known among the Democratic base. Poindexter and Cooke will need to hold on to undecided Democrats as well as independents to stay ahead as undecided Republicans settle on Miller and Van Orden.

Osborn is the only one of the four who cannot win without substantially more Republican support. For the three House candidates, the weeks until November 3 are likely to turn on Democrats who still aren’t familiar Brooks and Poindexter, working-class men in Cooke’s district as Van Orden claims the working-class mantle, and non-college-educated voters in Brooks’s district who now know he was a firefighter but haven’t yet moved into his column.

Each of the four economic populists we’re tracking has a real shot at victory in November. But even if they win, the bigger test for the theory of the case of progressive economic populism is whether they can deliver working-class votes above and beyond what any Democrat could in a highly favorable electoral context. That remains to be seen.

Methodology: The CWCP/Nira Data tracker interviewed 8,100 respondents from July 31 to August 25 and a separate sample of 9,510 from August 25 to September 24. Respondents were recruited through targeted online ads and assigned to districts by their self-reported zip code and county. Each sample is weighted to district demographics and 2024 presidential vote. Horse-race figures come from the September survey. Where we combine the two months to study smaller groups or small differences, we re-weight the combined sample. “Likely voters” say they will definitely or probably vote. In describing undecided voters, we count independents who lean toward a party as members of that party. In the September survey, the margin of error for each candidate’s share among likely voters ranges from about two and a half points in Wisconsin to four points in Ohio. We describe changes between August and September as changes only when they are larger than the margin of error, unless we say otherwise.