How Trump is giving the labor movement the blue-collar blues

Compared to Biden in 2020, Harris lost ground this year among the most broadly defined group of voters with ties to labor unions. The exit poll measures the voting behavior of union households, which includes both the members of labor unions and people who live with them. Among those union households, which comprised about one-fifth of all voters, the exits found Harris beat Trump by only 8 percentage points (53% to 45%), half of Biden’s 16-point advantage in 2020. VoteCast recorded a similar 10-point advantage for Harris over Trump among union households, down from Biden’s 14-point edge four years ago.

VoteCast also separates the results between the actual union members and others in their households. Through that lens, it found Harris very slightly improving on Biden’s 2020 performance among union members (winning 57% of them, compared to Biden’s 56% in 2020) but losing substantial ground among the other residents in union households.

But these overall results obscure the continued impact within union ranks of the educational polarization that has reconfigured the broader electorate, particularly since Trump’s emergence as the GOP’s leader in 2016.

In the exit polls, Harris beat Trump among college-educated White voters in union households by a resounding 67% to 32%; that was up from Biden’s 27-point advantage among such voters in 2020, and more than four times Hillary Clinton’s edge over Trump with them in 2016, according to results provided by CNN’s polling unit.

In almost mirror image, Trump routed Harris by 62% to 36% among White voters without a college degree in union households, a substantial deterioration for Democrats relative to Trump’s 17-point lead with those unionized blue-collar voters in 2020 and a comparable deficit to Trump’s 28-point lead over Hillary Clinton with them in 2016.

According to the exit polls, Harris also won Latinos in union households by just 9 percentage points, way down from a 44-point advantage among them for Biden and 22-point lead for Clinton. Harris carried just under four-in-five Black voters in union households; that was down from 2016, when more than nine-in-ten of them voted for Clinton. (Though the exit polls did not have a large enough sample to report separate results for Black union members in 2020, other results suggest Biden won over 85% of them.)

Harris, the exit poll found, also surrendered substantial ground among male voters in union households, losing them by 9 percentage points after Biden had won them by 10 and Clinton by 3. In sharp contrast, the surveys found Harris won female voters in union households by 28 percentage points, up slightly from Biden in 2020 (24 points) and significantly from Clinton in 2016 (15 points).

VoteCast produced very similar results among these groups of union households. VoteCast’s results among actual union members, tellingly, followed the same tracks. It found that Harris lost non-college White men who personally belong to labor unions by 29 percentage points, up from Biden’s 25-point deficit in 2020. Though Biden in 2020 won White women without a college degree who belonged to unions by 6 points, Harris lost them by 8, VoteCast found. The Democratic lead also notably declined among Black union members (from 82 points in 2020 to 61 in 2024) and Latinos in labor unions (from a 28-point edge last time to just 12 in 2024.)

By contrast, in the VoteCast results, college educated White women in labor unions preferred Harris by 44 percentage points (up from Biden’s 38-point lead in 2020). She also carried unionized college-educated White men by a comfortable 23 points, after VoteCast had showed them mostly supporting Trump in 2020.

“A lot of this is just the same cultural differences that we see outside the labor movement,” said Guy Molyneux, a Democratic pollster who has worked extensively for labor unions. “I think mostly this is just a story that unions can still have an influence with their members, but they can’t overcome this kind of social and cultural change.”

John Brabender, a top media adviser for the Trump campaign, is experienced in courting blue-collar voters for the GOP through his work in Pennsylvania politics. He says the economy was the most powerful force moving union members without a college degree toward Trump, but that concern about immigration and cultural change powerfully contributed, too. “Certainly the economy was the biggest part of their (choice), but they were feeling that they had blinked and the country they knew had changed overnight,” Brabender said.

These stark patterns extended into the key states that decided the election. Both VoteCast and the exit polls found that Harris won only about 35% to 40% of non-college Whites in union households in Michigan, Pennsylvania and Wisconsin. Among Nevada Latinos in union households, the VoteCast found Harris won just over half, while the exit poll results suggest she carried just below half of them – in either case, too few to win the state.

Union officials generally maintain that they did better with their members than the exit polls and VoteCast survey found. Pappageorge told me he believes Harris won about 7-in-10 of their largely Latino membership in Nevada. Weingarten said that AFT’s research showed the vice president capturing about the same share of their members. Ron Bieber, president of the Michigan AFL-CIO, told me that he did not feel the stark educational divide that the polls found among union households in that state. “It felt pretty good all the way through,” Bieber said. “I never got any reports of real trouble spots.”

But several officials who have worked in labor union politics agreed the widening educational divide even within union ranks captured in the exit polls and VoteCast tracks with their own experiences. “It doesn’t surprise me,” Weingarten said of the exit poll and VoteCast results. Based on their own internal research, she said, Trump ran better among lower-income AFT members than those who earned more.

Michael Podhorzer, a progressive strategist who previously served as the AFL-CIO’s long-time political director, says that most White labor union members without college degrees have been supporting Republican presidential candidates since around 2008. The gap between Democratic support among union members with and without a college degree, Podhorzer said, “wasn’t as big … 10 to 15 years ago” but “there was definitely always a difference.” Data provided by the CNN polling unit from earlier exit polls support his assertion: Those results show that Republican presidential candidates have carried most non-college White voters in union households in each election since 2004, but Trump’s margins have significantly widened the narrow advantage enjoyed by George W. Bush in that campaign (2 points) and John McCain in 2008 (6 points).

Podhorzer tends to see a silver lining in the 2024 numbers. He notes that unionized White men and women without a college degree, though mostly supporting Trump, still voted for Harris at somewhat higher levels (6 points for men, 9 for women) than the comparable voters who were not in union households. “To be clear there’s nothing for Democrats to be proud of” in the results among unionized non-college Whites, Podhorzer said. “But in a country where the Democrats have been doing increasingly (poorly) there, that’s a pretty big premium that being in a union still brings.”

Steve Rosenthal, another former AFL-CIO political director, also argued that, for the most part, “The only White working class voters who are still voting Democratic are in unions.” But he acknowledges that the combined effect of diminishing union presence in the labor force, combined with GOP gains among blue-collar union members, has weakened the movement’s ability to tip even the states where they remain strongest, like the key Rust Belt battlefields. With the combined effect of both of those dynamics, Rosenthal said, over the past few elections “you are looking at Democrats in a state like Michigan losing something like 450,000 union household votes for the top of the ticket.”

In important ways, the widening educational polarization evident in these results reflects the changing dynamics of the labor movement itself.

In recent decades, the most vibrant unions generally have been those representing workers in the public sector, such as teachers and government employees. Those public sector union workers are more likely to be female and hold college degrees than those in private sector construction or manufacturing unions, who are more likely to be men without college degrees. Union officials I spoke with acknowledge that Trump ran very strongly among private sector union members, while Harris led more comfortably among those in the public sector.

Rosenthal points to another internal union trend also cited by Harvard sociologist Theda Skocpol and co-author Lainey Newman in their recent book, “Rust Belt Union Blues”: As union membership has declined, so has the role of labor unions as a focal point of social life in blue collar communities.

“I like to tell a story about a local union hall in the Lehigh Valley in Pennsylvania that I visited in 2000,” said Rosenthal, who now runs a political program aimed largely at former union members. “They had a bowling alley and a bar. At one time, maybe 30% of the workers were in the union, another 10-12% in union households, and another 5-7% bowled in the bowling alley or hung out at the bar. We were reaching 50% of the voters in our echo chamber. Now we’ve lost that, and the right-wing echo chamber is much more powerful.”

What should be especially daunting for Democrats about Trump’s success with non-college union members and working class voters more broadly is that it came after the Biden-Harris administration had formulated, and largely executed, a systematic strategy to recover such voters.

Biden identified with unions more unreservedly than arguably any previous president. He was the first president to walk a picket line. Biden nominated union-friendly federal judges and members of the National Labor Relations Board, repeatedly took executive branch actions to bolster unions and workers (such as the Labor Department’s rule earlier this year extending overtime pay protections to many more workers), supported sweeping national labor law reform legislation (the Protecting Rights to Organize, or PRO, Act) and focused his broader economic agenda around bolstering economic opportunities for workers without college degrees. The principal domestic policy bills Biden signed – the bipartisan infrastructure bill and legislation promoting domestic production of clean energy and semiconductors – have touched off a boom in core blue-collar occupations. The White House recently announced that the economy has generated 1.6 million construction and manufacturing jobs under his presidency, while those bills have generated over $1 trillion in private sector investment and a doubling of spending on the construction of new manufacturing plants. All of these bills contain substantial incentives for companies to employ union workers.

As labor historian Erik Loomis wrote last year, “I would argue that Biden is the most pro-union president since Franklin D. Roosevelt.”

Harris also walked a picket line (while a US senator), cast the tie breaking vote for the legislation that sparked the clean energy investment boom and cast the deciding vote for legislation that saved pensions for millions of union workers. Trump, by comparison, appointed NLRB members and judges generally hostile to organized labor (a Trump-appointed judge recently blocked Biden’s overtime rule), has refused to rule out signing anti-union national right to work legislation and cheered Elon Musk during a joint appearance for breaking a strike by firing striking workers. (In an exception to this pattern, Trump’s second term labor secretary nominee, GOP Rep. Lori Chavez-DeRemer, is considered much more sympathetic to labor unions than most elected Republicans.)

Yet despite these stark differences on the priorities of organized labor, the VoteCast survey produced a striking verdict: Among White voters without a college degree in union households, it found that just 32% approved of Biden’s performance as president, while 58% retrospectively approved of Trump’s performance in office.

“To be honest, Biden has not been a credible messenger for a really long time,” Rosenthal said. “So despite the huge accomplishments, he was not communicating that. And I think it’s really difficult for a union to do that on its own. In the case of Harris, (rank-and-file union members) had no idea she cast the deciding vote on the pension bill, or she cast the vote that created tens of thousands of union jobs with the Inflation Reduction Act.”

Pappageorge said one issue essentially eclipsed everything else Biden delivered for working families and organized labor. “All of those things are important, but the No. 1 issue on the mind of working-class people, including union people, was inflation and the cost of living,” he said.

Pappageorge described Biden as “my hero” because of all of his accomplishments for working people, but he believes the president dug an insurmountable hole for the party by trying for months to emphasize the economy’s positive aspects rather than challenging companies raising prices and rents, as Harris forcefully did immediately after she entered the race. “Kamala had the best message we saw; we think her message was on point,” Pappageorge said. “If Biden had been running on Kamala’s program earlier, I think they would have reaped the benefit of that. But they didn’t. They tried to tell working class people the economy is setting records.”

Weingarten offered almost exactly the same diagnosis. Biden’s plan to build working-class support by promoting unions and creating non-college jobs “was a good theory,” she said. “But what Biden didn’t do was to feel the pain of people that it wasn’t happening fast enough. (Biden saying) ‘finish the job’ is different than being out there like FDR was in feeling people’s pain.” Harris’ shift in emphasis toward acknowledging the pain of inflation was valuable, Weingarten said, but came too late to change the outcome.

Brabender, the Trump media adviser, said the former president’s strong showing with blue-collar union members points toward a structural change that threatens the ability of virtually any organized constituency to influence its members’ choices in presidential races. Because voters have access to so much more information today, not only through traditional news sources but also via their social media connections, “it is much harder for any organization to have a singular message that trickles down to their members,” Brabender said. “They are hearing many more opinions than in years past. Let’s just say that union leaders’ megaphone isn’t as big and powerful as it used to be.”

Weingarten only disagrees with Brabender to a point. She argues that “our members trust their union” but that the education divide in support for Harris and Trump reflects a larger divergence, with Democrats running better among the voters who rely on (and trust) conventional news sources for their information, a group that tilts toward those with more education. “The education divide is also about what information is trusted … what news is trusted. It’s also about disinformation,” she said. “The divide is obvious, but it’s much deeper in terms of how you address it.”

For union leaders, as much as for Democratic elected officials and political strategists, addressing all the reasons for Trump’s gains with working-class voters looms as job one as he prepares to begin his second presidential term.