The New York Times reported Tuesday the good news that Kamala Harris leads in its polling for the first time since she entered the race. The Times/Siena poll is judged by FiveThirtyEight to be the most reliable, based on both its track record and its transparency. It is also one of the more helpful polls when you want to take a deep dive into the electorate, because it makes available to the public detailed responses from key subgroups among likely voters. The crosstabs in this latest Times/Siena poll offer something more surprising than the top line: mild encouragement that working-class voters, about whom I’ve been worrying a great deal, are finally warming to the Democratic candidate.
I say “mild” because there’s no evidence yet that working-class voters are warming to Harris in the key swing states of Pennsylvania, Michigan, and Wisconsin, which CNN has identified as “the most consistent tipping point in American politics.” During the past 32 years these states went Republican only once, when Trump won all three in 2016. If Harris wins these three, she wins the election. This isn’t her only path to the 270 electoral votes she needs to win (MSNBC has some alternative scenarios here), but it’s the easiest path. And among all seven swing states, Pennsylvania, Michigan, Wisconsin, and North Carolina have the highest proportion of blue-collar workers.
Back to the good news. The Times reported that Harris is leading nationally, and that she’s gained ground among some key constituencies, including older voters, who are hugely important because (if you define “older” as 50 years or more) they cast more than 60 percent of all ballots in 2024. Before he dropped out, Biden was leading with voters aged 65 and over, even though they usually lean conservative. (In 2020 they went for Trump.) But in mid-September the Times/Siena poll showed Harris losing this group to Trump by seven percentage points, 44–51. The latest poll shows Harris winning over-65s by two percentage points, 49–47. Hallelujah.
The Times did not report that Harris is gaining nationally among working-class voters, but she is. (Working-class voters are defined here conventionally as voters lacking a college degree.) Harris still lags Trump with this group in the Times/Siena poll, but since July she’s narrowed that gap from a dispiriting 15 percentage points to 11 percentage points. There’s still a lot of work to do, but as recently as mid-September, in a Times/Siena poll taken immediately after the presidential debate, Harris was losing this group to Trump by an alarming 18 percentage points.
Harris’s gain, surprisingly—though not if you’ve been following these minutiae since 2016—is not among nonwhite members of the working class. Because Harris is nonwhite herself, one would expect her to be increasing the Democrats’ traditional advantage here. But the nonwhite working class (principally Latino men) has been drifting toward Trump since 2016, in spite of Trump’s escalating denunciations of Latino immigrants. According to Times/Siena polls, Harris’s lead with nonwhite working-class voters (31 percentage points) is about where it stood in July (29 percentage points). Harris’s nonwhite working-class lead shrank briefly to 24 percentage points after her debate with Trump, but it’s now stabilized.
Where Harris is gaining some traction is with white working-class voters, a group that, with the sole exception of Bill Clinton in 1992, hasn’t gone for a Democratic presidential candidate since Lyndon Johnson. It’s doubtful that Harris will win white working-class voters outright in 2024, but since July she’s narrowed her deficit from 38 percentage points to 30 percentage points. As with Harris’s gain with working-class voters overall, the improvement is mostly recent; in mid-September, Harris’s deficit with white working class voters was 36 percentage points, and one week earlier (i.e., before her debate with Trump), Harris’s deficit was the same 36 percentage points. I would guess these white working-class voters are women.
Naturally, these polling results could be outliers—or even garbage. The death of the telephone call (which I eulogized shortly before the 2016 election) forced pollsters to alter their methods both before and after the 2020 election, and we won’t really know how everything worked out until next month. But if polls aren’t great with absolute numbers, they’re pretty good at identifying general directions, and the general direction of at least the white working class, which remains a plurality (if no longer the majority) of working-class voters, is shifting toward Harris.
Perhaps Latino males will follow. It may help that Trump turned down Harris’s challenge to debate again, since Latino males are the only demographic group that I’m aware of (judging from the nonwhite working class’s brief surge toward Trump in mid-September) that actually thinks Trump won the last one. This was the same debate, you’ll recall, in which Trump groused about “all the people that are pouring into our country and killing people.” The heart wants what it wants.
Because the new Times-Siena poll does not break out results by state, we don’t know whether in recent weeks working-class voters drifted toward Harris in Pennsylvania, Michigan, and Wisconsin to the same mildly encouraging extent they did nationally. An October 8 Wall Street Journal article by Ken Thomas and Catherine Lucey (“Kamala Harris Struggling to Break Through With Working Class, Democrats Fear”) reports that a private poll last week by Democratic Senator Tammy Baldwin, who’s running for reelection in Wisconsin, showed Baldwin up two points and Harris down three, with the difference attributed to working-class men.* The same story portrays Michigan Governor Gretchen Whitmer urging the Harris campaign to spend more time there; Harris’s last Michigan appearance was September 19. A September 30 New Yorker piece by Eyal Press about working-class voters in Pennsylvania has no new polling information and is mildly discouraging.
What we know from earlier Times/Siena polls is that working-class voters in these states were not moving toward Harris before October. A Times/Siena poll of likely voters in Michigan released September 28 showed virtually the same deficit (45–50 percent) that Harris had in a Times/Siena poll released August 10 (45–51 percent). In Pennsylvania, Harris did slightly worse in a Times/Siena poll released September 19 (42–55 percent) than she did in a Times/Siena poll released August 10 (44–52 percent), possibly because of the same peculiar Trump-debate bump from Latino men nationally.
In Wisconsin, Harris lost support from working-class voters in a Times/Siena poll released September 28 (42–54 percent) compared to a Times/Siena poll released August 10 (47–51 percent). Perhaps this too was a debate bump. But in all three states we know of no working-class movement toward Harris. Maybe it’s too recent to have shown up even by late September.
Multiple polls now show Harris gaining on Trump, and sometimes beating him, on the question of economic stewardship. Perhaps that’s why she’s gaining a bit among white working-class voters. There’s growing reason to believe Harris can win enough working-class support to win the presidency. But her pitch to these voters still needs to improve. Perhaps the small encouragement the new Times/Siena poll offers about the working-class vote will inspire her to do so.
* This post originally misidentified Tammy Baldwin.
