The replacement of the boring old man with the largely unknown younger woman has energized the Democratic electorate and completely changed the dynamic of the race. While I had been reluctant to read much into a handful of one-off polls, I agree with Ed Kilgore that “Kamala’s Gains Are Now a Trend, Not Just a Bounce.”
According to the FiveThirtyEight national polling averages, Harris is leading Trump by 1.4 percent (45.0 to 43.6 percent), with Robert F. Kennedy Jr. at 5.6 percent. When Biden dropped out, he was trailing in the same averages by 3.2 percent. In a contest as static as the 2024 presidential race had been, that’s a big swing.
The trend lines in national polls are equally telling. YouGov/Economist tested Harris against Trump back on July 16, showing Trump leading by five percent (44 to 39 percent). Then on July 23, after Biden’s withdrawal, the same pollster had Trump leading Harris by three percent (44 to 41 percent). On July 30 YouGov/Economist showed Harris leading Trump by two percent (46 to 44 percent). Similarly, RMG Research showed Trump leading Harris by two points (48 to 46 percent) on July 23, with Harris leading Trump by five points (47 to 42 percent) on July 31.
Polls comparing the Harris-Trump matchup to the earlier Biden-Trump matchup mostly show the same pro-Democratic trend. On July 17, the Morning Consult tracking poll had Trump leading Biden by four points (46 percent to 42 percent). On July 24, the same poll had Harris leading Trump by a point (46 percent to 45 percent). On July 16, Reuters-Ipsos showed Trump ahead of Biden by two points (43 percent to 41 percent). On July 23, the same poll gave Harris a two-point lead (44 percent to 42 percent). On July 2, the New York Times–Siena showed Trump leading Biden by six points (49 percent to 43 percent). On July 24, that pollster showed Trump leading Harris by one point (48 percent to 47 percent). Similarly, on July 2 the Wall Street Journal had Trump leading Biden by six points (48 to 42 percent), and Harris by just two points 49 – 47 percent) on July 25. Both Times-Siena and WSJ showed Harris ahead by a point when non-major-party candidates were included.
Given that these are relatively small movements within the margin of sampling error, there’s no reason to get excited by a single poll. But we’re at the point where it’s a trend, not a blip.
While I’m more than a little skeptical of the post-Nate Silver FiveThirtyEight’s election projection model, which is mostly driven by non-polling factors, I have no reason to doubt the validity of their polling aggregate. And it shows this:
Again, not exactly a run-away but this was a race mired in quicksand for more than a year with the two old guys.
The granddaddy of aggregators, RealClearPolitics, still shows a slight Trump lead:
But this is mostly a function of a single outlier:
While Rasmussen definitely has a Republican lean, it’s true that they’re one of the few “Likely Voter” polls; most of the rest are “Registered Voter” polls. The two are likely related but, in my judgment, their likely voter screen is outdated. Given the demographic realignment of the parties, it’s not obvious to me that we should expect Republicans to turn out at a higher rate than Democrats anymore.
Regardless, the national dynamic has clearly changed. Harris is generating more enthusiasm among Democrats, Black voters, and young voters (obviously, there’s significant overlap in those demos). Additionally and relatedly, support for Robert Kennedy Jr. has diminished considerably.
Of course, we don’t have a national election for President. And, while it’s possible that the number of “swing” states has gone up slightly with the new matchup, most analysts are still focused on the handful of Rust Belt states that went for Obama in 2012, Trump in 2016, and Biden in 2020.
And there’s more good news there:
Battleground-state data has been slower to arrive, but what we have shows Harris improving on Biden’s performance quite consistently. A battery of Emerson–The Hill polls taken from July 22 to July 23 of five battleground states showed Wisconsin tied at 47 percent and Trump leading Harris by five points (49 percent to 44 percent) in Arizona; two points (48 percent to 46 percent) in Georgia; one point (46 percent to 45 percent) in Michigan; and two points (48 percent to 46 percent) in Pennsylvania. What’s more significant are the trend lines since the last polls from Emerson in mid-July, testing Biden against Trump.
Most recently, and perhaps impressively, Bloomberg/Morning Consult has released a new batch of seven battleground state polls taken from July 24-28. Overall, they showed Harris leading Trump by one percent (48 to 47 percent), as compared to a two-point Trump lead over Biden in early July. The individual state gains by Harris were also striking: she led by two percent (49 to 47 percent) in Arizona, a real problem state for Biden; by two percent (47 to 45 percent) in Nevada; by two percent (49 to 47 percent) in Wisconsin; and by an astonishing 11 percent (53 to 42 percent) in Michigan. Harris was tied with Trump in Georgia at 47 percent, and trailed him by two percent (46 to 48 percent) in North Carolina and by four percent (46 to 50 percent) in Pennsylvania.
Three battleground states have enough post-Biden-Harris-switch polling now for FiveThirtyEight to compile averages, and all of them show very close races. In Georgia, Trump leads by 1.1 percent (45.9 to 44.8 percent), but Harris leads in Michigan by 1.8 percent (44.8 to 43.1 percent) and most surprisingly, in Pennsylvania by 0.4 percent (45.1 to 44.6 percent).
I’m a little more leery of making judgments on state-level polling, just given their relative scarcity and inherent sampling problems. But, again, the effects seem to be similar to the national dynamic.
The “double hater” phenomenon is gone, replaced by . . . I don’t know, exactly. Harris is a relative unknown. Thus far, though, she’s bringing an energy and joy to the campaign trail and has turned the tables on Trump with name-calling and jeering but in a way that’s less mean-spirited.
Trump’s seven-plus years of campaigning against “Sleepy Joe” is all for naught and there isn’t yet an obvious backup plan aside from racist appeals. But the demographic they appeal to was already in his camp.
