With four weeks to go until the election, the race for the White House is as tight as it’s been for the last two weeks. The New York Times polling average shows Kamala Harris and Donald Trump essentially tied across the seven battleground states considered likeliest to decide the presidency, with neither ahead by enough to count as even a modest favorite. (There was another reason the polls didn’t seem to budge in the last week: There just weren’t that many of them.) Updated Oct. 7
Nate CohnChief political analyst
With four weeks to go until the election, the race for the White House is as tight as it’s been for the last two weeks. The New York Times polling average shows Kamala Harris and Donald Trump essentially tied across the seven battleground states considered likeliest to decide the presidency, with neither ahead by enough to count as even a modest favorite. (There was another reason the polls didn’t seem to budge in the last week: There just weren’t that many of them.) Updated Oct. 7
Both Harris and Trump have states they can probably count on to get at least 200 electoral votes, but they need to get to 270 to win. Let’s zoom in on seven key states in the middle where the vote was closest in 2020.
If the polls change, or miss, in Trump’s favor
Trump could easily prevail if he gains ground in key states — or if the polls are already underestimating him. A shift or polling miss in his favor would not be uncommon, though of course that is not guaranteed.
Trump is within 1 point of the lead in this state.
Albert SunGraphics reporter
Polls in seven key battleground states are very close, and flipping just a state or two could lead to either candidate’s victory. Many outcomes are possible: Since the results in states often move together, and polls often miss in the same direction, it would take only a small shift — or polling miss — in either candidate’s favor to turn a narrow win in the Electoral College into a victory with more than 300 electoral votes.
| Election Results | |||
|---|---|---|---|
| Polling Leader | 2020 | 2016 | |
| 16 E.V. | Trump <1 | R +1 | R +4 |
| 19 E.V. | Harris <1 | D +1 | R +1 |
| 6 E.V. | Harris +1 | D +2 | D +2 |
| 15 E.V. | Harris +1 | D +3 | R <1 |
| 16 E.V. | Trump +2 | D <1 | R +5 |
| 10 E.V. | Harris +2 | D +1 | R +1 |
| 11 E.V. | Trump +2 | D <1 | R +4 |
| Election Results | |||
|---|---|---|---|
| Polling Leader | 2020 | 2016 | |
| 30 E.V. | Trump +4 | R +3 | R +1 |
| 10 E.V. | Harris +6 | D +7 | D +2 |
| 40 E.V. | Trump +6 | R +6 | R +9 |
| 13 E.V. | Harris +7 | D +10 | D +5 |
N.H. 4 E.V. | Harris +7 | D +7 | D <1 |
| 17 E.V. | Trump +8 | R +8 | R +8 |
Neb. C.D. 2 1 E.V. | Harris +8 | D +6 | R +2 |
| 5 E.V. | Harris +8 | D +11 | D +8 |
Albert SunGraphics reporter
Most pollsters are focusing their efforts on national polls and polls in the seven battleground states likeliest to decide the election. We can calculate polling averages for less competitive states once enough polls have been released, though many of the least competitive states are hardly being surveyed at all.
| Biggest recent polling miss | ||
|---|---|---|
| Wis. | 9 pts. (2020) | Range of polling miss |
| Mich. | 6 pts. (2022) | |
| Nev. | 4 pts. (2012) | |
| Pa. | 5 pts. (2022) | |
| N.C. | 6 pts. (2016) | |
| Ga. | 2 pts. (2016) | |
| Ariz. | 3 pts. (2022) |
Ruth IgielnikStaff editor, polling
In 2016 and 2020, the polls underestimated Trump. But in 2012 they underestimated the Democratic candidate, Barack Obama, and in 2022 Democrats outperformed polling averages in several key midterm races. Past polling misses can give a sense of the magnitude by which current averages could be wrong, but they are not a guarantee of which way today’s polls may be off.
Kaleigh RogersPolling reporter
We’re still waiting for high-quality polls fielded after the vice-presidential debate, but if you’re expecting a drastic shake-up in the race, don’t hold your breath. History shows vice-presidential debates have little to no impact, and it will be hard to disentangle any movement from other recent major events, like the aftermath of Hurricane Helene in the Southeast. There’s a decent chance that JD Vance and Tim Walz will see an increase in the share of voters who have an opinion on them, and perhaps even a boost to their favorability. But it’s unlikely the debate will have much effect on the numbers for Harris and Trump. Updated Oct. 4
July 21 Biden leaves race
Biden vs. Trump +3 Trump
Harris vs. Trump +3 Harris
Nate CohnChief political analyst
In late July and August, Harris made steady gains. Those gains seem to have slowed, suggesting she’s mostly consolidated her potential support. Any additional gains won’t be easy. If even a consensus debate victory can’t move the needle, it’s hard to see what would give either candidate a meaningful edge in the polls over the final stretch. Updated Sept. 23
| Current margin | polls | ||
|---|---|---|---|
| U.S. | Harris +3 | 62 | |
| Wis. | Harris +2 | 15 | |
| Mich. | Harris +1 | 15 | |
| Nev. | Harris +1 | 11 | |
| Pa. | Harris <1 | 23 | |
| N.C. | Trump <1 | 24 | |
| Ga. | Trump +2 | 14 | |
| Ariz. | Trump +2 | 14 |
Nate CohnChief political analyst
The polls have been remarkably stable over the last month or so since Harris successfully reunified her party after becoming the Democratic nominee. This stability can be a little deceiving. The poll averages could be so stable in part because many pollsters are using heavy-handed statistical techniques that reduce the variance of their results from poll to poll but that increase the risk of systematic errors. Updated Oct. 7
About our polling averages
Our averages include polls collected by The New York Times and by FiveThirtyEight. The estimates adjust for a variety of factors, including the recency and sample size of a poll, whether a poll represents likely voters, and whether other polls have shifted since a poll was conducted.
We also evaluate whether each pollster: Has a track record of accuracy in recent electionsIs a member of a professional polling organizationConducts probability-based sampling
These elements factor into how much weight each poll gets in the average. And we consider pollsters that meet at least two of the three criteria to be “select pollsters,” so long as they are conducting polls for nonpartisan sponsors. Read more about our methodology.
The Times conducts its own national and state polls in partnership with Siena College. Those polls are included in the averages. Follow Times/Siena polling here.
Maine and Nebraska award two electoral votes to the statewide winner and a single electoral vote to the winner of each congressional district. (Maine has two congressional districts, and Nebraska has three.) Historical election results for these districts are calculated based on votes cast within the current boundaries of the district.
Sources: Polling averages by The New York Times. Individual polls collected by FiveThirtyEight and The Times.
Credits
By Cam Baker, Laura Bejder Jensen, Ademola Bello, Dana Chiueh, Nate Cohn, Molly Cook Escobar, Annie Daniel, Ruth Igielnik, K.K. Rebecca Lai, Jasmine C. Lee, Alex Lemonides, Albert Sun, Rumsey Taylor and Isaac White. Additional work by Kristen Bayrakdarian, Asmaa Elkeurti, Andrew Fischer, Andrew Park, Jaymin Patel, Ethan Singer and James Thomas.