Good Polling for Harris
But it is all still close.

NBC News reports: Poll: Newly popular Harris builds momentum, challenging Trump for the mantle of change.
A double-digit increase in popularity, rising Democratic enthusiasm and an early edge for representing “change” have vaulted Vice President Kamala Harris forward and reshuffled the 2024 presidential contest, according to a new national NBC News poll.
With just over six weeks until Election Day, the poll finds Harris with a 5-point lead over former President Donald Trump among registered voters, 49% to 44%.
[…]
A double-digit increase in popularity, rising Democratic enthusiasm and an early edge for representing “change” have vaulted Vice President Kamala Harris forward and reshuffled the 2024 presidential contest, according to a new national NBC News poll.
With just over six weeks until Election Day, the poll finds Harris with a 5-point lead over former President Donald Trump among registered voters, 49% to 44%.
[…]
In an expanded ballot with third-party candidates, Harris leads Trump by 6 points, 47% to 41% — with Robert F. Kennedy Jr. at 2%, Jill Stein at 2% and Libertarian Chase Oliver at 1%. (Respondents were only able to pick from the major third-party candidates who will actually appear on the ballot in their states.)
Both ways of testing the race represent a change from July’s poll, when Trump was narrowly ahead of Biden by 2 points on the head-to-head ballot test and by 3 points on the expanded ballot test. The September results are the Democratic ticket’s best performance in the poll since the summer of 2023.
Via The Daily Beast: Kamala Harris Has Biggest Favorability Jump Since George W. Bush After 9/11.
Via CBS: Harris shows some gains and economic views brighten a bit — CBS News poll.

Plus from. late last week:
Meanwhile, via The Hill: More than 700 high-ranking national security officials endorse Harris.
Eyes on the prize. The Swing states. Eyes on the prize.
Madame Harris is proving surprisingly good, greatly “mis-underestimated” (thankfully perhaps?) – the Democrats (and USA in general) perhaps should have more ‘coronations’ and less of the ridiculously ineffective “vetting” by skewed pseudo-democratic primaries.
Keep the Orange Cretin out of presidency and the whole world shall be modestly better off.
Thank god for the EC otherwise all we would hear about is polling from a handful of states.
I know it’s petty but CBS is using an eight year old photo of Trump. Is that a choice or CBS policy to only use an Official Photo?
This is exactly why Lindsey Graham’s interference in Nebraska is so dangerous. Switching just that one vote, based on this polling data, would send the results to the House.
It should not be this close.
For us or against us, I just don’t have much faith in polls this year.
And then right on cue this morning, the NYT blasts out an email with the subject: “Donald Trump has pulled ahead of Kamala Harris in Arizona and is leading in tight races in Georgia and North Carolina, a Times/Siena poll shows.“
@MarkedMan: What you should have is a wise wariness of the actual meaning of point percentages in a probabilistic analysis – which one would profitably not understand as “predictions” in a point sense but rather a possible distriubtion of outcomes. This is very difficult for our brains which are frankly not well wired for such, being simpley overgrown chimps as we are – but one should force oneself to try to think in this direction.
One’s odds are improving right now but “up 5 pts” is actually a probabilistic cone and this is in reality knife-edge, knifier edge than it should be perhaps and knifier edge than desirable and knifier edge than even Madame Clinton.
But at least the dynamic is better and Madame Harris is showing herself to have a certain sense of the popular retail that l’éminance grise never had.
I do hope that the pressure on Trump who has never show great skill under real pressure generates more error that continues the slow acid bath erosion. Every molecule of vote eroded away counts.
@SKI: And?
Their poll says what it says.
Be glad for the motivation and glad that there is this to avoid complacency. It is merely reality in a tight race that polls, being not Future Predicators but displaying most likley range of probabilities, in such a tight race will bounce around.
Then there is this:
New York Times/Siena Sunbelt Polls
* Arizona: Trump 50-45% Over Harris, Gallego by 6 Leads Lake, Georgia: Trump 49-45% Over
Harris, North Carolina: Trump 49%, Harris 47%, Stein 10 points ahead of Robinson
* Trump Trusted More by Voters across AZ, GA, NC by 12-14 points on Top Issue, Economy
* Harris Trusted More by Voters across AZ, GA, NC by 9-13 points on Abortion
* Voters Evenly Split on Which Candidate Better on Democracy
Don’t get too excited yet.
Speaking of voting:
Department won’t provide election security after sheriff’s posts about Harris yard signs
Betcha he’s a “Constitutional Sheriff“.
I woke up this morning and went straight to the polls. As one does. I saw the NYT numbers and was depressed. Then I come here and see ‘good polling for Harris.’ As stated above, the topline number doesn’t mean much. A bunch of Californians can run up the national number without adding a single electoral vote.
I have never been a true-the-polls guy, but I do have one nagging question. I understand that pollsters adjust numbers based on other factors, age, race, location, whatever. But before all the algorithmic mumbo jumbo, they need actual numbers from actual respondents. My question is this: does anyone here know anyone under the age of 30 who answers an unidentified phone call? Ever?
I’m 70 and I don’t. For my kids, 27 and 25, it is simply inconceivable. I very much doubt that either of them has ever answered an unidentified phone call. So, I get that pollsters adjust the data, but do they have any actual data to begin with? Do they have even, say, 5 respondents in the 18-30 demo? And if so, just how desperately lonely are those five kids?
@Michael Reynolds: I’m also 70 and my phone is set to silence if a number is not in my contacts. No, I don’t answer any unrecognized phone number.
@Michael Reynolds: On the one, I think there are real issues with phone-based polling.
On the other, people were saying “I’ve never been polled” for decades–well before cell phones were a thing.
I think, ultimately, that the polling in the recent past has been empirically pretty good, if judged mathematically, not by vibes.
And BTW: if you all click through there is state-level good news for Harris as well.
But, as the sub-head says: it is all still close.
@Michael Reynolds:
One more thought: don’t fall for GOP talking points that the national numbers are just CA. (And Trump gets a lot of CA voters as well).
@Scott:
I find it difficult to believe that people who will vote for Gallego will also vote for Trump. That’s all; it just doesn’t make sense.
@Steven L. Taylor:
I know, the polls have been pretty accurate. How is a mystery. Who is answering a phone call? In 2024?
@just nutha:
I’ve had that same thought.
@Steven L. Taylor:
@Michael Reynolds:
Polling pretty good, polls pretty accurate. “Pretty” is doing a heavy lift here.
How Skewed Polls Fed a False Election Narrative (New York Times):
How 2024 pollsters are trying to avoid their 2020 mistakes (Axios):
I think polling has been “pretty” messy lately. Much worse than that, poll analysis has been shambolic. But there are folks (cable news talking heads, 538, Nate Silver’s substack) whose business models rely on passionately arguing otherwise.
Unfalsifiable probability models and election punditry are suspect enough on their own, as output. They can’t afford their consumers to start thinking the input is sus too.
@just nutha: The commonality there is that the two candidates who are down in the polls are women.
It has little to do with consistency, or policy.
@Jen: Okay. I hadn’t considered that factor. It works.
@Michael Reynolds:
Except there’s some correlation between the national numbers and the EC. In theory it might be possible for Harris to win the PV by 10 points and still lose the EC, but realistically we know that’s never happening. If she’s up by 5, I think we can safely assume she’s winning the EC. And the level of advantage the Republicans enjoyed in 2020, where Biden won nationally by nearly 5 points but his EC win depended on margins of less than a percent in the key battleground states, appears to have shrunk, when you look at the current national polls compared with those of the key battlegrounds. The polls could be wrong of course, but the polls did pick up the Republican advantage in the 2020 election before it happened (in fact they slightly overestimated that advantage).
Why this is the case is a good question. I think it’s partly because the Republicans have made gains in non-swing states like New York, while Dems have gained ground in the swing states. We shouldn’t discount how Dobbs and Jan. 6 may have changed the map, as well as the fact that Dems are no longer neglecting their ground game like they did in 2020 due to the pandemic.
Any poll modeling meant to factor in GOTV efforts has to be based on past patterns, does it not? But, everything I’m reading suggests there is unprecedented activity in new voter registration, small donor contributions, and imbalance in the respective campaigns’ field operations. Whatever adjustments pollsters are making to account for these enthusiasm indicators I suspect has more guesswork that usual.
Re: polling, there’s this at LGM:
They link to this tweet from Josh Chafetz that shows the net favorability of a range of politicians, entertainers, and economic systems. Tim Walz is just below capitalism and above Taylor Swift and…well, everything else, he’s literally second only to capitalism.
The consultants wanted Harris to pick Shapiro, but she went with her choice, which was the right one by light-years. It seems like whenever she has told the established political consultants to eff off, she was right and they were wrong.
@DK: IIRC, the polling error in 2020 was ~4%. While that is isn’t nothing, a little perspective is in order.
Likewise, the whole “red wave” thing is more about narratives than it is about polling, per se.
I am not saying it is perfect, but it is also the case that a useful tool with limitations.
And yes, the model stuff has gotten a tad out of hand.
@Lounsbury:
Not sure what you think you are responding to but it came from inside your own head, not what I wrote.
I posted what I did because it was a brand new poll that contradicted the theme of Steven’s article. The underlying message is that this race is way too close to take anything for granted.
Then there are folks that try to make sense out of it.
I mean, considering all the insanity of the past few weeks , it’s just got to be.
Trump is at 48 percent. How could this be possible but for widespread racism?
I guess I never really wanted to admit it to myself before, but there is a lot more hate than I had thought.
When I talk about Jim Crow governance, I try to make it clear that I don’t think this type of governance had its origins in the American South. I suspect it was prevalent as far back as our Neandertal ancestors. For a non-American example, here’s 18th century Frenchman Jean-Jacque Rousseau, as described in a NYTimes piece:
How in the F is AZ going for Trump? If economy is the big issue, how can those AZ voters not be completely in favor of a Dem. Why? The friggin’ CHIPS Act has literally made Maricopa county the fastest growing, with housing equity soaring with no end in sight as INTEL and TSMC add thousands of new, well paying jobs with a tertiary effect of adding 16 new jobs in the community (dry cleaners, markets, restaurants, construction) for every one semiconductor job. And it’s not just INTEL/TSMC. All the subcontractors (clean room, fab machines etc) are opening offices there. Arizona is a boom town. If it’s the economy stupid, why aren’t we winning in a landslide?
Because it’s also the border stupid. Gallego sounds like a Republican on this issue, hence he’s winning v Harris whose record as border “Czar” is hurting her in AZ.
Our border policy has cost us so many seats and POTUS elections. What’s so hard for Dems to say, “in order to protect our hard working blue collar and other lower wage employees, we need a secure, closed border….”
FACT: working class wages soared when border was closed for COVID for the first time since Carter was POTUS. Why? we stopped the invasion of millions of illegal laborers who have undercut wages for 50 years in the country, with the Dems fully complicit. Sanctuary cities? What a joke. If some red neck sheriff said “I am not going to enforce federal gun laws” we libs would have a fit. When our side says “I am not going to enforce federal immigration law” the trupers have a fit. And both sides are right in throwing a fit.
Note: On cue, Harris announced a visit to the border in AZ on Friday. Excellent strategy. Obviously, her advisors read my comment. (wink)
https://www.cnn.com/2024/09/23/politics/harris-campaign-possible-southern-border-trip/index.html
@MarkedMan: I hadn’t read that piece this morning. Thanks for the pointer, it’s spot on.
It’s a commonplace to say that race is the constant in American politics. The author is clear that Rousseau didn’t address race, but race “fits seamlessly” into divineness as a political strategy. And he’s right, the constant isn’t race, but rich and poor. Despite all their MAGA faux populism and other blurring of lines, Republicans are still the party of the rich and powerful.
Of late I’ve participated here in some discussion of NYT v FTFNYT. NYT sees themselves as the voice of the NY establishment, i.e., in their view, the world’s elite. Note this essay on Rousseau is by a political scientist, not an NYT staffer. For the attitude of NYT staffers, Eschaton links to an interview with Maggie Haberman.
How dare those Little People criticize ME? Living totally in a bubble. She also said,
She sandwiches that in praise of what a good job the press is doing. But when she talks of “process” and “system”, who the hell defines their processes and systems more than Haberman, Sulzberger, et al?
@Michael Reynolds:
It’s worth remembering what the polls measure: the preference of registered voters, adjusted by population demographics (well understood) and then adjusted again by expected turnout assuming that there are no unprecedented shifts in those demographics turning out to vote, and no other one sided boost (less well understood).
There’s likely an assumption that GOTV will not be dramatically more successful for Democrats than Republicans (Trump has trashed the RNC, but how this will affect state operations is a guess, and the Republicans are relying more on outside groups to do GOTV)
There’s also an assumption that women will make up N% of the electorate, based on statistics from the last few Presidential election cycles. If more are likely to vote (Abortion is an issue), the top-line polling numbers wouldn’t necessarily capture that.
The 2022 midterms didn’t have the expected Red Wave largely because the makeup of the electorate was different than previous off-year midterms.
All this goes to say — polls are tools that can and are widely misused. They are great at showing direction of movement, but the actual number… usually good, but a lot of ways to introduce error.
@The Q:
Local sheriffs don’t enforce federal laws. It literally isn’t their job.
@The Q:
If your objective is actually to protect jobs, then the increase in regulatory staff needs to be focused at….. the location of those jobs.
Want to actually protect native-US low-wage employees? Then do random raids of any/all businesses and issue instant $100,000 fines TO THE BUSINESS for every employee onsite who doesn’t have a green card.
But they don’t do that – and will never do that….. because this isn’t about US jobs at all, as every Chamber of Commerce knows. Restaurants, construction, food processing are just 3 of the many businesses that will immediately drop in volume without the ability to exploit low-wage immigrant labor. US citizens don’t want jobs that pay at best half of minimum wage.
Besides, Democrats did negotiate and do something earlier in 2024 to “fix” this…… but Republicans killed it even though many Republicans did sponsor and work towards passing the legislation.
@The Q:
False. The average wage briefly rose when millions of native-born Americans were briefly removed from the job market by COVID disruptions, thus briefly increasing the demand for and value of labor.
It had nothing to do with any paranoid “border invasion by illegals” xenophobic Fox News fearmongering. It was American citizens leaving and then reentering the workforce by the millions.
Despite the pandemic, wage growth held firm for most U.S. workers (Pew Research Centers):
@The Q:
Because Democrats aren’t into rightwing propaganda. The threat to hardworking blue collar workers is not migrants. It’s rich corporations and their political handmaidens.
Migrants are not responsible for the corporate and investment class’s choice to price gouge and pay slave wages while blocking paid leave, child tax credits, debt forgiveness, universal early childhood education, mass transit investment, and a public healthcare option.
But Trump Republicans *are* responsible for killing Biden’s bipartisan border bill.
Migration benefits the working class and all Americans, because our birth rates are declining and without with immigrant labor and tax money our public and private financials would implode.
Hence why whenever conservatives get a chance to keep their immigration restriction promises — like when they had total control in Trump’s first two years — their Chamber of Commerce overlords force them to spike the ball instead.
Democrats should urge voters to follow the money. They should not cosign the right’s favorite distraction tactic: scapegoating poor, powerless brown folk so the powerful, greedy billionaires who are actually harming us escape blame.
@Scott:
Has there ever been a more scathing indictment of American education?
@The Q:
Except that it’s not about the economy, or who would be better for the economy. It’s about perceptions of the economy, and of who would be better. And the GOP has been crushing the Dems in mistaken/false/disinformed perceptions of, well, pretty much everything, for many years now.
@DrDaveT: No. If the school systems were really failing, the split would be disproportionately for Trump, not an even split.