Friday Polls: What's new, who's the Veep, and what's next
Let's talk about the changing shape of the electorate, what a favorability score is anyway and how that effects the Democratic veepstakes... plus links
- How things have changed (so far) since June.
- What are favorability numbers, anyway, and what do they have to do with the veepstakes?
- General fundraising and get-out-the-vote skullduggery
How things have changed (so far) since June.
The debate was on June 27. Trump was shot on July 14. The RNC began on July 15. Biden stepped aside from the presidential race on July 21.
Today is August 2. Harris won the roll call vote of DNC delegates, clinching the nomination. She has raised $310 million in 12 days.
Where I live, we’re measuring enthusiasm via cup sales.
What do the polls say about all this change?
As a data source, I compared the June and July New York Times-Siena College poll cross tabs among Registered Voters.
In June, 22% of registered voters thought the country was on the right track. In July — after Biden stepped aside and Harris stepped up — 28% of registered voters said they believe the country is on the right track. That’s a big bump, suggesting that one of the things that made people feel bad about the state of the country was the state of their choices in the upcoming election.
Among those who say they intend to vote for the Democrat, that number increased from 41% in June (when Biden was the candidate) to 49% in July (with Harris as the candidate). Trump supporters are less likely to say it’s headed in the wrong direction, but most of those who defected from that position now say they’re not sure. The biggest bumps by demographics were among:
- 8-point bump for 45-64 year olds
- 15-point bump for African Americans;
- 8-point bump for college-educated non-whites
- 7-point bump for all those without a bachelor’s degree
There were other bumps, but these were the biggest.
In June 56% of registered voters said they would almost certainly vote in November. In July 62% of registered voters said so. This is also a pretty big bump — some of this is likely that people who hadn’t been paying attention before are paying attention now, but some of this is also renewed enthusiasm. The biggest bumps by demographics were among:
- 9-point bump for 30-44 year olds
- 12-point bump for Hispanics
- 7-point bump for those without a bachelor’s degree
- There was a curious 5-point decline for Black voters — so I’ll point out that the base sizes of Black and Hispanic registered voters were much smaller in July than in June, therefore the error bars on those answers are much wider than the overall survey margin of error.
In June 42% of registered voters said they would vote for Biden. In July 46% of registered voters said they would vote for Harris. Trump’s support stayed flat at 48%. Those who say they don’t know dropped 4 points from 10% to 6% between June and July. Here were the biggest bumps for Harris:
- 8-point bump for women
- 10-point bump for 18-29 year olds
- 10-point bump for Black voters
- 12-point bump for Hispanic voters
- 6-point bump for those with a bachelor’s degree, pretty evenly distributed between non-white and white college grads
- And a 10-point bump among those who voted for Biden in 2020
Trump’s net favorability went from -7 to -3. Biden’s went from -21 to -12, a pretty big improvement for a guy who dropped out. Here’s where they each got their bumps:
Trump’s “very favorable” bumps:
- 14 points among men
- 10 points among 45-64 year olds
- 11 points among 65+ year olds
- 10 points among white voters
- 10 points among Hispanic voters
- 15 points among those without a college degree
Biden’s “very favorable” bumps:
- 9 points among women
- 10 points among 45-64
- 22 points among Black voters
- 8 points among Hispanic voters
- 7 points among those with college degrees, and 5 points among those without — the lion’s share of this improvement came from non-white respondents.
You can see the shape of the electorate beginning to shift — and you can see the way the choice of candidates changed the way people felt about their choices, the direction of the country, and their enthusiasm. You can even see who thinks better of Biden for stepping aside, and who was impressed by the way Trump initially handled the attempted assassination.
If I were a predicting kind of gal, I’d say this election will be decided by an intersectional coalition of women under 50.
But — not enough time has elapsed to be sure about any of that.
What this data gestures at is the assumptions underpinning much political analysis are no longer up-to-date. The entire etch-a-sketch has been shaken, and as a result, you now see some of the poll trackers in an “archival” mode. The NYT polling tracker has set aside its trackers’ pre-July 21 data because the race has been fundamentally altered. One major assumption was that Pennsylvania would be the turning point for Democrats and Republicans to get to 270. It’s entirely plausible that this is no longer true. But for the next week or so, and definitely not until after the DNC, we simply won't know.
Polling averages need a little time, and a series of high-quality, reliable polls (that publish their cross tabs!) so that analysts and forecasters can tell you anything about where things are. If you’re seeing headlines about honeymoons and vibes, just know: the party that changed its candidate also changed the way its voters feel about the election, the country's future, and the sitting President.
What are favorability numbers really saying (about the veepstakes)?
We’re looking at a likely announcement next week about who Vice President Harris will select as her running mate. The contenders who have, according to recent reports,been actively vetted by her campaign include Pennsylvania Governor Josh Shapiro, Arizona Senator Mark Kelly, Illinois Governor JB Pritzker, Secretary of Transportation Pete Buttigieg, Minnesota Governor Tim Walz, and Kentucky Governor Andy Beshear.
Let’s take a look at some numbers for the potential VP picks from a very recent YouGov survey of US adults (note: not necessarily registered voters) based on total approval ratings:
- Mark Kelly: 44% (24 strong/20 somewhat)
- Pete Buttigieg: 44% (22 strong/22 somewhat)
- Josh Shapiro: 34% (13 strong/21 somewhat)
- Gretchen Whitmer: 39% (13 strong/26 somewhat)
- Gavin Newsom: 35% (13 strong/22 somewhat)
- Andy Beshear: 31% (12 strong/19 somewhat)
- Tim Walz: 25% (9 strong/16 somewhat)
- Raphael Warnock: 27% (6 strong/21 somewhat)
- Roy Cooper: 21% (6 strong/15 somewhat)
- JB Pritzker: 21% (5 strong/16 somewhat)
Let’s note how the list is sorted — it’s in descending order based on the “Strongly Approve” number, making Mark Kelly the seemingly best pick, followed by Pete Buttigieg and Josh Shapiro at 4-point intervals.
There are a couple of ways to calculate this, though — you could take the total of those who approve and subtract the total of those who disapprove. Those approval ratings look like this:
- Mark Kelly: +31
- Pete Buttigieg: +23
- Andy Beshear: +21
- Josh Shapiro: +20
- Gretchen Whitmer: +20
- Tim Walz: +13
- Raphael Warnock: +8
- Roy Cooper: +8
- JB Pritzker: +7
- Gavin Newsom: +5
Mark Kelly stays on top, Pete stays in second place, but Andy Beshear jumps from #6 to #3, and Gavin Newsom drops from #5 to last.
But we should also notice the “not sure” number. This is an option on the scale that is also sometimes labeled “Don’t know enough to say” — we can use this as an awareness or name recognition number. I’ll list those in ascending order because those with the lowest “don’t know” number have, in theory, the highest name recognition.
- Pete Buttigieg: 35%
- Gavin Newsom: 35%
- Gretchen Whitmer: 43%
- Mark Kelly: 44%
- Josh Shapiro: 52%
- Raphael Warnock: 54%
- Andy Beshear: 58%
- Tim Walz: 63%
- JB Pritzker: 64%
- Roy Cooper: 67%
For the sake of argument, let’s count the not sure number as a negative and call this something like approval + name ID score:
- Pete Buttigieg: -12
- Mark Kelly: -13
- Gretchen Whitmer: -23
- Gavin Newsom: -30
- Josh Shapiro: -32
- Andy Beshear: -37
- Raphael Warnock: -46
- Tim Walz: -50
- JB Pritzker: -57
- Roy Cooper: -59
So what do we really know about how people feel about the candidates? If it were me, in a branding exercise with one of my clients, I would essentially cut off any brand that had less than 50% name recognition. But if I do that, interesting things happen: Josh Shapiro, the odds-on favorite for the pick, just drops right off the map. If I’m a little more generous he stays on, but at the bottom. Regardless, there’s more work to do with Shapiro, from a brand building point of view, than with some of the other options, which may be politically riskier.
Of course, any candidate’s numbers will change if he’s selected because his name ID will go to 80%+ within a week. The question, strategically, rests on a few factors:
- How do any of the top 5 candidates fare with a coalition of women, especially women under 50, especially, frankly, white women under 50?
- Look at the whole map — does a candidate perform better in one swing state and worse in another? If the tipping point state is no longer Pennsylvania, for example, does Shapiro cost you anything in other states? Does Kelly or Buttigieg or Beshear?
- Look at the talents and backstory of the candidates — what do you have as raw material to build a brand in 90 days?
- Kelly has a great biography as a veteran, astronaut and husband of Gabby Giffords and comes from a nominal swing state, but isn’t regarded as an amazing campaigner and communicator
- Beshear looks great (and some say sounds great) but isn’t as tested on the national stage as other options and Kentucky is not likely to go blue this cycle
- Shapiro has a track record that pleases some and turns others off (to say nothing of what assumptions people are making — rightly and wrongly — about his Jewish heritage informing his political positions on Israel and Gaza) and is a reasonably popular popular first-term governor... and he’s widely regarded as a great communicator and campaigner, but his approval rating has slipped 10 points in recent polling
- Buttigieg has served in the Navy, run a national campaign in which he won a caucus and tied in a primary, and served in a cabinet-level role, is an excellent communicator and campaigner with high name ID, and calls a swing state (Michigan) home — but he’s gay and people feel some kind of way about that
I honestly have no idea who the pick will be — the press and pundit class generally prefer Shapiro because of a mostly apocryphal belief that the VP candidate can deliver a win in his home state and that Pennsylvania is still the tipping point state (recent performance suggests there is maybe a 1-2% home state bump, and maybe that will make all the difference in Pennsylvania). We’ll find out next week. We do know who that pick's chief of staff will be, however: Liz Allen, who currently heads the State Department's public diplomacy office (I have no earthly idea if that is a clue).
Skullduggery
Maybe this will become a regular feature — for now, let’s call it the links section. Administrative data and polling can be used to model turnout — but it’s tough to model people trying to mess with the system. These links will be about the ways various organizations are, in The Wire parlance, “juking the stats” to get the outcome they want. Even when the shifts in the results are rounding errors in the polls, they have the potential to significantly affect outcomes in a highly polarized electorate.
Elon Musk’s super PAC is messing around with people’s personal identifying information in swing states: https://qz.com/elon-musk-donald-trump-super-pac-election-voter-fraud-1851612153
"After a resident from New York, California, Wyoming, or North Dakota — historic blue and red states — inputs their email and zip code, they’ll either be linked to their state’s voter registration page or directed to repeat the process again. It’s unclear if this is due to website error or an intentional design.
But, if they live in a state expected to be competitive in the election, visitors can anticipate a lengthier process. After putting in a Pennsylvanian zip code, the site asks visitors for their full name, cell phone, home address, and date of birth. Then, rather than be directed to a voter registration page, they will be told “thank you” and provided no further help."
The Harris campaign has warned its supporters about the scam text messages we’re all receiving by the dozens these days (seriously, today I’ve received 10.):
“At first blush, the messages sent by these outside groups seem legitimate.
One recent text included a picture of Barbra Streisand, the famed singer, actress, and longtime Democrat, saying how “excited” she is “to support KAMALA HARRIS!” and offering a “700% MATCH ACTIVE” for donations to help “crush” Donald Trump.
But the text wasn’t from Streisand or the Harris campaign. It was from “Democratic Power,” a group started in October 2022 at an address that appears to be a UPS store in Southeast Washington, D.C.
Democratic Power has few traces online, save stray websites that contain records of its expenses or solicitations, like this one suggesting Hillary Clinton supports them while offering a “700% MATCH ACTIVE.”
That 700-percent match is almost certainly fake. As for the $2.2 million the group has raised since the start of 2023, federal records show that only a small portion ($363,000) was sent to candidate committees like the Biden-Harris campaign and its allied super PAC. Nearly $1.3 million went to operating expenses, often directed to entities with ties to Democratic Power.”
The other day the Harris campaign put on an event featuring Megan Thee Stallion (which you have surely seen) in Atlanta, and said the road to the White House runs through Georgia. But Republicans in that state are doing all they can to purge voters from the rolls: https://apnews.com/article/georgia-voter-removal-software-eagleai-266ead9198da7d54421798e8a1577d26
"Election officials say many challenges are powered by EagleAI. The tool was created by Dr. John “Rick” Richards Jr., a retired physician and entrepreneur who lives in suburban Augusta’s Columbia County.
Richards said in a Wednesday interview that people using his software are citizen volunteers, likening the work of finding ineligible voters to picking up roadside trash.
“No one is going to be denied the right to vote,” Richards said. “That’s a bunch of hooey.”
In online meetings and in-person appearances over the past year, Richards has pushed EagleAI as a sophisticated platform to cleanse dirty voter lists. The Associated Press found the platform is funded and used by supporters of Trump, some of whom worked to overturn the 2020 vote, and entwined with the Republican’s campaign.
An EagleAI document last year touted the system’s “use of AI” and “multitiered algorithms” to cleanse dirty voter lists, but Richards now says there is no artificial intelligence at work. The software instead draws in part from a database of “suspicious” voters hand-built by conservative activists, the AP found.
Over past months, an AP reporter joined online meetings publicized among activists before eventually being asked to leave. The AP also obtained additional meeting videos to glean a behind-the-scenes look at how the software is used in states including Arizona, Florida, Illinois, Nevada and Ohio.
“The left will hate this — hate this. But we love it,” Cleta Mitchell, a frequent participant, said during one presentation. Mitchell is a GOP election attorney who took part in the call when Trump implored Raffensperger to “find” more votes in the 2020 election. While Trump was indicted in Georgia for the call, Mitchell was not. Mitchell now is a leader in multiple organizations pushing to purge voting rolls.
Richards called Mitchell’s affiliations “irrelevant.”
Have a great weekend.