Down to the Wire in Pennsylvania
Independent Democratic groups' ground game may save the Harris campaign from its own hubris about the power of data analytics and paid media--a report from the field.
I am cautiously optimistic that Kamala Harris is going to defeat Donald Trump in this election. Broadly speaking, that’s because she has a higher ceiling on her potential vote, while everyone in America already knows who Trump is and they’ve made up their minds one way or the other about him. And in the last few days, the centerpoint of the election seems to have shifted from a referendum on the last four years to a referendum on Trump, the increasingly unsteady and dangerous wannabe fascist. As Mike Podhorzer, the former political director of the AFL-CIO, has been pointing out for some time, if voters see this as an election about MAGA and what Trump will do if he gets back into power, Democrats have the advantage.
But when she prevails, I don’t think the reason is going to be because the Harris campaign was so much better at getting out its vote than the Trump campaign. It will be because all the independent groups filling in the gaps in the many evident problems with the Harris campaign staved off disaster. That, and the evident incompetence of the groups in charge of Trump’s (missing) ground game, may be the real explanation for why enough of the critical swing states tip her way.
My reasons for doubting the brilliance of the Harris campaign range from the big stuff to the small stuff. As I wrote here last week, the Harris campaign is the Biden campaign organization with a different candidate. It’s very top-down, media-centric and data-driven. It’s not the Obama machines of 2008 or 2012, which were all of those things but also emphasized involving hundreds of thousands of grassroots Democratic activists as people with agency, empowered to take responsibility and partially own the campaign. Nor is it the Bernie Sanders or Elizabeth Warren efforts of 2016 or 2020, which by necessity were quite people-driven. As my friend Chris Rabb, a member of the Pennsylvania state legislature for the last 8 years, where he represents the heavily Black 200th district in North Philadelphia, says, “Harris is using a Biden infrastructure with Obama vibes.”
Nothing made this clearer to me than Monday’s front-page story (gift link) in The New York Times about how the Harris campaign is hunting down the remaining undecided voters in swing states. As Reid Epstein and Shane Goldmacher wrote, “Inside the Delaware headquarters of Ms. Harris’s campaign, analysts have spent 18 months curating a list of which television shows and podcasts voters consume in the battleground states. Her team has assigned every voter in these states a ‘contactability score’ from 0 to 100 to determine just how hard that person will be to reach — and who is best to deliver her closing message. The results are guiding Ms. Harris’s media and travel schedule, as well as campaign stops by brand-name supporters. For instance, the movie star Julia Roberts and the basketball great Magic Johnson earned high marks among certain voters, so they have been deployed to swing states.”
What is with this mad devotion to turning the messiness of human engagement into something that can be scored with the precision of a Wall Street trader assigning a risk level to a derivative? Well, actually, the analogy is kind of apt, though I’m sure the data whizzes driving the Democratic obsession with fine-tuning paid media think they are doing much holier work than Wall Street quants. The same kind of cold-blooded hubris dominates the decision-making of Future Forward, the $700 million SuperPAC that was first anointed by Biden’s team as its designated partner for mega-donors. According to a story a week ago (gift link) by Theodore Schleifer and Shane Goldmacher also in the Times, Future Forward has conducted more than 4 million voter surveys and “ranked the effectiveness of more than 300 ads that were run online and on television on behalf of both candidates — down to a tenth of a percentage point of precision in multiple categories.” A tenth of a point of precision is nonsense. People shift in intensity of their opinions all the time. Future Forward, they report, has a “dogmatic belief” in the power of late advertising, and so we are now in the midst of a great field test of that theory, along with the entire notion that it’s better to spend hundreds of millions on paid advertising even while leaving grassroots organizations in swing states severely underfunded.
All of this makes me shudder with PTSD about how the smartest people in the room who were also powered by the best data and the most money drove Hillary Clinton’s 2016 campaign into the trash bin of history. Let’s not forget that it wasn’t the fault of on-the-ground campaigners in Michigan that caused the Clinton team to ignore the warning signs of Trump’s rise. It was data whizzes in Brooklyn who told volunteers to stay in Iowa in the last week because their models showed Clinton winning Michigan by 5 points as late as Election Day. As Isaac Dovere-Smith wrote in Politico in a post-mortem that ought to be required reading for all budding politicos, “The only metric that people involved in the operations say they ever heard headquarters interested in was how many volunteer shifts had been signed up — though the volunteers were never given the now-standard handheld devices to input the responses they got in the field, and Brooklyn mandated that they not worry about data entry. Operatives watched packets of real-time voter information piled up in bins at the coordinated campaign headquarters. The sheets were updated only when they got ripped, or soaked with coffee. Existing packets with notes from the volunteers, including highlighting how much Trump inclination there was among some of the white male union members the Clinton campaign was sure would be with her, were tossed in the garbage.”
“The Brooklyn command believed that television and limited direct mail and digital efforts were the only way to win over voters, people familiar with the thinking at headquarters said. Guided by polls that showed the Midwestern states safer, the campaign spent, according to one internal estimate, about 3 percent as much in Michigan and Wisconsin as it spent in Florida, Ohio and North Carolina. Most voters in Michigan didn’t see a television ad until the final week.
Most importantly, multiple operatives said, the Clinton campaign dismissed what’s known as in-person “persuasion” — no one was knocking on doors trying to drum up support for the Democratic nominee, which also meant no one was hearing directly from voters aside from voters they’d already assumed were likely Clinton voters, no one tracking how feelings about the race and the candidates were evolving. This left no information to check the polling models against — which might have, for example, showed the campaign that some of the white male union members they had expected to be likely Clinton voters actually veering toward Trump — and no early warning system that the race was turning against them in ways that their daily tracking polls weren’t picking up.”
No, things aren’t the same as 2016. For one thing, every canvasser is reporting on their door-knocks using MiniVAN. But bad data in can lead to bad decisions out, or a false sense of certainty. I spent the last few days reporting a piece for The Nation on the Democratic get-out-the-vote operation in Pennsylvania. It just came out this morning and is titled “Trouble for Harris and Walz in Pennsylvania?” for a reason. There is a real disconnect between the Harris campaign and local elected officials and community groups; the campaign is too reliant on a DNC organizing infrastructure that values media messaging over street-level organizing; and while it has a big field operation with 50 offices and hundreds of paid staff, there are signs that the campaign’s managers are valuing the wrong metrics.
The disconnect with local leaders isn’t just in how the campaign has ignored and sidelined popular Black politicians like Rabb, whose North Philly legislative district is 75% Black and regular votes at higher rates than anywhere else in the state. (By the way, if you want to help him in these final days rebuild his street organizing team, which he is getting no help from Harris or the PA Dems to do, donate via this ActBlue link.)
Yesterday morning, too late for my Nation piece, I had a long off-the-record talk with an experienced organizer who has been working for the Harris campaign in the eastern half of the state. What they said was hair-raising. “To be honest, this cycle has been one of the more poorly run campaign cycles I’ve experienced as an organizer,” they started out. Among their complaints: that organizing staff have been blocked from contacting local community leaders, including board and committee people; that local volunteers seeking paid canvassing slots were overlooked most of those jobs going to outsiders; that multiple requests for a Latino-oriented field office were ignored and so were many of the Latino-heavy wards of Philadelphia. The same for other immigrant communities, like the Asian, Haitian and African, communities. Most shocking for a field nerd: that all the field organizing is based on how many phone calls are made and doors knocked, not people engaged and vote plans documented. A ward can be counted as canvassed if just one person in that ward is talked to, my source said.
In Bucks County, a purple rural/suburban area that has been at the epicenter of fights over school boards, critical race theory and transgender bathrooms, activists battling the “Moms for Liberty” crowd have fought back effectively. But a veteran organizer there tells me that the Harris campaign, which is organized and staffed by outsiders, isn’t interested in collaborating with local or state races, preferring instead to organize the many out-of-state volunteers flocking in by bus each weekend to canvass. Local candidate campaigns report that they can’t even get Harris literature from them.
It doesn’t have to be this way
As I report in my Nation piece, there is a big independent field operation running alongside the official Harris campaign in the state. For example, the Working Families Party is steering a coalition that includes One PA, which focuses on empowering Black communities and which says it has knocked more than 600,000 doors in Philly, Pittsburgh and Delaware County; Asian and Pacific Islanders of PA, who in addition to canvassing have tallied more than 4.2 million calls, including 18,000 in languages other than English and Spanish; Make the Road PA, which aims to hit half a million doors by election day, and unions like the SEIU and Unite Here, who together have already knocked over 1.2 million doors. The AFL-CIO’s numerous local labor councils are also coordinating a lot of door-knocking.
These folks are also doing some innovative things. The WFP has more than 20,000 people in its Pennsylvania Reach program. Unlike the DNC and DCCC, they’re using it to match to lots of local voters and then send them personalized postcards with photos from their friends. It’s also working with about 150 precinct captains in Philadelphia–people who live in the neighborhood whose voters they are responsible for engaging. And as in 2020, they’re running a Joy to the Polls program aimed at stimulating the early vote in Philly, as well as sending organizers to barbershops and salons.
But with some important exceptions, like Pennsylvania United’s deep canvass in western PA, most of this voter outreach is focused on quantity over quality. And as I noted in The Nation, people doing the kind of more in-depth conversations that are the heart of deep canvassing tell me that while many people may be telling door-knockers they’ve voting for Harris, enthusiasm levels vary widely and sometimes, perhaps often, these voters are just tired of being pestered and say whatever they think will get a canvasser to check a box and leave them alone. As Kipchoge Spencer, a veteran organizer from California who has spent the last week doing a mix of conventional and deep canvassing in and around Philly, told me, “Almost everyone I speak to in Philly, whether on the street or at the door on a walk list, immediately says that they’re registered and voting. In conventional canvassing, at this point you basically do some variation on making sure they have a way to the polls and thanking them. In deep canvassing, you engage them in story telling anyway and very often find out that they hadn’t been as committed to actually voting as originally expressed but have become substantially more so over the course of the connection.”
This past weekend, Changing the Conversation Together, an independent deep canvassing organization that has been working in Philadelphia and its northern suburbs since the 2020 election, ran 300 volunteer shifts, tallied 900 in-depth one-on-one conversations with voters, and nailed down 800 vote plans with them. (Notice how these metrics are different?) Adam Barbanel-Fried, the group’s founder, told me he often encounters voters who say they’ve been canvassed already by other groups. But often the training those groups give their volunteers leaves gaps.
“Yesterday I ran into a woman named Maria who opened the door saying she's already spoken to someone, and that she already voted,” Barbanel-Fried told me on Monday. “To which I said: ‘That's great how'd it go?’ Maria responded that she actually had received a call from the city saying her ballot had been canceled. Over time I sat with her, looked into the issue, figured out she was confused by different mailers she'd received (those from advocacy groups, campaigns she thought were from the city) and ultimately we got one of our volunteer drivers to bring her to the satelite election office where she was able to cure her ballot and vote.”
Barbanel-Fried’s canvassers have also come across group homes where dozens of people weren’t registered to vote; one volunteer focusing on these special projects personally registered over 50 people in three days. “To me what this speaks to is the time, energy, and patience it takes to find and engage with people who are at risk of not voting,” he says, as well as training that emphasizes patience over speed.
Last Saturday, I spent the day in Easton, a working class city of 28,000 hard against the Lehigh and Delaware rivers doing a deep canvass under the tutelage of Kathleen Campisano. The native of Louisville, Kentucky has been building her independent deep canvassing project Canvassing Connectors in eastern PA for the last two years. Dedicated volunteers, many from New York City and its northern suburbs, have been walking Easton’s streets for months–well before Harris was the Democratic candidate. Now Canvassing Connectors are neck deep in the city, aiming to deliver at least 2,000 low-propensity voters to the polls. Campisano teaches her door-knockers to take their time getting to know the voters they are trying to persuade, people who generally do not vote in every election and who feel ignored or exploited by the political process. “Remember we don’t know shit about these voters,” she admonishes her mostly white, middle-class, middle-aged and female volunteers. Meaning, if you approach people with humility and respect, they are more likely to open up and trust you.
I was very impressed by their whole operation, from how they onboarded volunteers, to the serious 90-minute training they required everyone to take, to their post-canvass review. With two weeks left to go, they are trying to get volunteers to come back multiple times to the same turf. Who does that? When I raise with Campisano all my concerns about how national Democrats and big outside groups approach getting out the vote, she’s both critical and sanguine. “There is not a built-in 'train everyone every time' with the Dems,” she says. But, she adds, “That's one of the things grassroots groups have – appreciation, respect for time, everyone-is-a-leader thinking.” Since 2016, such grassroots groups have been popping up all across Pennsylvania. “We don't have to abide by a set of pre-arranged rules,” says Campisano. “We get to invent as we go, based on our years of experience or first-timer ingenuity.”
These groups are Pennsylvania’s silver lining.
In other news:
—This coming Wednesday, October 30, from 9:30-11am, I’m moderating a panel at Fordham University in Manhattan on “Memes, Movements and Money: The 2024 Digital Campaign.” I’ll be joined by Kyle Tharp of FWIW Media, one of the country’s top experts on how and where the campaigns are spending on paid digital advertising; Glennis Meagher, the founder of Generator Collective, which works closely with social media influencers and political campaigns; and Prof. Kaia Niambi Shivers of NYU, founder of Ark Republic and expert on Black media. The event is free but RSVP for tickets here. It will also be streaming online, but you have to RSVP to get the link.
==The same day, from 1-2pm ET, Media Justice and Witness are holding a virtual briefing on “Elections in the Age of AI.” More info here.
—CNN reports on how deceptive online political fundraisers, mostly Republican but also Democrats, have preyed on the elderly, including people with dementia, extracting huge sums sometimes in the mid-six-figures.
ordered tickets. As you well know, there was a strong campaign to have the Democratic
Party's institutional structure harness the incredible talent of the grassroots constituencies. Small, very small, progress was made on that front. To the credit of the Grassroots community, as you describe so well in your article, we marched to our own drummer, unencumbered by so many layers of nonesense. Frankly, all I heard was how much respect the Biden campaign staff had for " ground game activists" but that turned out to be a lot of words but no strong engagement. I am praying you are right about the outcome, and fully agree as to who deserves the credit.
Great as always... As you point out, however, it is impossible to get accurate data from any campaign before election day and afterwards it is a crap shoot. There is really no third party verification for anything around campaign operations. So it really is up to the spin from Staffers and what they want you to believe. The only data that will matter will be votes and who wins and loses.