Tools for Technocrats, Not Citizens: Today's Political Tech Industry

Higher Ground Labs' 2022 landscape report suggests the field is yielding diminishing returns from its top-down, capital-intensive approach to organizing voters. Plus, more on AI and TikTok.

On the center-to-left of the political spectrum, there is probably no organization that has done more cutting-edge work in the last five years to understand and nurture the organizational and technological ecosystem than Higher Ground Labs. Its network of investors, mentors and start-ups stands out for its sophistication and attention to detail. Getting into its annual investment pool matters. And so too does its annual Political Tech Landscape Report, which just came out this past week.

I’ve been reading these reports since HGL’s first one came out in 2018, and a number of things stood out from this one. First are the findings that the report’s authors, Russell Mindich and Hillary Lehr (along with a few dozen expert contributors), chose to highlight:

• Paid political advertising continues to be the biggest engine driving the field and the current advertising landscape is in flux, as new streaming services are starting to embrace political ads while older platforms like Facebook stagnate. Prices are spiking, but that hasn’t stopped campaigns from a huge increase in spending on paid media.

• The fragmentation of the media landscape has also led to a full-on embrace of so-called “influencer marketing,” as organizations from the Democratic National Committee on down are choosing to use both paid and unpaid touts to reach voters online.

• People are starting to recognize that the text messaging gravy train is over, with declining contact rates, oversaturation and scams cited by HGL as concerns leading to a rethink of this tactic.

• Usable political tech for small-scale campaigns is still missing – “down-ballot candidates still lament the fact that they cannot find, afford or make use of other critical services, ranging from website development to digital content management to district data and demographic analysis tools.”

• And finally, the consolidation of key parts of Democratic infrastructure in the hands of private equity (NGP VAN’s acquisition by Apax Funds) has created a clear worry about the future of the party’s data infrastructure.

These are all valid findings and causes for concern. But I think the report also reveals a much bigger problem: tech-intensive ploys to prop up a declining Democratic brand are not working. Take this graph from the Democratic National Committee, included in the report on page 18.

The Y-axis of the graph isn’t included, so it’s not clear how precipitous a drop in phone and door-to-door contacts it’s actually showing. Elsewhere in the HGL report, the authors note that while voter outreach efforts have tripled since 2014, the last midterm with a sitting Democratic President, in part thanks to new tech like texting, scaled use of phone dialers and canvassing apps, contact rates for these traditional methods have been declining for multiple cycles.

Why are fewer people responding to the most basic way that campaigns try to connect with voters? One could argue that some of the decline is due to COVID, though that’s contradicted by the (temporary) improvement in door contacts shown for the 2019-2020 cycle. What this graph says to me seems clear: at the same time that spending on paid media along with a proliferation of new tech-powered tools and services has inarguably jumped, fewer people want to talk to Democrats when they get called or door-knocked. Is this just a result of over-saturation? Maybe yes--Americans received nearly 1.3 billion political texts in October 2022 alone, the report notes. Or is this a sign that the tech- and data-intensive approach to cultivating Democratic voters is exhausting the grass-roots, the same way that chemical-intensive crop cultivation leads to depleted soil and reduced returns?

There’s another section of the report that also ought to be highlighted in bright, blinking red, which starts: “This cycle, many practitioners lamented a shortage of volunteers out in the field.” Why? “A combination of lingering effects of the pandemic, a regionally-segmented enthusiasm gap, and the proliferation of digital advocacy tools that allow people to contribute to political efforts from their couches.” You could also add to that an over-reliance on easy-to-do low-tech but low-impact persuasion activities like mailing postcards to voters. The report suggests that what’s needed are new approaches to improve contact rates, asking, “How might campaigns be incentivized to prioritize long-term organizing and relationship building over burning out voters, volunteers and donors in pursuit of hitting shorter-term fundraising goals? How do we make higher quality contacts so people engage rather than tune out politics?”

The degree to which the politics industry relies on young mercenaries rather than grounded locals to staff campaigns also gets some attention from HGL, though again this deserves more highlighting. “In many cases, a significant portion of campaign staff are not from nor have ever lived in the city, town or district in which they work,” the authors. And they’re often recent college grads from out of state. “This presents challenges surrounding perceived authenticity and localized subject-matter familiarity.” In addition, the authors say debriefs of the 2022 cycle found burnout, staff shortages, and exhaustion “across the board.” Again, some of the causes are not inherent to political tech work: “the pandemic, enduring toxicity of emboldened conservative extremists, and recession-triggered layoffs” all come in for mention, along with the low wages, long hours and instability of political work. But the result is grim: “we lose tremendous talent, institutional knowledge and innovation to other industries each cycle.”

I’d go further. This is unsustainable. While a professional class of campaign managers, tacticians, analysts, message makers and pollsters will continue to make a lot of money from ever-more sophisticated methods of farming donors and volunteers for money and time, nothing is going to invigorate the base until something beyond the scope of political tech happens: a new generation of empowered leaders committed to a completely different party brand and a different brand of party-building has to come to power. And given how much big money and institutional inertia actually drive Democrats, we shouldn’t expect any of those things to come by 2024. Efforts to unionize campaigns and political vendors are the only thing that, in the short run, are moving in a different and better direction.

Oh, And One More Thing

Here and there in the HGL report, you can find signs for where new leadership and new vision might come from—Maurice Mitchell’s seminal piece on building resilient organizations gets a shout-out, Stac Labs’ pioneering work supporting state parties and down-ballot campaigns gets notice, the Donor Organizer Hub’s program to train volunteer donor advisors who are closer to actual givers gets a positive mention too. I’m not nearly as impressed by the many references to “relational” organizing that HGL points to; while the term has certainly caught on, in practice it isn’t about building real relationships with voters and volunteers as much as it is about tapping (exploiting?) people’s existing contact networks for better top-down targeting.

But overall, it’s also interesting to compare the landscape as HGL saw it in 2018 to what it sees from 2022. These two images come from the respective reports it produced.

What jumps out for me are 1)  There are a lot more companies doing messaging and media, data analytics and modeling, and research (and a lot lot more in subcategories like digital ad targeting and information integrity, to no surprise); 2) some categories that were barren, like volunteer analytics, have been filled out; and 3) and a whole section, called “organizational infrastructure,” has been reapportioned, with parts like “internal comms and productivity” added to the “movement wide” sector and others shifted into “organizing tools.” For some reason, a 2018 subsection on personnel hiring and training, which included organizations like Wellstone Action (now called re:Power) and The Arena, is gone. Similarly, membership governance, which was an empty box in 2018, is no longer even included on HGL’s landscape.

And that just shows, again, that the political tech industry has mostly focused on building tools for technocrats, not citizens. To be sure, there are a number of companies and consumer-facing tools on the map that ordinary people use to build their own power—ActionNetwork, Countable, Change.org, NationBuilder, Frank and Unified are all included. (Full disclosure: I’m a small investor in Unified, which is just coming out of alpha this spring.) But there’s really little here that someone running a local community organization would see as designed to grow or sustain their community. Like the rest of the society in which it is embedded, the political tech landscape is largely a creature of extractive, surveillance capitalism.

AI, AI, AI!

—If your hair isn’t on fire already about how simulated media made by AI is going to disrupt politics and campaigns, read this piece by Shane Goldmacher in The New York Times (gift link) along with this chaser by Kyle Tharp of FWIW.

—The “AI pause” letter signed by Elon Musk and a thousand-plus other computer scientists has gotten a lot of attention, but let me save you some time and suggest you just read this short and to-the-point response from Timnit Gebru, Emily Bender, Angelina McMillan-Major and Margaret Mitchell. The quick version: Instead of focusing on the potential harms of systems more powerful than GPT-4, let’s address the actual harms resulting from the use of AI systems today.

—Say hello to the AI Impact Lab, a new initiative from a good friend and fellow traveler in the world of progressive tech, Taren Stinebrickner-Kauffman. She’s planning to focus on how the AI tsunami may transform the work of social impact organizations, for good and for ill.

Odds and Ends

—ActBlue has laid off 54 employees, about 15% of its staff, the first sign of a shake-up at the Democratic fundraising clearinghouse under its new CEO Regina Wallace-Jones. The ActBlue union says its “disappointed in the mismanagement that has gotten us here,” and claims that management “was unwilling to explore the alternatives [to layoffs] we brought to the table.” It adds, “These layoffs jeopardize ActBlue’s ability to fully serve our entities, donors, and the larger Democratic movement ahead of a critical election cycle.”

—I tried to make the case for not banning TikTok last week, but as usual Cory Doctorow states it even better. His bottom line (and mine): “America doesn't need a Great Firewall to keep itself safe from tech spying – it needs a privacy law.”

—Robyn Caplan, a senior researcher at Data & Society, reports that last week’s congressional hearing on a TikTok ban did not go down well with the TikTokkers she studies. Most, she says, are both deeply knowledgeable about data privacy and deeply skeptical that anything can be done to really protect their data, either from Chinese or American snoops.

—Writing for Tech Policy Press, Alex Engler, a Brookings fellow, reports on last week’s Summit for Democracy and its implications for the tech sector.

End Times

Fundraising, in the days of the telegram?