Fixing Field: How Democrats Got Off Track and How to Rebuild

Since the 2008 Obama victory, Democrats have been enamored with the role that a big data-informed field campaign can play in propelling candidates across the finish line in November. Obama’s two campaigns reinvigorated field. Aligned with what was then highly-sophisticated data targeting, the Obama field program was disciplined and energized, with hundreds of organizing offices and a steady stream of volunteers in competitive states across the country. Soon after the impressive 2008 and 2012 victories, Democratic campaigns big and small moved to emulate the Obama model for grassroots, door-to-door campaigning.

More recently, though, some folks have begun to question Democrats’ reliance on field. David Shor has repeatedly flagged how other forms of media may actually be cheaper when all the costs of running a field campaign are taken into account. In a since-deleted tweet, Simon Bazelon noted that Coca-Cola, perhaps the most powerful brand name in the world, doesn’t go door-to-door encouraging you to buy a can.

Recent election results have also suggested the luster of fieldwork has been fading, too. Hillary’s 2016 campaign, despite a large advantage in the apparent size and scope of the field program, was shown to be a major black hole, where data being received by canvassers did not update the central campaign’s all-knowing model. When Virginia Democrats in October 2021 said that the election was looking tight and noted that Democrats’ field advantage would power us over the top, to the fatalistic among us, it indicated that we were going to lose.

Based on our time since 2017 working with over 200 state legislative campaigns, we agree that Democrats have lost their way on canvassing. Using the real story of the Obama 2008/2012 model for success, we highlight four ways that most campaigns, particularly at the state and local level, are currently missing the mark on field and how they can course correct to give themselves a fighting chance in 2022 and beyond.

1. Most field programs are time- and cost-ineffective approaches to data collection, rather than organizing.

If Coca-Cola relies on Super Bowl ads, what is the point of going door-to-door anyway? It lies in the benefits of actual, real human contact. There are two fundamental mechanisms at play. The first is social pressure. A door-to-door canvasser can hopefully encourage you to do something because other people in the community are doing it, too. Think of who tries to sell you things door-to-door, versus in an ad. New windows, solar panels. Keeping up with the Joneses stuff. That in-your-face conversation is about conveying a social norm to the other person.

At the same time, there’s also the powerful principle of reciprocity. By going to someone’s door, as a candidate or as a volunteer, and offering to listen to the issues and concerns of the voter in front of you, you can establish trust with a voter and move them in the direction of what you want them to do: vote for your campaign.

Those principles were at the heart of how the Obama campaign structured its field program. The campaign spoke of “values oriented” organizing, engaging volunteers on how to tell their personal story to connect with the voter at the door. Over a decade later, this style of campaigning has been rebuilt as deep listening. We’ll talk more about how to deploy that canvassing in a moment.

So, if that is what makes canvassing valuable, what have most campaigns been doing instead? If you have been to a canvassing launch in the last few years, your training has probably mostly consisted of a detailed explanation of how to fill in the walk packet (or MiniVAN app). Most canvassing “scripts,” the volunteer’s roadmap to the conversation, involve very little listening. Rather, the volunteer offers a quick comment and presses the voter for their support for the campaign.

So, volunteers are not being dispatched to make a human connection. They are being sent out as census takers in pursuit of so-called “1s and 2s,” the voters most likely to support the campaign. This is a very strange way to collect population-level data in a constrained environment.

2. Most campaigns don’t actually have armies of volunteers

The Obama campaign’s field success was powered by as many as 2.2 million volunteers, thousands of field offices, and the largest field staff in history up until that point. Most campaigns don’t have that, particularly at the state level. Our rule of thumb, when working with a state senate or house campaign, is to ask whether the campaign can get 30 reliable shifts of volunteers per weekend. This is, to most, a big ask, and to answer in the negative is to confirm that a broad-based field program is not really aligned with the campaign’s capabilities.

Meanwhile, others have largely encouraged campaigns to knock as many doors as possible. This creates a bit of a perverse incentive. With insufficient volunteers to do it right, campaigns are further encouraged to push canvassers to have the type of quick, ineffective conversations described in (1) above.

3. Most campaigns’ targeting isn’t very sophisticated

The Obama campaigns ushered in an era of more rigorous targeting of voters, using voter file and commercial data that had not been utilized by political campaigns in the same way before. And today, every campaign is plugged into an extensive CRM system, VoteBuilder, that can provide demographic information and modeled predictions about voters’ political preference, education levels, and ideology.

One of the biggest misconceptions outside of the campaign infrastructure is that the presence of this data informs much of the day-to-day workings of the typical state-level field program. It does not. That big shift in how people are voting in this country you’ve heard about – education polarization and/or racial resentment sending non-college voters away from us and college-educated voters toward us – is not something most campaigns are probing in their walk lists. That lingering concern that voters of color have trended away from Democrats since 2012? Also not being addressed through microtargeting within most state house or state senate campaigns.

Most campaigns use simple heuristics to identify voters and use their field program to work toward a “win number.” This number is an estimate of the number of votes it will take to claim victory in November. Those volunteers are sent out to try to help the campaign count up to that number. That results in a very broad, and, due to the resource constraints identified in (2), not very deep campaign. Campaigns aren’t focused sufficiently on the 2-3% of voters who might decide a competitive race and they aren’t getting enough specific data from the swingiest populations (like low turnout or highly cross-pressured voters) to update and inform campaign strategy to react and win.

4. Too much of our canvassing happens too late

Getting engaged too late is a broader problem in state legislative politics – we have documented how money comes far too late to most campaigns for them to deploy resources effectively. But it matters in particular on the field front. The research on persuasion indicates it has to start in a sweet spot: too late and too much other engagement has drowned out the persuasive conversation and too early and the voter loses the thread. So much canvassing happens after Labor Day. That means that even if campaigns are trying to persuade, they do not have the time to do it successfully.

And frequently, we mischaracterize who is a pure turnout door knock, who just needs a reminder, and who requires persuasion. Even “base” voters may need persuasion about why this election matters or about why their dissatisfaction with Biden shouldn’t outweigh supporting this candidate. In an age of rapidly shifting voter preferences, early knocks are a key way to get ahead of changing voters.  

What should we do instead?

We should, at once, turn on its head the way state and local campaigns are currently running their field programs and return to the core principles that drove the Obama campaign over a decade ago.

Statewide campaigns should focus on the broad communication needed to remind and engage the voters who just need a small nudge to get back out there. The local campaigns should set their sights very clearly on how to have repeated, effective contacts informed by the latest in campaign science, like deep canvassing. That means training volunteers better and creating substantially more focused field universes, particularly for the candidate and key super-volunteers.

Most importantly, campaigns should use early field efforts to understand what voters are saying and to test messages. In the wake of the 2016 and 2020 elections, we read repeated post-mortems that noted that Democrats were “surprised” by the demographic shifts happening in races, whether that is the bleeding of working-class Hispanic support or the gains in college-educated voters. A well-designed field universe focuses less on hitting every single person lightly and more on trying to figure out what we can learn about the electorate from the voters we do engage.

What is EveryDistrict doing to advance better field programs?

Since 2017, EveryDistrict has raised money for state legislative campaigns, supporting over 200 races across the country. This year, we’re diving deeper into helping our candidates run more strategic campaigns that help us win in the purple-pink districts that we need to build majorities. We’ve already endorsed 26 state legislative campaigns and counting in the battleground states of Arizona, Michigan, and Pennsylvania this year.

Are you excited by this work? Chip in here to make a donation.