It’s the Data, Stupid (Again)
Data-driven campaigning gets results at the margins, and tomorrow many Democratic professionals will perhaps congratulate themselves for another job well done. But if the people at Trump's rallies or at Sanders' taught us anything this year, it's that being a name in a file isn't enough.
Why Trump is Losing
On my way back from last week’s Code for America Summit in Oakland, I stopped in Denver to visit with family there and to spend a day on the ground watching how the presidential election is playing out in a swing state. I wanted to see how the Democratic party’s vaunted field operation was working (and to be honest, to add a few hours of my own shoe-leather to the Colorado Democratic party’s get-out-the-vote effort). So we were in a swing House district, in Aurora, where Democrat Morgan Carroll is challenging incumbent Republican Rep. Mike Coffman in a race that is rated as a toss-up. As luck had it, not only did I get to see Senator Bernie Sanders keynote an evening rally for Carroll, later Saturday night I got to attend Donald Trump’s campaign rally in Denver.
Going from the Clinton bubble to the Trump bubble was more than a trip across the ideological chasm now dividing our country. It also illuminated the tremendous differences in how the Democratic and Republican parties have approached political organizing in 2016.
My day knocking on doors for the Colorado Democrats was defined by voter data. My friends and I were given two walk packets with about 140 addresses covering a 10-block cluster of streets. At some addresses, one voter’s name was listed; at others there were as many as four. As anyone who has canvassed for a Democratic campaign knows, we had been assigned “turf” from the DNC’s vast voter file, using a tool called Votebuilder that lets campaigns create target lists of likely voters to contact. We were focusing on people who had been identified as regular Democratic voters, and our job was to remind them to get their ballots in, and to get them to commit to a plan for when they would vote. Intensive research has shown that this kind of door-to-door campaigning is the most effective way to get people to the polls.
We were also collecting information that the state Democratic party would use to update its voter file each night, including whether a voter said they had gotten their ballot in already, if they needed assistance in Spanish, or if they were on the fence or voting for Trump. That way, by the next day of canvassing the universe of voter targets would be adjusted accordingly.
Later, at the Sanders rally for Carroll, we saw another aspect of the data-driven campaign. Every single one of the thousand or so people who wanted to get into the rally was asked to fill out a paper form with their name, email, phone number and to check off times when they could commit to canvass or phone bank for the candidate. Even though I said I was from out of state, I couldn’t get in without filling out a form. No volunteer left behind!
The Trump rally was the polar opposite. I heard about it not by email, even though I am on the Trump campaign’s email list, but by reading about it in the Denver Post. The paper said attendees needed to register in advance, so I went to the Trump campaign’s website to sign up. But upon reaching the arena, the National Western Complex, I discovered that no one was collecting tickets or any other information. Perhaps three thousand people filled the old stockyard for the rally, but there wasn’t a single Republican staffer on hand collecting volunteer information either before, during, or after the event. And even though I RSVPed online, I have yet to get an email or text message from the Trump campaign or the Colorado GOP asking me to help turn out voters. (Impressively, someone from the Colorado Democratic state party just texted me Monday night as I was finishing this piece to ask me if I could volunteer again on Election Day and if I would bring a friend.)
As my friend Dave Karpf, a frequent contributor to Civicist and author of an excellent new book on the role of data in campaigning, Analytic Activism, just wrote, “This 2016 election was supposed to be an important step in the forward march of progress, featuring campaigns that made significant advances in data and experimentation….Instead, we got Donald Trump.” (Indeed, four years ago when I wrote on techPresident, “Election 2012: It’s Not Facebook, It’s the Data Stupid,” my argument was that campaigns cared less about social networking—which the press was over-covering—and much more about getting voter data into usable forms.)
He adds, “If Donald Trump manages to win on Tuesday, after running a campaign entirely free of the last decade of data-driven know-how, it raises questions about the efficacy of the entire system. A Trump win means, in effect, that decades of research designed to organize and influence voters can be overpowered by a chaotic, from-the-gut performance—if the performance is riveting enough.”
I don’t think Trump will win. And we’ll know very soon if that is the case. But while many of us are hoping for that result, and will savor the historic impact of defeating a proto-fascist and elected the nation’s first woman president, there’s something else that will happen tomorrow that we ought to lament and try to change.
Tomorrow, somewhere in the range of 5,000 to 10,000 very special people are going to lose their jobs, willingly. They are the people who have worked day and night on Hillary Clinton’s presidential campaign, plus the thousands more employed by state Democratic parties where the most critical votes were cast to elect Clinton and defeat Trump. If I’m right about the result of the election, it will be fair to say that these people were the heart of the machine that killed a fascist. Woody Guthrie would be proud of them. But despite the fact that Trump’s movement isn’t going away, and that the only thing that stopped it was a massive and well-organized political machine built by Democrats and their progressive allies, nearly all the people who made that happen are getting demobilized. With their consent.
The one thing that they will leave behind, in addition to all the people angling for a job in the new administration, is the DNC’s voter file. As UNC scholar Daniel Kreiss writes in his terrific book Prototype Politics: Technology-Intensive Campaigning and the Data of Democracy, “It is an undisputed fact that the Democratic Party has built the most powerful political database in the world, and serves as the ‘obligatory passage point’ through which all of its political campaigns must go to gain access to it.” As Kreiss explains,
Presidential campaigns essentially rent the DNC’s voter file, which is in essence a collection of 51 different state voter files (including Washington, DC) that states share with the national party and are all hosted in the same system, Votebuilder. As campaigns use VoteBuilder, they enrich the voter file though all of their canvassing and voter contacts. The voter contact and identification data generated by campaigns are firewalled from each other during elections. However, after elections, this voter contact and identification data created by campaign volunteers and paid staffers goes back to the DNC and state parties, and is ultimately made available to subsequent Democratic candidates. Campaigns do not own these voter files that they helped to enrich.”
Nor do the people who did all the legwork to update the DNC voter file. If you worked on the Clinton campaign, or if you volunteered on its behalf, or on the behalf of any number of down-ballot candidates, you built that. But you don’t own it and you can’t use it.
Why does this matter? Because political organizing isn’t just about identifying names on a list and turning those names into check marks on a ballot. It’s about ongoing relationships and mutual obligations. Today’s data-driven campaigning, at least as it’s conducted on the Democratic side of the aisle, is hyper-transactional. Unlike “big organizing,” which Bernie Sanders’ organizers Becky Bond and Zack Exley describe in their new book Rules for Revolutionaries as “asking people to do something big to win something big,” today’s Democratic Party asks people to do something small to win something relatively small, the election of a few figureheads.
Today, the party, through the manifestation of all the state-level coordinated campaigns like the one I volunteered with in Aurora, is at its puffed-up best. But tomorrow it will deflate back to a hollow shell, a holding company for a vast and valuable database, and little more.
That is, until the next election, when the whole process gets restarted. As Kreiss notes, it is highly likely that whatever tools and tactics the Clinton campaign developed in house to help power its 2016 efforts will go fallow. In 2012, the Obama campaign invested heavily in developing internal platforms like Narwhal, an effort to create a unified voter ID for every voter they had any kind of contact with, and Dashboard, a tool that was meant to give field organizers a unified way to cut turf and keep track of local precinct-level voter outreach efforts. Both tools were only partially realized by the time of the election and afterwards they were essentially abandoned.
Walking the streets of Aurora, knocking on the doors of a delightful mix of lower-middle-class whites, Latinos, Blacks and Asians, I was struck by how far these people were from the give-and-take of political life. Of the 40 or so people we spoke to, maybe two or three responded like really engaged citizens. The rest were polite and friendly, but more interested in enjoying their day off or the football game or cleaning their yard then talking about the election. A knock on the door was welcome, but when I said I was with the Colorado Democratic Party, I got blank stares or thin smiles most of the time, not enthusiasm.
This is the two-dimensional world that data-driven campaigning is optimized for. It gets results at the margins, and tomorrow many Democratic professionals will congratulate themselves for another job well done. If the people at Trump’s rallies, and at Sanders’, taught us anything this year, it’s that being a name in a file isn’t enough. A political party can’t just be a vehicle for gathering votes. And the work of politics isn’t over when a campaign ends. It’s just beginning.