Dan Ancona’s article “Power to the Edge: Obama’s California Field Operation from the Future” at Personal Democracy Forum‘s ‘techPresident’ tab contains valuable insights about the strategy and tactics that empowered Obama’s quest for the Democratic nomination. Although Obama did not get a majority of CA delegates, the tactics his campaign deployed there proved critical in his other primary victories, winning the Democratic nomination and building the coalition that elected him.
Ancona, Project Director of California VoterConnect, likens the Obama primary campaign to the British victory in the Battle of Trafalgar and Genghis Khan’s Mongol invasions, both of which involved unconventional techniques of precision targeting to overcome “a largley centralized and monolithic force.” The strategic and tactical implications for politics are far-reaching. As Ancona writes:

The Obama campaign is distributed and bottom-up in a way that is the clearest example of what a post-broadcast, distributed and participatory democracy is going to look like. The evolution in campaign tactics happening right now closely parallels what’s happening in the military, corporations, government and other large organizations. The dropping costs and increasing reliability and flexibility of information technology is having profound effects on how these organizations make things happen.
This transformation was dubbed “Power to the Edge” in 2003 by David Alberts and Richard Hayes, two Department of Defense researchers with the Command and Control Research Program. Their book is surprisingly readable and engaging, and available in its entirety on-line at that link. It may be the best written government document of the 21st century so far. The authors are unabashedly aware of their book’s broader ramifications, stating in the preface that “[T]his book explores a leap now in progress, one that will transform not only the U.S. military but all human interactions and collaborative endeavors.”
A good political analog to Alberts & Hayes is Joe Trippi’s too-often overlooked post-2004 tell-all, The Revolution Will Not be Televised, where he laid out the broad contours of the transformation from the transactional, broadcast, TV-based political era to the relational, participatory, distributed, internet-based one. Trippi’s a terrific storyteller and it’s packed with exactly the kind of inside dirt that both serious and armchair politics junkies love. But it goes beyond that, becoming something of a handy guidebook and roadmap grassroots activists working to align their local efforts into something larger. (Dean campaign veteran and TechPresident contributor Zephyr Teachout’s widely recommended new Mousepads, Shoe Leather and Hope looks like it takes up the similar line of argument.)

…The pre-Super Tuesday strategic challenge to the Obama campaign was enormous: smaller, easier to reach states with less expensive media markets were a better use of resources. Just punching through the name-ID barrier alone in California would have been exorbitantly expensive. Countering the soft but widespread early negative impressions generated by the email smear campaigns that were run against him were another factor. A ten million dollar ad buy is barely a drop in the bucket in California’s high-priced media markets. But how could they stay competitive in California, keep it from becoming a delegate-count wrecking blowout and force Senator Clinton’s campaign to contest it? How could the campaign leverage large numbers of supporters with hardly any paid media air support, limited visits from their A-list surrogate crew, and with only a skeletal paid staff?
The answer to their dilemma had actually been cooking in the California grassroots since 2004. I was involved with the 2004 grassroots Dean campaign in Santa Barbara, California, and from that experience (along with the dozens of other campaigns I’ve worked for both as a volunteer and, more recently, professionally), the outline of what a powerful web-based tool set would look like was starting to become clear.
First, a campaign would need to invest in a social networking tool that would let early supporters start to self-organize. Ideally the tool would let people find and connect with each other, blog, easily manage email lists, set up their own house parties and events, and maybe even do a little small-donor, person to person fundraising. (Or rather a lot of that kind of fundraising, even.)
The second component would involve some way to turn all that self-organized volunteer energy into a field program, and thereby turn volunteer energy into votes. The tool to do that was some kind of web-based, distributable voter file system.
This is exactly the set of tools the Obama campaign chose to deploy. The social networking tool on my.barackobama.com (an instance of Blue State Digital’s toolset, which many other progressive organizations also use) let people self-organize and opened up some strategic opportunities for the campaign. Then for the second phase, they tried something bold: a distributed deployment of the Voter Activation Network (or VAN), the voter file tool provided by the Boston firm of the same name. A little background on the VAN is necessary to understand this decision.
The Battle of the Boiler Rooms
In the late 1990s, Democrats and progressive groups were far behind the Republicans and their conservative counterparts in voter file management. The primary Republican voter file tool was the fearsome “Voter Vault,” which, while accessible only to a small group of party and campaign insiders, was deployed in all 50 states and enhanced with numerous commercial data fields, reputedly (but probably not seriously) including individual pizza topping preference. (“The GOP knows you don’t like anchovies“, LA Times, June 2006)
The VAN, founded by Jim St. George and Mark Sullivan in the late 90s and now employing upwards of 30 people, has origins as a custom-built tool for Iowa precinct organizing. It’s now in it’s fourth major generation and available in some form or another in all fifty states. The qualities of the VAN are a topic for another article, but in short, it is a truly remarkable piece of engineering. It has the flexibility to be deployed to individual users down to the precinct level and the power to deliver both precinct walk and call sheets and virtual phonebank capabilities efficiently and at scale. [Full disclosure: as Project Director of California VoterConnect, a mission-driven sort of reseller of the VAN in California, I stand to benefit from more people knowing how completely awesome this software is.] The Obama campaign clearly saw the opportunity the VAN offered, and they seized it. The deployment and field program was run by folks who had joined the campaign through the in-depth, personal-narrative based Camp Obama training sessions (see Zack Exley’s excellent coverage of these events from last August for more information). These training sessions were scalable and worked for serious pro-level volunteers making deep and early commitments to the campaign – sessions that took an entire long weekend – down to a 10-minute streamlined version that the campaign was running hundreds of people through in bulk trainings towards the end of the campaign. These effectively set up a basic framework of joint purpose and even light accountability among the organizers who participated in them.
A quick VAN run-through was also provided in these sessions. They took people through creating lists, making calls via the VAN’s virtual phone bank tool, and putting data back in to the system. The VAN voter file system itself was wrapped inside a site that also delivered training via video and written tutorials, and connected people to support contact information. Nearly all of this was written by technically inclined volunteers, with a minimal level of paid staff oversight.
This allowed the campaign to take maximum advantage of the power of in-group, neighbor-to-neighbor organizing. If you’ve made calls or knocked on doors both in another state and in your own neighborhood, you’ve seen what a dramatic difference the neighbor-to-neighbor approach can make. One of the shocking things about field organizing is learning how nice strangers can be, but it never hurts if you can start a conversation with “I live around the corner, and…”. Although some of the data seems to indicate that in-group contact does not make a significant difference with regard to race (see Niven 2004, for one example), experience seems to indicate that geographic proximity can make a significant difference.
Making the Numbers Count
The results of all this neighbor-to-neighbor campaign architecture border on the astonishing. In a brief post-primary conversation with Patrick “Map The Vote” DeTemple, one of the volunteers-turned-staff who built this organization, he laid out some of the numbers:
– The social networking tool created more than 105,000 contacts statewide.
– Of those, more than 10,000 signed up to be precinct captains
– 7,000 or so VAN login accounts were deployed
– Of those, four to five thousand became active
For the last ten days of the campaign, this voter contacting machine averaged more than 100,000 attempts per day, including more than 240,000 on one Saturday, yielding more than a million contacts total. 50-60% of the calls were generated by precinct captains, with the balance generated by more traditional phonebanks run out of the field offices. The campaign also cooked up an innovative, boot-strapped predictive analytic model for finding new voters. This model used a combination of results as they were being generated and demographic data from the voter file to find more demographic pockets to focus the volunteer energy on reaching.
Bottom line: a total of about 4.88 million votes were cast in the Democratic primary. More than a fifth of the voters got a personalized contact from the Obama campaign, and for more than half of those, that contact came from a neighbor. Those numbers for a small-town, field-driven city council campaign would be solid and respectable. But for a statewide campaign here, they border on the unreal.
This effort, massive though it was, was not enough to put Obama over the line first in California. Campaigns are complex systems embedded in a complex society, and winning is a function of controlling as many variables as possible. Senator Clinton’s campaign ran a nearly flawlessly executed absentee chase program in this state using early-vote data obtained by Catalist. The schedule certainly played a role as well, as the volunteer energy began growing exponentially only after the holidays and Obama’s victory in Iowa. Brent Messenger, who was often the only paid staff working out of the maelstrom of activity that was the San Francisco office, said they felt as if they’d “been handed two pistols and a bunch of machetes and told to hold off the invading hordes.”
This gambit enabled resources to go towards a clear win on February 5th, followed by 11 subsequent wins against the most powerful national Democratic machine since the Kennedys. All this happened after starting from practically nothing a year ago. Win or lose, the power of these tactics is clearly evident.
The Future of Distributed Campaigning
Many questions remain, and the answers to these questions will only come from further experimentation. The first and biggest challenge is how to grow the size and strengthen the capacity of the bottom-up, participatory system. Can this distributed, networked approach eventually become large enough and strong enough to repel smear attacks? Because person-to-person contacts take time to generate, how can volunteer energy be best created and harnessed before early and vote by mail voting starts? And finally, these developments aren’t happening in a vacuum. How will tactics evolve when two fully networked and distributed campaigns go up against each other? Senator Clinton’s campaign has utilized some distributed tactics, mostly for virtual phone banking and fundraising, but there is a lot of unexplored territory in this area.
One huge question facing activists in California and nationally is how other campaigns and organizations can deploy similar tactics, without the spectacle of a national campaign and a leader like Obama. The VAN is available in all 50 states, although usually through state parties that may or may not choose to deploy it in a distributed manner. But even without a high-profile race at the top of the ticket, thousands of smaller campaigns and individual efforts can add up if they were to distribute it. Activists could then better build on their efforts each cycle, because people who get involved have a tendency to continue to stay involved – if and only if an effort is made to keep them engaged. [More full disclosure: California VoterConnect’s purpose to enable exactly this scenario to happen.] “Power to the edge” has clear, positive implications for social justice and democracy. The Obama campaign’s California operation, while it didn’t lead to a victory in this state, enabled a highly successful national strategy and was a pointer towards what could be an exciting and epochal change for American democracy – if we choose to build it. This epic campaign season, with two amazing campaign machines based on radically different strategic emphases battling it out through all fifty states, should easily go down in history with the Battle of Trafalgar and Khan’s Mongol invasions.