Lessons from Burlington

I’ll be making a presentation at Harvard’s Berkman Center for Internet and Society next Tuesday, February 7. The Berkman Fellows’ Luncheon (a series) will be webcast and on IRC at 12:30 Tuesday. The description:

Dean Done Right

For two years, Britt Blaser has sought a way to duplicate what Zephyr Teachout and Jim Moore and David Weinberger and so many friends of Berkman demonstrated was possible at the Dean Campaign. The Open Resource Group is releasing ORGware, but informally it’s called “Dean Done Right.” Britt will preview what the Dean campaign would have used if it had started three years later, in the age of Web 2.0. Background material here.

Some of the last few posts have been intended as background material for the presentation. They were:

• Sliced Bread 1.5: Technical limitations of SEIU’s Since Sliced Bread idea contest.
• Since Canned Beer: Trusting your members, metrics of protest and 16 requirements for distributed policy formation.
• What that means is…: Definitions of the 16 technical requirements for distributed policy formation.

In addition to the lofty goals and tech talk, I appreciate the archaeological aspects, remembering what it was like to be on the third floor of 60 Farrell Street, South Burlington, VT in late 2003. They were heady times and I felt privileged just to carry a few bags for those magnificent people. The Dean Campaign was the latest in a series of unsuccessful attempts to forge grassroots activism into a tsunami deep and strong enough to overwhelm the cynicism of politics-as-usual. 

One lesson is that those 3-4 dozen tech-savvy people were the Dean web service, in ways that no other enterprise would or could have attempted. Understanding that is key to understanding why Dean Done Right has not been built before.

Self-Indentured Servants

Dean’s almost-unpaid net slaves were able to maintain the illusion of an automated web service at deanforamerica.com and blogforamerica.com with a lot of hands-on tech kludges, customized TCP/duct tape and IP/baling wire. Fortunately for the campaign, we net slaves rowed our own boats to Burlington and chained ourselves to the desks for the duration(many of us were even cheaper than slaves, feeding and housing ourselves). None of us was paid anything like what we’d earn in private enterprise. 

Dean’s suite of impressive and smooth-running web services were so dependent on this human network’s packets of inspiration that few other organizations could have pulled it off. The Dean campaign spent almost nothing on the Internet, yet it employed resources that would have cost about $280,000 per month in Burlington alone (c. 40 people x $7,000 incl. benefits). It would have been worth it, but they never would have made that investment.

And the tech? Fuhgeddaboutit! A hodge-podge of Movable Type, Meetup, Convio, many disparate databases, etc., etc. Those amazing people made it all look to the world the way a web service should, but the rules were that there had to be at least 2 people on site at all times, and a SWAT team on call. One Saturday evening in December 2003, Halley Suitt and I had dinner in Burlington and discussed this (among many other things, as you can well imagine). We decided to return to the office at 11pm – on a Saturday – to do a head count, which I guessed would be at least 45. The actual was 67!

When we see a successful web platform, we assume there will be a couple of even-better copycats on line within a month. Maybe that’s why Dean’s people-power web illusion set up the dismay that so many organizers feel when they reach for a tool like Dean had and they discover that there simply aren’t any available. Here’s Susan Crawford‘s recent amazement, and Harish Rao‘s revelation from a year ago, and even Nicco‘s not sure where this is going. 

Andrew Rasiej‘s campaign also foundered on the shoals of promoting a social network for New York City that couldn’t be found on the campaign’s web site, and which never formed around his candidacy. You may recall that I promised Andrew and Micah Sifry that Open Resource Group would produce a web platform to support his political platform, but we screwed up and couldn’t deliver. We were all amazed that there is no campaign-in-a-box.

That is the problem that ORGware aims to fix.

Back to the Syllabus

What are the lessons for Tuesday’s Case Study of Burlington?  I had not studied this problem space before 2003, so I took the thrilling possibilities at face value. I’m glad I did, or otherwise my midsummer enthusiasms might have been prematurely chilled by Micah Sifry’s harsh mid-winter explanation: these enthusiastic efforts crop up periodically and have always failed. Micah shared this insight with Doc and me over coffee a couple of days after Dean’s New Hampshire defeat, citing the Goldwater, McCarthy and Perot campaigns. 

I’ve been working on the mechanics of the solution since then. Without something no worse than ORGware, we’ll never send Mr. Smith back to Washington.

Here are the overarching ideas that can help the next idealist overcome this pattern of excitement and rejection:

• Have a Seed Crystal
  As with semiconductors, your social algorithm needs a seed to start crystallizing your social network. Howard Dean was such a seed crystal, but John Kerry was not. Barry Goldwater was conservatism’s seed crystal, but George H.W. Bush was a damper. Al Gore was nobody’s seed crystal. Bill Clinton is a force of nature, not a seed crystal – no movement persists. Harley-Davidson is, Buick isn’t.
• Smart = Busy = Distracted = Unfocused = Stupid
  Design for the Largest Common Denominator
• Roll the DICE
  Most users won’t adopt a new application or convene over at Me
  etup or Base Camp to do your work if you don’t provide all the functionality they’ll eventually need on your web site. Unlike techies and activists, they won’t go build their own blog or organize events at eVite. They need to be able to do everything on your site, so you need world-class programming in an interface that’s DICE-compliant:
      Deep
      Indulgent
      Complete
      Elegant
• Stepping Stones
  People need to move to be a movement. Online activism is the wild west for most people, so they need to move in baby steps (from browser screen to browser screen) by which their slight interest evolves into an Aha! moment, on to active debate, to recruiting their friends, to investing money, to voting. Even a city slicker might feel secure standing on a flat rock in a roaring river. If there’s a similarly hospitable stone a short step away, flat and dry and not too smooth or mossy, you might just step on it if it’s in a direction you’re a little interested in. Pre-build these small safe havens that are worth visiting for their own sake. Make each one comfortable enough to hang out for a while. Your future advocates don’t know they want to cross your river.
• Deputize Celebrities
  In the summer of 2003, People at Dean rallies wanted autographs from Zephyr, Matt and Nicco. The more celebrities you create in your movement, the greater its heft. Every celebrity is a new seed crystal and you need all you can manufacture.
• Lattice of Engagement
  Locate every member in identifiable relationships. It’s a movement, so everyone wants to know someone who’s more of an insider (mini-celebrity) in this exciting enterprise, and they want to be important to others who are newer to the movement. The movement wants constant news about its movement and the members creating motion.

• Strawberry Roots Activism
  Grass is nice, but your front lawn is dependent on you for seed, feed, water and weeding, each seed pushing out just a few blades for us to admire. Rhyzomes, like strawberries and crabgrass, are more creative. Once started, they shoot out opportunistic runners which put down roots in hospitable circumstances. If the new plant prospers, it puts out its own runners, and so on. Strawberry roots activism may be the future of politics.

• Federation of Hierarchies – the basis of strawberry roots activism
  Mass movements have to be egalitarian, which many activists confuse with hierarchy-free. But your movement needs all the hierarchies you can spin off, because work-group hierarchies are the only way to get things done. Let your people form hierarchies on-the-fly, where they can meet and do the work of politics by collaborating in the three areas that every movement needs: Discourse, Growth and Money.
• No one buys a Buick because GM needs the money, so . . .
  
• Govern Early and Often
  Discover each member’s values and maintain her personal values profile. Inspire each member to manage their values profile as assiduously as they manage their address book. Discuss values constantly and money peripherally (except when giving it energizes your members). Tabulate and aggregate the policy preferences of the members to transform unwelcome email broadcasting into a vibrant conversation. The Dean campaign’s professional policy advisers were supremely disinterested in polling the campaign’s most committed supporters and tabulating their policy preferences. They refused to have the candidate “tied to the explicit interests of his base.” Naturally, they got what they wished for.
• iTudes
  Movements need iTunes for Attitudes. Before iTunes, no one imagined that we’d need such a complex environment to buy and listen to music, but somehow we’re there now. It turns out that managing music was more complicated than we thought. So is democracy. With skill and luck, your people will spend as much energy expressing their political preferences as they now spend tweaking their music collection and publishing their favorites.

These are the background issues open for discussion Tuesday. I hope you’ll tune in to the webcast at 12:30 Tuesday and check in to Berkman’s IRC channel. We’ll save the chat log and will be setting up an ORGware site for all participants to extend the conversation.
1:47:28 PM