Grassroots Campaigns Inc's Great War of 2004 (p3): the Collapse of Leave No Voter Behind
by greg bloom, Wed Jul 19, 2006 at 07:11:23 AM EDT
In the first post of this series, I described my experience working as a field organizer for MoveOn PAC in the 2004 election:
The 2004 MoveOn PAC Leave No Voter Behind was not just a 'bad' experience. It was a soul-crushing experience.
I did not choose that phrase carelessly -- it is a sentiment shared by many of my fellow organizers who were out there on the ground. In the comments, they've called the campaign "ill-conceived,""cynical," and "disastrous." Since MoveOn has rehired the vendor that ran Leave No Voter Behind--Grassroots Campaigns, Inc--to run its 2006 ground operation, Operation Democracy, I believe it's important to open up a dialogue about what went wrong.
A number of GCI's managers from the campaign have taken issue with these posts -- they reason that all campaigns are "hard work," and that "kinks" and "snags" should be expected, especially since this particular campaign was the first time that either MoveOn or GCI had attempted such an operation.
But this glosses over some crucial context. GCI is in fact the youngest offspring of the Public Interest Research Groups family; Leave No Voter Behind was staffed by hundreds of PIRG/Fund for Public Interest Research managers, who took leaves of absence to work on the campaign. The PIRGs do, in fact, have experience in GOTV--on college campuses and state ballot initiatives--and though nothing on this scale had ever been implemented, in an important way this campaign was not so different from a standard PIRG/Fund operation. It was a new permutation of their well-worn model.
As I wrote in the "Strip-Mining the Grassroots" series (and it would help to read that post in conjunction with this one), the PIRG/Fund world is rigidly structured around this model:
The model is more than a set of guidelines -- it's a comprehensive campaign template that assigns the goals, schedules the time, scripts the interactions, and measures the progress of each participant. Those goals are defined entirely in terms of numbers -- numbers of recruits, numbers of members, numbers of dollars raised. It has three primary components -- recruitment, training, and canvassing -- which are compartmentalized and monitored in order to produce the optimal result.
In this post, I will explain -- in terms of this model itself -- why the Leave No Voter Behind campaign collapsed under a crisis of leadership that can be traced directly to the top of GCI.
From the outset of Leave No Voter Behind, it seemed that GCI had grafted the canvass model onto GOTV with relative ease. The basic infrastructure (group recruitment sessions, role-play trainings, and so on) remained the same. Instead of paid canvassers, we were recruiting volunteers; instead of asking for donations, the volunteers were identifying voters and encouraging them to vote. The scripted "raps" were rewritten accordingly.
Unlike the fundraising canvass, there were no 'quotas' -- but there were fixed (and large) goals, which were plotted out day by day, hour by hour. As organizers, our primary goal was to recruit precinct leaders by making cold-calls, primarily to MoveOn members. Supposedly, we would make six to eight contacts over the phone per hour, and schedule three or four of them for a recruitment meeting. Ideally, two or three would show for the meeting, and then one or two would sign on to be a precinct leader. So, according to the model, about one precinct leader would be generated per hour of calling. We would then train these precinct leaders to manage their own precinct team, working towards the goal of 90 contacted voters. Some of our volunteers would inevitably flake out. But if all went right, the organizers would call for four hours a day through the first two weeks of the campaign, then would 'reduce calling intensity.' We would spend the rest of the campaign training and leading our 25 precinct teams through the demanding triple-tiered canvass plan. From the very beginning, we had been warned that we would have to put whatever effort was needed into "getting the numbers."
As I explained in my last post, Leave No Voter Behind immediately fell far short of these numbers. The campaign launched a week late, and most organizers were without phone or internet connections for several days after that. So, things had gone wrong -- however, this just meant that we'd have to "kick it up a notch" in our recruitment calling, in order to bring in the numbers of volunteers that we needed. But then things got worse. Our 'Web Action Center' (WAC)--the internet-based application through which every step of our campaign was to be administered--began to severely malfunction, causing major interference in our recruitment, team management, and canvassing operations.
Now, a distinction needs to be made here between these dual setbacks of delayed launch and infrastructure collapse. The former entailed a significant decrease of the total amount of time in which we had to reach our goals. The latter represented a further decrease in total time (since the campaign was paralyzed through repeated computer crashes) but also, and more importantly, a significant and permanent increase of the time necessary to complete our tasks. This increase extended across the board: it now took longer to make and record a call, took longer to train a volunteer to work with a buggy system, and took longer to deal with those bugs ourselves. Out from these two compounding problems sprang the crisis of leadership.
In the comments of my last post, Mr E argued with this claim of 'crisis,' saying:
Anyone who has been through a campaign--of any sort--knows that plans often fail or fall-short. The key is to not give up, but make the most of what you've got.
Indeed. GCI now had somewhat less than five hundred organizers in the field, many of them with scant organizing experience, who had just barely enough time and resources to recruit perhaps fifteen precinct teams each. These organizers already had thousands of volunteer recruits who were struggling under an erratic system. But we were many more thousands short of LNVB's original goal.
'We can't change the date of the election,' our regional director told us, implying in the same breath that they would also not be changing our goals -- which meant that we still were expected to recruit the original number of precinct leaders.
After the WAC went down entirely, once it came back up with limited functionality, management increased our calling hours to six hours a day (or more: in a couple of unfortunate cases, struggling organizers were ordered to cold-call as much as twelve hours a day). Organizers were disallowed from working with their volunteers during the hours between 4 and 9pm -- the time when people (potential recruits and volunteers alike) are most likely to be available. (In another unfortunate situation, an office's management forced its organizers to turn over their cell phones during calling time -- so that their volunteers could not even get them on the phone.) This mandatory calling was extended all the way through to five or six days before the election.
Even under normal circumstances, this schedule would have been problematic -- it left us scrambling to find time to speak with our precinct leaders early in the morning, in the middle of their work days, or late at night. But with our 'wack-ass WAC,' it was disastrous. Remember that a key principle of the model is that "early success is essential to continued success." Mired in technical difficulties and facing a near-total absence of organizer guidance, our volunteers were dropping out of the campaign as quickly as we could recruit them.
The organizers followed them en masse. Within a few days of the WAC crash, anywhere from a quarter to a half of the staff in many offices quit; the entire Las Vegas office was summarily dismissed. Even after this wave of attrition subsided, organizers continued to drop out steadily, right up until the final weekend.
I actually think Lockse answered these comments best (in a comment that I urge you to go back and read in full):
The model is built so that people who have the will to do the work can succeed. If it wasn't working, and it never got fixed, then I can tell you it is not the fault of the people who signed on. Keep in mind that you're talking about young people who wanted more than anything to beat George Bush. Who were willing to do anything--if it would help beat Bush. Key word there is 'if,' people.
Remember that the PIRG/Fund model works on three essential components: recruitment, training, and canvassing. GCI chose to sacrifice training in order to meet its recruitment goals. As a result, attrition (the hidden fourth component of the model) skyrocketed, and the canvassing suffered.
Also note bschak's comment on the first post: even the handful of best-case offices that ran 'well'--with less than ten percent attrition--still ran into these fundamental problems and produced a compromised canvass. Off-blog, I asked Lockse--who has run many a canvass campaign--whether a fundraising canvass office would ever deprioritize training. The answer was: "Not if it wanted to raise money."
Lockse added that canvassing for votes is not all that different than canvassing for money -- the end products of a successful campaign are bonded teams that know what they're doing and why, and are able to do it in the most efficient way possible. That requires no small amount of leadership.
But GCI's senior management knows quite well how the model works -- they essentially built it. Why would they knowingly knock a leg out from under it?
I see two explanations -- I'll leave it to you to decide which is more disturbing. First, GCI's contract with MoveOn is likely to have stipulated the number of volunteers recruited (as opposed to the unknowable number of voters turned out). Second, the model itself--not the campaign--was GCI's true end. As Lockse wrote in the "Response to Strip-Mining the Grassroots," this model is like a giant game. So when Leave No Voter Behind was falling apart, with the model's future prospects at stake, and no dollar sign on the bottom line, GCI cheated at its own game in order to "win."
And yet this explains how, in the comments of the last post, Mr E could actually remark that "we beat our internal goals." Indeed, Leave No Voter Behind probably did manage to hit its original goal of volunteer recruits -- in much the same way that GCI raised twenty-two million dollars for the Democratic National Committee "to beat Bush." Neither campaign significantly helped us to win the 2004 election. Both turned away thousands of people who wanted to help and ended up feeling burned for it. Nevertheless, MoveOn, like the DNC, must have been pleased with the result.
Does this model seem like a success to you?