The ART of Facilitating Art:
How the DMV uses a Consensus Model to choose which Mutant Vehicles to invite to Burning Man
“Consensus is a process for decision-making that strives for higher ideals than other types of decision making processes.” - Tim Hartnett
How do you decide what art should be viewed by a particular audience? When there are many unique items to choose from and you can only choose a limited number, what’s the best method for choosing?
This is the challenge the Burning Man Department of Mutant Vehicles faces each year as we assess the hundreds of applications we receive. Over the years, we’ve evolved into a consensus model for our decisions that has been pretty successful.
Facilitating the decisions using a consensus model can be challenging, as it involves more time and effort than other methods, but using consensus can lead to better choices with more broad-based support.
Mutant Vehicles have become an indelible part of the visual fabric of Burning Man. Over the years, often inspired by the vehicles they’ve seen at previous burns, more and more people have created Mutant Vehicles and applied to bring them to the playa. We’ve reached a point where participants submit nearly 1,000 applications each year. The DMV is able to license around 625 vehicles on the playa, which means that to stay within our limit in recent years, up to one half of the vehicles that submitted applications to the DMV could not be invited.
Who Is The Department of Mutant Vehicles?
The DMV is the group within the Burning Man organization that oversees the licensing and operation of Mutant Vehicles and Disabled Persons Vehicles at the event. It is made up of 144 volunteers, 10 of whom are part of an oversight group called the DMV Council. Over the last decade the DMV Council has worked to try to find the best method for choosing which vehicles are invited to the event, with an eye to fairness and consideration of all aspects of each vehicle, and with the goal of inviting the vehicles that will add the most to the experience of the other participants at the event.
The rest of the DMV Volunteers begin their work in the spring, reviewing applications, and then work on-playa helping run the on-playa DMV, inspecting and licensing the Mutant and Disabled Persons Vehicles.
A LITTLE HISTORY
In the early days of Burning Man, people could drive around the event at will. A small group of people began bringing their Art Cars – cars that had been modified or decorated in some way.
As the event grew in size, driving had to be restricted for safety reasons. In 1997, after some vehicle-related injuries at the 1996 event, all non-essential driving at the event was banned, with the exception of the Art Cars. If you were driving an Art Car things were fine and if you were driving something else you were asked to stop and park it in camp.
But this “after-the-fact” oversight wasn’t really effective, so in 1999 the Department of Mutant Vehicles was formed. The Department asked anyone with an Art Car to come to an onsite kiosk where the DMV would take a look at their vehicle, determine if it was “arty enough” and then give them a “license,” basically a bumper sticker (later with a license number on it) confirming that this vehicle was allowed to drive at the event.
As the number of people wanting to bring Art Cars to the event grew, it eventually became larger than the Black Rock City could support. At that point the DMV refined the criteria of what would now be called “Mutant Vehicles” - art cars that were more significantly ‘mutated’. The DMV also began pre-registration: having people apply BEFORE going out to the event, so that DMV could review the vehicles and confirm they met the criteria before they were brought to the playa.
For the first few years, the DMV Review and Invitation process was fairly straightforward. Five or so volunteers would get on a conference call, review the applications, then decide whether the vehicle met the key criteria: “Mutated to the point it doesn’t represent a street vehicle.” While this criteria did not have a lot of nuance, it was intended to keep the DMV out of the position of having to “judge people’s art.” The review team fairly quickly decide if the vehicle met that criteria, or not, and decide whether or not to invite the vehicle.
As Burning Man continued to grow, so did the number of Mutant Vehicle applications. From 613 in 2004 to an all-time (so far) high of 1,032 in 2011.
During this time the Mutant Vehicle criteria continued to evolve, not in major ways, but each year there was more nuance and guidance added. As more people were applying, the average level of “mutation” of the vehicles increased and some vehicles that had previously been invited were now considered less worthy of an invitation as compared to the more involved, and more mutated vehicles. The Mutant Vehicle creators, on the whole, were raising the bar.
Finally, in 2015, the number of Mutant Vehicle applications that met the base criteria began to exceed the number of invites the DMV could bestow. In addition to this, we had some years when we reached the limit to the number of vehicles we could invite before we reached the end of the applications. And some of those later applications, the ones that were left out, were objectively “great” vehicles - ones that were really compelling, well-executed mutations.
Something had to change.
At first, the DMV considered “tightening up” the criteria - but the problem was, we couldn’t find a way to do so that didn’t have the unintended consequence of also excluding some vehicles that we felt SHOULD be invited. We also didn’t want to exclude smaller or simpler (or less expensive to produce) vehicles.
We also wanted to avoid a “First Come/First Considered” process, as that made it too random - it would favor the vehicles that got in line first and wouldn’t ensure that we were giving each application its due consideration, or picking the “best” vehicles. It could mean a “so-so” vehicle that barely met the criteria might end up getting a slot when a “great vehicle”, which just happened to apply a few minutes later, might not.
What we did instead was to change the overall process of review: Rather than the “First Come/First Considered” model where we make a final decision on EVERY vehicle the first time we look at it, we began reviewing the vehicles in separate “rounds,” In each round we would review ALL the applications in the order received, but not always make a final decision on each vehicle at that point. As we reviewed each application, if we felt the Mutant Vehicle described was a “no-brainer” - a well-mutated, well-executed vehicle, it would get an immediate invite. If we felt it met the minimum criteria, but was not in the top 50%, we would push it to the next round. This meant that we would be able to see ALL of the applications at least once, BEFORE we ran out of invites. Once we reached the end of the applications, we started again at the beginning, reviewing the remaining applications and again, considering them and inviting the top half (approximately) of what remained.
SKIMMING THE CREAM
This method of “skimming the cream off the top” involved more discussion during the review sessions. What was the cream? How do you determine the top XX% when you haven’t seen all the vehicles? Great questions and ones we tried to figure out as we went along.
Along with the other changes in 2015, we had a problem of too many DMV volunteers wanting to take part in review sessions. We sometimes had as many as 50 people reviewing applications on the conference calls. This became a bit unwieldy, and it became difficult to get to decisions on some of the vehicles, so the DMV capped the number of participants to 25, but the challenges in reaching decisions drove us to change the review process even further - to one that not only both considered input from the review participants, but also achieved more broadly accepted decisions than a simple voting process could achieve.
VOTING VS. CONSENSUS
The next key change was that we began focusing more on discussing the aspects of each vehicle, rather than simply voting. We soon found that when we had a structured discussion - one where the facilitator asked for opinions, then asked follow up questions - not only did it become easier to reach a broader consensus, but people were happier with both the outcome and the process itself. Part of the facilitators-follow up questioning involved specifically asking WHY someone felt the way they did about a vehicle. This helped others in the discussion better understand the person’s reasoning and they were then more accepting of an opinion if it differed from their own. Team members also noted that the process seemed more “fair” than it had in the past and that they felt their opinions were heard and considered more.
The gradual shift to the consensus model meant adjusting how much time we took with each application. Rather than making quick, 1-2 minute decisions, the average time spent on each application grew to 5-10 minutes. With particularly challenging vehicles, if consensus could not be reached after 15 minutes, we would table the review until another session (which would have a different makeup of volunteers), or push the vehicle to the next round.
THE QUAKER METHOD
In seeking out additional ideas and information on how we could improve our decision-making process, I came across some interesting information about our flavor of consensus-building: Even though we at the DMV had come to this model organically - by seeing what worked and what didn’t and changing things to align with our goals for the outcomes - the process we had developed closely mirrored a model of decision making known as the “Quaker Method” developed by Quaker communities.
In the “Quaker Method,” there is a neutral Facilitator called the Clerk who “…is theoretically a recording officer, but in practice he must frequently assume the duties of a presiding officer. He must be sensitive to all trends of opinion, including those not well expressed.” This is basically the role that the facilitator (usually me) has assumed in the DMV Processing meetings. While there are some differences between the “Quaker Method” and the “DMV Method,” the methods are strikingly similar, and our method has produced results similar to those reported by observers of the “Quaker Method.”
“It is surprising...how often real unity is reached, even though the discussion in its initial stages shows a wide variety of opinions, or a pronounced cleavage arising from strongly held convictions. As the consideration proceeds, unity gradually emerges and is finally reached. The decision may be along lines not even thought of at the beginning. This procedure takes more time and patience than the voting method, but the results are generally more satisfactory to all concerned.”
-Howard Brinton, Reaching Decisions: The Quaker Method.
In our newer process, we sometimes found that if we took a vote at the beginning of reviewing an application, after 5-10 minutes of structured discussion, when we took a subsequent vote, the tally was the inverse of the first vote - a 15-5 vote against a vehicle might become a 15-5 vote in favor of the same vehicle. The very act of discussion and consideration (and folks pointing out aspects of the vehicle that others might have missed) made a fundamental difference in people’s consideration of a vehicle.
Previously, we would have just stuck to the voting and if there was a majority for inviting, we would invite, and the discussion would be kept to a minimum.
As noted about the Quaker Method: “Key to this approach is the concept of dialogue, the creation of common meaning through an interactive process of listening, exploring assumptions and differences, and building a context for thinking together.” Malcolm Burson, Finding Clarity In The Midst Of Conflict: A Quaker Approach To Facilitating Dialogue.
This is what I see at play in the DMV process. Not always, and not perfectly, but in general. I think the key difference between simply voting and talking through to consensus is that when you just vote, you don’t really know why others voted the way they did, and why they didn’t see things your way. In the consensus model, everyone has the opportunity to voice their reasoning, and participants have the opportunity to understand WHY folks voted the way they did, not just THAT they voted that way. This seems to help folks accept the end result, even if they don’t agree with the majority.
Feedback from the DMV Volunteers (from our 2018 DMV Volunteer survey) seems to support that conclusion:
- 83% of the volunteers surveyed rated the DMV Application Review process “Very Good” to “Excellent.” The remainder rated it “Good.”
- 97% of the Volunteers felt the process was “Effective” to “Very Effective” in inviting the “best” vehicles to the event.
As for how participants view the outcome of our decisions, generally the feedback has been pretty positive. Of course there are always folks who don’t like one vehicle or another, or sometimes we have let something pass through that might not have shown up on playa as well-executed as their design plan, but overall, when it gets to be Wednesday night at Burning Man, and I look out across the playa at the illuminated kaleidoscope of amazingly mutated vehicles, I think we’ve done a pretty good job.
While the “DMV method” may be a little more involved, take a bit longer, and require more emotional investment than other methods, we tend to end up with better, more considered choices that better represent the community we’re choosing for:
We hope you think so too.
And if not, email us at email@example.com and let us know how we could do better.
For more on the DMV and the Mutant Vehicle criteria, see:
For more on general consensus building, see:
Consensus 101 Tim Hartnett
For more on the Quaker Method as it relates to consensus see:
Reaching Decisions: The Quaker Method
Finding Clarity In The Midst Of Conflict: A Quaker Approach To Facilitating Dialogue
Collective Intelligence and Quaker Practice