The Democracy Project

by David Graeber

  • Everyone who feels they have something relevant to say about a proposal ought to have their perspectives carefully considered.

  • Everyone who has strong concerns or objections should have those concerns or objections taken into account and, if possible, addressed in the final form of the proposal.

  • Anyone who feels a proposal violates a fundamental principle shared by the group should have the opportunity to veto (“block”) that proposal.

  • No one should be forced to go along with a decision to which they did not assent.

  Over the years, different groups or individuals have developed systems of formal consensus process to ensure these ends. These can take a number of different forms. But one doesn’t necessarily need a formal process. Sometimes it’s helpful. Sometimes it’s not. Smaller groups can often operate without any formal procedures at all. In fact, there is an endless variety of ways one might go about making decisions in the spirit of those four principles. Even the often debated question of whether or not the process of considering a proposal ends in a vote through some sort of formal show of hands, or other affirmation of consensus, is secondary: what’s crucial is the process that leads to decision. Ending with a vote tends to be problematic not because there is anything intrinsically wrong with showing hands, but because it makes it less likely that all perspectives will be fully taken into account. But if a process is created that ends in a vote yet also allows all perspectives to be satisfactorily addressed, there’s really nothing wrong with it.

  Let me give some practical examples of what I mean here.

  One common problem facing new groups is how one chooses a decision-making process to begin with. It can seem to be a bit of a chicken-egg conundrum. Does one need to take a vote to decide whether to operate by consensus, or to require a consensus that the group should operate by majority vote? What’s the default?

  To figure this out, it might be helpful to take a step back and think about the nature of the group itself. We’re used to thinking of groups as collections of people with some kind of formal membership. If you agree to join a group that already has a set of rules—a labor union, or for that matter an amateur softball league—you are, by the very act of joining, also agreeing to be bound by those rules. If it is a group that operates by majority vote, that means you are agreeing to be bound by majority decisions. If it is a vertical group with a leadership structure, it means you’re agreeing to do what the leaders say. You still have recourse: if you object to a decision, you can quit, or refuse to comply, which might cause the group to reconsider the decision but is more likely to mean you’ll be penalized in some way, or expelled. But the point is there is some kind of sanction. The group can coerce behavior through the threat of punishment.

  But if you’re talking about an activist meeting or public assembly, as opposed to a group composed of formal members, none of this is true. No one at a public meeting has agreed to anything. They are just a bunch of people sitting in a room (or standing in a public square). They are not bound by a majority decision unless they all agree to be. And even if they do agree, if a participant later finds a decision so objectionable he or she changes his or her mind, there’s not much the group can do about it. No one is really in a position to force anyone to do anything. And if it’s a horizontal or anarchist-inspired group, no one wishes to be in such a position.

  So how does such a group decide if they want to operate by majority vote or some other form of consensus? Well, first, everyone would have to agree to it. If there is no such agreement then it is fair to say that “everyone should have equal say and no one can be forced to do anything they strongly object to.” That becomes the general principle of any decision making.

  This doesn’t mean that one should never call for a majority show of hands at all. Most obviously, it’s often the best way to find out critical information, such as, “If we held an event at 1 P.M. on Monday, how many of you would be able to come?” Similarly, if there is a technical matter where it seems clear that no question of principle is likely to arise (“Should we table this discussion for now,” or, “Shall we meet on Tuesday or on Wednesday?”) a facilitator might simply ask if everyone is willing to be bound by a majority decision on that question and be done with it. More often, though, a facilitator will just ask for a show of hands as a “nonbinding straw poll” or “temperature check,” that is, just to get a sense of how people are feeling in the room. This can just be by a simple show of hands or a more subtle system where everyone either waves hands up in the air for approval, down for disapproval, and horizontal for uncertainty. While nonbinding, such tests can often give all the information you need to know: if sentiment is running strongly against a proposal, the person who submitted it might then withdraw it.

  When one is dealing with nontrivial questions, though, the four principles become more important. So how do you find consensus on more complex issues? There is a fairly standardized four-step procedure that has been developed over the years to ensure that proposals can be continually refashioned in a spirit of compromise and creativity until they reach a form most likely to be amenable to everyone. There’s really no need to be religious about it: there are lots of possible variations. And it’s important to remember that while those coming to a meeting might be presumed to have agreed to the basic principles, they have not agreed to any particular formal rules of procedure, so the procedures should adapt to the desires of the group. But generally speaking it goes something like this:

  1. someone makes a proposal for a certain course of action

  2. the facilitator asks for clarifying questions to make sure everyone understands precisely what is being proposed

  3. the facilitator asks for concerns

  a. during the discussion those with concerns may suggest friendly amendments to the proposal to address the concern, which the person originally bringing the proposal may or may not adopt

  b. there may or may not be a temperature check about the proposal, an amendment, or the seriousness of a concern

  c. in the course of this the proposal might be scotched, reformulated, combined with other proposals, broken into pieces, or tabled for later discussion.

  4. the facilitator checks for consensus by:

  a. asking if there are any stand-asides. By standing aside one is saying “I don’t like this idea, and wouldn’t take part in the action, but I’m not willing to stop others from doing so.” It is always important to allow all those who stand aside to have a chance to explain why they are doing so.

  b. asking if there are any blocks. A block is not a “no” vote. It is much more like a veto. Perhaps the best way to think of it is that it allows anyone in the group to temporarily don the robes of a Supreme Court justice and strike down a piece of legislation they consider unconstitutional; or, in this case, in violation of the fundamental principles of unity or purpose of being of the group.*

  There are various ways of dealing with a block. The easiest is simply to drop the proposal. The facilitator might encourage the blocker to meet up with those who brought the proposal, to join the relevant working group for instance, and see if they can come up with some kind of reasonable compromise. Sometimes, especially if others feel the block isn’t justified (e.g., “I don’t think it’s anti-Semitic to have the next meeting on Friday even though it’s a Jewish holiday. Most of us are Jewish and we don’t care!”†), there might be some process for challenging a block: for instance, asking if at least two other members of the group are willing to sustain it. (We sometimes speak of “consensus minus one” or “consensus minus two” to describe such a situation.) Or, if it is a large group, it is usually a good idea to have some fallback: if there is a strong feeling that most people want to go ahead, regardless of a block, one can turn to a supermajority vote. During our first August 2 meeting for Occupy Wall Street, for instance, we decided on a version of “modified consensus” where we could in the event of a logjam fall back on a two-thirds majority, but later, a few da
ys into the actual occupation, the General Assembly agreed on moving to a 90 percent fallback on the grounds that, with the movement growing so rapidly, the earlier system would allow proposals to pass that were opposed by hundreds or even thousands of participants. It is important though not to fall back on this automatically: if someone blocks, the most likely reason is a failure of process, that is, a legitimate concern was raised and not addressed. In that case, the group might do well to go back and reconsider the proposal. But, especially in a very large group, one will have to fall back on such expedients now and then.

  There are a few areas of consensus process that often cause problems or confusion that I will try to clarify a bit here.

  One is that one cannot very well base a block on a group’s principles of unity unless that group actually has principles of unity. Thus it’s always a good idea to come to some sort of agreement about why the group exists and what it is trying to accomplish as quickly as possible. It is best to keep these principles simple. It is also crucial, in framing them, to remember that any activist group exists to do something, to change the world in some way. So the principles should reflect both what the group is trying to accomplish and the manner in which it goes about trying to accomplish it—and the two (the ends and means) should be in as much harmony with each other as they can possibly be. But the smartest thing to do when it comes to defining the group is to keep it simple. It is much easier to write, “We oppose all forms of social hierarchy and oppression,” for instance, than to try to list every form of social hierarchy and oppression you think exists.

  One good thing about having principles of unity is not just that it clarifies blocking, but that it makes it possible for well-meaning participants to periodically remind everyone why they’re all there. This can be almost unimaginably helpful in resolving conflict, because, in moments of passionate conflict, people have a remarkable ability to forget why they got together in the first place. Which leads to another point: there’s nothing wrong with conflict, provided people do remember why they’re all there. This is another misconception about consensus. “But conflict is the essence of politics,” one often hears. “How can you try to eliminate it?” Obviously you can’t. Nor should you try to. Some of the confusion comes from the fact that in America (unlike many other places) activists were first introduced to consensus through the tradition of Quakerism, which has meant that for most activists, their first experience of consensus is rooted in gentle and, frankly, bourgeois sensibilities. Everyone is expected to be, at least superficially, extremely nice. After the macho histrionics of so much late 1960s radicalism, where jumping on chairs and pounding one’s fist was considered a normal way to behave, Quaker- and feminist-inspired consensus was a useful corrective. But before long, a desperately needed feminist emphasis on mutual listening, respect, and nonviolent communication began shading into a distinctly upper-middle-class cocktail-party-style emphasis on politeness and euphemism, on avoiding any open display of uncomfortable emotions at all—which is in its own way just as oppressive as the old macho style, especially for those who were not themselves of upper-middle-class origins.

  While the bourgeois style has hardly been put to rest, in recent years there’s been a shift away from it. The best facilitation trainers, for example, have realized it’s much better to say in effect: yes, we are passionate people, we are here because we care deeply and have strong emotions; displays of anger and frustration are just as important (and legitimate) as those of humor and love. Rather than trying to suppress all these things, we should instead understand that for a group to accomplish its goals, conflict between friends and allies ought to be encouraged, provided everyone remembers that this is, ultimately, a lovers’ quarrel. What that means in practice is that while it is perfectly legitimate to doubt the wisdom of another’s words or deeds during a meeting, or even to express outrage at their words and deeds, one must always give them the benefit of the doubt for honesty and good intentions. This can often be extremely difficult to do. Often one might have every reason to suspect that one’s interlocutors are not behaving honestly and do not have good intentions. One might even suspect they’re an undercover cop. But one could be wrong. And just as the surest way to guarantee people will act like children is to treat them like children, the surest way to guarantee people will start behaving irresponsibly during a meeting is to treat them as if they already are. Therefore, challenging though it is, everyone must be on guard for such behavior, and immediately call it out. It’s fine to tell someone they’re being an idiot, if you genuinely think they are. It’s not okay to say they’re intentionally trying to wreck the movement.

  If it turns out they are intentionally trying to wreck the movement, there are ways to deal with that. If someone does turn out to be a cop, or a Nazi, or is actively trying to block the group from achieving its purposes, or is just a stark raving lunatic, there has to be some way to get rid of them—though usually this has to happen outside the meeting. One problem we had in New York is that, even when people declared their purpose was to disrupt a meeting, they were often nonetheless allowed to take part. We eventually found the best way to deal with such people was by the equivalent of shunning: whatever they say, whatever they do, simply do not react. The approach was first developed, quite spontaneously, when using the People’s Microphone: if someone began saying something others found offensive, everyone would simply stop repeating it, and eventually, if the speaker continued in the same offensive vein, they found no one could hear anything they had to say.

  There are always boundaries, acknowledged or otherwise. If unacknowledged they become visible the moment someone breaks them. Just as “diversity of tactics” is based on the tacit assumption that no one would ever show up to a demo with a car bomb or rocket-propelled grenade, so assertions that no activist should be expelled from a meeting do assume certain parameters. I recently attended a Spokescouncil in New York where everyone had been engaged in a long debate over whether there should be a “community agreement” and a shared principle that if anyone violates that agreement, they should be asked to voluntarily leave. The proposal was meeting concerted opposition when, suddenly, someone noticed one of the delegates was holding a plaque saying “Aryan Identity Working Group.” He was immediately surrounded by people—many of those who had just been loudly insisting such a rule was oppressive—who successfully forced him to leave.

  This is only one of the various tools that have been developed over the years among activist groups to make consensus process work. There are many others (icebreakers, go-rounds, popcorn, fishbowls …) and detailed resources already exist on how to use them (they’re easily accessible on the web through a simple Google search). My own personal favorite guide on facilitation and process is by activist and author Starhawk, but there are many according to taste. There are also different models of organization (General Assemblies and Spokescouncils, for instance), each with its own merits. There is no single right way, exactly, or road map for how these models can be scaled up to organize all of society on a directly democratic basis. The beauty of consensus process is that it is so various and adaptable. So here are some practical considerations and common misunderstandings about the basic principles of consensus, which, hopefully, will make it easier for interested readers to participate in a process of figuring such things out for themselves:

  A QUICK CONSENSUS FAQ

  Q: But doesn’t all this “consensus process” just come down to manipulation by a tacit or hidden leadership clique?

  A: If you operate by consensus without any rules at all, then, yes, inevitably a tacit leadership will emerge—at least, as soon as your group grows larger than eight or nine people. The writer and activist Jo Freeman pointed this out back in the 1970s during the early years of the feminist movement. What we now call “consensus process” was created largely to address this problem in the wake of Freeman’s critique.

  The role of the facilitator is a perfect example here. The easiest way to know
you’re dealing with bad process is that the same person is (a) running the meeting, and (b) making all the proposals. In any horizontal group there will be a clear understanding that the facilitator doesn’t herself bring forward any proposals. He or she is just there to listen and become the medium through which the group can think. Usually, in fact, even the role of facilitator is broken up and divided among several people: one person to actually keep the meeting running, another to keep stack (count of those who’ve asked to speak), another to keep time, another as vibes watcher to ensure energy isn’t flagging and no one is feeling left out. This makes it even harder for a facilitator to manipulate debate, even unconsciously. Facilitators rotate, which allows the group to constantly maintain gender balance among facilitators, as well as in stacks.

  This doesn’t mean there won’t be cliques, especially in very large groups, or that some people won’t end up with much more influence than others. The only real solution is for the group to maintain constant vigilance against the rise of cliques.

  Q: But if you’re saying such influential cliques will tend to emerge, wouldn’t simply recognizing the fact that there really are leaders, and therefore creating a formal leadership structure, at least be better than having a secret unaccountable leadership no one acknowledges?

  A: Actually, no. People who do more work will, of course, have more influence. This does give a certain advantage to those who have more time on their hands. Inevitably, some will start coordinating together and this will mean some people have privileged access to information. This is the real problem. In any egalitarian group, information tends to become the limited resource: if hierarchies develop, it’ll be because some people have ways of finding out what’s happening that others do not. Formalizing this by declaring those with privileged access to information a “leadership” is not going to ameliorate the problem, it will only make it worse. The only way to ensure that this group doesn’t actually start imposing their will on others, even without intending to, is to create mechanisms that ensure that information is as widely available as possible, and constantly reminding the most active members that there is no formal leadership structure and no one has the right to impose their will.