The trouble with the consensus model

I write this post with a horrifying blend of psychological literature and my own personal observations. It makes my scientist side cringe. What it therefore represents is the subjective experience of a person with too many psychology degrees.

The consensus model for decision making is used large swathes of the protest movement. It is adopted as it is thought to be democratic, inclusive and non-hierarchical. A consensus meeting is facilitated by a person whose job it is to make sure that everyone who wishes to speak, will speak. Somebody will raise a point, and others will address the point, called upon by the facilitator when they indicate they wish to speak. Through hand signals, people express agreement and disagreement. Agreement is represented by “jazz hands”; disagreement by downward-facing jazz hands. People may make direct responses to any point, and anybody can raise a proposal. I may have done a poor job of explaining consensus meetings for those who have never attended one. This page gives a good overview of the process.

Many people give enthusiastic jazz hands of agreement to consensus decision making due to its leaderless, non-hierarchical nature.

This does not mean the model is without its problems.

First of all, as acknowledged by Maeve McKeown and Lisa Ansell , the model is open to issues with power relations. Both identify an issue which I have noted in my experience of consensus meetings: that typically the same voices will dominate a consensus meeting, and that these dominant voices will often reflect unequal power relations that are inherent in our society. To put it more bluntly, it’s often the loud white guys doing the talking.

In a situation like this, many people do not feel able to speak up. Some feel unable to speak because they feel as though they know less than the dominant voices wheeling out minor, inaccessible theoretical points. Others do not speak as they are afraid of being shouted down with a “direct response” from a dominant voice. Others, still, are overlooked by the (often white, male, able-bodied) faciliator and never get called upon to speak. Some people cannot even attend the meetings: for example, when I visited the old Anti Cuts Space on Bedford Square, the building was not readily accessible to people with disabilities.

Another very important effect, one which I experience in every consensus meeting I attend,  is that of feeling unable to speak because I am an “outsider”. In situations where consensus decisions are made–occupations, direct action groups, and the like–there is often an “ingroup”: a core group of individuals. These cliques are often highly cohesive–they share an identity as members of the group. To those in the group, and those outside, powerful effects emerge.

Outside the ingroup, it is difficult to sway the opinion of the group–the consensus. This makes it harder for outsiders to speak and to be heard.

Inside the ingroup, a sometimes toxic effect emerges: groupthink. The word “groupthink” is loaded, melodramatic, reminiscent of an Orwellian dystopia, but this does not mean it does not happen. Through analysis of historical decision-making, and observations of group decision-making, a well-documented effect emerges: cohesive groups, particularly those under pressure, often make poor decisions. Crucially, this tends to happen when the group is attempting to reach a consensus.

The theory behind groupthink proposes eight “symptoms”:

  1. Illusions of invulnerability creating excessive optimism and encouraging risk taking.
  2. Rationalizing warnings that might challenge the group’s assumptions.
  3. Unquestioned belief in the morality of the group, causing members to ignore the consequences of their actions.
  4. Stereotyping those who are opposed to the group as weak, evil, biased, spiteful, impotent, or stupid.
  5. Direct pressure to conform placed on any member who questions the group, couched in terms of “disloyalty”.
  6. Self-censorship of ideas that deviate from the apparent group consensus.
  7. Illusions of unanimity among group members, silence is viewed as agreement.
  8. Mind guards — self-appointed members who shield the group from dissenting information.

While I have not observed all of the symptoms, I have certainly noticed a few. Silence–and/or the absence of any form of jazz hands–is generally viewed as a sign of agreement. With all of the talk that goes on, warnings are rationalised. As I outlined above, self censorship can–and does–occur.

So how is groputhink bad for decision-making? Decisions made by groupthink may not be opimal for the following reasons:

  1. Incomplete survey of alternatives
  2. Incomplete survey of objectives
  3. Failure to examine risks of preferred choice
  4. Failure to reevaluate previously rejected alternatives
  5. Poor information search
  6. Selection bias in collecting information
  7. Failure to work out contingency plans.

One more concept from social psychology and the study of group processes is a concern for the consensus model: the Abilene paradox.  This happens when a group decides upon a course of action which is to the preferences of none of the members of the group. Again, it is linked to self-censorship: people do not want to rock the boat and speak up against what they mistakenly believe to be the consensus of the group as a whole.

While, theoretically, the consensus model should stop these group processes from occurring and allow the voices of people from marginalised groups to be heard, in practice this does not happen. For some, like Maeve, the consensus model still represents the best we can get, and it is better than alternatives. For others, like Lisa, the issues relating to inclusivity are insurmountable hurdles to reaching a decision which is genuinely representative of all involved.

I myself think that these problems with the consensus model need to be addressed. I hope that the movement will lead to widespread social change. This will not be good enough if voices are still marginalised.

An example of this is what happened in Egypt to the women of Tahrir Square. The women were part of the revolution, or at least they believed that they were. And yet, they were marginalised and silenced, put back into their place after the government had been overthrown.

If we cannot check our privilege within our own meetings, if we cannot genuinely ensure that all voices are heard as we make plans–from anything to the cleaning of an occupied space to a full-scale revolution–how can we expect to create positive, lasting social change?

Just because we say that the consensus model is non-hierarchical and inclusive does not mean that it is.

So how do we fight the group processes that lead to poor decision-making? How do we genuinely make sure that each voice is heard?

There are alternatives, some more palatable than others.

One could be to anonymise everything. The Delphi Method is used frequently in medical decision making to reach a consensus.  I have used this method in my own research. It is an iterative process in which the decision-makers fill in questionnaires anonymously, indicating their agreement or disagreement with particular courses of action. A facilitator will then provide anonymous feedback and the decision makers fill in the questionnaires again, eventually converging on an acceptable course of action. A strength of this approach is that the dominant voices are now given equal weight to those afraid to speak out. Another strength is that those with disabilities who cannot access spaces in which meetings are held can still participate. There are two limitations, though. Firstly, it is time-consuming, although arguably traditional consensus decision meetings can be long and drawn-out. A second limitation is resources: rather a lot of paper or webspace and internet is required for the Delphi method.

Or, perhaps, one of the proposed “cures” to groputhink may be useful–talking to others outside the group, formally or informally. A drawback to this is security concerns, but in making decisions regarding something which is safe and legal, surely it is good to get as much divergent opinion as possible.

A final suggestion may be unpalatable to many: giving priority to those who have not yet spoken. It may seem undemocratic to bump a person up the queue, but for those who do not feel empowered to speak, it may make the difference between being heard and not being heard. Affirmative action may be required to ensure that we are truly listening to every voice.

It is important that we become aware with the flaws in the consensus model. It is not good enough that it is “the best we have”. We must address these flaws, and be mindful of them. It is not the panacea of perfect decision-making. It is rooted in group processes and societal forces which allow the loudest and most privileged to express their opinions.

To fight this, knowing our enemy is a good place to start.

I warmly invite comment on this post. There is an alternative; the traditional consensus model is not the best we can ever have. Together, we can find this and make better, more inclusive decisions as a result.

Special thanks to Lisa Ansell and Ellen for great conversations which helped me write this.

Edit: I followed up on some of these thoughts and some things which emerged from the comments here.

22 thoughts on “The trouble with the consensus model”

  1. I agreed with pretty much all of that and yet it did nothing to alter my enthusiasm for the consensus process.

    It seems that your problem is not that the consensus model has inherent flaws but that (no matter how much we may kid ourselves that it does) it fails to remedy the flaws of society. Perhaps the message here is not that ‘consensus decision making is riddled with traditional society privileges’ but that ‘EVEN consensus decision making is riddled with traditional societal decision making’. Still, it’s good to be mindful.

    Now; alternatives…

    My main problems with the Delphi method are that I don’t like speaking anonymously and I don’t like being compelled to speak. If I’m going to contribute to a discussion or a decision I want to choose when and how to do so and I want to be held accountable for what I’ve said. (I offer this personal view just in case it hadn’t occurred to anyone reading that these objections might exist, because I can see how promising this method might look).

    Speaking to outsiders is great, although for our purposes surely anyone willing to engage with us about the decision making process should be invited to join? Once you approach an outsider for their input don’t they become an insider?

    Now, I have a question, which is why I started this response. Practically, how does the third option of giving priority to those who haven’t spoken, manifest itself, and why is that so distasteful? I thought it was fairly standard practise in any group discussion from a classroom to a meeting to overthrow the government that eventually you politely tell the one who keeps talking to shut up and (equally politely) ask people who aren’t contributing what they think.

    Speaking as an uncommonly privileged person, I think the problem is not with the model but in how people using it behave, and improving that behaviour is a far larger struggle. Put another way, an optimal decisions making process is impossible in our arse-backward society.

    1. I absolutely agree with you here, Mediocre Man, it *is* largely a behavioural problem. I am concerned, though, that with such behaviour taking place before the revolution, what will happen after? How do we nip all of this in the bud?

  2. Like you, I think there are some issues with the consensus process as it’s manifest in current social movements, and clear-sightedly acknowledging them is important. The process should be — and was always intended to be — an evolving model, and shouldn’t be looked on as sacred and unalterable, but we certainly run into situations where it’s held to be so, often as a consequence of justifiable paranoia about attempts to inculcate a centralised, hierarchical model of decision-making in a movement resistant to it. It strikes me that many of the issues you bring up should be apparent to an experienced facilitator, and that suggests that we need to spread those skills around a bit more.

    On the issue of power-relations, it’s worth reading the piece that does the rounds on the Tyranny of Structurelessness (, which identifies a number of these issues very clearly, though I think its implicit solutions to the problem are regressive and shouldn’t be adopted.

    Much of the problem I think you’re touching on is that what looks like consensus often isn’t, and I’d ascribe this to a concentration on the *formal* aspects of decision-making over the actual content of the discussion and decision itself. Hand-signals and the associated apparatus of ‘consensus’ aren’t actually consensus in themselves and there are situations in which they can be alienating and offputting to people who aren’t familiar with them, and even abused by people used to using them (abuse of direct points, or the leapfrogging of conflict by formation of an inadequate proposal.) Certainly, we’ve got to think about the usefulness of those signals in a context where we’re trying to bring new people into a discussion as equal partners.

    In fact, far more important than ensuring everyone knows precisely what they should be doing with their hands is a clear examination of the *purposes* of the decision, as well as the inculcation of a certain rigour about one’s own role as participant in a meeting – learning, in other words, when it’s time to be quiet.

    The model of the large meeting or general assembly is intimidating to some and unjustly advantages particular voices — especially those who are gifted with eloquence, or the education to frame their ideas in a particular way, or who are socially privileged. There should be a number of tools available to a facilitator to try to combat that, but we’re often not brilliant at acknowledging these problems exist. Indeed, there should be no particular attachment to the model of the single, large group meeting, and opening other avenues for discussion, in which other voices can be heard, should really be imperative.

    Of the problems you list, I suspect that the illusion of unanimity is the most pervasive and hardest to combat. Encouraging portions of a meeting to brainstorm on devil’s advocate positions can help allow some unspoken reservations to emerge, but it’s certainly not a perfect method.

    A little rushed for time, but I’d like to come back to this later — great post!

    1. Great reply, and thanks for the reading; I’ll come to that when I can.

      I agree with you, that the process in and of itself is not the problem, but how it is applied. Good facilitation is a comment I’ve run into a few times as a solution, and I do think that it can go a way to addressing some of these criticisms!

  3. I’ve long said consensus has limits and has a risk of becoming a fetish or an example of horrible piety in activist circles and it has does have problems (even philosophical ones and moral ones abstractly). However, I disagree with your analysis in this case for three major reasons.

    First, if hierarchies are occurring, people aren’t making enough effort. I am not saying they will disappear, but that people have come up with attempts to reduce their impact that work well. A lot of people recommend having a “vibes watcher” in the room – their role, separate from the facilitator, is to see if people are speaking too long, or enforcing their privilege or the discussion is generally drifting. They cannot get involved but can stop the meeting if they think people are doing this. Too few people use these figures, but in my view they should always be used in groups of a decent size. Important too is the “stacker” their role – and only role – is to make sure people who have their hand up are added to the queue of people speaking. This separates this from facilitator having to do this and thus missing people because they are too engaged as well as spreading the responsibility in the group. If people use direct responses and even technical points to jump the queue, then the vibes watcher should have the power to step in, and, with the process of trying to get people to co-operate basically stop this. These aspects of the process have been in place since the 60s, and probably informally longer.

    Second, as you say above, good facilitation is absolutely key. The facilitator is neutral to the topics of the meeting, but they are not neutral to getting the task at hand done well. It is their duty to bring people out when they haven’t spoken, to perhaps take people who haven’t spoken yet (I do this every time I facilitate), to notice (with the vibes watcher) if people’s body language is displaying frustration to bring them in, to embrace newcomers and make sure cliques don’t form. Let me bore you with an anecdote. During the 2009 wave of pro-Palestinian occupations we used consensus. We had one meeting that was hugely disastrous- we spent 45 minutes or so discussing if or if not we should have a phone emergency reinforcement as this meant having a list of numbers of everyone interested in the occupation lying about which was a potential security risk. It was so bad the facilitator stormed out – not being neutral was a problem here too. However, later that weekendwe had one of the guys from Smash EDO visit us. He’d been doing facilitation for years – the meeting ran so smoothly it was almost stupidly empowering – the way he did it was ask people the right questions, frame the debate (in a neutral as possible way), get people to speak up and get people to clarify things (and find out if people understood). Exactly the same process was used. People think, like chairing, anyone can facilitate just by doing it – potentially anyone can learn to, but it is a skill that needs honing.

    Third, it is important to have a time for people to reflect on the process itself – this can be hard – shitty even. But the facilitator, particular if the meeting is one in a series, should in my view say “in this five minutes we are going to assess our last meeting” and invite comments from the floor. This can attempt, gradually to check the problems you describe.

    Creating a non-hierarchical society marked by egalitarianism, freedom from exploitation and alienation and characterised by cooperation, reason and compassion, based on human need is not going to be easy. The fact that consensus sometimes mirrors the systems of domination is not at all surprising therefore. Though while I think people need to question it constantly, I do think it is an important tool in the box.

    1. Thanks for the input; with feedback, I am becoming more aware that what I have generally seen is consensus implemented poorly. Perhaps better training for all might be useful?

      1. People try and skill share it all the time. In America, where it is very common in grass roots work (even in large unions) there are training courses. Indeed, I think skill sharing in general is a problem for the movement thus far – if we believe, as I do, of destroying something like ‘the division of labour’ as well as making our activities more robust by allowing anyone to pick up any task then we should be working at this.

        Another little anecdote: I spoke to people from Croatia who had been involved in the Free Faculty of Humanities, who produced the brilliant occupation cookbook. Like us in the UK, they occupied, communised, had open meetings, working groups and so on. It was seemingly insanely dedicated, serious and powerful – they encouraged dock workers to form their own committees and direct democracy was part of the national political debate in a way that had not occurred before. They were supremely generous individuals and took me out to lunch. I enquired how they made their collective decisions (no one represented the occupation, no one was a spokesperson or leader), asking if it was consensus – what is consensus they said? They had never heard of it – but it didn’t stop them running one of the least hierarchical affairs in town. As someone says above, its an ongoing process and self-criticism is always important.

  4. Awesome article and timely. To me the system is great but has a few flaws. some of the flaws are process related and others are to do with group dynamics.

    In terms of the process it’s more than possible for the views of the minority or even discussion to be subverted by repeated use of ‘direct points’, ‘technical points’ etc. Only a highly skilled and confident facilitator ,(which I’ll get to in a minute), can chair what is in effect vexatious use of the system. Secondly, at the heart of it, when people are making decisions about circumstances which may affect personal safety, or those which are politically divisive can we ever expect full 100% jazz?

    And now for the psychological stuff…

    Whichever method is used, which ever form of consensus people are aiming to reach, you really do need a very very very very good facilitator. The reason for this is not because people like me sometimes get confused with the variety of hand gestures. The reason is nothing to do with the process at all. The reason why you need a very good facilitator is that facilitating meetings is actually really really difficult.

    Even when you’re not trying to smash any of the following:’the state’, ‘monogamy’, ‘capitalism’, ‘misogyny’ or ‘Jeremy Clarksons FACE’ group dynamics require skill to negotiate.

    I have seen people fail to reach a decisions on totally meaningless banal bullshit. (through work). The reasons are always the same:

    Poor listening skills
    Self interest
    Low level personality disorder

    and conversely

    Low Self esteem
    Feelings of powerlessness
    Passive Aggression

    All of those issues have existed in every meeting I have so far attended within ‘the movement’. (I do not exclude myself from the above, as I always manage to get my self heard. I am a ‘loud white male’ who is reasonably narcissistic).

    If you add into the mix that people are knowingly or unknowingly pushing their own personal boundaries it turns up the heat. If you also add to that people involved in dissent are probably more likely to have experienced mental health issues, or if i’m wrong that at least 1 in 4 absolutely will, what you are looking at could be described in slightly exaggerated terms as ‘a powder keg’.

    The system needs to improve to give power to those that remain silent at Jazz or other moments. Paperwork, probably isn’t the answer. I also think we need to run a session probably, post 26th on facilitation.

    In my day job, I’ve seen alot of facilitation and end up doing it myself but in all my years the best facilitation I have ever seen has been two people within this movement.

    R from arts against the cuts is simply the best facilitator I have ever seen, in or out of work. Also, someone who i think is called Sam – is brilliant also. Apart from them I think we all have something to learn. What ever system is used the people that don’t speak up need to be heard and big mouths like me need to respectfully shut up. If we don’t grasp this then no matter how many times people say ‘non-hierarchical’ or ‘leaderless’ it is hierarchical and it is dominated by leaders

    Social leadership is hierarchy in sheep’s clothing.

    1. Psychology high five!

      Certainly agree training facilitators is going to be of paramount imprtance–that, and psychological support!

  5. I don’t really agree with this desire for neutral facilitators. If it’s a discussion of any importance then neutrality will be, at best, a well disciplined pretence. Far safer, I think, for the facilitator to make their own views known (to say nothing of the fact that a facilitator attempting neutrality is denied a voice in the discussion).

  6. Thanks for this post Stavvers.

    The comments on group dynamics, such as the problem of outsiders and “groupthink” are really interesting.

    A few of my friends came to the UCL Occupation and were completely baffled by the jazz hands and the style of our meetings in general. We did explain the hand gestures at the start of each meeting, but if you missed the start then you wouldn’t know what we were doing. We probably looked like a cult! Although many people did join in over the course of the UCL Occupation, so while it was alienating to some, it wasn’t for others. But worth bearing in mind nonetheless.

    In terms of the problem of groupthink, I have witnessed all of the points you raise; especially failure to examine risks and assuming unanimity. As a facilitator, I always try to check everyone is showing approval before accepting a decision (but not all facilitators do this and sometimes, frustratingly, not everyone shows their hands or will voice their opinions). As a member of meetings, however, I have at times been frustrated by the lack risk assessment, or failure to listen to alternatives. These were things we did in the early days, but as the group has become more cohesive, they have fallen away somewhat. I will be more aware of and resistant to this now, because it would be a shame if we fell prey to this issue.

    The first alternative you suggest, the Delphi model, wouldn’t really work in an occupation situation, where decisions need to be made quickly and frequently. We wouldn’t have the time to draft a survey, distribute it and collate the data.

    Affirmative action is something we did use at the occupation and have continued to use throughout our meetings. All of the facilitators that have had some training are aware of it. We ask if there is anyone in the room who hasn’t spoken yet, prioritize those who haven’t spoken and try to give priority to minority participants. Of course, in the heat of a meeting you sometimes forget to do this. And sometimes, a person who has never facilitated a meeting before will take a meeting because there are no trained facilitators around, and they aren’t aware of this issue.

    You say,

    “it is important that we become aware with the flaws in the consensus model. It is not good enough that it is “the best we have”. We must address these flaws, and be mindful of them. It is not the panacea of perfect decision-making. It is rooted in group processes and societal forces which allow the loudest and most privileged to express their opinions.”

    I think everyone involved in the UCL Occupation would agree with you. We’re aware of the flaws of the model, we discussed it in meetings and everyone was pro-affirmative action. I think, like you say, we’ve just got to be ever mindful of the pitfalls and this was a useful reminder. Consensus decision-making certainly isn’t a panacea, but I still think it’s better than any other formats I’ve come across!

    1. Thanks for this marvellous comment, Maeve. Do you think it might be a good idea to get training facilitators?

      Also, perhaps it might be a good idea to put the “symptoms of groupthink” somewhere prominent during meetings to facilitate mindfulness?

  7. Really interesting post, and well-observed. I think I’ve noticed all of the above at consensus meetings – I find it especially frustrating that women are often reluctant to speak, more reluctant to facilitate and more ready than others to take on note-taking roles and other traditionally ‘female’ tasks.
    It shouldn’t just be down to the facilitator to notice such problems – facilitating a large meeting is a difficult task, and it can be very difficult to get a sense of the overall balance of the meeting. Part of this problem is intrinsic to consensus itself – if we want a rotation of facilitators, which is key to a non-hierarchical structure, then we will inevitably get facilitators who lack experience, and therefore find the task of balancing the flow of the meeting more challenging. Skillsharing plays a major part in remedying this, but I think some of the skill of facilitating does come from experience. It seems that for consensus to work properly, the whole group needs to be conscious of its own dynamics and not leave it to the responsibility of the facilitator, which should be how it works anyway.
    I think many of the problems within consensus, like the ones with gender, are that even in very self-aware groups, inequalities in society replicate themselves in a supposedly neutral, non-hierarchical space – white, middle class men taking the ‘lead’ in discussion and so on. So the only way to remedy it is a kind of ‘positive discrimination’ – actively inviting those who haven’t yet spoken to contribute, paying attention to gender/ social dynamics.
    On a practical level, something that’s worked in a discussion group I attend is starting the meeting with everyone introducing themselves to the group – speaking once to the room tends to build up confidence, and demystifies the people you’re speaking to, hopefully making it less intimidating. I’ve also heard that ‘blind consensus’ – hand signals with eyes closed – can be an interesting way of getting honest responses, although I haven’t seen it used.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.