Beyond ‘Consensus’ Decision-Making

Consensus decision-making claims to solve a major problem of majoritarian decision-making – that a minority can have its views or needs over-ruled. It carries the promise of a system where everyone can take part in decisions, and where outcomes are reached that have everyone’s full approval. Consensus decision-making is useful and innovative, particularly for the emphasis it places on deliberation and discussion.

But consensus decision-making process often fails to live up to its aim of producing decisions that all participants fully consent to, and needs improvement. It doesn’t allow us to be sure that we have made decisions which everyone is ok with. Whist it is possible that done well, by experts, it mighti, as generally practiced, it falls short. This is a problem, because a democratic decision-making process needs to be simple enough that everyone understands how it works – and that everyone can see that decisions have been made fairly – otherwise it won’t be democratic. If consensus decision-making only works between experts, it’s not fit for purpose, and needs rethinking.

I will discuss here, firstly, what ‘consensus decision-making’ is trying to achieve and why it fails, before discussing several major problems with how it is currently used, and suggesting some possible improvements to make it a more democratic and effective process. I will then discuss more generally the importance of disagreement and the need for a distinction to be made between the arguments against representative democracy and the arguments against voting.

What is consensus?

What does consensus mean? Whilst many would say it means that everyone consents to a decision, it is often, and is easily, confused with unanimity. Unanimity would mean everyone agrees with the decision reached, or even that everyone shares the same opinion, which is crucially different.

What’s the problem with voting?

Many people dislike majority-voting because it can involve a majority imposing their views on a minority. People are born free and should not have decisions made about them without their consent to these decisionsii. To get away from the perceived domination of majoritarian decision-making, consensus decision-making process claims to avoid voting, replacing it with a ‘check for consensus’; instead of a yes/no vote, participants are invited to agree, to ‘stand-aside’iii, or to block, vetoing the decision. Because there is no ‘yes / no’ vote, and the focus is on discussing problems to find a solution, with the option of a veto where participants have unaddressed concerns, it is argued that ‘consensus’ solves the problem of majoritarian decision-making; the coercion implied in a minority having their views over-ridden. But in fact by doing this consensus only disguises the problem.

To understand things better, we can make a distinction between voting and the ‘decision-rule’. Voting is simply a means of assessing everyone’s preferences – for example, by a show of hands. This is separate from the decision-rule, which is how we interpret these expressed preferences to determine what outcome has been chosen by the decision-making process: i.e., it is the criteria which a decision has to meet to be considered passediv. So, for example, a vote is often to show support either for or against, or for one of a number of different options: in a referendum, the vote is each person choosing which box to tick on the ballot paper. The decision-rule is how we interpret this vote: in a yes/no vote, for the motion to pass it might require a simple majority (50% of votes in favour), or a two-thirds majority: these are different decision-rules. Decision-rules need not be majoritarianv.

On this definition, it’s clear that the formal consensus decision-making process does involve voting: participants can vote in favour, to stand aside – or to blockvi. And the decision-rule is that for the decision to pass no-one must show disagreement: that no-one uses a block and there are not significant numbers of stand-asidesvii. So participants can disagree but it can be hard to do so, requiring a lot of confidence or bravery, and to express continued opposition by using a block you must place yourself in the position of preventing what everyone else is apparently agreed upon. It is even harder to take this step if you have not already shown disagreement in the discussion stage, if you perhaps felt unable to articulate your opposition. An absence of disagreement or of a block in consensus decision-making does not necessarily mean genuine support: it might simply mean that participants do not wish to hold up decision-making by expressing their disagreement.

Fetishising ‘not voting’, without a proper understanding of what we are opposed to, is harmful. Consensus decision-making does allow voting: it merely mystifies it, using different language, and gives no clear mechanism for no votes. Consensus decision-making can place pressure on people to agree and go along with what is perceived to be the leading option – and with no easy way to vote against, discussions can easily be swayed by whoever talks the loudest or the first, by dominant activists, or even by the way a facilitator pushes or frames the discussion. Without participants having had the opportunity to vote no, there is no way of knowing that the decision is genuinely the one supported by the most participantsviii.

One possible change which might solve this problem would be for ‘checks for consensus’ – the voting that follows discussion – to allow participants to vote in one of four ways: support, oppose, stand aside, or block. This actually gives participants five options, counting the possibility of abstaining. In this, opposing means ‘this isn’t what I think we should do, but I’m happy to go along with it if the group is in favour. Stand aside means ‘I don’t think the group should do this and if the group chooses to do despite me, I will not take any part in implementing it’. Block remains, as in consensus decision-making, a vetoix. This change would not result in participants having their views ignored; instead, it allows people to honestly register disagreement, whilst still being happy to go along with a decision. To avoid the assumption of agreement, instead of calling this a ‘check for consensus’, participants should be asked to show their preferences; or we should straightforwardly call it a vote. It is important to be honest, and refusing to call consensus checks a form of voting creates confusion and misunderstandings.

Making it easier for people to express their genuine opinions, including via clear voting methods, is necessary for us to make sure we reach decisions that have the greatest support and to which everyone consents.

Deciding between different options

Consensus decision-making often doesn’t have a clear process for considering and then deciding between multiple optionsx. This is partly because of its focus on discussion and collaboration, which allow different ideas to be merged, and which are important in their own right. Unfortunately however, the result of this can be a confusing, jumbled decision-making process; discussion flits in a disorganised way between different options under consideration, sometimes making the process frustrating. Moreover, it is often up to either the facilitator or whoever talks most which option gets taken forwards to a check for consensus. Majoritarian methods can suffer from similar problems.

Where different proposals can be constructively combined or improved, this is a good option. However, this will not always be the case and decision-making processes need a method for choosing between multiple proposals. One way of doing this would be for the facilitator to ask for proposals for ways forwards and for these to each be discussed in turn, allowing each one to be developedxi and amendments, before space is made for participants to express which should be taken forwards further to a final discussion and preference check. This could be quite time consuming, but more productive than trying to talk about several different topics at once. As for deciding which of the options has most support, there are several possible processes:

  • The simplest way of deciding which proposal is most favoured would be with a simple show-of-hands or temperature checkxii. This is, however, imprecise.

  • For more important decisions or where there is time and the ability to write out ballet papers, using a Condorcet method, Modified Borda Count, or Single Transferable Vote would ensure an outcome likely to work well for everyone. These methods involve voters numbering their preferences for different options in orderxiii.

  • Amendments to particular options could be made before or after a decision on which is the most favoured alternative, and accepted or rejected by the group. Experimentation by different groups will help determine the clearest process to follow.

Pluralism is a good thing and allowing a number of different options to be more fully developed and then decided between ensures a better quality of decision, with full awareness of all possibilities. Having a clear process to follow for deciding between multiple incompatible options would take power out of the hands of the facilitator, ensuring groups as a whole decide which possibilities are taken forwards.

Blocking change, or: time to veto the status quo?

One criticism often made of consensus decision-making is that the ability of people to block decisions can make change difficult. This can make it conservative and potentially prevent inequalities from being tackled, as changes can be blocked. However, once the decision-making process is adjusted to deal with decisions between multiple alternatives, there is a way out of this: the status quo should also have to argue its case. Certainly where historical inequalities remain, or where current practices are the result of previous un-democratic decision-making processes, it should not be assumed that this can continue if other options are blocked. To privilege the way things are, just because that is the way they have been in the past, contains its own form of violencexiv.

Disagreement makes us stronger

It’s important to be able to disagree, and to move forwards, constructively, despite these disagreements. Do we want to only live and work with people who share all of our views? I think it is important to challenge and to be challenged. This may take practice, and may at times be difficult, but it’s vital, vital to avoid becoming stuck in the bubbles and echo chambers of people who share the same views as oursxv. If we want to live in a democratic society, we need to be able to make decisions collectively, not just in autonomous, internally homogeneous groups. And when making decisions, having competing options and alternatives open makes our decision-making stronger – and is at the root of democracyxvi. A pressure for our groups always to agree can be stifling, hampering productive critique and debate. And the pressure consensus decision-making creates for everyone to agree encourages fragmentation: the creation of many different groups each with their own nichexvii, rather than the difficult building of alliances, of solidarity and collective strength, that we desperately need.

Voting and ‘representative democracy’

The arguments against voting can be mixed up with the arguments against representative democracy: Government and Parliament are associated with elections and voting, and voting comes to be seen as guilty by association. Confusing these arguments risks misunderstanding and mis-diagnosing the problems in our society.

Lack of democracy in our society comes from a number of places, including:

  • Many areas of our society – most people’s work places, much of the economy – are excluded from democratic decision-making: decision-making is seen as a private matter, with decisions-making powers bestowed by ownership.

  • State decision-making bodies are not genuinely democratic: they are ‘representative democracies’ rather than participatory democracies. The idea is that we choose them and they make decisions on our behalf…but even in theory, it has been shown that representative democracy is contradictory and doesn’t really work as democracyxviii. In practice, our MPs and ‘representatives’ are also highly influenced by money and other factors.

  • In Britain, particularly, the first-past-the-post electoral system distorts who gets elected. This means parties with a minority of the votes can get a majority of the MPs, and all the power in government. This is related to the problem of majoritarian decision-making. But the solution here is not for parliaments to use consensus decision-making: what would this result in? Whilst it would limit governments’ power to do harm, it would also harm the ability of governments’ with broad public support to take progressive steps. Would slavery have been abolished if the US congress had to agree it by consensus? Instead, the way states and democracy work, and how and where decisions are taken, needs to be fundamentally changedxix.

  • Decision-making is often highly centralised, even within supposedly democratic institutions, such as the state.

  • Some topics are not seen as matters for democratic, collective decision-making. Private property for example is seen as an absolute, rather than a creation of social institutions.

  • Media organisations, from whom we get much of our information and many of the ideas which we use to understand the world, are not democratic and their control is often highly centralised.

Hierarchies in our society are not primarily due to the use of majoritarian decision-making methods rather than consensus.

Spreading democratic decision-making

‘Consensus’ hand signals are a great innovation, and could be used fruitfully in other decision-making processes; as could be many of the other techniques sometimes used as part of consensus decision-making: using go rounds to give everyone a chance to talk, splitting into smaller groups, or using ‘fish bowls’xx. And consensus decision-making is great in many ways, particularly for emphasising the importance of talking in democracy, not just the voting; and when done well, for attempting to bring more structure to collective decision-making. However it has its own shortcomings and falls short of its own claims to produce decisions that work for everyone. It is crucial for us to find ways of making democratic decisions more effectively; otherwise, insisting that consensus is the only democratic decision-making process but struggling to make it work, people will resort to using sleight of hand to avoid having to make decisions as a group. Tricky decisions are hived off into ‘working groups’, or organisers simply avoid democratic decision-making completely, for fear of getting bogged down in endless debatesxxi.

It seems that many people who argue for consensus decision-making hate voting so much that they would rather groups resort to more hierarchical methods than consider using some form of voting. For example, the American activist Starhawk argues:

Consensus process can help a group find the best possible solution to a problem, but it is not an effective way to make an either-or choice between evils, for members will never be able to agree which is worse…flip a coin…In emergencies, in situations where urgent and immediate action is necessary, appointing a temporary leader may be the wisest course of action.xxii

Is a group taking some form of vote such a travesty that letting one person make the decision, or letting an inanimate object make the choice, is preferable? Or is refusing to consider using voting methods a defence mechanism that allows us not to have to think about different ways we could organise? Majoritarian voting is not the main enemy: forcing more decisions to be opened up to collective and participatory decision-making should be more of a priority than insisting that we never vote. Consensus decision-making should stop relying on a false distinction between voting and not voting.

Good collective decision-making involves synthesising many different needs and perspectives to find a way forwards that can gain the most support. It involves balancing the urgency and importance of the decision, and the various obstacles different group members face to taking part in the decision. Some may be less articulate or confident making their voice heard, perhaps because they are part of a historically-disadvantaged group, or because they are unfamiliar with the group, or the decision-making process. Decisions also vary according to the amount of time worth spending on them, and the number of options available. A group might need to decide which of ten varieties of porridge oats to buy, but members don’t think the decision worth more than a few seconds to decide on. Group size, location and resources, experience also vary: good and effective decision-making needs a balance all of these competing priorities, according to the groups wishes. Ideally this process of prioritising decisions, too, should be democratised and be up to the group as a whole, rather than just left to the facilitatorxxiii.

Similarly there is no point being perfectionist about our group decision-making methods if the cost of this is minimising the number of decisions made by the group as a whole.

A good democratic decision-making process must be understandable by all who take part in it, and must provide an effective means for people with a variety of different opinions to make mutually acceptable decisions. Part of being mutually acceptable is that it should be clear that the process itself meets certain minimum criteria. It must also strike an appropriate balance, aiming for thorough and participatory decision-making, but prioritising, even taking short cuts, to make sure that excessive time is not spent on minor issues. Whilst I have pointed to several ways the process might be improved, the proof of the pudding will be in the eating.

These suggestions are alterations to the existing consensus decision-making process, but if a new name is needed to avoid misunderstandings, it could be called ‘pluralistic direct democracy’.

Pluralistic Direct Democracy: Summary of process

This flowchart summarises how an amended decision-making process could look. This can be summarised as: Prioritise Agenda Items, Introduce Issues, Propose Solutions, Discuss & Amend, Focus, Vote.

This process is still complicated but avoids some of the mystifications of ‘consensus decision-making’, and hopefully it can be helpful for campaigners and others. This is only one of many possible democratic decision-making processes, and finding which work best in different circumstances will require experimentation.

Perfecting a decision-making procedure is not on its own sufficient for democracy. We need to develop a culture of consent, and of supporting people to take part in decisions that affect them; where people don’t dominate conversations, and make space for everyone to speak. We need more discussions of when things become public questions, of where private matters end and needed to be opened upxxiv. And we need to find ways of making bigger democracies, collaborating between groups and knitting them together, not just settling for internal democracy. And even in our movements, we need more democracy: too many events, too many organisations, and too many decisions, are not made through a democratic process. Hopefully by developing a satisfying, effective decision-making process we can demonstrate that participatory democracy is possible and necessary.

See also:

Jo Freeman’s article, the Tyranny of Structurelessness, can be read here.

Some useful introductions to consensus decision-making can be found on the Seeds for Change website. This is a longer explanation of how to use it.

A short discussion of an alternative to consensus can be read here: “Near-consensus alternatives: Consensus Oriented Decision-Making”.

Post Header Image: 'Straw Poll on Direct Action' by Collin Anderson, CC:By-NC-SA 2.0

References & Footnotes:

i Christopher Haug distinguishes between four different types of consensus decision-making in practice: see

ii Paul Wolff famously outlined the philosophical case for anarchism and against authority. See: Wolff, R.P. (1970) In Defense of Anarchism: With a Reply to Jeffrey H. Reiman’s In Defense of Political Philosophy. New York etc., Harper and Row.

iii The meaning of this has often been unclear; in groups I have been part of some have taken it as standing out of the outcome – ‘it can go ahead but I won’t be part of it’; others have seen it as a symbolic kind of no vote, misnamed and misunderstood.

v According to Wikipedia, the Borda Count ‘is often described as a consensus-based voting system’.

vi Even if in practice this check is often skipped and process is unclear. Haug (2012)

vii Haug, Christopher (2012).

viii We shouldn’t worry about no voting being ‘negative’. Honesty is more important than false positivity, and having no as an option means that when agreements do go through we can be sure that there is genuine positive support.

ix As in ‘consensus’, the extreme step of a block could be discouraged by requiring the wielder to explain their reasons: with great power comes great responsibility. This could perhaps be waived for historically-disadvantaged groups, those who may find it harder to participate in decision-making. Following a block groups should generally move on to discuss other options; is their an alternative course of action that satisfies everyone? Creativity is required to find an outcome that everyone can live with!

Groups can also split or choose to implement the decision as a subsection of the group, or where a decision must be made and participants cannot find a course of action that will not be blocked, it may be necessary to over-ride this – incredibly rarely. Formalised processes have been developed for dealing with these instances, for example, by the Quakers. See (p.3) for some ideas of ways to deal with blocks.

Stand-asides should also be taken seriously, and when this happens a group should decide whether to go ahead without the dissenter or whether to look for another option.

x Crowd-wise is an example of a method that has a clear process for deciding between different options,based on the Modified Borda Count:

xi Facilitators can use the amount of time allocated to making the decision to inform how long to give each proposal. Discussion could proceed in two ways; either the facilitator could ask first for proposals, and once these have all been drawn out, move on to discussion of each; or, as each proposal comes up, focus on that idea before asking for other proposals (which may be prompted by the discussion).

xii There could also be a show of hands and then a run-off second vote between the most favoured options. A further possibility would be, if full discussion had been had for each of the alternatives, to do a full ‘consensus check’- checking for agreement, disagreement, stand asides and blocks. The option with the greatest majority and without blocks or stand-asides would be chosen.

xiii Condorcet methods compare each option against every other option, examining these pairs to determine which of each is most preferred: Borda Counts allocate each option a number of points depending on how high up each voters preferences they are: These methods can produce different outcomes. For important decisions these methods are more likely to select an option on which everyone can agree than a straightforward show-of-hands will.

xiv Several thought experiments can be done to help us think through how decision-making procedures deal with different situations. One of these is found in a hypothetical post-revolutionary village. Residents have got together and the topic of today’s meeting is land reform. Most residents are arguing for redistribution, but one greedy land-owner, the village’s former lord, who has more land than the rest of the village put together, is ready to veto reform. Should this be allowed? Or should other residents be allowed to veto such an unequal state of affairs? Tradition can be oppressive and consensus decision-making risks allowing them to continue. Whilst this may be an extreme example, inequalities and privileges do exist within our social movements, just as they clearly do in wider society.

xv This is a phenomenon that new online technologies and phenomena are exacerbating: see e.g. Sunstein, C.R. (2007) Princeton, New Jersey ; Oxford, Princeton University Press. See also: and

xvi Mouffe, C. (2007) Articulated power relations: Markus Miessen in conversation with Chantal Mouffe. Available from: [Accessed: 19 March 2014].

xvii Many supporters of consensus implicitly accept this when they say that a prerequisite for using the process is shared views and goals within a group.

xviii Pitkin famously said there were four different, and often competing, forms of representation: . Previous to the modern period elections were thought of as the opposite to democracy. Instead citizen juries and other participatory processes were associated with the word.

xix This is far beyond the scope of this article to discuss. And even what we mean by the state is unclear, but unpicking it is closely linked to unpicking undemocratic power associated with it. See for example:

xx Splitting up into smaller groups for some bits of discussion is a useful method . Or for large groups a ‘Spokescouncil’ method can be used. See: for a discussion of some tools groups can use.

xxi This article discusses some of these criticisms of consensus – and challenges us to reach a consensus on using consensus decisions-making, arguing that if we can’t do that, we shouldn’t be using it.

We do see people looking for alternatives to consensus because of this. One alternative which has gained some popularity is ‘sociocracy’. I am sceptical about sociocracy for several reasons.

a) Sociocracy is both a method of having meetings but also a structure for organising meetings. That, in itself, is a good thing, and makes it unlike ‘consensus decision-making’ which is only about how you organise meetings.

It is important to think about wider structure, because a democratic society or a democratic organisation must be one where the relations between different groups or meetings function to make the whole democratic, not just the individual parts (meetings) on their own.

However, sociocracy, as I understand it, specifies that meetings (‘circles’) should be ‘double-linked’ – each of them should be linked by one person from each circle attending meetings of the other circles. Even if functional, this isn’t necessarily a good thing: for example, if this means that there will always be someone from the executive board of an organisation within meetings at other levels, existing power relations are likely to mean that what they say, goes. Sociocracy also makes no specifications about what the relations between different groups should look like, and therefore allows a situation where one ‘circle’ is given authority over another.

Structure needs to take account of, and often work against, existing power relations, to try and democratise organisations. That’s not to say that sociocracy couldn’t work in a democratic way, but perhaps it needs clearer guidelines to avoid being anti-democratic.

My preference is for coordination between different branches/groups/meetings to be done by collective meetings of recallable and rotating delegates – a spokescouncil-like model, perhaps. I also think there’s a case for decisions made more centrally to go back to ratification and even amendments at a more grassroots level.

b) The requirement in sociocracy for arguments to be framed in terms of the interest of the organisation as a whole could be quite harmful. This also puts a lot of power in the hands of whoever is able to define the interests of the organisation – often likely to be its executives. For many organisations, ‘the interests of the organisation’ is defined as whatever is best for its finances – which has profoundly anti-social consequences. People being framed as ‘traitors’ by nationalists, or socialists being kicked out of the Labour party for supposedly having campaigned against it previously, are examples of how organisational loyalty can have reactionary consequences.

xxiii This is easier said that done and may not always be possible. Facilitators must use their judgement, but their decisions should be challenge-able by participants, if they are rushing through a particular decision, for example.

Two ways of democratising this power of facilitation could be, firstly, going through the agenda as a group at the start of each meeting, and putting topics to be tackled in order of importance and agreeing how long to spend on each. Another way of doing it would be for each participant to allocate, with post-it notes or a pen, how long they would spend on each decision and this could simply be averaged.

Secondly, during decision-making, participants should be able to suggest a different way of proceeding, with a ‘process point’ hand signal (thumb and forefinger made into the shape of an O?). Process point suggestions could be accepted by default, or if challenged, put to an instant vote.

xxiv I recently had a discussion with someone who was setting up a new organisation aiming to be a platform for alternative media across the UK, but who, whilst professing support for consensus, was keen ‘not to make it too open too early on’. Whilst people should be able to do their own things and organise in their own ways, where a project claims to be for everyone or on behalf of a population, there should be a duty to facilitate democratic involvement in shaping this. At what point does a ‘private’ project need to be open to wider stakeholders? This is not a question that has been given much discussion. A culture where private poverty and wage slavery are endemic and where economic democracy is almost non-existent have allowed this crucial question to go ignored.