Consensus decision-making

Members of the Shimer College Assembly reaching a consensus through deliberation

Consensus decision-making or consensus process (often abbreviated to consensus) are group decision-making processes in which participants develop and decide on proposals with the aim, or requirement, of acceptance by all. The focus on establishing agreement of the supermajority and avoiding unproductive opinion, differentiates consensus from unanimity, which requires all participants to support a decision.

The word consensus is Latin meaning "agreement, accord", derived from consentire meaning "feel together".[1] Broadly, consensus relates to a generally accepted opinion,[2] but in the context of this article refers to the process and the outcome of consensus decision-making (e.g. "to decide by consensus" and "a consensus was reached").

Consensus decision-making is an alternative to commonly practiced group decision-making processes.[5] Robert's Rules of Order, for instance, is a guide book used by many organizations. This book allows the structuring of debate and passage of proposals that can be approved through majority vote. It does not emphasize the goal of full agreement. Critics of such a process believe that it can involve adversarial debate and the formation of competing factions. These dynamics may harm group member relationships and undermine the ability of a group to cooperatively implement a contentious decision. Consensus decision-making attempts to address the beliefs of such problems. Proponents claim that outcomes of the consensus process include:[3][6]

Consensus is not synonymous with "unanimity"– though that may be a rule agreed to in a decision making process. The level of agreement necessary to finalize a decision is known as a "decision rule".[3][7]

In very rapid decision making, simple consensus (computer science) rules are often imposed, such as:

Generally such rules assume a certain number of participants and thus would satisfy consensus thresholds stated in percentage terms. That is, if a majority of nine may override three in a group of twelve, that is a 75% consensus threshold but it is also a "unanimity minus three" threshold, and may be stated either or both ways, as in "75% or unanimity minus three whichever is harder to achieve", in a group's constitution. Such a statement allows for minorities to be more robustly represented in abstention or absence scenarios.

Even in rapid decision making contexts, minorities (the "minus") have the right to have dissenting opinion or negative outcome predictions recorded.

If there is any single simple rule that defines what is not consensus decision-making, it is censoring the dissenting opinion. Regardless of how decisions are made, dissents are always recorded in all consensus decision making systems, if only so that accuracy of predictions can be examined later so the group can learn. This principle can be applied in any system, but it is fundamental to all consensus. More controversially, systems that require unanimity are prone to hiding or intimidating, rather than recording, dissent (for example, groupthink). Many authors consider unanimity to be a sign of an inherently wrong decision. The Sanhedrin courts of ancient Israel were of this view, and biblical scholars note that the trial of Jesus was inherently unfair for being a unanimous guilty verdict.[8]

In groups of human participants, there are psychological implications to dissent, and not all participants are equal. For example, participants may:

For these reasons, most consensus decision-making emphasizes finding out why dissent occurs. In democratic contexts, political theory debates how to deal with dissent and consensus where violent opposition is possible (or even likely). Weale (1999) states the problem as:

Even with goodwill and social awareness, citizens are likely to disagree in their political opinions and judgments. Differences of interest as well as of perception and values will lead the citizens to divergent views about how to direct and use the organized political power of the community, in order to promote and protect common interests. If political representatives reflect this diversity, then there will be as much disagreement in the legislature as there is in the population.[9]

Rules and processes simply are never enough to resolve these questions, and a robust debate for millennia on political virtues has focused on what human characteristics participants must cultivate to achieve harmony under diversity.

To ensure the agreement or consent of all participants is valued, many groups choose unanimity or near-unanimity as their decision rule. Groups that require unanimity allow individual participants the option of blocking a group decision. This provision motivates a group to make sure that all group members consent to any new proposal before it is adopted. Proper guidelines for the use of this option, however, are important. The ethics of consensus decision-making encourage participants to place the good of the whole group above their own individual preferences. When there is potential for a block to a group decision, both the group and dissenters in the group are encouraged to collaborate until agreement can be reached. Simply vetoing a decision is not considered a responsible use of consensus blocking. Some common guidelines for the use of consensus blocking include:[3][10]

A participant who does not support a proposal may have alternatives to simply blocking it. Some common options may include the ability to:

The basic model for achieving consensus as defined by any decision rule involves:

All attempts at achieving consensus begin with a good faith attempt at generating full-agreement, regardless of decision rule threshold.

In the spokescouncil model, affinity groups make joint decisions by each designating a speaker and sitting behind that circle of spokespeople, akin to the spokes of a wheel. While speaking rights might be limited to each group's designee, the meeting may allot breakout time for the constituent groups to discuss an issue and return to the circle via their spokesperson. In the case of an activist spokescouncil preparing for the A16 Washington D.C. protests in 2000, affinity groups disputed their spokescouncil's imposition of nonviolence in their action guidelines. They received the reprieve of letting groups self-organize their protests, and as the city's protest was subsequently divided into pie slices, each blockaded by an affinity group's choice of protest. Many of the participants learned about the spokescouncil model on the fly by participating in it directly, and came to better understand their planned action by hearing others' concerns and voicing their own.[16]

The group first elects, say, three referees or consensors. The debate on the chosen problem is initiated by the facilitator calling for proposals. Every proposed option is accepted if the referees decide it is relevant and conforms with the Universal Declaration of Human Rights. The referees produce and display a list of these options. The debate proceeds, with queries, comments, criticisms and/or even new options. If the debate fails to come to a verbal consensus, the referees draw up a final list of options - usually between 4 and 6 - to represent the debate. When all agree, the chair calls for a preferential vote, as per the rules for a Modified Borda Count, MBC. The referees decide which option, or which composite of the two leading options, is the outcome. If its level of support surpasses a minimum consensus coefficient, it may be adopted.[17][18]

Groups that require unanimity commonly use a core set of procedures depicted in this flow chart.[19][20][21]

Once an agenda for discussion has been set and, optionally, the ground rules for the meeting have been agreed upon, each item of the agenda is addressed in turn. Typically, each decision arising from an agenda item follows through a simple structure:

Quaker-based consensus[22] is said to be effective because it puts in place a simple, time-tested structure that moves a group towards unity. The Quaker model is intended to allow hearing individual voices while providing a mechanism for dealing with disagreements.[6][23][24]

The Quaker model has been adapted by Earlham College for application to secular settings, and can be effectively applied in any consensus decision-making process.

Key components of Quaker-based consensus include a belief in a common humanity and the ability to decide together. The goal is "unity, not unanimity." Ensuring that group members speak only once until others are heard encourages a diversity of thought. The facilitator is understood as serving the group rather than acting as person-in-charge.[25] In the Quaker model, as with other consensus decision-making processes, articulating the emerging consensus allows members to be clear on the decision in front of them. As members' views are taken into account they are likely to support it.[26]

The consensus decision-making process often has several roles designed to make the process run more effectively. Although the name and nature of these roles varies from group to group, the most common are the facilitator, consensor, a timekeeper, an empath and a secretary or notes taker. Not all decision-making bodies use all of these roles, although the facilitator position is almost always filled, and some groups use supplementary roles, such as a Devil's advocate or greeter. Some decision-making bodies rotate these roles through the group members in order to build the experience and skills of the participants, and prevent any perceived concentration of power.[27]

Critics of consensus blocking often observe that the option, while potentially effective for small groups of motivated or trained individuals with a sufficiently high degree of affinity, has a number of possible shortcomings, notably

Consensus seeks to improve solidarity in the long run. Accordingly, it should not be confused with unanimity in the immediate situation, which is often a symptom of groupthink. Studies of effective consensus process usually indicate a shunning of unanimity or "illusion of unanimity"[40] that does not hold up as a group comes under real-world pressure (when dissent reappears). Cory Doctorow, Ralph Nader and other proponents of deliberative democracy or judicial-like methods view explicit dissent as a symbol of strength.

In his book about Wikipedia, Joseph Reagle considers the merits and challenges of consensus in open and online communities.[41] Randy Schutt,[42] Starhawk[43] and other practitioners of direct action focus on the hazards of apparent agreement followed by action in which group splits become dangerously obvious.

Unanimous, or apparently unanimous, decisions can have drawbacks.[44] They may be symptoms of a systemic bias, a rigged process (where an agenda is not published in advance or changed when it becomes clear who is present to consent), fear of speaking one's mind, a lack of creativity (to suggest alternatives) or even a lack of courage (to go further along the same road to a more extreme solution that would not achieve unanimous consent).

Unanimity is achieved when the full group apparently consents to a decision. It has disadvantages insofar as further disagreement, improvements or better ideas then remain hidden, but effectively ends the debate moving it to an implementation phase. Some consider all unanimity a form of groupthink, and some experts propose "coding systems ... for detecting the illusion of unanimity symptom".[45] In Consensus is not Unanimity, long-time progressive change activist Randy Schutt writes:

Many people think of consensus as simply an extended voting method in which everyone must cast their votes the same way. Since unanimity of this kind rarely occurs in groups with more than one member, groups that try to use this kind of process usually end up being either extremely frustrated or coercive. Decisions are never made (leading to the demise of the group), they are made covertly, or some group or individual dominates the rest. Sometimes a majority dominates, sometimes a minority, sometimes an individual who employs “the Block.” But no matter how it is done, this coercive process is not consensus.[42]

Confusion between unanimity and consensus, in other words, usually causes consensus decision-making to fail, and the group then either reverts to majority or supermajority rule or disbands.

Most robust models of consensus exclude uniformly unanimous decisions and require at least documentation of minority concerns. Some state clearly that unanimity is not consensus but rather evidence of intimidation, lack of imagination, lack of courage, failure to include all voices, or deliberate exclusion of the contrary views.

Some proponents of consensus decision-making view procedures that use majority rule as undesirable for several reasons. Majority voting is regarded as competitive, rather than cooperative, framing decision-making in a win/lose dichotomy that ignores the possibility of compromise or other mutually beneficial solutions.[46] Carlos Santiago Nino, on the other hand, has argued that majority rule leads to better deliberation practice than the alternatives, because it requires each member of the group to make arguments that appeal to at least half the participants.[47]

Some advocates of consensus would assert that a majority decision reduces the commitment of each individual decision-maker to the decision. Members of a minority position may feel less commitment to a majority decision, and even majority voters who may have taken their positions along party or bloc lines may have a sense of reduced responsibility for the ultimate decision. The result of this reduced commitment, according to many consensus proponents, is potentially less willingness to defend or act upon the decision.

Majority voting cannot measure consensus. Indeed,—so many 'for' and so many 'against'—it measures the very opposite, the degree of dissent. Consensus voting, in contrast, the Modified Borda Count, MBC, can identify the consensus of any electorate, whenever such a consensus exists. Furthermore, the rules laid down for this procedure can be the very catalyst of consensus.

Some formal models based on graph theory attempt to explore the implications of suppressed dissent and subsequent sabotage of the group as it takes action.[48]

High-stakes decision-making, such as judicial decisions of appeals courts, always require some such explicit documentation. Consent however is still observed that defies factional explanations. Nearly 40% of the decisions of the United States Supreme Court, for example, are unanimous, though often for widely varying reasons. "Consensus in Supreme Court voting, particularly the extreme consensus of unanimity, has often puzzled Court observers who adhere to ideological accounts of judicial decision making."[49] Historical evidence is mixed on whether particular Justices' views were suppressed in favour of public unity.[50]

Another method to promote agreement is to use a voting process under which all members of the group have a strategic incentive to agree rather than block.[37] However, this makes it very difficult to tell the difference between those who support the decision and those who merely tactically tolerate it for the incentive. Once they receive that incentive, they may undermine or refuse to implement the agreement in various and non-obvious ways. In general voting systems avoid allowing offering incentives (or "bribes") to change a heartfelt vote.

In the Abilene paradox, a group can unanimously agree on a course of action that no individual member of the group desires because no one individual is willing to go against the perceived will of the decision-making body.[51]

Since consensus decision-making focuses on discussion and seeks the input of all participants, it can be a time-consuming process. This is a potential liability in situations where decisions must be made speedily, or where it is not possible to canvass opinions of all delegates in a reasonable time. Additionally, the time commitment required to engage in the consensus decision-making process can sometimes act as a barrier to participation for individuals unable or unwilling to make the commitment.[52] However, once a decision has been reached it can be acted on more quickly than a decision handed down. American businessmen complained that in negotiations with a Japanese company, they had to discuss the idea with everyone even the janitor, yet once a decision was made the Americans found the Japanese were able to act much quicker because everyone was on board, while the Americans had to struggle with internal opposition.[53]

Outside of Western culture, multiple other cultures have used consensus decision-making. One early example is the Haudenosaunee (Iroquois) Confederacy Grand Council, which used a 75% supermajority to finalize its decisions,[54] potentially as early as 1142.[55] In the Xulu and Xhosa (South African) process of indaba, community leaders gather to listen to the public and negotiate figurative thresholds towards an acceptable compromise. The technique was also used during the 2015 United Nations Climate Change Conference.[56][57] In Aceh and Nias cultures (Indonesian), family and regional disputes, from playground fights to estate inheritance, are handled through a musyawarah consensus-building process in which parties mediate to find peace and avoid future hostility and revenge. The resulting agreements are expected to be followed, and range from advice and warnings to compensation and exile.[58][59]

Consensus-building and direct democracy experimentation was a feature of voter registration projects by the Student Nonviolent Coordinating Committee (SNCC) in the American South; the Economic Research and Action Project (ERAP) of Students for a Democratic Society (mid-1960s), some women's liberation groups (late 1960s to early 1970s) and anti-nuclear and peace movement groups (late 1970s and early 1980s).[60] For example, the anti-nuclear Clamshell Alliance and Movement for a New Society engaged in consensus decision-making processes.[61] The origins of formal consensus-making can be traced significantly further back, to the Religious Society of Friends, or Quakers, who adopted the technique as early as the 17th century.[62] Anabaptists, including some Mennonites, have a history of using consensus decision-making[63] and some believe Anabaptists practiced consensus as early as the Martyrs' Synod of 1527.[62] Some Christians trace consensus decision-making back to the Bible. The Global Anabaptist Mennonite Encyclopedia references, in particular, Acts 15[64] as an example of consensus in the New Testament. The lack of legitimate consensus process in the unanimous conviction of Jesus by corrupt priests[65] in an illegally held Sanhedrin court (which had rules preventing unanimous conviction in a hurried process) strongly influenced the views of pacifist Protestants, including the Anabaptists (Mennonites/Amish), Quakers and Shakers. In particular it influenced their distrust of expert-led courtrooms and to "be clear about process" and convene in a way that assures that "everyone must be heard".[66]

Consensus voting was advocated, among others, by Ramón Llull in 1199, by Nicholas Cusanus in 1435, by Jean-Charles de Borda in 1784, by Hother Hage in 1860, by Charles Dodgson (Lewis Carroll) in 1884, and by Peter Emerson in 1986.

Japanese companies normally use consensus decision-making, meaning that unanimous support on the board of directors is sought for any decision.[67] A ringi-sho is a circulation document used to obtain agreement. It must first be signed by the lowest level manager, and then upwards, and may need to be revised and the process started over.[68]

In the Internet Engineering Task Force (IETF), decisions are assumed to be taken by rough consensus.[69] The IETF has studiously refrained from defining a mechanical method for verifying such consensus, apparently in the belief that any such codification leads to attempts to "game the system." Instead, a working group (WG) chair or BoF chair is supposed to articulate the "sense of the group."

One tradition in support of rough consensus is the tradition of humming rather than (countable) hand-raising; this allows a group to quickly discern the prevalence of dissent, without making it easy to slip into majority rule.[70]

Much of the business of the IETF is carried out on mailing lists, where all parties can speak their views at all times.

In 2001, Robert Rocco Cottone published a consensus-based model of professional decision-making for counselors and psychologists.[71] Based on social constructivist philosophy, the model operates as a consensus-building model, as the clinician addresses ethical conflicts through a process of negotiating to consensus. Conflicts are resolved by consensually agreed on arbitrators who are selected early in the negotiation process.

The United States Bureau of Land Management's policy is to seek to use collaborative stakeholder engagement as standard operating practice for natural resources projects, plans, and decision-making except under unusual conditions such as when constrained by law, regulation, or other mandates or when conventional processes are important for establishing new, or reaffirming existing, precedent.[72]

The Polish–Lithuanian Commonwealth of 1569–1795 used consensus decision-making in the form of liberum veto ('free veto') in its Sejms (legislative assemblies). A type of unanimous consent, the liberum veto originally allowed any member of a Sejm to veto an individual law by shouting Sisto activitatem! (Latin: "I stop the activity!") or Nie pozwalam! (Polish: "I do not allow!").[73] Over time it developed into a much more extreme form, where any Sejm member could unilaterally and immediately force the end of the current session and nullify any previously passed legislation from that session.[74] Due to excessive use and intentional sabotage from neighboring powers bribing Sejm members, legislating became very difficult and weakened the Commonwealth. Soon after the Commonwealth banned liberum veto as part of its Constitution of 3 May 1791, it dissolved under pressure from neighboring powers.[75]