Avoiding Oligarchy Through Sortition

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 15.3

a food co-op board

Let’s look at the way that a typical consumer co-op might select its governing board, in its attempt to be democratic. The members of the small board of directors might have overlapping terms of two or three years, with some portion of the board elected each year. If there aren’t term limits, many board members probably are re-elected to multiple terms. It is likely that there is a nomination committee, usually appointed by (and often drawn from) the existing board. In some unusual cases they may propose more candidates than there are seats to fill, in order to give an actual choice to the members, but more often they nominate the same number of candidates as seats to be filled. From time to time, there may also be renegade members who get over the necessary hurdles to get their names on the ballot. Nonprofit boards often have a dearth of candidates seeking to serve. Those who do put themselves forward may happen to be excellent, but alternatively may have problematic traits or conflicting interests. Some may simply be seeking to bolster their resume by adding some “leadership” credentials.

The general members typically get a bit of information, such as a brief bio and candidate statement and perhaps even answers to some standard questions, and a photo of each candidate. It is rare to have any real campaign or opportunity for candidates to challenge each other’s assertions in their statements, and no fact-checking. Then, a small percentage of members take the time to vote, generally with little or no understanding of the issues facing the co-op and little or no reliable knowledge about substantive differences between the candidates, generally relying on superficial impressions. If things are running smoothly at the co-op, the incumbents are likely to be returned, and if there are serious problems there is some chance they may be replaced — if there happen to be any alternative candidates. Many members are unaware that the election is even taking place. Any efforts to improve turnout (perhaps a raffle prize?) have only a small effect, prompting a few more, still uninformed, members to send in their ballot, or attend the annual meeting. This is probably the best that can be realistically hoped for under an electoral scheme.

In some cases, it may be the intention of the co-op leadership to maintain an undemocratic status quo, by intentionally minimizing member participation. But more often it is the lack of interest and lack of knowledge among members that is the crux of the problem. In the eighteenth century, English philosopher Jeremy Bentham used the term “active aptitude” to describe the conscientious attention to duty that a decision-maker needed to have. He wasn’t referring to some inherent character trait of people, but rather the results of various possible procedures and designs of legislative bodies. It is this active aptitude that citizens will generally fail to exhibit when acting as a rank and file member of a large “democratic” association. While the pejorative term “apathy” may be applied, “rational ignorance” is a more appropriate term. It just isn’t reasonable to expect most people to devote the amount of effort needed to remain “up to speed” on the large number of issues that face every would-be democratic association, especially when their one vote as part of a mass is extremely unlikely to actually make any difference (especially when a slate is proposed that matches the number of seats to be elected). This rational ignorance applies whether voting in an issue referendum or candidate election.

Sortition can solve these problems. The solution that sortition offers is asking a subset of members to give full attention to the needs of their co-op for a relatively short period of time, with the expectation that another subset will take a turn subsequently, and then another, etc. Not all members can have active aptitude, but some can for a limited duration. A large statistically representative sample might be needed on occasion (for very big decisions), but much smaller random samples may be sufficient for certain governance tasks. Unlike governmental mini-publics, which sometimes need to balance fundamentally incompatible societal interests, members of democratic associations generally have a common shared interest that permit a somewhat smaller sample to suffice. Also, members of a smallish board of directors are more akin to delegated agents, rather than constituting an accurately representative body. This concept is affirmed by the commonly used term “board of trustees.”

However, I need to be clear that the idea is not to randomly select a small board of directors. There are many obvious problems with such an approach. If the only concern were impartiality and protection from corrupt manipulation (such as for a court jury), random selection of a small sample might suffice — but democracy also requires some degree of representativeness. Such a small sample size has a high risk of being unacceptably unrepresentative of the membership. Statistical representativeness is tied to the law of large numbers.  Another, perhaps more serious problem with selecting a small board by random selection, is the issue of commitment. A board would need to serve longer terms than a typical court jury — probably a year at an absolute minimum —  and many members simply wouldn’t have the necessary level of commitment, free time, or energy to give adequate attention. Further, a typical board of directors ideally would have an assortment of traits, some of which might be fairly rare in the pool of members. These desired traits would probably be included in a largish mini-public, but could easily be missing in a small board. There is a distinction between a mini-public that only makes broad decisions about policy direction, as opposed to the ongoing detailed oversight of a business’s finances, etc. typically required of a board.

The better way to employ sortition, if a relatively small board is intended, is through a two-step process. A larger, more representative short duration mini-public could be randomly selected to act as a sort of nominating or hiring committee to recruit and select a smaller board of directors, or in lieu of a board of directors, adopt strategic plans while evaluating, as well as hiring and firing, a manager directly. My focus here is not whether a small board of directors, or large member assemblies make policy, but rather how the general membership can exercise informed control of the organization. To maximize the willingness of ordinary members to participate in a mini-public, the jury would be of short duration (perhaps only a few meetings) and its members would be compensated in some manner (e.g. a discount, catered meals during meetings, or direct payments). Short duration representative juries are an excellent way to assure ultimate authority and regular oversight by the ordinary members as a whole. Very few members will keep on top of their coop's issues week by week, year after year, but most members would be willing to focus on their coop governance matters for a small amount of time (perhaps with some compensation), knowing other regular members will do likewise in turn. 

Relying on pure volunteerism for this task is dangerous, since self-selection bias can allow unrepresentative special interests to dominate. However, one impetus toward oligarchy, which Michels discussed, was the control of information by the existing staff or board. It is entirely possible that a temporary mini-public could be artfully manipulated if they were dependent on such “oligarchs' ' for their information. For this reason an effort to use sortition as a democratic tool also necessitates independent facilitation and staffing, not selected by the existing staff or board. The procedure for ensuring independence needs to be institutionalized at the outset, and not be an ad hoc effort at the time of each lottery drawing. There are many possible ways of managing this. Perhaps a separate oversight board made up of members from previous mini-publics would be charged with arranging facilitation for the next mini-public. If sortition became widespread, a cottage industry of experienced mini-public facilitators might grow (indeed has grown in many countries) who would be crowd-source-evaluated for such things as competence and impartiality, etc. Chosen facilitators would need to have free access (with appropriate confidentiality commitments) to all of the co-op’s records and information, to assure the mini-public got fully informed. These facilitators would be evaluated by the staff and members of the mini-public, and file a report about problems and successes at the conclusion of the mini-public that the subsequent facilitator could use to make improvements.

To demonstrate another superiority of a sortition recruiting process for boards, over elections, let's take a hypothetical example of a cooperative whose members all want to have a balanced board that reflects the diversity of the membership. They all want a board that also includes a member with a legal background, one with budget and bookkeeping experience, and one with lots of media skills. Suppose plenty of candidates meeting these criteria decide to run for the board in an election. Because of the problem of voter coordination, regardless of whether a block plurality, ranked-choice preferential, or other voting method is used, it could easily turn out that the mix on the board that gets elected fails to meet most of these criteria that all of the members believe are important. The board may turn out to be all white males without any media or bookkeeping experience. The way that some coops try to overcome this problem is by having a nomination committee select a favored “slate.” For this to work, the election itself must be nominal or token, and it is the selection of the nominating committee that is the point of democratic challenge. With the jury model, a large representative nominating committee can be selected by lot. This randomly selected mini-public could interview potential board members and seek to come to a consensus about the best mix of people to form a balanced board. The election could be dispensed with or maintained merely pro forma or to give the membership a veto option. While most democratic associations have bylaws that mandate the use of elections, they rarely mandate the details for the process of nomination, leaving the door open for easy implementation of a sortition process. Once the system has proven itself, however, it is desirable to amend the bylaws to institutionalize the sortition procedure, because the iron law of oligarchy will always remain a threat with some current group in authority seeking to cement their power.

However, even if the board is democratically selected using a sortition recruitment body, there is another potential locus of “elite control.” An executive director is generally hired and fired by the board. But the board is very part-time and gets most of their information from the executive director, who promotes a friendly “same-team” relationship. It is common for nonprofits to actually have the executive director or staff manipulate the board and preserve oligarchic control. For this reason, it is better to have the executive director hired and fired not by the board, but by single-purpose lottery selected review bodies that have no relationship with the executive director, but can interview supporters and detractors.

The lottery selection process itself must be completely transparent and public so that there is never any question within the membership whether the mini-public selection was honest or manipulated. In Classical Athens, a simple mechanical device was invented called a kleroterion, which was used to randomly and publicly assign citizens to mini-publics tasked with making public decisions. It might be wise to display and honor this lottery drawing as the foundation of the organization’s democracy with some sort of public celebratory event.