The Sortition Solution

First: Clearing Some Weeds

Before I dig into the potential benefits of sortition, I want to clear away a few assumptions, or “weeds”, that may prompt some to prejudge or even dismiss the concept before considering the arguments. This book focuses on the many negative features of democracy based on elections. However, there are certain common functions of elective office, which are generally accepted as positive, that we would remove from lawmakers in a sortition democracy. Three functions that elected legislators are commonly expected to fulfill are political leadership, constituency service, and providing legislative expertise.

Leadership

I discussed leadership in some detail above in the section on executive elections in Chapter 2, but it warrants a cursory reexamination here in a legislative context. A sortition democracy would certainly still have many policy leaders with broad appeal seeking to gain public support for particular proposals—they simply wouldn’t be elected officials. Within an elected legislature, partisan teams are akin to soldiers constantly preparing for battle against the enemy in an opposing political party. In that milieu, leadership like that of a general is valued and even necessary. But a sortition-based lawmaking process would have none of that. There, the goal would be problem-solving and the search for common ground. Outside policy leaders might be invited to give testimony, but would be prohibited from exercising leadership within a mini-public. Think of an outside leader seeking to steer a court jury’s direction. That is appropriately deemed to be criminal. It is my contention that partisan political leadership would do far more harm than good in a sortition democracy.

Constituency Service

A major function that American legislators take on is constituency service—such as helping citizens wend their way through a confusing government bureaucracy to get services they are entitled to. A significant portion of congressional staff time is devoted to such constituent service. Rather than accepting this as a positive, we should recognize this as a failing of the government. An important point to recognize is that it is in the interests of legislators to preserve a faulty system that constantly feeds a stream of voters to them seeking help. Like the various patronage schemes of old, constituency service allows incumbents to curry favor with voters in hopes of receiving their votes down the road. Incumbent legislators actually have a built-in incentive to not fix systemic problems, in order to keep this source of voter help and reciprocation in place. An opportunity their election challengers can’t match. An impartial system of ombudsmen combined with a steady improvement of systemic flaws would be a natural alternative in a sortition democracy.

Legislative Expertise

Many people assume that legislators exhibit a certain kind of expertise. After all, society has grown so complicated that surely our representatives must develop an appropriate level of expertise, at least in the policy area of their committee assignments. Elected politicians tend to become experts in the fields of negotiation, public relations, and campaigning, while leaving to lobbyists and their staff the task of mastering policy and bill-drafting expertise. In an article titled “Three Reasons Congress Is Broken”, the author and former Washington Post journalist Robert Kaiser quotes Sen. Ted Kennedy from his 2009 memoir. Senator Ted Kennedy estimated that 95 percent of the drafting and negotiating in Congress is done by staff rather than legislators (Kaiser 2013). Legislators learn sufficient terminology and memorize enough key facts and budget numbers to seem well-informed, but only occasionally become bona fide experts on policy matters (and when they believe they have expertise, they generally suffer from the illusion of knowledge bias I discussed at the end of Chapter 8). When legislators (or their hired staff) bring in bona fide policy experts, they give inordinate attention to the public relations and fundraising implications, rather than the substance of policy.

If modern society is indeed so complicated that average citizens simply don’t know enough to govern themselves, then elected legislators are even further from that level of adequacy. In short, professional politicians are frequently ego-driven professional public relations people rather than policy experts. Average citizens in an allotted decision-making body at least have an inherent interest in finding genuine and impartial experts for advice, without the overwhelming diversion of electoral campaign, fundraising, and public relations concerns. Thus, putting “average” citizens in the driver’s seats enhances the likelihood that true and appropriate experts will be called upon as needed. This is a key point. When proposing the idea of allowing everyday citizens to make complex policy decisions, many people recoil with fear—imagining this would mean a loss of policy expertise. I contend that just the opposite effect would predominate. Politicians also have minimal actual policy expertise, but have the distorting incentive to seek out advice from like-minded partisans, special interest lobbyists (with large campaign donation potential), and media manipulation experts. Everyday citizens, on the other hand, would have an incentive to seek out advice from a range of genuine policy experts.

Using sortition for the crafting and review of legislation, and separately for final adoption, offers many benefits over election. The electoral problems below have been addressed in the foregoing chapters. But the inadequacy of elections does not automatically show that sortition would perform better on each point. While I addressed many of these problems in Chapter 4 (about election reforms), the summary below shows how sortition digs out and eliminates each problem at its root. As I mentioned in the Introduction in Chapter 1, the transition to sortition is unlikely to be sudden, and elections and sortition will likely exist side by side for a time. This will weaken the sortition solutions to various degrees during that transition. But this book is intended to be a “North Star” showing the direction democracy needs to go. So, the discussion below is about the impact of sortition at the end point, or when elections are merely a vestigial remnant. I will discuss ideas about how the transition to a sortition-based democracy might work in the final chapters of this book.

Benefits of Sortition-based Democracy

Okay, now that we’ve cleared away those weeds …

Representativeness

The ability of random selection or scientific sampling to generate a far more descriptively representative body than elections is self-evident, and confirmed by hundreds of actual sortition implementations. This representativeness may not be perfect, but will not be symbolic or token—but deep. In addition to mirroring the gender, race, age, income, and ethnic makeup of the community, the values, life experiences, class, religious beliefs or lack thereof, occupations, political philosophy, cognitive style and interests will also be a far better match than in any elected legislative chamber. This result is dependent, however, on either extremely effective stratified sampling or the use of quasi-mandatory service (as with jury duty in many jurisdictions). The risk of self-selection bias in non-mandatory mini-publics, in terms of personality and cognitive style, means that more research still needs to be done on the effectiveness of stratified sampling. Unfortunately, such research using parallel mini-publics, where some are drawn using stratified sampling from those willing to serve, and the others have quasi-mandatory service, could only be conducted by a government with statutory authority to compel participation. Regardless, the accuracy of the representativeness would be dramatically greater than any elected body. Such a mini-public would be, as John Adams described an ideal legislature in 1776, “an exact portrait of the people at large. It should think, feel, reason and act like them” (Adams 1776). It should be noted, however (as I discussed in Chapter 6), that descriptive representativeness was not a goal of electoral systems. Elections were based on the aspiration (or fantasy) that elected politicians would be superior men, with civic virtue, who were authorized to rule on behalf of the people.¹

Diversity

Diversity is a companion trait with representativeness, but not synonymous. I have stressed the tremendous value of diversity for policy drafting and filtering bodies previously. Sortition assures wide-ranging diversity in the deliberative process. Elections return individuals who, despite substantive political and philosophical differences, exhibit significant homogeneity in other attributes. In Chapter 10, I examined how the collective intelligence of a diverse group can be more beneficial than even a homogeneous group made up of experts. From the time of Aristotle to the most recent scholarly research, diversity has been shown to ramp up group creativity and problem-solving ability.

Campaign Finance and Lobbying

Obviously, an absence of campaigning eliminates the elitism filter and corruption problems associated with raising campaign funds. Under a sortition system, the flagrant corruption facilitated by campaign fundraising would vanish. Independent campaign spending by wealthy interests to sway undecided voters would also disappear. However, independent expenditures could theoretically still be made, but they would need to influence the public in general on topics where a jury is likely to be called. Jury tampering laws, and perhaps even sequestering of some mini-publics, would preclude targeted spending by special interests. Society-wide expenditures would be dramatically less efficient and less effective for wealthy interests than traditional campaign spending. In an electoral campaign situation, with voter rational ignorance, there are standard advertising methods to turn enough ill-informed swing voters to affect the outcome. If applied in a sortition context, however, this sort of superficial persuasion would be easily negated once individuals were selected through a lottery. This is due to the intensive learning, presented with a broad range of information, as well as the focused attention fostered by being in a relatively small, randomly selected decision-making group.

It is helpful to use an analogy of a traditional court trial. Imagine if jurors and judges solicited and received contributions of money from the prosecutors and defense team in a trial, and had private meetings with them outside of the official court proceedings. That is the situation we have in our existing legislative process, where elected legislators serve effectively as judges and jurors, but on matters of policy rather than guilt or innocence. It might be decided that once a jury is called, such legislative juries should be sequestered as a defense against jury tampering (lobbying), as is often the case in criminal trials.

Gerrymandering

Obviously, all problems related to gerrymandering could not exist in a system based on sortition rather than election. Multi-member districts using proportional representation voting methods can reduce the effects of gerrymandering, but sortition eliminates it completely. With no elections, there are no district lines to be drawn, and gerrymandering cannot exist. The analogous challenge for sortition is the creation of a scientifically accurate and publicly transparent random selection procedure.

Participation

The term participatory democracy is often used to refer to a system in which all citizens can vote in numerous referendums. Sortition is not “participatory” in this sense. In order to overcome rational ignorance, it is essential that a relatively small number of citizens be deeply engaged in investigating and then deciding on each piece of legislation. Thus, sortition can be considered a form of representative democracy. However, compared to a traditional electoral “representative democracy”, sortition can be structured so as to spread not only participation, but power, widely throughout the population—ruling and being ruled in turn.

The number of citizens participating depends on the details of how the system is designed. If Policy Juries meet for a relatively short duration, on a single bill, for example, the turnover would be very large. On the other hand, if a lottery-selected legislature simply replaced an elected legislature (a design that many people imagine when they first hear about sortition, but that I strongly oppose for many different reasons), in terms of numbers and terms of office, the only increase would be due to the improbability of serving multiple terms.

Sortition democracy can also open the flood gates for citizen participation by allowing all who wish to, to work on designing draft legislation that can be delivered to randomly selected Review Panels to consider and possibly pass on to randomly selected Policy Juries to adopt. Since this “raw material” participation would be self-selected, and thus unrepresentative, it is essential that such efforts be targeted at coming up with drafts and arguments, rather than deciding— proposing rather than disposing. Such open participation would also be substantially more significant, in terms of developing a democratic culture, than the kind of mass participation available to citizens in an electoral representative system. In a competitive election, participation is largely confined to putting up campaign signs or giving campaign donations. The vast majority of citizen “participation” is merely voting, with a few people also engaging in campaign (marketing) tasks. Only the post-election letter writing or social media posts may have any meaningful civic thought involved.

A sortition system, especially if it exists at local, state, and national levels on an issue-by-issue basis, could involve orders of magnitude more citizens in meaningful democratic deliberation and decision-making. Such democratic participation spreads democratic skills and appreciation of the challenges of good decision-making—fostering the “civic virtue” that republican theorists hoped for.

Rational Ignorance

In actual practice, electoral systems provide an attenuated and stilted range of choices. But even if the options included genuinely good choices, an Achilles’ heel of mass participation schemes, whether electing candidates or voting on referendums, is the inevitable fact that most voters will simply not be adequately informed about the choices the elites present. This then links to the ability of campaigns to influence voters based on factors totally unrelated to merit. It is not rational, nor even possible, that millions of people will devote the amount of time needed to adequately understand the ramifications of the options when their single vote is like a drop in the bucket. By reducing the number of decision makers on a particular issue to those who could fit in a single room, the evidence from hundreds of citizens’ assemblies suggests that most of the participants will have sufficient motivation to devote the effort and focus necessary to learn facts and use Kahneman’s system two “slow thinking”. Manipulation of inevitably ill-informed voters in elections by those with access to wealth and with control of media has always been a problem, but has now become a precise scientific endeavor. The attention garnered by social media algorithms, disinformation, and AI-generated fakes has made this a topic of discussion. But in mass elections, this problem of manipulation has always existed. It has never not existed, simply due to voter rational ignorance combined with determined partisans and special interests. Rather than tapping the wisdom of crowds or collective intelligence through deliberation, mass elections elicit the cliché herd-like “madness of crowds”.

With sortition, there is still a danger that members of even a relatively small body might fall back on blindly following “leaders”. from inside or outside the body, whether from partisan, ethnic, religious or other “tribes”. Thus, while sortition offers the best chance for overcoming rational ignorance, cognitive biases, or default heuristics, which plague elections, the exercise of deliberative thought will depend on the structure, culture, and various supports that are provided.

One small practice that might have some value is having those who are randomly selected commit to some sort of juror oath, such as this:

I agree to treat all those present with respect. I will listen with an open mind, and delay making any final decision on the matters under consideration until I have heard all of the evidence and arguments. I recognize that some things that I believed to be true may turn out to be wrong, and that other things that I did not accept may turn out to be valid. I commit to making a final decision that on balance I genuinely believe is in the best interests of all.

But even if we assume that most members of a randomly selected body will ignore the common good (or we assume that the very concept of a “common good” is a vain myth), and instead vote according to each member’s selfish interests, at worst we arrive at the ideal outcome championed by adversarial competitive (electoral) democracy, which is finding the majority preference among competing interests.

Self-Selection Bias

A frequent problem with electoral democracy is self-selection bias among decision makers. The electoral nomination process is based on self-selection of those with inflated egos, partisan loyalty and a tendency toward overconfidence and a level of certainty, making them less open-minded and thus poor deliberators. Those who advise the elected legislators (lobbyists) are also self-selected, often based on how much money stands to be made or lost from some public policy decision. Through random selection, sortition can virtually eliminate this problem (or sharply reduce it, if stratified sampling is used). The dynamic is not eliminated completely simply due to the fact that certain people are also more likely to participate intently if selected, while meeker individuals may play a less active role. Good facilitation has been shown to successfully encourage input from shyer members. A system for professional advisory staff can also avoid self-selection (self-interest) bias by presenters, with a professional civil service that structures balanced presentations to juries, rather than by reliance primarily on self-interest-motivated lobbyists, as in the existing electoral system.

Corruption

Sortition (combined with regular rotation in office) was historically employed in ancient Greece and the Italian city republics for centuries, from the medieval era through the Renaissance, as a defense against corruption and tyranny² (Dowlen 2008). It is worth thinking through exactly why sortition increases the ability to resist corruption using a small-scale example. Run-of-the-mill corruption, such as giving a bribe to a public official in order to get prompt action on some needed permit, is a sort of market transaction in a monopoly situation. Only one office can issue the permit, and the implied threat of moving an application to the bottom of the pile can be lucrative for the official, while the opportunity to move it to the top of the pile can be beneficial to the citizen. The corrupt official’s superior may also be taking a cut, and elected politicians may even be in on such corrupt deals. In many countries, such penny-ante corruption is standard practice. At this retail level, the briber, the official being bribed, and perhaps the supervisor, have an immediate interest in the success of the bribe. An official elected or appointed to put a stop to the corruption is subject to the temptation to instead participate in the ongoing corruption. The common perception of certain Mexican police taking bribes while nominally fighting drug cartels is an example.

But viewed from a distance, even bribers (when the potential bribers have no application pending), and society as a whole, have an interest in ending such corruption. Giving oversight authority, with the power to fire corrupt officials, to a randomly selected group of citizens allows this societal interest to overcome the individual interest of a briber or an ongoing official to engage in corruption. In other words, when facing a corrupt situation at a retail, one-on-one level, there is a compelling interest for each citizen to participate, if that is the local “norm”. But when gathered in a collective capacity, these same average citizens have an overriding interest in ending such corruption. Random selection in this circumstance can solve what economists term a “collective action problem”. This same anti-corruption potential of sortition can also be employed on a grand scale throughout a society.

Multi-Body Sortition

In Chapter 16, I will present one design for democratic control of a legislative process. While not a catchy name, I coined the term “multi-body sortition” in a 2013 paper I published (Bouricius 2013). For smaller municipal or regional implementations, some of the elements might be modified or eliminated, and the size of the bodies would likely also be adjusted. The implementations embedding sortition as a permanent part of governments in East Belgium, the city of Paris, and elsewhere have incorporated some aspects of this multi-body sortition design. Chapter 16 will present an overview of the design concept.

There are a variety of distinct tasks involved in developing, assessing, adopting and implementing government policies. It is a mistake to have each of these separate tasks performed by the same group of people, or by the same procedures. Not only are different design characteristics appropriate for different tasks, but the performance of one task actually hobbles the ability of that group of people to perform certain other tasks.

In this design, the final decision-making body for new laws would always be a large Policy Jury, functioning much the way nomothetai did in Classical Athens. In my multi-body design they would be convened to hear pro-and-con presentations about a proposal coming from a separate sortition body (a Review Panel), and vote it up or down. These juries would each deal with a single proposal, meet for a relatively short duration, have quasi-mandatory service (like court juries), and be quite large (perhaps a thousand members for national Policy Juries). Most of the other bodies in the multi-body design (including Review Panels that craft final versions of bills to submit to the Policy Juries, Agenda Councils that decide what issues need new legislation, and Rules and Oversight Councils) would be selected using random stratified sampling. A large number of small Interest Panels would be open to any citizen who wanted to help develop proposals or arguments in response to the call from the Agenda Council. Each body would have terms of service, procedures, and size of membership suited to their particular task in the lawmaking process.