Government by Sortition

From "The Trouble With Elections: Everything We Thought We Knew About Democracy is Wrong," Chapter 12.4

The court jury is a useful reference that can lend legitimacy to sortition for people who have never heard of it. However, the similarities between existing jury systems and randomly selected policy-making bodies are few, and we should not hew too close to the current court jury model. As I discussed in chapter 6, on the history of juries, there would need to be fundamental differences in terms of selection methods, size, procedures, and functions. Policy panelists must be free to ask questions, and bring in opposing experts of their own choosing, rather than being manipulated by “experts” (lawyers) who are seeking a particular decision.

Sortition uniquely allows for the melding of deliberative and participatory democracy, so long as there are numerous allotted bodies at national, state and local levels. Some might be short-duration single issue policy juries, while others might review executive performance, set agendas, deliberate to generate draft proposals, oversee the lottery process, deliberation procedures, and assure staff impartiality, etc. Anyone would be able to directly participate in actual self-government at various times throughout their life. 

There are distinctly separate decision-making tasks of government that are inter-mixed in electoral regimes. First is the matter of deciding what societal problems need to be addressed, or what goals should be pursued. In many respects this is a subjective question, where experts need to be consulted, but expertise cannot be decisive. Experts within each domain tend to think that the issues they know about are more important than issues they don’t know about. Once well-informed, the people, through sortition, must set the agenda. Second, is the matter of how to accomplish the chosen goals. Which possible government policies are most likely to succeed while causing the least harm? Here again disagreeing experts need to be consulted. Again, this weighing or tradeoffs requires well-informed assessment by the community through sortition. Feeding into the “how” decision, there must also be a process of developing alternative proposals for consideration. It would not be wise to limit the generation of proposals to groups selected by lottery. The best option will likely be the brainchild of a person or group not selected by lottery. However, sortition bodies can hear from diverse advocates for and against each option. Third, there will be a whole set of implementation decisions – the classic executive function. Fourth and finally, there needs to be an evaluation process, which looks back on past performance, so adjustments can be made with the seeking of new proposals.

In standard elective governments, legislators take on all of these decision-making tasks (except the executive implementation function in non-parliamentary presidential systems). The experts consulted are generally like-minded partisans – the opposite of diverse. This is a very bad design, which was not originally intended for this use. As I discussed in chapter 6, all-purpose legislative chambers for lawmaking in an elected legislature, are a design passed down and re-purposed from the advisory chambers created to advise the monarchs in Europe. While it is a bad design for elected law-makers, it is an even worse design for a sortition-based system. The mere act of drafting a bill, as well as the act of advocating for or against the bill, makes those people unfit to then judge their own handiwork. Pride of authorship, information cascades, or group-think warps their ability to judge the final product up for adoption, or to evaluate its performance once passed. Simply substituting lottery selection for election in filling such an all-functions legislative body would be a mistake. I’ve written a journal article on this topic, entitled “Why Hybrid Bicameralism is Not Right for Sortition.” A sortition-based law making process needs to be designed from the ground up, rather than merely replacing elected legislators with legislators selected by lot. Various separate sortition bodies should take on distinct functions.

A crucially important fact about sortition, which is often ignored by deliberative democracy advocates, is that this tool can be used for a variety of tasks other than creating descriptively representative deliberative bodies. As Oliver Dowlen pointed out in his book The Political Potential of Sortition: A Study of the Random Selection of Citizens for Public Office, using its arational (not irrational) “blind break,” sortition is also useful for its impartiality and anti-corruption characteristics. The current jury system that selects only 12 jurors uses a sample that is too small to have adequate probability of creating accurately descriptive representation. But that is not its goal. Random selection in this case is seeking to select impartial citizens that are diverse and only roughly typical of the population.  A society might, for example, randomly appoint agency heads or judges from a pool of highly qualified people, to avoid partisan manipulation through political appointment. Indeed it is currently common to have judges randomly assigned to particular cases to avoid corruption. Not all panels in a sortition-based democracy would necessarily need to achieve, nor even seek, descriptive representation. The final decision-making policy juries should achieve this, however, in order to earn the needed public legitimacy.

There have been a variety of notions on how to incorporate democratic lotteries as a permanent part of modern democracies. At the end of this chapter I will briefly discuss my preferred multi-body sortition design. But first I want to challenge two common design concepts.

Most of the ideas about institutionalizing (or embedding) sortition into government have focused on the deliberative benefits of having diverse and representative bodies engaged in developing or winnowing policy options. But rather than having any real power, these citizens’ assemblies would serve as advisory bodies, submitting proposals for elected “decision-makers” to consider. This has been the basic scenario for the vast majority of citizens’ assemblies convened in recent decades. While helpful for establishing the legitimacy and workability of sortition, this model isn’t a desirable end-point. Merely advising elected decision-makers, maintains most of the failings of the electoral system discussed throughout this book. Adding a veneer of public participation through citizens’ assemblies may serve as a democratic fig leaf for the fundamentally un-democratic, oligarchic electoral elites, allowing politicians to pick and choose which recommendations to accept and which to ignore.¹

A more radical proposal involves simply changing the way legislators are selected – from election to lottery. As discussed above, a design that uses an all-purpose body (that sets agendas, drafts bills, advocates for them, and then adopts them) is terrible. It activates many of the problematic cognitive biases, and becomes a focal point for ready corruption. We want to avoid having a single all-powerful chamber, regardless of how the members are selected. An all-purpose sortition legislature might easily decide that they need more power, longer terms in office, and more money, etc. It is my strong belief that any decision proposed by one allotted body needs to be approved by a different allotted body – as a check and balance. As I describe in my multi-body sortition paper, a separate Rules Council (selected by lot of course) could propose procedures for all other allotted bodies to use,  but those proposed revisions would need to be ratified by an independent sortition jury.

There is no “one size fits all” design that is suitable for all sortition bodies. Different tasks and functions require different procedures and designs in terms of size, duration, deliberative procedures, etc. It is also important to recognize that sortition need not be limited to policy and deliberative tasks (though that has been the focus of the vast majority of citizens’ assemblies convened in recent decades). A representative sample of citizens could be empaneled periodically to assess the performance of department heads or chief executives. These panels might have the authority to remove poor performing executives and initiate the formation of a randomly selected recruitment and hiring panel to find a replacement. Sortition could also be used in a similar manner to select board members or trustees of NGOs, unions, or co-ops, where elections are also often extremely problematic (this will be discussed in a subsequent chapter).

The next post will set out the design, and briefly describe the bodies ideally employed in multi-body sortition.

1. Some implementation groups, such as the newDemocracy Foundation in Australia, require governmental sponsors of citizens’ assemblies they run to make a commitment to take action on the sortition-generated proposals, before they will organize such an assembly.