Edmund Griffiths  

E d m u n d   G r i f f i t h s

Rough notes towards The Modern Discorsi

Seven potential problems with sortition

I take it that the advantages of sortition—the system whereby members of representative bodies are picked at random, like jurors—are well known; but, in brief, they are these:

      • it is likelier than any other system to produce representative bodies that are sociologically representative of the people;

      • it removes the need for any specific positive discrimination;

      • it forces political parties, campaign groups, etc., to address themselves to the public as a whole if they want to have any consistent influence on policy;

      • it transforms political representation into a genuine public service, carried out by people who would often not have chosen it: a matter of duty, not ambition;

      • it ensures that representative bodies can draw on a very wide range of experience;

      • it prevents representatives from believing that they are better, wiser, or more deserving than the people they represent;

      • it makes representatives largely immune to pressure from party machines, lobbyists, etc.: re-election is no longer in anyone’s gift, so the threat to withhold support becomes empty;

      • it achieves full proportional representation of parties based on their levels of support, including (unlike rival systems) independents, ‘floating voters’, and people who oppose all the parties;

      • it means failure to be re-elected can no longer be seen as an insult or a rejection;

      • it guarantees a regular turnover of representatives, without any need for formal term limits;

      • continuity of experience can easily be provided, if required, by having representative bodies picked (say) a third at a time;

      • local or regional representation can easily be provided, if required, by having representatives drawn at random from single-member or multi-member districts: democratic Athens, the state with which sortition is most associated, actually used multi-member districts (the ten demes);

      • since membership of representative bodies under sortition is a duty and a burden that the members did not seek, it tends naturally to go with short terms in office and frequent rotation;

      • it is not likely to produce representative bodies that are significantly less competent or knowledgeable than the general public.

These advantages are real, and convincing; but there are also a number of difficulties that advocates of sortition need to address. (There is no reason to think they are insoluble.)

1. The problem of trust
At the heart of sortition is the generation of random numbers: and verifying that a random number generator has been set up correctly and is working according to its specifications requires technical skills that most people presently lack. Even if the hardware and software were open to public scrutiny, only a small minority would be able to check them as effectively as most of us can check that ballot papers are being put onto the proper pile. The majority would be obliged to take the experts’ word for it—an unfortunate start for an experiment in radical democracy. And it seems quite likely that many people would end up refusing to do so. It is in the nature of random processes that they do not always look random, especially to observers who are untrained in statistics and probability. There are runs of results; there are clusters and disproportions. It would be terribly easy, under sortition, to point out oddities or anomalies, to insinuate that particular groups were over- or under-represented on purpose. ‘What, you really believe it’s random?’ It has been put to me, and it is certainly true, that people trust the National Lottery: but this is not as sanguine an omen for sortition as might be hoped. For one thing, the lottery is optional. People who mistrust it are free not to buy a ticket. But there is also no substantial group that could stand to gain from calling the lottery into doubt. Only a handful of people can win the big prize each time. Indeed, if the pleasure of playing the lottery is—as I suspect—derived to a great extent from the licence it gives to daydream about a prize one has no real anticipation of winning, then players are likely actively to resent any attempt to question the lottery’s bona fides. Under sortition, by contrast, many people might be open to believing they or some group with which they identified had a grievance; and there would be ample incentive to spread distrust on the part of anyone who, say, disapproved of randomized democracy and wanted to bring back the old system based on elections. There is thus a risk that sortition, even though it actually represented the people very faithfully, would not be accepted as legitimate.

2. The problem of fraud
The possibility cannot be excluded, however, that the random number generator really will include deliberate biases. Allowing people to examine the machine and/or the programming makes fraud harder, but cannot make it impossible—it merely creates a challenge for the talented and ingenious fraudster. And, unless the skewed results are blatantly skewed, suspicion is likely to be met only with reproaches along the lines suggested in the last paragraph: evidently you are not well-trained in statistics and probability, you don’t realize that there are always going to be anomalous little runs and disproportions, perhaps you want to undermine faith in our democracy so you can bring elections back, etc.

3. The problem of error
Deliberate sabotage is not, in point of fact, the only way in which complicated machines and programs can fail to work as expected. They can simply contain subtle mistakes, mistakes that sometimes only come to light after quite a long period of apparently correct functioning. The discovery of such a mistake might lead to a collapse of confidence in how sortition was implemented, especially if it followed some close and emotive votes where it might be thought that the results could have been different if recent sortitions had been conducted using a true random number generator rather than an inadvertently biased one.

4. The problem of population size
At Athens, where women, immigrants, and slaves were denied the political rights enjoyed by (free, male) citizens, the Council consisted of 500 men chosen by sortition from the demes: something like 2% of the total eligible population. Since the Council was picked every year, and no-one could serve on it more than twice in a lifetime, the average citizen had quite a high chance of being called upon to serve. Almost every citizen must have known at least one Council member, almost all of the time. Under circumstances like these, sortition feels a lot like straightforward rotation: you take your turn on the Council just as your friends and neighbours do. But in a modern state with a population of, say, twenty-five million adult citizens, the proportion serving on a 500-member Council would be not one in fifty but one in fifty thousand. It would not be unusual to go your whole life without anybody you knew ever being picked. Your chance of being picked yourself in any given year would be so small that it might feel negligible. It is hard to believe that sortition would continue to feel as collaborative and participatory as it probably did in the polis. On the contrary, there is a risk that the political process would come to seem something abstract and remote, without even the formal participation offered by an election. If that meant people stopped following politics and no longer tried to keep up with current events, the result is that the assemblies formed by sortition would consist of ignorant, apolitical, and uninformed ‘private individuals’ instead of active citizens.

5. The problem of advisers
A representative body formed using sortition would naturally want to take evidence about the bills that were before it, to hear submissions from experts and interested parties, to entertain different outside proposals for solving particular problems, etc. But who would have the right to be heard? It would hardly be reasonable to expect the representatives themselves to know the names of everyone they might want to hear from. The only part of our present constitution with an element based on sortition is a law court, and there the jury depends on its advisers—the barristers paid to advise it by the parties, and the judge paid to advise it by the state—to select the evidence that will be presented and the witnesses who will be called. One can imagine an analogous system growing up around a national assembly formed by sortition, with paid advisers and orators retained to steer the representatives one way or another. And one can imagine worse than that: the assembly, surrounded by paid lobbyists and paid advisers, might come to resemble a feeble king in an absolute monarchy, nominally the sovereign of the whole state but in fact easily manipulated by a clique of flatterers and hangers-on.

6. The problem of legislative initiative
Given an assembly whose members were mostly independent of any political party, who would decide which matters required legislation? Who would decide how much time to allocate to one or another bill? Who would decide which provisions to include in the first draft? Once again, it seems possible that the sortition-based assembly would find itself in the position of a jury—reduced to accepting or rejecting proposals drawn up elsewhere. This would still represent a substantial injection of democracy into the political system; but it would raise further questions about how the advisers, steering groups, drafting committees, and so forth were elected or appointed.

7. The problem of organizations with activists and supporters
So far, we have been considering sortition either in the state or in an organization that resembles the state in having a definite, fixed body of citizens / members. Many organizations, however, have both activists (who may be rather few) and also supporters: people who are willing to attend a demonstration but not to leaflet for it, or to vote for a party’s candidates but not to canvass for them, or even to pay a few pounds a month but not to attend any meetings. Organizations of this kind do sometimes make some attempt to involve their supporters in the decision-making process—by holding an open primary, say, or by having committee elections ratified at a public rally where large numbers of supporters are expected to be in attendance. Whether these electoral devices really are very democratic is a matter for another day. At best they do give the supporters some say; at worst they let the supporters feel they are being given a say. But it is not apparent how anything of the sort can be done under sortition. Of course everybody who attends a demonstration can be offered a ‘raffle ticket’ entering them into the draw to become a member of the committee that organized it: but, if they do not have the time or the inclination to become activists, they will refuse to take a ticket or else they will simply not do the work if they happen to be picked.

Tuesday, 14 June 2016