Add a pinch of sortition and stir…

I’ve finally had the chance to read Filimon Peonids’ interesting paper “Bringing Direct Democracy in a Representative Assembly: The Case of Allotted MPs.” The paper, as I mentioned before, can be found here–

The paper’s novel argument is that people should have the chance to cast a vote for “select somebody at random” when they elect their legislatures. After each election is held, a number of legislators should be randomly drawn that is proportionate to the number of citizens voting for this option. (Obviously, this solution requires some form of proportional representation.)

Two things struck me while reading the paper. First, the author has a somewhat unusual understanding of “popular sovereignty.” As far as I could tell, the author is completely unconcerned about the fact that legislatures don’t “look like” the people they serve. The fact that sortition ensures the presence of a diverse cross section of people in a legislature is therefore not a reason for liking it. What bothers the author is the fact that ordinary people have effectively zero chance of getting into the legislature. If popular sovereignty means that everyone who wants to participate in government has a chance of doing so, then indeed sortition can help. Of course, in the United States sortition would only give the average person one chance in several million of being selected for a randomly-selected seat in Congress, which doesn’t sound much better than the chance that a real-life “Joe the Plumber” can get elected to Congress. (It’s rather like the fact that by not buying a lottery ticket, my chance of winning is only slightly smaller than my chance of winning if I buy one.) But this argument is less applicable in a smaller country like Greece, where there will be many fewer potential candidates in the pool.

Second, the author seems to have the all-too-familiar fear of popular participation. If voters were sufficiently disenchanted as to elect a randomly-selected majority, then said majority might run amuck and do nutty or evil or fascist things or something. Peonidis seems convinced that this is more likely to happen with a randomly-selected legislature than with an elected one. I for one am skeptical that this is anything but a prejudice. Governments get elected with “mandates” all the time, and they manage to do astonishingly stupid things that betray the constituencies that elected them. (See also: Obama, Barack.) Would a randomly-selected majority do worse? Maybe, but it’s far from an obvious point.

In any event, some version of this danger will always be present so long as the randomly-selected legislators hold the balance of power in the legislature. (And if any party has an effective majority without randomly-selected legislators, expect the former to ignore the latter completely.) But if the inexperience or ignorance of randomly-selected legislators is really a concern, there are ways around it. Why not select a new crop of randomly-selected legislators for each major piece of legislation? Say that X% of the people vote to have candidates selected by sortition. Then before the legislature takes up the topic of health care, a number of legislators equal to X% of the legislature gets selected. Give them the kind of resources that Jim Fishkin/Ned Crosby/Peter Dienel give their randomly selected bodies, and then let them join the legislature as voting members, but just for that one bill. Then when the legislature considers the fight against global warming, repeat the process all over again by filling X% of the legislature with an entirely new random draw. Surely this could provide the advantages of random selection while assuaging the fears of Peonidis. And it would increase the chance that any volunteer would get randomly selected at some point, which would surely advance Peonidis’ conception of popular sovereignty.

As you can see, I don’t really agree with the paper’s conclusion. But it’s an awfully good effort at stimulating further thought–even among people who who are already interested in sortition.


5 Responses

  1. Peter,

    > After each election is held, a number of legislators should be randomly drawn that is proportionate to the number of citizens voting for this option.

    What do you think about the idea itself, apart from the ideological stance that accompanies it? As I wrote in my comment to your post linking to the paper, I think it is attractive as a way to legitimate sortition in the current situation when it is likely to meet some resistance due to unfamiliarity with the idea. It would pose some problems as well.

    > the author has a somewhat unusual understanding of “popular sovereignty.”

    I think that, in fact, most people do not see “popular sovereignty” as implying a representation that “looks like” the citizenry. To the extent that a representation that “looks like” the citizenry is valued this is because it is seen as having a better chance of representing the interests of the citizens.

    Some exceptions to this rule (that statistical representation implies representation of interests) are widely accepted. For example, most people, I think, would rather not have the “incompetent” sector of the population represented in parliament. This may account for people’s willingness to put up with “Big Men” (such as those that sit in elected houses) – they are supposed to be highly competent. Removing children from the electorate is also justified by arguing for representation of interests: children’s best interests are (supposedly) best represented by adults, not by children.


  2. Re: Popular sovereignty. Yeah, this concept is a really fuzzy one. I actually think it’s the most interesting topic that proponents of sortition need to consider. The question is, just what the heck is the relationship between “rule by the people” and rule by a random sample of the people? On one level, it makes intuitive sense, but at another level, electoral democracy–where the people as a whole get to elect their representatives–seems more consistent with popular sovereignty. Complicating matters further is the fact that we don’t want to count literally EVERYONE (children, Charles Manson) as a part of the “people” who should rule, either directly or indirectly. I’d really love to see the Kleroterians produce a clear opinion on how we should think about this.


  3. Five years later, I take note (only now having run across this entry) …

    1.) I haven’t seen any further discussion on this blog of the ballot choice: ‘select a representative by random selection’. I suppose it gains no interest because implementation seems highly improbable?

    2.) Not having read every blog post, I have the impression that discussion is deadlocked — or dead? — regarding WHO should be in the pool for allotment. I and others suggest that the pool be only of ‘the willing and the able’. That is: those who volunteer and who pass a basic civics test (for the U.S., the Naturalization Test for immigrants seeking citizenship).

    3.) I went to Wikipedia to confirm my understanding that the idea of juries for one bill at a time … is demarchy. To my surprise I found that the ‘demarchy’ entry had been replaced by a stub indicating that ‘demarchy’ had been moved into the page on ‘sortition’. I’m wondering if everyone on this blog would agree with that conflation?

    4.) That Wikipedia search leads to something new for me … John Burnheim’s suggestion that the jury also be responsible for executing its decision.
    I’m surprised because I’ve been locked into thinking separation-of-powers. Namely, sortitional selection for the policy-deciding legislature and meritocratic selection for the implementing executive.
    Here is Burnheim’s quote cited in Wikipedia:
    “Let the convention for deciding what is our common will be that we will accept the decision of a group of people who are well informed about the question, well-motivated to find as good a solution as possible and representative of our range of interests simply because they are statistically representative of us as a group. If this group is then responsible for carrying out what it decides, the problem of control of the execution process largely vanishes.”
    That reminds me of the Athenians who used sortition — with the exception of the military and the fiduciary — for both executive and legislative functions.
    I wonder how much support there is for this idea? Has it been implemented anywhere — a jury implementing its decision?


  4. 1) We briefly talked about the possibility of linking a none-of-the-above option to random appointment. So if 12% of people vote none-of-the-above 12% of seats would be assigned to citizens drawn randomly. In a conventional legislature it probably wouldn’t be a good idea, because the size of the sampling probably wouldn’t be statistically significant and the potential for bribery would be extreme. The balance of power between parties could easily end up being random, rather than reflecting popular support. I would favor the option in elections for advocates.

    2) I guess it depends on what you want. If you would like to use the lot to prepare a statistical sampling of the population, then limiting it to those who want to serve would make the sampling representative of that subset of the people only, not the whole population. I accept there are certain practical issues. Infants matter, but we can’t exactly draft them into the legislature, for example. I would argue that these practical limitations should decide who is a part of the lottery pool. At least for voting offices, that is. There’s a case to be made for a more limited pool (and perhaps volunteerism) in the allotment of specialized offices. If we had a citizens’ assembly to prepare a list of candidates for a supreme court appointment, it might make sense to have a partially non-representative stratification system. You might want to ensure a certain percentage are constitutional law scholars, for example.

    3) It was my understanding that demarchy is the use of many small allotted committees in parallel, rather than ad-hoc samplings per se. In any case there are many untried sortition proposals. Whether this one merits having it’s own page is not something I have any business holding an opinion on.

    4) Practically, how would this even work? You need a substantial civil service to administer a state and different subdivisions of the civil service need to coordinate their activities. There’s the danger of conflict between the different samplings as they attempt to execute conflicting policies in parallel. Just because all the samplings will be statistically significant does not mean they will think the same way or have the same priorities, given that they will be tasked with solving different problems. We need to decompose separation-of-powers into two components: separation-of-personnel and separation-of-purpose. Separation-of-personnel is not a bad thing as far as I’m concerned. Separation-of-purpose is indicative of a serious design flaw and must be avoided at all cost. One strategy to is to drop separation-of-personnel as well, as in a conventional parliamentary system. But this is probably not necessary. As long as the process of appointing and dismissing executive officers is congruent with the process of making laws and raising and expending funds there should be no problem. If you want the laws to be made by a highly diffuse network of committees acting in parallel… you’ve got a lot of problems that go way beyond separation-of-purpose between the executive and legislative processes. I’m of the belief that centralization and modest fixed terms are both absolutely necessary.


  5. David

    That’s certainly not John Burnheim’s current view. He has a new book, The Demarchy Manifesto due to be published with us (Imprint Academic) early next year. He still advocates parallel ad-hoc demarchic councils that are staffed by a sample of people drawn from “all affected interests” (i.e. self-nominating persons). The only reason he still advocates sortition is in order to winnow down the numbers in an impartial way (his ideal council would only have a dozen or so members) and he no longer has any interest in statistical representation. In fact he strongly opposes it, as he thinks it will re-introduce power relations by the back door, and his concerns are purely Habermasian — i.e. the forceless force of the better argument. He certainly does not propose that demarchic councils should have any executive function, in fact he locates them squarely in the field of the informal public sphere. The only way that their decisions would have any effect would be indirectly — by helping to shape and inform public opinion, politicians will be encouraged to follow the decisions of the councils as they will perceive this to be electorally advantageous. John does hope that in the long run the middle man (i.e. the political state) will wither away, leaving the demarchic councils in effect in the driving seat, but the “withering away” is probably just the residual Marxist/anarchist in him.

    I tried to persuade John that the back-cover blurb should refer to him as the founding father of the modern sortition movement but if anything he wants to distance himself from sortinistas as he thinks we are mostly barking mad! Perhaps it’s all the mud-slinging he’s witnessed on this site.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: