“Lottery Voting: A Thought Experiment” by Akhil Reed Amar

Lottery Voting: A Thought Experiment” is a 1995 article by Akhil Reed Amar, perhaps the most influential Constitutional Law theorist in the US.

In the article Amar proposes using the lottery to determine representation only after a standard election campaign has determined the percentage of support each candidate has received.

In this case then, the ‘statistical representation’ (which he champions) would be only of those citizens willing to go through the process of standing for election.

This is much less representative — and less attractive to me — than eliminating campaigning altogether… and using only two parameters to enter the lot: 1.) registration for it, indicating a willingness to serve if chosen; 2.) a simple civics test (such as the Naturalization Test in the U.S.).

35 Responses

  1. David,

    I don’t think Amar is championing statistical representation in the sense that kleroterians do (i.e. that the legislature should be a portrait in miniature of the citizen body). His concern is only that the legislature should better reflect its electoral preferences — i.e. as an alternative to proportional representation. At the end of the article he draws a contrast between his proposal and juries: “unlike the jury, we want leaders in the legislature rather than ordinary citizens, so that’s why we have voting“.

    Liked by 1 person

  2. *** A point is noteworthy in Common Lot Sortionist comment: the idea of a “simple civic test” necessary to entering a minipublic. If the minipublic is large enough to be entrusted to exercise popular sovereignty, any test of any kind is clearly contrary to the democratic principle of general franchise. The system becomes a « wide aristocracy », with the corresponding classical bad points – among others the non-neutrality of the test along race, ethnicity, gender, class … (we know very well the use of literacy tests for ethnic discrimination), and the dropping of the democratic principle in favor of the weaker principle of « moderate selection » – weaker, because of the unavoidable arbitrariness of the limits and the loss of the strong feeling of equality: such a moderate selection implies, for instance, that a citizen may be conscripted to war without having any right to decide about war.
    *** The Second Athenian Democracy, which gave strong political role to the big juries, legislative and judicial, did not establish any test to be member of such juries. There was some testing (dokimasia) but only for small executive juries, and for the Council, which had only an auxiliary role (if important). As far as we know, this testing could include only basic morality and involvement in antidemocratic undertakings; it did not include any idea of competence, even a minimal one. Dokimasia was useful to reduce the risks of randomness for small juries, and to exclude antidemocrat conspirators from the Council, the political one permanent body. Dokimasia was a very very moderate selection and even such selection was excluded for the legislative and judicial big juries which took the sovereign decisions.
    *** To conclude, as a minipublic exercising democratic sovereignty must be the perfect mirror of the citizen body, the percentage of ignorant people must be the same as in the citizen body – and likewise the percentage of supposed dumb people. Any citizen the society considers able to make major decisions in personal life (taking a job, maybe a dangerous one, entering a contract, marrying, engendering children, enlisting in army, emigrating …) is entitled in a democracy to participate in the collective sovereign decisions – whether in general votes or as member of a minipublic.

    Liked by 1 person

  3. I agree with Andre. As long as the size of the mini-public is adequate, we don’t care if EVERY member is competent, just whether the group as a whole is competent. But in addition, there are other benefits to NOT having any qualifying test…which is the diversity achieved. Even a person who is generally incompetent may have a crucial piece of knowledge to share that would be lost if a qualifying test were employed… Even just the presence of a member with Down’s Syndrome in a deliberative group may help remind the other members about the wisdom of structuring some law so that an inability to read at a 7th grade level won’t unfairly hurt such citizens… a thought that might not have occurred to members if that one member were not present, etc.


  4. “Any citizen the society considers able to make major decisions in personal life…”
    That statement obviously implies making a limit.
    I would not want automobile drivers to be on the road without a qualifying test of competence.
    The Naturalization Test is the equivalent of a driver’s test (at least in the U.S.)
    I am fully aware and concerned about the dangers inherent in tests. The literacy tests of the U.S. South being an apt case in point.
    I support whatever determinant is needed to assure the lottery is fully open to any and all ‘citizens that the society deems able to make major decisions in personal life’.


  5. I support the sentiments in the comments about exclusions.


  6. There would be no need for exclusions so long as the job of the (large, randomly selected) assembly is to listen to arguments and then vote, as the relatively small number of nut-jobs would be unlikely to affect the decision outcome. But David and Terry make reference to individual tasks (speech acts) where the issue of competence and civic virtue is of greater relevance. And there’s no good reason to believe that random selection is the best way (or even a moderately good way) of achieving Terry’s epistemic goals — note that Landemore, in particular, is emphatic that cognitive diversity is not the same thing as the wisdom of crowds. I can’t think of a single example of a large group of randomly-selected conscripts sitting around and brainstorming the answer to political problems for which they may have no particular experience/expertise. Of course then you end up with a small group of volunteers (as in all of Landemore’s examples) but this is not compatible with stochation. So I think we all need to be a little clearer about what sort of group we are advocating before setting out the selection criteria. (Amar, of course, would have no interest in any of this, as his proposal is for a way of selecting political leaders. He fully accepts the principle of distinction, but argues that lottery voting is a better way of achieving aggregate representation than PR.)


  7. Terry, I’m currently reading Dryzek and Niemeyer’s 2008 APSR article ‘Discursive Representation’ and came across this passage (commenting on Liebe, 2004) indicating why random selection is a poor way of constituting the cognitively-diverse input that epistemic democrats like you value:

    “The problem with random selection is that large numbers are needed to guard against the possibility that a relevant discourse might be missed. However, the larger the number of representatives, the harder it becomes for them to deliberate together. This is why large-scale processes such as deliberative opinion polls and citizens’ assemblies subdivide their participants into smaller deliberative groups of no more than 20 or so each. Thus, we need a procedure better than random chance to ensure that all discourses are effectively represented in each group.” (p.486)

    Dryzek’s Wittgensteinian perspective on discourses is unsympathetic to the collective-intererest assumption of descriptive representation (women, ethnic minorities etc).


  8. To Yoram Gat
    Thanks for indicating the former post about “exclusions”. I did not follow the blog at this time.


  9. *** I find interesting Common Lot Sortitionist’s analogy between civic test and driver’s license.
    *** In France at least (sorry) some rights, which I will name « general rights » (taking a job, maybe a dangerous one, entering a contract, entering a sexual relationship, marrying, engendering children, enlisting into army, emigrating …) are given automatically except to mentally disabled persons, children and adolescents – I will cover the three classes as « minors » in an extended sense. Some of these rights may be allowed to some kinds of minors by specific rules, but it is not automatic.
    *** The rights to drive a motorcar, to pilot a boat, or to pilot a plane are considered as belonging to a realm of technical rights, and the State checks corresponding abilities. (Some minorities have problems, as the colorblind minority, allowed since 2001 to pilot a private plane, but with restrictions, only at day and visual flight. The reason is our world is made by « normal » persons – if the color-blind were a majority, or even a big minority, there would not be night signals differing only by color.)
    *** Therefore we see clearly the difference between the democratic general franchise idea, where participating to the exercise of sovereignty is a « general right », and the « wide aristocracy » model corresponding to the idea of a « civic test », where the franchise is a « technical right ». I understand the reasons of the supporters of « wide aristocracy », but I think stronger the reasons for democracy.
    *** Note that if someone proposes to extend the franchise to 7 years children (the « age of reason »), even a perfect democrat as me is not compelled to approve it as « more democratic »; it would be a democratic requirement only in a society where these children would not be minors, and would be given all “general rights”.


  10. About lack of some points of view by luck.
    *** Let’s suppose a minipublic undertaking face-to-face discussion, it would be possible only by small fractions, and often a deliberating fraction will lack, by pure luck, any member of a class of persons whose points of view is interesting – for instance strongly concerned persons.
    *** The problem was lesser in an ancient Greek democracy, where many kinds of people might be encountered on the agora. But even there it did exist. And it was very serious when the concerned ones were foreign peoples.
    *** In a modern dêmokratia, the deliberation system could include a mandatory step with face-to-face deliberations between the minipublic exercising sovereignty and alloted samples of especially concerned classes. For instance let’s imagine a US minipublic studying the case of Guantanamo jail: the deliberation should include a step of face-to-face discussion (by telecommunication) between fractions of the minipublic and alloted samples of Guantanamo inmates (well, some parts of this samples could be fanatics with low epistemic value; but maybe some others would have some interesting things to say).
    *** But, clearly, that does not mean including these inmates in the minipublic as deciding body (except eventually as US citizen chosen by the lot).


  11. Andre

    Your Guantanamo case study is a good example of my general argument that the role of the deliberative assembly is to listen to diverse arguments (“discourses” in Dryzek’s terminology) and then vote. The status of the Guantanamo inmates is that of the expert witness or advocate in their own cause, not a participant in the minipublic, unless, as you say, they happened to be a US citizen chosen by lot. The decision to include the inmates has nothing to do with random selection, so is orthogonal to Terry’s case for cognitive diversity. Discursive democrats like Dryzek have little interest in sortition as their inclusion mechanisms are Q methodology questionnaires, interviews and other social science techniques.


  12. *** I agree with Bouricius that a benefit to NOT having any qualifying test (democracy instead of wide aristocracy) is the greater cognitive diversity achieved.
    *** This diversity will be nevertheless limited, because random sampling effects will lessen diversity in small discussion groups, because interesting voices may belong to a very small minority, or because interesting voices may be external to the citizen body – as in my Guantanamo example, or in foreign policy). We must add “orthogonal” procedures, as the one I proposed (with practical difficulties to get voices from foreign peoples).
    *** But we must not look down to the minimal diversity due to democratic lot. Especially by contrast to many discussions in today elite circles, where we can see often tendency either towards simplistic binarity or towards homogeneity. At least in France I encountered often more diversity in discussions in plant canteens than in the elite newspapers. The lack of idea diversity in elites may be explained: the elites wish to rule, but in modern societies they cannot hope to be obeyed from reverence, to be true aristocratic elites. They must rule by ideological control of the masses, which implies some level of ideological unity in a given elite. Democratic minipublics will allow escaping this drift towards “one thought”.


  13. *** Keith Sutherland writes that Dryzek’s perspective is unsympathetic to descriptive representation. I will say more widely that Dryzek’s thought is unsympathetic to dêmokratia. Hence Dryzek’s tendency to exclude as irrelevant to modern situation the idea of dêmos and his fuzziness about the idea of sovereign decision, kratos, and the corresponding legitimacy (the opponents of a modern dêmokratia, logically, are prone to exclude from the political thought the ideas of dêmos and kratos). Dryzek, as many others, should present themselves as proponents of “deliberative commonwealth”, not “deliberative democracy”.
    *** In a dêmokratia there are two kinds of basic political materials: the speeches of “orators”, articulate, with precise personal responsibility, and the elaborations of ordinary citizens through face-to-face discussions; they are the material allowing for an enlightened choice (by general vote or by a minipublic). The “discourses” in Dryzek thought are the “ideological discourses” of polyarchy, i-e fuzzy sets of ideas elaborated by the factions, the lobbies and the elites, with few place either for precise responsible personal speeches or for creative discussion by common citizens.
    *** The proposals by Dryzek are internal to the polyarchic model and aim to heighten, in this system, the influence of the culture elite, especially able to produce the “discourses” he wants to be better “represented”.


  14. Andre,

    That’s very interesting and confirms my suspicions. I gave a talk at a PSA conference in which I argued that deliberative democracy was an oxymoron: John Dryzek was in the audience and I could see the steam coming out of his ears. I agree that there is no such thing as discursive democracy, but would extend the argument to include most deliberative democrats, not just Dryzek’s clique. The Oxford Political Theory handbook claims that deliberative democracy is the child of the Frankfurt School (of cultural Marxism) so I guess that accords with your perspective on the lack of interest of ultra-leftists in sortition-based democracy. (Most people on this forum would claim that deliberative assemblies convened randomly are democratic, but I would argue that ongoing representativity is compromised by full-mandate deliberation for the reason given by Iris Young — the deliberations would be dominated by random loud-mouths who nobody elected, or the group would be so small as to no longer be statistically representative.)

    >But we must not look down to the minimal diversity due to democratic lot.

    I think we need to be more ambitious than to be satisfied with minimal diversity. There are far better ways to establish this than just drawing straws. Selection by lot establishes the (aggregate) wisdom of crowds, but is a very poor way to establish the cognitive diversity necessary for active problem-solving. These are two entirely different functions that shouldn’t be conflated.


  15. Keith,

    Random selection is a good method for achieving both cognitive diversity and demographic diversity… simply because no other method and no expert can know in advance exactly WHAT diverse traits may prove beneficial to the group decision. While obviously if the group is too small, diversity will suffer. For larger groups it would be necessary to limit duration and frequency of speakers, and rotate (and call on those who haven’t spoken), etc. to achieve better equality of participation than you fear (loud mouths dominating deliberation). A large group can still deliberate by listening to witnesses, asking questions and breaking into smaller groups that re-mix periodically for more discussion, etc. This allows for good problem-solving (proposal review, or even development). But like you , I think any final adoption should be by a separate random group that did not craft the proposal. Many psychological factors, even the simple “pride of authorship,” can weaken the skepticism necessary for good assessment.


  16. Terry:

    no other method [than random selection] and no expert can know in advance exactly WHAT diverse traits may prove beneficial to the group decision.

    That’s certainly true in the case of general-purpose standing assemblies. As you know my own preference is to model more closely 5th century Athenian practice, which limited random selection to ad hoc juries. Any citizen who wished (not just the tiny number selected by lot) could speak or make a proposal and the challenge is to find a modern (representative) analogue for this. Crowd sourcing, information markets, election, votation and direct-democratic initiative are much better ways of establishing cognitive diversity, while still adhering to the principle of (representative) ho boulomenos. Why would you want to limit this to a tiny number of persons conscripted by lot and does it not trouble you that there are no precedents, either historical or modern, for your proposal?


  17. Keith,

    Recall that my ideal design proposes that the function of ho boulomenos (ANY citizen may propose) is enshrined in what I termed “interest panels” which could be formed by random selection of volunteers or probably more often by those with special interest or expertise. I like the idea of a wide net for proposals, but then a representative and fully deliberative body for winnowing them down to a final piece of legislation and a mute jury for final adoption or rejection. Like you, I don’t think a single all-purpose legislative body is ideal…and that applies both to one selected by lot and ALSO to one selected by election. But once an elected legislature HAS a monopoly on legislative functions it is hard to get them to give away certain functions to other representative bodies.


  18. Terry,
    I’m concerned that a winnowing process would have the exact opposite effect. The net may be cast wide, but if you have a thousand redundant (and entirely reasonable) proposals for a given problem it will probably be impractical to give them all a fair hearing before a statistically significant assembly, to say nothing of the unreasonable proposals. And even if you could give them all a fair hearing there would be no way to choose between the reasonable proposals by rational means alone. All things being equal, people are more comfortable with proposals that don’t challenge their beliefs/prejudices/ideological leanings. Proposals that don’t fit comfortably within the beliefs/prejudices/ideological leanings of the average person will be winnowed away very early on.

    Whether the winnowing process is for candidates (as Jon has proposed) or for proposals I believe it constitutes one of the most extreme forms of majoritarianism ever seriously considered.

    I believe it would be better to limit the number of people with proposal generating powers so that they can effectively coordinate with each other. If several people with policy initiation powers want to make the same basic proposal they will simply collaborate on the same bill as they do in legislatures now. The elimination of redundancy makes time for people with unusual perspectives to have their say.


  19. Terry,

    Examining your three stages in turn:

    >1) “interest panels” which could be formed by random selection of volunteers or probably more often by those with special interest or expertise.

    This is the ho boulomenos stage but, given that the winnowing down happens at the second stage, what is the point of the panel? (especially in the light of Naomi’s concerns regarding extreme majoritarianism). In my competing proposal this stage is open to literally anybody and is fully pluralist, in that no official panel is involved. Any citizen can put their hand up to make a suggestion, the only constraint being that to move on to the decision stage the proposal has to gain public support (a democratic principle that you reject).

    >2) a representative and fully deliberative body for winnowing them down

    As you know I consider the combination of (statistical) representativity and full deliberation to be oxymoronic, and would once again ask you for some historical precedent (our knowledge of the internal workings of the allotted secretariat to the Athenian assembly is so scant that this would not constitute historical precedent). Fully-deliberative groups typically involve 12-20 people and are, consequently, entirely unrepresentative. You are trying to have your cake and eat it.

    >3) a mute jury for final adoption or rejection.

    Given the above two convoluted stages, with their faux-democratic mandate, an allotted jury would be under enormous pressure not to reject. Note also that all of the agonism (explicitly encouraged by the 5th century Athenian alternative) would have been, in Naomi’s words, “winnowed away” by the first two stages which lead to an (effective) consensus to be rubber stamped by the mute jury who will be happy not to have their beliefs/prejudices/ideological leanings challenged. My principal objection to epistemic “democracy” is the attempt to overcome agonism — I recall on an earlier occasion that you argued that the deliberative process is more a case of problem solving, the “political” decisions having taken place at an [unspecified] earlier stage. This is the antipathy of our (Greek-derived) notion of politics and is the reason that Nadia Urbinati’s last book includes the epistemic procedures that you lionise as the principal disfiguration of modern democratic practice.


  20. Keith and Naomi,

    1. The proposal is that ANY citizen who wished could join one of the interest panels on a particular issue and work with perhaps a dozen people to try and come up with a draft proposal. Many or most of these interest panels would probably fail to agree on a draft, but at least some would likely develop a draft to pass to the next stage (so not thousands).

    2. Keith wrote: “the only constraint being that to move on to the decision stage the proposal has to gain public support (a democratic principle that you reject).” Like Yoram, I consider mass election (or issue popularity contests through “votation”) to be fairly UN-democratic as it puts undue power in the hands of those who control the media or have wealth. Only smaller groups of average citizens given time and incentive to examine an issue closely can fairly reflect the informed judgement of a citizenry.

    3. Review panels would not be fully descriptively representative simply because the work load and complexity of the task would cause many citizens to decline to serve. However, they would be MORE representative than any existing legislature, making them a better body for deeply analyzing draft bills, combining and amending them (without the corrupting dynamics of partisan blaming, or re-election imperatives). In other words, these are not attempts at full blown “statistical representation” but rather an attempt at creating diverse groups of citizens without the systematic bias that is exacerbated by partisan elections. ONLY when people are given time and incentive to dig into issues deeply is there any possibility of moving from what Daniel Kahneman calls “system 1 thinking” (knee-jerk and superficial) to system 2 thinking (rational and deliberative).

    As for historical examples… the clearest are recent citizen assemblies, such as the British Columbia Citizens’ Assembly of 2004 or the Irish Constitutional Convention of 2012.

    4. As for agonistic process, this is what the final jury stage is all about, with sharply opposed pro and con presentations prepared by supporters and opponents of the final proposal generated by the review panel (which had operated more along the lines of “deliberative democrats” problem-solving model.) Both processes have great value and both should be employed…but they cannot be employed by the same group of people, because they are incompatible.

    Basically, I see a serious flaw with injecting electoral imperatives and partisan negativism (always suggesting the opposition party is motivated by evil or foolishness) into a democratic process. This animosity will spread through the population, poisoning the deliberative process as average citizens rely on party label heuristics, rather than their own independent judgement.


  21. Terry,

    >Only smaller groups of average citizens given time and incentive to examine an issue closely can fairly reflect the informed judgement of a citizenry.

    This is little more than a statement of faith or (in Yoram’s case) the result of a deductive syllogism. It relies on the notion of the “average citizen” (as distinct from the member of an elite), whereas the concept of average that I rely on is a statistical one, and this doesn’t apply to the sort of small deliberative group that you are proposing. The alternative I propose is subject to the partisan distortions that you describe, but in a large-scale pluralistic polyarchy these influences can be ameliorated, especially given a wide variety of mechanisms and sources for isegoria. But your proposal to completely deny the general public any say in the decision process would be viewed as undemocratic by anyone other than diehard sortitionists.

    >sharply opposed pro and con presentations prepared by supporters and opponents

    And where would the opponents come from? Clearly not the review panels (given their deliberative problem-solving mandate). In the classical and modern electoral case agonism is built in but it looks like you would need to create some sort of token devil’s advocacy. Either you follow an epistemic or an agonistic approach, you really can’t have it both ways.


  22. Making participation in the drafting/winnowing process voluntary opens up other problems. Serving on a panel represents a considerable amount of work that will almost certainly be discarded in the end. So the people drawn to the process will be the most passionate and, by extension, the least persuadable. Such a process will also select in favor of those with time on their hands (read: retirees). In the US, the proposals to reach a final vote would probably look like they were written by the Tea Party. If later steps in the process are more representative, then the groundwork is laid for perpetual gridlock. A winnowing process is a very rigorous way of reducing a large number of proposals down to only those that meet the ideal preferences of those doing the selecting. Chances are good you’ll have problems if you switch the personnel selection process at the end.


  23. Let’s say 1% of people want to participate in one of these panels in a given year. Ignoring qualification issues we are looking at 3million people. If we assume 15 people to a panel then we have 200,000 panels. If we have a hundred issues to discuss in a year, then we have 2000 panels per issue. If 10% of panels produce something that moves on to the next step, then we have 200 proposals to manage per issue. Perhaps “thousand” was a bit excessive, but not terribly so. Certainly many issues will end up with a thousand reasonable proposals if participants can choose their topic.


  24. Agree, that’s why (in a democracy) the winnowing has to be by public votation. The (acknowledged) problems of elite influence, rational ignorance and partisanship are minor (and addressable) compared to the problems that Naomi has highlighted.


  25. Keith wrote of my multi-body sortition concept that it would
    >”completely deny the general public any say in the decision process”

    Under this design, all citizens who wished would be able to contribute at the proposal development stage of the public decision-making process, and all citizens would have an equal chance to participate in the review stage and at the final decision-making jury stage. Compare that to an elective scheme where only an elite few who have sufficient connections to win election play ANY meaningful role in public decision-making. Being allowed to cast one of millions of votes for or against a prescreened list of candidates (often just one or two candidates) does not constitute any meaningful role in public decision-making. And votation (mass election for choosing issues) suffers all of the failings (rational ignorance, elite domination, system one heuristics, etc.) that electing candidates suffers from.

    Your math only deals with one level of government…Ideally, this model would be employed at the neighborhood, municipal, state and national levels, so there would be FAR more issues to distribute willing citizens among for interest panels. Certainly far more citizens would volunteer for a panel dealing with military spending, than whether to expand the local water district, but the numbers are completely reasonable, with “hotter” issues generating more draft proposals.


  26. Terry,

    I accept that in principle ho boulomenos would apply to the proposal stage of your model, but I share Naomi’s scepticism regarding the practicality of the numbers involved. But the biggest problem is the total exclusion of the general public from the winnowing-down process. Notwithstanding the problems of election/votation that you outline, the overwhelming majority of political scientists agree that, in a well-functioning electoral system, political parties are responsive to public preferences. Our earlier debate over the Gilens article shows that you disagree, but I think this is largely because the US (where you have honed your own views) could not be described as a well-functioning electoral system. Given the potential for representative isonomia of election/votation, the flaws need to be ameliorated, rather than just rejecting the isonomic principle tout court. The trouble with epistemic theorists is that they don’t really like democracy, as it means risking bad decision outcomes — that’s why Urbinati views this perspective as a disfiguration of demokratia. A better term for the sort of epistemic “democracy” that you champion would be plebian Platonism.

    5th century democracy was a three-legged stool (ho boulomenos, all-inclusive isonomia and allotted nomothesia), and stools with only two legs are likely to collapse, in particular when the leg that you have chosen to amputate is the only element of demokratia that we have retained in modernity. Do you really believe that a system of pure multi-body sortition (in which the vast majority of citizens are disenfranchised) has a chance of being taken seriously, or is your project purely of a utopian nature?


  27. Yes, exclusively sortition-based democracy is Utopian…I expect elections will persist in some form for generations to come… but my goal is to help democrats understand that elections are one of the least democratic part of the system, and sully democracy. As many sub-elements of public decision-making that can be removed from elections and moved to sortition the better. As to your repeated assertion that democracy through sortition in my design means the “total exclusion of the general public from the winnowing-down process,” I counter by saying this instead brings the general public INTO the winnowing down process in an informed way for the first time. Selecting among proposals for a jury to choose is something that any person who thinks about it for a few minutes will agree SHOULD be done with careful thought, facts, and without consistent bias. Elections and votation can NEVER achieve that. Only a smaller group that has the opportunity, time, incentive, and information (a representative group) can exercise INFORMED consent. Letting ALL citizens have an equal chance is the ONLY democratic way to achieve this since elections (and votation) are inevitably unduly influenced, if not outright controlled by elites with a bias.


  28. Terry

    Most citizens who currently have a vote (albeit one possessed of miniscule causal power) would argue that removal of that hard-fought-for right amounts to disenfranchisement — after all that’s what the word means. The Athenian democrats would have viewed such a move as an act of treason, as all citizens had a vote on changing the laws in both the 4th and the 5th century.

    In your modern proposal, the general public are not involved in the winnowing-down process, as this is performed by a tiny clique of self-selected persons. Your claim that these persons are “representative” is either a) an act of faith or b) the claim that there is such a creature as an “average” citizen (as opposed to a member of an elite). As Naomi pointed out, there is a good likelihood that such self-selecting persons would generate a portrait in miniature of the Tea Party. All your other arguments (“informed”, “careful thought”, “facts”, “without consistent bias”, “information”, “INFORMED”) are epistemic and have nothing to do with democracy per se (which may well lead to dreadful decision outcomes). Andre has argued that Dryzek and his friends should refer to their programme as discursive commonwealth (rather than discursive democracy) and I suggest you follow a similar path and adopt a term such as that suggested above (plebian Platonism), in order to ensure that we no longer talk past each other.

    The trouble is, as Quentin Skinner pointed out in a 1973 article in the opening volume of Political Theory, “democracy” is both a descriptive and evaluative term — a hurrah word often used more for its “commendary force”. Skinner has had the honesty to admit that he is a neo-republican, not a democrat, and I think you should follow his example. This would also put you in the excellent company of John Burnheim and Helene Landemore, who have also acknowledged that they are not democrats as their concerns are primarily epistemic.


  29. Keith,
    While I broadly agree with your points I’m concerned your tone might be unnecessarily confrontational. Please forgive me for saying so.

    I find it difficult to believe that the panel members would be inclined to set aside their own biases/ideological leanings/misc. baggage for the sake of quality deliberation any more easily than the participants on this blog. But I could be wrong. I’m not very familiar with the deliberative democracy literature. I can see how in the absence of systematic bias one might expect policies agreeable to most citizens to float to the top. What Keith and I are arguing is that the baggage that the citizens bring with them into a panel is potentially just as harmful as a systematic bias. Perhaps harmful is the wrong word. In any case both aspects need to be addressed.

    Elections are a specialized tool that is currently being used in a general capacity. If you take the power to choose out of the hands of elected officials a lot of problems go away. Elections are a very lossy mechanism. Relying on them for a task that requires absolutely precise representation (such as deciding the balance of power in a conventional elected assembly) is probably not very democratic. It’s hard to see how they are less democratic than any other tool where imprecision makes no difference at all. In order to thwart the democratic will in a system where elected officials propose and an allotted body chooses you’d need *unanimity* from a very ideologically diverse group of people who in turn must retain a measure of popular support to remain in office. Even in the US (with our anachronistic electoral system) there are a few members of Congress I like. Can you say there is not a single member of either the House or the Senate of whom you genuinely support? What’s more, by retaining elections in the system we retain mass participation. I’m inclined to believe this has intrinsic value, if only in encouraging political awareness.


  30. Naomi,

    Yes, I am confrontational, but I don’t agree that this is unnecessary. The trouble is many entirely different (incompatible?) perspectives are subsumed under the banner of sortition and deliberative democracy, so it’s important to make firm distinctions. And when extreme one-trick-pony proposals are made for the design of the political institutions under which we live then this needs to be exposed as utopian. My firm view is that the arguments coming from Terry and Yoram are extremely dangerous and need to be confronted as such. Fortunately nobody in the real world is going to take them seriously, but then the whole sortition movement becomes tarred with the same brush, and we’re dismissed as a bunch of cranks. As most of the work that I’ve been doing over the last 10 years has been on sortition, this gets me very upset, hence the confrontational nature of my language. If sortition isn’t your day job then I guess you can take it more in your stride.

    >What Keith and I are arguing is that the baggage that the citizens bring with them into a panel is potentially just as harmful as a systematic bias.

    So long as the selection process is truly random (i.e. quasi-compulsory attendance) and so long as the panel is large (c. 1,000) and limited to voting only, then I have no problem with any kind of baggage. This is because I’m more of a democrat than a Platonist.

    >In order to thwart the democratic will in a system where elected officials propose and an allotted body chooses you’d need *unanimity* from a very ideologically diverse group of people who in turn must retain a measure of popular support to remain in office.

    Yes, this is the essence of the Harringtonian argument, in which there is a cybernetic relationship between proposing and disposing. Of course the unanimity of beliefs/interests of the political class (ditto with the “masses”) is the foundation myth of the archaic political anthropology that underlies the views of those who are not prepared in principle to have any truck with election.


  31. Naomi,

    While I am not a pure “deliberative democracy” guy, there is merit to some key points about their criticism of our NON-deliberative so-called “democratic” system. Deliberation is different than mere bargaining, negotiation, or compromise, which are the MOST that exist in a representative body selected by competitive elections. Of course, mostly what passes for deliberation in an elected chamber is only showmanship to try and frame the public perception of the evil of the opposing side for the next election. But they do occasionally actually negotiate from their relative power positions. However… they NEVER deliberate, which means give and honestly receive arguments intended to get at underlying truths, needs and likely outcomes with a hope of finding mutually agreeable or even win-win policies. Since we never see actual deliberation in the political process based on competitive elections, people often can’t quite imagine it. We only get to see it in mini-publics and similar citizen assemblies. For DELIBERATION, which is a sort of problem-solving activity (as opposed to grand-standing or debating to WIN over the opponents), diversity has been shown to be extremely valuable. But deliberation has severe weaknesses as well and can sometimes mask real and deep conflicts of interests. That is why I agree that once the problem-solving effort has been made (which is virtually absent in modern electoral democracies), we still need pro and con presentations for any policy before a descriptively representative jury for a final decision. Simply because of the workload, duration and the nature of the work, it is likely that those who volunteer to work on an interest panel or who accept an appointment to an review panel will deviate from accurate representation (though being dramatically more representative than any existing legislature), so while these bodies provide optimal proposal development and refinement, in the interests of democracy they are not suited to making final decisions…only a jury is.


  32. Terry:

    >those who volunteer to work on an interest panel or who accept an appointment to an review panel will deviate from accurate representation (though being dramatically more representative than any existing legislature)

    “Dramatically” assumes that there is no truth at all in the claim that in a well-functioning electoral system politicians will be responsive to the preferences of those who elected them. The representatives may have little in common with their constituents, but that doesn’t mean that they can’t represent their interests, preferences and beliefs.


  33. Terry,
    The trouble is you need to fill the panels with people of the right mindset to get proper deliberation. Maybe a truly random sampling of citizens would work for most panels. I don’t know. But low turnout favors the extremes in elections and I see no reason to expect any different here. Volunteers will be more passionate about their issue, start with a more ridgid idea of the outcome they wish to obtain, and be less interested in listening to reason than a random sampling.

    How does deliberation drive quality policymaking? Is it the casual brainstorming process? Someone makes a suggestion, and then someone else makes another and so on and these suggestions either get a warm reception from the other members or they fizzle out. Many who make a suggestion become stubbornly attached to their suggestion. And of course not all members of a deliberative panel offer alternatives and engaging in discussion. Enough to decide the issue listen and judge.

    It might be worthwhile to see if we can get the same benefits of active deliberation in a format that preserves statistical representivity. The system I favor has the two major deliberative roles crystallized out. If the terms of the voting members are fixed, rather than ad-hoc, bills can be built-up piecemeal. There is room to consider every potentially objectionable detail. For an aspect of an initial proposal to survive it must be favored above all the amendments that would be introduced by the ideologically diverse set of advocates. These advocates would no doubt be looking for reasonable alternatives to any prevailing policy points which would leave their constituents at a disadvantage. If the number of advocates is fairly large, their unwillingness to reevaluate their positions and let go of their proposals is not important. Someone else will step in.

    So wherein lies the difference? It’s social in nature. The individual voting members will not get to know each other. They can’t be expected sympathize with each other and favor win-win alternatives to the same extent. That said, they should still sympathize with the speakers (advocates and their guests). After all, in a conventional deliberative setting, a member is not likely to sympathize with another member who does not speak either. Another effect is that the members would be much less likely to become emotionally invested in one particular option, preventing the breakdown in active deliberation that tends to follow.

    I suppose another difference is that isolated members who have a problem with the course of the deliberative process would be unable to speak out. There’s less room to make all the voting members happy. To speak out they need representation among the advocates. However, the same relationship between the advocates and the voting members also exists between the assembly/panel and the population as a whole. To have a voice in an allotted assembly or panel one must be drawn into it. If a subset of the population fails to find representation due to the luck of the draw it is entirely irrelevant to them if the assembly/panel found something they could agree to by broad consensus. As I’m sure Keith would point out, being drawn in is not necessarily enough. They would actually have to speak out rather than fearfully pouting in the corner. I’ve been there before. I imagine most people have.


  34. One of the problems with the notion of “all affected interests”, privileged by deliberative democrats, is that such (self-nominating) people are unlikely to deliberate in an open-minded way, for the reasons that Naomi has outlined. This is why Madison’s ideal legislator was disinterested (i.e. a professional, rather than a a businessman or a landowner). Nowadays we’re rightly sceptical about the existence of such godlike creatures, hence the attempt to arrive at a dispassionate conclusion via the democratic mechanism of the aggregate wisdom or crowds. But volunteering is anathema to any kind of disinterested outcome, hence the need to ensure that policy proposers are clearly labelled as advocates and that the partisan nature of their contribution be modulated by the need to gain prior public approval. Partisanship needs to be quarantined, it can’t be abolished by fiat.


  35. Naomi,

    Since interest panels are open to all who wish (an important principle of democracy—- that anyone who wishes can play a part) and some might be self-organized by single-issue advocates (while others are cobbled together from general interest civics minded volunteers like League of Women Voter types), many of these panels will NOT use good deliberative processes and end up generating extremist proposals. But the key is that they have no decision-making power, and each interest panel is only generating one of a large number of draft proposals that go to the next level … a randomly selected deliberative body with professional staff. But the interest panels, KNOWING they are only generating draft proposals, have an interest in mediating and balancing their proposal because extremist proposals will likely be thrown out at the next review level. The desired genuine deliberation occurs primarily at this review panel stage…taking pieces from the various drafts, amending them and cobbling together a final draft in search of win-win possibilities. Because members of these Review Panels (focused on one issue area) might serve for several years and having an intellectually challenging task, many drawn citizens might decline to serve… so they will inevitably be less descriptively representative than a short duration semi-mandatory service final jury The jury, after hearing pro and con presentations on a final bill adopts or rejects it.


Leave a Reply

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

WordPress.com Logo

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

Facebook photo

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

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: