The Blind Break and the Invisible Hand, Part 1

Last week’s sortition workshop at Queen Mary, University of London was entitled Sortition and the Consolidation of Democracy. The host and academic convenor was Oliver Dowlen, who was a research fellow at QM, and most of the papers adopted the Dowlen-Stone ‘Lottery Thesis’ that the primary value of sortition is as a prophylactic to protect the political system from factionalism and corruption — a principle that Dowlen and Stone attribute to the arational ‘Blind Break’ introduced by the lottery mechanism. The main focus of my commentary is Peter Stone’s (draft) paper, ‘Democracy and Good Government’ (Stone, 2013), but my arguments are addressed to all three signatories of the ‘Dublin Declaration’ (Delannoi, Dowlen and Stone, 2013),* which, whilst acknowledging that sortition is also a way of instituting ‘descriptive’ representation, concluded that this was a ‘weak’ use of the lot. My claim in this commentary is that the ‘Blind Break’ is a) politically conservative (sortition is capable of protecting the integrity of any political system, not just democracy) and b) philosophically tautological (its ‘strength’ is entirely reliant on the pre-definition of the Lottery Principle).

Stone’s paper was based on Bo Rothstein’s comparative study of the quality of government (QoG) in modern states (Rothstein, 2011). QoG is measured in terms of economic, political and social outcomes (legitimacy, social capital and [non]corruption) and Rothstein finds that there is no necessary connection between democracy (an input function) and benign outcomes. One chapter makes a direct comparison between Singapore and Jamaica: both were former British colonies and equally underdeveloped, but Singapore (an authoritarian state) now ranks considerably higher than Jamaica (a democracy) in all the QoG factors. Rothstein’s book, or at least Stone’s portrayal of it, does not attribute an inherent value to democracy (other than arithmetical equality and Amartya Sen’s argument for institutional safeguards) and, in terms of measurable outcomes, ranks it lower than authoritarian alternatives. So what value could sortition bring to the table? Clearly, given Rothstein’s emphasis on impartiality in the exercise of power and the lottery’s ability to select public officials without fear or favour, this is a vindication of the argument for the Blind Break.

Although it might be unfair to use this one paper as a criticism of the whole workshop this led me to conclude that Sortition and the Consolidation of Democracy was something of a misnomer — it might have been better to call it Sortition and the Consolidation of Good Governance or something similar. The arationality of the blind break is capable of enhancing impartiality in any political system — democracy, aristocracy, epistocracy (noocracy) or even monarchy (so long as the monarch does not define the polity exclusively in terms of her own will). Indeed Stone acknowledges this in his paper, arguing that impartiality (which has no essential connection with democracy or even equality) is normally couched in terms of the rule of law, a concept that is applicable to any political system. No doubt sortition would also have enabled the German government to implement the holocaust in a more impartial way, given that the resources were not available to achieve the Final Solution in its totality.

This is the reason why the inherently conservative Stone-Dowlen Lottery Principle has proved unattractive to those of us who argue that democracy has important intrinsic properties, rather than judging it purely in terms of outcomes. Stone does make an egalitarian case for democracy (input-level equality in Rothstein’s terms) — we all have an equally miniscule role in selecting the government and we are all (theoretically) equal in our chance of securing public office. This numerical equality could be secured by sortition better than it can by election, on account of Manin’s (1997) principle of distinction. But democracy (in it’s non-Schumpeterian form) also suggests an attempt to ascertain the political preferences of (the majority of) the demos (the ‘popular self-rule’ argument rejected by both Stone and Rothstein). And democrats should also attend to Rousseau’s requirement that the preservation of the freedom that is our birthright presupposes that we are the author of our own laws. Neither of these factors has anything to do with (epistemic) outcomes; meanwhile how to achieve freedom and our own policy preferences in large-scale societies is a demanding problem.

Can sortition have anything to contribute to the intrinsic case for democracy, other than the establishment of arithmetic equality and the diminished likelihood of famine in developing countries (Sen, 1983)? I would like to argue that in addition to the Blind Break (an entirely negative principle) there is an additional positive principle, which I will refer to here as the Invisible Hand.** Decision-making by all citizens is not feasible in large political communities on account of the problem of rational ignorance, so a case can be made for decision-making by a representative microcosm — a portrait in miniature of the entire citizen body. But why not just use stratified sampling to ensure proper representation of age, gender, occupational category etc, why do we need to resort to sortition? If, however, politics is attending to the general arrangements of a group of people thrown together by choice or necessity (I paraphrase Oakeshott here), there is no a priori way of knowing what the relevant criteria are for the stratified sample (I’m leaving aside the further need for impartial selection, secured by the lottery). But this is where the law of large numbers comes into play — if the sample is large enough then there are good mathematical reasons to believe that random selection will select for all salient features (including those that we are entirely ignorant of) in proportion to their presence in the target population. This is an entirely rational use of the lot (it’s the ratio that we’re interested in), hence my argument that the Invisible Hand is an entirely distinct use of sortition to the Blind Break.

Dowlen does discuss representation in his 2008 book, but concludes that this is a ‘weak’ use of sortition. But this is only tautologically true, in so far as sortition has been pre-defined in terms of arationality. In his 2011 book, Stone also deals with representation in a cursory manner (the term fails to make an appearance in the index) but he subsumes this function within his one hegemonic Lottery Principle. In the present paper he also repeats Bernard Manin’s argument that sortition cannot possibly be compatible with the need for democratic consent. I thought I had successfully refuted that canard in my paper for the Manin workshop in Paris, so there is clearly a need to return to that subject but that will require a further blog post (Part II, forthcoming).


* The Dublin Declaration (2013) should not be confused with the Dublin Proclamation (1916).

** Yves Sintomer has argued that there are more like five or six lottery principles, but our concern here is for the political potential of sortition in large modern states, as opposed to archaic uses such as rotation.


Gil Delannoi, Oliver Dowlen and Peter Stone (2013), The Lottery as a Democratic Institution (Studies in Public Policy 28, Policy Institute, Trinity College Dublin).

Oliver Dowlen, (2008), The Political Potential of Sortition (Imprint Academic).

Bernard Manin (1997), The Principles of Representative Government (Cambridge University Press).

Bo Rothstein (2011), The Quality of Government: Corruption, Social Trust, and Inequality in International Perspective (University of Chicago Press).

Amartya Sen (1983), Poverty and Famines: An Essay on Entitlement and Deprivation (Oxford University Press).

Peter Stone (2011), The Luck of the Draw: The Role of Lotteries in Decision Making (Oxford University Press).

Peter Stone (2013), ‘Democracy and Good Government’, draft paper presented at the workshop Sortition and the Consolidation of Democracy, Queen Mary, University of London, 10-11 October 2013.

12 Responses

  1. Two quick responses. First, I’m not sure what you mean when you say “most” of the papers at the workshop endorsed the position you associate with Oliver and myself. What is your basis for saying that? Gil Delannoi did not say anything about it at all, nor did Mindy Peden or Conall Boyle or our guest from Politeia 2.0. You’re welcome to disagree with me and with Oliver, but I would appreciate it if you would not depict us as some kind of monolithic and ruthless conspiracy, to use JFK’s turn of phrase.

    Second, I addressed the point you raise her at the close of the session. I take it you are not convinced, but I’m actually not sure you understand my point, so let me try again. My point was that ANY defense of the lottery is going to require reference to one or more values. If two people are committed to different values, they will find different things to do with lotteries. So there’s nothing inherently democratic about lotteries; it all depends upon how you specify what are democratic values and relate them to lotteries.

    For example, you believe that lotteries are valuable because they secure descriptive representation. OK, fine. But there’s no inherent link between descriptive representation and democracy. An aristocracy could use the lottery to achieve descriptive representation with regard to the 1% on its decision-making bodies. There has to be a substantive commitment to democratic values in order to link democracy to sortition, for you as well as for me. So I see no reason to think my position less democratic or more conservative than yours.

    Now, we do disagree on which values we regard as essential to democracy. You stress majority rule, I stress political equality. That’s fine–there’s a worthwhile argument to be had about this, and I certainly don’t have a fully worked-out position on this yet. But it is just wrong to act like it is true by definition that democracy means majority rule. Consider the debates over Classical Athens (which I mentioned in my talk) and whether its institutions were designed to protect political equality or to ensure the supremacy of the Ekklesia, or both. There’s room for debate here–debate between people equally committed to democracy–which you simply seem to be missing.


  2. Peter,

    Apologies for tarring everyone with the same brush. I guess this was prompted by the hegemonic nature of the Dublin Declaration, which did not strike me as a fair summary of the debate in the workshop. It’s true that Gil’s short talk only outlined a typology of forms of governance but his signature was on the Declaration, hence my depiction of him as a member of the faction that treats statistical representation as a weak use of sortition. Conall has always disdained the use of lot as a form of representation and Mindy states in her paper: ‘At Dowlen’s instigation I mined [sic] for interpretations where the introduction of a “blind break” could conceivably and practically [undermine elite bias]’.

    I agree that sortition will only provide a statistical portrait of its target population. The underlying assumption of my approach (this is also true of Yoram, C&P and the other advocates of statistical representation) is the (universal) franchise that is a characteristic of modern democracies. It’s true, as you point out, that a lottery can “describe” any population — for example an aristocracy, or one based on gender or racial exclusions. So I suppose the universal-franchise assumptions that we are working under are somewhat parochial. The commitment to democratic values is also assumed, so the system of representation that we are advocating would probably not be suited to, say, Saudi Arabia. As for whether your position is less democratic than mine, the thrust of your paper appeared to be that from a QoG outcomes perspective democracy was a piss-poor method of organising human affairs and that all sortition could do would be to secure impartiality (aka the rule of law). I’m not sure that there is anything remotely democratic about this conclusion so, yes, I genuinely do question your commitment to democracy. As for our respective conservatism, the Blind Break is simply a way of making any system of politics less partial, so it has no progressive implications whatsoever. As a self-confessed conservative I’m amused to find myself in an uneasy alliance with those who seek to blow up our political arrangements and replace them with something better.

    I’m equally committed to the principle of equality as I am to the principle of majority rule, it’s just that I don’t think numerically-equal impotence is a form of equality worth having (that’s why I don’t bother to buy tickets for the National Lottery). The case of classical Athens is a red herring as all citizens stood a reasonably chance of obtaining office — power (rather than numerical impotence) was shared on a rotational basis. My commitment to preserving the equality of those who in a modern sortition-based polity would be excluded from exercising power is the reason for my stubborn insistence that the allotted assembly should not indulge in speech acts or any other behaviour that adversely affects its collective representativity. This is the argument that I will be dealing with in part two of this post.


  3. As I see it, the capture by the ‘elite’ (a shifting group, but handy shorthand for the !%) has happened despite, or maybe because of the process of electing representatives. MPs and Civil Servants seem to be putty in their hands.

    I picture the sortitionally selected representatives as being lambs to the slaughter in the face of elite pressure — unless, as with juries-at-law, strict rules govern access to jurors. The Doge of Venice was kept a virtual prisoner to prevent undue influence being brought.

    Even with safeguards, I find it difficult to see how sortition would change the balance of power and wealth between the elite and the citizens. Can you help me here?


  4. Conall,

    In my proposal jurors are restricted to voting in secret, so as to make them less open to corruption by lobbyists. I have also been converted to Robert’s proposal for policy-making by e-petition, with a lower threshold for acceptance than that of the current UK Government (100,000 signatures). Yoram has described the constraints I would place on jurors as being akin to choosing between two elite proposals. There is some truth in this, but jurors could opt to reject all the options, thereby constraining elites to make proposals that are likely to secure the support of the allotted jury. At the very worst this would amount to a modest improvement on our existing political arrangements.


  5. A few points;
    1. Compulsory sortition is used mainly to distribute burdens equally. It may easily be unfair, because what from an abstract point of view is the same burden, eg serving on a jury, is concretely a much greater burden for some rather than others.
    Whether or not sortition is appropriate from this point of view in a political context depends very much on the role that the selected are expected to perform. So, in Keith Sutherland’s proposal they operate under constraints very similar to a jury in a civil court, with no scope for initiative or negotiation. That is a burdensome role, and veery many people would rather not be lumbered with it. So sortition is appropriate.
    On the other hand, if those chosen are supposed to fill the sort of functions that parliamentarians now fill, initiating and refining legislation, monitoring the administration and providing community leadership over the whole range of public affairs, it is extremely unlikely that a random sample of the population would contain enough people of suitable education, motivation and capacities to fill those roles, not because they are stupid, but because the focus of their lives is on the development of their particular goals in life, often at a stage where those goals demand almost their whole attention. It is unjust to a promising artist who is struggling to find her voice or to a mother striving to balance family and career to attempt to impose such a burden on them, interrupting their lives for a substantial period.
    We have first to decide what we expect people to be able to do in a certain role before we can say whether it is appropriate to assign them to it by compulsory sortition.

    2. Choosing by sortition among those who are willing to perform a role is a very different matter. It certainly will not achieve the sort of abstract equality that some seem to want, even if everybody is equally entitled to nominate, since the chattering classes and retirees would supply most of those who are interested. It would still be possible to ensure that all the relevant different interests were represented if the procedure of choosing were based on an analysis of those interests and candidates were scored on appropriate criteria to give a representative sample of those affected. Once again, it is necessary to decide first what they are supposed to do.

    3. Discussions of sortition seem invariably to focus on the myopic concerns of nation states. It might be more profitable to think of the problems of transnational governance. We need a host of regional and global specialised bodies to regulate the chaotic areas of environmental concerns, trading in all sorts of thing, especially money, international crime and human rights. Some sortitional arrangements offer a way of getting such authorities off the ground without their becoming victims of the internal politics of states.


  6. >Discussions of sortition seem invariably to focus on the myopic concerns of nation states…

    Jesus wept.


  7. John,

    I salute you, as always, for the clarity of your analysis. My overriding concern is representativity, hence my insistence on a constrained and mandatory role for sortition, as in jury service. As for the Burnheim model for demarchic committees, I’m not sure how we would know what the “appropriate criteria” were. I acknowledge that this would be a lot easier for subject-specific committees, but I worry then that this leaves the final decision-making in the hands of people with interests (albeit declared interests). I would agree with Madison that the disposing function is better left to people who have a more dispassionate perspective but share modern scepticism as to whether individuals with this higher faculty (“wise and virtuous citizens” in Madison-speak) are available for the task. This being the case, the best we can do is the aggregate judgment of a representative sample (the allotted jury).

    The ideal solution might well be a combination of the demarchic committee and the allotted jury; perhaps Terry’s hierarchical model has something to offer here — voluntary demarchic committees would deliberate together before presenting their policy proposals to an allotted jury, which would have final disposal rights. This might even be a compromise we could all agree on as it would combine the best of both approaches and would be based on a clear analysis of the different roles that allotment might play. The Blind Breakers would approve of the use of sortition for voluntary demarchic committees (as it would sanitise the appointment process) and the democrats would be satisfied that the disposal function was in the hands of a statistically-representative jury. This would also resolve the tensions contained within the oxymoron “deliberative democracy”, as the demarchic committee would assume the deliberative role (agreeing on and formulating proposals) and the jury the democratic disposal role. And there need be no recourse to partisan posturing and the manufacture of artificial cleavages.

    I’ve always viewed sortition in the context of the nation state, as this is the existing political forum. But you are right to point out that many (most?) of our pressing problems are transnational. I’m currently reading an interesting book by Nick Rengger, who takes an Oakeshottian approach to international relations (IR). In Oakeshott’s view the problem with IR is the lack of a non-instrumental civil society perspective — perhaps sortition at the international level would be a way of addressing these concerns, especially as it would not displace any existing bodies with clear democratic mandates.


  8. Keith, have you read the paper in question? That might help. I agree that Rothstein’s commitment to democracy is questionable because impartiality, the value he highlights as central to good government, is almost indifferent as to the type of government involved. (I say “almost” because it is hard to imagine a fascist regime applying laws impartially, even though it is theoretically possible.) My argument was that impartiality, at both the input and the output side, needs to be fleshed out by adding explicitly a set of egalitarian values. On the input side, that means political equality, which I take to be the essential core of democracy.

    Now, this is all pretty preliminary; the paper was intended as exploratory. I don’t have a fully worked-out theory of this, but frankly, neither do you. I really haven’t heard a clear picture of just what set of values you want advanced, and why. Certainly, the following claim of yours has big problems:

    “I’m equally committed to the principle of equality as I am to the principle of majority rule, it’s just that I don’t think numerically-equal impotence is a form of equality worth having (that’s why I don’t bother to buy tickets for the National Lottery).”

    Please permit me to point out that if a nation the size of the U.S. selected a legislative body by lot, nobody would have ANY input into the selection of the legislature–not even the “numerically-equal impotence” that we have now. Now you would like to keep one elected legislature alongside an AC (assuming that’s still your position–I understand it’s evolved a bit). But if you keep the elected legislature around because of its democratic credentials–which I gather is your reasoning–then 1) you are accepting the rationale for our current institutional arrangements (talk about conservative!) and 2) you have not made any DEMOCRATIC case for sortition. Sortition does not generate rule by the people, whatever its other virtues. And so if sortition is to be regarded as a democratic institution, the case for it has to be made on other grounds.

    Finally, this might not have been clear in my talk, but it’s just a caricature of my position to say that having a vote exhausts the demands of political equality. I don’t have a fully worked out theory of political equality, but I would take it to involve much more extensive rights of political participation. Those rights cannot be fleshed out meaningfully as having an “equal say” in government for precisely the reasons you cite–in a nation of millions, having an equal say is tantamount to worthless. And so that set of democratic rights will almost certainly justified in nonconsequentialist terms. It’s the right way for the system to treat people, setting aside the question of the system’s effects. (Obviously, the system’s effects matter. Within the set of institutions compatible with political equality, we should select one that leads to good outputs. And that leads us to considerations of collective wisdom, diversity in decision-making, etc.)


  9. Peter,

    I’ve read your paper (three times) but not Bo Rothstein’s book, hence my reliance on your interpretation of it.

    My overriding concern is the separation of the proposing and disposing functions. Given the nature of ho boulomenos, the most we can hope (and wish) for in the proposing function is equality of opportunity. I’m agnostic as to the best way of securing this — preference elections, e-petitions or demarchic committees (see my response to John Burnheim, immediately above your last comment), or a combination of all three. The democratic imperative is that “(s)he who wishes” should be able to offer herself as an electoral candidate, join/form a political party, make an e-proposal or volunteer for a demarchic committee and that opportunity costs should be kept as low as possible for all these options. This would equate to the “rights of political participation” that your rightly point out as an essential component of democratic equality (of opportunity).

    A democratic model for disposing presupposes an entirely different form of equality. I’m no expert on the egalitarian literature, but I suppose this would be more akin to equality of outcome. My concern here is how to ensure that all people have an equal vote in determining the outcome of legislative proposals and this presupposes direct democracy. However referendums and other direct-democratic mechanisms in large political communities do not return a considered verdict, as voters don’t take the considerable time and trouble required to study the arguments (as their vote is of negligible causal significance). Hence the need for a statistically-representative microcosm with a tightly-constrained mandate (listening to balanced information/arguments and then determining the outcome via the secret vote). Sortition performs a dual role in this scenario — both the blind break and the invisible hand — and impartiality and statistical representativity is further secure by the secret vote and the prohibition of speech acts on the part of randomly-selected jurors.

    Whether or not this delivers a high level of QoG is orthogonal to the argument for outcome equality, but the wisdom-of-crowds (WoC) literature provides some hope that the output quality would be no worse than that of an elected assembly — unlike in elected legislatures allotted members would have to listen to the debate, rather than turning up when the bell rings to be herded through the division lobbies. (Note that the argument for cognitive diversity is more applicable to the proposal function as opposed the aggregate WoC — Helene and her colleagues like to keep the two issues separate.)

    I hope this clarifies how it’s possible to have your cake and eat it — how to combine political equality and majority rule. Most of the confusion is caused by the combination of different functions (proposing and disposing) within a single “body of men” and the failure to distinguish between equality of opportunity and equality of outcome. John Burnheim’s comment to this thread is a good example of the analytic clarity that is needed to make progress in this debate.

    In part two of this post I’ll make the case for sortition both as a way of establishing “rule by the people” and for “consent-by-proxy” as a better alternative than the existing Lockean model. But that’s for another day.


  10. Keith,

    Your summary above is the closest thing to reflecting my own views you’ve posted yet. My vision is that any and all who wish may participate in the processes of generating proposals. A diverse allotted body that is representative of those with sufficient commitment to be willing to research and deliberate to filter down, amend and recombine the widely generated proposals to a final draft that is subsequently voted up or down by a fully representative jury through a secret ballot.


  11. Terry,

    Great! Three sequential principles:

    1. Ho boulomenos
    2. Demarchic agenda/review committees
    3. Allotted juries

    Advocates for the last stage will be drawn from the first two. The expertise that stage 2 would uncover is an improvement on my vague talk of a House of Advocates, although I’m still concerned about the absence of peer review or any exogenous way of identifying expertise. I know we are all concerned about elite dominance but are there not more reliable ways of uncovering expertise than self-nomination?

    I would hope that John Burnheim might also find this compromise agreeable.


  12. […] view sortition is perfectly compatible with the democratic case for equality (which I addressed in part 1 of this post) but can in no sense can be taken as an indication of democratic consent. (Of course […]


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 )

Connecting to %s

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

%d bloggers like this: