How to design a democratic legislative system – order of questions

I think it will help us, and could help many other people, to have a useful order of questions for designing a democratic legislative system. I’m not saying “the right order of questions,” or even “the most useful order” – only “a useful order.” I’m also not suggesting that we should follow this order in our conversations. Instead, I think it could act as a useful reference point for those times when someone says, “Wait a minute – it’s premature to talk about x before we’ve settled y.”

Here’s what I’ve come up with so far:

  1. Criteria – What criteria should define a “democratic” and “good” legislative system?
  2. Categories of actors – Which broad categories of actors (e.g. all the people, allotted representatives, elected representatives, all-purpose versus limited-purpose representatives, staff) should play important roles in the legislative process? What roles should they play?
  3. Activities – What are the main activities that must (or should?) be carried out in a democratic legislative process, and in what order? In some cases order will matter, in others it won’t.
  4. Bodies and offices – Which specific bodies and offices (e.g. allotted chamber, single issue panels) should carry out each activity, playing what roles?
  5. Processes – What processes should be used for each activity?

What do you think? I look forward to your ideas, and I’m hoping that maybe together we can create a simple structure that will not only help us, but many others as well.


60 Responses

  1. I would have thought that 3 (activities) would logically follow after 1 (criteria) and I’m not sure whether it’s possible to distinguish usefully between 2, 4 and 5. The particular activity (3) would suggest an appropriate category/body/process. This would require a list of only three:

    1. Criteria (normative analysis)
    2. Activities (functional analysis)
    3. Implementation (actors/bodies/offices/processes)


  2. Keith,

    You wrote, “I would have thought that 3 (activities) would logically follow after 1 (criteria).” I think you’re right – in fact, that’s how I was thinking about it when I started this. Certainly, questions about criteria are logically prior to everything else. And I think it makes sense to identify the necessary activities of lawmaking (and dependencies between them) before specifying who should carry out each activity.

    You also wrote, “I’m not sure whether it’s possible to distinguish usefully between 2, 4 and 5” (categories of actors, vs bodies and offices, vs processes).

    I haven’t been able to fully articulate this yet, but let me try to explain what I’m thinking. Before designing specific bodies like your “House of Advocates” or Terry’s “Interest Panels,” I’d first want to decide the general principles of what allotted representatives are good for, what elected representatives are good for, and what mass voting is good for. And then I’d want to decide the pros and cons of general-purpose representatives (dealing with all legislation, like the allotted chamber that Yoram advocates) vs limited purpose representatives (dealing with one issue, or even one bill, like the panels that Terry advocates). Then I’d want to design a set of bodies and offices, then work out which bodies and offices perform which tasks, and lastly work out the protocols they use (what meetings processes? What voting systems? Etc.)

    I’d love to hear what other people think as well.


  3. Yes that makes sense, but don’t the general principles fall under 2? If so then 3 should perhaps have subdivisions a) pros-and-cons b) offices c) protocols


  4. David,

    > What criteria should define a “democratic” and “good” legislative system?

    Those are two completely different issues. “Good” is a completely subjective matter. Different people will consider different policies as being good. “Democratic” on the other hand is a much more objective term. Robert Dahl has pretty good discussions of democracy in several of his books. I wrote an extended review of (part of) his book “On Political Equality”, which deals with this issue.

    As for a connection between “good” and “democratic”, one can hope to argue that a democratic system is generally better than non-democratic ones, but anything stronger than that is clearly out of the question.


  5. David isn’t asking about criteria for a good legislative OUTCOME, but rather a good system. Due to unintended consequences and the general unpredictability of the future, we can’t worry about BEST decisions for society, but rather what decisions reflect what the whole people would choose if everyone fully participated. We could go too far afield defining a good legislative system, if we were to weigh the value of Plato’s Guardians, for example vs. some variant of “democratic” process.

    But, if we limit our discussion to legislative traits that can at least arguably be government of, by and for the people, there is some room for helpful discussion. Some of the desirable criteria will be mutually exclusive, and it will be a matter of finding the sweet spot or balance point. For example:

    What is the relative value of “expert,” or experienced legislators? Can experience and corruption be separated? If expertise is important can it be provided in another way. What is the value of diversity or descriptive representativeness? Does diversity also mandate incompetence? etc.


  6. The recent work on epistemic democracy does claim that there is a correlation between cognitive diversity and getting the “best” outcome (some would go so far as to claim even the “right” outcome) and this is the core of the epistemic case for sortition. The key authors on this are David Estlund, Scott Page and Helene Landemore. Obviously the notion of the “right” outcome is contestable but sortive democracy has the advantage of combining this with the normative argument, so you can have your cake and eat it. This happy coincidence doesn’t apply to any other political system (Plato’s guardians being the obvious example).


  7. Terry,

    A good system is one that produces good outcomes.

    It seems to me you are simply redefining “good” as (some interpretation of) “democratic”. Again, I think we should avoid conflating those two terms. This is all too often done in modern political discussion (both by experts and by laymen), and it serves only to obfuscate matters and entrench misconceptions that are part of the dominant ideology.

    (I will admit that upon analysis it turns out that, due to epistemological difficulties, “good” and “democratic” are not as separate as I made them out to be. But such dependencies – which are not identities – should be derived rather than assumed at the outset.)


  8. Yoram, thanks for your comments. I agree that we should avoid conflating “good” and “democratic,” and I now think it would be simpler just to ask about criteria for “good” legislative systems (which, for most of us, would include some version of “democratic,” although our definitions would be different).

    Terry, I actually do think that a “good” legislative system should produce “good” laws – but I’m much *more* interested in it’s effects on things like democracy (the extent to which the people rule), and human development and learning, and the ability to get better at those things over time.


  9. Here’s a revised version of the order of questions, based on feedback so far.

    1. CRITERIA – What criteria should define a “good” legislative system?

    2. ACTIVITIES – What are the main activities that must (or should?) be carried out in a legislative process, and in what order? (In some cases order will matter, in others it won’t.)

    3. CATEGORIES OF ACTORS – Which broad categories of actors should play important roles in the legislative process? (for example, is there a role for elected representatives? Allotted representatives? Representatives of interest groups? Limited-purpose representatives? Self-selected volunteers? Permanent staff? Subject mater experts?)

    4. ROLES OF ACTOR CATEGORIES — What general roles should each category of actor play? (for example, should elected representatives propose laws and allotted representatives vote on them?)

    5. BODIES AND OFFICES – Which specific bodies and offices (e.g. House of Advocates, Interest Panels) should be involved in each activity, playing what roles? For each body: what selection method, term of office, salary?

    6. PROTOCOLS – What protocols should be used for each activity? (deliberation processes, voting systems, etc.)

    By the way – is there a topic in this forum devoted to the question of criteria for legislative systems?


  10. Terry, you wrote, “What is the relative value of “expert,” or experienced legislators? Can experience and corruption be separated? If expertise is important can it be provided in another way.”

    I think you have a good point. I’ve been habitually thinking of elected legislators as “more experienced,” and allotted legislators as ” less experienced,” but there’s no necessary connection between amount of experience and selection method.


  11. I’m struggling to get my mind around the third design question I posed,

    “3. CATEGORIES OF ACTORS – Which broad categories of actors should play important roles in the legislative process? (for example, is there a role for elected representatives? Allotted representatives? Representatives of interest groups? Limited-purpose representatives? Self-selected volunteers? Permanent staff? Subject mater experts?)”

    I’m starting to think that maybe this is really not about categories of actors – maybe it’s really about methods for selecting decision makers – what methods are worth using at all, and what’s the best role for each one. For example:
    • “all the people” (mass voting, Town Meeting, other methods that can at least potentially accommodate the whole demos)
    • sortition
    • election
    • self-selection (yes, I realize that to some extent both elections and Town Meeting are also examples of self-selection, but I’m thinking of groups that can’t include a whole demos)
    • interest group (“stakeholder”) representation

    Any thoughts?


  12. The logical extension of “all the people” for large societies is referenda and initiatives (which don’t appear on your list) as opposed to self-selection. I think there is an inherent contradiction between self-selection and democracy.


  13. Referenda and initiatives is what I meant by “mass voting” – I probably should have spelled it out.

    Do you see no useful role for self-selection in agenda-setting? Proposing possible issues and bills? Civil society? Opportunity for “the loudmouths” to make their voices heard on issue that they’re interested in?


  14. I don’t have a problem with self-selection iff the proposals then have to undrgo a democratic screen (ideally a votation) prior to consideration by an allotted chamber. This, after all, is the basis of liberal democracy — individuals self-select and then offer their candidacy to the public. To go straight from self-selection to sortition would be the worst kind of oligarchic rule as ordinary citizens would have no opportunity to make their preferences known.


  15. What about interest-group representation? If I remember right, you did a lot of thinking about that between your first book and your last, and changed your design accordingly.


  16. Yes that’s right. My first book called for the abolition of party politics and their replacement by sortition alone, so my position was very close to some of the more “purist” members of this forum. But I was subsequently (and reluctantly) persuaded that this didn’t allow for the representation of interests, so at that point I turned to Madison’s distinction between interests and judgment, with the former function provided by political parties and the latter by sortition. When political parties are entrusted to represent interests this means that all citizens have the opportunity to register their preference and all votes are exactly equal.

    But, as Peter has pointed out, this leaves the sortive chamber with only a weak form of veto power, hence the need for other ways to influence the agenda-setting process. This is where loudmouths come in — via initiatives, petitions, or Terry’s model of volunteer sortition committees. However these still need to go through a democratic filter (ideally votation) where all citizens can choose which loudmouth they prefer. So I have no problem with self-selection, but only if all citizens get to choose which loudmouth they prefer. I fully accept that this allows media and other elites to influence the debate and that the public’s choice will be unconsidered, but the former can be rectified by breaking up media monopolies and the latter by the final deliberative judgment of the allotted chamber. It strikes me that this is a reasonable attempt at a modern analogue of Athenian political practice, but that to miss out the intermediary stage — to go straight from the loudmouths to the sortive microcosm — would be akin to Athenian democracy without the general assembly.

    BTW: Jeremy Waldron pointed out to me recently that the interests-judgment distinction is much older than Madison — the doctrine of nemo iudex sua causa is basic to Roman law (although it originated in Aristotle) and is also important to Coke and the English common law tradition. See his paper “Legislatures Judging In Their Own Cause” (Legisprudence, III, 1).


  17. Keith wrote:
    “To go straight from self-selection to sortition would be the worst kind of oligarchic rule as ordinary citizens would have no opportunity to make their preferences known.”

    The “opportunity” would exist, in that any citizen can sign a petition, initiate a petition, or volunteer (self-select) for an interest panel. Your point is that such a system does not actively solicit the less interested citizen to express a preference. I question the value or desirability of referenda on a large number of issues where most citizens have no opinion, let alone a “considered” opinion. A filter like “votation” seems like it would empower demagogues with media access, while stifling important topics that would be valued ONLY after information and consideration (such as would be achieved in a smaller sortition group). For example, an unpopular minority group might have no chance of getting their top concern through a votation filter, but might persuade a smaller group in an allotted chamber of the justice of their request.

    The concept of democracy I am using is coming to the decision that the whole society would, if the whole society were able to deliberate with attention and information. Votation, does not seem to help here.


  18. Terry, you wrote, “The concept of democracy I am using is coming to the decision that the whole society would, if the whole society were able to deliberate with attention and information.”

    I think this is critically important. I also think it’s insufficient, for two reasons. First, if we want democracy to mean “rule of the people,” what you describe is a *proxy* for rule of the people. And second, if the vast majority of people have no meaningful role to play, then they miss out on the human developmental benefits that come from exercising voice and power in the decisions that affect their lives – what the Athenians called “paideia.”

    Because of this, I’m interested in strategies that would increase the number of people who can participate in governance. For example:
    1. Having a large number of randomly-selected bodies instead of one allotted chamber, which increases the number of citizens who get to exercise some voice and power (you’ve proposed this, and it’s the basis of demarchy)
    2. Having something like the New England Town Meeting at local levels
    3. Having a role for initiative and referendum, as Keith proposes
    4. Use of large scale deliberation (as in America Speaks “Twenty First Century Town Meeting”), and/or ways to aggregate small scale face to face deliberation


  19. David,

    I agree on your points one and two above, but am not yet convinced about three and four.

    In a large society representation is inevitable simply due to the need for division of labor (there are too many decisions for me to participate in very many of them). As a citizen, here’s what I want from a political system…

    A. A way for me to weigh in on issues that I care about. (I propose the creation of Interest Panels by lot that include EVERY person who wants to be involved…this may mean a large number of them).

    B. Confidence that the countless other decisions that need to be made will be made by groups that include “people like me,” who share my interests, without corruption or concentrations of power, and with good information (many allotted decision-making panels).

    C. Confidence that administration and implementation decisions that genuinely require expertise will be made by people selected and accountable to people like me who will have the opportunity, time and motivation to be well informed. (allotted popular scrutiny of bureaucrats).


  20. Terry,

    I agree with everything you wrote here. I would add one more point to your list of what you want from a political system as a citizen,

    D. The developmental benefit that comes from the experience of meaningful engagement in governance – which will probably change my ideas about what issues I want to weigh in on.


  21. Terry,

    To what extent do you think it’s important to represent interest groups as well as individuals? And if it’s important, how would you propose to do it?


  22. I haven’t thought this through yet, but here is my off-the-top-of-the-head reaction… It is inevitable that interests will organize (forming associations, parties, or the like). But if these entities are formally inserted into the political structure they hobble the ability of people to compromise. Once organized into an entity, the particular interest becomes the sole priority. Such associations (like corporations) are not natural persons, and will inevitably distort democratic decision making if empowered to participate formally. I would favor having interests be involved only THROUGH natural persons within a structure limited to natural persons.


  23. David, of course I agree with you regarding establishing a system that would “come to the decision that the whole society would, if the whole society were able to deliberate with attention and information.” Unfortunately sortition can only do that via voting (for all the reasons that we have been debating, ad nauseam). It is true, of couse, that most people are indifferent to most things, but that does not justify priviliging those with strong views. I’m fully aware of the dangers of demagoguery, this is a problem that the Athenians had, but they did not choose to address it by abolishing the assembly. Unfortunately what Terry and yourself are suggesting is tantamount to the latter. By all means continue to advocate such a system but let’s all be fully aware that it has nothing to do with democracy.


  24. Sorry, anonymous was me and I should have been replying to Terry, not David (am in the middle of a ski trip and not concentrating properly!



  25. It seems to me that the combination of initiative and referendum plus large scale deliberation methods could create a pretty good modern-day equivalent of the Assembly — especially if people also had the experience of participating in face to face assemblies with real power at the local level.


  26. I do not think referenda (or votation) are actually that analogous to the Athenian ecclesia. Ecclesia participants had to listen to presentations from both sides before voting. Presumably they had some rules for assuring that both sides could be heard. Since the space was limited, only a minority of citizens (around 6,000) could participate in any given ecclesia assembly. The Athenian ecclesia is more akin to an assembly of the self-selected and interested citizens forming a semi-representative body. In many respects this is more similar to a sortition body that hears both sides, than to a referendum in which voters can show up and cast votes without hearing a word of argument.


  27. Terry,

    The Athenian Assembly was to a large extent a mass political body, and was therefore quite similar in many respects to the modern mass political arena – favoring the rich and famous, with almost all those present being passive observers. The fact that the allotted Boule set the agenda and proposed legislation was the main democratic aspect of the work of the Assembly.

    Having to listen to “both sides” means very little. These days, as well, we are usually treated to “both sides” of “every issue”. Of course, the elites determine what “the issues” are and what their “two sides” are. Once these parameters are set, the rest is mostly political theater.

    (As for self-selection, the electorate today is a self-selected subset as well whose members are probably more politically engaged and more “informed” (i.e., exposed to more political propaganda) than the rest of the group. How does this make it more like an allotted body?)


  28. The key factor about the ecclesia is that every citizen COULD attend if he chose to and everyone had the right to speak. This has always been seen as the primary institution of Athenian democracy and a modern analogue has to make SOME provision for every citizen to have a say. We are all aware of the flaws of the ecclesia hence the need for a modern equivalent of the legislative courts. Insisting that the flaws require the wholesale rejection of mass democracy is throwing out the baby with the bathwater. Even if we ignore these arguments, in practice there is zero likelithood of a sortition-only system ever meeting with the approval of governing elites and the general public. Mixed government is here to stay, so we’d better just get used to it.


  29. Keith,

    I agree that a necessary element of a good democracy is the opportunity of anyone who wishes to weigh in on a matter. I think that mass voting (in a referendum or candidate election) only provides this symbolically. I also agree that any system that does not provide AT LEAST an equivalent ability to participate for any who choose, is doomed. That is why I propose, a superior means for any citizen who wishes to participate — through allotted Interest Panels of volunteers. The participation here would involve gathering facts and persuasive arguments, as well as compromise and seeking win-win opportunities (unlike elections, which are more akin to massing resources and armies for war).


  30. Yes I agree that voting in a mass democracy is largely symbolic, but one shouldn’t underestimate the power of symbolism! The vote of a single Athenian citizen at the assembly was largely symbolic, especially as they were not even counted. As you well know, it will only be a tiny subset of people who volunteer for interest panels and volunteers will never be typical of the whole population. Just as sortition will provide statistical representation at the judgment level, democratic norms require something similar for the agenda-setting process, even though this will entail the dysfunctions that are inevitably associated with mass democracy. But modern information technology, allied with statutory provision for plural media can go a long way towards reducing the dysfunctions. Even if you could overcome the theoretical obstacles re. democratic legitimacy, I can’t see how you can get round the need for symbolism (one man one vote etc. etc.) Note that I’m v, sympathetic to your call for interest panels, but can’t see how you can get round the need for the democratic filter (votation). I accept that this will mean that majority views predominate, but iminoritiy rights are normally protected by liberal institutions, not democratic. Unforunately I can’t see what I’m typing at the moment oexcue any gobbledygoodk


  31. > a superior means for any citizen who wishes to participate — through allotted Interest Panels of volunteers.

    If sortition takes place then presumably some people would not be selected by the lottery and would not participate. If no sortition takes place, then how is it different from voluntary participation in a special interest group (such Greenpeace or the ACLU or many, many others) in the current system?


  32. My idea of Interest Panels is that EVERYBODY who wants to contribute to formulating proposals for the Allotted Chamber WOULD get to join a panel. The role of the lottery at this stage is not to select WHO gets to participate, but rather simply assigns them to a random group, so that the panels will be diverse. For a particularly hot topic there may be a huge number of interest Panels, all working to develop a proposal that they think can pass through the next filter of a more fully descriptively representative body, before going on to a final vote by the AC. Some (many) Interest Panels will simply not be able to agree on a proposal (maybe they would have a super-majority requirement to finalize a proposal).


  33. Terry, do you really believe that a system of government by sortition alone would ever meet with the consent of both governing elites and mass voters (who would lose at a stroke the hard-gotten gains of the franchise battle)? Bear in mind that there are no historical examples of sortition as a system of democratic representation, so you’re asking rather a lot.

    If not then is this not then just a self-indulgent scholastic exchange, akin to estimating how many angels can dance on the head of a pin?


  34. I don’t think Terry is suggesting government by sortition alone. I think he’s suggesting a combination of sortition and self-selection.


  35. Terry,

    Again – if no selection is involved, we are back in the mass politics situation producing its well known elitist effects.

    As things stand today, anyone can participate in various activist groups that are able to propose bills. How is what you are proposing different?

    Why would the average person bother to join a panel if their huge number guarantees that the effort-to-effect ratio is extremely large?


  36. Self-selection + sortition is John Burnheim’s formula, but John has had the good grace to acknowledge that he is no democrat. I find myself agreeing with Yoram but would argue that the elitist effects need to be ameliorated rather than abolished by fiat.


  37. Yoram,

    Like Keith, I am less concerned about a elitist effects when they are limited to the proposal stage (rather than the decision stage) AND there is an alternate sortition route for agenda setting and proposal as well. Elites may have good ideas, as well as bad ones, and should be allowed to put ideas into the hopper, as long as that is not a monopoly right.

    My concern is about the extremely negative spill-over effects that elections have on society generally (demonizing opponents, fostering hate, etc.) and thus also on the deliberation within an allotted chamber, since its participants will have been exposed and thus poisoned by election campaign propaganda.

    Keith makes an obviously valid point, that eliminating the franchise entirely is a vain goal…So are there any ideas for ways to minimize the us vs. them negative effects I see if elections continue?


  38. The first thing would be to put a strict limit on party campaign funding and to ban TV advertising (as in the UK). Given that politicians would not be forming a government, they would simply be advocating policies, at least the negativity would be relevant to the issues, rather than focusing on the candidate’s haircut or birth certificate.

    But I think there is a philosophical issue at stake, whether or not conflict and disagreement is a valuable way of arriving at the right answer. There is some evidence from the UK showing that legislation resulting from an all-party consensus is usually poor quality. This would also be backed up by the groupthink evidence (Thaler, Sunstein etc).


  39. Terry,

    It is still unclear to me in what way the panels you suggest are different from present-day voluntary membership and activity in various political groups. Such groups can deliberate and come up with legislative proposals.

    > Elites may have good ideas

    Sure – members of the elites can use their seats in the allotted chamber to promote any such ideas. I don’t see why the elites should get a privileged position (even if not a monopoly) through their domination of mass politics.

    > eliminating the franchise entirely is a vain goal

    I also don’t see why we should accept electoralism as a fact of life any more than we should accept monarchy as a fact of life.

    > are there any ideas for ways to minimize the us vs. them negative effects

    As far as I am concerned, the electoral us-vs.-them political theater is not a problem – it is a ridiculous smokescreen for the underlying unity of the electoral elite.


  40. Keith,

    I make a distinction between conflict that arises inevitably from different interests and artificially manufactured conflict generated for the sole purpose of demonizing an enemy in order to win elections. MOST of the visible political conflict in the U.S. is “fake,” and fractures people into warring camps that I fear would continue to refuse to cooperate in an allotted chamber.


    The us-vs.-them-political-theater, even if “fake” still is a REAL problem simply because it can poisong attitudes and block potential compromise within an allotted chamber.


  41. Terry, I agree regarding the dysfunctional effects of artificially manufacturing conflict. I accept that elections have this effect but would argue that we need to minimise it, rather than seek to abolish it.

    >Sure – members of the elites can use their seats in the allotted chamber to promote any such ideas. I don’t see why the elites should get a privileged position (even if not a monopoly) through their domination of mass politics.

    Now I’m completely confused! In an earlier exchange you chastised me for falsely attributing to you the view that elites were not just the produduct of elections. It would appear from the above comment that you do in fact seek to divide the population into preexisting categories, so I think your earlier attacks on me were a little disingenous.


  42. Keith,

    So your goal is to minimize rather than abolish the harmful effects of artificially generated conflict of elections. Is this because you see counter-balancing positive effects of elections? Is it simply a pragmatic recognition that elections are too entrenched to advocate abolishing them (in order to be taken seriously)? Or is it merely adherence to the Aristotelian definition of democracy requiring a mass role for “the many,” even if that role is counter-productive?


  43. It’s for a variety of reasons:

    1) Pragmatic (exactly as you put it).
    2) Accountability (albeit imperfect, it’s better than nothing)
    3) Equality in the agenda-setting process (ditto).

    Two provisos:

    1) I’m not committed to elections, the bottom line for me is votation, with some of the items on the ticket being promoted by political parties, and some by petition, initiative etc.
    2) Aristotle’s positive case for democracy is more a justification for sortition (the wisdom of crowds).

    Where I differ from you and Yoram is that I don’t have a visceral aversion to elites and partisan behaviour, which I think is colouring your judgment.


  44. Keith,

    I get #1 (pragmatism), but not the next two reasons.

    The accountability reason seems like chickens wanting to keep a system of electing foxes to protect the hen house, so that they can vote out a particular fox after they see it eating one of the chickens, as a means of “accountability.” Why not get rid of all of the foxes?

    And I’m not sure I understand your point about equality of agenda-setting…Is that specific to votation?


  45. Terry,

    Unfortunately you can’t get rid of vermin, all you can do is drive interests back underground. Electoral accountability is weak, but it’s better than nothing. As for the equality of agenda-setting, the point is that all citizens should have an equal share in choosing between alternatives for the agenda, and should also be able to propose alternatives themselves, to be selected or rejected by their peers. This would require votation. Obviously some citizens would be more equal than others, so tight campaign finance limits and media plurality are also necessary in order to minimise this form of inequality.

    I must also acknowledge an error in my last comment when I said that Aristotle’s positive case for democracy was a justification for sortition. This is only true in the modern context, when sortition is being proposed as a means of descriptive representation. This was not the case with the ancients, where democracy meant mass affirmation in the assembly. The process was as flawed then as it is now, hence the need for deliberative scrutiny, but a constitutional system cannot describe itself as democratic without an opportunity for everyone to indicate their preferences directly.


  46. In terms of electoral accountability, I think it’s useful to distinguish between electing representatives versus decisions about laws (such as votation, initiative, referendum, Town meeting). I can see the logic behind arguments to replace elected representatives with allotted ones. But it’s hard for me to imagine a democracy with no role for “all the people.”

    Certainly, the role of “all the people” in the systems we have now is a sad facade of democracy — but I’d hate to use that as a reason to eliminate processes like votation entirely. Why not improve them instead?


  47. Agree 100%. If something is sub-optimal it’s better to try and improve it rather than eliminate it. Given that there are no historical examples of sortition as a system of democratic representation, some degree of caution is in order, partly on account of what would be acceptable (both to voters and politicians) and partly on account of the law of unintended consequences.


  48. P.S. Accountability in a votation-based system would function according to Bayesian principles — a group/person/party that made proposals that resulted in good legislation would have a higher probability of having their proposals accepted the next time round (no doubt Yoram will tell us if this is a misrepresentation of Bayesian learning). This would be how advocacy groups would morph into political parties and the long-term fortunes of the latter would depend very much on results. Parties that screwed up would more than likely disappear rather than simply sitting back and waiting for their turn in the next electoral cycle. Similarly an ad-hoc group that proposed a successful initiative would more than likely evolve into a political party. This would provide a workable combination of ex-post accountability with an initiative process that was open to all.

    It remains an open question as to where opposition advocacy would come from. If it was from the losers this would just reintroduce the yah-boo negative exchanges that characterise existing legislatures. My preference would be for a large, permanent advocates’ house that contained representatives of all the key organisations in civil society (employers’ groups, trade unions, environmental advocates, medicine, sport, education, religion etc etc). Some time ago Mark made the interesting suggestion that the advocacy house should also be appointed by sortition:


  49. Keith,
    If “all the people” use votation to filter the agenda, doesn’t this imply that interests that the majority are not currently sympathetic to — though they WOULD be, if fully informed — get their proposals blocked from thoughtful consideration — where they would have passed?

    It seems that votation simply undercuts the ability of the people to make considered decisions.

    I might prefer to use a lottery to select agenda items rather than votation ;-)


  50. Yes I agree this is a problem. Unfortunately democracy in practice cashes out as majority rule, with minority rights being protected by constitutional, judicial and other liberal safeguards. The problem with directly empowering minority views is that it is highly unlikely to meet with majority consent — even if, as you suggest, the verdict might be different after deliberation (although it might not). The views of the political class are also often poorly-aligned with popular opinion, but at least electors have the opportunity to eject politicians who step too far out of line.

    I do think the issue of majority consent is a serious problem with proposals that rely on sortition alone. I would refer to my earlier thought experiment that the judgment of every allotted assembly would, in theory, be the same in every case, and would therefore rightly be perceived as representing the informed consent of the whole political community. Not so with policy initiatives, which depend on happenstance, hence the need for a votational filter — albeit with all the flaws that you point out. Random selection for the judgment function leads to predictability, not so for policy initiatives — which would be highly unpredictable. The former is democratic, the latter anything but. Although votation reintroduces rational ignorance, unfortunately that’s the price we have to pay to ensure consent — and we can be confident that poor-quality populist policies will be rejected after deliberative scrutiny. This was the function of the Athenian legislative courts (ditto the UK House of Lords) and if we take this as our model at least we have some historical precedent that might give us hope that it would work. I’m sorry to repeat myself but there are no historical examples of sortition-only constitutions.


  51. Keith, you wrote, “Unfortunately democracy in practice cashes out as majority rule, with minority rights being protected by constitutional, judicial and other liberal safeguards.”

    I think that cumulative voting, where everyone has multiple votes rather than just “yes or now,” might go a long way towards reducing this problem. With cumulative voting, the basic voting principle changes from “winner take all” to “winners take most, but all substantial minorities take some.”

    I also think there are ways to add more deliberation into a process like the votation your propose.


  52. That’s interesting — how would cumulative voting be applied to votation, as opposed to proportional (party-political) representation? It’s also good to hear that votation could become a more deliberative process as hopefully that would go some way to meeting Terry’s concerns.


  53. The idea of using proportional voting to set an agenda is interesting (the majority selects a majority of agenda items, but minorities can select a proportionate share based on their numbers). There are more than a dozen proportional voting methods, with cumulative voting being termed a semi-proportional method (because it requires coordination by the minority to concentrate their votes on agreed on choices to win a minority of seats). Whether cumulative voting, ranked ballot, or range voting, etc. could be used on ISSUES rather than candidates is a controversial topic among election method experts…but something to consider.


  54. Keith, in cumulative voting, each voter gets multiple votes to split among multiple alternatives (whether candidates for office or candidate issues or bills). You can choose to divide your votes equally, or you can give more votes to alternatives that you feel more strongly about and less to others. Because of this, members of any kind of minority can, if they choose, concentrate their votes on agreed-upon priorities. They won’t win most of the decisions or hold the majority hostage, but they will win on some things – a huge difference, in my opinion.

    Terry, could you suggest any references about the pros and cons of using different proportional voting methods for issues or bills as opposed to candidates?


  55. This sounds like a very promising compromise — My ignorance on voting systems is total, so I’m glad to learn that this could apply to votation as well as conventional representation.

    The other issue that David mentioned was the possibility of making the voting process more deliberative. I’m a little sceptical about this, given that the maths would still gravitate towards rational ignorance — surely it’s an inevitable consequence of universal suffrage? The problem would be compounded by the fact that there would be a significant number of proposals on the votation paper (given a threshold of, say, 100,000 signatures), so I have a hunch the votation would be little more than a visceral reaction. But at least nobody could claim that they had not been consulted and, given the acquiescence of supporters of losing parties in elections, we can have some confidence that the winning proposals on the votation ticket would indicate the (uninformed) consent of voters.

    To my mind public (and media) deliberation would be more likely to occur before and during the deliberations of the AC. But this would feed back into the proposal-making process as those who made populist (but unworkable or unjust) proposals would (hopefully) be rejected at this stage and would suffer in the long term from a tarnished reputation. I appreciate it seems perverse to introduce rational ignorance only to have it overturned later, but I can’t see any other way of ensuring that proposals have a democratic mandate. That’s certainly the way it worked in Athens, where the role of the legislative courts (appointed by sortition) was to overturn Assembly decisions, so I think that’s about as good as it gets.


  56. David,

    Unfortunately I can’t refer you to any particular web site or document dealing with the difficulty of using alternative voting methods for policy decisions. In short, candidates are fixed (you can’t amend the candidate to make him more acceptable), but policy proposals are infinitely adjustable. This is why legislatures use deliberation and amendment with a long series of sequential votes on each element, rather than referendum style up/down voting. By controlling the precise wording of a proposal (or the number, or order of proposals on the ballot, etc.) an election administrator can significantly shape the outcome under most voting methods.

    Can you set out a hypothetical votation process? How long would a proposal ballot item be? Would proposals have a clear intent expressed (rather than merely being an open-ended agenda topic)? How much detail? It seems they would need to have specific elements expressed so voters could decide whether to vote no or yes. (For example a given voter might vote yes for a health care reform proposal if it stated that she could pick any doctor, but vote no if it was silent on that matter).

    But then, the legislative drafting/amending process might amend the proposal to remove the ability to pick your own doctor. So the voter will regret having advanced the proposal to that next stage.

    I just don’t get what the ballot would say.


  57. Terry, you wrote, “legislatures use deliberation and amendment with a long series of sequential votes on each element, rather than referendum style up/down voting.”

    That’s a very important point. A final up/down vote has “embedded” in it the answers to a series of prior questions, e.g.
    • “Is there a significant need for health care reform legislation?”
    • “If so, what should be the objectives and criteria that a bill should meet?”
    • a long series of decisions about features of each alternative bill (hopefully there will be more than one!)
    • “Which bill is best?”

    In your proposal, who takes the long series of sequential votes? An all-purpose allotted chamber? Single-issue citizen panels of limited duration? A combination of the two?


  58. Terry

    Those are all very fair points. I’m intentionally vague on details as I have no experience of working legislatures, so would defer ultimately to people like yourself. Votation/referendum issues do have to be short and punchy but then, as you rightly point out, can be watered-down or otherwise corrupted when turned into legislative bills. Two suggestions which might help resolve the problem:

    1) Winners of the votation would convert the proposal into a legislative bill with the aid of both parliamentary/congressional draftsmen and an AC. The job of the latter would be to ensure that the spirit of the votation proposal was retained. Given that it’s the same population that is being sampled in two different ways (votationally and descriptively) the AC would be the obvious candidate to ensure that the bill accurately reflected the spirit of the original initiative. Note: this would be an additional sortive body to the one that decided the final outcome, so I’m starting to be persuaded by your policy for multiple sortitions!

    2) Interest groups or political parties who deceived voters by deviating from the spirit of the votation proposal in the detailed legislative bill would lose credibility. This would mean their bill would almost certainly be rejected by the final-decision AC and that they would no longer be trusted in future votations. This principle alone might well be sufficient to retain the integrity of the process without suggestion (1) and shows the importance of considering the system as a reflexive whole, functioning over a longer time-span.


  59. David,

    My concept has four distinct kinds of bodies.

    A. A “Meta Legislature” that deals only with rules, process, ethics enforcement, overseeing staffing of other bodies, selecting some generic agenda items (recommended by experts, interest groups, and the like) to establish Interest Panels on — but no direct legislation as such. This Meta Legislature would be selected by lot– perhaps from a pool of people who have previously served on some other allotted body described below (so that they have some practical knowledge of what they are overseeing). They would serve longish overlapping terms, so that institutional memory is maintained.

    B. “Interest Panels” would include any and all volunteers seeking to contribute to a particular issue. They would be placed on panels by lottery to assure mixing. They would prepare first drafts of legislation. There should probably be a super-majority requirement to advance a bill to the next level. They would do work similar to legislative committee work. An Interest Panel could continue as long as they maintained a quorum of participants, or finished and voted out a bill.

    C. Deliberative Panels would be selected by lot, but because their work load is substantial, and might serve for months or a year, it would be relatively easy to excuse yourself from service. This means they would be less descriptively representative, and have some flavor of self-selection volunteerism. However, they would NOT get to select WHAT issues they were going to work on, so the self-selection will not be content-based. They would be assigned a content area by the Meta Legislature to review relevant legislation. They would review all of the proposals from relevant Interest Panels and prepare a final draft bill for the Allotted Decision Panel.

    D. Allotted Decision Panels would be the most descriptively representative bodies, and each would serve for mere days or weeks. They would operate as Keith has suggested…hearing pro and con presentations and voting by secret ballot without internal debate. They would be shielded from lobbyists (beyond the official pro/con presentations), with the equivalent of jury tampering laws. Perhaps they could also make motions to refer a bill back to a Deliberative Panel with specific objections, or to refer a bill to a general referendum.

    By the way, I need better names for some of these bodies, especially the last one.


  60. For the allotted decision panels, I’d suggest something like “policy juries.” These bodies would operate very much like the juries we’re familiar with, but they would be deciding about laws instead of criminal cases.


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: