Would sortition help against corruption?

Political parties and institutions in Australia and the US are increasingly dominated by interest groups representing the few, leading to a large policy-induced increase in inequality in recent decades and a long raft of new policies favouring the few by giving them the tax receipts of the many. One hypothesised solution to the problem of having these coalitions strangle the public interest is to make more use of sortition, ie to have randomly selected citizens decide more things in public life.

Currently, citizen juries are used in criminal cases, but one can in principle use sortition to select ‘policy juries’, to select members of parliament, to select local councillors and mayors, etc. It was quite normal in ancient Athens to do so, hence we know it can work in some circumstances.

The general pros and cons of sortition have been debated ad infinitum over the centuries, with the main recognised advantage that one gets more involvement of the citizenry and less coalitions within the set of the governors. The main disadvantage is that you get people deciding on things they are fairly ignorant about, which is a problem when long-experience and knowledge is called for. The ancient Athenians knew this tradeoff too: whilst they selected their councillors randomly, they were more picky when it came to their war leaders.

I want to think out loud just what is likely to happen to the problem of special interests in Australia in two different scenarios: if we’d select our MPs randomly, or if we’d decide on mayor policies via citizen juries.

The short answer is that the problems are likely to be worse with sortition. The long answer is over the fold.

                Random members of parliament.

Technically speaking, we could easily fill our parliaments with random citizens, subject to some minimal screening such that people who don’t want to do it or are incapable of doing it don’t have to. Random MPs would kill off the system of political parties we have at present and it would prevent special interests from having influence on whom enters parliament: when they come in, they would come in all eager and idealistic. But that is just step 1 of what would happen.

Since parliament selects the government, we’d hence have 150 random individuals having to decide who amongst them will be the PM and the ministers. Even if the citizens are well-intentioned, that will lead to a lot of bickering and lots of votes of no-confidence! But perhaps we’d re-jig those rules to make it harder for MPs to change their minds about who would form the cabinet. Still, the 150 selected MPs would quickly realise that they are competing with each other for desirable jobs. Step 2.

Since the members of parliament are by design one-term members, their interests in doing the best for the country versus doing the best for just themselves would be relatively low: at least long-run political parties have an incentive to keep up their brand-name, and individual MPs currently can hope to get back into parliament in future periods if their electorate approves of them. One-term MPs who are random members of the citizenry don’t have such constraints and should normally be expected to sell their votes to the highest bidder. This would also go for members of the Senate. Step 3.

The special interest group, meanwhile, still have the same interests and the same deep pockets as before. In stead of being able to intertwine with political parties wherein they form coalitions and beliefs that support them, they now face a different landscape, one where every sortition cycle they face a new batch of individuals with tremendous power, normal wealth levels, and no incentive to be honest.

Hence, given that parliaments decide on the allocation of roughly 400 billion dollars of tax-payer money per year, that is the amount that these one-off parliamentarians could make off with and allocate to the highest bidders. They surely wouldn’t steal all of it, but a large chunk should be expected to be allocated to the special interest groups, much like they are at present.

Why was this not a problem in ancient Athens? Well, for one, the ancient Athenians had nowhere to run with ill-gotten money. Other places didn’t have their language or culture, nor was the idea of tax havens widespread, so life wasn’t that pleasant if one outstayed one’s welcome in Athens. Perhaps more importantly though, there weren’t that many Athenians. Maybe 30,000 male citizens. That is small enough for a lot of social control to be present, i.e. for the individual ‘lucky’ citizen who finds himself temporarily in power to have a relatively large group of friends and family looking over their shoulders who will have had an incentive to keep the lucky boy in line.

Neither of these social control mechanisms are relevant in Australia today: given the fantastical amount of taxes to steal and the ubiquitous use of English and banking secrecy in other parts of the world, the constraints on random members of the public in parliament to stay honest are pretty low.

So in truth, I don’t think a parliament consisting of random members of the public is more likely to do the bidding of the population: their willingness to sell out to the organised interests would be even higher than members of long-standing political parties, since those parties have at least some incentive to keep the wheeling-and-dealing of their current members to an acceptable minimum.

Surely, having randomly allotted MPs would move many other factors, such as the relation between politicians and ministries, and the relation between media and politics. It is hard to know how all that would pan out, but the basic point remains: if you temporarily give a lot of power to randomly chosen individuals, they have even less incentive to do the right thing than individuals from within long-lived clubs who depend on a brand-name with the population.

 

                Random policy juries

What about random citizens being asked to adjudicate on particular policies then?

First off, it is far less clear how to actually do this than to have a random selection of MPs: for which policies would one form citizen juries and for which not? Would decisions to go to war be subject to 20 random citizens? How about levels of taxation or the nitty-gritty of health-and-safety laws? There is no easy rule on this in sight, nor is it clear who would have the final say on what things to pass onto a citizen’s jury. After all, why would, say, parliament or government assist to give some of its own powers to citizen juries? And if government and parliament is not going to decide which policies to pass on to citizen juries, who then?

So right from the get-go, the idea of citizen juries for particular types of policies seems a dead duck from an organisational point of view.

However, let’s plough on and suppose that we manage to embed the idea of citizen juries into ministerial and parliamentary decision making: suppose we manage to have citizen juries formed only for the specific job of deciding yes/no on policies already voted upon by parliament, as a kind of additional check on the system (perhaps in lieu of the Senate).

Since the jury is temporary and facing the full glare of the nation, you’d probably have quite motivated individuals and they would be hard to bribe because it wouldn’t be so easy to get to them and offer them a bribe in the short time they have to convene. So Step 1 is, once again, looking bright from an anti-corruption point of view.

Then the question of course is who is going to feed that jury the information needed to make decisions. On anything as complicated as laws and policies, which often involve hundreds of pages of legislation and tens of thousands of pages of background legislation and documentation, the question of who gets to feed the citizen juries their summaries will be crucial: whomever gets to whisper the ‘soundbite version’ to the citizen jury effectively controls them. That is the crucial Step 2.

In the best-case, it would be a job for ministries to feed the citizen juries the information they ‘need’ in order to make a decision, much like courts currently oversee what information reaches juries. In turn, this makes the ministry more powerful than before. Crucially, though, the ministry itself is doing the bidding of the politicians.

So, would citizen juries be harder to co-opt for those special interest groups? No! It would add another hoop to their task, for sure, but not one that seems awfully high or narrow: since the special interest groups must already co-opt the top of ministries and of politics in order to get their way currently, feeding their preferred slant to a citizen jury via the ministry wouldn’t take much additional effort at all.

How about other means of feeding them information, such as just leaving it up to the jury to gather their own information? Well, for one, the longer the jury can roam about, the easier it is to get to them and bribe them. But more fundamentally, legislation and policies are very particular tasks on which most of the expertise is by design in the state system that has lead to legislation and policies. So I expect these juries to trust the system on which they rely for most of their information, rather than the millions of punters disagreeing with each other outside of that system.

Again, citizen juries would change the nature of the game in many directions that are hard to foresee, but, again, the basic point that they are amateurs relying on the information fed to them by a state machinery means that their presumed independence is fake.

 

                Summing up

I used to be quite charmed of the idea of citizen juries for policies and even for deciding on who would be in parliament. It sounded so democratic, such an elegant solution to the problem of special interest groups worming their way into our democratic institutions. It seemed like a magic solution for hard problems.

On reflection though, I find myself on the side of Edmund Burke and Socrates, who both denounced the idea as silly and unworkable. I agree with them: it is hard to see what use small random groups of citizens would be for policy-making in modern Western institutions.

This entry was posted in bubble, Cultural Critique, Economics and public policy, Ethics, Geeky Musings, History, Information, Law, Libertarian Musings, Life, Philosophy, Political theory, Politics - national, Public and Private Goods, regulation, Social Policy, Society. Bookmark the permalink.

64 Responses to Would sortition help against corruption?

  1. Paul,

    I’m happy to try to look for the downside of deliberative democracy. The fact that I’m so strongly in favour of it means I’m on the lookout for ways I might be kidding myself.

    But I really can’t credit your idea of the motives of the chosen few. Assuming pure, unenculturated self-seeking from the citizens seems really wide of the mark to me. Most people try to do a good job of what they’re doing. They ‘get’ the trust that’s being put in them. Ask people on a jury what they thought of those they were on it with. Many will say they can’t believe how undisciplined they were in addressing evidence, how uncomprehending they were of the judge’s instructions, but very few are nakedly self-interested.

    When you look at documented experiences with the outcomes of citizens’ juries and surveying people who’ve been in citizens’ jury situations you find how flatterd and uplifted they are by the opportunity to give the citizens voice, how much they try to do the right thing.

    We’ve had another quite strong test of this with the randos in the Senate recently. It was cleary in the interests of several of them to avoid a double dissolution, but many didn’t (did any?).

    I think your hypothesis of self-seeking makes more sense in the party system simply because the party system has accommodated itself to political realities for a couple of centuries now. And parties are ruthlessly self-seeking and, because of that can impose pretty ruthless discipline on their representatives through the channel of career incentives.

    Having said all that, I wouldn’t support simply going from one system to another one. I’m a (Burkean) conservative in that sense – suspicious of my own or anyone else’s knowledge about how such a wholesale thing would turn out. But we could take a leaf out of the democratic theorists of the eighteenth and nineteenth century who sought to balance different perspectives – for instance in different houses of parliament/congress often with elaborate mechanisms for managing a deadlock (why we’re even experiencing some of their thinking in that regard as I write these words.) So the way I see it, we should be trying to think how (say) a Senate chosen by sortition might add to (or detract from) our democracy with government still made and lost from the House of Reps. I confidently predict that such a setup would have rejected the cockamamie climate change policy of the current government – moving from a low to high-cost means of delivering a target.

    I could go on, but I hope to write something on this for publication in the not too distant.

    As an aside, Socrates was a thoroughgoing anti-democrat (having probably been a supporter of two extremely brutal Spartan backed putches run by his aristocratic students and relatives – which is probably why they indicted him). Edmund Burke was writing at a time when the vast majority of the educated thought democracy would, in no time degrade into the unworkable rule of the mob.

  2. Alan says:

    I am sure Themistokles, who ended his days at the Persian court squandering both his ill-gotten gains from his time as the main leader of the Athenian polity and those paid to him as bribes by Artaxerxes I, would have been astounded to learn that Athenians on the take had nowhere to go. Those of us who are content to do a little research are aware that a considerable number of ancient Greeks, far from having nowhere to go, ended their days in other cities, or at foreign courts.

    I am equally sure that one can conveniently denounce sortition by claiming that the prime minister and cabinet would have to be drawn from a sortitive assembly. That was not the case at Athens or any of the other Greek cities, so there is no reason to imagine that it would apply in the twenty-first century.

    • yes, I was aware of some of the leakage in those ancient times, but for the reasons given in the blog, I think that social control was much stronger then than it is now. Indeed, Rome hardly made use of sortition, perchance precisely because the world had already become a much bigger place by then!

  3. Hi Nick,

    yes, I knew you had a different opinion than me when I wrote it and happy to hear you are writing more on this. My motivation for the above post was that I too have just written a piece on how useful randomisation could be for allocating prizes in our economy (grants, jobs, student places), which forced me to think where I thought the limits were to the usefulness of randomness in public life.

    The key bit we disagree on is on motivation. You say

    “When you look at documented experiences with the outcomes of citizens’ juries and surveying people who’ve been in citizens’ jury situations you find how flatterd and uplifted they are by the opportunity to give the citizens voice, how much they try to do the right thing. “

    Sure, it’s a common finding that people feel good about participating in experiments of short duration with high public exposure. And of course people will try to do the right thing if that is expected of them and there’s no obvious way to benefit from doing the wrong thing. But does that tell us what the system would look like if it becomes normal? Hardly. Vested interests have no incentive to interfere with short-lived experiments on things that dont matter, but they have every incentive to interfere with sortition if it becomes the norm.

    Random citizens in parliament would be there for several years. They wouldn’t be the highly motivated ‘randos’ that you see in parliament currently (and that have constituents!), but average members of the public. And they wouldn’t be as relatively powerless as the current set of randos, who are dominated by the established party members: they would be running the country. To expect them to display the enthousiasm and integrity you see in juries in criminal cases or in ‘citizen democracy’ experiments that last a week is naive. Power and money corrupt, and 400 billion a year going through their hands will corrupt very quickly, particularly if thousands of lobbyists and advisors will be telling them how to do it and that it’s not really corruption.

    I agree that political parties are ruthlessly self-serving, but at least they are there for the long haul and hence have quite strong reputational incentives. Random members of the public don’t even have those. And, to be fair to federal politicians, I would tend to think that they are at least as ‘pure’ as random members of the public when they get into the game.

    Your proposed proposed system of government relies a lot on the innate and persistent goodness of those who would find themselves in power by accident. Goodness is not a good thing to be relying on when it comes to the use of power. We wouldn’t need any separation of powers otherwise!

    I still like the idea of more participation of the citizens in our democracy, but just don’t see mileage in sortition.

    • Alan says:

      Why would random citizens be there for several years ? Most proposals I haves sen suggest very short non-repeatable terms such as 3 months or 6 months. Again the classical and medieval sortitive republics almost always imposed very strict term limits so why would a twenty-first century sortitive assembly be different? We are surely not in the business of electing random straw men to a sortitive assembly.

      I will see you Rome, in your answer above, and raise you Venice. Both republics operated across the Mediterranean basin and the Black Sea. Rome never used sortition even when it was a strictly local affair dwarfed by the surrounding Etruscan and Greek cities. Venice used a rather elaborate system of sortition to elect the doge for around 600 years.

      • Why not a shorter term in my first hypothetical? Because as the term shortens from the length of the current parliament to a few months, it starts to look more and more like …… my second hypothetical, ie a citizen policy jury!

        I will challenge the relevance of your ‘Doge of Venice’! You see, its not sortition in the sense of letting fate decide who governs. Yes, a fairly complicated system of randomness was used to select the people who elected the Doge, but all the decision makers who ended up doing the selecting were still from the Great Council, i.e. the aristocrats of Venice. And the Doge was still elected by those aristocrats, though only if he was the preferred candidate of rather a lot them.

        I will grant you that its a pretty spectacular example of an odd voting system, apparently designed to force everyone to agree on an acceptable outside focal point (such as the oldest shrewd person they could find).

        For the readers, from Wikipedia:

        New regulations for the elections of the doge introduced in 1268 remained in force until the end of the republic in 1797. Their object was to minimize as far as possible the influence of individual great families, and this was effected by a complex elective machinery. Thirty members of the Great Council, chosen by lot, were reduced by lot to nine; the nine chose forty and the forty were reduced by lot to twelve, who chose twenty-five. The twenty-five were reduced by lot to nine and the nine elected forty-five. Then the forty-five were once more reduced by lot to eleven, and the eleven finally chose the forty-one who actually elected the doge. None could be elected but by at least twenty-five votes out of forty-one, nine votes out of eleven or twelve, or seven votes out of nine electors.

        wow, what a system!

        • Ken Parish says:

          I don’t want to be a party pooper, but note that random selection/sortition would be unconstitutional at federal level. Sections 7 and 24 require both Senators and Reps members to be directly chosen by the people. A referendum to change this would never pass, nor should it. Those sections are also the foundation for constitutional protection of the right to vote and freedom of political communication and association.

          However, you could legislate for a few random MPs at state level without constitutional problems. it would be quite an interesting experiment.

          Personally I prefer multi-member seats elected by Hare-Clark PR as the best way to get more diversity in our Parliaments and therefore hopefully more deliberative discourse and constructive compromise.

        • paul frijters says:

          Hi Ken,

          of course we are talking about extremely unlikely hypotheticals taht are politically difficult to imagine. The blog in that sense wonders whether sortition could be useful in principle. I used to think ‘yes’, but now I see it more as a red herring.

          • Alan says:

            I do not see sortition as either a red herring or a universal solution to all known problems including the common cold.

            I think there is a case for a small number of random MHRs on short and unique terms, say 3 months. They would not vote except on tied votes, procedural motions and electing the speaker. They would have a suspensory veto on some matters, like accountability legislation in general and the electoral act in particular and an overridable veto on secondary legislation. They would be disqualified from parliament for 3 years after their term ended.

        • Alan says:

          Ken

          I agree with you that STV multimember districts as solution to many of our existing problems. I don’t see that as a reform that necessarily excludes the addition of a small number of random MHRs. Since we are in the land of George Bernard Shaw’s why-not we may as well go the whole hog.

  4. Moz of Yarramulla says:

    I’m not sure why you assume jurors shouldn’t be trained, that does seem like the most obvious way to improve their ability to make good decisions. Even if by having them observe their predecessors for a period, but ideally by having time to study the requirements of their position. As we have seen particularly with Attorneys General, it is desirable that specialist positions within government have specialist training.

    I wonder whether replacing or leavening the Senates with randos might help. We get a few of them now, and I think they do help, if only by preventing one party having control of both houses. Making that explicit by requiring that 1/3 or 1/4 of the members be jurors might be better than the current system. I think that at the very least, having a broader selection of the community in there would help – random selections are unlikely to give us 80% rich, straight, white, men. Which would bring wider perspectives, if nothing else, and hopefully reduce the amount of anti-majority bias in some parts of the system (if you ever want proof of the gulf between law and justice, look at the percentage of legal professionals who would recommend to a rape victim that they were close to, that she involve the legal system).

    Finally, could you clarify how your arguments don’t also apply to voters? We let a wide variety of people vote, and for a single voter the benefit of accepting an inducement vastly outweighs the loss of voting according to the inducement. For that matter, one advantage of online voting is that it will be easier to prove that someone did vote per instructions, making corruption more efficient (from the payers). Ubiquitous smartphones make that possible now, but online would make it easy.

  5. ChrisB says:

    And Florence –
    Florence was governed by a council called the signoria, which consisted of nine men. The head of the signoria was the gonfaloniere, who was chosen every two months in a lottery, as was his signoria. To be eligible, one had to have sound finances, no arrears or bankruptcies, he had to be older than thirty, had to be a member of Florence’s seven main guilds (merchant traders, bankers, two clothe guilds, and judges). The roster of names in the lottery were replaced every five years.
    The main organs of government were known as the tre maggiori. They were: the twelve good men, the standard bearers of the gonfaloniere, and the signoria. The first two debated and ratified proposed legislation, but could not introduce it.To hold an elective office, one had to be of a family that had previously held office.
    There were further complications, but the main point of the analogy is to clarify that however many random elements were introduced the city was still ruled by the Medici fiddling all of them.

    • Alan says:

      The Florentine republic was founded in 1115 when Florence overthrew Matilda of Tuscany. The Medici did not become significant until 1434. 319 years is rather a long time for an electoral system to persist and a longer time than any of our contemporary system of competitive elections has lasted.

      Moreover, the Medici seized power by the simple expedient of stuffing the purses used for the random selection of officials. It is no different from stuffing ballot boxes in a conventional election. It proves not that sortition is peculiarly vulnerable to bankers who wish to become hereditary autocrats, but that bankers with ambition will do what they must to seize absolute power.

  6. ChrisB says:

    On the other hand, random election/appointment to purely ceremonial positions, such as Head of State, makes obvious sense, and I’ve proposed it before.
    It is now generally accepted that
    • The majority of Australians want an Australian to be head of state:
    • The majority of Australians are averse to any very large changes in our constitutional arrangements;
    • The web of accumulated, derived, and developed constitutional roles and functions of the crown in Australia cannot be transferred from the royal family unless they can be defined, and any attempt at definition brings forth such disagreement as to split the parties into mutually repugnant and uncooperative camps.

    This being the case, only the arrangements I propose can square the circle and allow Australians to have what they agree they want.

    1. Australia shall be ruled by a titular monarch.
    This permits all existing constitutional structures, understandings and conventions to be carried on unaltered, with the governor-general standing in for the monarch and his or her powers and duties to remain as they have been, whatever that might be, with all existing ambiguity and uncertainty retained unaltered.
    2. The monarch shall be chosen by computer by random selection from all persons on the Australian electoral roll born on a randomly chosen day .
    This ensures that
    a) the monarch will be an Australian citizen and
    b) the election or appointment of the monarch will not cause divisions among the populace.
    3. The identity of the chosen monarch shall remain in the custody of the computer, neither the governor-general, the government, the public, or the person concerned being informed.
    This means that
    a) every Australian could not only aspire to being King or Queen, but every 365th Australian could believe that they might already be King or Queen – producing that pleasant tickle existing at the back of the mind in the time between buying a lottery ticket and the draw, only indefinitely prolonged for no expense
    b) the person chosen would not be stressed by sudden fame or corrupted by unexpected power.
    4. The only possible objection to this plan would be that as the law now stands the monarch cannot be tried in his or her own courts, and unless appropriate arrangements were made every 365th person brought into court could plead that as it could not be proved they were not king or queen the matter would have to be dismissed; this defect could be cured, however, by introducing a constitutional fiction – the only significant change in the constitutional fabric required by my scheme – to the effect that Australian citizenship involves the waiving of any rights under this head.
    Any nation that can give a real Queen an imaginary birthday should have no problem giving a real birthday an imaginary Queen.
    While it might be objected that this proposal is ludicrous, its great merit is that it is considerably less ludicrous either than the existing system of privileging the heirs of Guillaume le Conquerant or the alternative proposal of going through all the trouble and expense of electing a president empowered to do no more than open fêtes.

  7. paul walter says:

    Yes, just let me at ’em. I’d sortition them like they’d never need sortioning again.

  8. Jon Roland says:

    This post seems to assume direct random selection from the general population, when historical sortition systems have never worked that way. They alternated sortition with a sceening stage, a system I call fetura, from the Latin word for breeding, which involves such alternation. The systems of Venice and Florence are examples of this. I have one, as part of a model constitution.
    1. Selection of officials is multi-stage, beginning at the most local level (wards), but those selected at random are not limited to selection themselves for the next round.
    2. As for policy juries, there would be more than one on any policy, and policies would go through the same kind of fetura, using competing juries that would decide how to present the policy options to one another. Some such juries, which I call inquestries, would function like grand juries to empower prosecutors to propose and defend policies to further juries.
    3. Corruption of members would be be inhibited by audits of the members of one jury by another. There would have to be full disclosure of assets and income. That does not prevent payoffs to cronies, but the same process could be used for that. Everything checks and balances everything else.
    4. One inhibition to corruption would be that selection is not confined to the selectees of the last random round, so that reputation for integrity would be important. One might call the system reputaracracy (combining Latin and Greek). Everyone would have an incentive to acquire and maintain a good reputation, because not all selection is random. Our modern system of credit scoring is an example of this, although it can be manipulated.
    One can describe the system used by political parties in their nominating process as a kind of fetura, although it is vulnerable to dominance by cult leaders.

  9. Your objections to random members of parliament are persuasive and indicate the idiocy of simply replacing elections with sortition. As for random policy juries you’re right to emphasise the problem of who is going to feed the jury the information needed to make decisions. But your answer (the government) overlooks the 20-year research programme into deliberative polling which endeavours to give equally-balanced information on both sides (as in a judicial court and the Athenian nomothetai) and also presupposes large juries (c. 300) in order to ensure accurate descriptive representation and to protect against corruption. So this part of your article is aimed at a straw man.

  10. Jon Roland says:

    Historic systems have not been simple direct sortition from the general population. Here are two examples:

    Venetian system

    New regulations for the elections of the doge introduced in 1268 remained in force until the end of the republic in 1797. Their object was to minimize as far as possible the influence of individual great families, and this was effected by a complex elective machinery. Thirty members of the Great Council, chosen by lot, were reduced by lot to nine; the nine chose forty and the forty were reduced by lot to twelve, who chose twenty-five. The twenty-five were reduced by lot to nine and the nine elected forty-five. Then the forty-five were once more reduced by lot to eleven, and the eleven finally chose the forty-one who actually elected the doge. None could be elected but by at least twenty-five votes out of forty-one, nine votes out of eleven or twelve, or seven votes out of nine electors.

    Florentine system

    Florence was governed by a council called the signoria, which consisted of nine men. The head of the signoria was the gonfaloniere, who was chosen every two months in a lottery, as was his signoria. To be eligible, one had to have sound finances, no arrears or bankruptcies, he had to be older than thirty, had to be a member of Florence’s seven main guilds (merchant traders, bankers, two clothe guilds, and judges). The roster of names in the lottery were replaced every five years.
    The main organs of government were known as the the maggiori. They were: the twelve good men, the standard bearers of the gonfaloniere, and the signoria. The first two debated and ratified proposed legislation, but could not introduce it. To hold an elective office, one had to be of a family that had previously held office.

  11. Naomi says:

    “their willingness to sell out to the organised interests would be even higher than members of long-standing political parties”

    Which is why a very different parliamentary structure would be needed. There are so many wonderful shades of grey between a conventional parliamentary structure and ad-hoc policy juries. Have a pool of elected officials who make proposals. Also have a short-term statistical sampling of the people which would be responsible for voting these proposals up or down after listening to the for/against arguments made by the elected officials.

    Proposals can’t be tailored to suit the whims of individual randomly drawn members if the sampling is generally drawn after proposals are introduced. Selling votes to the highest bidder would be difficult if they use a secret ballot.

    There’d still be elections and parties. But no party would have control of the government at any given moment. Everything a major party wants to do would effectively be subjected to a jury trial. No party would be free to implement its platform entirely and no party in parliament would be shut out completely.

    And there’s certainly no need for the PM and the cabinet to be selected from the sampling.

  12. paul frijters says:

    John, Keith, Naomi,

    glad to see you take the post in the deliberative spirit in which it was intended! It is useful to mull over the shades of grey Naomi talks about.

    John talks of pre-screening of candidates for assemblies and juries, presumably using recognised signals, like whether someone works for the public service or holds a good job. He seems to put some faith in multiple juries checking on the members of the other ones, deciding how to present cases to each other, and in the reputational incentive coming from the possibility of being selected more than once for a jury.

    If you’d want to prevent random members selling out, you’d need to select people from a pool that is international in the sense that they’d care about their international reputation. That is a bit like government-by-international-experts. Whilst I see the benefit of random expert fly-ins to decide on an ad-hoc basis on things of great local importance (like property rezoning), this is different from effectively handing over sovereignty from the population to the experts of this world deemed worthy to survive John’s fetura. It wouldn’t have legitimacy, which is also the obvious problem with the systems of Venice and Florence. Hence I think John oversells the degree to which we could use (or want to use) pre-screening.

    Similarly, I think juries watching other juries is not an obvious improvement on thousands of voters and a dedicated opposition watching the existing members of parliament.

    Naomi’s plans to have officials make proposals to random members of the population sounds pretty much like the policy jury talked about in the post. It would be experts putting cases in front of amateurs, and it needs a pre-screening of proposals by the system. Doesn’t sound very workable to me.

    Keith seems to have rather large juries in mind for policy proposals. Given the high volume of legislation currently going through parliament, we could be talking a lot of people then. It would need a large bureaucracy to set them all up and have them meet on time, etc. Not easy to see how that could work, but maybe Keith can elaborate?

    Also, as a general comment, I think that in order to appeal to the population as a feasible alternative to the current arrangements, any system of sortition that is championed must be easy to understand and its benefits must be easy to see as well. Randomly selected members of parliament is easy to understand, systems of multiple juries feeding each other information and checking each others bank balance is less easy to understand and thereby less likely to capture the imagination.

    • What Naomi suggests makes a lot of sense to me.

      Paul, here is a narrower suggestion along the lines of what Naomi says (and which I first put forward in published form in articles in 1998). See if you disagree.

      The power that legislatures now have to legislate contrary to the informed views of the people (contrary to the views a majority of the people would have were they/we to become well-informed about the law in question) is an undemocratic power that would not exist in a well designed system of legislative democracy.

      When a law is passed by a majority in parliament it can go to a legislative jury (one large enough to be a statistically accurate miniature of the public, with stratification used to further ensure such representativness) for review. The parliamentary majority that passed the law can present the case in favour of the law to the jury (or someone they elect/choose to do so can be in charge of conducting the case in favour). Who would conduct the case against is I think a trickier question but one option (not necessarily the one I think best) is that the members of parliament who voted against the law will elect/choose someone to be in charge of conducting the case against it. Both sides will be given equal time.

      The legislative jury will hear the case and when ready to make an informed decision will do so. This might take a matter of days, or it could take weeks. Jurors should be paid so that all can afford to serve.

      The hearing given by the legislative jury needs to be well designed to ensure an informed result, and that the yes and no sides will be on a level playing field. A “democratic design directorate” (or whatever we may call it) can be tasked with working out the best possible design, and improving on it over time. This directorate can be overseen by one or more directors chosen by jury, and can be limited to advisory power, with all final decisions about design being made by jury. The point of this is to have a design process that is highly democratic, well informed, and fully independent from politicians, political parties and special interests. Politicians must be excluded from the design process because of the conflicts of interest they have, in particular their interest in maintaining their monopoly over lawmaking.

      This is a sortition/jury analog to the veto referendum, and is in my view far superior to it. (I do not by the way think that a jury analog to the veto referendum is sufficient to make lawmaking democratic. I also believe we need a jury analog to the ballot initiative, as I proposed in articles in 1998 and since, and some other sortition-based reforms as well.)

      (I am opposed to jurors serving for terms of longer than one year, and do not think choosing MPs by random selection rather than popular vote is a good idea.)

      • paul frijters says:

        Hi Simon,

        I see you have put a great deal of thought into it. So let’s explore your proposal, which is pretty close to the second hypothetical in my blog in that its an ex-post citizen jury whose role is very similar to that of the Senate, somewhat akin to a referendum.

        What you describe is a large jury with considerable resources at its disposal to create a design directorate, and that in principle works like a court, in the sense that it invites proponents and adversaries of a view to make their case, deciding yes/no on a policy.

        Some preliminary remarks:
        1. Legislation on important areas now is seldom stand-alone because we have integrated systems requiring changes in many of them if we change something small in one spot. A law about schools is for instance likely to involve urban planning, property rezoning, union involvement, commonwealth/state negotiations, changes to teacher training, the system of pupil allocation, etc. The interdependencies make a lot of legislation complicated, voluminous, and not amenable to yes/no decisions. The alternatives to proposed legislation will often be more slight alterations to various parts of the proposed legislation rather than something completely different. This fits a Senate-type set-up where legislation before it can be amended, rather than a jury-type set-up that just says yes or no.

        2. Part of the job of government is to horse-trade and thus combine choices for which there is no majority in favour individually, but for which there is a majority when bundled. This coalition-building job would be frustrated if every single item in a horse-trade would be subject to a yes/no jury vote. So you don’t want to make it compulsory that everything that in principle could be stand-alone is treated as stand-alone.

        3. Consideration 1 and 2 combine to make it very hard to see how we could get a workable rule as to what goes to citizen juries and what does not: threatened by citizen juries, government and parliament will use either the ‘too complicated’ card, or else the ‘we’ve made it part of a large package’ card to prevent its decisions being questions. The only counter-move would be if the decision to go to a jury is out of the hands of government and parliament, such as perhaps triggered by a certain number of citizen signatures. This in turn means that citizen juries will be rare, as their triggering has to be outside of the regular government/parliament machinery: if it were inside, it would be circumvented.

        4. If I think about clear yes/no decisions, such as going to war, selling something off, adding something to the constitution, a change in the legality of something (gay rights?), or buying something big (like submarines!), then we are talking about subjects on which there will already be many commissions, reports, parliamentary inquiries, etc. There will be large constituencies in favour and against: the issue will be divisive and long-running. What does yes or no mean in such a situation? Would we get the same thing sent to a jury every year? If I think of something like gay marriage, the likely answer is ‘yes’. And would that be an improvement on the current system where the issue flares up in parliament and, after so many years, the population is pretty well involved and most people have a considered view? The answer has to be ‘not much’.

        5. Consider now the kind of decisions you’d ideally want a citizen jury for: allocation decisions where a lot of public money is spent and there is frantic lobbying by special interests to get their way. That’s where the corruption is and where the current system is failing us. But these are exactly the deeply embedded decisions, parts of large packages, like the budget and whole ministerial plans. They are the compromise decisions, involving horse-trading across regions and constituencies throughout the country. Very hard to see how one could siphon off something manageable for an ad-hoc jury with the ability to gain independent expertise to circumvent the special interests.

        So perhaps we should think of this the other way around: think about what we want to prevent or highlight, and then devise a system that does that.

        For instance, whilst yes/no decisions on legislation seems the wrong thing to put in front of citizen juries, it would be useful to have signals about corruption, and juries can generate those.

        Hence, in stead of asking citizen juries to decide for or against legislation, we can perhaps set them up to judge whether a whole set of decisions and legislation is in the benefit of the general population or serves special interests. This would be not too dissimilar from the function one would want journalism to have, but one can argue that the current business model of journalism has made it too dependent on private funding to be able to perform that function.

        Similarly, one can perhaps think of citizen juries as an alternative to Royal Commissions, where the juries would not be charged with judging legislation, but rather with sending out a signal as to what future programs/legislation they want to see. A kind of a citizen wish-list. These would be alternatives to current forms of inquiries that are easy to circumvent by the executive. Still, it would require careful thinking about how that could be made to work and, in general, I think some kind of democratic trigger outside of parliament would be needed at the minimum. Something for Keith and Naomi to explore perchance? Are you their supervisor or something, Simon?

        • I know Naomi and Keith through the Equality by Lot blog. We have not met, though my guess is we eventually will.

          I’ll reply to your interesting comments when I have a chance.

        • Paul makes some very interesting points on the difficulty of putting theoretical proposals (like Simon’s and my own) into practice.

          The only counter-move would be if the decision to go to a jury is out of the hands of government and parliament, such as perhaps triggered by a certain number of citizen signatures.

          That’s an interesting idea, and similar to my own proposal for direct-democratic initiatives, although I would be hesitant to base it purely on the number of signatures, due to the ease by which pressure and lobby groups (and media) can gain this level of support. That’s why I combine a 100,000 signature threshold with an annual public votation (an instrument of Swiss democracy), in which all citizens make their choice from those proposals that have exceeded the threshold. My proposal (for a public enquiry with balanced information and advocacy and a large jury) would certainly work best for referendum style decisions (should Britain remain in the EU?) and the result should be binding for a generation.

          Part of the job of government is to horse-trade and thus combine choices for which there is no majority in favour individually, but for which there is a majority when bundled.

          Most of us kleroterians view that as part of the problem. For example Tony Blair’s New Labour was able to smuggle in a whole load of social and cultural reforms that would have been otherwise opposed by their core voters. What is wrong, in principle, with disaggregating these bundles? Why should socially-conservative working-class voters have to accept multiculturalism, gay marriage and high levels of immigration because this was the only way of returning a government that was felt to deliver policies in their economic interests?

          PS Simon is a lawyer, Naomi a research chemist and I’m a student in political theory. We have never met.

        • paul frijters says:

          Hi Keith,

          Ok on triggers for juries. What one wants is a hurdle that needs widespread popular support to jump.

          Part of the job of government is to horse-trade and thus combine choices for which there is no majority in favour individually, but for which there is a majority when bundled.

          Most of us kleroterians view that as part of the problem. For example Tony Blair’s New Labour was able to smuggle in a whole load of social and cultural reforms that would have been otherwise opposed by their core voters. What is wrong, in principle, with disaggregating these bundles?

          What is wrong is that disaggregating bundles means you make it impossible for winners and losers of parts of the bundle to negotiate with each other: the bundle is like a negotiated outcome and if you unbundle you are essentially outlawing trade.

          The core issue is that something can be in the public interest without commanding a majority because the winners feel more strongly about it than the losers. By bundling several things, the total package represents a win for a larger group.

          Simple example: Mary feels passionate about gay rights and doesn’t like a new highway or more free-market labour laws, but doesn’t care that much about those two. John feels passionate about the new highway, but in principle weakly opposes gay rights and freer labour laws. Robin feels passionate about new labour laws and weakly opposes gay rights and new highways. So each of the three innovations (gay rights, labour laws, highway) is opposed by a majority if it came to a vote just on that. But as a bundle, all three voters support the combination of the three new policies because they all three benefit.

          The example illustrates the benefit of horse trading, deriving from the fact that government is about maximising utility and not about following the majority of the people on single issues. If you like, its about preferences over all combinations of possible policies, not preferences for single-item proposals. Another way to put it is that you want to allow trade in the political sphere between groups wanting different things, and bundles represent trades.

          • Alan says:

            I would think a deliberative body, such as a citizens assembly, eminently qualified for precisely that kind of coalition-building in a way that our highly agonistic political parties and legislatures are not. A deliberative citizens assembly (particularly with Keith’s adversarial structure) is certainly a better way to resolve such issues than a referendum or initiative. It may be that rights questions should be dealt with differently to economic and planning questions.

        • Finally returning to this thread 8 days on.

          I don’t support the petition or signature requirement of the initiative and veto referendum, preferring instead that juries fulfill that role (deciding if a citizen-proposed law or a government-passed law is referred to the ballot or a legislative jury). Yes I see legislative juries as an improved version or better analog of a veto referendum and initiative vote.

          Re 1 (a law about schools which might well involve a whole slew of things). The Utah legislature passed a law establishing a voucher system for Utah schools. The law was defeated and stopped in a veto referendum. There is no problem with it being stopped that I can see. (I would prefer the veto be exercised by a legislative jury than by a signature drive and referendum vote, but there is no problem I can see with the public being able to veto government legislation on schools, at least not on the informed and democratic basis of a legislative jury doing it).

          Re 2 (horse trading). I don’t see any inherent virtue in laws based on horse trading. For example if a conservative religious party (perhaps in Israel) is able to get a conservative religious law (that most people oppose) by horse trading, I don’t see the benefit of such a law being immune from a veto of the people.

          If a law that has resulted from horse-trading is one which most people would support overall (even though most people would oppose each of the horse trades by itself) that has a certain democratic quality to it, and a legislative jury might well approve it. (I do think a legislative jury needs the power to break conglomerate laws passed by the legislature into their individual topics, but I’m willing to entertain the idea that a jury might sometimes prefer not to in order to protect a “horse trade law” that results in something that is overall what most people want.)

          3 (government trickery in the laws they pass). Every law passed by parliament or a legislature needs to go to a legislative jury to ensure that it is not contrary to the informed judgement of the people. One effect of this will be that legislatures will have an incentive to pass laws that have a good chance of being in accord with the informed judgement of the people. Playing games to try and sneak things past the jury that are contrary to the informed judgement of the people is likely to backfire as it will tend to annoy people, and also the jury needs the power to break “conglomerate laws” into their separate parts/topics.

          Re 4. Every decision of a legislature is a yes/no decision in the end. If the laws a legislature passes go to a yes/no decision of a jury I don’t see a problem. It just means the final yes/no decision will be more democratic. (I was just talking about legislative juries having the final say on laws passed by politicians. The system could be designed to prevent the government from continuously passing the same law in hopes they can finally convince a jury to support it. The arrangements and procedures for legislative juries need, as said, to be decided independently from politicians.)

          Re 5. The point of legislative juries is to ensure the supremacy of informed rule by the people over rule by politicians. This is basic democracy. One of the more specific reasons for this is that politicians have conflicts of interest (re election rules, paying off their financial backers, government secrecy, whistleblower protection, concentration of power in themselves, for example).

          Thinking of the U.S., I don’t see how it would have been a problem for example for a legislative jury to have blocked the repeal of Glass Steagal or the Bush tax cuts for the rich or the billions being spent to “modernize” the U.S. nuclear arsenal and develop “miniature” “battlefield” nuclear weapons or privatized prisons.

          I am not in favour of juries being limited to the role of being “talking shops” and only advisory, just as a few hundred years ago supporters of parliamentary democracy did not see a parliament limited to advising the king as what they were going for.

    • >experts putting cases in front of amateurs, and it needs a pre-screening of proposals by the system. Doesn’t sound very workable to me.

      If it works for judicial trials then why not the trial of legislative proposals before the High Court of Parliament?

      >Keith seems to have rather large juries in mind for policy proposals. . . It would need a large bureaucracy to set them all up and have them meet on time, etc. Not easy to see how that could work, but maybe Keith can elaborate?

      As a political theorist (currently finishing a PhD on sortition) I leave the grubby details for others! My proposal is for a large number of ad hoc juries, who would serve only for the term of their own “trial”. I guess it would work best for big-ticket items, with statutory instruments (the vast bulk of bread-and-butter issues) nodded through via the usual channels. I proposed a similar mechanism recently (a judge-led public enquiry with adversarial exchanges and a large jury determining the outcome) as an alternative to Britain’s EU referendum which is more likely to be decided by whether or not voters like Boris Johnson’s new haircut. The main point that Naomi and I would like to make is that it’s not a case of either/or — a complex mix of sortition, election and direct democracy will be required for a well-functioning system that is perceived to be democratically legitimate.

  13. Jon Roland says:

    Juries watching and checking other juries, and passing on proposals to them, is better than what we do now because the juries would be encouraged to be focused, with specific propositions to decide on, confined to a room, with no distractions, until they make a decision. If that were not better, then why not have guilt or innocence be decided by asking for votes of court audience members, who mostly come with their minds made up.
    As for being understandable and durable, the Venetian system lasted for 529 years, until conquered by Napoleon, and the Florentine for about 300, until it was corrupted by the Medici. That sounds like some kind of legitimate and understandable to me. If we got 200 good years out of it that would be better than what we have done in the U.S.

    See http://constitution.org/reform/us/constitution-us-model.html for a detailed proposal.

    • PW1202 says:

      “then why not have guilt or innocence be decided by asking for votes of court audience members, who mostly come with their minds made up.”

      Jon, agree with this. I have always had troubles with the “twelve good men and true” approach. While it may be random to some degree (and the concept is sold on its randomness) , the overall profile of the 12 may vary considerably, from “dispassionate and analytical” to “the bastard is guilty, and why do courts use these big words”. What do you do with a jury member of poor perception and analytics. How does this jury member cope with the judges directions to the jury, which in a complex case, may have many inter-dependent issues.

      And as Jon has indicated, many jurists believe the the police would only bang “guilty bastards” in court, so these silly proceedings we are about to consider are a mere formality. The Police have better things to do than put innocents in court.

  14. ChrisB says:

    I don’t know that ranking political systems by how many years they endure is a knockdown argument, if only because that would make the Pharaonic system the undisputed leader, followed by Chinese emperor system and the papacy; which is two hereditary monarchies and one limited electorate.

    • paul frijters says:

      Indeed, longevity of systems devised and manned by a small elite are not a selling point.

      Besides, the previous comment thread already dealt with Venice and Florence. Alan already raised their examples. John clearly hasn’t read the thread, but I don’t see why the rest should repeat ourselves.

    • Alan says:

      If you reread my comment I did not rank political systems, but electoral systems, by longevity, Nor did I cite longevity as an overarching principle but simply as an answer to repeated claims that sortition is inherently unstable. In any case the papacy of the Coptic church is at least as old as the Catholic papacy, and there a large church assembly reduces the candidates to three by repeated votes, and then the Holy Ghost votes for one of the three by causing their name to emerge from a random draw.

      Much of this argument consists of claims that Sortition Example X was frightfully simple. If anyone answers that Sortition Example X was actually not so simple the original commenter generally comes right back by arguing that Sortition Example X was frightfully complex.

  15. ChrisB says:

    Though, actually, the papal election system might have a point. If you valued the opinion of people who’d already been selected for being on top of their game in the vital area, how about having the heads of commonwealth departments (or commonwealth and state departments) vote to choose the head of state? With the expectation being that they’d generally, but not invariably (see Hadrian the Seventh) choose one of their number.

  16. ChrisB says:

    On the difficulty of detaching any single decision from the background, I was once charged by an assembly with designing a draft system of university governance.
    Any such system was going to be so complex that a take-it-or-leave it vote would necessarily include a lot of things that any single voter didn’t agree with. To minimise that, I set up a decision tree where the assembly could decide between A and B; if B, B1 or B2 or B3: if B2, B2a or… and so on, with debate and discussion at each point.
    It was a total failure, because
    A) most the people who wanted (for example) A refused to accept the legitimacy of the view of the majority of the Assembly who voted for B, and (even more, and more basically, destructive)
    B) the people who were outvoted at the point B2civ* then said OK, now I know that’s what’ll happen I want to go back and vote B3 instead of B2 and go up that branch. That might have been repairable if we’d been able to run the exercise ten times (though then the decision on when to end the exercise would be the important one) but we weren’t paying the voters and thus couldn’t.
    If a complex outcome has five components that could go one of five ways then there are 5 to the 5th possible variants, and none will have a majority. Decisions are meaningful only in the entire context, and the person who sets the fixed points that are off the table has the real power.

  17. If alternatives to electoral democracy are to be rejected on the basis that they contain flaws, then surely the “flaw test” must also be applied to electoral democracy itself?

    If our acceptance of electoral democracy is based on the idea that it is the least-worst system, and we accept that electoral democracy is continuing to degrade in terms of its ability to protect the weak from the aristocratic, then surely we must constantly review the “flaw” scorecard of each alternative to verify that we are actually using the least-worst system?

    Of course there are flaws with sortition, but how do these flaws stack up against the flaws in our current system? How serious a concern is it that the meshing of media, political and corporate interests is now so advanced that power can quite literally be bought? How serious a concern is it that electoral democracy has no answer to trans-national catastrophes like Climate Change and the international refugee crisis? How serious a concern is it that in each electoral cycle, more and more people are refusing to participate?

    Picking apart any political system will produce no end of doubt and whataboutery, but if you are to pick apart one system, you must apply the same test to every other system. The question is not “Is sortition a pain-free antidote to all the ills of the modern world?” The question is “Has electoral democracy become so degraded that serious consideration of non-electoral systems is now warranted?”

    • paul frijters says:

      of course it’s about relative ills. Which is why the post speaks about whether the particular innovations looked at are likely to make things better or worse compared to the status quo.

  18. Sorry, the b-quote function in my last last didn’t work, so here’s it again:

    Paul makes some very interesting points on the difficulty of putting theoretical proposals (like Simon’s and my own) into practice.

    >The only counter-move would be if the decision to go to a jury is out of the hands of government and parliament, such as perhaps triggered by a certain number of citizen signatures.

    That’s an interesting idea, and similar to my own proposal for direct-democratic initiatives, although I would be hesitant to base it purely on the number of signatures, due to the ease by which pressure and lobby groups (and media) can gain this level of support. That’s why I combine a 100,000 signature threshold with an annual public votation (an instrument of Swiss democracy), in which all citizens make their choice from those proposals that have exceeded the threshold. My proposal (for a public enquiry with balanced information and advocacy and a large jury) would certainly work best for referendum style decisions (should Britain remain in the EU?) and the result should be binding for a generation.

    >Part of the job of government is to horse-trade and thus combine choices for which there is no majority in favour individually, but for which there is a majority when bundled.

    Most of us kleroterians view that as part of the problem. For example Tony Blair’s New Labour was able to smuggle in a whole load of social and cultural reforms that would have been otherwise opposed by their core voters. What is wrong, in principle, with disaggregating these bundles? Why should socially-conservative working-class voters have to accept multiculturalism, gay marriage and high levels of immigration because this was the only way of returning a government that was felt to deliver policies in their economic interests?

    PS Simon is a lawyer, Naomi a research chemist and I’m a student in political theory. We have never met.

  19. Paul,

    >The core issue is that something can be in the public interest without commanding a majority because the winners feel more strongly about it than the losers.

    OK, but why does that automatically equate with maximising utility? Why is it necessarily in the public interest to allow those with a passion in a particular cause to predominate? Interestingly your approach is close to the proposal of John Burnheim (Sydney University) for demarchic councils, selected by sortition from those with the strongest interests in any particular domain. But John is resolutely opposed to any form of horse-trading or policy aggregation. My concern is to privilege the views of the silent majority and I share Madison’s search for impartial decision-making (i.e. those with the least interest in the bill under consideration). In this scenario Mary, John and Robin would have to secure the change they are seeking through their persuasive arguments, rather than pork-trading, which is a camouflage for the defrauding of the interests of the general public.

    >the total package represents a win for a larger group.

    Unfortunately politics is a zero-sum-game and for every winner there is a loser, as working-class followers of New Labour have discovered to their (considerable) cost as wages have been forced down by wholesale levels of immigration.

    • paul frijters says:

      “Unfortunately politics is a zero-sum-game”

      I guess that is where we differ in core assumptions then. Trade is not zero-sum and incompetence is not zero-sum. I agree that there are no policies that have zero losers, but I do think there are better and worse policies.

      Utility is just a measure of strength of feelings about issues. The example I constructed is one where all individual policies represented an increase in average utility even though a majority saw their utility decrease. By allowing the voters to horse-trade, a package supported by a majority could be found.

      I am not in favour of sortition based on deep interest. But I do want the politicians, however they are elected or sorted, to be able to horse-trade. That is a core part of their job.

      • Paul,

        >Utility is just a measure of strength of feelings about issues. The example I constructed is one where all individual policies represented an increase in average utility even though a majority saw their utility decrease.

        There speaks the economist. In moral philosophy utility is regarded as the greatest good of the greatest number rather than the province of those with the loudest voices (especially on account of the overlap of the latter with the agency privileges resulting from education, wealth and other such factors).

  20. Naomi says:

    Paul,
    I’m one of the few kleroterians who share your position on policy trading. I don’t see it being any different in principle from trade in an economy. All the participants end up better off than they were before.

    On a related note, I’m not too fond of pure ad-hoc juries. I think there should be one jury/statistical sampling of a few hundred people who vote on all the matters to come up during their short and preferably staggered terms. They’d have more of an opportunity to get their heads around the affairs of the day. It will be a lot for them to take in, but we Condorcet’s Jury Theorem working in our favor. To get consistently reasonable results we just need the average voting member to do slightly better than a coin-toss at determining which proposals are reasonable. We don’t need to train an assembly of policy wonks.

    To make up for the fixed terms (in part anyway) I also support a two-vote system for proposal approval. If a proposal is approved by a supermajority then it should go into force immediately. If it only passes by a simple majority then the sponsors of the proposal should have to wait until the entire jury/statistical sampling turns over, three months, or a year, or whatever. Then, with a fresh jury/sampling, it can be passed again by a simple majority to go into force. So to get the passage of a proposal opposed by a majority you’d either need to bribe a huge portion of the voting members (despite their use of the secret ballot) once or somehow bribe two statistical samplings in sequence. And the members of the second sampling would be unknown to the people who made the proposal. The arguments made by the supporters and opponents in the first round would be known months before the second round, so the chances of a more rigorous debate would be greater. The chances of a proposal passing due to a statistical fluke in the sampling would be very small.

    The fixed terms allow for proposals to be built-up gradually in chamber over time rather than being an all or nothing affair. Amendments to active proposals can be introduced and debated by the elected members and approved or rejected by the voting members. If a package deal makes sense and leaves the average person happier with the overall outcome… then the case for bundling can be made to the voting members. Politicians already have to justify their package deals to the electorate today. Here they have to justify them to a better informed statistical sampling of the population, one that is largely free from rational ignorance or time constraints.

    Simon mentioned proposal introduction by parliamentary “majority.” I’m less inclined to require the support of a majority of elected members. If we were to allow any elected member to introduce legislation for a vote, then no party would ever hold a monopoly on lawmaking. The diversity of proposals would be maximized and the chances of collusion would be minimized. With no formal majority there’d be no chance of a hung parliament regardless of the electoral system. We could even encourage fragmentation to increase electoral competition and voter choice.

    This is, I think, one of the major advantages of sortition. If you have a majority government (be it a single party or a coalition) then all the affairs of the day—every crisis, every issue to come up regardless of the scope of the campaign manifesto—are viewed through the ideological lens of that majority and its leadership. The reasons a party might win an election are manifold. Maybe most people preferred their platform. Maybe they have a relatable leader, or a leader who has a more trustworthy smile. Perhaps the other guys did a bad job and needed to be punished. Whatever the reason, once a party or parties hold a majority, then the whole government is tinted their color. They can be penalized for their actions to an extent limited by the memory of the electorate and countless other competing factors. Is it more important to punish corruption or avoid economic misfortune? Only a few matters ever become major campaign issues. This is completely understandable. The public’s attention is limited. We all have busy lives. But an election decides on a litany of other issues, major and minor, that get squeezed out in a busy election cycle. Elections are undoubtedly all-important in a democracy. The trouble is they are a low-resolution mechanism.

    On the other hand, if any elected member can introduce legislation there’s the danger the agenda would get too crazy for the jury to keep up with. I’m inclined to split the difference and simply require a large number of co-sponsors for the introduction of proposals and amendments. Perhaps a third or a quarter of all elected members would be an appropriate threshold. That should get us the best of both worlds.

    • Naomi,

      >If a package deal makes sense and leaves the average person happier with the overall outcome… then the case for bundling can be made to the voting members.

      It depends what you mean by package deal — if it’s a question of trade-offs (e.g. between cost and optimisation, or between keeping the lights on and saving the environment) then a rational case for the bundling can be made. But I thought Paul was referring to pork-trading, i.e. the need to accept costly or sub-optimal policy proposals in order to build alliances in the legislature? How can such (corrupt) behaviour be justified, and why would it be necessary if the policy jury voted on each issue, rather than on the manifesto commitments of a political party (holding one’s nose over the commitments that one dislikes). The only moral case I can imagine is the empowerment of minority interests, but for every empowered minority there is a disempowered majority, hence my argument that politics is a zero-sum game.

      • Naomi says:

        Keith,
        I guess there are two angles here. One is the pursuit of a political equilibrium that takes into account the strength of preferences as well as simple numbers. If a third of people strongly support Policy A and only weakly oppose Policy B and another third strongly supports Policy B and only weakly oppose Policy A (with the other third opposing both) then stitching the two together leaves a majority happier with the overall result than they would have been if every issue had been taken one at a time. It’s a more optimal result.

        The other angle is the chance of some subset applying undue leverage to try to extract more goodies. We certainly see this behavior often enough when an elected assembly lacks a robust and disciplined majority. It’s a problem, for sure, but going all the way to a ban on political negotiations is a bit extreme. Mechanically carrying out the preferences of the majority without some opportunity for accommodation and give-and-take only ensures the alienation of ideological minorities who would have no hope of even a token victory. Their only recourse would be action outside the political system. If politics is the civilized alternative to armed conflict (jaw-jaw instead of war-war) then room for conscious negotiation and compromise between the leaders of the different factions is necessary.

        I suppose the goal should be to minimize undue leverage while still allowing a negotiated equilibrium to be reached. The sort of hybrid we’ve been talking about might come pretty close. Later samplings would have to vote down challenges to components of a deal for it to hold. No one is going to stand by an obvious bit of pork when examined in isolation. Also, courting individual votes would be quite impossible, as the members of the sampling would not be known before they were drawn. One would be limited to pursuing subsets of the population large enough to be reliably represented in the sampling.

        • >If politics is the civilized alternative to armed conflict (jaw-jaw instead of war-war) then room for conscious negotiation and compromise between the leaders of the different factions is necessary.

          I guess I still have a residual affection for Burke’s view on virtual representation and deliberation. So long as there are one or two advocates for the minority perspective (ensured by your PR-based selection process), and so long as the jury is schooled in the art of listening, then benign outcomes for minorities can be achieved via the forceless force of the better argument without the need for pork and other forms of corruption. I would hope that the decorum and gravitas of the institutional setting would impress itself sufficiently on the members of the jury, who would be chosen (by god) to legislate and have a solemn duty to act on behalf of the whole nation. The connection of lottery and religious belief is not entirely coincidental and my own experience of jury service would lend me to believe that most citizens would take their task very seriously.

  21. ChrisB says:

    If large citizen juries also have long terms – longer than about a month – and have to decide a number of issues, then I can’t see a way of preventing parties forming – either ad hoc or based on existing allegiances – and placing us back much where we were. The party system is simply a trading convenience, where I get the votes for my project by agreeing to yours, with a saving in time from not having to reach every delegate individually. It makes sense to get 51% of the delegates together and say “anything that gets 51% of us will bind us all”.
    (and then, of course, it also makes sense to have a faction that could reliably deliver half of them; and so on)

  22. Alan says:

    We have one very concrete example of sortition wiping out corruption at a stroke. The California Citizens Redistricting Commission is chosen by a process that a Venetian would surely recognise.

    Applicants must meet a rather stringent list of disqualifications such as holding public office or contributing more than $2000 in campaign donations in the preceding 10 years. A selection committee of three qualified auditors who meet the same disqualification list is determined by lot. The commissioner selection process mist be open and transparent.

    There were 5000 applicants.

    The selection process reduces this to 20 Democrats, 20 Republicans and 20 Independents. 4 party leaders in the Legislature then get 2 peremptory challenges from each pool. There is then a random draw from the remaining applicants in each pool to select 5 Democrats, 5 Republicans and 4 Independents.

    Under this dangerous, irresponsible, baroque, untrustworthy, opaque, complicated, corruption-prone, positively Venetian and almost Byzantine system California went in one election from one the most gerrymandered states in the Union to one of the least.

    I would think it simpler to get rid of the peremptory challenges and party representation and simply have the selection committee certify candidates as competent or not competent and then select the commission from the competent applicants by random draw.

    • paul frijters says:

      nice example. How did they get the transfer of power to this commission to do the re-zoning? Referendum? It is a nice case of a stand-alone issue that you would want to have dealt with outside of politics.

  23. paul frijters says:

    Naomi,

    if I can summarise your proposal: you want it to be possible for about a third to the quarter of our current parliament to send legislation to a citizen jury that either in one go or in a staggered manner considers the legislation and then accepts/rejects it. You want those juries to live for short periods of time and for the current pool that would make up a second jury to be unknown to anyone. You hope this makes it hard to bribe them and that the barrier to invoking them is high enough to ensure only serious things go to them, but low enough so that you undo many of the advantages of being the largest political party, i.e. you’d hope it constrains the power of the executive.

    Let’s try and think this through. In general, the counter-move by the executive would be to pack huge amounts of legislation in one bill and to generally swamp parliament such that any jury would be overwhelmed. To a large degree, this is already what is happening to parliament right now, with MPs swamped by legislation that they don’t understand such that they are forced to trust the executive and its feeding bureaucracy. They’d try the same trick with a jury and its hard to see how a jury can effectively respond as it is not part of the legislating machinery.

    Let’s then take a good topical example of a policy that shouldn’t be implemented in Australia, but that has passed the parliamentary hurdle: the plan to buy dozens of submarines, to be built in Adelaide and serviced in Sydney, delivered in the far future, at costs far above world-levels. I am not 100% certain, but easily 90% certain this is a classic case of special interests defrauding the population. The issue is whether and how a citizen jury could have stopped it.

    How did these submarines interests do it? Well, they pre-cooked the plan for years, building up their coalition. They will have spent months thinking of how to sell it to the government, employing former ministers and former generals. They will have spent millions on fake studies ‘showing’ how much we need these things, and tens of millions on success fees to middle-men who are deep into the cabinet and both political parties. By the time the public heard about it, many months of frantic lobbying would already have been underway.

    They won over the executive by the story that the places in Adelaide building it are marginal seats. With the same story, they ripened the minds in Adelaide, making it hard for both political parties to ‘withhold’ jobs from poor labourers. They muddied the waters by means of dubious but hard to disprove claims, like how they were safeguarding the manufacturing base of Australia. Even professional lawyers on this very blog, who are very well meaning and unusually well-read, fell for it hook-line-and-sinker. Generals said we need them, both political parties fell in line.

    How would the special interests have dealt with a jury? In much the same way: a media blitz, a procession of insider witnesses making up the top of the army and the ministry claiming we need these things, and huge volumes of legislation that would have to be mulled over. You’d need a team of professional lawyers spending months on it to unpack it. Amateur citizens wouldn’t have a chance wading through it. Of course they could summon some counter-experts from economics or from defence agencies abroad who’d tell them its a con-trick that is all about siphoning off money from the public purse, but the machinery of the state would be on the pro-side.

    What would be needed to prevent the submarine scandal from occurring? Well, the citizen jury would need to be able to cross-examine witnesses under oath about success fees and future favours they would expect to get as a result of their lobbying efforts. They would need to have full access to top-secret documents where the ministry insiders and foreign experts were mulling over these purchases. In effect, they would need to have more resources and powers than a Royal Commission. And even then, I think they’d likely fall in line anyway as the media campaign to convince the population was raging.

    How could you stop it then? One way is to have stronger heads of ministries being able to resist these things: they often know what is really going on, but they lack the power to resist the machinations, or they themselves are part of the con-trick. But having the right people with some independent power at various places in the state bureaucracy might help. Would would also help is if there was some ex-post oversight so that the politicians and the special interests could be held to account later on when it became clearer what really went on.

    But I don’t see how a normal citizen jury would succeed in the submarine case. Since there is bi-partisan support, it wouldn’t even go to them in your proposal, Naomi.

    • Ken Parish says:

      I don’t know whether or to what extent a 12 submarine capability is a genuine priority need for Australia’s defence. Nor I suspect do you. Seemingly knowledgeable independent academic experts say they ARE a priority need.

      Assuming that to be the case, does it make sense to spend an extra $8-10 billion in a $50 billion project to build them in Australia? Again I’m not sure. I certainly haven’t done an economic cost/benefit study, nor do I have the capability even if all the relevant facts were knowable (some will undoubtedly be classified information). However it’s reasonable to think that if the subs ARE a critical part of our defence capability then we will need to be able to maintain them in Australia and that probably means building at least some of them here.

      There is also the factor of using this huge defence expenditure to leverage development of Australia’s ongoing/reborn manufacturing capability. I don’t share the universal hostility of many classical and neoliberal economists towards governments “picking winners” and subsidising industry sectors. If the capabilities are multifarious and able to be applied across numerous sectors (as would no doubt be the case), then it might well be an excellent way to preserve and enhance Australia’s high tech industrial capacity (ditto with car manufacturing about whose abandonment I am also doubtful).

      The reality is that the US, Britain and large parts of Europe (Germany, France, Sweden) piggybacked a large part of their industrial capability development through the 20th century by leveraging off military spending. They don’t refer to the “military-industrial complex” for no reason. I suppose it’s conceivable in theory that a first world country could develop advanced industrial capabilities solely through the marketplace endeavours of rugged private entrepreneurs, but I can’t immediately think of any country where this has actually occurred. It’s certainly fair enough to be wary/skeptical of “rent-seeking” behaviour by corporations intent on gluing themselves to the public nipple, and the economic distortion that can generate, but that’s no reason to reject the entire idea outright without examination or analysis. It would be instructive to see such analysis if it was done (I hope it was), but as I say one would suspect significant relevant aspects would probably be classified.

      • paul frijters says:

        Hi Ken,

        the article you linked to is a great example of the kind of lobby-effort I so dislike. The writer was involved in the defense white-paper and a consummate insider. To call him an independent expert is not believable.

        And how does he argue his case? With what I think of as con-tricks, playing on habits of obedience, being super-vague where it matters and hush-hush-trust-me.

        Example: when arguing for why Australia needs submarines, he says

        Among the long-standing priorities in Australia’s defence strategy are protecting critical lines of trade and communication for essential national transport and military operations, and denying the use of the sea to a potential adversary.

        Because of their unique characteristics, submarines will play an essential role in these endeavours. Their ability to restrict the actions of any would-be aggressor in the maritime domain remains unmatched. And despite the rate of technological change, they are unlikely to be challenged for at least a generation.

        Submarines have the ability to operate covertly for extended periods and to attack with devastating lethality without warning. This means they can create uncertainty in the mind of an adversary about where they are and whether it is safe to sail ships or submarines. And the larger the submarine force, the greater that uncertainty.

        I can get really worked up about paragraphs like this. He first talks about ‘protecting lines of trade’, but then argues that submarines are not about protection of own trade but about threats to other people’s vessels. So he has an offensive scenario in mind rather than defensive.

        And who is the supposed enemy for which we need these submarines? He doesn’t say anywhere in the paragraphs or in the article, nor is there is a logical one in sight: against a big country like China, our forces would be ludicrously small anyway such that we’d need a coalition, and virtually everybody else is our friend at the moment and for the foreseeable future.

        So on the basis of unspecified threats of unknown enemies whom we would want to hurt but not protect ourselves from, we should spend billions in marginal seats on nice shiny machines? Yeah, right.

        And wouldn’t the same argument of a pivital role for submarines go for other countries? You would think so but you don’t see our European partners investing the same amount in submarines.

        So it walks like a duck, talks like a duck, and sh*ts like a duck. I give it 90% chance that it is a duck :-)

    • Naomi says:

      Paul,
      I prefer to think in terms of a single assembly. Some members would be elected by the general public and serve a several year term and some would be drawn by lot from the general public and serve a several month term. Proposal introduction and speaking powers would be held by the elected members. Voting powers would be held by the members drawn by lot. Let’s call them advocates and jurors respectively. One house. Two estates. Speaking power should belong to all the advocates individually. Proposal introduction should require a large number of co-sponsors to keep things from getting too crazy. I’d prefer if the terms of the jurors were staggered, like the terms of the American senators, to ensure that the behavior of the body does not oscillate on a several month cycle with the more experienced jurors being replaced with a fresh set en masse.

      We need to remember the point Garreth made above about relative ills. If someone can get the public on-board with something, good or bad, then almost any government derived from the people will go along with it as well. Also, a government consisting of a relatively small number of individuals will usually be susceptible to manipulation to some extent. The issue is one of degree. A conventional elected legislature with near unanimity behind some corrupt piece of legislation will surely enact it into law. At least here the few opponents would have a some chance of prevailing in the in-chamber debate and convincing the jurors to oppose the legislation. Ideally the advocates should be diverse enough to where we can safely assume some subset will be in opposition to any particular proposal. A highly proportional representation system is probably best for advocate elections. Fragmentation discourages collusion, both willful and coincidental.

      Your possible solution to the submarine issue is to strengthen the heads of the ministries. Appointment and dismissal of senior executive officers by a hybrid assembly should move us in exactly this direction. Furthermore, the jurors themselves would face no consequences for voting down a piece of legislation. This is in stark contrast to the current situation where MPs can conceivably face an election or, at the very least, make enemies in the leadership if they vote the wrong way.

      • paul frijters says:

        your single assembly is a regular parliament (which would have regular political parties), plus a citizen jury that votes on policies and, presumably, government, pretty much exactly like in the second proposal in my blog. Putting them in one house empowers the insiders and eases coalition formation. I don’t really see why you would want that, or that it makes it easier to argue for sortition. Best to keep the set-up as pristine as possible when arguing for a change.

        In your proposal, you wouldn’t get a large field of applicants for top-jobs in ministries because its the insiders who would do the pre-sifting and thus come up with the candidates. So you wouldn’t be able to strengthen the ministries with that system. You need the citizen juries to be able to set the parameters themselves for job applicants, which is the power of initiative you envisage to be held by the executive.

  24. Jon Roland says:

    There is another kind of corruption we haven’t discussed much. The corruption of the general population, continually demanding more and more government services at ever greater cost, until the government collapses into insolvency and the economy collapses from excessive government interference. All election systems are susceptible to it. THis is discussed by Frederick Bastiat, The Law http://constitution.org/cmt/bastiat/the_law.html

    Perhaps the best solution to runaway socialism is sortition, so that decisionmakers are not tempted to yield to limitless public demands for public goods. A sorted body might be better able to resist such demands.

  25. ChrisB says:

    Jon, Bastiat died in 1850; in the intervening 166 years no French government has collapsed into insolvency; neither has any English government; neither, since 1900 – admittedly a mere 116 years – has any Australian government. How nigh, precisely, is the end? When do you currently think this collapse is going to happen? How long, more or less, can a prophecy go unfulfilled before it is accepted that it’s just not going to happen?

  26. Jon Roland says:

    Bastiat was looking at longer timespans than 116 years. Such as the 20 civilizations that preceded our own, and eventually failed, most of them not democratic, but popular pressures operated much the same way. But technology and modern finance markets accelerate the process. We have already seen insolvency in Weimar Germany, Argentina, now Venezuela and Puerto Rico. These nations eventually recovered, as did the world from the Depression of 1931, because there were external sources of assistance. But the world is now one large market, now in a bubble of reckless speculation, driven by endless demands for consumer goods and government services. It doesn’t go on forever, and there is no outside assistance.

    No one can know when the collapse will come. Perhaps some new technology an delay the inevitable, but none is in sight. Probably in only a few years. Perhaps tomorrow. When it happens, the people (the ones who survive) may be more receptive to reforms like fetura.

  27. ChrisB says:

    Confining myself, with extreme difficulty, to items with remote connections to the point of the post, I say only that if the collapse of twenty previous civilisations from (apparently) the sole cause of lack of enough fetura hasn’t persuaded us to adopt fetura in heaping quantities as a feature of all elections from President to Masterchef it seems unlikely that our surviving descendants will be moved to feturaphilia by the collapse of twenty-one.

  28. Jon Roland says:

    It is up to us to prepare people to accept new solutions when circumstances provide the opportunity. After all, Venice and Florence didn’t adopt their systems in a time of peace, but out of bitter experience with worse alternatives. Just as the United States did. Some people do learn from history. Let us be among them.

  29. paul frijters says:

    John,

    I think that the main force that keeps bureaucracies and populations in line is competition between countries. That operates as a disciplinary device with almost any system of government.

    Whilst it is clear that sortition has a political mountain to climb, I don’t want to consign it to the hopeless set of ‘things we will learn after armageddon’. Systems do evolve and countries to experiment with alternative systems of government and democracies. Australia is a great example of that, with its fairly unique system of preferential voting.

    I still can’t see the great value of citizen juries for policies, but perhaps as means of selecting people for jobs it has more mileage. After all, that is what the Venice and Florence systems of random juries were about: not to decide on policies itself, but on who was going to decide on policies.

    • Alan says:

      I suspect it is competition between systems. Comparing the dates for the collapse of the Soviet Union and the ascent of neoliberalism to a fixed political theology as unchangeable as gravity is interesting. It is as though the Washington Consensus has been able to dictate the terms of the Melian dialogue to the whole planet since 1989.

  30. ChrisB says:

    Raising another objection to sortition, I understand that in game theory the prisoner’s dilemma – where it makes sense for each individual to inform on their colleague despite the fact that the best outcome would be for neither to inform – can only be solved if there exercise is repeated regularly and an informer can be influenced by the prospect of their bad behaviour being held against them in later rounds. A citizen jury seems to me to have an analogy with that, in that every single-issue juror would have very little incentive to compromise. They would accrue no favours from any compromise and no suffer reputational damage from refusal. They would have no incentive to consider the bigger picture, or to say “I vote for A if, and only if, B.” Imagine, say, two citizen juries convened to decide on the desirability of a carbon tax and the desirability of compensating low income earners for changes in taxation structure. Each jury might wish to say “I vote for A if, and only if, the other jury votes for (a)”, but they can do nothing to affect the other outcome.

  31. Jon Roland says:

    @ChrisB
    Your argument is why I keep emphasizing that “sortition” should not be conceived as only a cramped. one step random procedure. Yes, the prisoner’s dilemma applies, or at least the iterated prisoner’s dilemma, discussed by Rapoport in The Evolution of Cooperation.
    A better model is provided by genetic (or evolutionary) algorithms, in which random steps (such as reconfiguration) alternate with a fitness screening. That is why I use the term fetura, which is another name for the process.
    Another key concept is self-organizing systems (part of chaos theory). Society is a self-organizing system composed of other self-organizing systems, which in turn are composed of other self-organizing systems, and so forth.
    The argument we should be making is not over just the random step, but over the entire system in which it is just one step of many. A well-designed system is composed of many self-organizing systems composed of other self-organizing systems in which fetura is a key element.
    What we are discussing is how to design a subset of society that serves society as well as society could if it were reduced to a single wise decider, who is himself composed of other systems. In other words a microcosm of society that somehow represents society, and makes decisions for the best interests of society.
    Stretch your imaginations a little (okay, a lot).

  32. Jon Roland says:

    Following through, a ell-designed system cultivates virtue among its decisionmakers, which can be assessed by other decisionmakers. This relates to one of the three types of morality: deontological (duty based), consequentialist, and aretaic (virtue-oriented). The last depends on reputation and admirable qualities. “We become what we admire.” But admirable qualities are not just a matter of taste, because they have consequences.
    So a well-designed system cultivates good reputations and the perception of such.

Leave a Reply

Your email address will not be published. Required fields are marked *

Notify me of followup comments via e-mail. You can also subscribe without commenting.