Could sortition help against corruption, part II

In part 1, I looked at whether it made sense to have random individuals inserted into parliament, or to let policies be decided by juries full of randomly chosen individuals. Both were argued to be unworkable and likely to lead to more corruption, rather than less: policies that favour the special interests are so entwined with large legislative programs that it is not feasible or desirable to have them judged by juries of amateurs. And random members of the public spending years in parliament have even less incentive to be honest than members of long-standing political parties with brand-names to protect.

Here I want to mull over the possibility of using random members of the jury for appointments and for generating signals about corruption. The end result of my deliberations is that I can see use for random juries in major appointments within the public sector, but that I cannot see them working as a means of generating information.

For those who missed the earlier discussion on why having parliaments made up of random members of the public is not a good idea: a good example of problems you get with excessive diversity in parliament is Brazil, which has dozens of political parties and as such without long-established major players who have some incentive to care about reputations for honesty. Their parliaments are pretty close to random groups of populists, a diversity that has been strongly argued to make their politicians sitting ducks for special interest groups. Their parliamentarians have been argued to sell out to the highest bidder.

The fundamental problem, which is that without big players no-one has enough incentive to care about corruption, is an old argument you also see in other arenas: to keep big companies honest you need large shareholders with enough skin in the game to monitor and punish; to keep rogue countries in line and do something about international crises, you need large countries big enough to worry about and organise an answer to these problems.

For sure, there are some policy-areas where citizen juries might be useful, such as when it comes to decisions about politics itself, such as electoral boundaries, but there too the West has plenty of examples of well-functioning institutions that do the same job without citizen juries.

The post drew some knowledgeable commenters though who have been wedded to the idea of sortition for quite a while, and the proposals they brought up did make me wonder whether sortition could help with other decisions in our democratic systems. Mainly they pointed out the use of randomly selected groups of citizens who got to decide on appointments rather than policies. Sortition systems in Florence and Venice operating for many centuries were examples of how randomly chosen people voted in the ultimate leaders, though in those cases the ‘randomly’ chosen people were all members of the elite themselves.

The hope is that random members of the public would be pretty good at judging the bias and character of people. After all, that is pretty much what criminal-trial juries already do and in which random members of the public are arguably better than specialist insiders who cannot see beyond their own little corners.

So let us come from the problem from the other end and look at whether having the wrong people in charge is indeed a problem within our democratic system, then see how sortition would be applied to hire the right people, and then try and think through whether sortition would truly help if we consider the counter-moves of special interests.

One could describe our current Western democratic system as having a strong executive that is heavily aligned with special interests, feeding slanted proposals to parliament, dominating the civil service that has lost its independence. The reason this system has gone off the rails in recent decades are varied, but I see three mayor stylised problems:

  1. The political executive has managed to politicise the civil service, and as a result the ministries lack the independent power to resist special-interest initiatives that get the backing of the minister.
  2. The media no longer has the ability and the interest to hold the government to account such that corruption and misconduct are uncovered and kept in the public’s eye until the problems are sorted out. Rather, a lot of the media is co-opted or is indifferent: stories are still being unearthed, but only a small minority of the population gets exposed to them. One suspects that in the background, newspaper owners are selling silence to those major interests willing to pay (ie you can pay to not have your corruption exposed! Probably more lucrative than chasing the advertisement dollar).
  3. The executive and the top of politics has got locked into a material status-race such that former politicians and top ministerial staffers are obsessed with getting their slice of the pie. They then become co-opted by special interests to lobby the current executive and thus to defraud the population.

How can sortition help with this? I think the third problem is somewhat out of the hands of the sortitioners as the influence of the old party members is an inevitable by-product of the party system wherein the old party heroes still have influence on the new members. The material status-race is also a general trait of our culture and though I do have theories on how to reduce it (see here), for all intents and purposes, it’s not something we can do much about. You get status races with every elite.

The second and first problem though do seem to lend themselves to an appointment-via-sortition intervention. The role of providing believable signals to the population about corruption and misconduct can be taken up by citizen juries; and one can think of citizen juries as means of selecting top bureaucrats so as to give them an independent mandate versus the politicians. Let us think through both in turn.

                Citizen juries selecting bureaucrats.

Currently, appointments of senior civil servants, like Deputy Secretaries of major departments, are partly political and partly driven by the bureaucracies themselves. We could mandate that the top civil servants just below the minister get appointed via a citizen jury that selects candidates and chooses one amongst them in open hearings. This would be akin to the top appointments in the US that have to go through the US Senate (such as for the Supreme Court), but I would envisage both the short-listing of candidates and the selection of a single one to be outside of the political and bureaucratic system.

There are many ways to do this, but the essential idea is that one takes 20 or so random members of the public, let’s them sift through thousands of candidates for top-bureaucrat jobs to come up with a short-list for each position, and then select another 20 random members of the public to make the actual decision of who gets what job by means of open hearings and open verdicts. By separating the short-listing from the actual decision, one takes away most of the opportunities for gaming the system, and by having the hearings in the open, one gets accountability of the jury process.

Using such a sortition process for appointments to jobs is similar to what was done historically in Venice and Florence, where random sets of important people decided on who got the top political jobs. Yet, I don’t think that we have an undemocratic or bad way to select our top politicians and that we should hence fully copy those historical, elitist, examples (in Venice and Florence only the insiders got to vote): the current problem is more that the whole layer of executive politicians has too much power to pervert the bureaucracy and that the bureaucracy has too little incentive to resist. So the obvious solution would be to use citizen juries to select the ‘inbetween’ layer that links the executive with the bureaucracy.

The purpose would thus be to empower the top civil servants and hence to make the whole bureaucracy more independent of the designs of the politicians, but still beholden to the population that is then involved in choosing them.

Let us think how such a system could be gamed if it were institutionalised.

Ideally, one would probably want to get honest public-oriented civil servants winning most of the top-posts: people who have the insider knowledge and expertise needed, but who are not part of the political tribes or special interest groups. You would want them to float to the top of a citizen-jury selection system. But would they?

How would the special interest groups react to such an appointment system? They would certainly push some of their cronies to apply, but they would find it hard to bribe the juries, particularly if it takes more than 1 jury to actually appoint someone. And anyone obviously partisan or in the pockets of special interests would be easily unmasked, even by a jury of amateurs (members of the public may not be able to judge complicated policies, but they can spot a fraud when they pay attention, I think).

The special interest groups could try and harm the reputation of particular candidates they really don’t like, feeding the media information on their supposed incompetency. Citizen juries would find that hard to ignore. But then, the current politicians and civil servants also find smear campaigns hard to ignore. And citizen juries in principle could make the effort to unearth the truth about candidates, if given enough powers to investigate, such as having people giving evidence under oath.

When it comes to positions that are worth billions of dollars to special interest groups (like tops of mayor procurement ministries such as Defence), one should expect the lobbyists to professionalise the way they try to game citizen jury appointment, running slick marketing campaigns that try to prevent people from applying whom they don’t like, including vindictive campaigns against people within the ministries who applied in the past, just in order to set an example.

What the system would evolve to under that kind of pressure is not easy to see, but I do think it would be hard for special interests to really dictate who gets the top jobs. They will probably be able to scare off any identifiable group (like their enemies inside the ministries), but not the whole population of competent potential candidates, particularly not the ones outside of the state system.

So, I would expect that the people who survive the smearing by the interest groups and the independent scrutiny of the citizen juries to be relatively dull but competent independent characters. Perhaps not the ideal outcome, but likely an improvement on what we have now.

As a system of appointments, it would probably not be all that costly to set these juries up: currently the costs involve headhunters and god-knows-what for each individual appointment, so juries that appoint a whole lot of them in different areas after open hearings might be relatively cheap. The appointments decided upon this way could include lots of positions currently decided by politics, including semi-public positions like Vice-Chancellors and the executives of public hospitals.

In short, I can see it work in practise, though the political opposition to it might be immense, both from ministries and from politicians who, rightly, would fear loss of influence.


                Citizen juries judging previous policies, decisions, and politicians.

Currently, the systems of oversight on politicians and top civil servants seem ineffectual. A good example is the former Crime and Misconduct Commission in Queensland, which got into the habit of asking the very people accused of wrong-doing whether there was any merit to the accusation, and where the previous government tried to appoint one of their cronies at the top to make sure no investigations happened on their watch. In economic parlance, the supposedly independent monitoring institutions became captured by special interests.

The oversight by the media is also severely wanting. Perchance this is because one cannot effectively charge for expensive investigative pieces. Perchance this is because there is more money to be made by newspapers and TV-channels when they sell out on particular topics to local vested interests (just think of the infomercial on medicines!). Perchance this is because the internet allows everyone to exclusively nurse their tribal biases, immunising them from real information. Whatever the reason though, the media is not very successful in holding politicians and special interest groups to account. This aggravates the problem of capture by independent agencies.

How can citizen juries help with this, apart from directly appoint the tops of independent scrutinising agencies?

Citizen juries in principle can become an institutionalised means of generating the strong and open signals we want from the media: they could decide after-the-fact on whether decisions have been corrupt or not and openly communicate their findings. Like courts currently, they would be able to decide on what evidence they gather and to have hearings under oath.

I am here not thinking of deciding that something is corrupt in a legal sense, but more in a public-morality sense: whether the random members of the public honestly think upon reflection that a decision has been made against the public interest, and whether the motives of key decision makers have been good or bad. Whether there is individual criminal culpability is a different question, and more naturally one that one wants the criminal justice system to deal with.

One would want to have a fairly high barrier to these citizen jury courts, such that they are not used willy-nilly: one should perhaps only instigate them once a reasonable percentage of the citizens has signed a petition for them. Yet, one does want the verdict they reach to be made publicly available and broadly advertised via the national media channels, so that their verdicts become an unmistakable public signal of what the citizenry on reflection thing about previous decisions and politicians.

Let us think how this might work out if the citizen juries would be able to look at current politicians and current policies. The worry is that it would turn into a witch-hunt or a white-wash, the hope that it holds decisions and groups accountable who would otherwise remain out of sight because they are so strongly inter-woven with the state-system.

The stakes for individual politicians or interest groups would now be very high, much higher even than in the case of appointments because the citizen juries are now threatening their existing positions. Hence one should expect intense gaming and attempted manipulation of these citizen juries, even if they start out with a clear mandate from a set of concerned citizens who have gathered enough signatures to make the hearings happen.

Politicians and interest groups would pull out all the stops, including trying to prevent the juries from getting information by saying it would violate privacy and national security. Vilification of jury members and their families who veer towards ‘guilty of corruption’ verdicts should be expected. And accusations of incompetence and of biased intentions would be commonplace: all the effort that politicians, interest groups, and bureaucracies now put into keeping their dealings secret and unnoticed, would then go into preventing and disrupting citizen juries. One can just imagine what worried politicians might do to distract the population!

Whilst the vilification would be less if we’d restrict the citizen juries to judging former politicians, the important cases would by design always be those that are still highly relevant today. Hence, reducing the applicability of citizen juries to generate signals wouldn’t help: if their judgment matters, they will give rise to vendettas.

Would citizen juries be able to function properly with a full vendetta against them? Unlike judges, who are appointed for life and have spent decades toughening up and getting the support of their whole profession for their roles, we’d be talking about random members of the public with vulnerable families and impressionable communities. The pressures to exonerate powerful interest groups and politicians from any hint of malfeasance would be huge, and the individual benefits to reaching the right conclusions would be small.

By and large, it would seem doubtful that citizen juries would dare to do a proper job when it comes to judging politicians and interest groups. They would much more likely give everyone in power the benefit of the doubt rather than risk their wrath, encouraged to do so by the information fed them by the state bureaucracy.

In short, whilst I do think citizen juries would have the ability to judge who has been corrupt and what previous policies have served special interests rather than the general public, I am sceptical that they would have the guts to do so because the risks to them personally are one-sided. I would expect them to go with the flow of public opinion and state propaganda, which is not useful.



In can see citizen juries as a useful means of making appointments in the ministries and semi-public institutions, strengthening the independence of the state bureaucracy versus the politicians and special interests. But I cannot see citizen juries being useful in uncovering corruption or denouncing bad previous policies. That will need something else.

This entry was posted in Cultural Critique, Democracy, Economics and public policy, Education, Ethics, History, Information, IT and Internet, Journalism, Law, Libertarian Musings, Life, Miscellaneous, Philosophy, Political theory, Politics - national, Print media, regulation, Social Policy, Society, Web and Government 2.0. Bookmark the permalink.

13 Responses to Could sortition help against corruption, part II

  1. Moz of Yarramulla says:

    I cannot see citizen juries being useful in uncovering corruption or denouncing bad previous policies

    I suspect that your suggestion putting corruption etc bodies behind the shield of a citizen’s jury would make that unnecessary. You would have to give them the power to override government budgets, lest you find that while they have the power, they have no funds and the people they appoint can’t be paid. Similarly, I suspect a constitutional amendment laying this stuff out would be absolutely necessary, otherwise they become just another advisory group.

    A third use is as a publicity tool for a government that wants to do something unpopular. The opposite of a plebescite – the government publicises and funds a citizen’s jury to enquire into a topic and make recommendations. After much public debate, submissions and carry-on, the jury announces that on reflection they think that yes, Australia should implement the UNHCR recommendations in law or try to reduce AGW or whatever. Much furore ensues as the opposition realise that time is running out to defeat the proposal. With any luck a (mandatory?) free vote ensues in parliament and the requisite legislation passes. Government run round looking hapless and saying “oh well, ’tis the will’o’t’people” and other expressions of dismay (as our current lot are so very good at). We can but hope.

    Again, the underlying requirement there is for a general consensus that is unable to be expressed by parliament (like gay marriage or abortion), or a solid set of facts that can be expressed through the jury process (the negative effects of violence on children, or non-enforcement of hate crime bans). I’m thinking of NZ’s anti-smacking law in particular, which both major parties opposed because parents have a right, nay an obligation, to beat their children. But then public opinion swung and suddenly “don’t hit children” had always been the position of everyone involved.

  2. Ravi Smith says:


    Oregon, has an institutionalized form of Citizen Initiative Review that has worked out pretty well ( Several other states are in the process of copying it. The citizen jury hears arguments for and against a citizen’s initiative and then creates a one page document summarizing the arguments for and against and how many of the 20 jurors support the initiative. A small book containing all of the summaries is mailed to each voter with their ballot. Studies show that around 80% of voters read and are strongly influenced by the pamphlet. The information generated seems to be of relatively high quality. They have legalized marijuana, assisted suicide, same-sex marriage, GMOs, and several other politically fraught subjects. There is lots of money spent by the special interests on advertising, but none have attempted to go after the citizen juries. This system probably only works if you have a separation of powers (so you can vote on each issue) and citizen’s initiatives.

    • paul frijters says:

      Hi Ravi,

      yes, that belongs to the issue of how citizen juries can support referenda and direct democracy. Citizen juries as independent advisers on a referendum issue are fine, though in the longer run, I think you can do better than citizen juries for that particular role: I have long thought that the secret of the success of the Swiss referenda is not really the fact that they have so many referenda, but that a whole system of citizen-advise institutions has developed alongside it that sifts complex information and comes up with voting advise. So the Swiss have voting advise generated by the business community, consumer groups, labour groups, etc. Those long-standing advise generating institutions have a real incentive to be true to their constituency, but this combine that with the advantage of having specialised knowledge.

      Whilst a fuller analysis of the Swiss system is something for another day, it is an interesting thought that the combination of referenda and advising citizen juries is a quick way to have an embryonic sifting institution preceding more longer-run advisory institutions around referenda. Citizen juries as an intermediary step towards more professional but targeted advising institutions. Hmmmm.

      • Ravi Smith says:

        Hi Paul,

        I’m studying in Switzerland right now and I definitely agree with your take. It’s an amazingly complicated system, with an abundance of civic associations and seems to work like clockwork. I’d also add that being very decentralized is another information generating mechanism (you can see the effects of policy at the local level). It will be interesting to see if more professional citizen-advise institutions emerge in Oregon, Washington, and Colorado.

        • paul frijters says:

          yes, the Swiss system is very developed and cannot be simply adopted elsewhere: it takes time to develop and the intermediary period needs to seem better than what came before. You might well be right that the citizen juries in the US will gradually give way to the more professional bodies you have in Switzerland.

          A big difference is that mobility seems to be much lower in Switzerland, which gives the citizens more of a reason to care about local politics, which strengthens their democracy. The high internal mobility of the US makes that commitment much lower. Even commitment to the state will be lowish in the US. Commitment to the US as a whole is very high though, so perhaps the local initiatives will work much better once they grow into systems at the federal level.

          Perhaps Australia has more chance to develop a Swiss-type system. People don’t move around here as much. On the other hand, migration is huge and that won’t help.

        • Ravi Smith says:

          The US is certainly a very different cultural context from Switzerland or Australia. The US constitutional structure has too little accountability, too many veto points, and requires unanimity to change the senate. This makes it almost impossible to stop rent-seeking at the federal level .

          Most of the innovative and effective policy changes are being implemented from the bottom-up. Washington and California are voting on a $15 minimum wage and Colorado is voting on installing single payer healthcare. Utah and Massachusetts are even starting their own immigration programs. The best states have social mobility rates on par with the Scandinavian countries and higher income (, so we know it’s possibile to tackle tough issues from the states.

          Despite high mobility levels, more Americans are involved in their local democracy each year than vote for Congress.
          The Bay Area, with high levels of mobility and migration, is fragmented into 101 cities, each with their own taxes, schools, hospitals, and police. The region has more variations in sales tax than Europe. They fiercely resist amalgamation even though there are no major cultural differences, many were founded after WW2, and almost half of them are immigrants. Despite this unparalleled fragmentation, Tiebout competition seems to be the main driver of reforms and investment.

          I would bet on American parochialism, restlessness and the interjurisdictional competition that produces more than the US ability to build consensus and implement thoughtful federal reforms.

        • paul frijters says:

          the Chetty paper you refer to is not convincing on the mobility rates. It uses a very particular age band and limited income aggregation, things that matter an awful lot when calculating mobility rates. So don’t get too hung up on that paper and its widely misinterpreted findings.
          Besides, I was thinking more of physical mobility and its implications for community cohesion than income mobility.
          Interesting to hear about the diversity of experiments though.

  3. paul walter says:

    It seems to sound a bit like the Agora of ancient Athens, something fundamental to democracy as an early system, inviting rather than discouraging participation.

    I ran from the idea of plebiscites and popularly elected presidnts after the republic failed referendum of some time back, on the horror of “Ray Martin or Alan Jones for President”, but Packer is dead and could Murdoch really swing something for the populist right any more, anway?

    But wasn’t Parliament itself a sort of sortition or at least until its ossification and “capture” by vested interests?

  4. Nicholas Gruen says:

    “The Swiss have voting advice generated by the business community, consumer groups, labour groups, etc. Those long-standing advice generating institutions have a real incentive to be true to their constituency.”

    Well Paul, you obviously know Switzerland a lot better than me, but that second sentence floored me. Having vested interests advise their constituencies on how to vote – what could possibly go wrong?

    • paul frijters says:

      One way to put it is that essentially all interests have become vested in the Swiss referenda. In that kind of Beckerian world, you dont have a few well-organised vested interests who can defraud the rest of the population when they are not paying attention. Everyone is paying attention.

  5. Nicholas Gruen says:

    Yes, I get the idea, but they’re also campaigning non? That seems problematic to me.

    • paul frijters says:

      yes, they are campaigning. But the Swiss seem to have developed a strong bond of trust with the advisory institutions that they feel represent them, and as a result seem to discount the campaigns of marginals. Whether they resist the siren call of particularly powerful special interests, like the security apparatus, the pharmaceuticals or the banking industry? I don’t know. Maybe not. They probably figure that the main negative of those is on other countries anyway.

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.