Roslyn Fuller: Don’t be fooled by citizens’ assemblies

Highly recommended post by Roslyn Fuller on

I believe that the biggest threat to democracy is the belief among the current societal elite that what they want and what democracy is are the same thing – and that tweaking the rules of the game to get what you want is therefore right, just and somehow unto itself democratic.

This trend that sees democracy as a set of particular decisions, rather than just as a method for making decisions, has been well under way for some time and tends to divide the world between ‘informed’, ‘correct’ decisions, and ‘uninformed’ ‘incorrect’ decisions. ‘Correct’ decisions are automatically democratic; ‘incorrect’ ones are not.

One of the ways that is currently in vogue for ensuring ‘informed’, ‘correct’ decisions is to hold so-called citizens’ assemblies, a democratic ‘innovation’ that many leaders currently feel assured will bring them the results that they want.

This is followed by a nuanced (and sceptical) examination of the use and abuse of randomly-selected citizens’ assemblies, focusing on (wilful) misunderstandings of Irish CAs:

British politicians and intellectuals apparently feel themselves entitled to just blithely repeat these myths as a justification for holding such assemblies on all manner of decisions in Britain.

34 Responses

  1. Fuller’s criticism of the rhetoric around citizen assemblies is quite good. Her attack on sortition, on the other hand, is rather weak (“it’s identity politics”).

    But since Fuller discloses that to her mass-voting is the only legitimate way to make decisions, then that rules out sortition at the outset:

    [A]ny law depends a great deal on its real or perceived legitimacy and I’m at a loss to understand how that can be achieved without some process that allows each citizen a vote, whether that be directly (my preference) or through voting for elected representatives.

    It doesn’t matter if we are happy with the outcome or not, she says – vote we must.

    Democracies are intentionally constructed to only require a majority of votes cast. Thus, that one is unhappy with a result is no indicator of its democratic or undemocratic nature.

    This also completely ignores the crucial question of who decides what we vote on, but that question seems to be lost in the all-encompassing rush to vote.


  2. Oh for crying out loud, Keith, is it really so flipping hard for you or Roslyn to admit the bleeding obvious–that when it came to Brexit, nobody had a plan, and nobody had the vaguest idea what they were doing? Is that really not obvious NOW, with Britain ready to crash out of the EU? If there’s one thing the Athenian democrats did right, it was NOT treat the voice of the people like it was the voice of God–at their best, they were quite capable of recognizing when popular rule produced mistakes. It would be pretty sad and pathetic for sortition advocates to forget that. As for Brexit–if that isn’t a pure and simple democratic mistake, I don’t know what is.


  3. Peter,

    If you read Roslyn’s post in full you will see that although Brexit is featured in the introduction, her principal target is the Irish citizens’ assemblies, but I think your rather hysterical reaction nicely illustrates her general argument against epistemic “democracy”. Her new book In Defence of Democracy which includes a (partial) defence of sortition “in its proper place”, defends electoral democracy against referenda and other populist initiatives.

    >If there’s one thing the Athenian democrats did right, it was NOT treat the voice of the people like it was the voice of God–at their best, they were quite capable of recognizing when popular rule produced mistakes.

    Yes indeed. As the external examiner of my PhD thesis you will recall that I argued for a sortition-based alternative to the Brexit referendum that was closely modelled on 4th century nomothesia.


  4. I’ve just read the section “Sortition — The False Democrats” in In Defence of Democracy and it confirms my worst fears regarding how poorly-designed sortition proposals are already giving the whole field a bad name. Roslyn’s principal targets are citizens’ assemblies designed to correct “wrong” popular decisions (Nick Gruen’s and UCL’s proposals for (anti)-Brexit CAs, overseen by “a board of respected citizens”), “pure” sortition (Terry Bouricius and his derivatives), and epistemic and deliberative theorists who believe “every informed person would make the same decisions they themselves would” (p. 127). I also agree with her that agonism is an essential feature of democratic politics and any attempt to move to a rationalist consensual model is counterproductive (and unhistorical, from an Athenian perspective). Deliberative democrats are simply wrong in claiming classical Greek provenance.

    She believes (wrongly IMO) that sortitionists believe their proposals would increase levels of public participation (Carole Pateman told me that she believed participatory and deliberative democracy had nothing in common). But her section on representation rightly claims that the tiny bodies advocated by UCL, Sortition Foundation, newDemocracy etc cannot claim to accurately represent their target population. I would be curious to learn if my own Brexit proposal would suffer the same fate as Nick’s proposal.** My calculations would require a jury of around 6,000 to return a decision with the necessary confidence to match the referendum results. A body of this size would be equally representative of any size population (assuming true random selection and quasi-mandatory participation). As for the provision of “balanced” information, this would be entirely in the hands of the rhetors of the Brexit and Remain camps, constrained only by the modern version of the nomothetic water clock.) And of course this can be tested empirically — repeat the experiment several times and see if the decision is the same (within an agreed margin of error).

    It will be interesting to see how Roslyn’s book is received (it was only published yesterday).

    ** PS I voted for Brexit in the referendum (sorry Peter).


  5. PS to Nick Gruen and UCL we need to add Rory Stewart MP, Compass, The Archbishop of Canterbury, Gordon Brown and various superannuated Cool Britannia luvvies and pop stars.


  6. Let’s cut the crap, shall we? If the Brexit vote had narrowly gone the other way, you wouldn’t be saying any of this. And if you’re still looking at the Brexit situation, and think everything is fine, well, I’m not quite sure what to say about your own personal epistemic abilities. Your political house is a raging dumpster fire right now. And if your response when someone points that out is to call them “hysterical,” well, that’s just pathetic.


  7. Peter,

    My concerns are purely with the decision process not substantive issues and my advocacy for a sortition-based Brexit alternative was before the referendum. Those who are now belatedly calling for a CA are resolutely anti-Brexit.


  8. Te only thing I can say about your proposal is that we, when proposing sortition, are experiencing the same as when we defend referenda. Referenda at the initiative of governments are no democratic instrument, it is an instrument of dictators. Please don’t mix referenda at citizen’s initiative with plebiscites and are not allowed in Switserland and some other states with direct democracy. This way you are choosing the wrong enemy. And wrongly designed panels, sortition or not, are also not democratic instruments. I think we agree on that.

    Liked by 1 person

  9. Given that Rosyln lumps sortitionists in with arch-enemies of democracy like Brennan, Caplan, Somin, Rauch, Wittes and Bell, and is an advocate for electronic participative democracy I was amused by this paragraph in the section of her book that addresses some of the problems with mass deliberation:

    Indeed, various methods could be used to encourage the maximum amount of deliberation among participants. People could, for example, debate in smaller randomly assigned groups. These could be of any size (ten people, a hundred people), and could be composed completely randomly or on the basis of representing certain segments of society (e.g., age, gender, ethnicity). Fuller, In Defence of Democracy, p. 185

    Sounds familiar? In fairness, Rosyln is arguing that every citizen who wishes should be able to participate (rather than those selected by lot) but is unfazed that “many of these participants come from typically active, time-rich segments of society: retirees, students or people who were always particularly politically interested” (p. 167). The advantage of quasi-mandatory sortition with large samples is that it improves the vicarious participation of Joe Average rather than the usual suspects listed above. Roslyn, however, rejects the representative principle tout court, arguing that the difference is scale between ancient city states and large modern states is of no consequence, on account of the ability of modern electronic platforms to extend the boundary of the polis beyond those who are able to hear the herald’s cry “who wishes to speak?”


  10. I’ve just finished reading Roslyn’s new book In Defence of Democracy and recommend it to anyone on this forum. The first section The Terrible Truth: People Aren’t All That Stupid or Evil will appeal to Yoram as it confirms his view that there is no such thing as “rational ignorance” — voters only abstain from participation as there are no real choices available to them. The second part Fixing Politics the Anti-Democrat Way lumps sortitionists in with high-profile free-market and epistemic (anti)democrats largely on account of calls by Nick Gruen, David van Reybrouck and an assortment of superannuated celebrities to adopt small randomly-selected deliberative fora to undo Brexit, Trumpism and other populist evils (this is why I keep banging on about the danger to our whole movement of poorly-designed and unrepresentative voluntarist sortition proposals). The third section A World You Might Want to Actually Live In (Fuller Democracy [pun intended] is one of the most passionate defences of participatory democracy since Carole Pateman’s Participation and Democratic Theory (1970) and Benjamin Barber’s Strong Democracy (1984). As she dismisses Downsian rational ignorance she views electronic platforms as a viable way of recreating 5th century democracy in large modern states.

    The one thing that puzzles me is her unwillingness to accept that the 4th century reforms amounted to an improvement on 5th century mass democracy. While it’s true that the majority of Athenian politics had to do with decrees rather than laws, that doesn’t mean that 4th century nomothesia should not become a template for modern usage. I agree with her that all nomoi first had to run the gauntlet of the Assembly — that’s why Alex Kovner and myself are currently working on a modern analogue of the assembly stage and also an attempt to ensure that isegoria in large modern states is as representative as isonomia. But, as I mentioned in my previous comment, representation is not something Roslyn is interested in — so long as ho boulomenos then that’s OK. I think the vast majority of denizens of the basket of deplorables would not see that as an improvement on ousourcing their voice to the orange demagogue.


  11. Fuller starts of correctly – if incomplete – by defining democracy not as a set of particular decisions, rather than just as a method for making decisions. Already her next sentence leads her astray, she wrongly rejects the concept of a democratically “right” decision, due to her problematic definition that this would be the belief among the current societal elite that what they want.

    As so often, I maintain that we must define a “right” democratic decision on a completely different level. A right democratic decision is merely falsifiable by stating exactly the empirical outcomes which can falsify the democratic decision. No more but also no less. The opposite, a “wrong” decision, we could just as well call a “metaphysical” decision.

    Obviously, a “right” decision according the democratic demarcation criterion does not tell us anything whatsoever with the degree of being right or wrong on its various subject matter criteria which is a fuzzy value in the range between 0 and 1.

    So I restate the initial definition to something useful: “Democracy is a method for making the right societal decisions.”


  12. Hubertus,

    If so, then what makes the decision a democratic one, as opposed to one taken by an oligarchy, autocracy or whatever? Or are you saying that democracy is just more likely to take (espistemically) “right” decisions than other decision mechanisms?


  13. Thanks for asking. The second: As soon as we ensure “right” decisions, democratically, we will be on a path to “better” decisions, epistemically.


  14. Thanks for the confirmation Hubertus. Most democratic theorists would take the opposite view (democracy [should] operationalise the equal right to do the “wrong”** thing). Good that we clarify this, so that we don’t continue talking past each other.

    ** scare quotes as one man’s right is another man’s wrong.


  15. At the risk of confusing you:

    The right democratic method could producing the wrong subject matter decision. This is unavoidable at times, but will improve, due to the proper democratic method’s inbuilt falsifiability.

    Whereas, if you use a metaphysical democratic method (Fuller’s definition) you will be stuck at the same level of error, or worse, errors will accumulate over time as bad decisions pile on prior bad decisions. Such a metaphysical democracy must collapse.

    As they say: Most problems start with their solution.


  16. Hubertus,

    I’m sure that democracies (metaphysical and otherwise) will collapse on account of bad decisions (take a look at 5th century Athens), but that doesn’t render the decision process undemocratic.


  17. Like a rotten apple full of worms is still an apple?

    The term “democratic” per se is not helpful as it applies to many methods or systems of very different quality. The quest is to define a better variant of democracy than today’s.


  18. >The quest is to define a better variant of democracy than today’s

    Agreed. But the principal criterion is representative fidelity — epistemic considerations are the icing on the cake.


  19. Disagreed. The better democracy is the one which yields the best societal decision. This will survive against the others. (And for your next question, please refer to what I wrote above at 10:48.)

    You shovel mud and want a process to mix it representatively.
    I shovel mud and want a process to find the gold nugget in it.


  20. For sure Cleisthenes and his followers reformed the Athenian Polis for quite similar reasons as the professional politicians of our time try to do.So I can see no point in Fuller’s argument. The motivation does play no role but only the functionality of the democratic process itself (same power to all people, room for interaction for people from very different parts of society, etc). As the German political historian Christian Meier emphasized, we can see it as a fact, that the Athenians never ever wanted to build a “democracy” – until they had one. Even the term “democracy” occurs much later then the Clesthenian Reforms.

    But what I miss even more in Fuller’s defence for “virtual tools” in the democratic process is her total ignorance for the role that empathy plays in democratic deliberation and decision making. Plus the fact that empathy doesn’t take place from the distance. People have to be in one room together, in small, ever changing groups to have an opportunity at all to develop empathy whith each other. Espescially if the selected group is not very homogeneous.

    One of the key failures in the institutional setting of the ancient Athenian Democracy, the reason why it failed at last and left us with roundabout 2200 years of no serious democratization attempts was the big role that “rhetoric” still played in it. Professional speakers, trained by sophists able to “move the masses” were still much too important. If I look at our modern attempts in using sortition in comparison I see already much better processes than the Athenians ever had.

    For all this reasons I find Fuller is totally wrong. She doen’t have a clue why democracy works and why it fails when it fails.

    Fuller’s proposals in the end lead to the same consequences that made Plato reject democracy as a constitution that would always cause the “rule of the most stupid”, not using the wisdom of the (empathetic connected) crowd but the manipulability of the disconnected mass. And I myself really don’t want to see a second rise of epistocratic philosophy for the next 2200 years.

    Every sorted citizens’ assembly of our time shows what politics could be like and that its processes are much superior to everything else we’ve got in politics up to now. There have not to be “perfect, uncorruptable processes” right from the start to convince a broader audience that a constitution with sorted citizens’ assemblies in its heart is a realistic alternative that will work for anyone. Even for the epistocratic aristocrats. ;)


  21. Ardalan: A constitution with sorted citizens’ assemblies in its heart is a realistic alternative that will work for anyone.

    What about the overwhelming majority of people who do not take part directly in the empathetic deliberation? Claiming (without an adequate statistical foundation) that tiny groups of volunteers are “representative” (because nobody chose them) is evading the problem. Besides which, most deliberative democrats (aka epistemic aristocrats) are concerned with reasons not representation (essential in large states), so this doesn’t strike me as a democratic solution.


  22. Keith,

    Though you acknowledge that (for many reasons) elections fail to elect relatively small bodies of accurate representatives, you often also fall back on dismissing small RANDOM bodies by saying “nobody chose them.” Elections almost never involve much REAL “choice.” A handful of self selected and prefiltered candidates can only offer the legal fallacy of voters “choosing.” I won ten elections (five for city councilor and five for member of the state legislature). Usually the voters had a choice between me and one other resident, and a couple of times I ran unopposed. I either was or was not a good representative, but it had nothing to do with voters “choosing me.” Perhaps (if they weren’t inevitably victims of rational ignorance), the voters could reject a WORSE representative of the two on offer, but that is a ridiculously low bar.

    A relatively small (let’s say under 100 member) stratified random mini-public can fairly represent the general population in cases where they come to a common agreement on policy that 80% (a huge majority) of them agree on. If the mini-public is nearly evenly split on a policy, then the process needs to be repeated with a larger mini-public.

    Liked by 2 people

  23. Terry,

    I agree that electoral democracy as currently practiced is sub-optimal — Alex Kovner’s proposal for a heavily revised version — The Jurga Manifesto: Towards a Sortition Democracy — is currently in its second draft and will be published in a few months time. My main criticism of the current work on minipublics is not just sample size but the fact that the participants choose themselves rather than being there at the behest of their constituents. This is the worst of all evils as such minipublics benefit from neither of the forms of representation that Pitkin outlines in her book, hence the view of Roslyn (and others) that they should be lumped in with the current backlash against democracy. For some reason this doesn’t seem to bother deliberative/epistemic democrats, born out by the fact that successive drafts of my paper on this topic have been turned down by The Journal of Public Deliberation without even being sent out for peer review.


  24. “The fact that the participants choose themselves rather than being there at the behest of their constituents.. This is the worst of all evils”

    While in today’s “democracies” it is true that self-selection is a big issue, as loud activist groups can distort (the perception of) the general will, self-selection is certainly not the “worst of evil” in a well designed voluntary system of sortition.

    In fact there is zero evil, as we can show with some simple maths.

    Let’s assume that an activist group has 2,000 loud people. There are 66 million people in the UK of which 40 million are voters. Best practice of over-recruiting for a citizen jury needs a factor of 100. So, for a mini-public of 100 we will invite a raw sample of 10.000 citizens. We expect 1,000 to respond, of which we again draw the 100 we need, while improving representativeness with a second draw from those. Hence, the chance of having even one of these loud people in our first sample round is only 50%. That person will volunteer and have a chance of 10% to be drawn in the second round for a 5% chance overall.

    And even in the one case (in 20 citizen juries) where one of the loud activists makes it that person can shift the vote by less than 1%, whereas we see significant supermajorities of 66% or more in typical citizen jury votes.

    Ergo, it is a tad exaggerated to paint a 1:20 chance of a vote distortion of below 1% as the “worst of evil”.


  25. Hubertus,

    If, as Plato, Oscar Wilde and Arthur C. Clarke believed, interest in political matters is a significant and deleterious population parameter then a voluntary random selection that over-represents such persons by a factor of 96:4 will produce a highly distorted sample, which no amount of gerrymandering (stratification) will correct.


  26. But Keith, essentially 100% of the elected legislators are this voluntary self-selected activist sort you are concerned about. A random selection among the entire population, with good inducements, pay, child care etc. (and a deepening sense of civic responsibility), will be vastly LESS problematic in this regard than with an elected body. Your usual come back is that at least the elected representatives were “chosen” by voters to be their agents. I pointed out that this is a legal fallacy. Voters are ONLY able to choose between volunteer activist sorts.

    Liked by 3 people

  27. Terry:> Voters are ONLY able to choose between volunteer activist sorts.

    Yes, and quite rightly so, that’s why Alex and myself refer to the political party as the best (or possibly only) way of instituting representative isegoria in large modern states. As usual you are conflating two entirely different forms of representation — the active representation of interests (Chapter 6 in Pitkin’s book) and descriptive representation (Chapter 4). As in classical Athens, isegoria is the province of volunteer activists, but the role of isonomia is for all citizens (or a descriptively representative sample) to judge between the policies on offer. We agree with you that political parties in their current form do not perform the isegoria role very well, that’s why we are proposing major reforms. But both varieties of representation are, and always have been, necessary and sortition only applies to the isonomia role.


  28. The conflation of the two forms of representation originates not here in our discussion (from Terrill) but is a fact, a severe flaw, in current party systems. Only those who are actively representing particular interests will be nominated by their parties and can be elected into legislator positions which by rights and logic would need descriptive representation (the domain of sortition).


  29. Hubertus,

    We all agree (including Terry, Andre, Alex and myself) that it is a mistake to conflate the representation of particular interests (“parties” in Madisonian parlance) and aggregate judgment (the domain of sortition) in a “single body of men”. Isegoria and isonomia were distinct in both theory and practice in ancient poleis and the challenge is to implement them in large modern states. This will require distinct mechanisms for each (as was the case in antiquity). I need to emphasise that Alex and myself are not defending existing party political systems but we don’t want to throw out the baby with the bathwater.


  30. Which baby?


  31. Representative isegoria (in the form of competition between political parties). We both consider this to be an essential element of demokratia (pace my first book The Party’s Over, which I now believe to be a good candidate for the bonfire).


  32. Thanks for clarifying, Keith. And agreed, with one caveat: current political parties are not designed for isegoria. In a future of isonomia by sortition, parties would be specialised “one issue” organisations competing for citizens’ approval (and tax money), and politicians would be respected innovation leaders, not just today’s manipulative talking heads trained in (irrational) mass psychology, as Ibrahim correctly pointed out above.


  33. That’s quite close to Alex’s model (currently in its second iteration). He also incorporates a Bayesian strand whereby parties live and die (partly) on the basis of their epistemic competence.


  34. […] A citizens’ assembly on Brexit was widely discussed in the UK. […]


Leave a Reply

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

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

Facebook photo

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

Connecting to %s

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

%d bloggers like this: