Dedieu and Toulemonde: Taking political chances with sortition

Franck Dedieu, a professor at the IPAG Business School, and Charles Toulemonde, a research engineer, write in Le Croix.

This short and readable essay is critical of sortition, or at least of the proposals currently discussed in France, but is not completely hostile to the idea. The authors avoid some of the most common knee-jerk anti-sortition arguments and make some interesting and valuable points.

Taking political chances with sortition
29 November 2016, Franck Dedieu and Charles Toulemonde

Machiavelli attributed to chance more than half of human actions. Free choice and individual will would therefore control the minor part of history. Miserable fate! And yet, over the last several years, and more so over the course of the present presidential campaign, the idea of drawing by lot representatives of the people made a breakthrough in the political agenda.

A proven system

In 2012, Ségolène Royal imagined citizen juries supervising the elected officials. Today, Jean-Luc Mélenchon, the leader of France Insoumise, has relied partially on chance last October for selecting delegates to his convention.

Arnaud Montebourg, a candidate in the primary of the left, wants to do the same for selecting the members of the senate. The environmentalists of the EELV, the activists of the Nouvelle Donne party and the members of the Nuit Debout movement of the spring of 2016 have crowned sortition with all the political virtues. This idea of a horizontal Republic is based on a simple argument: the elected, having become the professionals of politics, are living in a closed vessel in an increasingly inbred system and do not represent the social and sociological realities of the electorate. However, as usual in politics, we must be wary of silver-bullet ready-made solutions.

Controversial legitimacy?

On reflection, this stochocracy (from the Greek stokhastikos, randomness, a term used by the philosopher Reger de Sizif) moves away from democracy, rather than approaches it. There is a risk that sortition would strengthen the foibles of the very “electoral oligarchy” is denounces. How will the two political classes share power? Will the “elected deputies” regard themselves as equal to the “loto-senators”? The “chosen” will have the upper hand of the electoral legitimacy while the “commoners” will only have the legitimacy of the lucky draw. What a distance between the Oath of the Tennis Court of the deputies of the Third Estate and the oath of the casino of the Mélenchonists!

Fundamentally, the political women and men, the advocates of sortition, not only are unlikely to lose their power but are finding in the pretense of sharing it a convenient tool for extinguishing the fire. The casino game is their means of redemption… Machiavelliesqe. Furthermore, politicians drawn from civil society, supposedly the source of a better representation of the people, are more often lost in the partisan maneuvers and end up deceived.

Democracy upended

Not democratic then, and in addition not really republican. Sortition carries a part of the “divine”. Providence guides the hands of the innocent, and it is Providence that arranges the grains of sand that determine the lives of each. Théophile Gautier saw “chance [as] the pseudonym for God when He chooses not want to give a sign”. This horizontal democracy contains above all a vertical, supernatural, miraculous dimension. This reverses the most basic republican principles. In fact, universal suffrage is the opposite, being grounded in the real and the tangible, the counting of votes, in the the enumeration of the ballots according to a simple law of arithmetic rather than the law of heavens.

Representativity undermined

More trivially, sortition itself can be contested according to its own principal argument, that of the faithful representation of society. In fact, allotting representative is extracting at random balls from a bag. Like in all good magic shows, the public’s attention is directed toward the drawing of the balls, but not really toward the nature of the bag. Or, the “container” plays a key role in this story of the the democratic “content”. If the bag is rather large, the results of the allotment are therefore very faithful, and representative as argue the advocates. The power is therefore with those who determine the contents of the bag from which the sortition is to be carried out.

When Jean-Luc Mélenchon decided to select the majority of the delegates to the convention by allotment among the activists of France Insoumise, it certainly diluted in theory the power of the leaders of his movement, but it doubtless reproduced the sociological biases of the activists (under-representation of certain professions and of age cohorts). This raises therefore the questions of representation of different groups of citizens such as those of the gender parity or of minorities.

As a solution, France Insoumise has therefore to apply statistical methods that are more advanced than the simple allotment procedure also known as the “Monte Carlo” method. Alluding to the games of chance in the a district of the principality, not known for its socialist values, this method of sampling long used by engineers and researchers. Applied to politics, the more modern methods create a sampling method according to quota rules in order to assure a more faithful representation of each minority. This fragmentation could be used without limit. Thus, it applies equally to voting!

Advertisements

23 Responses

  1. As another indication that sortition is becoming a force in French politics, Daniel Cohn Bendit, who reportedly used to dismiss the idea, now sees it as the way toward the renewal of political representation.

    Liked by 1 person

  2. Dedieu and Toulemonde make a good point about how sortition can be used in a restricted way that allows elites to maintain significant control of outcomes: “Fundamentally, the political women and men, the advocates of sortition, not only are unlikely to lose their power but are finding in the pretense of sharing it a convenient tool for extinguishing the fire.”

    The way the jury is used in the judicial branch is suggestive of how sortition could be introduced in highly controlled ways that largely result in reproducing the status quo.

    It’s rare to see a proposal or an experiment that completely cuts the thread with hierarchical control. What would recursive sortition — sortitive control of sortition — look like?

    Relatedly, it’s interesting how a number of NGOs that run sortition projects and citizen juries are not themselves internally structured democratically but in an authoritarian manner.

    But even highly controlled use of sortition can help mainstream it and lay the foundation for its expanded use. It’s fascinating that these discussions are even happening in France.

    Liked by 1 person

  3. Jonathan,

    Thanks – good points and interesting questions.

    > What would recursive sortition — sortitive control of sortition — look like?

    Yes. A sortition-based system cannot have its rules set by a non-allotted (i.e., elite) body. It must have an allotted rule-setting body. To what extent such a rule-setting body should be separate from the allotted parliament handling day-to-day decision-making is a detail that must itself be considered by the rule-setting body.

    > it’s interesting how a number of NGOs that run sortition projects and citizen juries are not themselves internally structured democratically but in an authoritarian manner.

    Yes. That is a very good point, IMO, which may be revealing that those NGOs may not be aiming at a radical democratization of the oligarchical status quo.

    To be fair, bootstrapping a sortition-based system is not easy since sortition is really suitable to a large group, high-resources, high impact situation, and less suitable when those conditions do not hold. If, say, newDemocracy (which is probably one of the more high-resources NGOs in this space) wanted to democratize its internal structure, it is not obvious how it would go about doing it.

    > But even highly controlled use of sortition can help mainstream it and lay the foundation for its expanded use.

    Possibly. But there is also the risk that such misapplication of sortition can make it appear as just another tool used by the elite to buttress its power, which can delegitimize sortition itself.

    It is very important, I think, that those who really want to see sortition used as a tool for radical democratization make a very clear distinction between situations where sortition results in democratic power and situation where it is no more than a tool for manipulation.

    Liked by 1 person

  4. *** I agree with Jonathan Crock and Yoram Gat comments, except about one point : I am disturbed by too close a relationship between democratic sovereignty and the direction of a political group. There is a basic difference. No citizen must be excluded from juries from his political behavior and sensitivities (even suspectedly antidemocratic), whereas it is logical that an activist group includes only those with the same basic aims, which implies some « authoritarianism » – who disagrees with the basic aims must choose « exit » and create a new group.
    *** An activist group without this « authoritarian » element would be actually an « aristocracy-to-be ».

    Like

  5. *** Jonathan Crock says Dedieu and Toulemonde make a good point about how sortition can be used in a restricted way that allows elites to maintain significant control of outcomes: “Fundamentally, the political women and men, the advocates of sortition, not only are unlikely to lose their power but are finding in the pretense of sharing it a convenient tool for extinguishing the fire.”
    *** We must consider the known political affinities of Dedieu (I suppose the same for Toulemonde). He belongs to networks linked to French « hard-republicanism » : supporters of the polyarchic formula, but with a strong stress on a strong national State grounded on general suffrage, « laïque », « souverainiste », against free trade and economic globalisation (« which gives power to multinational companies »), against unrepublican use of the « separation of powers » …
    *** Their model is therefore a biased form of polyarchy, with a basic stress on national general suffrage. The ultra-republicans are eccentric among the French elites (and among the Western elites !), but nethertheless have a clout because they are attuned to the French intellectual tradition and to values grounded in the history and anthropology of an important part of France. A noticeable fact is that the hard-republicanism clout extends from the right to the left, orthogonally to the right/ left division.
    *** The idea of a strong State grounded on general suffrage makes the French hard-republicans ideologues very suspicious of sortition, and very sensitive to the potential uses of it to allow the dominating political Establishment to « extinguish the fire », and to dilute the effect of general votes it does not trust anymore.
    *** But alongside the sympathizers of hard-republicanism value ideas as civic equality and real sovereignty which may lead them to sympathy with dêmokratia (ortho-democracy), which in a modern society implies practically minipublics through sortition.
    *** The relationship is therefore complex. The ideologues of hard-republicanism criticize sortition, but we see for example Montebourg alongside courting hard-republican sensitivities and proposing an allotted Senate.

    Like

  6. *** Among the thoughts which may come by reading Dedieu et Toulemonde, we must notice the deep mixing of two concepts : the concept of mini-public through sortition (with stratification possibly used to lessen the error margin), and the concept I would name « mimicry» : through quotas the governing body must be « alike the general society ».
    *** The clearer example of this confusion occurs when they mention the case of « France Insoumise », the leftish-populist movement of Melenchon, who « decided to select the majority of the delegates to the convention by allotment among the activists of France Insoumise », which « doubtless reproduced the sociological biases of the activists (under-representation of certain professions and of age cohorts). » And they suggest quotas to have « a more faithful representation of each minority ». It is clear that if Melenchon follows this idea, the selected group will be neither statistically representative of the movement, nor statiscally representative of the French dêmos. It will be a sample of France Insoumise activists biased to mimic the French dêmos along some parameters.
    *** A strange fact. This confusion, by intellectuals without totalitarian affinities, leads to a kind of « representativity » which reminds us of former totalitarian movements : a German national-socialist party which could claim to be representative of the entire Volk (at least of good blood and true race) because it could claim followers from all the social classes ; or a French communist party which could claim to be, by the sociology of its militants, the party of the working class – whereas a noticeable part of the workers did not follow the Stalinist ideology.
    *** Mimicry is not representativity.

    Liked by 1 person

  7. Andre,

    >a sample of France Insoumise activists biased to mimic the French dêmos along some parameters

    Yes, the distinction between a randomly-selected minipublic and mimicry is a valuable one. Although both bodies might look the same, the most important factor (ideology) would be overlooked, unless one assumes that ideology mirrors socio-economic factors (or that activists are more representative participants as they are less prone to false consciousness). This is why, from a liberal perspective, the default principle for any mini-public has to be quasi-mandatory participation.

    Like

  8. Keith mentions the thesis « activists are more representative participants as they are less prone to false consciousness ». Yes, ideas of this kind are more or less explicitly used against sortition. Activist women represent better women than ordinary women ; or women more educated, with university degree ; or women belonging to more conscientized classes of the society … These ideas may be elaborated and spread inside different ideological spheres, totalitarian or polyarchic, and can circulate between these spheres, strenghthening a diffuse hostility versus dêmokratia and sortition.

    Liked by 1 person

  9. Andre,

    >These ideas may be elaborated and spread inside different ideological spheres, totalitarian or polyarchic.

    My concern is that those advocating non-mandatory sortition on this blog are part of the same mindset. If a mini-public is to provide accurate statistical representation then the default position has to be that all those called to serve should take up the calling, and this will require a combination of the stick and the carrot. And even if everyone chosen turns up, the more educated, those with university degrees and the “conscientized” are more likely to engage in speech acts (which may well be better received by fellow jurors), hence the need to limit the participation to silent deliberation and voting. It’s interesting that this is the perspective on deliberation — “This would allow voters time to deliberate in their home” — adopted by the author of the Wired article recently posted by Peter Stone on the Kleroterian blog: https://www.wired.com/2012/05/st_essay_voting/

    Like

  10. Sure, nothing promotes democracy like forcing people to keep silent.

    Like

  11. Yes it is paradoxical, but the preservation of the equal freedom of the vast majority of citizens who don’t get to participate is vital.

    Like

  12. >nothing promotes democracy like forcing people to keep silent

    PS I recommend Andy Dobson’s Listening for Democracy in which he criticizes the focus of the deliberative democracy movement with talking, rather than developing apophatic listening skills:

    Although much prized in daily conversation, good listening has been almost completely ignored in that form of political conversation we know as democracy. This book examines the reasons why so little attention has been paid to the listening aspect of democratic conversation, explores the role that listening might play in democracy, and outlines some institutional changes that could be made to make listening more central to democratic processes.

    Like

  13. *** Keith Sutherland writes : « If a mini-public is to provide accurate statistical representation then the default position has to be that all those called to serve should take up the calling, and this will require a combination of the stick and the carrot. And even if everyone chosen turns up, the more educated, those with university degrees and the “conscientized” are more likely to engage in speech acts (which may well be better received by fellow jurors), hence the need to limit the participation to silent deliberation and voting. »
    *** I agree with the first point, not with the second one. No sovereignty will have value without deep deliberation, including face to face deliberation which has its drawbacks but likewise its benefits : tendencially, at least, more empathy between people, less influence of factions, lesser risk of conformity .
    *** Face of face deliberation was a common reality of the Athenian democracy – on the Agora ; or of Geneva described (idealized ?) by Rousseau – in the coffee shops. It was not possible in general assemblies or in (large) juries, and any trend towards deliberation would have created a risk of mass opinion movements, moblike, engineered by factions. Rousseau favored silent deliberation of the assembly, but that was after the face to face deliberation in coffee shops. Probably in Athens there was a possibility of face to face deliberation in the Council – and in the Second Democracy the seats in the Council were allotted, to lessen the factional risk.
    *** In a modern dêmokratia, electronic technology allows large juries made from a string of small juries, and therefore allows an element of face to face deliberation.
    *** There will be anyway in a modern dêmokratia something akin to face to face deliberation in Agora : the talks and arguments in Internet. But we know their strong drawbacks, whereas face to face deliberation in small juries may be organized with smarter procedures. And real face to face has a specific efficiency to heighten empathy.
    *** A modern, dynamic, society is logically a strongly politicized one, as the society has to make new choices corresponding to new situations. Low politicization can occur only if the citizens think there is really no choice ; but this feeling will be rarer in dêmokratia. Then the political life will be high on internet. The problem is not whether to have face to face arguments, we will have ; but it will be better to add to the unregulated Internet a strong element of face to face deliberation in the minipublics, with good procedures. And good procedures, designed to avoid that the argument is conficated by the more rhetorically able and therefore allow better deliberation, will, as byside effect, lessen the differences of influence.
    *** That said, sure, there will never be equal influence of citizens in dêmokratia ; not only because of differences in rhetorical skills, but likewise because of differences in the skills in reception and evaluation of messages. If we are democrats, we must accept this fact. That may disturb some hard-equalitarians. They have to remember that anyway the unequality of political influence between citizens will be always much lesser than in any other known political system.

    Like

  14. Of course, Sutherland is not proposing avoiding deliberation altogether. Deliberation is unavoidable in the agenda-setting phase of decision making. So in a process where elites set the agenda, elite deliberation must take place. It is only the hoi polloi that Sutherland proposes to keep silent.

    Like

  15. Yoram is right, I am not proposing abandoning deliberation. In the isegoria chapter of my thesis I distinguish between deliberation in the public sphere (Andre’s examples of the agora and the coffee house pertain) and deliberation in political institutions. To argue against the former would be both illiberal and plain silly — appropriate only to a society like North Korea.

    But the problem that should concern us is how to ensure that deliberation in political institutions is properly representative. Habermas’s early work focused on the public sphere and he has been heavily criticised for his later focus on deliberation in political institutions. I’m doubtful that a 500-strong allotted body ever engaged in meaningful face-to-face deliberation (the boule was a collective secretariat, not a deliberative council) and see no way of ensuring that it could be conducted in a representative manner, for reasons that I’ve provided at length in earlier posts. I’m puzzled that Andre claims that such a discursive style would not lead to conformity as the empirical evidence would suggest otherwise. I’m all for empathy, non-conformity, motherhood and apple pie (in the public sphere), but if we want our juries to be representative of what everyone (or at least the plurality) would think under good conditions then we need to follow very closely the original Athenian model. In Athens any of the hoi polloi had the right to speak (though few did), but that is simply impossible in large modern states, hence the need for representative isegoria.

    Like

  16. Andre,

    >In a modern dêmokratia, electronic technology allows large juries made from a string of small juries, and therefore allows an element of face to face deliberation.

    That’s an interesting idea and might well work for big-ticket items like Brexit as it would be very expensive. Given the likelihood that face-to-face deliberation would adversely affect the statistical representativity of each juror (as they could easily be swayed by a single highly-persuasive member), it would be necessary to view each jury as a single individual from a statistical point of view. Assuming that each jury had (say) eight members, and statisticians (and pollsters) tend to view a random selection of 1,000 as the minimum (depending on the majority threshold), this would mean an aggregate jury size of 8,000. For Condorcettian reasons, it would also be necessary to ensure that each jury deliberated in camera, that there was no communication between them, and each jury would need to listen to the same balanced information and advocacy before deliberation (as with a trial jury). If Brexit had been decided in such a manner then it would be fair to say that the outcome indicated what everyone (or the majority) would have decided under good conditions.

    PS each jury would also be free to call on outside advocacy as it saw fit. The final (electronic) vote could either be the majority verdict of each minipublic or else each juror could vote independently. The former would be more “deliberative” but the latter would better reflect aggregate public opinion so would be more likely to be perceived as legitimate from the point of view of the vast majority of citizens who didn’t get to deliberate and vote themselves. Hopefully this would assuage Yoram’s concern about the exclusion of the hoi polloi from deliberation and my concern about the danger of deliberation affecting statistical representativity.

    Like

  17. > statistical representativity of each juror (as they could easily be swayed by a single highly-persuasive member

    Sorry, I should have said: “statistical representativity of the aggregate jury (as each deliberative minipublic could easily be swayed by a single highly-persuasive member

    Like

  18. Andre makes an important point about the miss-application of sortition in advocacy organizations such as political parties …”who disagrees with the basic aims must choose « exit » and create a new group.”

    Sortition is appropriate in large organizations where membership is mandatory or quasi-mandatory (where exit would be a hardship)….For example, a society, a condominium association, a union, a worker-owned business, a consumer co-op, etc. Sortition is not an all-purpose democratic method.

    However, a two-layer sortition process COULD be appropriate in a large advocacy organization, if the member mini-public was charged with hiring and firing the governing board for the organization (they can become better informed than the general membership facing a leadership election).

    Like

  19. Terry,

    >Sortition is appropriate in large organizations where membership is mandatory or quasi-mandatory (where exit would be a hardship)

    Agree, and the nation state is the best example of such an organization. The quasi-mandatory criterion appertains to membership of the microcosm for the same reasons that it applies to the macrocosm. I think Andre, you and I agree on this regarding the sovereign body, and I’m coming round to his perspective on the representativity of a large number of small deliberative minipublics (although it would be very expensive to implement).

    Like

  20. *** Keith Sutherland writes : « I’m all for empathy, non-conformity, motherhood and apple pie (in the public sphere). »
    *** These things may be generally good by themselves in any context (although at least some non-conformities may be very unpleasant, and accept them in deliberation may be a rough duty) ; but I was speaking from a systemic political point of view.
    *** Empathy in face to face deliberation is useful in dêmokratia.
    *** First, empathy in face to face relationship helps deliberation, because it leads to listen even to persons with low rhetorical skill, and it prevents to exclude too quickly some points of view. Real empathy here goes along with minimal face to face respect. It is easier to scorn and ignore a point of view in internet than in a face to face argument.
    *** Second, some level of empathy is necessary between the members of a collective sovereign, who must consider himself as a collective person and, likewise, will be divided by high politicization. In an aristocracy that can be achieved by many face to face contacts in a small group, and by self-consciousness in front of the low classes. In a democracy it is a problem. Frequent wars with military service could be useful, with face to face warrior companionship in front of the ennemy; but they have much lower probability than in ancient Greece. Therefore a high level of political work with an element of face to face deliberation would be a positive factor (with other institutions).

    Like

  21. Andre,

    >empathy is necessary between the members of a collective sovereign, who must consider himself as a collective person and, likewise, will be divided by high politicization.

    Good point. I agree that empathy is essential between the members of the collective sovereign, my concern is not to accidentally recreate the mechanism that applies to an aristocratic sovereign (the only difference being that the aristoi are selected by lot rather than by birth, or wealth). I’m increasingly persuaded by your model of aggregating the decision output of multiple small deliberative minidemoi, but cannot see how this can work outside big ticket items as each aggregation would require around 8,000 jurors in order to ensure representativity (as each jury would have to be viewed as a single person). The numbers are determined by the ideal size of a small deliberative group (I’ve assumed 8, but I guess it could be up to the high teens) and statisticians’ and pollsters’ views on the minimum size of a representative sample (I’m going by John Garry’s view of 1,000 as the absolute minimum). I’m also assuming that moderators/facilitators are ruled out by the quis custodiet? principle.

    Like

  22. Keith,
    There is no particular “absolute minimum” for a representative sample. you will not find other statisticians using 1,000 as some magic number. It all depends on the frequency that you can tolerate a significantly unrepresentative draw. The same is true of your concern about the influence of an unusually persuasive speaker. you can NEVER get 100% certainty of exactly what the entire population would decide if given the same opportunity, motivation, and resources….It is always a matter of probability and the risk (potential harm) of a “bad” decision COMPARED to the risk of “bad” decisions by elected, or hereditary elites.

    As for having multiple small mini-publics replicating a deliberation, even if the total number were modest, there would be added confidence if they all reached the same conclusion, but cause for re-examination if they reached different decisions (like the super-majority requirement discussed on some previous post, where a second minipublic would be drawn if the vote was close, e.g. 51-49, )

    Liked by 1 person

  23. Terry,

    No-one is looking for certainty, probability is quite sufficient and it’s fairly straightforward to make the calculation trading off sample size and decision threshold. But I do think illocutionary imbalance in a small deliberative group is 100% certain, hence my argument that we would need to treat the decision outcome of each group as that of a single individual.

    >the risk (potential harm) of a “bad” decision COMPARED to the risk of “bad” decisions by elected, or hereditary elites.

    I don’t see how such epistemic considerations can be operationalised until some considerable time after the decision has been put into practice. As such we need to focus on the decision rule and its perceived legitimacy. One may choose to argue that (for example) Brexit or the recent US presidential election result was a bad decision but (absent concrete evidence of sinister influence) the decision is still perceived as legitimate. We both agree that aleatory legitimacy will be greater the more minidemoi reach the same conclusion.

    I’m hopeful that Andre’s model for aggregating the decision outcome of multiple small deliberative forums may lead to some common ground between us all. If so then we need to focus on issues like the maximum size of a deliberative group — Coote & Lenaghan 1997, put the number somewhere between 12 and 24, whereas Dahl argues ‘I think that the optimal size for a working committee of actively participating members can hardly be more than ten or a dozen and is probably less.’ (Dahl, 1990 p. 53)

    Refs
    ====

    Coote, A., & Lenaghan, J. (1997). Citizens Juries: Theory into Pratice. London: IPPR.

    Dahl, R. A. (1990 [1970]). After the Revolution: Authority in a Good Society. New Haven: Yale University Press.

    Like

Leave a Reply

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

WordPress.com Logo

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

Twitter picture

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

Facebook photo

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

Google+ photo

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

Connecting to %s

%d bloggers like this: