This post is a rebuttal to the conclusion of Keith Sutherland’s 2013 paper “What sortition can and cannot do”, whereby sortition is deemed inappropriate for the “advocacy role” of representative legislature owing to alleged improper representation.
Since posters here address each other with their first names and especially since I will likely be chatting with him soon, I will refer to Keith Sutherland as Keith despite having never chatted with him yet and despite that going against the convention of referring to scholars one is discussing by their last names.
I have seen that there is a long standing feud of sorts between Keith and Yoram, and I find it appropriate to mention this here given that what seems to be at the origin of the conflict is precisely largely their differing views on the extent to which sortition should pervade the selection of policymakers compounded by a more general disagreement on political ideology according to a fairly standard left-right antagonism. I have no dog in this fight nor a particular affinity to either’s position whether on sortition or on political ideology more generally as far as I can tell, appearing to hold an intermediate position in both respects. As such though I am posting on Yoram’s site against Keith, this should not be construed as an attack by “team Yoram” against “team Keith”, as further evidenced by the fact that I had never communicated with Yoram until a few days ago when a renewed focus on sortition led me to make a few comments on his site. I imagine that Keith has heard all my arguments here before, and that most others have too, but since I couldn’t find a similar post on the topic I figured it would at least be useful to have a post dedicated to it. I am presuming that readers are familiar with the paper and the concepts it discusses and so I am not reintroducing them here.
Starting with the somewhat lesser objections to sortition for the advocacy role, it is typically considered that for successful deliberation groups should be rather small, in which case we suffer from typically unrepresentative samples. However this can be solved by having multiple smaller advocacy bodies whose conclusions are then all presented to the judgment body. And while this goes against the general sortition orthodoxy, I consider a small group to already be typically sufficiently representative, as supported by the fact that it’s considered fine to have a jury of only 12 members which is borne out by, e.g., the verdicts on the highly publicized and politicized shooting cases in the United States which I found to be systematically normatively right and thus impressively non-partisan. In the absence of political parties, it’s probable that issues would be much less polarized as well which makes it even more likely than small groups would reach the right decision even without being particularly representative. This ideal is largely what motivates the Habermasian model of advocacy bodies that Keith speaks of which sacrifices representativity for discourse, which serves to help people learn of and consider the different aspects of an issue, but the latter which also could be done on one’s own given sufficient time.
Keith’s criticism of the flip side of the tradeoff, sacrificing deliberation for representativity, which characterizes the Fishkinian model that he also speaks of, also falls apart quite easily. The main objection here apart from the lack of good deliberation is that this model requires moderation which brings up the question of who is charge of it. This can be solved by either having the body itself make the rules of moderation, or else leave that to another body as in Terry Bouricius’s multi-body model.
The argument that sortition makes legislators more susceptible to bribing owing to lacking the pressure of re-election is valid, but it can easily be fixed by increasing the harshness of punishment of accepting bribes (and of offering them) until the total disincentive is equal to what it was with a milder punishment plus the risk of losing re-election.
And now we reach the core argument, the lack of isegoria. To start off, isegoria as defined by simply the “amount talked” could in theory simply be ensured by having requirements that everyone speak the same amount and propose the same amount of bills, which Keith acknowledges. But it’s true that in a body composed of many legislators, such requirements are impractical, and in their absence, certain voices would indeed come to dominate, and even with these requirements certain legislators would speak more eloquently and thus convincingly than others, as Keith points out. But in fact this does not matter, because loquacity and eloquence are simply a descriptive characteristic like any other, meaning that with a large enough group, there should be equal distribution of loquacity and eloquence among different interest groups/constituencies (italicized because this is the actual rebuttal of Keith’s main argument), provided that this is equally distributed among the groups, which admittedly is probably not quite the case, with, e.g., women tending to be more reserved and poorer, and foreign people tending to be less eloquent. Ignoring this latter issue for now (which makes sense since this issue is not part of Keith’s main argument) all that is required to ensure an equitable amount of advocacy is to have a large enough sample that groups tend to have a similar proportion of loquacious/eloquent people; in other words all that this means is that there is an extra variable that requires equal descriptive representation, meaning that the sample needs to be somewhat bigger, nothing more. The failure to realize this appears to be the most important reason that Keith opposes sortition for the advocacy role, who goes as far as to claim that this inequality in advocacy is a priori ex hypothesi necessary in the sense of inevitable, which is in fact false.
Admittedly this doesn’t hold if we consider that there is only a small number of voices that can dominate, which means that the sample of such voices which are the ones that will be most impactful is going to be necessarily small and thus typically fairly unrepresentative (and although as I said before that shouldn’t matter since 12-person juries work well, in this context it would be more problematic because it would tend to involve loudmouths trying to impose their point of view rather than having a reasonable discussion). However if that is the case then the same would hold for an elected body as well, which means that while due to this phenomenon sortition would be poorly representative, it would still be no less so than elections. And while indeed sortition would presumably select a wider number of quiet legislators which implies a greater disparity in loquacity which would yield a greater imbalance in speech, if there is a limited number of voices which can dominate, this effect would be nullified by simply having a sufficient number of legislators that all that the available “slots” for dominant voices are filled.
However though I said we were ignoring the differences in loquacity and eloquence between groups, this is actually important, since they cannot be corrected by increasing sample size. And there are other factors as well which could impact the success of advocacy not through direct persuasion as with differences in verbal skill but rather through the effect of biases: this includes anything that influences the perceived value of the advocate (e.g., prettiness, dominance), of the advocate’s interest group (e.g., cultural similarity, wealth), or the interests themselves (e.g., the extent to which they are beneficial or harmful to those not part of the interest group). And these would tend to differ between groups as well. So it turns out that there are quite a few factors that would influence success of advocacy which probably differ between interest groups and thus not be corrected by larger sampling.
And so, to sum up, the appropriateness of substituting election for sortition for the advocacy role of legislatures (and here ignoring the occasional unrepresentative sample) hinges on the degree to which the differences between these groups impacts the success of advocacy. If the advocacy is so greatly impacted by this that different groups have less equitable advocacy than with election, then indeed election remains preferable as Keith believes, though not for the same reason. However, again given the success of juries and likewise of citizens’ assemblies, I am quite confident that this is not the case and that even in the advocacy role sortition would be fairer than election, though it is admittedly not certain.
There is quite a bit more that I initially wrote on the subject that I ended up discarding because it wasn’t quite as central, for instance my addressing the different concepts of representation which ultimately proved irrelevant, so if some feel there are things that remain unexplained, I can cover those in another post.
Adam Cronkright
Ali
Ahmed R. Teleb
avanderven
avergne
Alex Kovner
azakaras
Brett Hennig
campbellwallace
Simon Threlkeld
Nick Coccoma
codyhipskind
Common Lot Sortitionist
Conall Boyle
constitutionalism
davidschecter
graytmatter
gtridimas
H.J. Hofkirchner
John Burnheim
jgarry20151
Jochen Krattenmacher
johnedwardtaylor
johnhuang865
Jonathan Crock
keithsutherland
Jacob Richter
kkoreilly
lukewhito
SimonPek
Nathan El
Nicholas Gruen
Owen Shaffer
Patrick Chalmers
Paul Gölz
peterstone
Pierre-Etienne Vandamme
pronfeld
Emil
rcaze
richwbrown
Roger Knights
roslynfuller
Pierre-Etienne Vandamme
shaharavin
tbouricius
Ted Wachtel
Joshua Laferriere
theodorestathis
towardsautonomyblog
Wayne Liebman
Yoram Gat
Oliver Milne
Equality by lot 
Two comments:
1. I don’t see how elections can be justified as having any role in a democratic society. This is an inherently oligarchical mechanism which grants power to an elite which then naturally goes on to use this power for self-promotion at the expense of the rest of the population. Agenda setting, advocacy or final up-or-down decision making – it does not matter. Elections are an anti-democratic procedure.
2. The notion that some groups are unable to represent themselves – and again it doesn’t matter whether we are talking about agenda setting, advocacy or final up-or-down decision making – is anti-democratic. The democratic ideology asserts that people, as individuals and groups, are the best representatives of their own ideas and interests. The notion that they need to have others speak for them or make decisions for them is a classic guardianist position. Wrapping these claims in paternalistic language about the need to take care of women who are reserved, or immigrants who are ineloquent is also an age old guardianist rhetoric. If someone wants to recruit a professional speaker to help them out, let them decide that they want to do so, and how they want to do so, rather than shove it down their throats via the oppressive tool of elections.
LikeLiked by 1 person
Two somewhat personal asides:
1. Sutherland’s tired tactic to counter the obvious point that elections do not serve the weak but rather amplify the advantages of the powerful is to assert that elections are inherently democratic since they give each person a vote (while sortition supposedly disenfranchises all those who are not allotted). Hiding behind such false formalistic formulas of political symmetry is of course the standard tactic for those who want to keep the current oligarchical arrangement while pretending to support democracy.
2. While it is true that Sutherland is a right-wing zealot while I am a socialist, and while it is true that Sutherland supports authoritarian oligarchical government while I am a democrat, the real cause of our “feud” (as opposed to substantive disagreement) is Sutherland’s mendacity which has been on display on this blog many times. Sutherland’s attempt to present his proposals as being motivated by values which are similar to ours, while in fact they are motivated by authoritarian oligarchical sentiments, is one aspect of this mendacity. But there have been many instances where the mendacity is much more superficial and obvious. If someone (other than Sutherland himself, who is immune to reason and facts) is interested in examples, I can refer them to some choice exchanges in the archives.
LikeLike
I haven’t delved enough into your exchanges to determine whether or not Keith indeed is a habitual liar, though I have already come across a few such accusations by you. You can link some examples if you like, but based on what I have seen I get the impression that it’s probably not exactly lying but a combination of uncharitable hyperbole and genuine belief that his policies are the most beneficial for society. And I do also get the impression that you too engage in uncharitable hyperbole, precisely due to calling him an authoritarian liar rather than something more nuanced as I have proposed. And obviously one’s exaggeration invites a symmetrical response from the other.
Though I do believe that sortition is probably much better, there are some valid concerns about it, which I believe Keith overstates, but I likewise consider it excessive to outright label as authoritarian someone who favor elections out of these concerns while nonetheless remaining quite open to sortition and indeed advocating for it partially.
I don’t consider it anti-democratic to recognize that some groups tend to speak more or better; this is simply recognition of a (likely) fact, which would impact the advancement of their interests, so this is a genuine concern precisely for ensuring democracy. As for elections being elitist, in fact sortition could precisely be seen as a kind of elitism itself in favor of the more volubile and eloquent groups, which could be theoretically corrected with elections which would enable the poorer speaking groups to select advocates for themselves.
Legislators recruiting professional speakers is something I hadn’t thought of and that does in fact seem like the best solution. If done well that should equalize the “advocacy power” between legislators and the groups they represent, and as such this would in fact resolve this problem with sortition of unequal advocacy skills. So I consider this an important breakthrough; thanks Yoram (though I wish I had thought of it myself!)
LikeLiked by 1 person
I want to add that the fact that you gave me a new idea very relevantly showcases the value of discussion between people rather that just thinking about things on one’s own. And hence why I consider comment sections/discussion forums a particularly useful feature of the internet.
LikeLiked by 2 people
Legislators recruiting professionals to speak for them seems like an enormous can of worms. Funding issues aside, the speakers may bring more to the table than their rhetorical powers – their reputation and institutional backing may carry as much weight as their words, and may likewise give them the means and motive to influence the legislators looking to hire them.
As for the eternal beef between Yoram and Keith, in my experience Keith is the more reasonable interlocutor, despite his rightwing politics. This is not to disparage the sterling work Yoram does for this blog, nor his remarkably even-handed editorial practices given the low esteem in which he publicly holds some of its contributors. Perhaps I am just biased by Yoram’s more stubbornly combative demeanour. His bizarre ideological deference to the results of opinion polls in one-party states and dictatorships seems about the same level of misguided as Keith’s anti-wokeness.
In terms of actual, practical proposals, Keith and Alex Kovner’s ‘democratic diarchy’ – in which an elected chamber presents multiple proposals to a deciding jury, with each proposal requiring the support of only a fraction of the elected chamber – represents an obvious potential stepping stone to the sort of fully sortitional system Yoram wants, opening the door to the transition while offering established parties enough of a role that, under pressure, they might conceivably back it. But with both scenarios firmly in the realm of political fantasy, I guess the only cost of their continued hostilities is in hurt feelings.
LikeLiked by 1 person
I’m flattered that Nathan assumes that members of the Yoram Forum will be familiar with the arguments of my 2013 paper ‘What Sortition Can and Cannot Do’ and am glad of the opportunity to respond to his thoughtful comments:
1. The only commonality between trial juries and legislative panels is that they are both randomly selected. The purpose of the trial jury is to determine the facts of the matter (did she or didn’t she do it?), whereas the legislative panel is intended to uncover informed preferences, only epistemic “democrats” believing that there is a right or a wrong answer to political questions. That’s why Jim Fishkin uses the term deliberative opinion poll.
2. A group of 12 established by quasi-mandatory random selection will exhibit a degree of diversity (and the “output” of many small groups can be aggregated), so that may well be acceptable for the final up-down decision function. But when it comes to making policy proposals, what matters is that the group (or aggregation of groups) reflects the ideological diversity of the full electorate and there are good reasons to believe that sortition is not suitable for such a role. The first problem is that, typically, only 4-7% of those selected by lot take up the offer, and this is likely to result in the overrepresentation of activists and those predisposed to public life. Whilst that may or may not make sense epistemically, there is nothing democratic about it. If voluntarism is a significant population parameter then this will mean 93-96% of the target population will be denied any form of political representation. And there is a substantial literature that demonstrates that small-group deliberation is subject to information cascades and domination by more persuasive individuals, which will distort the “output” even more. (But I think we agree on this last point.)
Not so. Participation would need to be quasi-mandatory. It’s worth noting, of course, that voluntarism is the one population parameter that cannot be rectified by stratified sampling.
Absolutely. Moderation is in breach of the quis cusodiet principle. I’ve debated this point with Jim Fishkin, but he insists that (active) deliberation is an essential part of preference transformation. I think this is because he still views himself as a Habermasian DD, even though the evidence (e.g. from the Bloomfield Track poll) demonstrates that it is the information stage that is most important. My preference is for a large silent jury, as with the 4th century Athenian nomothetai.
Hmm, this might well bring the lottery acceptance rate down catastrophically. And lobbyists are pretty skilled at camouflaging their influence even when working with tiny numbers of professional politicians who are constantly in the public eye.
Oliver:> In terms of actual, practical proposals, Keith and Alex Kovner’s ‘democratic diarchy’ – in which an elected chamber presents multiple proposals to a deciding jury, with each proposal requiring the support of only a fraction of the elected chamber – represents an obvious potential stepping stone . . .
My partnership with Alex started long after I wrote this paper. His Superminority Principle is explained at https://alexkovner.com/2020/09/07/superminority/. Note that we do not view it as a halfway house to a sortition-only system, as we believe that elections will always be required for policy proposals. In a forthcoming paper Cristina Lafont and Nadia Urbinati argue that elections are the modern incarnation of the isegoria principle, which was primarily the role of the rhetores, rather than the “masses”. Whilst some like to view that as a support for oligarchy, Alex and I claim that it is the only way of ensuring that the demos have the exclusive right to make policy proposals. (Unpacking that paradox requires a book chapter).
Oliver:> with both scenarios firmly in the realm of political fantasy.
The advantage of Superminority is that it requires nothing more than a change in the decision threshold of the elected legislature. That sounds pretty feasible to me, especially as it means that ordinary elected politicians would (potentially) be in power all the time, rather than only the leader of the majority party some of the time. The only constraint on the power of politicians would be the ability to divine the informed preferences of the electorate (as revealed by the randomly-selected jury), and that sounds pretty democratic to me. As James Harrington pointed out (in 1656) you can have your cake and eat it after all.
LikeLiked by 2 people
Nathan,
> I don’t consider it anti-democratic to recognize that some groups tend to speak more or better
I am not sure what speaking “better” means. Surely some groups speak more. In fact, some groups are essentially deprived of any opportunity to speak in the public sphere and many people have their speech restricted even in private spheres, while other groups and people have their voices amplified. That is a major part of the way our oligarchical society is arranged. I don’t see how the fact of this oppression justifies perpetuating it.
The point is that, again, the democratic assumption is that people and groups are the best representatives of their own interests. People should be empowered to represent themselves (including, if they choose to do so, by recruiting others to speak for them). Sortition does exactly that, while elections are a tool for amplifying the voices of the elites and suppressing the voices of average people. The guardianist argument that some people are better off being forced to select speakers for themselves among the elite (via elections) is anti-democratic.
LikeLike
Nathan,
> I haven’t delved enough into your exchanges to determine whether or not Keith indeed is a habitual liar, though I have already come across a few such accusations by you. You can link some examples if you like
Since you ask, and as promised, I’ll provide some evidence – it is merely a sample. A frequent pattern of lying by Sutherland is his making strident false assertions, often about supposed positions taken by various people. When confronted about the falsehood of his statements, he makes no effort to substantiate his claims, but rather casually shifts his ground and moves on, treating the matter of their veracity as being petty.
A typical example is here. A more extreme example is this one, and a particularly ridiculous one is this one. Note that in all three examples the issue is not about Sutherland’s positions but about straightforward factual claims. In two of those cases the claims are being made with great conviction, and in one of them the claim was made repeatedly over multiple threads. Many more examples can be furnished with some additional searching.
Such behavior is unfortunately often condoned here by saying these are merely honest mistakes rather than a deliberate disregard for the truth. This excuse cannot be accepted if the errors are made repeatedly and no attempt is made to explain how why they were made or to avoid making the same kind of errors in the future. Casually moving on after making a gross misrepresentation of the facts is an indication of indifference with regard to the truth – the attitude of a liar.
Finally, regarding my accusation of Sutherland possessing an authoritarian mindset. This is not hyperbole but simply stating the facts. Sutherland’s authoritarian leanings are expressed in his institutional proposals (forcing the allotted to sit silently while the elite-selected speakers get to dominate the discourse, feeling that it is only natural and fair that elites get disproportional political power) and in his sometimes farcical obsequiousness toward any kind of supposed expertise or “scholarly consensus”. But even more directly it is expressed in his demeanor on the blog, where he repeatedly issues edicts that discussants avoid certain subjects, utilize only terminology that he approves of, or avoid being disrespectful toward the high and mighty in order not to bring disrepute on the forum.
LikeLike
Oliver,
> His bizarre ideological deference to the results of opinion polls in one-party states and dictatorships
My deference is to the ability of people to judge their own situation. This applies to one-party states and dictatorships and even to two-party states and electoral oligarchies. Again, I see this as being part (maybe the most fundamental part) of my democratic commitments.
If your claim is that opinion polls are an imperfect tool for observing people’s judgements, or that in certain countries they tend to be more flawed than in others, I am very much open to this claim and would like to hear specifics. I would note, however, that any observation has its limitations, so this is a matter of comparatively weighing the pros and cons of the method of observations we have rather than dismissing one of them because it is not perfect. It seems to me that the alternative method of empirical observation that you lean so heavily on – namely, new reports – is fundamentally flawed and serves as a very poor basis for forming a justified opinion on any government or society.
If, on the other hand, your claim is that people (the Chinese supporting Xi, the Russians supporting Putin and the Muslims in India supporting Modi, were the groups in question in our previous exchanges on this matter) can be duped into thinking that they are doing well while in fact they are being oppressed, then this seems a-priori problematic and requires a great deal of self-assurance on your part which is most likely derived from ethnocentrism or elitism rather than real superior judgement. I have yet to hear from you or anyone else how you would be able even in theory to substantiate such a claim, let alone to see evidence that this is indeed the case. But in this case as well, I am open to change my mind if either a good argument or good evidence is presented.
LikeLiked by 1 person
Getting the more unpleasant part over with first, Yoram, all the links you provided were just more of the typical exchanges I had already come across whereby Keith makes a statement that you consider blatantly false which prompts you to call him a liar, but as previously I can’t find any indication that Keith’s statements aren’t in good faith or that they are even particularly false for that matter. These examples don’t even constitute the “uncharitable hyperbole” I mentioned by which I meant more his labeling of you as an extremist fanatic or suchlike which I have also noticed a few times. And Keith certainly does have the more conciliatory attitude as Oliver mentions here.
LikeLike
Oliver and Keith, thanks for the responses, likewise presenting valuable new ideas for me.
Oliver, good point about recruiting professionals being problematic. The obvious solution that comes to mind is a rule that only speakers that didn’t have some degree of prior fame could be hired. Determining that is naturally going to be somewhat arbitrary, but it doesn’t seem to me to be a complete deal-breaker.
Keith, to your first point about juries serving to uncover facts unlike legislatures, my own retort is that I actually do believe there is such as thing as a factually optimal policy which means that legislatures would be striving to uncover facts as well.
Then as for the issue of the unrepresentativity of voluntary participation owing to self-selection – yes indeed, I wasn’t taking that into account in my post, and indeed I feel rather silly for not having thought of that issue in that context. However that does not appear to be the fundamental argument in your paper about the lack of isegoria; indeed as you state here this issue can be corrected with mandatory participation, whereas your argument was rather that there was an inherent inequality of “advocacy power” between representatives owing to the fact that some people speak more or more convincingly than others, and the purpose of my post was to point out that this specific argument is false. And indeed you have not countered my rebuttal to this very specific argument. And thus funnily enough, if intentional this would constitute a form of “pivoting”, of eluding the topic, thereby lending more credence to Yoram’s accusations against you, though not for quite the same reason.
That being said, you have raised here a quite major problem with sortition, which is that one either has to contend with an unrepresentative sample made up of activists, or a policy of conscription which would be particularly oppressive if we had long-term positions as is likely preferable for some roles. Being reminded of this issue does in fact make me rather less confident about sortition for the advocacy role.
Oliver and Keith, the diarchy/superminority model does indeed strike me as quite promising. But this could also be applied to sortition, allowing for several groups to form within an agenda council/review council or equivalent body who then go on to present their different ideas to the final jury, as presumably you already recognize, so this would just be about improving elections in the context of retaining elections owing to the previously mentioned issues with sortition.
Tell you what, the above issue of the tradeoff between a representative sample vs lack of coercion, combined with possibility of improvement to the electoral system, is sufficient to make me open to election for the advocacy role of legislature, though I still have misgivings about it given the major problems with electorialism. So, rephrasing the title of my post to be more nuanced and inclusive without actually recanting, though I still believe sortition can probably do it all well and even best, it likely doesn’t have to do it all, and it may well even be somewhat preferable to not have it do it all; Yoram, I am partially on the dark side now :) The point being that we shouldn’t intransigently rule out this partial-election model. I think it’s likely the different systems will likely just have to be tried out extensively, and which system is best will likely just be determined empirically a posteriori rather than theoretically a priori as people here are trying to do. And likely even then we won’t have a clear answer since there should in fact be a clear winner in theory for there to be a clear winner in practice.
LikeLiked by 1 person
Hi Nathan, I think we have a meeting of minds. Let me address the problem of perlocutionary imbalances (J.L. Austin’s term) in speech acts:
Persuasive skills are generally biased in the favour of the highly educated (rhetoric used to be a key element in the classical university curriculum) and this is also the case with deference to high-status persons and professions. I’m sceptical that the law of large numbers will ensure that all groups/constituencies are equally represented, especially given the pressure on contemporary elites to comply with dominant social norms. Under our current arrangements the Average Joe can select a rhetor who makes a persuasive representative claim. All Alex and I are suggesting is extending this right beyond the two warring tribes and enabling the (virtual) demos to choose the most plausible. This is compatible with the Argumentative Theory of Reasoning (Mercier and Sperber 2011), whereas you seem to adhere to the classical alternative.
>there is such as thing as a factually optimal policy which means that legislatures would be striving to uncover facts as well.
Sure, uncovering the facts is the principal job of deliberation. But, according to Bernard Manin’s modified position (2005), this is best done through adversarial, rather than Habermasian exchange. Interesting, of course, to note that Manin’s background would be the French legal system, which is far less adversarial than the Anglo-American.
>allowing for several groups to form within an agenda council/review council or equivalent body who then go on to present their different ideas to the final jury
Yes, that’s true, but only if the different ideas accurately reflected the ideological diversity of the target population. Existing legislative assemblies do to a degree (depending on the electoral rule) represent the ideological diversity of the electorate, but that diversity is destroyed by the need to obtain a majority. Reduce the decision threshold and the proposals will better reflect the diversity. Here’s an excerpt from a forthcoming paper by Alex:
It’s hard to see why a randomly-selected agenda council (a la Bouricius) would do a better job and, more importantly, Superminority has the added benefit of allowing all citizens a (small) say, rather than just the aleatory elite who draw the golden ticket. Oligarchy comes in all shapes and forms.
>long-term positions as is likely preferable for some roles.
Most sortition theorists are very wary of long-term appointments.
>which system is best will likely just be determined empirically a posteriori rather than theoretically a priori
True, but scientific progress is based on testing hypotheses, rather than just doing stuff and seeing what happens. I’ve made the case several times that the “repeatability” case for adjudicating the output of sortition bodies can be put to the test. If the decision outcome varies between different samples, then which one can claim democratic legitimacy? Whether the consistency reflects the forceless force of the better argument or procedural considerations (quasi-mandatory participation, balanced advocacy, deliberation within) is another matter.
Thank you for the opportunity to debate these important points in a collegial manner!
Keith
LikeLiked by 1 person
Nathan,
> Keith makes a statement that you consider blatantly false which prompts you to call him a liar, but as previously I can’t find any indication that Keith’s statements aren’t in good faith or that they are even particularly false for that matter
You seem to have missed the point. The fact that those statements are false is not disputed by Sutherland. The bad faith is that he doesn’t care whether his statements are true or false. All he cares about are his conclusions – which are preconceived. The arguments are just post-hoc rationalization. If they turn out to be false, there will be others, quite possibly also false. It doesn’t matter.
LikeLike
Yoram, so your main point is that Keith doesn’t actively dispute your allegations. I have indeed noticed his tendency to just let them slide which I found a bit odd but which I ascribed to a mix of innate indolence, rational apathy since it seems unlikely that he would be able to change your mind, escalation avoidance, and ban avoidance given that you are the owner and moderator of this site. And I recall he did in fact dispute your claim in the first link, insisting, against your objection, that it does seem that you consider sortition the only form of democracy, which does seem very plausible given that you call him an authoritarian for supporting election. So I really feel you are at least overstating Keith’s dishonesty and likely outright imagining it in many instances. But I do now agree there appears to be a kernel of truth to your claims given that he has again failed to address my actual main argument against his paper, even after my having called him out on it in my first reply. Interestingly this matches my general experience with the psychology of “leftists” vs “rightists”, with the former seeing ill intent where there is the none and the latter acting rather placid and ignoring accusations against them being both very typical of each respective side.
LikeLike
Keith, I am myself glad that a pre-eminent scholar on sortition is willing to converse with an average-Joe ignoramus such as myself :)
As I mentioned in my reply to Yoram though, you still haven’t addressed my main argument, so this evasion does appear to be deliberate, though I suspect the fault likely lies in your subconscious and thus eludes your awareness as appears to be the case for many people who pursue such a tactic. You now address the issue of perlocutionary imbalances of my post which is indeed an argument against sortition for the advocacy role of legislature, but that is not the argument in your paper, or certainly doesn’t appear to be so. The argument in your paper is that it’s the differences in illocutionary force between individual legislators which matter, (presumably) even if the average between groups is the same; that is what I contest. As such there are indeed good arguments against sortition in the advocacy role of legislature, it’s just that the main argument of your 2013 paper is not one of them.
> Sure, uncovering the facts is the principal job of deliberation. But, according to Bernard Manin’s modified position (2005), this is best done through adversarial, rather than Habermasian exchange. Interesting, of course, to note that Manin’s background would be the French legal system, which is far less adversarial than the Anglo-American.
(How does one do the quotation box?)
Interestingly this is very much in line with and analogous to another issue I have been researching which is precisely the question of which among civil law and common law (or more precisely inquisitorial vs adversarial judicial process respectively) is best, a debate I found well covered by Zywicki 2007 on Tullock – though the difference there is even more extreme given that the fact-finding in the inquisitorial system is done by a single person rather than a deliberation group. Advocates of the inquisitorial process claim that it is superior due to the judge being only intersted in representing the truth while the attorneys of the wrongful team deliberately seek to obfuscate it, but advocates of the adversarial process counter that the latter generates a broader range of viewpoints and thus ultimately yields a more complete and truthful picture despite the deliberate attempts at obfuscation; the adversarial process also tends to expend more which makes it more costly but which could ultimately be worthwhile if it helps bring out the trurth. So here too there are good arguments for both positions and it doesn’t appear that one of the two systems can be considered indisputably superior.
The other items you bring up I have already covered.
LikeLike
Nathan,
This discussion regarding Suthreland’s lying has probably gone on longer than is interesting for either of us. It seems we don’t agree on this matter. But I do have to correct you on one point.
> you call him an authoritarian for supporting election
That is very far from the case. I gave a list of reasons (at the end of this comment) which substantiate my claim that Sutherland is an authoritarian. The fact that he supports elections is not one of them.
LikeLike
Yoram, ok I guess I misunderstood and thereby misrepresented your position on authoritarianism vs election. And yes I certainly don’t feel the need to pursue the question of Keith’s potential mendacity.
LikeLike
Hi Nathan,
I confess it’s a long time since I read my 2013 paper — perhaps that’s an example of my “innate indolence”! My views on sortition have evolved considerably over the last 20 years — in his more charitable moments Yoram would describe that as inconsistency or opportunism, but I prefer to view it as being a good listener and open to persuasion. Forgive me if I defend the speech act entailments of my current position, rather than 10 years ago.
In our (Alex and myself) view, modern large-scale democracies should minimise the role of individual legislators — we agree with Lafont and Urbinati that isegoria should be representative, rather than providing a platform for Pericles, Demosthenes and their modern equivalents. The formation of policy groups under Superminority will be a black box and no doubt some legislators will have more influence than others, but the prime constraint will be the need to anticipate the preferences of the citizens’ jury. There will inevitably be differences in the persuasive powers of the advocates, but this can be equalised to some extent by time limits (the Athenians used a water clock). There will be no perlocutionary imbalances within the jury, as all they get to do is listen and then decide the outcome. We view the reduction in the role of individual legislators as one of the (democratic) advantages of our hybrid model. Given that the term “descriptive representatives” does not parse in the singular, I hope this explains our position.
>Advocates of the inquisitorial process claim that it is superior due to the judge being only interested in representing the truth while the attorneys of the wrongful team deliberately seek to obfuscate.
Bear in mind that most citizens will be sceptical about the notion of political “truths”, especially given our current highly polarised public culture. As sortition will lead to the effective disenfranchisement of the vast majority of citizens, it’s essential that the proceedings should be seen to reflect the widest range of ideologies and preferences. And I highly recommend the Mercier/Sperber paper on the Argumentative Theory of Reasoning, although Helene Landemore summarises it nicely in her earlier book.
LikeLike
It’s worth examining Yoram’s idiosyncratic definition(s) of democracy:
>The democratic ideology asserts that people, as individuals and groups, are the best representatives of their own ideas and interests.
but, on the other hand
>This applies to one-party states and dictatorships and even to two-party states and electoral oligarchies.
Hmm. The word democracy actually means that “the people” have “power”. “The people” references a singularity (albeit in an aggregate form), but some deliberative democrats and sortition activists have conflated this with the plural term “persons” or “citizens”. Jon Elster, for example, defines democracy as “any kind of effective and formalized control by citizens over leaders or policies” (Elster, 1998, 98), without any apparent concern as to which citizens are in control. Sortitonists, however seek to empower a representative sample, which “stands for” the entire citizen body.
But what does this mean? Obviously the votes of a large mandatory random sample will mirror the preferences of everyone else to an acceptable degree of accuracy, the only issue being (from an epistemic point of view) the rational ignorance threshold. But what about the other tasks that politicians are required to do, that depend largely on speech acts?
That’s why Alex and I argue that the role of the randomly-selected jury be limited to listening to arguments and then voting. Anything more contravenes democratic norms. I don’t see what is authoritarian about this conclusion.
LikeLike
> “most citizens will be sceptical about the notion of political “truths”, especially given our current highly polarised public culture”
I think this is a misreading of the situation. Our highly polarised political culture is in part a product of voters and party members believing passionately that there *are* political truths, and that the other side is wilfully denying them for nefarious reasons. And they’re often right about that! Very few people, in my experience, describe their political preferences as preferences, except out of politeness in mixed company; many more describe them as a matters of right and wrong, or (at their most relativist) good strategy and bad. Even when their preferences boil down to what’s good for them personally, people like to tell themselves what they’re *really* after is the general good, which just happens to coincide with their own interests. Thus they may approve in principle of a system they can see will be good at ascertaining the truth, even if, once implemented, they object to its decisions in practice.
LikeLike
If I might clarify my last comment with a parallel from reproductive biology, where the female is the “limiting factor”. In a democracy any number of votes can be aggregated, but if everyone chose to exercise their isegoria rights then the result would be an unintelligible cacophony. That’s why isegoria needs to be representative, hence Lafont and Urbinati’s claim that the political party is the modern incarnation of the classical-era rhetor. Superminority is intended to ensure that isegoria better represents the ideological spectrum of the whole electorate — just how representative will depend on the decision threshold.
LikeLike
Oliver,
My point is that politics is not like science, there are no “right” answers, pace the view of (some) epistemic democrats.
>Thus they may approve in principle of a system they can see will be good at ascertaining the truth, even if, once implemented, they object to its decisions in practice.
Yes, but that’s why it’s important to focus on procedure, rather than outcomes. Until very recently citizens were prepared to accept the outcome of elections, so long as they were seen as free and fair. The challenge for democratic innovators is to get the procedures right (Estlund would agree), and I’m not persuaded by John Burnheim’s claim that this can be ensured by publishing the committee deliberations on the internet. What people will want to see is that the party representing their preferences/prejudices is given an equal hearing vis a vis the other guys, and that would incline us to the debate format in which advocates employ whatever rhetorical tricks they think will bring them success.
LikeLike
Keith, thanks for the further comments. Indeed I assumed that the paper I was referring to was fresh in your mind which was a bit naive of me, and I obviously would have forgotten much of what I wrote 10 years ago. I do still get the impression that you haven’t grasped my argument since the Pitkin quote is the one in your paper and it precisely seems to indicate that disparities between individual legislators are considered a problem in and of themselves even in the absence of perlocutionary imbalance between groups. Since you are addressing my complaint I am satisfied that there is seemingly no voluntary dishonesty on your behalf here. And like I have alredy acknowledged, there are still sufficiently strong arguments in favor of your system that it appears to me a legitimate alternative to a full-sortition model.
Oliver, good point about most people considering that there are objective political truths. That being said an adversarial system where different actors have their own entrenched opposing views of the truth could in fact be precisely best for uncovering the objective truth by presenting more varied viewpoints to the final jury/voters.
LikeLike
Nathan,
I’m still struggling to understand, but I hope our downgrading of the role of individual legislators will, in practice, overcome any disagreement that we have. Disparities in status, economic power, cultural clout etc are a fact of life, but Superminority is intended to give the underdog a fighting chance. It’s also the case that voters will elect representatives who manage to overcome structural disparities (Manin’s principle of distinction) and/or select representatives who claim to speak for them, even in the absence of any (descriptive) affinity. For many years in the UK the Rt. Hon. Anthony Wedgwood Benn, Viscount Stansgate, was seen as a key representative of the underdog.
LikeLike
Well I’ll give it another shot, providing an actual example.
Let’s imagine that the only 2 kinds of people are women and men who are of equal numbers in the population, and have equal advocacy power (illocutionary force), such that half of each sex are powerful advocates.
If we have a small legislature, say of only 10 members, then it’s of course likely that there won’t be even a roughly equal number of both sexes, it’s quite likely for there to be 7 and 3, 8 and 2, or even 9 and 1 or 10 and 0; more relevantly here, even if there are 5 men and 5 women, again because of the small sample of each sex, it’s quite likely that there won’t be an equal advocacy power in each group, so for instance among the 5 men there could likely be 4 powerful advocates and among the women only 1 or vice-versa. This would be a problem because then even if the sexes were equal in number, the sexes would be unequally represented owing to the difference in advocacy power of the combined legislators of each group, which is your precise concern.
But now let’s say there the legislature is now a large body composed of 1000 members. Because of how statistical sampling works, it’s much more unlikely to have the proportions of women and men deviate much from half-and-half, so typically there will be about 500 of each (though it would still be unlikely to have exactly 500 each). And because we have a much larger total sample, the sample of each sex is also much larger, which means that one would expect the amount of powerful advocates in each sex-group to also be about half within the group (since we postulated that half of the total population of each sex are good rhetoricians); in other words we would expect to have about 500 men, about 250 of whom are powerful advocates, and likewise about 500 women, likewise about 250 of whom are powerful advocates. This means that there would in fact typically be an equal distribution of powerful advocates among both sexes, meaning that the interests of each sex are equally well advocated for.
That is my argument: with a large enough sample, provided the populations of different groups are equal in advocacy power, the groups will typically have equal advocacy power owing to the typically equal distribution of powerful advocates. And this means that individual differences in advocacy power between legislators is not a problem, contra your thesis.
Sure there are probably in practice differences in advocacy between population groups, but your position is that even if there weren’t any such differences there wouldn’t be equal representation, and what I am saying is that that is false.
And sure there are many more variables than just sex, but the point still holds: with a large enough sample, on average, again ignoring differences between population groups, there will be proportional advocacy for groups. Admittedly for small groups, there won’t be such equity, since small groups will have too few representatives to typically have a proportional number of powerful advocates. But in fact that is not much more of a problem than the more general one which is that sortition won’t typically select a very representative number of legislators from those small groups anyway. And I get the impression that this is in fact something that you don’t quite get either, despite the fact that you precisely acknowledge that individual allotted legislators aren’t representative and thus shouldn’t be called representatives individually. Individual legislators don’t actually represent any group, they don’t represent their fraction of the population amounting to 1/n, with n being the number of legislators. Individual legislators have no constituency. Rather it’s just the body as a whole that represents the population, that’s it. And yes for big enough groups there will also be a proportional representation of those groups. But that does not hold for small numbers. For a group that makes up 1/1000 of the population, in a body of 1000, often you will get 1 legislator of that group (who is then a “representative” of sorts), but often you will also get 2, less often higher numbers, and often 0. So there already isn’t typically a proportional representation of that group on average. Often that group will have double the advocacy power they deserve by having 2 legislators instead of 1 (presuming average advocacy power of these legislators), and often that group will have no advocacy power at all. Note that we have this exact problem just as much with the final “legislative jury” with sortition. The difference in advocacy power between individual legislators will merely compound the variability in advocacy power of the group depending on the draw, which is already much too high anyway for them to be considered to have a consistent equitable representation anyway. Now in slightly larger groups, you could have representation that passes a certain, necessarily arbitrary, threshold of mean equity of representation, but then because of this extra issue of difference in advocacy power between individuals, the value of mean equity of advocacy power would fail to pass the threshold, which could be of course made to pass the threshold if the size of the sample was enlarged, which is why I brought that up. And anyway it doesn’t matter that small groups don’t have equitable representation, because even if they did it wouldn’t do much for them since ultimately the policies that get passed are those supported by the majority, and if they are small they are unlikely to tip the balance on their own. If it did matter then that would be an argument against holding sortition at all even for the final policy jury. But you yourself rightly recognize that it doesn’t matter.
This is about as clear as I can make things. Hopefully it helps.
LikeLike
Thanks for the clarification Nathan. A couple of questions:
1. Is your allotted legislature modelled on existing elected assemblies, and if so why? The two tasks of a democratic legislature (making proposals and deciding whether to approve them) are entirely distinct. Theorists as disparate as James Madison, Robert Dahl and Terry Bouricius have argued that combining the two functions in a “single body of men” is a gateway to corruption.
2. What is the proposed size of your assembly/ies? I think we both agree that the wee pretendie parliaments of some sortition advocates are (statistically) unrepresentative, but how can you have a deliberative group of 1,000? Bagehot argued that the House of Commons c. 1865 was too large for deliberation, which had moved to the much smaller cabinet. People often cite the example of the Athenian Council of 500, but most historians argue that this was an administrative body — the secretariat for the Assembly — and it’s principal role was to protect the sovereign body from factional domination.
>sortition won’t typically select a very representative number of legislators from those small groups anyway.
True, but that’s the beauty of the Superminority Principle (SP). If a minority group can establish a coalition with other groupings, then the proposal that they formulate will have an equal opportunity of being passed into law as that of a hegemonic grouping. And remember Burke’s claim that every corporation (interest group) only requires one virtual representative.
>ultimately the policies that get passed are those supported by the majority, and if they are small they are unlikely to tip the balance on their own.
Of course they still have to pass muster with the citizens’ jury (under the majoritarian principle), but the SP provides minorities with a unique opportunity to present their case and there are myriad contemporary examples of junior coalition partners punching well above their weight.
LikeLike
There is no “my” allotted legislature, I don’t have a favored specific model like you or Terry Bouricius do. But if you want me to propose something as a likely viable alternative to your system, then we can take Terry’s model which I find quite compelling. Terry has bodies of over 100 for the “bill proposing” and “bill deliberating and specifying” roles each, which is perhaps too large for deliberation. But we could also maybe let small groups form in line with your suggestion, and we could also maybe have a single body where Terry has 2 or more. Or maybe even have multiple (very) small bodies for a given role, whether we have bodies take care of the whole proposing-deliberation role as you would or we split the different aspects like Terry, and this small size would enable the most frutiful discussion, and the unrepresentativeness would actually be an asset here, since it would enable the different groups to come up with a wider array of suggestions, which would each be presented to the final voting jury body.
As for the rest, you are mostly preaching to the choir at this point :) I don’t dispute that there are advantages to your system and I agree with what you have written in favor of it.
So again, my point is rather just that there is this one extra argument of yours against sortition which I consider faulty, and which seems to be what makes you deem sortition unviable for the advocacy role, and I am just trying to get you to agree that that argument is faulty and thereby hopefully see that sortition for the advocacy role is not clearly inferior to your model.
I will add while I don’t have a favored specific system for a representative body, I do have a favored specific overarching system of governance, which has popular votation override the representative body like in Switzerland. And of course that means that I don’t “need” to be too preoccupied about the representative body, since if it works poorly it can just be altered by popular votation. I am of course aware of the drawbacks of votation, in particular rational apathy for both the question of voting at all and of researching the issue well to make an informed vote. But even in a large representative body this issue is prominent too since one’s vote is still unlikely to be the deciding one. And the reason I favor votation is that I deem it to have more conceptual legitimacy given that it’s direct democracy, it’s the actual people who take the matters that concern them into their own hands, which of course wouldn’t matter if we could guarantee a representative sample in sortition, but one of the features of drawing by lot is that sometimes one gets a completely unrepresentative sample. Obviously votation only is viable when there are only a few important policies that make it to the vote, since people won’t bother to turn out to vote for more, which is of course why there still needs to be a representative body. Also since the really important issues get voted on by the population, it means that the issues that get voted on by representatives are of lesser importance, which also makes it less important to have very representative bodies. This means that it could even be acceptable to have an elected body make the final vote as is done now, which would have the advantage of eliminating sortition’s drawback of having to choose between compulsory participation and unrepresentative samples. So at this point I would say that for the representative body we could have all sortition, all election, a combination of both, or even some other system like selecting only experts (by sortition) as I have seen proposed by Hubertus Hofkirchner here. I am now open to pretty much anything since I don’t think there is any system that is very clearly all in all superior or inferior to another.
LikeLike
And in fact I realize that my conclusion, viz that in the presence of overriding votation there is no clear best system of representation (though I do still lean towards short terms with mandatory sortition for the final voting body at least), concludes my whole current foray into the topic, since the question of the best system was its main motivation. I am glad to have had the opportunity to discuss things with you Keith as well as the other people I discussed with, viz Yoram, Terry, Oliver, and Nicholas Gruen, both in and of itself and because it helped me develop my ideas and reach my conclusion, so thanks to all.
LikeLike
Oh and of course there is obviously also no clear best system of representation in the absence of an overriding entity, rather what I meant is that with an overriding entity the question of the system of representation doesn’t matter much.
LikeLike
Though that doesn’t necessarily have to be the case, there could be an obvious best system of representation in the absence of an overriding entity but not in its presence, eg if it were deemed crucial to have an allotted body for the final decision-making in the absence of an overriding entity but not in the presence of one. It’s just that given the flaws of sortition for any body I don’t consider that to be fully the case.
LikeLike
I had a look at the Mercier and Sperber article, interesting stuff, though not exactly surprising, it’s obvious from general experience with people that people tend to confirm their beliefs rather than try to reach the truth, but it’s good to have studies on it. And this further supports the idea of having multiple work-groups in the advocacy role and a separate body for the judgment role. Though again it doesn’t preclude selecting the advocacy body through sortition.
LikeLike
Nathan:> sortition for the advocacy role is not clearly inferior to your model.
Leaving aside the theoretical arguments (I’m still struggling to understand the lacuna), let’s view it from an implementation perspective. If the policy-proposing/advocacy role was arrogated to a body, or bodies, selected by lot this will lead to the disenfranchisement of 99.9% of citizens, along with elected politicians losing their jobs. This is unlikely to be welcomed by either party, whereas the Superminority Principle (SP) will increase the power of all voters and all elected politicians. Unless you share Terry and Yoram’s visceral antipathy to elections it’s hard to see what there is to dislike about it. And on the principle of Occam’s Razor one should always go for the simple option, rather than a multi-body system of Byzantine complexity. And all you need to do is change the decision threshold in the House of Commons and replace the House of Lords with large legislative juries constituted by quasi-mandatory sortition. The first change is just numerical (there’s noting magical about 50+1) and everyone and his dog wants to replace the Lords with something more democratic.
I agree that popular votation would provide a necessary safety valve, but bear in mind that many, or even most, Brexit voters now regret the outcome of the referendum. It has also been argued that the 4th century nomothesia reforms were a reaction against the wild fluctuations in 5th century mass democracy.
>with an overriding entity the question of the system of representation doesn’t matter much.
Dahl (rightly) insists that the demos must be in control of all stages of the legislative process. That would be true for SP, but less so for alternative proposals. Terry acknowledges that the demos would not be in control of the small long-term voluntary panels in his proposal. He claims that they would be more representative than elected politicians but two entirely different forms of representation are involved (Pitkin, 1967), so this is comparing apples and oranges.
LikeLike
Keith, yes of course your system is less of a radical change so it would be easier to get people to agree to implement it. Your specific suggestion for the UK seems like the obvious solution for the legislature there at least initially. And again, your overall diarchy-with-superminority model seems decent, and I would be willing to back it, I just don’t find it clearly superior to a full-sortition model. It seems that you want me to reject full sortition just like Yoram wants people to reject your partial election model, but that’s not going to happen, there are convincing arguments for and against both.
Regarding regretting Brexit: with a proper Swiss-style popular votation system people could just backtrack. Though then of course that causes fluctuation as you mention. But I don’t think that’s all that much of an issue and eventually things would stabilize. And anyway there is little fluctuation in Switzerland.
I’ve explained to the best of my ability why sortition isn’t inherently particularly less representative even in the advocacy role, so I have nothing more to add on that.
Like I already wrote, I think I have discussed everything I felt the need for, I don’t think my writing more would be fruitful for either of us or anyone else. I appreciate your feedback and exchange :)
LikeLike
Thanks Nathan, as I spent seven years developing the democratic diarchy model, I guess I have a lot of psychic investment in it (hence your argument for subconscious factors). And I’m also an apostate from lottocracy (my first book was called The Party’s Over). Perhaps that’s why Yoram has pronounced a fatwa on me.
It’s also worth noting that there are no historical precedents for a sortition-only polity — the only literature being utopian science fiction. Barbara Goodwin, who invented the mythical country Aleatoria for her book Justice by Lottery is (wearing here political theorist’s hat) deeply sceptical about it’s application to democratic politics (as, surprisingly, is Surowiecki).
And thank you for making me examine my argument more carefully. This forum would benefit enormously if other correspondents followed your thoughtful example!
LikeLiked by 1 person
Good points, and thank you too :)
LikeLike
Oh and by the way I have a psycho-cognitive block myself that makes it hard for me to read which is why I have little erudition, so it’s not like I don’t have my own flaws in that domain.
LikeLike
Nathan,
> a quite major problem with sortition, which is that one either has to contend with an unrepresentative sample made up of activists, or a policy of conscription which would be particularly oppressive if we had long-term positions as is likely preferable for some roles
> sortition’s drawback of having to choose between compulsory participation and unrepresentative samples
The solution to this problem is not compulsion (remember, you can lead a horse to water, but you can’t make it drink) but pecuniary incentives. Just give the allotted jurors a hefty salary and, most importantly, a lifelong pension after just one mandate (because they won’t have a second one) and everyone will suddenly realise that they don’t have anything better to do with their lives than being full-time legislators for the next four years.
Admittedly, this rationale might not apply to the richest 1%, but I am not particularly concerned with them being underrepresented for once.
LikeLiked by 1 person
Arturo,
Yes, that’s why we usually preface mandatory with “quasi”. I guess some people prefer bribery to compulsion, but I prefer to think of it in terms of civic obligation. One of the problems with the modern sortition movement is all the talk of the equal rights of the self-governing individual, whereas the ancients privileged the defence of the polis, politics being more a case of jaw-jaw rather than war-war. The subject of Alex’s PhD is the needs of the state, rather than “human rights”.
LikeLike
I have so much to say… but I need to finish reading first (I’m only up to August 30 so far). I just wanted to post this, so in case anyone was checking and concluded the discussion had ended on Sept. 1… that you haven’t seen my pearls of wisdom yet. I’ll try to finish reading the discussion and make a substantive post within a day or so.
LikeLike
I’ll just make a few points instead of responding to piles of things in this long discussion.
1. The continuation of elections with campaigning, etc. poisons the well… If most members of the jury have voted for one or another of the parties in the last election, when they are presenting options under the super-minority plan, they become biased jurors. Rather than making independent judgments. We all tend to defer to people we trust (are invested in) who seem like they know more than we do. Jurors might even self-organize into caucuses based on party allegiance, etc. The benefits of sortition and random selection are forfeited if election campaigns persist. Parties have a distorted and sinister hidden agenda, that supersedes the public interest (they want to WIN), and that often requires hiding some facts and lying about others. Parties that have to appeal to ill-informed (rationally ignorant) voters, is a recipe for disaster. They are a dangerous place to put advocacy and proposal development functions.
2. There are problems with giving a jury multiple choices related to Arrow’s Impossibility Theorem… such that making a single choice from more than two choices (such as yes/no) always leaves the door open to strategic voting. The most troubling is called “raising turkey” where a voter is motivated to rank a choice they think is worst, but has no chance of winning, very high, because they are concerned that their true second choice might defeat their true favorite choice, so they also rank their true second choice low so as not to defeat their true first choice. There are dozens of different tabulation systems (Condorcet, variants, score, Borda count, etc.) but EVERY voting method trying to pick one winner from several choices can be gamed or encourage insincere voting. The Condorcet method that many favor also can suffer from the Condorcet paradox, where option A, beats option B, B beats C, but C beats A.
3. The goal of policy development, or advocacy and deliberation, is not to get the perfect one and only version of a policy that society as a whole would produce if everyone magically could deliberate… The goal is to avoid domination by a powerful subset, and develop policy that is motivated by the public interest, rather than special interests… to be “good enough.” Even an elected chamber may develop different proposals in a parallel universe, or if the same people started the process one week earlier or one week later. There is no need to see if two different deliberating mini-publics would generate the same proposal or have identical speech acts, etc. Exact replication is a fantasy that can never happen in any democracy. Imperfect proposals will arise, be adopted by a jury, and the next mini-public will not be invested (no pride of authorship, as with parties), so will undo them and replace them with something better. Sortition in both advocacy bodies (policy development) and judging bodies will exhibit reversion to the mean.
LikeLike
a very interesting website about multiple choice can be found here https://ncase.me/ballot/ What is not mentioned is that multiple choice produces “weak outcomes” because it is not supported by any majority. In most cases it even is a minority decision.
LikeLike
Terry,
>The continuation of elections with campaigning, etc. poisons the well
If you believe the end of elections means the end of political parties and campaigning, I’m afraid you’re in for a nasty surprise. Campaigning will still occur, just on issues rather than for election, and professional politicians will still exist, in the form of professional campaigners and rabble-rousers who spend their time riling up the public and soliciting donations and cushy post-politics jobs. If there are no formal political roles for them to take, they will simply create informal ones and lobby for more power to be handed over to them. Democracy must be able to endure the presence of these phenomena. It cannot be conditional on their not showing up.
LikeLike
That’s a good point Oliver, one could argue that election is a way of ensuring that rabble rousing is conducted in the full glare of publicity (and constrained by laws designed to minimise the role of sinister influences). The argument to abolish elections is not unlike the case for Prohibition (of alcohol), and look where that got us.
Terry:> If most members of the jury have voted for one or another of the parties in the last election, when they are presenting options under the super-minority plan, they become biased jurors.
Jurors under Superminority do not present options, they just judge the outcome. Although US politics is going through a Two Tribes phase, this is not the case in the rest of the world, and Superminority will increase the degree of pluralism. Alex and I anticipate that the allegiance of jurors will be to policies, not parties, and that allegiances will shift according to the issues being debated.
>There is no need to see if two different deliberating mini-publics would generate the same proposal or have identical speech acts, etc. Exact replication is a fantasy that can never happen in any democracy. . . Sortition in both advocacy bodies (policy development) and judging bodies will exhibit reversion to the mean.
That’s akin to saying it doesn’t matter who wins the election, as the other guys will get in next time.
I’ll leave it to Alex to respond to your point on Arrow’s Theorem.
LikeLike
Oliver,
There is a fundamental difference between issue disagreements, advocacy in a generic state and in a partisan state. Of course groups will always be advocating policies they prefer, and seeking to educate the general public (especially since some of them ay be called to jury service). But the nature of this advocacy is very different in a competitive electoral environment, where the chief goal is to WIN ELECTIONS (more than it is to adopt particular policies), which involves whipping up the base, distorting facts, and demonizing the other parties as the “enemy,” convincing supporters that no other groups can be trusted to tell you the truth, etc. Many hot issues are essentially “fake.” Parties USE these issues to mobilize supporters. This is just good campaign strategy. Parties and politicians are about POWER more than they are about policy. If there weren’t competitive elections around, many of these issues would be boring and could be dealt with rationally without party manipulation. This may be tamed somewhat by the super-minority plan, where politicians can’t unilaterally assert power without jury support…. BUT this also may motivate parties to distort, lie and demonize EVEN MORE, because they can’t exercise power without manipulating the jurors. Thus, continued elections will cause parties to try to shore up tribal loyalty as much or MORE than they do right now.
LikeLike
Terry,
>> The continuation of elections with campaigning, etc. poisons the well… If most members of the jury have voted for one or another of the parties in the last election, when they are presenting options under the super-minority plan, they become biased jurors.
There is surprisingly little evidence for this. British conservatives voted for Brexit in droves despite David Cameron urging them to vote remain. In Florida in 2018, voting rights for felons passed by nearly 2-to-1 in the same year that Republicans, who universally opposed the measure, won the legislature and governorship. Voters have consistently shown that they will vote their conscience even when it conflicts with their political party.
>> The benefits of sortition and random selection are forfeited if election campaigns persist
American elections are the worst possible style for elections. We vote directly for candidates instead of proportionally for parties. We vote at a single moment in time instead of in a rolling fashion. We have no public financing of elections, severe gerrymandering, and a host of other problems specific to our system.
The challenge for the proposing body is that it must be a roughly proportional representation of the ideological diversity of the country. What happens if we randomly select from volunteers? Vested interests will scour social media and other places finding people who are sympathetic to their cause, then recruit them to volunteer. Such a body will be stacked with people favorable to such interests. That will be an arms race every bit as much as elections are currently.
I understand your antipathy to elections as they are now, but try to separate the absurdities of our current system with elections in general. A rolling, proportional election for parties instead of candidates minimizes the craziness, while giving a proportional snapshot of the political spectrum. How do you plan to get proportionality out of self-selection?
>>There are problems with giving a jury multiple choices related to Arrow’s Impossibility Theorem…
There are problems with everything. Of course if you increase the number of choices, the mathematical complexity of aggregation increases, and brings pathologies along with it. But restricting choice is far worse. Take Brexit: there should have been at least three options: remain, a hard (no deal) brexit, and a negotiated brexit, The lack of these options turned it into a circus, allowing brexit advocates to sell pie-in-the-sky. Cognitively, people need three or four options (at least) to be able to make an informed and free decision. We can argue about aggregation methods all day, but the need for more options is essential in any democratic advance.
>>The most troubling is called “raising turkey”…
Strategic voting requires a lot of knowledge about what other voters are thinking. Our juries never meet as a whole. At most they work in small groups, and many will work alone. They will not even be allowed to meet other jurors, except possible being assigned a small group. They work in their home cities and meet only virtually. Also, single use juries make it harder to vote strategically since doing so takes some understanding of aggregation which jurors will not have.
Terry, you are putting too much weight on pathologies. Pathologies exist everywhere in every system. We should decide what decision making structure works best at a human level and then work to mitigate the pathologies, otherwise the tail wags the dog.
>> The goal is to avoid domination by a powerful subset…
Absolutely. I think that’s one of the strengths of our system. Self-selection opens the door to special interests, as it creates an arms race to recruit sympathetic volunteers.
>>There is no need to see if two different deliberating mini-publics would generate the same proposal or have identical speech acts, etc. Exact replication is a fantasy that can never happen in any democracy
Exact replication is a fantasy, but you still want policy to converge to what ordinary people want. Over time, similar issues should be resolved in a roughly consistent way. Deliberative (noisy) mini-publics don’t offer even that much.
LikeLike
Terry,
I think you are probably the only member of this forum who has served as an elected legislator, and this obviously gives you experiential insights that the rest of us can only imagine. But it’s important to realize that the electoral model that Alex and I are proposing is very different from current practice. The danger of throwing the baby out with the bathwater is you can end up with something far worse. I think it’s also worth pointing out that your own ideological hinterland is not well disposed to reformist approaches.
LikeLike
>>There are problems with giving a jury multiple choices related to Arrow’s Impossibility Theorem
We aren’t totally wedded to multiple choice — there’s no reason in principle why the final jury decision should not be a simple majority vote. Superminority is to ensure there is a range of policy proposals that better match the ideological diversity of the political community, but a simple up/down vote could still be the decision rule (and the jury would have to listen to all the arguments first).
LikeLike
Not to belabor things… just a few brief responses…
I admit that since the role of parties and elections in your plan is so different it is hard to have any confidence in how they would impact jurors, etc. But, if they aren’t gaining real power (only gaining permission to make proposals to juries), it also seems likely that voter participation will plummet. Then their vote for parties REALLY doesn’t matter (and many or most already think it doesn’t), and you won’t have any meaningful societal proportional (ideological or other) representation. Only random selection can generate proportionality (and not from “volunteers”).
You give the example of Brexit and Florida re-enfranchisement… but those are somewhat “gut” issues, perceived by voters as “simple” (though they of course weren’t), so with rational ignorance as their guide they voted for what “felt” right. If a random group of people had DELIBERATED on these issues (as Keith suggested), they could have made INFORMED decisions instead. But a key point is to think what would happen when competing proposals on the appropriate way to spend money on infrastructure upgrades, or some other “dry” policy domain without any “gut” reactions is presented to a jury. If elections are a big thing (and I am wrong that few people will bother to vote), then they WILL defer to the party policy from the party they voted for, rather than do the heavy lifting of making up their own minds, and the sortition benefit is lost.
Finally on the multiple choice voting problem… You state that “Pathologies exist everywhere in every system.” But ONLY when trying to aggregate through VOTING among multiple choices. If instead there is active deliberation by a random panel (not of “volunteers”) they can take a piece from this proposal, a piece from that and often are able to hone a final product that a large majority of members think is pretty good. If a final yes/no jury is then presented with a yes or no choice all of the voting pathologies evaporate and do not apply. This idea of giving a jury several competing options is the core of your plan, but also its Achilles heel, because aggregation (voting) is only appropriate for two choices, not multiple. When you have multiple choices deliberation is the appropriate tool instead.
LikeLike
Terry:> if [parties] aren’t gaining real power (only gaining permission to make proposals to juries), it also seems likely that voter participation will plummet.
We would anticipate the opposite. It seems likely that the number of political parties would approximate to the Superminority decision threshold. This would mean that voters would have a choice between (say) five parties, each of whom would have an opportunity to put forward policies that approximated to their beliefs and preferences, and they would all have equal chance of becoming law. At the moment most votes are wasted, so I can’t see why Superminority should not appeal to both voters and politicians.
You are right to put volunteers in scare quotes, but all three of us agree that descriptive representation depends on quasi-mandatory participation.
On the topic of multiple choice, bear in mind Fishkin’s Zegaou DP, which seems to have worked pretty well, from the perspective of both citizens and party leaders. (In this poll citizens were asked to rank infrastructure projects in order of preference.) It’s a bit like Yoram’s example of a group of friends deciding what restaurant to eat at — at the end of the day this is decided by a vote, which may or may not have been preceded by deliberation. What the friends don’t get to do is to pick and mix menu items from all the restaurants.
LikeLike
Terry,
>> This idea of giving a jury several competing options is the core of your plan, but also its Achilles heel, because aggregation (voting) is only appropriate for two choices
Multiple options are a necessity for any human voting system. Binary voting IS the pathology. It splits any group into two angry, hateful camps no matter how the groups are chosen. Human pathology is what matters, and reform efforts should primarily focus on the effects on ourselves.
Every decision making process has pathologies. The pathology of up-or-down, black-or-white, good-or-evil voting is that it polarizes people. That’s a deadly pathology. The fact that it doesn’t show up in a game theory textbook doesn’t mean that it can be ignored. Polarization and negative partisanship are the principle pathologies of contemporary democracy, and they occur in ANY system where only two choices are available.
LikeLike