I’ve spent the last few days at an international workshop on ‘Representation in Historical and Transcultural Perspectives’ organised by the Centre for Political Thought at Exeter University. On the final day it was suggested that representation was an ‘ideological and legal fiction’ and none of the participants disagreed. Yves Sintomer gave the final presentation and suggested that his fieldwork comparing Chinese and French systems of representation showed little difference between the two and that the lack of effective representation was an existential crisis for democracy. I had a question for Yves, but we ran out of time, so will ask it here (and draw it to his attention):
One system of representation that would clearly be non-fictional is delegates with legally-binding instructions but this was rejected at the time of the American and French revolutions. Trustee representation (with free mandate) may have worked for a time but didn’t long survive the extension of the franchise and is now rejected by the populist uprising in Europe and America. Virtual representation in Burke’s sense was always fictional due to the dissimilarity between voters and the political class. This would suggest that the only form of non-fictional democratic representation would be when final decision power is vested in a statistically-representative minipublic. Concerns might be raised both about the accuracy of the descriptive representation and the epistemic consequences, but such a system would be non-fictional so long as the microcosm retained ongoing descriptive representation vis-a-vis the target population. This would require large juries, quasi-mandatory participation, short-term service, balanced (exogenous) information and advocacy, and silent deliberation and secret voting, but the representation would not be fictional if it could be demonstrated that multiple samples of the same population generated closely-matching decision outputs. Might such a system be the only way of establishing genuine political representation?
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
johnedwardtaylor
johnhuang865
Jonathan Crock
keithsutherland
Jacob Richter
lukewhito
SimonPek
Nathan El
Nicholas Gruen
Owen Shaffer
Patrick Chalmers
Paul Gölz
paulnollen
peterstone
Pierre-Etienne Vandamme
pronfeld
Emil
rcaze
richwbrown
Roger Knights
roslynfuller
Pierre-Etienne Vandamme
shaharavin
tbouricius
Ted Wachtel
Joshua Laferriere
theodorestathis
Wayne Liebman
Yoram Gat
Oliver Milne
Equality by lot 
The only representational system is a “town meeting” of the entire population. Any “jury” can only be a statistical approximation of “representative”. The issue is how much unrepresentative is tolerable. Any statistical approximation is a fiction.
LikeLike
Mostly yes Keith, with the exception of „large juries and mandatory participation“. Frequentist statistic representation is an unsuitable approach to collective decision making. There the difficulty lies not in diagnostics where frequentists estimate what is. It is all about predicting the future, i.e. a set of beliefs, on which a collective decision hinges. In this field Bayesian statistics is preferred: no big samples are needed as every belief updates the prior one.
LikeLike
To test a Bayesian model one needs a testing protocol. Otherwise what is predicted is indeterminate.
LikeLiked by 1 person
Jon, Not sure how this relates to my comment?
LikeLike
Such a system is the only way of establishing genuine political representation with the addition that 18 should not be the cutoff age and provision should be made for representation of very young children and future generations. Otherwise adults will go on abusing and exploiting current children and future generations on a scale that with climate catastrophe will be unprecedented.
LikeLike
Dear Keith Sutherland,
If you really have something to say, say it briefly in no more than 100 words and say it clearly. Everything can be said in a few words, even the general relativity of Einstein.
LikeLike
Jon:> The only representational system is a “town meeting” of the entire population. Any “jury” can only be a statistical approximation of “representative”. The issue is how much unrepresentative is tolerable. Any statistical approximation is a fiction.
John Adams’s “portrait, in miniature” was only an image, not the real thing. A town meeting is a form of direct democracy (that privileges activist citizens), not a system of representation. Given that the role of a decision body is to take (binary) decisions then the size of the sample can be adjusted to take account of the margin of error, decision threshold etc. Statistical representation is imperfect, but it’s certainly not fictional.
Hubertus>: Frequentist statistic representation is an unsuitable approach to collective decision making . . . It is all about predicting the future
I acknowledged that there might be epistemic objections to my proposal, but the issue at hand is to find a system of representation that is non-fictional (even though it may not necessarily lead to benign outcomes).
Ad:> If you really have something to say, say it briefly in no more than 100 words and say it clearly.
The complexity of the issue is why I didn’t have time to ask Yves the question at the seminar, but are you saying that 192 words is too long for a post on an internet forum? Not sure how you would expect me to abridge it as each sentence deals with a different point (e.g. delegate, trustee and virtual representation are the topics of the first three sentences, and each would require a separate paragraph to explain). It would be a shame if our forum had to accept Twitter-style constraints.
LikeLiked by 1 person
PS, to put the “fictional” claim in context we should bear in mind that in 1997 Bernard Manin published a much-cited book claiming that, despite it’s metamorphoses, representative government was in rude health. Although Manin views this as a “mixed” form of government, Nadia Urbinati went on to claim in 2006 that it was a well-functioning form of democracy. But Yves claimed at the workshop that this was only true for a few decades and now the consensus among political theorists and legal scholars appears to be that it’s just a fiction. This is beginning to sound like a paradigm shift, although there was little, if any, suggestion as to how the problem might be rectified — hence the question that I pose in this thread.
LikeLiked by 1 person
Keith>> I acknowledged that there might be epistemic objections to my proposal
Mine was not an objection. It was intended as an encouragement and simplification because large sample sizes and mandatory participation are not necessary conditions for Bayesians.
From a Bayesian perspective, the path to a correct decision in not about the (entirely illusory) frequentist representativeness by large samples, it is all about knowledge and as many different perspectives flowing into the decision process. The step-wise update of beliefs – a form of consensus – regarding the best path forward. (Which we can never know for certain, in advance.)
LikeLiked by 1 person
Hubertus,
Words like “correct decision” are relevant from an epistemic point of view, but entirely orthogonal to democratic representation. There is a high probability that the informed decision of a large minipublic would reflect the general zeitgeist, warts and all. Whilst that may be terrible from a deliberative and/or epistemic perspective, it still qualifies as a non-fictional form of representation. Democracy is all about the equal right to take the wrong decision and Urbinati is instructive regarding the epistemic disfigurement of democratic theory.
LikeLike
Keith,
We had this discussion, Keith. Let’s agree to disagree.
I say, we must look for the correct decision, it’s out there. The process is as good as the decision which it finds.
Others may say, what’s right is defined by the outcome of today’s general voting process where all people participate mandatorily.
I find the latter position tautological and not useful at all.
It reminds me of physicians in the dark ages. They did their procedure, proper bleeding of the patient and all. If the patient died, it was the patient’s fault. After all, the physician hath dutifully employed the prescribed process.
LikeLike
Hubertus,
What is at stake is the meaning of words (democracy, representation etc), not the correctness of decisions or whether the process is good or bad. A benign dictator might well come up with wise policies that are better for the public (i.e. in their interest) than the ones they chose themselves, but the process would not be described as democratic. And an accurate representation of an ugly person would be equally ugly. You are right to view this as tautological as that’s what’s involved in the analysis of the meaning of words.
LikeLike
Disagree. Not even “might”. A centralist power (dictator or any few) will never be wiser than the (deliberating) many or come up with better policies for the public.
See Ludwig Mises’s prediction about central planners fallacy, and Hayek’s brilliant expansion of this principle, due to insufficient use of knowledge in society.
Counterargument: You say “democracy” (a very ambiguous word, once you get to the nitty-gritty) is not about the correctness of decisions.
If that were true then a society with a dictator which organises better decisions (for more economic and military power) would prevail over democracies which is clearly not the case if we look in the real world. Democracies are currently winning BECAUSE they produce better decisions than the other systems.
LikeLike
Hubertus,
“From a Bayesian perspective, the path to a correct decision in not about the (entirely illusory) frequentist representativeness by large samples, it is all about knowledge and as many different perspectives flowing into the decision process. The step-wise update of beliefs – a form of consensus – regarding the best path forward. (Which we can never know for certain, in advance.)”
This is the essence of generating ideas, proposing legislation and advancing candidates for office. But you still need a “blink break” to keep this proposing function honest, which can be accomplished by a large, randomly chosen group of citizens. The verdict of this panel provides the feedback to update the (Bayesian) proposing service.
Bayesian and frequentist approaches are not necessarily in opposition. They can work together!
LikeLike
Hubertus,
I’m referring to logical possibilities and the meaning of words, whereas you are referring to the evaluation of decision outcomes. Unfortunately democracy just means power (kratos) in the hands of the people (demos). Whether this generates better outcomes than other constitutions is beside the point. By all means start a thread on that topic, but it’s of no relevance to this one.
LikeLike
Let’s see what’s being proposed here, so harmoniously:
Let’s take a small Bayesian group, say 50 to 200. Entirely sufficient.
Then, let’s take a “large group” as frequentist believe is needed.
If the two “work together” you get a “large group”.
Nice try. But no, thanks. :-)
LikeLike
Hubertus,
Not what I had in mind. If you separate proposing from disposing, then the proposing service operates in a Bayesian manner, and the disposing in a frequentist manner. The disposing function provides the “observations” that are used to update the proposing service.
LikeLike
My 2C: Proposing (i.e. creative) can be done even by a single person, if it’s a genius. no statitics needed there. Disposing is more tricky. A Bayesian sample design will suffice in well above 90% of cases, but for the remaining 10% e.g. constitutional matters – we could skip those large samples which frequentists need – just verify the Bayesian sample’s decision with a full referendum, Swiss style. Do it properly rather than mucking around.
LikeLike
Hubertus:> Disposing is more tricky. A Bayesian sample design will suffice in well above 90% of cases . . .
Given that your Bayesian sample design translates into decision-making by small demarchic committee, do you seriously believe this would be viewed as democratically legitimate by the vast mass of citizens excluded from the decision process? We should all bear in mind that deliberative and epistemic “democrats” like Helene Landemore are unconcerned with the issue of perceived legitimacy and their theoretical models simply assume that citizens will vote according to the general good, as opposed to their own interests. And demarchists like John Burnheim believe that democratic legitimacy will be adequately provided by publicity — i.e. making the committee deliberations available for everyone to read on the internet.
I put it to Helene that she was taking Rousseauian civic obligation into Buzz Lightyear** territory, but I think it’s more a case of cloud cuckoo land.
** To Infinity and Beyond!
LikeLike
Keith: “do you seriously believe this would be viewed as democratically legitimate”
Good question, Keith.
Personally, I have come to the opinion, yes. Two reasons:
Psychologically, the critical number seems to be the number sitting in parliament. This number has (kind of) worked for the people. So for Austria, the GILT party proposes a number of 183.
Second: Let’s look at the nuclear deposit case in South Australia Iain Walker here can surely tell us more, but I understand that a very powerful political leaders could not dismiss a citizen jury of 300 by going against their decision. Even spending serious money to promote a referendum, he could not convince the people. So, if we can agree that a referendum (a best practice one) is as democratically legitimate as it can get, then it follows logically that the SA Jury was viewed as democratically legitimate.
LikeLike
Hubertus,
Both Alex and myself endorse the view that a final decision arrogated to a large randomly-selected citizen jury will be viewed as democratically legitimate by the vast majority of excluded citizens iff it can be demonstrated that the decision would be invariant between different samples of the same population. Why would anyone object if it makes no difference which empirical individuals (including themselves) were included in the sample? Note that any factor that diminishes the statistical representativity of the sample (including the exclusion of poor predictors) will invalidate this condition — it may well be of epistemic value,** but will not meet the conditions for representative democracy (the topic of this thread).
** although it’s worth noting that Surowiecki specifically notes (c.f. Tetlock) that the inclusion of stupid people in the group does not diminish it’s collective “wisdom”.
LikeLike
Keith>> Why would anyone object if it makes no difference which empirical individuals (including themselves) were included in the sample?
Exactly, so why should anyone object, if we can show that the correlation of outcomes is just as good (I suspect even slightly better) for small Bayesian samples as with large frequentist samples.
Actually, GILT is planning this in case of getting funding (by the “party” being elected) as the first of three ways to verify if citizen parliament decisions are correct: “replication”.
So if and until then, who can first run parallel juries at different sizes? Any volunteers?
LikeLike
Hubertus:> Exactly, so why should anyone object, if we can show that the correlation of outcomes is just as good (I suspect even slightly better) for small Bayesian samples as with large frequentist samples.
I’m not a statistician (unlike Alex) but the only other statistician I know working in this field (John Garry at Queens Uni Belfast) posits 1,000 as the absolute minimum for an adequate level of descriptive accuracy. And I imagine the Athenians had good reason to require legislative juries in the 501-5,001 region, especially as they had to pay them.
PS The issue of correlation is with the informed preferences of average citizens, whereas I seem to recall that your demarchic model excluded those with poor prediction ability.
LikeLike
Keith>> your demarchic model excluded those with poor prediction ability.
Yes, this is a working assumption. As you read Tetlock (actually Berg found this much earlier) stupid people cancel out, just like white noise. So logically, if stupids don’t add to an improved decision, we might just as well not waste time and money on them.
However, being Austrian school Popperians, GILT will of course run experiments to falsify that theory. For this, we will vary the chances of stupids from 100% to 0% in the second round of replications.
LikeLike
Hubertus:> So logically, if stupids don’t add to an improved decision, we might just as well not waste time and money on them.
Although it might be viewed by some as a waste of time and money, democratic equality is the spirit of our age — what you are suggesting would have put you up against the wall in fourth century Athens and wouldn’t have made you very popular in postwar Germany. (And experiments on the exclusion of stupid people also have an unfortunate provenance). In addition, the notion of an “improved” decision makes no sense from the perspective of democratic theory and has no relevance at all to representation (the topic of this thread).
LikeLike
Feeling duly threatened. Keep in mind, shooting, burning and killing people may be a kind of argument but not really convincing. Also, I strongly object to the disturbing allusion in the brackets and must sadly resign from a debate carried out in such a style. Goodbye.
LikeLike
I’m sorry if you feel threatened Hubertus, but making blatantly undemocratic proposals (disenfranchising the stupid) on a forum dedicated to democratic equality was never going to go down very well.
LikeLike
“As you read Tetlock (actually Berg found this much earlier) stupid people cancel out, just like white noise.”
This captures the essence of the matter. Including everyone involves statistical noise. Deliberate disenfranchisement–well intended or not–is biased toward whatever outcome the guardian class wants.
In addition, a “Bayesian” process assumes the observer is unbiased. Simply being a part of a powerful committee changes a person. Even if selected at random, a committee of citizens will diverge from the population simply by the fact of their selection. Choose a new committee to dispose of every political proposal, and you avoid this problem, at least to the extent it can be avoided.
LikeLiked by 1 person
Alex:> Simply being a part of a powerful committee changes a person. Even if selected at random, a committee of citizens will diverge from the population simply by the fact of their selection.
That’s very true, but for some reason demarchists and deliberative democrats (including the majority of sortition advocates) fail to appreciate this important (and rather obvious) point. I’m puzzled as to why this should be the case — one possibility being the view that beliefs are the product of interests that are conditioned by entrenched socio-economic factors. So, if you are one of the 99% then class loyalty will determine an outcome that favours the interests of your peers. I appreciate this focus on the economy is somewhat anachronistic since the rise of the New Left — in the words of Cody Hipskind on a recent thread on this blog:
What backs up my hunch is the widely-accepted view that ‘deliberative democracy, when properly conceived, is the rightful heir of the early Frankfurt School [of cultural Marxism]’ (Scheuerman, 2006, pp. 86). This being the case, one would expect former Marxists like Burnheim and critical theorists like Dryzek to ignore or downplay the social psychology of committee participation.
Ref
===
Scheuerman, W. E. (2006). Critical Theory Beyond Habermas. In B. Honig, J. S. Dryzek & A. Phillips (Eds.), The Oxford Handbook of Political Theory (pp. 84-105). Oxford: Oxford University Press.
LikeLike
There is no such thing as “class loyalty”. It is redefined from one moment to the next. It is only an aggregated behavior observed by others.
LikeLike
Jon:> There is no such thing as “class loyalty”.
Yes that’s right, it’s an archaic Marxist concept that is long past its sell-by date.
LikeLike
quote: “”This would require large juries, quasi-mandatory participation, short-term service, balanced (exogenous) information and advocacy, and silent deliberation and secret voting, ..Unquote. I think that our proposals are in accordance with this criteria https://independent.academia.edu/PNollen/Sortition-for-a-real-citizens’-representation (and presented also on this blog https://equalitybylot.com/2019/03/15/two-proposals-for-representative-representation/ )
LikeLike
Paul,
Ideal. These criteria are so much more important than the somewhat technical concerns addressed on the parallel thread. I think we both agree that the specification of an impartial selection mechanism is only the first step in the creation of a democratic decision-making body and that ongoing representativity is the principal concern. If I’m not mistaken, Yoram would disagree with all of these criteria and this presents a serious difficulty for those of us seeking to implement sortition.
LikeLike