[This serves, in a sense, as my response to the discussion on the French climate assembly post.]
Now that minipublics are no longer limited to local level “experiments” but are regularly involved in consequential political occasions, constitutional amendments (Ireland), long term city planning (Australia, Germany, US), responses to major political crises (France, Iceland, Ireland), institutionalized checks within representative government (East Belgium, Oregon)–to name a few—the question of their “representativeness,” and, more fundamentally, their legitimate democratic role is no longer academic. Given the response rate problem, those who accept invitations to a citizens’ assembly or jury (however scientifically sampled) are different in some respects from those who do not, and the number of participants in any such minipublic will, regardless of sampling, be exceedingly small compared to the population. Sortinistas and participatory democrats have raised the question of how a not entirely representative, unelected minority could legitimately affect political outcomes for the overwhelming majority who do not take part in the minipublic. In contrast to the “allotted citizen,” with the implication of egalitarian empowerment, some would disparagingly label participants in a minipublic chosen by lot an “aleatoric elite”–ignoring the standard implication of non-ephemerality in the term “elite.” But this focus on strict representativity misses the strongest reason for using minipublics chosen by lot in the first place, and it distracts us form their most promising participatory democratic uses.
After summarizing the strongest arguments articulated by both sortinistas and participatory democrats for the strengths and political potential of minipublics, I suggest another dimension on which they can function. Allotted minipublics can serve as unique spaces of political action and contestation, different from the space of electoral struggle, the space of confrontation in protest, or “enclave” spaces within activist groups and political parties. An electoral campaign is mostly fighting, a protest mostly “manifesting” strength or conviction, a party/union/organization meeting mostly strategizing or venting; but a minipublic provides a rare opportunity for the “everyman,” in a time of cognitive and political “bubbles,” to confront or act with a plurality of points of view, no one of which she/he can anticipate.
First, let me begin with the strongest rationales for the use of lot mentioned by sortition theorists. Oliver Dowlen maintains that neither in Athens, nor in Venice and Florence, nor even in Harrington’s fictional Oceana, was lot used for “representation.” In Athens it was a way to temper the potential domination or the revenge taking by “the many” against the aristocracy, and to avoid factions in general. In Venice and Florence, lot was used to avoid domination by any one powerful family or faction. Harrington suggested that lot be used for allocating soldiers to units in order to avoid dangerous military factions. Moreover, Peter Stone suggests there are only two legitimate uses of lottery, and by extension allotted minipublics: 1) when there is no other legitimate selection criterion, and 2) in order to avoid corruption or undue influence. If the main dissatisfaction with contemporary representative government is domination by special interests and the economically (or socially) powerful—which seems to be the case—then minipublics are a democratic tool against domination by elites and systematic bias due to the overrepresentation of special interests. This conforms with much of the discussion of the participants in the German citizens’ assembly on democracy. “Lobbyism” was a major concern and “citizen councils” were seen as a potential “counter-lobby for the citizen.”
Second, resonant with the “citizens’ lobby” sentiment, the focus on representativeness turns our attention away from the unquestionably democratic uses of minipublics that participatory democrats see as their democratizing political potential. Political theorist and fifty-year advocate of a “participatory society,” Carol Pateman suggests that minipublics can only be democracy prompting in view of the society they are in. A truly democratic society requires reform of social and political structures. Democracy must be practices in “everyday life,” such as the workplace. People other than those lucky enough to be chosen for a minipublic need to have the opportunity and the ability to participate in the decisions that affect their lives. Christina Lafont, also once a critic of minipublics came to accept them under the condition that they are not used as “short cuts” to reaching better decisions by majority rule through improving public opinion. The issue, for Lafont, remains democratic legitimacy. If democracy ultimately relies on self-rule, then even a perfectly representative, informed, an organized minipublic still must persuade the general public that its recommendation/decision is the best choice.
The “deliberative systems” approach argues that minipublics are but one kind of a “variety” of discursive forums, but does not tell us what role a minipublic in particular ought to play within the “system.” Lafont sees the potential in a minipublic in its ability to claim its result is a “considered opinion” of the public It can do this so long as selection by lot assures inclusive diversity of perspectives and non-capture by elites/special interests, while adequate facilitation and unbiased information assure that the process is indeed considered/deliberative. I would add a fourth condition. Transparency allows independent scrutiny and ensures legitimacy in the eyes of the public.
In fact, we can understand this claiming that a minipublic represents the “considered opinion” of the wider public as representation in the sense of Michael Saward. Representation here means neither “descriptively” nor “authoritatively” representing the public; but it is a claim by those who would use the minipublic’s result as representative that the person making the claim still had to defend politically, publicly.
Given that, Lafont’s classification of the democratically promising political uses of minipublics is helpful both in resolving some of the theoretical issues and in interpreting recent events like the citizens’ assemblies. Case 1) the recommendation of the minipublic is contrary to current public opinion surveys. This case, she calls, contestatory. Although it would be illegitimate to force the result on the public, the result can be used as a signal to the public to further consider the issue, and as a rhetorical tool and encouragement for the minority opinion to continue to fight to change public opinion. Case 2) when the minipublic agrees with prevailing opinion, the vigilant use, the results can be sued as a further argument by the public against the government to change current policy. Such was the case, contrary to Roslyn Fuller’s view, in Ireland on abortion. It wasn’t that the government “manipulated” the tool of the citizens’ assembly to do what they wanted to do anyway; rather, they used the result politically to prevent an illegitimate minority opinion from blocking action. Such is also the case with the German citizens’ assembly on democracy in Leipzig. Public opinion is already in favor of more participation and more direct democracy. Those in favor of that can now argue that the Assembly shows that the prevailing opinion is also a considered one. The process was open and transparent and one can see that a variety of positions was presented to the minipublic in an even handed way, and that facilitation did not allow any one view to dominate discussion.
Case 3) there is no current public opinion on a matter, the anticipatory use of a minipublic. Here a minipublic can be an impetus for the public to start thinking about an issue that they had not until then seen as impacting their lives. A good example might be a new technology, otherwise obscure or known only in special circles, such as cryptocurrency. The findings and reasoning of a minipublic could begin a necessary public conversation to prevent a technocratic elite from taking action without public oversight. Lastly, case 4) empowered minipublic, which unfortunately dominate the discussion on Equality by Lot. An empowered minipublic is an institutionalized body that makes a decision in conjunction with other government institutions. Such a body can check other agencies or co-create policy. Here it derives its legitimacy from its institutionalization and that it can be contested politically just like any other government decision. Again, in none of these uses is strict representativity crucial, although in the case of empowered minipublics representativity becomes more important.
Furthermore, we can infer from Lafont’s suggested uses that minipublics could play a salutary role on public opinion because better information and better reasoning will be seen as something that could affect the results of a minipublic. The minipublic, or the organization running it, would also be incentivized to be transparent and independent, in order for its results to be taken seriously by the public, to maintain its status as “independent evidence.”
Finally, let me turn to another political dimension of minipublics. In their everyday lives, people rarely interact politically with others they may disagree with on important public issues. As one sociologist argues, “political” is almost an offensive or tabu in everyday life. Electoral campaigns are mostly spectator sports, besides perhaps for the small number of very self-selected political volunteers. Demonstrations are not venues of give and take but pure “shows” of force. Union or activist meetings are again for a small number and revolve around strategizing not genuine discussion and confrontation. Town Hall meetings are not meetings between equals but a place either to gauge or demonstrate raw sentiment. Minipublic are an almost unique space where without foreknowledge of anyone’s political positioning or background, ordinary citizens must confront, deal with, and act with difference politically.
The same characteristics that allow minipublics to claim, if done right, to embody “considered public opinion” also make them unique venues for encountering difference. Lot assures diversity, inclusivity, and that they are not captured by elites; facilitation (done right) means that no voices or perspectives dominate. Moreover, even when, maybe especially when, minipublics such as citizens’ assemblies do not make final decisions but only generate ideas, reasoning, and concerns, they encourage the healthy discussion of a variety of views. That they are outside the distorted spaces of party and parliamentary politics, make them not only more “considered” and “deliberative,” but also more genuine and more plural. Absent the filters of party coherence and the electoral game, minipublics are literally “refreshing,” as participants often comment. This aspect should not be underestimated and should be further investigated in itself.
[Edited, 10/10/2019.]
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
Nicholas Gruen
Owen Shaffer
Patrick Chalmers
Paul Gölz
paulnollen
peterstone
Pierre-Etienne Vandamme
pronfeld
qcollective
rcaze
richwbrown
Roger Knights
roslynfuller
Pierre-Etienne Vandamme
shaharavin
tbouricius
Ted Wachtel
Joshua Laferriere
Wayne Liebman
Yoram Gat
Oliver Milne
Equality by lot 
When I read that the highly manipulated and uncontrollable ‘allotted’ mini-publics are ‘legitimate’ in any way I find it highly disturbing. They are even less ‘controllable’ then political parties. If you take the trouble to evaluate a mini-public with the criteria mentioned in our proposals (or the evaluation grid) or try to evaluate the project with the code of good conduct explained here https://independent.academia.edu/PNollen/Sortition-for-a-real-citizens’-representation (and discussed at this blog) , you will see that most mini-publics are far from any legitimation. Even more, most essential information is simply missing.
LikeLike
Ahmed,
I agree with Dowlen and Stone that sortition has not been used in the past as a form of representation — this is hardly surprising given that representative democracy is only a couple of centuries old and sortition has not been used seriously during this period. But, for better or worse, we are now in an age of representative democracy, so sortition projects need to be evaluated by representative criteria — that’s why organisations like The Sortition Foundation and newDemocracy (wrongly) talk of “representative samples”. Thank you for acknowledging that this is not your concern, let’s hope this honesty is contagious.
>Lafont sees the potential in a minipublic in its ability to claim its result is a “considered opinion” of the public It can do this so long as selection by lot assures inclusive diversity of perspectives and non-capture by elites/special interests, while adequate facilitation and unbiased information assure that the process is indeed considered/deliberative. I would add a fourth condition. Transparency allows independent scrutiny and ensures legitimacy in the eyes of the public.
Who is to decide which perspectives to include and how they should be weighted? Elites come in many forms, including activists and counter-cultural warriors (Dryzek agrees that deliberative democracy is the product of the Frankfurt School of cultural Marxism). Facilitation is open to the quis custodiet objection and the notion of unbiased information is simply laughable. And who are the independent scrutineers to be?
Paul:> When I read that the highly manipulated and uncontrollable ‘allotted’ mini-publics are ‘legitimate’ in any way I find it highly disturbing.
Ditto
LikeLike
Good one. “Representativeness” is a rather elusive concept in real life projects, so it’s a good idea to start a talk about deemphasising it.
We could also just tune this requirement down, to make sense from an Bayesian viewpoint. A minipublic should give us diversity of knowledge, by including a “reasonable sociodemographic mix” as a fully sufficient alternative to “representativeness”. As long as a piece of private information are likely present in a crowd, it will sufficiently influence collective decision making by its weight of logic, independent fron its weight by frequency.
On the four cases, however, I take sides for this forum’s emphasis on #4 empowered minipublics for a simple reason: without a subsequent action and falsifiable expected consequences, such minipublics would be of a metaphysical nature.
LikeLike
Hubertus:> As long as a piece of private information are likely present in a crowd, it will sufficiently influence collective decision making by its weight of logic, independent from its weight by frequency.
Yes that is the argument of Burke, Habermas and most deliberative, discursive and epistemic “democrats”. Unfortunately it fails in the case of empowered minipublics on conventional majoritarian criteria. As for “falsifiability”, one just has to hope that citizens have a long memory.
LikeLike
Ahmed,
> in none of these uses is strict representativity crucial
I am not sure what is the reasoning behind this assertion. To me it seems that if an allotted body is to have any legitimacy, regardless of its position, it must be representative. If it is not, then how is it better than a think tank, which could be entirely in the service of interested parties?
LikeLiked by 1 person
Let me address the comments so far together, because they fall into two categories: the desirability of “representation” or “representativeness,” the manipulability of a minipublic.
My position, which is a paraphrase and modifaction of Lafont’s, is that IF minipublics execute the three conditions she mentions (sortition, independent adequate information, and adequate facilitation) and my extra condition of transparency, THEN they can CLAIM to represent a “considered public opinion.” The claim MUST still be justified and argued politically by the person making the claim. I do not ASSUME that any particular type of minipublic is ideal. Every minipublic must STILL JUSTIFY ITSELF as fair in inputs, as fair in process, and as sufficiently inclusive. One of the ways it can do that is to be transparent so that it can be scrutinized, contested, or defended.
This dovetails into my answer to the question on representation. This CLAIMING to reflect the hypothetical “considered public opinion’ in itself amounts to the sort of representation that Michael Saward argues is more coherent than either the “mirror” (descriptive) or the “statesman” (authoritative) form of representation.
Lastly, I claim that minpublics (again, done right, fair, inclusive, etc.) CAN be unique places for humans to be “political animals” in a way the overwhelming majority is deprived of. If this is true, then we should encourage MORE minipublics even when they are “only” making recommendations.
LikeLike
Ahmed,
Presumably Lafont’s justification of sortition is that it is less immune to domination by the rich ‘n powerful; the counterclaim being that voluntary participation makes the minipublic subject to an even stronger form of domination (by self-selected activists, loudmouths and cultural elites). As for “impartiality” this is not true ex-post, where minipublics are more open to corruption as they are not protected by party discipline and/or constituent accountability.
>hypothetical “considered public opinion’ in itself amounts to the sort of representation that Michael Saward argues is more coherent.
If the minipublic is not representative the opinion may well be considered but is not (by definition) that of the public. I’ve read Mike’s book several times and have engaged with him at several seminars and I don’t recall The Representative Claim being a ringing endorsement of sortition.
>CAN be unique places for humans to be “political animals” in a way the overwhelming majority is deprived of.
Sure, but the overwhelming majority are still deprived, that’s why Carole Pateman told me (personal communication) that there is an unbridgeable gulf between participatory and deliberative democracy. The only way of overcoming deprivation and alienation in large modern states is through some sort of representation (which, according to the title of this thread, you abjure).
LikeLike
P.S. Saward’s “representative claim” is the critical focus of the isegoria chapter of my PhD thesis.
LikeLike
Ahmed,
> Every minipublic must STILL JUSTIFY ITSELF as fair in inputs, as fair in process, and as sufficiently inclusive. One of the ways it can do that is to be transparent so that it can be scrutinized, contested, or defended.
It is true that every decision making body must justify itself in various ways – and having fair inputs, fair process, being sufficiently inclusive and being transparent are all important components necessary for such a justification. However, even if all these conditions are met, a non-statistically representative body could easily produce self-serving decisions that do not represent the public interest. Therefore, statistical representativity is a necessary condition that cannot be disposed of by meeting different conditions.
> This CLAIMING to reflect the hypothetical “considered public opinion’ in itself amounts to the sort of representation that Michael Saward argues is more coherent than either the “mirror” (descriptive) or the “statesman” (authoritative) form of representation.
Simply making the claim means very little. Power can claim anything – electoral bodies claim to be representative as well.
As for the claim made by allotted bodies: these bodies’ claim to legitimacy or to being democratically representative is via their claim to be representative of the beliefs and interests of the public. In this way self-representation of their own interests can be extended to representing the public interest.
LikeLike
All these comments bring me right back to the points in my 2013 paper on multi-body sortition….. There are different desirable traits of mini-publics that are mutually exclusive and cannot reside in a single body. If a body is small enough to have deep and meaningful deliberation to work out policies, this inevitably means that the workload is such that many, or most citizens would decline to do that much relatively thankless work. But a small random body will not have sufficient statistical accuracy to reliably represent the interests and attitudes of an informed population. Thus, we need at a minimum at least two very different kinds of mini-publics. Small intense active deliberating ones that serve for perhaps a year and take expert testimony and craft final proposals, and separate large mini-publics that serve only briefly, refrain from advocacy and “debate,” but instead hear pro and con presentations on final bills and vote by secret ballot. As Ahmed said, the small deliberative mini-publics are suitable for figuring things out because they can be diverse, impartial and relatively free from elite domination if well run, and can thus be good for refining recommendations… but they are not adequately representative to stand in for the general population to make final decisions. (In my paper I propose several other sorts of bodies as well, such as meta-legislative oversight bodies, agenda bodies, etc.)
LikeLiked by 1 person
Terry:> small deliberative mini-publics are suitable for figuring things out because they can be diverse, impartial and relatively free from elite domination
1. Self-selection (imo) makes for a singularly undiverse group in that stratification only make an unrepresentative group of activists, policy wonks and self-opinionated loudmouths look “like [say] America” on gender, age, ethnicity and other relatively insignificant criteria.
2. The blind break only works ex ante and will do nothing to ensure ongoing impartiality.
3. There are many forms of elites, and self-selection will privilege cultural elites, for all the reasons Andre has outlined.
Just take a look at Yoram’s new post to see how this approach is discrediting the entire sortition movement.
LikeLike
Keith, your interpretation of Pateman’s position is odd. Here, from her APSA address in 2011:
“One way of looking at the new expansion of participation is that in poor coun tries it can help improve governance, and in rich countries it can help bolster the legitimacy of the present system. If citizens can participate in mini-publics or can decide on the disbursement of some public funds then, in another piece of ubiquitous jargon, they “own” those decisions. More optimistically, it is possible now that citizens are obtaining some practice in new ways of exercising their citizenship” Pateman, Carole. 2012. “Participatory Democracy Revisited.” Perspectives on Politics 10(1): 14
Your reference to “party discipline” is also quite add for a sortinista–in general it is often hard to see whether you are at all one. “Party discipline” is a major reason many people call for the
Lastly, as for your (or Andres) use of the word “elite,” that goes beyond the pale of a reasonable interpretation of it. People who RESPOND are not activists or loudmouths; in fact we don’t know what “bias” the response rate bias is. It could simply be people who are not afraid of talking to strangers, or people who do not have weekend obligations such as coaching a team.
LikeLike
Yoram, I am not against “statistical representation,” just against making it THE touchstone for fairness in ALL participatory processes. What we do not want, is a systematically biases sample such as “those who show up to a town hall meeting on Monday at 7:30pm.”
That said, you as the statistician, what criteria will you judge representativeness by? Sex, age, gender, level of education, ethnic membership, economic class? How large would a sample have to be then for a country the size of the UK, US, India?
When you say: “Simply making the claim means very little. Power can claim anything – electoral bodies claim to be representative as well,” you are making my point for me. The overwhelming majority of people DISAGREE with the electoral body’s claim to represent; and they are BLATANTLY un-representative on sex, gender, ethnicity, and class.
Minipublics the size of 100 or 200 are already WAY better along these dimensions.
But at the end of the day, we cannot bypass politics based on a process, no matter how well thought out. If democracy is about self-rule, whatever minipublic you have still has to do some political work.
And, raising the general level of political ability and political awareness must be a goal of a democratic society, along with democratizing everyday institutions like the work place. Beyond this social and economic STRUCTURES must be democratized if you want to claim you have a democracy. Having extreme inequality and then putting on a minipublic once in a while will not magically make society democratic.
LikeLike
Terry, interestingly the CIRs in Oregon incorporate some of your suggestions. The advisory committee that oversees the work of the CIR is made up of 8 former citizen jurors and only 3 politicians.
If every minipublic has an advisory and an evaluation committee; one can say that a best practice is that those committees have some citizens chosen by lot, e.g., from the list of those who respond but exceed the size of the event.
LikeLike
Ahmed,
> What we do not want, is a systematically biases sample such as “those who show up to a town hall meeting on Monday at 7:30pm.”
How is “those who accept an invitation to serve under terms that 95% of the people reject” any less of systematic bias?
> That said, you as the statistician, what criteria will you judge representativeness by? Sex, age, gender, level of education, ethnic membership, economic class?
Representativity according to age, gender, education, etc., are all necessary but insufficient conditions for representativity of ideas and interests – which is the criterion that really matters for a decision making body. The only way to make sure a body really represents the public statistically in terms of ideas and interests is by having equi-probable sampling and very low rejection rate. (Of course, this is also a necessary but insufficient condition for having a democratic body.)
> How large would a sample have to be then for a country the size of the UK, US, India?
As is often pointed out, the size of the population is essentially immaterial. A good size of an allotted body is determined by different considerations that may change depending on the settings. However, an upper limit on the size is set by Dunbar’s Number – the maximum number (~200) of people who can engage in an all-to-all communication.
LikeLike
Ahmed:> Keith, your interpretation of Pateman’s position is odd
This was her response to me:
The implication being that participatory and deliberative democracy (“the current fashion”) are two different subjects.
>Your reference to “party discipline” is also quite odd for a sortinista–in general it is often hard to see whether you are one at all.
I’m not. I believe in democracy and that requires a combination of electoral and allotted institutions.
>in fact we don’t know what “bias” the response rate bias is
That’s true, but seeing as the typical refusenik rate is 95-96%, the precautionary principle would suggest that this may be a statistically significant population parameter (imagine the objections to an allotted body that was 96% men and 4% women [or vice versa]). Unfortunately no amount of stratification can overcome this bias, the only solution being a judicious combination of the stick and the carrot. Fishkin clearly believes it’s a significant population parameter, that’s why the DP methodology goes to such lengths to ensure that as many of those who are allotted actually participate.
>That said, you as the statistician, what criteria will you judge representativeness by? Sex, age, gender, level of education, ethnic membership, economic class?
None of these — just make sure that everyone who is selected participates.
>How large would a sample have to be then for a country the size of the UK, US, India?
The size of the country is irrelevant, but for an acceptable level of confidence the sample would need to be 600-1,000.
>[electoral bodies] are BLATANTLY un-representative
They may fail on the descriptive representation criterion but they pass muster because they were chosen by voters. Pitkin is emphatic that descriptive representation is only one variant of the general principle. Personally I don’t give a damn whether my lawyer, doctor etc is male or female, I just want them to act in my interests.
LikeLike
>That said, you as the statistician, what criteria will you judge representativeness by? Sex, age, gender, level of education, ethnic membership, economic class? How large would a sample have to be then for a country the size of the UK, US, India?
What I am missing so far is the notion of ‘diversity of life experience’. Imo this can only be reached with, at least, a statistical sufficient number of participants.
LikeLiked by 1 person
Paul:> What I am missing so far is the notion of ‘diversity of life experience’.
That’s very nicely put and it clearly cannot be mirrored by either small groups or stratification. The implication that beliefs and preferences are in some way determined by gender, ethnicity, age, education and income levels is patronising to say the least.
LikeLiked by 1 person
Agreed. What we do know about the effect of the response rate is that the better educated are usually over represented in minipublics even when stratified against that possibility.
LikeLike
> What I am missing so far is the notion of ‘diversity of life experience’. Imo this can only be reached with, at least, a statistical sufficient number of participants.
Without some approximate definition of this notion, it would be impossible to say how it can be attained. How would we know if 10, 100, 1000 or 1000000 is a sufficient number?
The numbers regularly thrown around (e.g., 1,000) are at best completely arbitrary.
LikeLike
While not open to formal definition, It can be tested in practice. If two minipublics judging on a specific political topic and presented with identical information/advocacy come to the same decision (within an agreed margin of error), then the diversity of the minipublic matches that of the population that it describes. The figure of 1,000 (which I think originates with Dahl) is advocated by a statistician working on sortition (John Garry of Queen’s University Belfast). Robert Luskin, a statistician from U. Texas, Austin, is prepared to deal with smaller samples (c. 300), but Deliberative Poll sizes are constrained by practical considerations (money) and have no binding mandate.
But I think we all agree that volunteer bodies under 50 are not representative in a way that would be meaningful to any statistician working in the polling industry and sampling from a diverse multicultural society.
LikeLike
The core of the problem being discussed is that there is no ideal all-purpose mini-public, size, deliberation style, acceptance rate, term of service, etc. There is no Goldilocks mean. IT DEPENDS ON THE TASK! No single mini-public can actively deliberate face to face (be small enough), develop and refine a proposal (be willing to do intensive study and serve a long enough duration), be statistically representative (be large enough and serve a short enough term so as not to exclude people who CAN’T make a long term commitment), etc. These various REQUIREMENTS cannot be fulfilled by one body as they are mutually exclusive… We need different mini-publics for different steps of the process of making laws.
A few more responses…There is a distinction from standard “self-selection” (such as candidates willing to run for office, or activists turning out for a public hearing on their own) on one hand, and the low acceptance rate “self-selection” from a lottery, where people who did NOT step forward on their own to volunteer, but instead were CALLED and decided to accept. Recent implementations proves that a room full of lottery selected people will not be predominantly “activists.” They will not have strong pre-determined policy views. They will be open to learning and making up their minds based on what they learn and the arguments they hear. A smallish (30 -150) random body with stratified random sampling will exhibit FAR MORE diversity and representativeness than any elected legislature ever. And this diversity is vital to good decision making. They will probably be far more representative IN EVERY WAY (interests, life experiences, demographics, etc.) than any elected legislature. But they may not be sufficiently accurately representative of the population to stand in for the general population to make FINAL DECISIONS unless they have at least many hundreds of members, serve for a short duration and have adequate carrots and sticks to achieve very high acceptance rates.
LikeLike
Terry:> We need different mini-publics for different steps of the process of making laws.
That’s assuming minipublics are the only show in town. The alternative that Alex and I are working on is a judicious combination of elected and allotted bodies.
>diversity is vital to good decision making
Absolutely, when it comes to choosing between options. But diversity of options available may well be better served by the competition between internally-homogeneous elite policy-making bodies. The argumentative theory of reasoning suggests that strong pre-determined policy views are essential for representative isegoria.
LikeLike
*** The debate about « minipublics beyond representativity » is very interesting, but I think we must distinguish two concepts which are very different.
*** First, which I would name a « lay panel », where sortition put many people which are not always members of an elite or an activist minority, and which is able of autonomous brain-storming, of issuing proposals which sometimes may not belong to any partisan agenda. Such panels may be « spaces of political action and contestation”. In such a panel majoritarian rule is not necessary : an interesting idea of proposal may be selected if it is considered serious by some defined part of the panel. Representativity is not paramount. Actually discrepancy from representativity may be useful, even necessary. If the subject is prison, such a panel must include in higher proportion prisoners, wardens etc (including non-citizen prisoners). If the subject is about schools, or about teen age sexuality, the panel must include an enhanced proportion of teen agers, including below the « civic age ». Etc …
*** Second concept, the mini-populus (Dahl’s wording) which ideally must be the perfect mirror of the populus. Quoting Dahl (Democracy and its critics, p 340) « The judgment of a minipopulus wouId “represent” the judgment of the demos. Its verdict would be the verdict of the demos itself, if the demos were able to advantage of the best available knowledge to decide what policies were most likely to achieve the ends it sought. The judgments of the minipopulus wouId thus derive their authority from the legitimacy of democracy. »
*** Such a mini-populus could make decisions which are different from a referendum on the same subject. We may imagine that a British mini-populus could have decided on Brexit differently from the referendum, or an American judicial mini-populus could judge Weinstein with a different result from the result of an hypothetical verdict by a general vote of the citizenry. In both cases because the mini-populus would have better information and deliberation. Maybe the results would have been the same, but with better value. But if the mini-populus is not mirroring the populus, any discrepancy with the potential referendum vote will be interpreted as coming from this lack of representativity, and the mini-populus choice will lack democratic legitimacy (and undergo strong attacks along this line). The representativity of the mini-populus, as Caesar’s wife, must be above suspicion.
*** To blend the two kinds under the same name of « mini-public » is confusing. To say that « mini » in « mini-public » is alluding only to the size is very debatable. It can be understood as « public in miniature », which implies representativity. I think that the word of mini-public should be restricted to an allotted body mirroring the corresponding public, another use of the word could be seen as a misleading trick.
LikeLike
Yes, we are constantly coming back to the lack of standard terminology, and no single authoritative body to agree upon terminology.
I agree that “mini-public” SHOULD mean a statistically accurate representation of the public as a whole ( “a portrait in miniature”). Keith is adamant that maintaining statistical accuracy requires that they not actively deliberate (debate)… but this is at least arguable, since a goal could be to see what the public as a whole would possibly settle on WITH active deliberation – knowing that changes them away from their original perfect representativeness… and that is the INTENT of the process…. A more thoughtful portrait in miniature. (I actually tend to agree with Keith that these large statistical mini-publics should not debate between members, but am far less insistent on that point).
Many practitioners of “deliberative democracy” seem to have settled on the term deliberative “citizens assembly” for a much smaller random body that must actively deliberate, probably with stratified sampling.
Both are examples of sortition (random selection of public officials). And of course there are many OTHER uses for sortition, such as drawing from original pools that have a unique character, etc.
LikeLike
Terry:> a goal could be to see what the public as a whole would possibly settle on WITH active deliberation
I would go further and say that this is the only goal for a representative mini-public, the issue being what derivation of “deliberation” you opt for (Latinate or Germanic) — in both cases active debate being necessary. Although Jim Fishkin opts for the Latinate derivation, he contradicts himself by insisting (personal communication) that small-group deliberation is an essential component of preference modification — although I’m not aware of a DP experiment that tested the outcome without small groups and Goodin and Niemeyer’s paper on the Bloomfield Track CA suggests that the heavy lifting is performed at the information rather than the deliberation stage.
The problem is what to do when three identical minipublics (the Texas utilities DPs) come to wildly different conclusions — which one represents the considered preferences of the public? My hunch (it’s nothing more than that) is that the small-group deliberation causes the variance, but this could easily be put to the test. Convene three identical large, quasi-mandatory minipublics, provide them identical balanced advocacy (“balance” being achieved by dialectical means), then let them decide the issue without small-group deliberation and see if they come to the same decision (within an agreed margin of error). Then convene three more DPs and allow small-group deliberation and see if statistically-significant variance results.
Whilst I’d love to see the experiment done (anyone got the funds?), I think the results could be predicted with some degree of confidence. If in the first case there was statistically-significant variance then the samples cannot have been large enough. And it’s hard to see, given the experimental design, what factor other than small-group deliberation would cause the variance.
I appreciate that those who claim a “dialectic” approach to information advocacy only serves to entrench the power of the rich ‘n powerful (as the difference between the two elite advocacy camps is just Tesco vs Sainsburys) will reject the experimental design. But you’ll just have to wait until Alex Kovner’s proposal for his New Model Parties is published (currently in its third iteration after extensive feedback from a number of political theorists), to see how genuine representative isegoria is possible in large multicultural states.
LikeLike
Keith,
Seeking the same outcome from duplicate mini-publics is an unrealistic standard (or the acceptable margin of error needs to be very large). Even the exact same group of people can come to different decisions on different days of the week.There are countless unrecognised factors that can change attitudes of members beyond the presentations. One study (I don’t have the citation) found that a referendum vote with a religious connotation (could have been abortion, helping the poor, I don’t remember) has a statistically different result from different polling places depending on whether the polling place was in or near a church, as compared to in a school of public building. Likewise whether a legislative body meets in a grand ornate chamber, or in a crowded shabby auditorium can change outcomes. Simply whether it was a sunny or rainy day can affect mood and attitude. In short I feel there is a broad RANGE of legitimate public policy positions that can be said to be the considered public judgement, and duplicate mini-publics can differ but all be valid. A good democracy is a perpetually self-correcting process with outcomes that swing around the mean. Decisions will regress to the mean.
LikeLike
Terry:> Decisions will [over time] regress to the mean.
Yes, and in the long run we’re all dead. Whilst it’s true that all sorts of decisions (including voting) are affected by the weather and who won the football game, democratic theory and practice requires that the majority/plurality vote at any particular time succeeds. But in the case of multiple concurrent minipublics returning radically conflicting decisions we just wouldn’t know what the informed will of the public is. If the votes in a general election were randomly divided into three samples before counting and (albeit highly unlikely) the vote tally for each sample were significantly different, then the election would be declared null and void. The same principle applies to multiple samples of the same public.
I’m surprised that you think a common outcome would require a large margin of error — to my mind that would indicate the need to increase the size of the sample. If there are location-specific effects then each assembly would simply need to gather in a setting that had a similar ambience.
LikeLike
For a simple yes/no vote on a policy presented to a mini-public, and without active deliberation, then, yes any mini-public with adequate accuracy or representation should make the same decision. But in the case of refining of proposals to develop a final draft bill it is okay if different randomly selected bodies end up generating different final proposals. That fact would not invalidate the “public judgement.”
LikeLike
Terry:> in the case of refining of proposals to develop a final draft bill it is okay if different randomly selected bodies end up generating different final proposals.
Agreed. This is Alex’s model for the random selection of proposing agencies, the only difference being that the agencies are constituted by election, not sortition. This is because the proposal stage is under the auspices of representative isegoria, not isonomia.
LikeLike
*** The effect of deliberation on reproductibility could be evaluated experimentally, within the established system, for the criminal juries. In the USA, they deliberate without magistrates, and judge about 5% of the crimes, with usually twelve jurors, but in some cases less. Let’s imagine some trials being processed with two different juries, seated in the courtroom, getting the same informations from the trial, but deliberating separately. It would be interesting to see how often there are discrepancies.
*** In such an experiment, discrepancies would not be a legal problem, Given the official rule of doubt acting for the accused, the rule should be that the more lenient verdict prevail.
*** A famous film « Twelve Angry Men » described the influence of one juror on the final verdict. Even if the script is somewhat unrealistic, it gives force to the idea of lack of reproductibility. It would be interesting to study how much thinking there was about this aspect in discussions about the film.
LikeLike
Andre:> A famous film « Twelve Angry Men » described the influence of one juror on the final verdict.
This was also my own experience when I did jury service. The case was one of intent to defraud and, because I have personal experience in this area when my own company went bankrupt, I managed to persuade a couple of jurors to acquit after a few days tough deliberation. When the case was retried with a fresh jury they convicted unanimously after less than an hour’s deliberation.
My presence on the original jury made it (statistically) unrepresentative and I don’t see that as a problem when it comes to judging matters of fact, but it’s a serious problem if the role of the jury is to reflect the informed preferences of the public that it “describes”. That’s why in my thesis I criticise Helene Landemore for using this movie as an example of the value of active deliberation.
LikeLike
*** Keith says « My presence on the original jury made it (statistically) unrepresentative and I don’t see that as a problem when it comes to judging matters of fact”.
It seems to me a problem if a jury can send somebody to the gas chamber and if the verdict depends strongly on chance !
*** I think Keith knows that, even in English style criminal trials where the jury is supposed only judging matters of fact, many verdicts may depend on juror choices which are deeply political: opinions about “the reasonable behaviour” of people, for instance; or opinions about “the reasonable doubt”.
*** A judicial jury says a chemical society did wrong to a worker by using a specific chemical product ; a legislative jury bans this chemical; are the information and deliberation problems really very different?
*** Many problems are common to judicial and legislative juries.
LikeLike
Andre:> Many problems are common to judicial and legislative juries.
Yes that’s true. Judicial juries are a blunt instrument — we just haven’t worked out a better way of establishing the facts of the matter. My point is that in legislative juries there are no facts of the matter (only informed preferences), so if someone were able to behave in the way I did at the trial jury this would result in a corruption of the democratic process.
I recently read a draft paper on how the (lack of) descriptive representativity in judicial juries in the UK “appears to imply no injustice either to minority defendants or to minority victims of crime”. I would be very surprised if that were the case with legislative juries.
LikeLike
Keith,
Stop and think a moment and you will see your mistake in stating, “My point is that in legislative juries there are no facts of the matter (only informed preferences)…”
A dispute about “the facts” is very often the ESSENCE of the legislative issue. As one example…All political parties have a preference for full employment, but partisans disagree about what the facts and evidence of past policies actually indicate. They engage in motivated reasoning and suffer confirmation bias and fail to figure out what the true facts of the situation are.
LikeLike
Terry:> All political parties have a preference for full employment
Not so. UK macroeconomic policy in the 1980s was to use high unemployment as a way of bringing down inflation, killing off unproductive companies and beating up the unions. I agree that partisans of all stripes suffer confirmation bias, that’s why decisions should be taken by large quasi-mandatory allotted jurgas. But such bodies would be driven by informed preferences (employment -vs- inflation) not establishing the “facts” of the matter and would decide after evaluating the motivated reasoning of opposing partisans. If there were such a thing as “facts” in public policy then there would be no need for politics as we could just consult Google (or another oracle).
LikeLike
[…] First posted on Equality by Lot on Oct 8 […]
LikeLike