I’ll start by introducing myself. My name is Henry Tuttle. My background is outside political theory. I spent about thirty years in the casino business, where designing systems around intelligent people looking for exploitable rules is normal. I developed most of what follows from first principles with very little initial exposure to the sortition literature.
I’ve been developing a legislative architecture around a fairly simple question: “what is each political mechanism actually good at, and where should it get out of the way?”. My answer is FILA: Functional Integrity Legislative Architecture.
FILA uses sortition for informed judgment, but not for everything. Proposal generation is open. Filtering is designed to handle large numbers of proposals without unnecessarily eliminating potentially valuable ones. Expertise and advocacy are provided by people selected for those functions. Binding decisions are made by temporary, single-task citizen juries that hear the competing cases, vote, and dissolve.
The basic idea is to separate functions rather than asking one institution to do everything, and to avoid creating persistent decision-makers or unnecessary gatekeepers where they provide little value.
The abstract to my paper follows. Links to the full paper and to a shorter specification are below. I’d be very interested in criticism, particularly structural problems or failure modes I’ve missed.
For centuries, debates over legislative reform have largely focused on choosing the right people or the right method of choosing them. This paper argues that these are secondary questions. The more fundamental question is whether that function is best served by a permanent body.
A legislature can lose its function without looking broken. Committees convene, votes are recorded, and procedures run as written. What is lost is functional integrity—the capacity to do what the institution exists for. A permanent body can be studied, profiled, courted, and worn down from outside, while incentives and institutional self-preservation deform it from within. Call that functional deformation: sustained divergence between what an institution exists to do and what it actually does, whether driven from outside or generated from within. Capture is one member of the family, not the family. Functional integrity is the property it consumes. No law need be broken. Permanence does not guarantee capture, exploitation, or drift, but it guarantees the conditions in which they accumulate.
If this diagnosis is correct, many familiar debates over elections, campaign finance, term limits, and representation are addressing symptoms rather than causes.
Two things are foundational: dissolution and informed judgment—the requirement that whoever holds the final vote attend to the full competing case before casting it. Dissolution is not a universal institutional virtue. Drafting, advocacy, administration, and institutional memory remain permanent because they require continuity. Permanence is confined to the layer that holds no vote; the vote is confined to the layer that does not persist.The paper develops an architecture that applies these principles by separating drafting law from judging it. Legislative judgment is expressed through temporary citizen panels that hear the strongest competing cases, complete a single task, and dissolve. Unlike fixed-size legislatures, one, several, or dozens of independent panels may evaluate the same proposal simultaneously. Confidence can scale with the importance of the issue while each panel stays small enough to compel attention, reduce groupthink and social contagion, and reveal unexpected disagreement between panels as a signal that something upstream may have gone wrong.
Sortition is the default because it combines unpredictability, reduced self-selection and correlated interests, and deciders who cannot be cultivated in advance. It also delivers the broad representativeness and democratic legitimacy most of its advocates value. But the properties, not sortition itself, are fundamental. Where election, appointment, or another method would serve them better, that method can be used.
Objections, failure modes, and residual costs are treated where each mechanism is set out rather than deferred, and a final part collects the problems that remain unsolved. They are published in the hope that others will solve them: the author would rather see these problems answered than hold credit for having found them. The proposal is comparative. Its standard is not perfection, but whether functional deformation becomes less frequent, less severe, less persistent, and more reversible than it is in existing legislative systems.
Specification: https://doi.org/10.5281/zenodo.22004098
Full paper: https://doi.org/10.5281/zenodo.21996367
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
Ethan McCutchen
graytmatter
gtridimas
Henry Tuttle
H.J. Hofkirchner
HpR
James Kierstead
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 
Hi Henry, intrigued to learn you have a background in the casino business — gives a whole new meaning to the phrase “gaming the system”. I agree with you that sortition is one procedure among many, rather than a magic bullet. I wouldn’t worry about your lack of a political theory background, in my experience the people who understand these things best are data scientists like Alex, Kai, Jordan and yourself.
LikeLiked by 1 person
Thanks, Keith.
My family was in the casino business from before I was born. You could almost say I was born in a casino. My father had an “unlicensed casino” in his house in Vietnam while he was an American soldier stationed overseas, and we lived upstairs. I was technically born in a hospital, but close enough.
I grew up watching my father in casinos throughout East and Southeast Asia, and eventually went on to start casinos myself in several countries.
I’ve also lived in 14 countries and through revolutions and major changes of government. One thing that kept recurring, regardless of the political system, was that institutions and positions seemed to deform over time despite the safeguards around them. That experience probably has a lot to do with why I ended up thinking about the problem in terms of permanence rather than simply how people are selected.
I have no objection to elections. I think they’re a relatively weak selection mechanism, but one with very high legitimacy. Same with meritocratic selection – it can be useful. My concern is what happens after either produces people who occupy positions of power for long periods.
I’m actually using both ideas in the executive architecture I’m still working on.
LikeLike
Thanks, Keith.
My family was in the casino business from before I was born. You could almost say I was born in a casino. My father had an “unlicensed casino” in his house in Vietnam while he was an American soldier stationed overseas, and we lived upstairs. I was technically born in a hospital, but close enough.
I grew up watching my father in casinos throughout East and Southeast Asia, and eventually went on to start casinos myself in several countries.
I’ve also lived in 14 countries and through revolutions and major changes of government. One thing that kept recurring, regardless of the political system, was that institutions and positions seemed to deform over time despite the safeguards around them. That experience probably has a lot to do with why I ended up thinking about the problem in terms of permanence rather than simply how people are selected.
I have no objection to elections. I think they’re a relatively weak selection mechanism, but one with very high legitimacy. Same with meritocratic selection – it can be useful. My concern is what happens after either produces people who occupy positions of power for long periods.
I’m actually using both ideas in the executive architecture I’m still working on.
LikeLike
Very appropriate biog for a lottocrat!
LikeLiked by 1 person
Those who make their living from the permanent body will not kindly to its necessity being questioned.
LikeLiked by 1 person
Hi Henry – welcome!
I have not read your paper but the post and the abstract make some good points and others that I find questionable. The framing of sortition as a way to reduce the opportunity for exploitation reminds me of an old post on this site which you may find interesting: https://equalitybylot.com/2010/12/03/the-delegation-game/. I also agree that sortition, like any mechanism being considered should be evaluated by its expected outcomes and not based on other properties, such as the formalistic properties often invoked (“embodying consent”, “equal say”, “mirroring”, etc.).
But this brings me to questions regarding your statements about elections. First, you write: “Where election, appointment, or another method would serve them better, that method can be used.” What in your mind is a good use-case for elections? To me it seems that elections are an inherently oligarchical mechanism which can never be expected to produce good outcomes.
Second, you write: “I have no objection to elections. I think they’re a relatively weak selection mechanism, but one with very high legitimacy.” This seems to endorse elections because they have “high legitimacy”. But “legitimacy” (which is really just convention and propaganda in this case at least) is not a substitute for good outcomes. Again, why would a democratic society ever use elections?
LikeLiked by 1 person
Henry,
I look forward to reading your paper! The focus you have is very similar to the one I set out in my journal article from 2013, Multi-body Sortition: Athenian Lessons for the Modern Day, which you can access here without a paywall
https://delibdemjournal.org/articles/abstract/10.16997/jdd.156/
I also proposed many different bodies, selected by different procedures, and with different size, duration, etc. based on their task.
Keith, and I (though not Yoram) fundamentally agree with you that the final decision-making body must NOT be an ongoing one, but more like a large jury selected by lottery for that single task.
Keith and I disagree about the harm vs. benefit of using elections at all, however.
In brief, There are four overarching concerns I have about elections (and many secondary concerns)
Due to rational ignorance, in mass elections voters will never have adequate knowledge about their choices and thus cannot effectively use elections for accountability. Those who SEEK power through elections are far more likely to be narcissists and psychopaths than the general public, and that self-selection bias of elections enshrines oligarchy. And thirdly, campaign incentives prompt parties (which form in any electoral scheme) to seek to polarize the population into warring tribal camps, which is detrimental to civil society. And fourthly, parties select agenda items based on how effective they are at mobilizing voters (especially using hatred and fear), causing vital (but very complex, or boring) issues to get ignored until they reach crisis level.
LikeLiked by 1 person
Thanks, Terry.
I actually read your paper after I developed my architecture, when AI pointed out that we had significant overlap. I think the most obvious similarities are the separation of proposing from deciding and the use of a lottery-selected body to make the binding decision on legislation.
I’m more skeptical of using randomly selected panels for some of the other functions in your architecture. Random selection has disadvantages as well as advantages, particularly the lack of prior qualification or expertise. My inclination is to use it where its advantages are especially valuable – independent judgment and oversight – and use people selected for expertise where expertise is actually the function. For binding legislative judgment, I also think dissolution after the decision is extremely important.
I don’t substantially disagree with any of your four objections to elections. Where I’m less categorical is whether those are properties of election itself or, at least partly, properties of the way we normally implement elections.
A mass electorate making low-information choices is probably close to the worst case. In current elections, candidates can succeed based heavily on characteristics that have little to do with performance in office. If the voters were fewer, much better informed, and choosing among candidates who had already demonstrated relevant qualifications, perhaps the result would be substantially better. Whether it would then outperform lottery or another selection mechanism is a completely different question. In a sufficiently low-information environment, I think there’s even an argument that election could perform worse than random selection.
My default is lottery, but I don’t think any selection method should be used simply because it is democratic, meritocratic, random, or anything else. I want to use whichever mechanism has the best characteristics for the particular function.
That’s actually one place where I think our architectures differ substantially despite the similarities. I’ll be particularly interested in what you think about that after you’ve read the paper.
LikeLike
Thanks, Yoram.
I probably stated my position on elections too broadly.
I think the usefulness of an election scales considerably with the quality of the informed judgment behind the votes. In a mass election there is so much noise, rational ignorance and superficial decision-making that I’m not at all convinced the result is much better than random selection. But that’s also close to the worst possible implementation of an election, and I don’t think it tells us that elections can never be useful.
If prior knowledge, experience or demonstrated competence is important for a particular function, restricting the candidate pool or using some form of merit selection may improve the result. Then of course you immediately get another problem: who decides what constitutes merit and who applies the criteria? Perhaps a randomly selected panel belongs somewhere in that process.
Even with excellent information and qualification, I don’t assume election wins. Lottery is my default for a reason. But if two mechanisms produce roughly comparable expected outcomes, I think the additional perceived legitimacy from participation can matter. I suspect most people would rather accept a somewhat worse result they genuinely participated in choosing than have a slightly better result imposed on them. Obviously that stops being true as the difference in outcomes becomes large.
So I don’t regard legitimacy as a substitute for outcomes. I regard it as one of the things that affects the outcome.
LikeLike
Well done Henry. Oddly, I found the long paper more readable (in terms of following the ideas and your argumentation) than the short one.
You may want to explicitly cover two additional parameters: how you avoid manipulation (i.e ‘designing defensively), and where you think you could first apply this (I was picturing a local council). Adding practicality and ‘making it real’ can assist the reader to picture the solution.
Noting what you have already created, I hope you turn this into a short ~120pp book and get the idea in circulation. If you’ve read David van Reybrouck’s Against Elections you’ll get a sense of the tone and level of detail which appeals to a wide readership – and its that broad awareness of new ideas which is the holy grail.
Nicely argued. I’ll be poaching some of the lines and anecdotes for use in our own advocacy :)
LikeLiked by 1 person
Thank you!
I put together the shorter specification largely because Keith thought the full paper was WAY too long. I can’t really argue with him. I developed this organically from first principles with very little exposure to the existing literature, so I had no good sense of what people in the field would already take for granted and what needed to be explained. Consequently, I found myself explaining virtually everything. I suspect Keith and many of the people here don’t need quite as much hand-holding.
On your suggestions, yes – I’ve been thinking along exactly those lines. My instinct is that the first real application should be a town or small municipality. I originally had a population of around 50,000 in mind: large enough to provide a good citizen pool, but small enough that the machinery could actually be built and tested without rebuilding an entire state. I should probably make that more prominent.
The book suggestion is interesting because this is actually the tip of the iceberg. I’ve been developing a broader method I’m calling Functional Integrity Institutional Design. The basic idea is to start with what an institution is actually supposed to accomplish, look at how and why that function deforms, and then use whatever mechanisms best protect it rather than starting with a preferred form of governance.
The legislative architecture is the most developed application so far. I’m working on an executive architecture now, but that has far less of a proven analogue behind it, so it requires considerably more willingness to entertain unproven mechanisms.
And by all means poach the lines and anecdotes – just leave a few breadcrumbs back to me. As you know, besides the fame and big bucks that come along with alternative governance, reputation is where the real score is: alternative governance groupies! ;-)
And if you know anyone else who might enjoy kicking back with a beer and a 40-page paper on alternative governance, please pass it along.
LikeLike
Sorry, one more question:
What did you mean by “how you avoid manipulation”? When? At what stage? In which specific mechanism? Or any at all?
LikeLike
Glad to see there is some measure of agreement emerging. Regarding the ongoing need for elections, is there any alternative method of ensuring that the discursive diversity of the full citizen body is proportionately represented than a PR based electoral system (combined with a reduced parliamentary proposal threshold)? Added to this, perceived legitimacy is vital, given that the lottocratic coup would effectively disenfranchise the overwhelming majority of citizens. Yoram’s focus on outputs overlooks Dahl’s requirement that the demos should have complete control of the inputs and I genuinely don’t see how this is possible without elections (noting Urbinati’s observation that the political party is the representative reincarnation of ho boulomenos).
LikeLike
Keith, I have to admit I had to run that through AI to translate the political-theory terminology before I was sure I understood your question. But I don’t know enough of the terminology to know whether the translation was right, so please feel free to tell me if AI translated you incorrectly.
It seems you haven’t read the latest version of my agenda mechanism. Open proposal generation was already part of it when we corresponded, but I subsequently developed the filtering and competition mechanisms considerably.
My objection to using PR to guarantee diversity of inputs is that it protects diversity we’ve already identified and organized. What about an intense minority that cuts across party lines, a small group affected by something almost nobody else cares about, or a constituency nobody knew existed until an issue arose?
I’d rather not protect minorities by first having to define which minorities deserve representation. And open origination potentially gives us far greater diversity of proposals than filtering political preferences through a limited number of parties in the first place.
Anyone can propose. A relatively low threshold of scarce citizen endorsements gets a proposal through the first gate. A salience jury – or possibly a sufficiently robust survey – then asks whether the issue is important enough to justify the resources required for serious evaluation. It isn’t deciding whether the proposal is good. This is intended to solve the resource-allocation problem that makes completely open systems like California’s difficult.
The other major addition is that a jury can reject a proposal but decide the underlying question deserves better answers. That opens the question to competition, and anyone can submit an alternative. Nobody constructs the menu of acceptable solutions.
So if I’ve understood you correctly, we’re trying to protect the same thing differently. You’re trying to ensure that the citizen body’s political diversity is proportionately represented in generating proposals. I’m trying to avoid having to define that diversity at all – while hopefully allowing considerably more of it through.
LikeLike
One concrete example occurred to me after posting this. Everyone participating here presumably thinks sortition has some useful political role, although we probably disagree considerably on many other political questions.
Under a party-based system, who represents that preference? Which party do we join? Do we need to agree on taxation, immigration, foreign policy and everything else before our shared view on sortition gets represented?
This is part of what I mean by losing diversity when inputs have to pass through parties. A political constituency can exist around one issue without being a constituency around anything else. With open proposal origination, it doesn’t have to become a party first.
LikeLike
Henry,
When designing a democratic system, there are two overriding priorities: 1) simplicity and 2) all citizens must have a role in authorising the laws. The modification to the status quo proposed by Alex, Kai, Jason and myself relies on two principles that everyone is familiar with: 1) election and 2) decision making by juries. Our proposals have three interrelated elements:
Electing representatives to the legislature
Lowering the proposal threshold for new laws
Decision making by large juries
Citizens would be free to elect representatives or parties (depending on the electoral rule) and to choose their own principle of distinction. Proposals will be made by ad hoc (and dissolving) coalitions which will vary depending on the issue, so there is no need to find a party that you agree with on everything. And the jury holds the whip hand, so successful proposal coalitions will seek to predict the considered preferences of the citizen body, issue by issue.
This really is as simple as it gets, and is far more likely to gain public approval than a byzantine hierarchy of voluntary panels, committees and juries, which will tend to privilege organised interests and those with the resources to collect signatures. Remember also that elites come in all shapes and sizes. Arguably the culture elite is just as dominant as the rich and powerful. Remember that Habermas was a member of the Frankfurt School, hence the wish to further empower the chattering classes.
I’m still waiting for someone to argue against Urbinati’s claim that the political party is the modern (representative) incarnation of ho boulomenos.
>I’d rather not protect minorities by first having to define which minorities deserve representation
I’m not sure where this comes from, it’s certainly not part of any PR system that I’m aware of.
LikeLiked by 1 person
Well done Henry, an interesting and worthwhile contribution.
I did come across some expressions that were sufficiently strange that I wondered if they began in an AI. I have no purist objection you or anyone else using AIs, but they’re very clever at appearing to say things while actually being evasive. For instance your longer paper says this “Pooling makes additional panels narrow the estimate rather than create additional veto points, so panel count becomes a dial for confidence rather than a hurdle count.”
Then there’s “Agenda intake has no standing gatekeeper.”
I haven’t read your paper carefully yet, but I genuinely don’t know what these passages mean.
OK – just writing out both passages and then staring at them, I have figured out what they mean, but I think they can be expressed quite a bit more simply.
LikeLiked by 1 person
Yes. Those were written by AI to express ideas that originated from me. Thank you for pointing them out. I just thought, “that’s how you guys speak, I guess.”
Pooling is supposed to mean that the individual votes from multiple panels are combined rather than treating each panel as a separate decision. So if one panel votes 6–5 and another votes 11–0, the result is 17–5. The 6–5 panel shouldn’t have the same influence as the 11–0 panel simply because they are separate panels.
The “dial for confidence” part is supposed to mean that you can run more panels when you want greater confidence in the result. More independently selected voters should reduce sampling noise without creating additional veto points.
“Agenda intake has no standing gatekeeper” just means that there is no permanent board, committee or official with discretionary power to decide which proposals are allowed into the system. There are rules and thresholds, and later a salience jury or survey can decide whether something warrants more resources, but that body is temporary and newly selected rather than a permanent agenda authority.
I certainly don’t like the way my AI phrased either of those. I’m not sure my versions are elegant either, but at least I hope they’re clearer.
AI claims this happened largely during my attempts to compress the paper. It also claims you’re right and that it should have written:
“Votes from additional panels are pooled together, so adding panels increases confidence in the result rather than creating additional vetoes.”
and
“No permanent body decides which proposals are allowed onto the agenda.”
Those seem considerably less painful.
BTW, I specifically noted in the paper that my collaboration has been almost entirely with AI. Its contribution has primarily been arguing with me by asking stupid questions, making stupid assumptions and suggestions, and forcing me to figure out exactly why I think it’s wrong.
After several back-and-forths, we’ll usually reach an understanding and I’ll say, “OK, summarize that.” It cleans up the conclusion, and then I edit it.
Some things like these apparently slip through when I ask AI to compress areas it assures me have “cost me nothing.” Obviously that wasn’t true here. I’ll be more careful.
AI did occasionally suggest substantive mechanisms as well; I’ve tried to identify those in the paper rather than pretend every useful idea originated with me.
LikeLike
I think this gets to our actual disagreement.
I agree completely that simplicity is valuable. My original version was basically proposals, signatures and a jury. Most of the added machinery appeared because I kept finding ways to break that version.
But I think your simplicity comes partly from narrowing who gets to generate proposals. How low can your proposal threshold go before flooding becomes a problem? Alex uses scarce tokens; your superminority gives minorities a relatively easy route. What limits volume in your version?
My instinct goes the other direction. In an ideal system I’d like almost every idea to have a chance to surface. We’ll shovel a lot of manure, but we’ll also search a vastly larger space and occasionally find things a few hundred professional politicians never would.
Open origination also doesn’t remove any existing source of sophisticated proposals. Parties, corporations, NGOs, experts and elected politicians can all keep producing them. They just don’t own the starting line.
That matters because a permanent elected legislature is not spending most of its time searching for the best laws. Its members are also campaigning, fundraising, managing party relationships, serving constituents and doing everything else politicians do.
The real problem with open proposals is filtering. My answer is to treat that as another bounded judgment problem: cheap entry, scarce endorsements, then a temporary salience jury before serious resources are spent.
So perhaps the difference is simply this: you reduce the filtering problem by limiting proposal generation to representatives. I’m trying to preserve the much larger search space and solve the filtering problem instead.
LikeLike
Henry:> You reduce the filtering problem by limiting proposal generation to representatives.
Yes, representative of the discourses/ideologies etc prevalent in society at that time.
>How low can your proposal threshold go before flooding becomes a problem? Alex uses scarce tokens; your superminority gives minorities a relatively easy route. What limits volume in your version?
I think there would be an element of trial and error. I imagine the threshold would lead to between 3 and 5 proposals per agenda item. It’s all very well allowing a thousand flowers to bloom, but somebody has to pick them. Citizens who had a novel proposal would be able to lobby a sympathetic MP/Congressman.
LikeLike
All seems reasonable on paper. And certainly better than the existing system.
My problem is what I think those 3–5 proposals look like in real life: mostly proposals written or heavily shaped by lobbyists, campaign donors, organized interests and people who already have political access. Basically, the same pork-laden, special-interest-catering garbage we have today — only now we get three to five competing versions of it.
Still far better than the existing system, because those 3–5 proposals ultimately have to get through a lottery-selected jury.
But in mine, those same professionally produced proposals can still exist. They’re just occasionally joined by an actual proposal that didn’t need a sponsor, donor, party or political connection to get through the front door.
And “lobby a sympathetic MP/Congressman” is exactly where I think the problem gets worse. If you have a news crew following your cause, belong to an organized constituency, have money, staff, connections or some other reason for a politician to listen, fine.
But an ordinary citizen does not have meaningful equal access to this mythical sympathetic Congressman. Technically the door exists. In practice, attention is rationed by exactly the kinds of resources I’m trying to keep from controlling proposal access.
In that respect, I actually think the resource allocation is worse than California’s open system. California has plenty of problems, but at least ordinary people can get through the front door.
I think we’re making a very different tradeoff. You’re limiting the number of flowers allowed to bloom so somebody can choose among 3–5 ugly ones. I’m trying to let a thousand flowers bloom and build a mechanism good enough to shovel through the manure until we find the beautiful ones.
LikeLike
Henry:> California has plenty of problems, but at least ordinary people can get through the front door.
The mistake lottocrats make is to claim that democracy is the rule of “ordinary people”. The demos in democracy is a collective entity, albeit a deeply-divided one. As I mentioned before, the proposal threshold would be decided on a trial-and-error basis in order to get the optimum trade-off between discourse diversity and outcome — get it high enough and the lobbyists cancel each other out. Democracy in large poleis requires representation (either descriptive or the active representation of interests). What lottocrats advocate is aleatocracy.
Speaking personally, I hate restaurants with long menus, I’d much prefer a limited choice. I wouldn’t like that choice to be constrained by self-selecting busybodies (I prefer professional food critics), and the restaurant knows it has to keep its customers happy or they will decamp to the opposition. And anyone is free to set up their own restaurant. That’s (more or less) what liberal democracy means.
I’m no expert on Californian open-initiative democracy, but the impression we get from this side of the pond is that you don’t live in a well-functioning state, especially given that the need to collect signatures privileges the well-connected.
LikeLike
Keith, I’m starting to wonder whether we’re actually discussing the same system.
Some of your objections seem directed either at lottocratic systems generally or at an earlier version of mine, before I developed the current agenda mechanism. I’ve mentioned the newer mechanism several times in our correspondence and tried to summarize the relevant parts here, but I realize it’s considerably less simple than the architecture you’re proposing, so perhaps my summaries haven’t made the distinction clear.
The old version really was much closer to “anyone proposes, collect enough support/signatures, then send it to a jury.” I found too many ways to break that, which is why the agenda mechanism changed substantially.
In the current version, self-selecting proposers do not determine the menu, California-style signature collection is not the main gatekeeper, and an individual proposal can be rejected while the underlying issue is opened to competing solutions. So when you criticize self-selecting busybodies constraining the choices, or the resources required to collect signatures, I’m not sure whether you’re criticizing my current mechanism at all.
Likewise, I’m not trying to defend every system that substitutes sortition for elections. If your argument is against lottocracy or aleatocracy generally, I don’t know enough about those systems to defend them, and that isn’t really the argument I’m making.
Could you tell me which system you’re actually criticizing here – the general lottocratic approach, my earlier version, or the current agenda mechanism? If it’s the current one, I’d be much more interested in which specific part you think fails and why.
P.S. I wouldn’t blame you for not wanting to dig through the full paper to find the revised agenda mechanism. That’s why I also sent it to you separately by email. Here it is again:
https://doi.org/10.5281/zenodo.21923261
LikeLike
Henry, I’ve tried to read it (it was a struggle, on account of the complexity). You make Terry’s byzantine hierarchy of agencies look comparatively simple. And it doesn’t clear Dahl’s first criterion:
“The demos [a collective entity] must have the exclusive opportunity to decide how matters are to be placed on the agenda of matters that are to be decided by means of the democratic process.” (Dahl, 1989, p.133)
The advantage of election is that it’s simple — voters just choose who they want to speak for them, according to whatever criteria they deem appropriate. The existing flaws (that we all agree on) can be ameliorated by reducing the proposal threshold and leaving the final decision to a representative jury (along with the introduction of PR, campaign finance limitations, and a modern version of Athenian euthynai, graphe paranomon & dokimasia). There is no chance of a system as complex as you (or Terry) proposes being accepted as a legitimate alternative to citizens exercising their free choice (for the reasons that Manin outlines in his book).
LikeLiked by 1 person
First, thank you for trying. My apologies that it wasn’t clear. Please let me know where you had the most trouble so I can fix it.
The basic agenda mechanism is actually simpler than the document makes it look:
1. Anyone can propose an idea.
2. Citizens have only a limited number of endorsements to give each year, so flooding the system requires persuading actual people rather than simply spending more money.
3. A temporary group of citizens, or a random survey, then decides whether the issue is important enough to deserve the resources of a full hearing.
4. If the proposal itself is rejected but the problem is worth solving, the question can be opened up and anyone can propose a better solution.
5. The rest is mostly safeguards against gaming those steps – making sure proposals are understandable and decidable, endorsements can’t simply be bought, and nobody quietly gains control over which alternatives get considered.
Most of the complexity in the document is me trying to break those few steps and then adding defenses against the ways I found.
On Dahl, I don’t know his work, so I can only respond to the criterion you’ve quoted. But I’m not sure I understand why your system obviously satisfies it and mine doesn’t.
In mine, any citizen can originate something, citizens decide which proposals are worth spending their scarce endorsements on, and citizens ultimately decide which issues deserve serious consideration. There is no permanent group with the exclusive power to decide what the public is allowed to consider. What’s more “demos” than that?
In yours, citizens elect representatives, but then those representatives control which proposals reach the system. An ordinary citizen with a new idea has to persuade one of them to take it up. Perhaps Dahl considers that the demos controlling the agenda through representation – I don’t know enough about Dahl to say – but it certainly isn’t obvious to me that this gives the demos more control over the inputs.
Your “lobby a sympathetic MP/Congressman” answer seems to require considerably more faith than I have that an unorganized citizen with no money, constituency or political connections will receive meaningful access to that gatekeeper.
On legitimacy and acceptance, though, I agree with you. I don’t think my system has ANY chance of replacing a functioning democracy.
But I’m afraid I have to extend that problem to yours.
Your elected representatives don’t make the final decision either. Your lottery-selected jury does. That’s the important legislative power.
So if citizens will refuse to accept my system because the people making their laws weren’t chosen through their “free choice,” why will they accept yours? And if they’ll accept a randomly selected jury having the final say on legislation, then you’ve already crossed most of the legitimacy bridge that I’m being told is impossible to cross.
That’s why I’m not sure your compromise with elections buys enough additional acceptance to justify what it preserves. You preserve a permanent political body, along with the campaigning, parties, lobbying, organized interests and everything else that attaches itself to permanent political power – but the elected body still doesn’t make the laws.
I agree that mine is more complicated. I wish it weren’t. My original system was much simpler. I kept adding machinery because I kept finding ways to break the simpler versions.
And this is probably where our fundamental disagreement lies. I think a permanent body with discretionary control over the agenda will eventually deform. I don’t know how to prevent that. Everything I have ever seen that is adaptable and continually interacts with other adaptable agents eventually gets studied, pressured, optimized against and changed by that interaction. The longer the same target remains in place, the more time every interested actor has to learn how to influence it. Politics seems like an especially hostile environment in which to assume otherwise.
So I’m unwilling to preserve a permanent agenda-setting body merely because it makes the architecture simpler.
As for adoption, perhaps I’m too pessimistic, but I think democracy has become a religion in the modern world. It’s based in deep faith that it will work regardless of any evidence to the contrary. People don’t abandon it simply because it performs badly. I suspect they abandon it only when the existing system has lost nearly everything.
Look at the United States. Whatever label we put on its present condition, could we ask for a much more dramatic demonstration that the institutions are failing? And yet the overwhelming response is still to march back to the polls and try to elect the right people next time.
So I don’t expect FILA to replace a reasonably functional Western democracy. Maybe a small municipality would be willing to experiment, and that pilot might provide some evidence that everyone will then proceed to ignore. The more realistic large-scale opportunity, though, may be a country rebuilding its institutions after revolution, collapse or some other failure severe enough that returning to the old system no longer seems obviously preferable.
LikeLike
I’m afraid my eyes glazed over on account of the sheer complexity of the detail. If it had that effect on me, then you’ve no chance with Joe Public.
Your reference to “any citizen” shows how you are missing Dahl’s point — the demos is a collective (albeit deeply divided) entity. As soon as an individual citizen performs a speech act she is only representing herself or those who authorised her (see Pitkin’s book).
>I don’t think my system has ANY chance of replacing a functioning democracy.
I don’t have any time for ground zero approaches, hence my preference to make do and mend. As you pointed out in our first exchange I am conservative by nature.
>So if citizens will refuse to accept my system because the people making their laws weren’t chosen through their “free choice,” why will they accept yours? And if they’ll accept a randomly selected jury having the final say on legislation, then you’ve already crossed most of the legitimacy bridge that I’m being told is impossible to cross.
For two reasons: 1) familiarity with the democratic jury as a valid decision-making entity. 2) The LLN. Iff it can be demonstrated that different samples will come to the same conclusion (within an agreed margin of error), there is no difference between the jury decision and that of the population it describes.
>You preserve a permanent political body
Not really — Alex’s proposal is that there should be rolling elections, so the representativity is constantly updated.
>I think a permanent body with discretionary control over the agenda will eventually deform.
We’ll have to reserve judgment on that. In our view the three elements would interact in such a way as to minimise this risk of capture.
>Look at the United States. Whatever label we put on its present condition, could we ask for a much more dramatic demonstration that the institutions are failing?
We can agree on that. The disagreement is whether it’s best to fix it up or junk it in favour of an uproven (and barely comprehensible alternative). The precautionary principle would suggest the former, rather than eagerly awaiting system collapse. In 1968 I wanted to enrol at the Sorbonne and learn how to hurl cobblestones at les flics, but I’m a little more cautious in my dotage.
LikeLiked by 1 person
Henry.
I read the full paper, and agree with your fundamental principles and strategy. We do have an awful lot of overlap.
A key element, I think you might want to incorporate is a rotating Rules Council that proposes improvements and fixes to the initial design. Your design allows for refinements, but I think a body focused on that task is worthy.
If you are interested, my recent book (which is open access, so the e-book versions are free) Democracy Without Politicians: Government By the People is available here:
https://terrybouricius.com/
However, the bulk of that book is about why elections are not-democratic, with relatively less dealing with better design (using sortition).
The little bit you have in your paper on strategy to get there (a bi-cameral way-station) is problematic in my assessment. I urge you to read this other paper of mine as to why.
https://www.academia.edu/37578530/Why_Hybrid_Bicameralism_is_Not_Right_for_Sortition
That paper also proposes a different transition strategy, I refer to as “peeling” — removing one policy domain at a time from elected governments, and entrusting that domain to the new system. My vision is that eventually the elected politicians will be like the remaining monarchs of Europe… perhaps still existent, but with all real power having been peeled away.
If you want to write another paper about the executive function, a short piece I co-authored on how to manage the executive role of government is here:
https://www.academia.edu/11673705/An_Idealized_Design_for_Government_Part_2_Executive_Branch_Accountability
In brief, a purely administrative executive without a policy role, and subject to regular audit and review.
LikeLike
Terry,
Thank you. I’m delighted that you read the whole thing, and especially interested that you see the overlap at the level of principles and strategy.
The Rules Council immediately raises one of the central issues behind my methodology. My starting assumption is that any human institution – really, any adaptive component repeatedly interacting with other adaptive actors – will tend to deform over time. Where continuity is necessary, I think we have to accept some of that risk and contain it through restricted powers, oversight, separation of functions, rotation, etc. Where continuity is unnecessary, my instinct is to dissolve the body when its task is finished and reform it when needed.
That’s also why I’m cautious about long-lived randomly selected bodies for expertise-heavy functions. Random selection buys some very valuable things, but if a body has to remain for years in order to acquire the expertise necessary to do its job, it may combine the qualification disadvantage of randomness with the deformation risks of persistence.
Your “peeling” strategy sounds immediately more attractive to me than the bicameral transition, and I’m going to read that paper closely.
I’ve also begun developing the executive side. My assumption there is that executive continuity cannot be eliminated the way legislative continuity can, so the problem becomes how to improve the candidate pool, improve what we know about candidates before selection, create conditions in which good executives can perform well, constrain abuse, and allow successful executives to continue while still making removal possible. I’d love to get your input or help developing it.
If you’d be open to it, perhaps we could continue by email or have a call sometime.
LikeLike
Terry:> [Henry and I] have an awful lot of overlap.
Yes indeed, you have described your project as a utopian blueprint and Henry acknowledges the need for a revolution or total system collapse before his proposal would be taken seriously. Meanwhile Helene Landemore seeks to foster “civic love” and James Fishkin’s aim is no less than the deliberative transformation of society. What worries those of us who are just trying to make things a little better is that the sortition option will no longer be considered by those seeking to work with the crooked timbers of mankind.
I do, however, agree with Terry’s “peeling principle” for issues that elected politicians find difficult to resolve (abortion, climate change etc). My original proposal was for a sortitional alternative to the Brexit referendum, but the input advocacy would still be in the hands of political parties (Leave, Remain, Renegotiate). Anything else would be a travesty of democracy.
LikeLike
Henry,
I’ll wait until you’ve had time to read some of my stuff on the executive role, before going into any detail… but my model is the city manager system. A sortition formed search committee that would seek out an executive team. As a general principle, anyone who SEEKS power should be excluded, so no self-promoting applicants. With periodic performance reviews by sortition panels, it is possible a good executive team would serve for a long time. By regularly having freshly drawn performance review panels, each is independent and did not appoint the current executives, so does not need to defend them to save face.
LikeLike
Terry,
I’ve now gone through both pieces. I think your “peeling” strategy is substantially better than the bicameral bridge I sketched, and I expect to revise that section and cite your paper. I still think a bicameral version may have some value as a pilot simply because even a compromised experiment can produce information, although the reputational cost of failure may outweigh what it teaches. I’ll probably retain it as an option rather than the preferred transition.
On the executive side, I think we may differ more fundamentally. Elections as currently practiced strike me as a terrible selection mechanism – largely a political circus conducted among poorly informed voters. But the vote still confers public legitimacy, and I suspect elections select somewhat better than random choice despite all their defects.
Your search-panel model improves the information available to the selectors, but I’m less sure it solves the candidate problem. A randomly drawn panel is unlikely to personally know many people qualified to run a government, so in practice someone still has to construct or heavily influence the pool it searches. Otherwise it may simply gravitate toward prominent executives, media figures, academics, military figures, etc. Those people have also self-selected for status and visibility; they just did so outside politics. In some respects that may be a worse filter than candidates who at least spent years demonstrating an interest in public service and building a relevant record.
I also share your distrust of people who seek power for its own sake, but excluding everyone who actively seeks the job seems likely to eliminate many people who deliberately prepared for it because they genuinely want to do it well. My instinct is to try to make the path to executive power itself filter motives rather than treat willingness to seek the job as disqualifying.
I’ve been working through many of these same problems in my own executive architecture, from a somewhat different starting point. I’d be interested in how you think about these tradeoffs.
LikeLike
This is a note from Kai O’Reilly: “I have also redesigned my agenda system in the updated draft I am currently working on, and I think my new design of a proportional agenda queue based on legislator backing is much more dynamic than my previous agenda jury system.”
I’m genuinely puzzled as to why pragmatists like Henry and Terry will not consider an agenda-setting role for elected legislators (Yoram has ruled out election on principle). Both perceived legitimacy and Occam’s razor would make it the obvious solution.
LikeLike