notes-groupDecisionMaking-liquidRepublic-todo

modifications that i plan to make in the bylaws:

council = jury

juries ideas:

a council + a judge is a court. in this case the council is acting as a jury.

the court must decide thing unanimously. the judge is supposed to vote in agreement with the jury unless they feel that the decision must be the other way as a question of law, however. (so e.g. the judge can throw out a case obviously without merit)

in a given constituency, the only eligible candidates for delegate are those whose council score is positive, and also greater than the median of their constituency. for elect, greater than the median of the country.

notes: the >median requirement ensures that each higher layer of the delegate pyramid will tend to have delegates with higher council scores

in order to bring a court case, you must put at risk/temporarily spend 6 points of your council score. if the case looks like it will take a long time (time = the average amount of time a juror must serve per score point gained (note: this time includes time in unsuccessful councils) * 6 jurors), the council can ask another council (an appeal council) to review and assign/demand more points for this case. each council member has a preset limit on how large a case they are willing to accept (or mb do this according to next scheduled vacation), and if the point score for this case is too high for any member, the case will not be assigned to this council. the system tries to keep together council members with high limits (e.g. willing to hear long cases). the council must come to consensus within the time limit set for the case (otherwise the issue is undecided and automatically goes to another council, and the councilmembers are penalized for indecision, see below).

if you win the case, you get your 6+ points back, and the defendant is charged them instead. you cannot bring a case with points you do not have, but defendants without enough points just go to zero.

you can have just a trial by a single judge if you only want to risk 1 point. the defendant, however, can always insist that it be a jury trial, in which case the prosecution is forced to risk more too.

other things that cost points (generally 6 points):

each person's score goes up 1 whenever the council they are sitting on decides a case (which is always by consensus) (regardless of if the decision is 'accept' or 'reject'). the score goes down 1 whenever a council they are sitting on cannot come to consensus within the time limit set for the case, or down by 4 when they resign from a council (even if no case is pending). if a member resigns, the issue at hand is undecided and gets reassigned as if the council could not come to consensus, but the other members are not penalized.

there is no 'findings of fact/findings of law' distinction, although the judge (who is not anonymous) is bound to vote against the rest only if there is an issue that is a question of law. mb the judge should only talk about questions of law, too.

judges are not anonymous, although the other jurists are.

issues can be appealed, but only if the court decides other than the 'default' for that issue type. for trials, the default is 'innocent'. For resolutions, the default is 'accept', and for flagged posts, the default is 'unflag'.

an appeal costs points as if the issue were re-brought, minus one (e.g. there is little reason you would re-submit, rather than appeal, if your main motion were rejected). when an issue is appealed, it it sent to a council where the minimum score over all members is greater than or equal to the median score in the old council.

appeal court levels are created by taking the median score over all councils in the next lower level

we try to place people on councils with similar scores to their own. so the more experienced councilmembers dont have to deal with educating noobs. this makes the system less miserable for the most dedicated council volunteers.

when a court is reviewing a proposed main motion, they can ask for amendments and the proposer can make them, or not. the court may refuse to accept the motion if their demands are not heeded, or they can make suggestions that are not demands if they choose to. The suggested criterion for approving main motion and amendment texts is:

You may also reject motions if the submitter rejects amendments to it that the council has demanded. The council should only demand amendments accomplishing what seems to be the main aim, while that greatly improving it, or while making it much more pallatable to its opposition.

A candidate for delegate (even of a low constituency) or elector must have their jury record reviewed by jury to ascertain whether they actually participated or whether they just agree all the time to get points. If the latter, there is a note on them on the ballot that says so, and they are listed on the last page along with the candidates that a jury judged to be likely corrupt or criminal or violent (censored candidates, see below); but if they are a candidate for judge, this blocks their candidacy (if not appealed). However, the record stays anonymous (clearly, in a small enough group, this effectively breaks anonymity, however -- i guess that's fine).

It suffices to meet the council score condition at the beginning of a run for elected office; after that point, if one's score drops, it does not disqualify. a candidate remains a candidate in their own recall election regardless of their council score. however, to be re-elected, one's council score must satisfy the condition.

having the councils as 'courts' allows us to use the 'positive votes only' variant of reweighted score voting, which gives little incentive for negative campaigning. to take care of the equivalent of 'flagged comments', the courts adjudicate the flag (which they might actually do when the system is being used on an online forum; or, there might be a 'decorum' category of court for this; actually i guess we should call them 'censors'. note: no PRIOR restraint; also, all banned comments can still be viewed by browsing at -1).

what about the actual situation of crazy or evil candidates, those for whom negative campaigning might be appropriate? We could allow them to be 'flagged/censored' on the ballot if a court determines that they should be. This doesnt mean they're hidden, it just means that they're listed is a special section of 'censored candidates', along with a short statement of why the council censored them, and a rebuttal by the candidate.

'refer to committee' refers to courts. a court that takes a motion referred to them cannot 'reject', they can only approve, by consensus, a rewritten version. BOTH the original version and the rewritten version become alternatives in the final vote (which i guess is decided by approval voting?, with a runoff between status-quo and the winning motion).

court category ideas:

or should censorship and trials be mixed in with legislation, to give legislative committees more 'hands on experience' with the effects of legislation? maybe trials should all be mixed in with 'justice'. and maybe censorship legislation should be mixed in with the activity of censorship. so:

court category ideas:

note: motions by the reps (the delegates and the electors) ARE NOT reviewed by councils. the parliamentarians do that. delegates and electors can appeal parliamentarian decisions to the High Court, who can overturn with a vote of consensus - 1.

perhaps only the highest-level (e.g. appeal) councils get referred to by refer to committee. e.g. you work your way up reviewing member motions? this suggests that motion reviews should be subcategorized by topical area, so that committeemembers can become experts by the time they are elgible to be referred to.

refer to committee always contains a time limit. if the committee cannot come to consensus by this time limit, the topic is simply returned to the Forum as-is.

subreferral: if a committee gets a big topic, it may wish to divide it into parts and have each part be sent to another committee, and then to integrate the parts together.

"parliamentary categorization" refers to determinations like which voting threshold category a given motion falls within

now, can courts still influence legislation? i propose yes, and i propose a different mechanism than the one i envisioned before: when there is a tie among the 3 houses of the legislature (that is, 1 passes a resolution, one rejects that resolution), then councils can break the tie; each council can, if it reaches consensus, cast a vote one way or the other, and the sum of the votes, times the council scores of the councils casting them, determines the vote of a fourth, tie-breaker house. If the third house accepts or rejects, however, then the council votes have no effect.

councils decide via majority vote whether to begin discussing tiebreaking legislation. there is no score penalty if they fail to come to consensus on this, although they still get +1 if they do come to consensus.

council scores are just the sum of the council scores of everyone on the council, with the exception that councilmembers with negative council scores only contribute a 0 to the sum (actually i think i decided that there can be no negative scores..).

councils could also produce a 'council summary' of a legislative debate, to be shown alongside the for- and against- summaries created by voting up.

the system randomly assigns incoming requests to juries (must be random to avoid stacked jury attack, where one political party tells a small proportion of its members to vote for other parties, just to have a few juries whose members are all really from one party -- with stacked attack, this only helps in proportion to the votes that the large party is giving up)

if there is no queue of appeals, appeals courts are always routed to first -- if there are not enough incoming requests+appeals to saturate them, then the lower courts get no action (because we are guessing that the experienced appeals courts produce the quickest and best results)

maybe # of appeals levels is determined by queue; the goal is to add more and more levels until waiting time is minimized (e.g. until the reduction in waiting time at the next appeal level is balanced by the increase in waiting time at the lower level because of the loss of jurors to create the new level)

in appeals, the court first makes a decision and records it, then they can look at what the lower level said, and consider changing their decision.

should we let councilmembers stay with one council, and try and form bonds with them and bargain and such (i'll agree on this case if you agree with me on that legislative tie-breaker)? i'd like that, but i think it makes the logistics too complicated when people go on vacation -- when they're on vacation, the rest of their council would have to wait for them (or accept a temporary replacement member, leading to more complex logistics). also, then you have to have rules about how many vacations you can take. it's logistically simpler to just ask each person how long they have to continuously serve, then place them randomly on juries with people with a similar length of time currently available.

should we let juries refuse cases assigned to them? i'd like to say no, but if we said that, people cant recuse themselves. on the other hand, since identities are anonymous, we cant easily enforce recusation, so maybe we should not have it because it would be uneven. on the third hand, we could disclose identities to the judge, and they could ask ppl questions relating to recusement.

written jury deliberations are public but identity of members is anonymous, even to the other members

the judge's identity is not anonymous.

resigning from a jury, or being kicked off by being absent too often, subtracts 4 from your score. this can go negative (only skill: see below). it costs 1 to bring a case. corporate entities can bring cases only using their officers' pts (so the corporation better have at least one officer who is a good citizen -- hiring someone to only do that (only provide points), without having other powers comparable to the company's other officers, is illegal). points are non transferable and cannot be bought.

but since it costs 6 pts and only 5 ppl get 1 pt, and there are negative pts for nonagreement, the system will run out of pts and no one will be able to propose anything or bring a case. so, give ppl pts for offering to serve when no case is assigned. split pts into 'skill' pts and 'effort' pts. only effort pts are incremented just for offering to serve. otherwise, both are affected similarly by actions. pts are per-category. everyone starts with 0 points in each category at the beginning.

defending a case cant give you negative pts: the price you pay when you lose a case that you were defending cant push you below 0 (effort pts); and the price for being on a council that didnt come to consensus cant push you below 0. if you were already below 0, it cant push you lower. skill pts can go below 0.

effort pts are the things that you spend. skill pts are used in selecting appeals courts. remember, both skill and effort are normally incremented and decremented together depending upon the outcome of things, except for the effort below 0 exception, and the effort points awarded for offering to serve but not being assigned.

effort pts can be decremented by voluntary actions such as resignation or too many unexcused absences for council sessions.

effort points are awarded for offering to serve but not serving only until you have 6 pts in the category. you can only earn more pts than this by actually serving.

hmm now that councils are so great, i'd rather have them immediately, rather than the three phases of org size. how would we handle that, though? it seems silly to tell people that if there are 6 people in the organization, 5 of them must be delegates. usually in that situation, 1-3 of them are more involved or respected than the others. i guess the answer is to allow 1 person to hold multiple elector or delegate seats. disallowing this could be an option that the group could enable later. i guess that term limits should be similarly made optional, to allow small projects to adopt this system immediately in full form. and similarly the same person can be an elector and a delegate. maybe these restrictions could automatically come into being when the group got sufficiently large.. or maybe not.

this is a little dangerous though.. if the same person holds a majority of the elector seats, and a majority of the delegate seats, then they are a dictator and can't be overruled by anyone else. if they hold a supermajority of both, they can amend the constitution. maybe let that go though... if 66% of the voters want that to happen, how could anything be done about it? That's a constitution-amending supermajority already. another answer would be: this is one cases where the rules for a government should differ from the rules for a private organization.

btw did i ever write down anywhere that p should be 5 (not 7 or 11), because it's just too hard to get anything done with lots of people, and

open question for amendable code: should non-top-level delegates have open or secret ballots? split it somehow (some levels open, some secret, or maybe some delegates on each level are open, some secret)?

each delegation can pass resolutions with a 66% vote. the set of such passed resolutions become the 'platform' (as in 'party platform') of that delegation. delegates are encouraged but not required (e.g. they are allowed to disagree and to vote their conscience) to try to get their delegation's platform resolutions passed at the next higher level.

resolutions may not say anything about individual candidates for higher levels of the delegate pyramid -- they may only discuss general issues, not individual people.

each candidate must say if they 'agree as stated, disagree, sort of agree (agree but think there's a minor problem with the wording of the question, or think that their opinion on this issue cannot be reduced to a yes/no on this question)'. each candidate must also rank-order their priority of the resolutions for their delegation.

if constituents of a constituency can complain if they think that their constituency's parliamentary procedure was not followed. a candidate with more than 40% complaints is not elected (e.g. that constituency has no delegate).

delegates are provided standard software for doing all of this by the organization as a whole, but each constituency can, if it chooses, adopt its own parliamentary procedure and not use this software. in fact, all this stuff about platforms and voting thresholds is itself only a default. all the top-level organization cares about is the registration of who is with which delegation (each delegation submits a list, and so does each persion, and a person is considered to be with a delegation only if both the delegation and the person say that they are), and which delegate was selected. recall that to initially form a constituency you need 50% more than the minimum threshold below which the delegate will be removed; so if 1/3 of the ppl leave, then the delegate is removed. this means that at least 66% of the candidacy needs to agree to support whoever is chosen even if they weren't their first choice; if the constituency's parliamentary procedure has turned into something unfair, or if its chair is not enforcing it correctly, the remedy for the constituents is to resign. as a side-effect, have the std/default parliamentary procedure requires a candidate to get at least 66% approval.

if a higher-level constituency passes a resolution, this is (in the default parliamentary procedure) put to a vote in each supporting delegation. this gives supporting delegations a chance to express dissent by not passing the resolution or even by passing its opposite.

architectural principals: -- competition between constituencies -- no mass campaigning for delegates -- delegates chosen bottom-up, no top-down influence or circle of power in the delegate pyramid -- little incentive for negative campaigning -- small groups -- when no firm agreement, usually better not to pass legislation (which is restrictive) or (negative) judgement -- in the absence of strong opposition or procedural obstacles, the executive should just decide things unilaterally -- the chair's role is solely negative -- the chair opposes the government -- the chair is a position of individual conscience, not a position where one relays the views of those who elected you -- judges are a position of individual conscience, not a position where one relays the views of those who elected you -- electors and delegates are free to vote their conscience, but the system encourages them not to via the threat of losing the next election or being recalled

a condition for elgibility of a person to be initially appointed a judge in a circuit is that they, as a councilmember, have more than the median number of skill points out of councilmembers in the highest appeals level of that circuit (e.g. the same condition to be on a higher appeals court, except that here there aren't any). in addition to their other confirmation requirements, they must be 'confirmed' by a court (starting at a low level court and working their way up via appeal if they lose). a censor court looks at their past juror history and confirms that they truly participated, just like for political candidates, and that they maintain decorum. another court, in the subject area the judge is being appointed for, confirms that they seem to be knowledgable and correct in their application of the law.

sentencing: in a trial, if the verdict is guilty, then judge proposes a sentence and a majority of the jury must accept it.

(all) appeals courts decide by majority vote which cases to hear.

judges are supposed to only vote based on determinations of law, not determinations of fact; that is, they should vote in agreement with the non-judge-councilmember consensus, even if they personally disagree with it, unless they think it is wrong on a point of law.

perhaps every time you go up 2 appeals levels, you get 2 more judges added to the case. these judges vote and do not have to be in consensus with the non-judges for a decision; but the majority of the judges must be in consensus with the non-judges for the non-default decision to be made.

in trials, the default in innocent. for resolutions, the default is to allow them. for flags of abusive posts, the default is not to flag.

a new judge, after their nomination is confirmed by 2 courts, as above, can be appointed by an Act. they are appointed for a designated subject area. They can be removed for any cause by a unanimous - 1 action of the High Court (e.g. 4 out of 5 High Court judges must vote to remove; or 3 out of 4 if there are only 4 of them; etc; and judges cannot be removed otherwise, until their term expires; their term must be as long as the longest other term).

when an appeals case with a jury assignment comes up (or when a judge comes up, depending on which one is more busy), a judge is assigned to it based on a random seniority-weighted selection process from among the judges in that subject area; that is, there is a higher probability that an experienced judge is assigned to an appeals case, but not necessarily. if there is an overabundance of judges, the chance of a judge being assigned at all is also proportional to seniority. seniority is the sum of skill points earned by jurors in cases that the judge presides over, that is, the judge might feel pressure to serve as a facilitator helping the jury come to consensus. eh, strike that, the rich get richer dynamic is too strong. same for juries -- even if there is sufficient unused jurors to constitute an appeals court when a case comes in, a normal court is started with. this both gives novice jurors a chance to accumulate cases, as well as giving the parties a chance to first come before a uniform random distribution of volunteer jurors in their community (although, if this is a reputation franchise, should weight by reputation).

high court selection process: reweighted approval voting (binary score voting) by the electors and delegates where all judges are candidates. the process is continuous, in that electors and delegates always change their approvals, and the system always reports who would win. To actually reconstitute the High Court with the current winners, a Reconstitute the Court motion is made. It automatically contains the names of the winning judges at the time that the motion is made. It must be passed as an Act with a 2/3 threshold. (if the winning judges based on approval voting would change after the motion is made but before it is passed, the original roster is still the one that takes effect). The court is reconstituted only for cases which have not yet asked the High Court to hear them; for any cases which the High Court has already been asked to hear, the old roster decides whether to hear them and if so, decides. the new roster, however, is immediately and simultaneously available for new cases.

high court or a majority of chairs can force a higher voting threshold (eg the decision of a chair to allow a lower threshold can be appealed to the high court); the court must act with the higher threshold being considered to allow the lower threshold

change the name to just Amendable Code. Fluid Democracy is way too confusable with Liquid Democracy.

hmm, for pietrust groups, reputation should be incorporated somehow into juries, to deter attackers from infiltrating the jury system and then either blocking it or making crazy decisions when they find a bunch of themselves on the same jury (it deters because it makes this more expensive, because each attacker would have to spend the time accumulating a high reputation before they can damage the jury system). So in addition to the minimum jury skill score for appeal courts, have a minimum reputation score.

military selected/promoted by ceoi, but commanded by ceoe.

active military is forbidden from taking public political positions or from intervention, even to prevent corruption

ceoe must resign upon either a declaration of war, or upon ordering violence.

eliminate electors -- to simplify, but also to enhance the prestige of chairs.

chairs elected by multiseat reweighted score voting.

shall we allow the chairs to veto legislation? no, i think that gets them too involved. they can veto judges, though (but they can't veto more than some fraction of the available judges, perhaps?)

if you appeal from the decision of the chair, you appeal to the 'en banc' all 3 chairs, i guess. this is majoritarian (e.g. if a minority faction controls only 1 of the 3 chairs, they dont have the ability to prevent the majority from misinterpreting the rules), but so is Robert's Rules Appeal from the Decision of the Chair, so i guess that may be unavoidable.

when one high judge (out of the 3) resigns, that forces their 2 colleagues to resign at the same time, and forces an immediate replacement election after 1 week (e.g. as if the 66% majority had been achieved).

government groups also the rules about not holding multiple offices, reporting of outside income, and term limits. term limits includes not reelecting a judge. mb their judges should be 4 life.

since we got rid of electors, allow the tribunes to decide when the forum and the senate deadlocks (one passes, one vetos).

so now the high officials are:

and there's no dicrete scaling transitions.

chairs, judges, boardmembers, forum topics, and highlighted forum proposals among un-selected forum topics are all selected using multiseat proportional voting (reweighted score). CEOs election and forum voting uses single-seat score voting (forum voting also has a followup simple 2-choice yes/no vote after each proposal is selected from among the alternatives).

forum standing committees and citizen court/councils both use the mechanism of board-coloring (5 person committees) and consensus decision making. Comment-flagging in the unofficial forums uses a similar mechanism to forum proposal submission, flagging of potentially corrupt candidates, and screening of elgible candidates.

non-governmental groups have no limitation on holding multiple offices or term limits, making the scaling of the Board unnecessary, as a small group could just have 1 or 2 people holding multiple seats, and just one person as Chair.

the only set electoral term if that of the chairs. boardmembers, judges, presidents are recalled at will (by constituencies, the board and the board respectively).

the complexity of token contests is gone. the electors are gone. scaling is gone, except for scaling the delegate pyramid.

so i think (once all this gets incorporated into the bylaws), we'll have simplified things a lot. the system should be easy to scale down for small groups, too.

minority reports: on judicial decisions and committee reports, each member (assenting or dissenting) may write a 'minority report' but it can only be 1/15 the length of the main opinion/report. So e.g. the maximal summed length will be 5/15 = 1/3 of the length of the majority document, = 1/4 of the length of the total document (e.g. the majority report + the minority report). For committees, the entire committee must agree to the majority report, for the judges, a majority must agree with the majority report (e.g. if 4/5 judges vote No, then 3/5 judges (a majority of 5; the 4/5 doesn't affect this number) must consent to the majority report). If they cannot agree reasonably quickly, there is no majority report, just a decision without explanation (and the minority report length is bounded by 1/3 the length of the proposition being decided).

quorum for board is >50%. note that a minority cannot therefore block business by fleeing quorum.

note that all 5 of the board seats may not be filled

initial conditions: before judges have been selected, chairs fill in. before internal and external PMs have been selected, senior chair is internal and next chair is external. if judges or PMs are incapacitated, this is also the order of succession for interim judges/PMs.

3 judges

unanimity among high judges to remove (impeach) another judge or a chair, whether for incapacity or for bad behavior. delegates are recallable by their constituency and hence cannot be impeached by judges.

the board can and should informally discuss whatever they wish in closed session amongst themselves (with no chair present) as much as possible. however, in order to take action (to have binding votes), they must enter a formal open session run by a Chair. the formal session gives them the chance to make speeches on the record for public consumption.

anonymity

have both a transparent and an anonymous ballet, and separate scores for each; but only the transparent one counts (or, take the min of both? configurable option; only transparent, only anonymous, min of both, multiply both)

i think multiplying both together is a good default (with the default being that your transparent ballot is also your anon ballot). This way, there is an incentive to bother to fill in the anon if you have something anonymous to say, but there's also an incentive to fill in the transparent accurately, and an incentive for them to try to match, so that others can still delegate to you based on what they see in the transparent one.


another name: concord? accord, concur, cohere, harmony, concert


ooo -- fractal consensus minority reports in forum

---

todo: remove the remaining forum emergency scheduling motions -- the Board/Senate can handle things that need quick action, and the potential to veto is now automatically forced onto the schedule

draw up a new emergency procedure. how about: if the senate is killed, then all of the chairs can act together to make a temporary exception. if the senate's still there, they can already pass a temporary exception, so i see no need for anything further. if the chairs and the senate are killed, the judges can unanimously make the exception. Note that temporary exceptions cannot permit the delay of the reestablishment of whoever was killed by the usual means (e.g. cant prevent constituencies from replacing the senators; cant delay a snap election for new chairs).


on the difficulty of amending the us constitution: http://www.ivr.uzh.ch/institutsmitglieder/diggelmann/lehre/usconstitutionallaw/Ackermann_The_Living_Constitution.pdf


people with similar interests:

this guy claims to be writing a book in line with Bagehot, the Federalist Papers, etc: http://www.usconstitution.org/

http://www.ivr.uzh.ch/institutsmitglieder/diggelmann/lehre/usconstitutionallaw/Ackermann_The_Living_Constitution.pdf : Bruce Ackerman

Lawrence Lessig might be interested

surely the ongoing nomic games and the people involved with machine nomics such as PerlNomic?

http://metagovernment.org/wiki/Main_Page and similar a ton of links: http://metagovernment.org/wiki/Active_projects


IDEA FOR RECONCILING WITH C-CORPORATION:

C Corporations often have mandatory minimum quorums for their annual meeting. In Delaware the quorum may be set by the bylaws but must be at least 1/3, at least for the purposes of certain mandated types of votes. People who send in proxies and vote on anything are marked as 'present' for the whole meeting. Translating this to our domain: