Wikipedia talk:Arbitration policy proposed updating

Introduction

This page represents a proposed discussion draft for community revision and updating of the Wikipedia:Arbitration policy. In addition to a general updating of the policy in light of experience in the four years since it was drafted, I have incorporated some of the input from the RfC on the Arbitration Committee that took place over this past summer. Issues relating to the selection of arbitrators (in particular, length of term) are discussed on Wikipedia:Arbitration Committee but can also be discussed on this talkpage for consideration of a modification of that page.

In order that all editors can work off the same discussion draft, I suggest that for a few days, editors refrain from editing the proposed policy, except for correcting any typographical or formal errors. Once we have a consensus on whether the discussion draft provides a basis for updating the policy, and what areas require further work, the revised policy can then be opened for consensus editing under the ordinary wiki model. In the interim, I suggest that comments be placed on this page and arranged by topic so that we can see whether we have a consensus or not for any change.

I will add that although I mentioned that I would be undertaking this project to my colleagues on the Arbitration Committee, and some of the ideas I would be raising, the proposed revision is my own work product. No claim is made that other arbitrators support it or any of it, and I expect to see sitting and former arbitrators, along with lots of other editors, here to comment on various aspects. Newyorkbrad (talk) 19:36, 11 October 2008 (UTC)

At first glance, I like this, will spend cycles digesting it. — Rlevse • Talk • 19:51, 11 October 2008 (UTC)

I've now read it three times, and I like what I am seeing overall. I believe this is a workable and reasonable proposal that takes into account the majority of relevant issues raised separately. Risker (talk) 23:15, 11 October 2008 (UTC)

Comments concerning specific drafting and policy issues

Issue 1: Case opening criteria and calculation of majority

I agree with the change to the rule on how cases are deemed ripe for opening; I had suggested substantially the same change some time ago (to little feedback).

I still think that my second change is worth consideration; specifically, that arbitrators are not considered active on a case unless and until they either positively state they are or act on the case. The presumption that all arbs not otherwise specifically recused or marked inactive are, in fact, actively monitoring and considering the proposed decisions has been shown to be inaccurate in practice; and cases have often languished without a decision for lack of arbitrator participation, making attaining the requisite absolute majority of sitting arbs difficult or impossible.

It seems perverse that if the only, say, six arbitrators that have actively participated on a case unanimously agree on a proposed decision that it would then been scuttled because of lack of interest/participation by the other arbitrators (whatever the reason(s) might be). — Coren ^(talk) 21:20, 11 October 2008 (UTC)

Although the first change you reference (from "net four" back to "four accepts") may not have received a lot of feedback when you raised it, rest assured that it has been discussed in great detail a number of times. I believe can probably find a number of the discussions, if you are interested, in the archives of Wikipedia talk:Requests for arbitration.

Your second proposed change has been discussed from time to time as well. In my experience, there have been relatively few cases in which arbitrator inactivity has wound up ultimately affecting the result. More common is that cases sit around well after they should have been closed, waiting for the last few arbitrators to straggle in. But the latecomers often do contribute valuable comments or additional proposals when they arrive, I'd rather see more active and prompt participation brought about by the presumptive time limits, rather than making it easier to count the additional arbitrators out of the cases altogether. Newyorkbrad (talk) 22:07, 11 October 2008 (UTC)

The problem is that, as currently worded, the proposed policy doesn't create actual deadlines so much as a statement of intent to keep the cases under certain timelines — not that I would support strict deadlines either, mind you, since that leaves little room for extending the deliberations when the case is genuinely complex or difficult to reach consensus on.

Is it really that much of an imposition to require that, for an arbitrator to be counted when calculating majorities, they need to have made one positive act at some point in the case? (Whether by opining on the case opening, voting on one of the measures or simply stating somewhere that "I am active on this case")?

I don't want to sound callous, but frankly if an arbitrator can't be bothered to write a one-liner at some point during the case, the presumption that they are not paying attention to it is certainly warranted. — Coren ^(talk) 22:44, 11 October 2008 (UTC)

Issue 2: Jimbo-related policies

Prefacing with absolute wonderful work, Brad. This took lots of time and I think you did a great job.

The one tiny bone I'd pick-- if I ran the circus, phrases like "may be appealed to Jimbo Wales" would be replaced with "may be appealed to the individual delegated by the Wikimedia Foundation Board to handle such appeals, Wikimedia Chairman Emeritus Jimbo Wales."

It's a minor thing, but you don't have to go far to hear accusations of dictatorship at Wikipedia. My own experience is that these accusations are utterly baseless. No such dictator exists, Jimbo has exercised his authority with total responsibility.

But, as long as we're getting sitting down to write one article in the constitution of Wikipedia, we might pause for just a second to put our best foot forward.

In the current draft, it's unclear why arbcom appeals go to Jimbo. A reader could ask "Why Jimbo? Why him? Why not somebody else???"

The answer is, of course, that he's the person we all trust-- he is THE most trusted neutral broker the project has. It's not just that he was a founder-- it's that time and again he has earned the trust of those around him.

It would cost us nothing to spell this out by explicitly-- by mentioning that appeals go to the person who is the most trusted member of Wikipedia-- the person who has been entrusted by the board (and the community!) to handle such appeals.

It's pure semantics, but semantics do matter. People like leaders who lead by example and have earned respect and trust. People despise dictators who have the final say for no apparent reason. Let us make it clear at every opportunity what we in the community already know-- Jimbo leads with earned respect and trust.

When weighing how much participation dedicate to the project, when deciding how much money to donate to the Wikimedia Foundation, or when writing up a news story that will influence the opinions of the public-- perceptions and semantics DO matter.

If the allegations were true-- if Jimbo were a secret tyrant who clings to power at all costs, well, we'd just have to hope for the best, try and sweep the issue under the rug. But that's not the case. But the allegation are NOT true in the slightest. Jimbo has not dictated his role, he has had his role thrust upon him by virtue of the trust he as earned.

So, let's take half a sentence clause, here and there, to state the truth that makes reflects positively on the project and the foundation-- that Jimbo has been rightfully entrusted with his role by his colleagues, and that he has not simply dictated his role (which, as former owner, he certainly once could have done). --Alecmconroy (talk) 23:30, 11 October 2008 (UTC)

Thanks for the kind words. As for the Jimbo Wales issue, I understand your semantic point completely. But as a historical matter, it really isn't true that Jimbo is "the individual delegated by the Wikimedia Foundation Board to handle such appeals." To the best of my knowledge, the Board has made no such delegation nor has anyone asked them to (and in fact, the Board would be very unlikely to take an action of this type that would affect just one project). As I understand it, Jimbo's special role here, whatever it is, is a function of history and community acceptance rather than of Board action. Newyorkbrad (talk) 23:35, 11 October 2008 (UTC)

I stand corrected-- I realized the board had never explicitly said so, but I assumed that was their position and that were Jimbo to overrule Arbcom, he would do so on behalf of the board.

Though the answer isn't as simple as I like, I think it would be excellent if we could find some sort of explanatory language to give the reader an answer the question of why Arbcom rulings can be appealed to one particular individual. An answer beyond "That's sorta just how it's always been."

So why CAN arbcom decisions be appealed to Jimbo? Is it because the board or the bylaws says so? Is it because community consensus says so? Or to put a finer point on it-- if the board wanted to end Jimbo's ability to overrule arbcom, could they? If the community wanted to end Jimbo's ability to overrule arbcom, could they?

It doesn't much matter to me which the answer is, but I think somehow we should find an answer and write it down. The important thing is to debunk the smear I see, from time to time, that Jimbo simply "owns" Wikipedia and can do whatever he wants just because he started it. That's not the way Jimbo himself wants it, that's not the way EnWikipedia works, that's not the way the Wikimedia Foundation runs things. The further we can get from that urban legend, the better. --Alecmconroy (talk) 23:57, 11 October 2008 (UTC)

The genesis of Jimbo's role is that he founded wikipedia and gave himself final authority. The continuation of Jimbo's role is that the community has at no point reached a consensus that it wants to change this traditional status-quo. But it is entirely a matter for the English wikipedia project and its community, not for the board. If you can get a consensus to have someone or some process other than Jimbo as a "final authority" then that's what would happen.--Scott MacDonald (talk) 00:03, 12 October 2008 (UTC)

So I ask the audience: "Is Scott right?". I hope so-- I think that's a great answer-- "delegated by community consensus to handle appeals" is even less 'dictatorial' than "delegated by the board to handle appeals", and encoding that in policy would go along way towards killing the "jimbo is an evil dictator" meme. --Alecmconroy (talk) 00:28, 12 October 2008 (UTC)

I didn't say it was delegated by the community. That would imply the community had a power and decided to give it to Jimbo. That's not what happened. Jimbo had the power from the beginning, and the community has never changed that status-quo (which, in theory, it could by consensus).--Scott MacDonald (talk) 00:36, 12 October 2008 (UTC)

Granted-- I'm making a minor leap of logic there to assume that the community has made such a delegation through consensus just by it's inaction and overall endorsement of the status quo. I don't think there's any doubt that Jimbo's appeal power does have consensus-- but if we wanted to be sticklers, we could always make a proposal to ratify that power so there's no doubt whatsoever it has consensus-- such a proposal would probably pass without a singe opposition !vote. If we all agree that Jimbo's arbcom appeal power is community-determined, I'd just as soon ignore-all-rules and declare it to have consensus, rather than go through the extra step of asking the community a question we already know the answer to. --Alecmconroy (talk) 00:50, 12 October 2008 (UTC)

No, I think you'd get lots of objections. However, again that's the wrong way round, it does not need consensus. We don't ratify the status-quo; rather it is anyone looking to change it need to get the consensus. There are many parts of wiki-process that wouldn't pass "ratification of consensus" but they remain until some alternative does pass it.--Scott MacDonald (talk) 00:55, 12 October 2008 (UTC)

Well, this complexity underscores how important it is for us to figure out what the deal with Jimbo's arbcom-appeal power is and enshrine it in policy. There are a lot of anti-wikipedia pages (which I shan't link to) which put forward the jimbo-as-dictator meme. Since we're all here writing a policy which mentions Jimbo's arbcom appeal power, now would be a great opportunity for us to cut that smear off at the knees. --Alecmconroy (talk) 01:01, 12 October 2008 (UTC)

Em? You'll never stop such sites. We don't need to "figure out what do to with Jimbo's appeal power" - it exists. It will continue to exist unless someone gets consensus to do something else.--Scott MacDonald (talk) 01:15, 12 October 2008 (UTC)

By "figure out what to do with Jimbo's appeal power", I just mean-- get it written down somewhere what you said earlier-- that "yes, community consensus could decide whether to grantJimbo that power or not."

The alternative view is that "No, Jimbo's appeal power comes from the nature of the foundation, the nature of the project, or the decision of the board, and is not up for alteration by community policy".

I've heard both views expressed on-wiki at one point or the other-- I don't know what the answer is-- but WHATEVER the answer is, this is a great place to write it down, so that people will know in the future. --Alecmconroy (talk) 05:26, 12 October 2008 (UTC)

Should we propose a JIMBO policy?

As best I can tell, Jimbo has two "special powers" on EnWP that he alone possesses-- he can override Arbcom decisions and he can veto Arbcom candidates. The current draft does not specify why this is so. I had thought these were granted by the foundation board, but apparently not. I had also thought it might be by community consensus, but apparently the community never actually enacted such a policy, so perhaps we can't say it's due to community consensus?

What would people think about just proposing a policy that says should have the powers he has now: Arbcom Appeal and Arbcom Candidate Veto? It would certainly reach consensus, and the matter would be settled.

Or is such a step unnecessary? Perhaps it's just abundantly evident that yes, Jimbo's powers do in fact, have community consensus, and we can just state it as such without hopping through the hoop of a formal ratification? --Alecmconroy (talk) 10:18, 12 October 2008 (UTC)

(and I know, yes, it's kind of a minor point, but as long as we're updating the Arbcom Policy page, let's make it absolutely as good as possible). --Alecmconroy (talk) 10:18, 12 October 2008 (UTC)

Ah, what you're proposing is a codification of reserve powers! This fits in excellently with this analysis of mine way back from March last year, which posits that the governance model of the English Wikipedia (the other Wikimedia projects are largely in a different position) is essentially that of a constitutional monarchy. --bainer (talk) 08:35, 15 October 2008 (UTC)

Issue 3: Off wiki-deliberation

The arbitrators may discuss and deliberate on cases on their private mailing list, particularly where warranted by privacy or similar factors, but the substantive basis for the final decision should be apparent from the decision itself or from arbitrators' comments on it.

This seems to mean that arbs don't have to discuss on-wiki at all, as long as final decision is clear on-wiki. Sometimes it can seem like there are two discussions, one on the workshop without much arb involvement and one behind closed doors - where arbs try to get agreement before proposing decisions. I would have thought that 1) arbs should engage with community discussion on the workshop where they can and 2) they should show their working (including disagreements) on the proposed decision page. That is DO NOT initially make proposals on a mailing list, make them on wiki!!! The proposed decision page is the arbcom's workshop, not merely a place to disclose what's been thrashed out in secret. (I am here speaking of cases where no privacy is involved.)

I'd suggest changing this to:

Where warranted by privacy or similar factors, the arbitrators may discuss and deliberate on cases on their private mailing list. Beyond that, where practical, deliberation and discussion by arbitrators should mainly be either on the workshop, or on proposed decision page.

Even that is an encouragement to on-wiki discussion rather than an imperative.--Scott MacDonald (talk) 23:34, 11 October 2008 (UTC)

I see no problem with them discussing the issue in secret. I believe that as long as the final decision is made on wiki, and can be veirfied by looking at the history of the page, their internal discussions should be allowed to take place in secret. עוד מישהו Od Mishehu 09:18, 12 October 2008 (UTC)

I really don't like the way arbcom discusses things in secret either. I would much prefer to see them wrangle it over than just one day up and announce the decision. I never really know what led to what, and the reasonings that are posted in contentious cases usually don't actually explain to me how it came to pass that arbcom came to the decision it did.

The "findings" and "remedies" usually read more like a verdict than a reasoning. They summarize the principles, the facts, and the remedies, but in complex cases, I generally am left wondering who 'really' thought what when, what their reasoning was, what comments or evidence they considered critical or influential. Did anyone change their mind? and if so, what made them change it?

When that happens, I'm always left wondering about my own participation in a case. Was I helpful? Were my comments just distractions? Was I counterproductive? For that matter, I wonder which of the arbs even read the evidence, the statements, or the commentary. Given the volume, I can hardly imagine all them read everything, and I certainly wouldn't hold it against them that the couldn't get through it all.

But the deliberations-in-private make it hard to know how best to interface with Arbcom. Certainly the decisions they post have all the required justifications and conclusions, but for any case where arbcom seemed divided the posted decision always leaves you feeling a bit bewildered, no matter what the decision is. It's as if someone took the entire script to 12 Angry Men and threw out all the dialogue, all the details, and just gives you as a synopsis the phrase "Not Guilty". You can't help but feel you're missing out on the 'real story'.

BUT-- having said all that, I don't know how we can reasonably create any sort of policy that limits deliberations in private. The best we can do is say "Please be aware that any time you use the mailing list to deliberate instead of the workshops, you are setting the community up to not understand how this decision came to be. Beyond that, use your best judgment."

I guess what I'm saying is-- I know I find the private-deliberations disorienting, but I don't honestly presume to know which style is really best for the project, and I think the best we can hope for is to tell arbcom how private deliberations affect us, and leave it up to them to do the balancing. --Alecmconroy (talk) 10:36, 12 October 2008 (UTC)

My impression is that very little substantive disagreement is "hidden", and certainly not with the intent of presenting a united front. Sam Korn ^(smoddy) 10:42, 12 October 2008 (UTC)

Well, I don't mean to say "hidden" with the sense of "mysterious, creepy, cabalistic" or anything like that. But just as a for-example the most recently closed case RFAr Everybody and their brother. For three months, it seemed impossible to ascertain the status of the committee-- were they deadlocked? busy? just widely inactive?

At some point, NYB came back and we had a proposed decision which was, basically, unanimously supported. We then had an additional remedy proposed, which also was unanimously supported!

So, what went on for four months? Were there deliberations? A deadlock that got resolved by some critical piece of evidence or some novel argument? Did people change their minds, did people make up their minds, and do some of the arbs disagree with the decision but decided it was best to present a united front? Who knows. There is no knowing.

And maybe we don't need to know. We almost certainly don't need to know now, after the case has closed, so nobody take this as an invitation to dig up hatchets best left buried. It's just a matter that the whole process is somewhat bewildering, and while they did a fine job of posting a valid reasoning, we don't 'really' know how the decision came to pass.

That's okay-- it's just something to be considered. The upside of private deliberations is that it glosses over intra-arbcom disputes and lends credibility to the final decision. The downside is that it makes it harder for the community to understand how arbcom works and how best to help arbcom do its job. I leave it to wiser minds than mine to figure what is best overall. --Alecmconroy (talk) 11:08, 12 October 2008 (UTC)

It's an interesting point. My take on it is that the arbitrators are generally uninterested in playing to the gallery -- proposing solutions that they do not believe can pass even though they may personally endorse them: their object is to present a proposal that is both effective and can get support enough to be passed. This may give the impression of dissent being forbidden, but I would rather attribute it to a mature and sensible decision-forming process.

The duration of the case you cite is, I think, a result of many members of the Committee being quite frankly stumped -- not knowing how to structure a decision that was a) useful, b) fair and c) relevant. I find it hard to blame them for that in this instance, though I strongly recognise that the delay in the case was unsatisfactory.

In short, I attribute the unanimity common to AC decisions a result of the members of the Committee working together to deal with other members' concerns, rather than anything more suspicious.

Sam Korn ^(smoddy) 11:26, 12 October 2008 (UTC)

No, definitely not suspicious, and almost certainly mature, and maybe even ideal.  :) Just confusing. With arbcom, I think people throw anything they can think of at the committee because they have no idea what will stick. --Alecmconroy (talk) 12:17, 12 October 2008 (UTC)

From the period when I was on the Committee, I know that it was helpful to informally discuss proposals before formally proposing them. This means small issues can be ironed out without having to change already-voted on proposals (non-transparent) or add multiple options (confusing not least in terms of knowing which proposal has passed). While it would be good if this discussion was able to take place on the Workshop or in some other public forum, I think it is a matter of practice rather than of policy. It would be good if there could be some kind of commitment from the Committee to be as transparent as possible. This seems a more practicable solution than enshrining it in policy, only to be ignored. Sam Korn ^(smoddy) 10:42, 12 October 2008 (UTC)

I think it's disengenous to talk about the committee's discussions being "transparent" and "public" as if they were the same thing. Transparency is absolutely crucial for the functioning of the process: we have to know who supports what and have confidence that their decisions are free and fair. That doesn't necessarily require that all the background discussion has to be public. As long as we can safely assume that we know what is being discussed on the mailing list, I don't think we need to be too concerned over its exact content. As long as I know whether I'm being talked about, I'm not going to lose sleep over who's saying what. I'll know what conclusion they come to in due course. Happy‑melon 16:57, 13 October 2008 (UTC)

Issue 4: Scope, evidence, private communications

Nothing here addresses the scope of cases. Because no formal charges are brought, an editor may not realize that some aspect of his or her actions are being scrutinized and that rebuttal evidence is needed. By the time that the decision is drafted and charges are laid out, it is essentially too late. It would be helpful to parties if the ArbCom could at some point or points explain which issues they are addressing. It might be at the beginning, when a case is accepted, and/or midway, after initial evidence is presented and reviewed but while there is still time to add more evidence.

In a related manner, this policy allows for the private submission of evidence but it doesn't mention sharing that evidence with involved parties. So a party could have evidence presented against him or her but not even be aware of the accusations much less have an opportunity to rebut them. While some evidence may be so confidential that it can't be shared in its entirety, the confidential parts could be redacted or the evidence summarized. Communications with the ArbCom are a topic that seems to generate frustration - parties don't know who to email, and they may not receive replies or even acknowledgements. Would it make sense to designate an ArbCom member or two as the point of contact for emails?

Lastly, an unrelated item - the 7-day periods appear to be illusory. To make editors hurry to gather evidence within a week, and then to have the ArbCom spend months deciding the case is unfair to volunteer editors. If the ArbCom cannot follow schedules like this it's better not to raise expectations. I don't see anything in this proposal which will help the ArbCom handle cases more expeditiously than it has in the past. ·:· Will Beback ·:· 23:43, 11 October 2008 (UTC)

Issue 5: (case duration)

I realize the WP:RFAR/Everybody and their dog case went on for way too long, but that doesn't mean you have to over-correct by imposing firm time limits (either 2 or 3 weeks depending on how I read this). A month and a half is fine, so is two months, as long as there's steady progress being made. Really, it depends on the case. The number of people involved will in most cases exponentially increase the time needed. — CharlotteWebb 02:20, 12 October 2008 (UTC)

Agree in spirit. The adoption of a tight timeline may lead to it being broken quickly, and then completely ignored. I still like a guideline - but let's be generous, and then really expect to live within it. The draft has 7 days for evidence, 7 until a proposed decision is posted, and 7 until closure. These are modified by "Except in unusual circumstances" and "their best efforts." How about 15-15-15? or 20-20-20? (or 20-20-7?) Jd2718 (talk) 12:58, 12 October 2008 (UTC)

Or how about "as much time is it takes, but no longer"... — CharlotteWebb 13:08, 12 October 2008 (UTC)

Issue 6: Formal correction

Arbitration policy is WP:OWNed by Jimbo, and (possibly and to a lesser extent), by the Committee itself. While the proposed policy wording itself isn't affected, the introduction is.

The community only proposes matters that relate to the Committee and the Arbitration process. For example, in the Arbcom elections, the community votes and Jimbo decides who to appoint based upon that vote. Arbcom itself never was at any stage, other than Jimbo's ownership as project leader. The most accurate description of policy change is probably that it would be some kind of tripartite decision between Jimbo, the community, and the then-committee. Probably in realistic terms it would end up like the Arbcom elections, ie communally discussed and Jimbo final say.

Formal correction:

OLD - "by consensus of the community."

NEW - "by consensus of the community, as endorsed by Jimbo Wales and the present Arbitration Committee."

I'm putting together other lesser and informal comments (including hopefully improvement to readability), but that's a formal correction. Jimbo needs to clarify the wording here.

FT2 ^{(Talk | email)} 12:27, 14 October 2008 (UTC)

Issue 7: Review, and refinement

I have gone through this proposal, which is broadly speaking, good. I found what seem to me like omissions, and also some areas where past experience suggests that minor clarification may help. I have also organized it into sections for readability, to break up the long list of bullet points.

The version posted is very close to the version discussed here, and I have highlighted the edits in bold (with explanation via footnote) to make it easier to review. Any or all of these changes may be important in the final version. Although I am usually wordy, mostly these are very short. Quick summary:

Omissions/additions

1. Membership and selection - should be described by policy: Arbitrators appointed by Jimbo following advisory elections held in a manner determined by the community. Current process for interim appointments is briefly stated for completeness but this will need updating when a more satisfactory decision on this is reached.
2. Arbitrator conduct - When I drafted WP:ADMIN policy it included as a new section the communal standards we had for admin conduct. I cannot accept an Arbitration policy that does not make clear high conduct standards are expected of Arbitrators, too.
3. Inactivity and resignation. (Also covers participation if inactive, long term inactivity, and return after resignation.)
4. Explanation of breached time limits - If time limits are broken or a case drags on, the community should have an update no less often than every 7 days, of where it's at. Will help the community, and will focus the minds of the Committee on the passage of time.
5. Past decisions - The old policy stated that past decisions were useful but not binding. This seems to have been omitted. Re-added in a better form.
6. Interim decisions pending communal consensus, in truly exceptional cases. Broadly speaking, for dispute resolution purposes, if there is a major dispute between admins on some important policy issue, that will not (yet) go away, it's sometimes going to be better if Arbcom states an interim solution until communal consensus is reached, than that we have an "admin war" over it.
7. Policy change process - Arbitration Policy should say how it can be amended, and also say where Jimbo fits in, and why.

Significant clarifications and "problem fixing"

1. Fix an inconsistency between "Responsibilities" and "Scope", by amending the responsibilities to start To act as last resort (other than Jimbo Wales) for Wikipedia disputes....
2. Clarify "content dispute" (perennially recurring issue).
3. Fix a point where proposal is incorrect according to mediation policy, and propose a reasonable solution.
4. Off-wiki may be needed for serious real-world harassment cases, as well as privacy and urgancy. While this could be considered part of "privacy" it's worth saying since "privacy" could be taken to just mean cases involving WMF private data.
5. Redraft of "personal writings/emails as evidence" - the original had serious issues when looked at. Simplest solution is to not try and be prescriptive, but to state that text by communicating parties and quoted people need consent to be openly posted, and if there is a question ask an arbitrator what they would consider reasonable to post (including redaction etc). Drawing a "bright line" on what is and isn't needed to be redacted or summarized is never going to work; this is far easier and also much less gameable.
6. Redraft of "recusal" to explain the principles here, to require higher standards of self-scrutiny, and to foreground that their ability to reach a decision neutrally and fairly is paramount.

Points outstanding (non crucial)

1. "The Committee reserves the right to experiment with new forms of page structure for RFAR cases, consistent with the present aims, concepts and approaches."
2. Matters flagged for discussion - (a) may move over time to a new process for alternate Arbitrators, (b) may wish to review case acceptance voting criteria.

Link.

FT2 ^{(Talk | email)} 10:56, 16 October 2008 (UTC)

Comments concerning non-policy issue(s)

(S)electing arbitrators

It is not clear if most issues relating to ArbCom properly belong to the Committee, to the Community, or to Wales. Ironically, the group of items that it is clearest do not belong to the committee, that is election frequency, eligibility, term length, etc, that group of issues is absent. Where does that discussion belong? Jd2718 (talk) 20:12, 11 October 2008 (UTC)

Per my comments at the top of this page, on a revision of Wikipedia:Arbitration Committee. I'll hopefully get work up a proposed discussion draft for revising that page, as a companion to this one, in a day or two. Newyorkbrad (talk) 20:14, 11 October 2008 (UTC)

Thanks, I missed it. And for this draft... which demands a close read. Jd2718 (talk) 20:21, 11 October 2008 (UTC)

Term length

Actually, as I think about it, it would probably be better off to have the term-length discussion here. My own view is that there is no need at all to change the current terms or selection process. However, a reasonable alternative would be to go to two-year rather than three-year terms (meaning tranches of seven or eight arbitrators). I could not support any reduction of the term length below two years, as it takes time for some of the arbitrators to get up to speed, and I could not support holding ArbCom elections more often than once a year given the amount of community time and attention they divert from other work. Newyorkbrad (talk) 20:31, 11 October 2008 (UTC)

It is important that you mention the time to get up to speed; the opposite poll of discussion is the awful attrition rate. If I read the chart correctly, only three arbitrators have completed 3 year terms (The Epopt, Raul654 and Neutrality, though three more are only 12 weeks away Charles Matthews, Morven, and James F.. (Raul and James F. remarkably have over three years each.) The tension between learning curve and attrition pushes me to think that "no less than two years, no more than 3 years" contains as close to the 'right' answer as we will get. Jd2718 (talk) 20:43, 11 October 2008 (UTC)

Well, arbitrator certainly is one of the more draining assignments we have, there is no doubt about that, as any sitting or former arbitrator can attest from experience. What we really need is 15 (or however many) fully participating arbitrators to spread the workload around, but for a variety of real-world and on-wiki reasons that can never be guaranteed. Newyorkbrad (talk) 20:47, 11 October 2008 (UTC)

I suspect Sam Korn nailed it when predicting two-year terms, yearly elections. I've heard a lot of support for 18 month terms, but the election schedules get hectic and the sittings arbs seem pretty unanimous in that 18 months is just too short.

In addition to election schedule and term duration, the other big question is how many seats. Aside from noting that it's widely held we need more arbs in order to get more active arbs, I haven't the foggiest clue how to even begin to suggest what that number should be. I start by assuming how many active arbs we need, the multiply that by the product of term length and arb attrition rate, divide by a constant, then get a headache.

As such, I'd urge NYB to be bold and just pick a number of seats, based on his own experience and judgment, and shout it out. (Of course, this is probably what he was already gonna do before I and my longwinded comments suggested he do it-- but, you can never have enough 'Huzzahs of Confidence' urging you to be bold, so, I'll add one more.) --Alecmconroy (talk) 00:14, 12 October 2008 (UTC)

The current number is probably about right. With more arbitrators, we would never come to a consensus or majority decision on anything (unless we started acting by panels or subcommittees, but there's no reason to consider doing that given the current low caseload). With fewer, there wouldn't be enough people to do the work given that a few arbs are always inactive or busy at any given time. Somewhere between 12 and 15 is probably optimal and we are at 15 now so I don't see a need for any change. Newyorkbrad (talk) 00:16, 12 October 2008 (UTC)

Well, no. There are currently 13 committee members, two of whom are inactive. Having said that, I agree that 15 sounds about right, and wouldn't recommend a change in that number. I would suggest 2 year terms, two uneven tranches of 7 and 8 members. Risker (talk) 04:13, 15 October 2008 (UTC)

Groovy by me. I had interpreted your earlier comment to suggest that you thought the total seats should be inflated to try to guarantee a solid 15 active arbs. hehehe but when looking over the discussion to see who brought up adding more arbs, I guess in fact _I_ did-- so... just ignore that. :) --Alecmconroy (talk) 00:22, 12 October 2008 (UTC)

Election frequency

Copied from Newyorkbrad's comments, at the top of the section: "I could not support any reduction of the term length below two years, as it takes time for some of the arbitrators to get up to speed, and I could not support holding ArbCom elections more often than once a year given the amount of community time and attention they divert from other work."

I am concerned about the frequency of resignation (or going completely inactive). In the last two elections (the only ones I have participated in) there was a gap between the approval voting for the top few candidates, and the next few - meaning that allowing Jimbo to choose from further down the list means appointing arbitrators who do not possess the same community consensus approval that say you or Kirill or Flcelloguy or Paul, or, or, or enjoyed. Also, one year is a long time to wait to fill in a missing slot; by-elections should be a complete non-starter. Frequent elections would be a problem, but so are empty seats. We should consider slightly more frequent elections. Jd2718 (talk) 16:24, 12 October 2008 (UTC)

We should at least designate backup arbitrators before a membership crisis arises. — CharlotteWebb 17:03, 12 October 2008 (UTC)

Agreed. Selecting alternates at election time is a wonderful idea. --Alecmconroy (talk) 12:03, 14 October 2008 (UTC)

Disagree about "selecting" alternates. I do agree that there should be a process for filling a seat that comes open within the first 6-8 months of a year; however, after that time the learning curve is too steep for a "temporary" member to get up to speed. It strikes me that only those who have stood as candidates in the most recent Arbitration Committee election (thus having publicly taken positions on questions posited by members of the community) and who achieved a minimum of 51% approval, should be eligible. Past members whose terms have not yet expired would be an acceptable alternative, if those individuals could be persuaded to return. Risker (talk) 04:35, 15 October 2008 (UTC)

51% seems like an awfully low threshhold. Jd2718 (talk) 05:58, 16 October 2008 (UTC)

Better to do the picking before hand, rather than after hand. We shouldn't need that many picks anyway-- how many arbs have we ever lost before in a single year?

Another possibility would be to have the alternates, like alternate jurors, view all the same material as the regular members, but just not participate in deliberations or voting. That way they could, through the mailing list and other sources, begin to get up to speed in case they are needed. I don't know if I'd actually endorse that idea myself, but it occurred to me so I thought I'd shout it out. --Alecmconroy (talk) 06:10, 16 October 2008 (UTC)