User:SMcCandlish/Adminship reform

Over at m:Wikimedia Foundation Annual Plan/2015-16#Wikimedia Foundation Risks (which has a talk page) and wmf:2015-2016 Annual Plan##Wikimedia Foundation Risks (which does not), the Wikimedia Foundation included the following in the list of "advancement" risks to the entire project's goals and future:

“ Risk: Community leadership does not develop effectively, damaging community participation, governance, and responsiveness ”

Below, I've laid out why this is not a "risk" but a reality already, and what to do about it.

This risk already came to pass a long time ago

Wikipedia's adminship system is a dismal failure (on en.wp, anyway). Hardly anyone wants to be an admin any more, the community sorely abuses most applicants (unless they are near-noobs who've never ruffled a single feather, but don't really know what they're doing), experienced admins are leaving, and a good ol' boys' club of ancient admins with a wiki-aristocracy mindset are increasingly in control. Administrative backlogs are piling up, with basic maintenance (see the CfD backlog for example) lagging months behind. Well-meaning admins shy away from anything that smacks of controversy out of fears that they'll make hounding enemies that the community and the arbitration system will never really do anything about. Meanwhile a few admins who revel in playing cop do nothing useful, but spend all their time in dramaboards looking for people to block and ban. Add to this the fact that external special interests (government of Pakistan, etc., etc.) are planting "long game" CIVILPOV editors who become admins and then help gradually warp content to suit external agendas by favoring the desired side in disputes, and you have an overall recipe for very predictable, very obvious disaster.

The Arbitration Committee, while several changes in membership and attitude have eased some of the community's distrust of and scorn for it, is a legalistic morass more interested in nit-picky proceduralism, despite WP:NOT#BUREAUCRACY, and no one ever goes there unless they're desperate or stupid. Their WP:AE "enforcement" corps (they actually named it that? with a straight face?) and its "discretionary sanctions" (read: admins just punish whoever they like any way they like on no grounds other than what the topic is) is a Judge Dredd "I am the law!" farce, and almost entirely unaccountable, because neither the admin pool nor ArbCom will ever contradict what another member of "the brotherhood" does unless the abuse is so glaring they community will not let them get away with ignoring it.

The Mediation Committee is useless and toothless. Virtually no one ever uses it (other than to perpetuate a fight and refuse to abide by the mediation). It should either be shut down, or given ArbCom-level authority to require compliance (and be able to take cases that are internal as well as content-based). A guard dog with no teeth and no voice is just a pet.

The other dispute resolution systems are hidebound in their refusal to ever address anything but article content disputes, so the worst and most disruptive behavior problems on WP, not to mention the long-term "sneaky abuse" ones, like special interests trying to shift WP policy, never get addressed.

Bureaucrats (they actually named them that? with a straight face?) basically have nothing left to do, and no one cares to be one or vote for them any longer.

Wikiprojects are increasingly owned effectively by whoever is the most charismatic + tendentious, and they are increasingly trying to behave like independent sovereign entities, WP:LOCALCONSENSUS policy be damned.

What other "leadership" is there on Wikipedia? We have some people who help shape opinion by writing Signpost articles; others who seek to polish content in the Good Article nomination and Featured Article candidacy processes (the latter of which is also turning into a corrupt, walled garden); others who watchlist and shepherd our policies and guidelines, on the alert for innocuous-looking but ill-conceived alterations; people who actually go to real-world wikiconferences, editathons, and other WP/WMF focused meetings to network, plan, and get stuff done; a shrinking pool of serious, overworked admins trying to do the right thing (mostly, we hope); and a large stockpile of competent, long-term, community- and project-focused editors, most of whom would not touch adminship with a ten-foot (approx. metre-long) pole.

We're not fixing it

It's become painfully clear over the last decade that virtually no one is satisfied with the current adminship system, and that most of them even agree on what the problems are and how to fix them, in the abstract, but that nothing will be done internally by the community to fix it. Every single reform proposal of any kind is stillborn for three primary reasons:

• Good Ol' Boys' Club admins will not tolerate a threat to their elite "first class citizen" power.
• Just plain good admins sometimes seem to just feel that it is the way it is, and if they had to go through a world of shit to become an admin, everyone else should, too.
• The community at large are terrified that any change will just make things worse, no matter how bad things are now, or what obvious improvements are proposed.

The last of these is the important one, since changing it or working around it will make the other two irrelevant, because the entire existing admin pool would be vastly outvoted. A decade has proven, however, that the community fears cannot be assuaged. It's genuinely pathological: The worse the situation gets, the worse the anxiety about and resistance to reform gets.

It's already proven remarkably successful to unbundle various admin "bits", like template-editor (a bit upon which much of my best WP work depends). This points the way.

The solution: WMF mandates that the community devise an alternative system, within certain constraints, in a timely fashion

The solution is obviously for WMF to take the same approach it took with ArbCom: Require the community to come up with a reformed administration system, by a specific date (give it 1 year, no more) or be subject to one being imposed on them. Further, require that the solution have some requisite features. I'm not sure what all of them should be, but the obvious ones are the following:

• All admin permissions/authorities will be unbundled that do not require admin-level trust (some do, of course, like blocking, issuing topic bans, access to deleted pages, and other serious matters). All other bits should be able to be applied for, like template-editor; most importantly:
  • Ability to delete pages, only for maintenance purposes, e.g. deleting redirects for moves to clear out the RM backlog, deleting emptied categories to clean up after the CfD backlog, and deleting pages with expired PRODs, but not for CSD reasons, which requires admin trust level) Other uses of the bit would be limited to admins, just like closing contentious RfCs cannot be done by non-admin closure. I.e., We stop confusing access to a tool for every imaginable kind of permission to use it.
    • Update: A small part of this has already happened: unbundling of the page mover bit in 2016.
  • Whatever else it takes for any competent user in good standing to be able to perform work that needs to be done and isn't debatable. If it's subject to a backlog because the matter has already been decided it should never require an admin to do it.
• Adminship will be based on competence and good standing, not popularity, nor subjective considerations. Essentially, the same reasoning that is applied to whether an editors' input is constructive in mainspace will be applied in spirit to whether they'll be permitted access to admin tools.
• Because competence requires the ability to function well in a collaborative system (both in a "don't be a jerk" sense and in the WP:ENC vs. WP:NOTHERE sense, behavioral concerns are valid. However, the entire culture of RfA will be rebooted, on a WP:ASPERSIONS basis: Any claims of wrongdoing must be backed by diffs, that a reasonable person would find actually evidentiary, or the accuser may be sanctioned.
• Determination of success by Bureaucrats will be less tied to the head-count vote, and determined primarily the same way we judge all other consensuses: By weighing the actual merits of the arguments provided, and especially discounting poorly-supported venom spewed by individual who have a personality-based dispute with the candidate.
• Candidates may request (and be held to) access to only specific tools, or the community permission to some of them only for certain purposes, on a probationary basis. A large part of the current problem is that all tools are bundled together in a carte blanche what that the community feels requires an excessive amount of trust. The system must permit anyone capable of being a good admin in certain areas or ways to become one.
• The goal should be that most regularly active editors with around 18 months under their belts and no obvious mental problems should practically automatically become admins, and that great participants can do so sooner with more ease. The model would be that of the free software community: If your code is good, and you know what you're doing, you get commit permission.

What we'd gain

The results of this reform would be:

• No more admin backlogs.
• No ability of "cabals" of admins, either 2007 ones who feel "vested", or social-engineered ones set up by outside interests secretively, or even internal PoV-pushing groups engaged in jackassery, to ever strongly affect outcomes.
• No more fear by admins who knew a decision was wrong to contradict another admin, because adminship would no longer be special, precious, or threatening again.
• Unbundling more tools and contextually limited remit to use them provides training of potential full admins and a proving ground to judge their performance that the community can evaluate.
• No more rancorous divide between admins on high and lowly editors.
• Ability of admins to remain more active content editors, without any pressure to always just do admin tasks (also a reduction in "career admins" (one of several administration-related cancers now affecting the system).
• Active and sensible community members who'd rather eat their own feet that be drawn into admin drama cesspools could become admins and, by specifically disavowing things like block and ban ability, make it impossible that anyone can draw them into such stuff, and be confident that they can do only the admin work that they feel is needed and which they'd be comfortable with. (I'm in this camp.)
• An end to the "civility exists everywhere on WP except at RfA" shitstorm problem.
• Long-term administrative stability.
• No more "dramaboard" effect of too many noticeboards (especially AE and ANI) being dominated by admins itching to deal out heavy-handed sanctions (i.e. "please quit being a Wikipedian" offers that are hard to refuse).
• Enough admins around that disruption by COIs and the like can be reigned in more quickly, and that recourse to noticeboards much less ArbCom is less likely.
• Because it will obviously work better than the current system, years of community anxiety about the future will melt away.
• Abuse of administrative tools will be dealt with more promptly and with less favoritism toward never questioning an admin, because we'll have so many of them, and getting the tools back per a "WP:ADMIN-STANDARDOFFER" will be so routine, that ArbCom and AN will drop the pretense that it's better to keep a terrible admin than lose any admins, and that an admin should not be sanctioned unless what they did is so beyond the pale it's astounding.

If WMF doesn't do something like this, en.wikipedia is headed for a lot of trouble.  — SMcCandlish ☺ ☏ ¢ ≽^ʌⱷ҅_ᴥⱷ^ʌ≼  10:07, 25 January 2016 (UTC) (originally posted at Meta)