User:Herostratus/Administrator reconfirmation

TL;DR SUMMARY: This is a (complicated and difficult) procedure for requiring an admin to undergo a reconfirmation RfA (or recall RfA if you prefer) even if that admin has not agreed to to be subject to recall (per WP:RECALL).

Everything in this procedure is intended and designed to be within both the letter and spirit of Wikipedia policy, and to be congruent with the best interests and guiding principles of the Wikipedia. If you think that the material could be improved to better meet this intent, please do so, or leave a note on the talk page.

Everything in this procedure is intended and designed to produce a result with a minimum of drama. By "drama" I mean edit warring, out-of-process deletions, blocks or threats of blocks, hurt feelings, improper use of venues, appeals for smiting, and that sort of thing. Extensive discussion is not "drama", it is normal for the Wikipedia.

For a real-world analogy, see National Popular Vote Interstate Compact. Here you also have a situation where 1) many people believe that a certain situation is non-optimal, but 2) the dead hand of the past makes it impossible to change (it's very difficult to amend the Constitution; just as with changing Wikipedia policies, it requires a huge supermajority etc.). Therefore a clever (and legal) work-around is devised. If you think the National Popular Vote Interstate Compact and devices like that are corrupt, wrongheaded, illegal, or sneaky, you probably won't like the idea on this page so you can maybe stop reading now.

Improvements to streamline the procedure or clarify its description are welcome, either through editing this page or suggestions on the talk page.

In my opinion, admins ought to all make themselves available for recall. However, some don't. In cases where the admin hasn't done so, and a reasonable case can be made that his continued status as an admin should be subject to scrutiny, the following procedure is suggested.

Everything suggested here is perfectly within policy, but the first time this is done, it may be contentious. That's too bad, but you can't make an omelette without breaking eggs, and it will just be for the first time. Once the principle of reconfirmation is established (if it is), opposition will turn to more stately methods, and there will be plenty of volunteers to smooth out the rough patches.

So here's what you do.

Before engaging this procedure

Before getting to this this point, you have presumably worked through other steps to resolve the issue, if it's the sort of thing that can be resolved (rather than a general pattern of abuse). You might consider asking the admin to voluntarily submit to Wikipedia:Administrator review. And as a final step you should ask the admin to voluntarily submit to a reconfirmation RfA. The following procedures are only if all this fails.

Reconfirmation petition

Create a section on your own talk page asking for six editors in good standing (and with one month's service and 500+ edits) to sign on to the following statement:

"We request that Admin X be subject to a reconfirmation RfA."

You need six signatures. That is the number required by the default recall procedure. (In my opinion, six may be too low, it should be eight or ten, but this is what we have. If this procedure gains acceptance, the default reconfirmation procedure can be discussed and changed later.)

(You may also wish to advertise the petition in some neutral, friendly manner at some neutral messageboard(s) such as the pump.)

Point out that this is not the place for discussion of the merits of Admin X. All you are asking for is six signatures, and the only useful discussion would be trying to convince signers to withdraw their signatures (or you to withdraw the petition). Anything else would be wasted bits, save it for the RfA. The petition has to stay open a week, even if you have six signatures, as signers may withdraw during that time.

It's possible that someone will try to delete this section of your talk page. Don't stand for this, and insist on a proper MfD. Since it's your talk page and it's a perfectly respectful petition, you'll likely win this point. If the MfD succeeds, game over. If not, and after a week you have six valid signatures, continue to the next step.

Finding a closer

You need to determine who will close the RfA before you create it, because you must specify in the RfA itself who will close the RfA (either "a bureaucrat" or the name of a particular editor), to avoid confusion and drama.

The best person to close the RfA would be a bureaucrat. Approach the bureaucrats at the Wikipedia:Bureaucrats' noticeboard with a proposition to this effect:

Dear Bureaucrats, we are about to initiate a reconfirmation RfA (recall of an existing admin). However it goes, it won't require the use of any bureaucrat powers (no one is going to be sysopped, and bureaucrats can't desysop anyone except for inactivity), so closing it is technically outside the scope of your duties. Nevertheless, we would very much like a bureaucrat to close the RfA, because 1) you have experience closing RfA's, and 2) you are respected by the community as being especially fair-minded and skilled. Will one of you do the community this service? We don't need all of you to agree to this, just one of you. If you like, you may consider that you are doing the close as an editor who happens to be a bureaucrat rather than as a bureaucrat per se.

In any case, please don't otherwise interfere with the RfA. It is outside the scope of bureaucrat duties to do so. An MfD will be available to any editor who disagrees with the premise of the RfA.

This will result in either a particular bureaucrat agreeing to close, a number of bureaucrats agreeing that one of them will close, or no bureaucrat agreeing to close. If you can't get a bureaucrat to agree to close, you will have to find someone else. I would suggest finding some experienced editor who is uninvolved and who has a reputation for evenhandedness and who is willing to do it. A mediator type maybe, if you can get one.

(Once the principle of reconfirmation is established, a guideline specifying who closes reconfirmation RfA's will be established, but we'll leave that for the future.)

There has been some discussion about whether the assumed-to-pass percentage for reconfirmation RfA's should be lower than for regular RfA's, but this has never been decided. The person closing the RfA has full discretion to close as they see fit, including using a lower assumed-to-pass percentage. It's not anything that you can control.

Creating the reconfirmation RfA

Create the RfA page normally, but if there's going to be an MfD it'd probably be best to create it in your userspace (see below).

You don't have to (and probably shouldn't) say very much beyond "Here is a reconfirmation RfA", and describing what that is. Explain that voting is the same as a regular RfA, with "Support" meaning "this person should be an admin" and "Oppose" meaning "this person should not be an admin". You can describe, in a succinct and neutral manner, specific reasons for the RfA, but be evenhanded - you are not a prosecutor here. Leave the arguments to the the commentors.

You need to inform the subject, but it is not necessary for the subject to accept (he can if he wants to) or for you to include the usual three general questions, as neither of those requirements are appropriate for reconfirmation RfA's. If the subject does accept, you can probably skip the "the MfD" section below and publish the RfA normally. (the MfD will have to be done the first time a subject doesn't accept a reconfirmation RfA, though.)

If there's also to be an MfD (see below) add a prominent note to the top to effect of "This is not an active RfA, please don't vote yet."

In your userspace?

It's possible that the RfA should be placed in your userspace, with a second prominent note to the effect of "This page is in userspace, but for the purposes of this MfD discussion treat it like it's in Wikipedia space, where it will be placed immediately it survives the MfD (if it does)". I really don't know whether that'd be better or not.

Some advantages to holding it in userspace for the MfD process are:

You'd be more able to delete premature votes or people messing with the content without being accused of WP:OWN. (Improvements to the content for clarity etc. are welcome, I'm talking about messing it up on purpose.)
It would be a lot harder for a person to claim "MfD or no, admonitions to not vote yet or no, it expired one week after posting, period" and be justified as closing the RfA on the basis of no consensus (because no votes). (That'd be egregious Wikilawyering, but that happens.)

Some disadvantages to holding it in userspace for the MfD process are:

It might be confusing.
A person could claim that you can't run an MfD on a page that's not actually in Wikispace as if it was in Wikispace and that he'd be justified in running a new separate MfD immediately it's moved to Wikispace, which is what we're trying to avoid, an MfD concurrent with the voting. (That'd be egregious Wikilawyering, but that happens.)

If it was me I'd put it in my userspace and move it to Wikipedia space if it survives the MfD.

The MfD

You need to file and pass an MfD on the RfA before you publish it. This is 1) to prevent someone just deleting the page and possible subsequent drama, and 2) to prevent someone filing an MfD on the RfA when it is active, which would cause confusion.

Advisory XfD's are perfectly legitimate and are seen from time time. Your MfD should say something to this effect:

Here is a page for discussion. This is an advisory MfD and I as nominator don't support deletion. This is a reconfirmation RfA (also called recall RfA). It is essentially similar to a normal RfA, except that 1) the candidate is already an admin, and 2) the candidate has not accepted the RfA. It is because of these two differences, which have not before been seen in an RfA, that I initiate this opportunity for discussion.

and provide whatever ancillary data is required for people to understand the background, such as what will happen after the RfA is closed.

Since this is an important MfD, I would strongly suggest posting a short neutral notice pointing to the MfD at a few appropriate places - the WP:RFA talk page, say, maybe village pump, and definitely at the administrator's noticeboard. WP:MFD itself is not that heavily populated, and you want a full quorum and a legitimate MfD. This is not canvassing, and since you posted at the administrator's noticeboard (unlikely to a nexus of enthusiasm for the this procedure) you can't fairly be accused of that.

For the MfD to result in a "delete" decision, one of these conditions will have to occur:

A general overall consensus that the RfA should be deleted, or
A clear supermajority of commentors advocating deletion of the RfA, or
A clearly stronger argument (as a fair neutral observer would acknowledge) being made for deletion than for retention.

and it's not a given that any of these will in fact occur, although its certainly possible.

Although one can't anticipate everything, I think that the two "delete" arguments likely to be the strongest are:

This page is improper, since there is no policy or guideline describing or allowing a reconfirmation RfA, and no consensus to create one.
This is a malformed RfA, since WP:RFA says "Nominations may only be added here... after the candidate has signed their acceptance of the nomination", and that hasn't happened here.

You may be be able to come up with better counterarguments than I can (you shouldn't be doing this if you are not clever and subtle as well as brave), but some thoughts follow.

In response to the first point, you could make these counterarguments:

WP:RFA doesn't state that subject of an RfA can't be an admin. (And it's happened at least once before, albeit semi-voluntarily.)
And per User:Herostratus/Administrator reconfirmation#If the person fails the RfA the RfA can't create or mandate any state. It only creates a data point to be presented to the subject, the ArbCom, and the community.
And there is no policy or guideline prohibiting the existence of such a page creating such a data point.
And per WP:IAR we are allowed and even bound to perform such actions as are not prohibited and which enhance the Wikipedia. And while some people don't think that this enhances the Wikipedia, others do.

And probably other arguments that I haven't thought of.

In response to the second point, you could make these counterarguments:

The rule is designed to prevent pointless RfA's; reconfirmation RfA's are a different matter not envisioned when the rule was written, and they are not pointless if the subject doesn't accept, and so the rule doesn't apply to them. And WP:NOTBUREAUCRACY, which states "Do not follow an overly strict interpretation of the letter of policy", is a policy.
And as it says at Wikipedia:Guide to requests for adminship, "The RfA nomination process has evolved and continues to evolve", which sure seems to discourage reading each instruction as scripture in situations where they clearly aren't appropriate.
But if you want to quote scripture, you have
- this at Wikipedia:Guide to requests for adminship: "When nominating someone, it is generally a good idea to ask them if they would like to be nominated" (a good idea, not a requirement), and
- this at Wikipedia:Requests for adminship/Nominate: "Verify with the person you are wanting to nominate. This might save the embarrassing situation of them declining" (intention of rule is to save you embarrassment, which you are not concerned with).

And probably other arguments that I haven't thought of.

You'll either win these points and the MfD in general, or not. If the MfD results in deletion, fair enough, game over.

If the MfD results in a "keep" or "no consensus to delete" decision, post {{Template:Oldmfd}} template on the RfA talk page. You are now immunized against out-of-process deletion, edit warring, an MfD being filed against the active RfA, and similar drama, and may continue, publishing the RfA per the normal procedure.

If the person passes the RfA

End of story. The system worked, and this is certainly the best outcome. Nothing more needs to be done, except congratulating the reconfirmed admin.

If the person fails the RfA

If the person fails the RfA, the next step is to remove the person's admin rights. (The Bureaucrats can't do this, they have the technical ability to remove admin rights but are only allowed to so in certain specific narrow circumstances.)

Removing admin rights - first option

First, ask the person to resign their adminship. In my opinion, this would be the honorable thing to do if one fails an RfA, and the great majority of admins are honorable people, so problem solved. If they're not honorable, or if they just feel that the process was not legitimate, they might still be persuaded to resign just so as to not look silly, or just for the good of the Wikipedia, so problem solved.

Removing admin rights - second option

Second (if the person won't resign), file an ArbCom case, more or less to this effect:

Dear ArbCom, please adjudicate this matter. It won't take long. We assert that the following three statements are true: first, that Admin X was properly subjected to a reconfirmation RfA, by the signatures of six editors properly entitled to so sign. Second, that Admins X's RfA did not succeed. Third, that Admin X has refused to resign his administrator status. [Provide diffs; it will require the ArbCom just minutes to determine that these statements are true.]

Given the above statements, we ask that you make a decision on the following proposition: that it within the scope of ArbCom's duties to direct the stewards to remove Admin X's administrator rights. (And, if you decide in the affirmative, to so direct the stewards.)

We sincerely hope that you will at least accept this case, to clarify your stand on the matter. We also sincerely hope that your decision on the proposition will be in the affirmative, as otherwise we will have the very unsatisfactory situation of an administrator who has been formally rejected by the community continuing to hold administrator powers. In our opinion this would be most unhealthy situation.'

This needs to formatted as a proper Request for Arbitration. Who would be "parties" to the request I don't know. It could be just the RfA initiator (you), the RfA closer, and the un-reconfirmed admin. But maybe you should add everyone who commented in the RfA to be on the safe side. "Confirmation that other steps in dispute resolution have been tried" would point the declined (or ignored) request to the un-reconfirmed admin to resign voluntarily, I guess.

You can also point out that this won't increase their workload, as the few cases that are likely to come before them are simple technical matters requiring them to simply verify that the proper procedure has been followed.

If the ArbCom decides in the negative - that this is none of their business - or rejects the case (which amounts to the same thing), that would be too bad, and in my opinion terrible misfeasance on the part of ArbCom. An argument on narrow technical grounds could be made that adjudicating this matter could be said to be not strictly within ArbCom's scope. However, the ArbCom are not drones or dullard paper-shufflers, they are Wikipedians, and particularly committed and erudite ones at that, so I think it very likely that they will accept the petition and the proposition; it would be astonishing if they didn't, I think. But if they refuse to help out, you can't make them.

(It'd be great if the ArbCom would indicate their stance before we go through all this, but the ArbCom doesn't accept hypothetical cases.)

Removing admin rights - third option

If ArbCom fails the test, you're now out of good options, and I'm not sure exactly what to advise you.

You could just wait him out, send him to Coventry as much as possible and hope that after awhile he'll realize that clinging to the bit is not contributing to his overall happiness in life.

Or you could more or less force the ArbCom to take a case involving the person's use of his admin rights. (I'm not going to describe how you might do that; you can probably figure it out. It'd be painful, disruptive, and difficult all around.) Once the ArbCom is forced to take a case involving the person, they'll have no choice but to decide either that the person's admin rights are held legitimately, or that they're not. If forced to decide, It's hard to image them siding with a person who's been voted out, but I suppose it's possible.

Or -- and this also would be painful, disruptive, and difficult all around -- you could make a case to the admin corps that the presence of the person -- voted out by the community but still holding admin rights -- is inherently disruptive to the project, and we can deal with disruptive persons however we like, by community ban for instance. It'd be a reasonable case but I don't know if the admin corps would go for it. Maybe. Of course that results in the loss of the person altogether, which wasn't the original intent, but at this point it'd probably be for the best anyway.

I know none of these are good options. You've run out of those, which is too bad, so you'd have to decide if you want to pursue these bad options or just give up.

Removing admin rights - fourth option

There's no fourth option. You've done all you can, and it didn't work. That happens in life. You tried, and there's a saying that the only failure is not try. Maybe some other reform will come out it.

Some notes

If ArbCom won't help out, it might occur to you to go directly to the stewards, pointing to the failed RfA. However, I assume that the stewards only honor requests from properly constituted authorities (which is as it should be), so I would forget that idea.
Note that, by precedent, an RfA can consider everything. Even if the reconfirmation was triggered by a specific action or series of actions, all of the person's actions - including non-admin edits, actions performed before he was an admin, off-wiki behavior, an ugly userpage, and so forth - are up for discussion. I'm not saying that this is good or not, just that by precedent it's true. (Down the line, I think that a proposal to consider only certain classes of actions in a reconfirmation RfA would likely be made and would garner much support. But leave that for the future, you can't fix everything at once.)
The question of what to do with "Support, don't know anything about the case but am opposed to this process in principle" comments would be up the person closing the RfA, and its not anything that you can control. If it was me, I would disregard them as being out of scope for the venue.
Probably a few, or even many, reconfirmation RfA's will be attempted for insufficient reasons. Reading over the history, I see several attempts at recalling admins who have marked themselves as available for recall that failed to garner six signatures, so I think this is a sufficient barrier.
And a note to the community at large: for God's sake, remember that this is likely to be painful for the subject. It's just a fact that some people aren't cut for adminship, just as many people aren't cut out for management. I'm not (or at least the community doesn't think so) and there's nothing wrong with that. Whatever qualities of kindness that we have, this would be the time to deploy them. We need to do everything in our power to convince the person that we value them as an editor.

More blather

Expanding on the above point, its quite possible that a person going through this is likely to feel annoyed and will quite Wikipedia altogether. Sometimes this happens (when people are de-adminned through other methods) and sometimes not. If it does happen, that's a shame -- we lose a good editor. It's a quite understandable reaction (if a bit immature), and I don't know what to do about that. Certainly it's another reason not to initiate this process lightly. But we can't be held hostage to this. Functional member-run organizations need to police themselves.

Regarding the question of whether all this is a good idea or not: I don't know for sure (and neither do you). On the one hand, we need admins, and making it easier for us to have fewer admins might seem not so good. On the other hand, some people feel RfA is too rigorous, and if there's a procedure in place for getting admins to step down if they don't work out this rigor can be reduced, resulting -- in theory -- in more admins in the long run.

I don't know about that. My guess it'd have neither effect -- neither much reducing the number of admins directly (I assume its use would be rare) or much affecting how RfA is run. However, we need editors as well as admins. If the awareness of the existence of this process casts a shadow of doubt in the mind of other admins about to treat an editor overly harshly, that'd be good, and is pretty much the point of this whole exercise. If it casts a shadow of doubt in the mind of an admin about to perform a good action, that's no good. Would that happen? Probably not never. But nothing's perfect.

If it turns into a tool for trolls and malcontents, that's no good, but that can probably be handled by reform (e.g. an editor can sign just one petition a year or whatever). So I wouldn't worry about that too much. We can't reform it if it doesn't exist.