Jump to content

User:Tamzin/ACE2023 guide

From Wikipedia, the free encyclopedia

This is a single-issue guide on the topic of admin accountability, intended for voters who generally agree with my view that we need more accountability for admins and functionaries, and more transparency about that accountability. I have asked every ArbCom candidate this question:

Do you think ArbCom should be more transparent about the outcomes of private inquiries, especially regarding admins and functionaries? This question is motivated by the admin meatpuppetry situation in September, but it's up to you whether to discuss that situation in particular.

I have also considered responses to other questions about admin accountability.

I am making this guide because ArbCom, sadly, remains the only body with the power to desysop. Arbitrators' willingness to hold admins accountable, and philosophy for doing so, is one of the most important considerations in fitness to serve. I am not comparing candidates answers to a single "right" answer, but more interested in the philosophy of admin accountability that their answers show. That said, this assessment is not written from a neutral point of view. My analysis is based on my experience with ArbCom, as someone who has interacted with the Commitee extensively since becoming an admin in May 2022, and who has been a party to two cases this year. There are no right answers, but there are (in my view) wrong answers, and I have called them out as such.

Because this is a single-issue guide, I consider a positive finding here a mostly-necessary but not sufficient criterion for supporting. That is to say, if I find someone's commitment to admin accountability insufficient, I will (quite likely) either vote neutral or oppose, but if I find it satisfactory, that does not necessarily mean I will support.

Disclaimers

[edit]

I have a generally positive or neutral view of every candidate in this election, both as human beings and as editors (and administrators where applicable). I do not have a COI or INVOLVEment with any candidate in the policy sense, although Firefly was my RfA-flightmate-to-be (in the end we ran a few months apart due to my tardiness) and I would describe us as friendly. I talked for a bit with Z1720 at WikiConference North America 2023. I've exchanged emails or Discord messages on a few occasions with ToBeFree, Maxim, Cabayi, and I think Wugapodes, generally more "business" than "pleasure". HJ Mitchell indeffed my alt once, but based on an assumption that it was an account made to harass me, so I suppose that could be read as biasing me either for or against him.

I have strived to not let personal views influence me here. The three "insufficient"s I have cast are all for editors I have positive opinions of, and none was fun to write, but I feel some obligation to make something out of what I've learned in my interactions with ArbCom this year.

Assessments

[edit]

These are written to be read both by someone reading this page from the top down, or someone reading across all guides candidate-by-candidate. For that reason, some of the wording in each analysis may be a bit repetitive. Skim as needed.

Candidate Assessed level of commitment to admin accountability Analysis
Aoidh Gray check markYg Sufficient Aoidh's answer to my question was a bit light on details, but did address one of the most important details of admin accountability in private settings: ArbCom needing to be transparent and accountable while also needing to keep private information private is an unfortunate dichotomy, but I would say that ArbCom should be as transparent as privacy considerations permit. I was more impressed by his answer to Red-tailed hawk's "worst remedy or finding-of-fact" question, in which he called out the vague findings in Stephen as needing elaboration, rather than punt on a difficult question. The responses to WereSpielChequers' question about middle-ground admin sanctions and Xaosflux' question about community CU/OS management were also adequate.
Cabayi Red X symbolN Insufficient Cabayi was the only candidate to ask me for more details about my question. I responded by asking whether ArbCom had acted appropriately in handling my private report of meatpuppetry by Mark Ironie and CorbieVreccan. Cabayi's response implies that, by mandating disclosure of the shared IP aspect, ArbCom adequately addressed the situation. However, ArbCom was at that point in possession of information that Mark had blocked a user for edit-warring with Corbie, among several other admin or admin-adjacent actions, while the two shared an IP, including a vote to TBAN an opponent of Corbie's five days before I emailed the Committee. ArbCom took no steps to ensure that this matter would see the light of day after it chose not to impose sanctions, and if I had simply walked away (as I was tempted to given the stress of the situation), the users who were targeted by this long-term admin meatpuppetry would have never been aware of it.

That is, of course, my perception of the chain of events, and I do not pretend to be a neutral narrator here. I would not fault a candidate who said that ArbCom handled things well, if they could explain that position in a way that acknowledged that the IP-disclosure sanction failed to address the concerns of tool misuse. Cabayi's statement does not, and instead makes strange references to Mark and Corbie's overlap perhaps being coincidental (despite sharing an IP) and a sanction-free resolution that we and they hoped & anticipated. The responses to WereSpielChequers' question about middle-ground admin sanctions and Xaosflux' question about community CU/OS management are adequate. (WSQ's question is tough, so I won't fault a non-answer.) Combined with the answer to my question, I don't love that, in declining Red-tailed hawk's "worst" metric (to be clear, a reasonable time to "fight the question"), Cabayi's first alternative is to change "worst" to "harsh" and to give the example of stripp[ing] privileges from good people as a necessary act in maintaining community confidence that advanced tools were only being used in accordance with policy. To me this perpetuates the notion of advanced tools as a right, rather than privilege.

Firefly Green checkmarkY Exemplary Firefly offers a clear solution to the problem implied by my question: Perhaps in cases where there is both a private evidence component and a public/on-wiki component, the Committee could handle only the private component, give a suitably redacted summary of its findings, and pass the remainder of the matter back to the community. Naturally, there will be cases where the Committee feels the community is not able or not best placed to resolve the full matter, but I believe it usually owes the community a chance. This is an excellent idea, and one I hope that the Committee considers.

The responses to WereSpielChequers' question about middle-ground admin sanctions gives a good summary of the existence and appropriateness of such sanctions but also why they are rarely employed. During AlisonW, the first of two case( request)s to which I was a party this year, I watched the Committee wrestle with the question of when to impose such sanctions, and wrestled with it myself. I'm glad that Firefly will keep an open mind on that front. Firefly's answer to Xaosflux' question about community CU/OS management is reasonable and I appreciate his flexibility.

HJ Mitchell Red X symbolN (Barely) insufficient This one comes really close to the line, but I ultimately cannot support a candidate whose preferred approach is one that prioritizes the feelings of admins engaged in serious misconduct over the community's right to know. To be clear, I agree that ArbCom should always be as courteous and humane with parties as possible—both philosophically and out of my own self-interest, having been the subject of at least two (speedily dismissed) accusations of admin misconduct—but that cannot supersede enforcing core admin policies. In the case of my private report of meatpuppetry by Mark Ironie and CorbieVreccan, ArbCom was in possession of information that Mark had blocked a user for edit-warring with Corbie, among several other admin or admin-adjacent actions, while the two shared an IP, including a vote to TBAN an opponent of Corbie's five days before I emailed the Committee. HJ's suggestion that ArbCom should have been more transparent about how it compelled that IP disclosure is what moves this closer to the line, as it's close to a situation where the community could have handled the case with full information, but it's still not quite there, and implicitly treats adminship more like a right than a privilege.

The mostly-non-answer to WereSpielChequers' question about middle-ground admin sanctions is fine given the difficulty of the question, and I like the redirection to the matter of re-RfAs. In AlisonW, the first of two case( request)s to which I was a party this year, it was discussed how the candidate—by all accounts a likable person committed to the Wikimedia movement—would be a good re-RfA candidate someday, but also how it has been a very long time since a re-RfA after a for-cause/under-a-cloud desysop succeeded. The response to Xaosflux' question about community CU/OS management is a good example of a comment that veers away from maximum accountability but is nonetheless reasonable and evidence-based, and so I won't at all hold it against HJ. So overall HJ fails my standard for this single-issue guide, but only barely; I would definitely understand why one would vote for him nonetheless, and I think that on a committee that is adequately accountability-minded, there wouldn't be much downside.

Maxim Green checkmarkY Exemplary Maxim responds to my question that if it was a formal case with some elements of private evidence, we would summarize and publish that evidence on-wiki to the extent possible—fairly similar to Firefly, although framing the matter more as a past practice than a new proposal. I'm not sure which framing is correct, but as long as it gets done, that satisfies my criteria for increased admin accountability in private matters. Maxim stresses the interplay between privacy and transparency concerns in ArbCom business, without presupposing whether the correct decision was made in the Mark/Corbie case. That's really all that can be asked for.

It's no surprise that Maxim, as the maintainer of User:Maxim/ArbCom and desysops, gives one of the best answers to WereSpielChequers regarding middle-ground admin sanctions, highlighting both the exceptional nature of GiantSnowman and the circumstances under which such a sanction could be seen again. Their response to Xaosflux' question about community CU/OS management is a reasonable analysis of community, ArbCom, OmbComm, and (previously) AUSC jurisdictions, albeit somewhat dodging giving a personal opinion by leaning on what would be likely to gain support.

Robert McClenon Gray check markYg (Barely) sufficient The response to my question is a bit strange, in that Robert focuses on there being just enough public information for the case to provoke interest without satisfying it, when in fact all relevant information was public by the time I started the AN thread. However, he does say In general, ArbCom should maintain the maximum transparency that is consistent with privacy, and doesn't say anything to imply that adminship is a right rather than a privilege, nor that admins' feelings should be put ahead of the community's wellbeing, so this barely passes the mark. In response to WereSpielChequers' question about middle-ground admin sanctions, I would have liked to see more discussion of the implications, with respect to fitness to be an admin, of admins committing sanctionable misconduct.

Robert's answer to Xaosflux' question about community CU/OS management is light on details, but falls on the side of greater accountability, so that's good. I also see it as a strong plus in terms of admin accountability that Robert is not himself an admin. I will basically always vote for an experienced non-admin arb candidate, for this reason.

Sdrqaz Green checkmarkY Exemplary The words when we keep matters completely private, decisions may be perceived as carrying less weight than in public get to the heart of the issue with private resolution of admin accountability concerns. Examples of the other layers of scrutiny to the Committee's actions contemplated would have been great, but I didn't ask for them in the question, and Sdrqaz' answer satisfies what I'm looking for: a commitment to admin accountability to the greatest extent possible without conflicting with other priorities.

Sdrqaz' answer to WereSpielChequers' question about middle-ground admin sanctions manages to best Maxim at their own game. Sdrqaz gives a great analysis of the role of admin responsiveness to feedback, and of the fact that admin conduct cases are often brought to the Committee quicker than others. Sdrqaz makes a good case for drawing a distinction between editing and adminning restrictions, and makes a fair criticism of the enforceability of the sanctions in GiantSnowman (although for what it's worth I've never observed GS to violate them). Sdrqaz' answer to Xaosflux' question about community CU/OS management strikes me as a missed opportunity to explore the nuances of accountability and transparency, but does not contain anything I actually find objectionable.

ToBeFree Gray check markYg Sufficient ToBeFree's answer to my question is short but says what needs to be said. He doesn't equivocate in response to a follow-up by Faendalimas about balancing transparency and privacy. He makes a good point, in response to WereSpielChequers' question about middle-ground admin sanctions, that the only difference between an admonishment and a restriction is the threat of a block. His response to Xaosflux' question about community CU/OS management is fine, although I can't believe he missed the obvious "arbitrary committee" pun.
Wugapodes Red X symbolN Insufficient; outright oppose for lack of candor I have to disagree with Wugapodes' answer to my question from its very first sentence: The committee is generally very transparent regarding the outcomes of private inquiries, especially as they relate to admins and functionaries. The last entry in the WP:ACN archives giving the outcome of a private admin conduct investigation is Level II desysop of Stephen (Nov. '22), which the Committee was required by policy to post; the last I can find where no sanction was imposed (i.e. ArbCom was not required to notify the community) is... actually, I don't know. I've read back to the start of Wikipedia:Arbitration Committee/Noticeboard/Archive 11 (Dec. '15) and can't find a single one. I know that I've been the subject of a private ArbCom report in that time. It was non-credible—an accusation of misusing private information that I never had access to to begin with—but I'm also aware of credible inquiries, including ones that have resulted in checkusers being directed to not run checks under certain circumstances, and one in which an admin was credibly accused of logged-out vandalism. So I would not call 0 publicized inquiry results, other than bans or blocks, since 2016 or earlier, "very transparent", and I don't know where Wugapodes, generally a very reasonable editor, would get the idea that it is.

[Addendum: Shortly after I posted the initial version of this analysis, ArbCom suspended Beeblebrox for off-wiki disclosure of private information, and in doing so disclosed that Beeblebrox had been warned for the same in 2021. I struggle to reconcile my view of Wugapodes as an honest and thoughtful editor with the fact that they, while actively discussing a motion that acknowledged that 2021 warning—issued to Beeblebrox without any disclosure to the community, to someone operating at the highest level of trust, in a year that they successfully ran for reelection—was willing to say with such certainty that, I quote again, The committee is generally very transparent regarding the outcomes of private inquiries, especially as they relate to admins and functionaries. Regardless of whether one agrees with my view on accountability and transparency, I hope the community can recognize this as an evasive or even deceptive statement for Wugapodes to have made while staring in the face of evidence to the contrary. I will make an exception to my framing of this guide as a single-issue one, and adjust my assessment to outright oppose. I do not think Wugapodes has shown sufficient candor to be trusted to serve on ArbCom any longer, nor should they be serving on ArbCom right now.]

[Original analysis continues.] Wugapodes continues that my message about Mark Ironie and CorbieVreccan was "a narrow question pursuant to language at SPI". My email concluded with a recommendation of blocks of one month if mitigation existed, indefinite if not, so I wouldn't call that a narrow question, or a question at all. As Wugapodes knows, ArbCom never responded to me on the matter. When I was informally notified of the mandated self-disclosure, I was told that ArbCom would not stop me from bringing the matter to AN, and so I did that, taking on what ArbCom had refused to. Given Wugapodes' statement that the Committee did not want the matter publicized at WP:ACN, it's hard to believe that the Committee simultaneously expected me to actually take it to AN. This is further borne out by the Committee later tending toward sanctions in public that it had just rejected in private, based on substantially the same evidence.

Ultimately, the statement Generally I think we strike a good balance between concerns for privacy and the need for transparency so that the community can handle affairs on its own is only true if "the community" refers to me. I have brought or spurred two of the four case requests this year that resulted in an admin's desysop for-cause or under-a-cloud. One of those (AlisonW) was a case I stumbled on completely by chance. This does not sound like a working system. I do not think Wugapodes has made a good case, or any case at all really, that it is one. The appeal to privacy is particularly ironic in the Mark/Corbie case, given that, as far as I know, not a single detail was ever actually eligible for Oversight.

The answer to WereSpielChequers' question about middle-ground admin sanctions is well-reasoned, and I like the deference to the community. The answer to Xaosflux' question about community CU/OS management is thorough and one of the best anyone has given.

Z1720 Green checkmarkY Exemplary Z1720 lays out a clear and reasonable approach to admin accountability in private cases: I will advocate for transparency unless there are outing concerns, personal harm, legal reasons why this information cannot be public, or other serious considerations. If necessary, I will encourage the committee to announce at WP:A/N when the committee is evaluating a user's conduct, give specifics if appropriate, and ask editors to submit evidence in the appropriate channels. Notifying the community not just at the conclusion of a private investigation but even midway through is something that no other candidate has suggested, and is a great idea. I do bristle slightly at the phrase "personal harm"—not because I disagree, but because the term is nebulous, and to different arbs could mean anything from "would lead to attempts on the subject's life" to "would make the subject sad". But overall I think this is a great idea.

I'm curious about Z's point in response to WereSpielChequers' question about middle-ground admin sanctions. He says The fear of desysop prevents great admin from commenting in contentious areas. This leads to fewer admin maintaining Wikipedia policies in these topics, leading to burnout, or these topics spiralling out of control until a case is brought to ArbCom. Is it true that some admins feel that way? If so, it's not a well-founded concern, as I'm not aware of any case where an admin has been desysopped for unINVOLVED routine moderation of a contentious area. David Gerard's sanction in Manning naming dispute (2013)—an exceptionally poor sanction that I doubt would ever be levied today—is the only one I can see since 2010 that might qualify as justifying Z's concerns. That's not to say that Z is wrong that some admins fear this; but the solution might be dispelling a myth rather than changing anything. Anyways, this is mostly tangential. The overall answer is fine. I appreciate Z's willingness to take Red-tailed hawk's "worst remedy or finding-of-fact" question by the horns and call out ArbCom for insufficient guidance to AE admins. (Maybe a bit biased here as an AE admin.) I also appreciate his willingness to explicitly call for some kind of community CU/OS appointment process in response to Xaosflux' question.