User:Maxim/The dark abyss

It has occurred to me that we don't have a very up-to-date description of what happens with email received by the Arbitration Committee. It's no secret that there is a perception that ArbCom can be a bit of black box, particularly with respect to the emails that are sent to it.

A top-level summary is roughly as follows. The email gets acknowledged, and then several arbitrators opine. There may be some sort of follow-up with the original sender or the subject of the complaint. Where things get a bit muddier is in the response. Unless an issue can be reasonably handled by an individual arbitrator acting in an "individual capacity", that is, as an ordinary administrator or an ordinary checkuser, it requires multiple arbitrators to agree on an action. It might not be a strict majority, but at a minimum, it requires a sense of where the committee as a whole stands (think of how public requests for clarification are handled). But, depending on what the matter is about, it may be difficult to get to the "multiple arbitrators to agree" step, not as much because there are too many competing ideas, but because none are acceptable (or suggested).

The specifics of what happens really depend on what the original email is about. I move to a bulleted list to describe there:

• Appeals: as of a month prior to writing this essay, we have very little appeals traffic now, because of appeals reform. If an appeal does stall, it suggests that the appeal is not obviously bad, but a good case to unblock has not been given. Especially when we had large numbers of appeals, my thoughts on how to deal with that is to decline appeals that have not gained consensus to unblock after several weeks (or months).
• Cases: especially with private evidence submissions, these are noted. Unless we need to follow up with the submitter, such a chain stops at the acknowledgment.
• Socks: either we dump this off to the checkuser team (but if it's sent to ArbCom, often enough it is for cause), an individual arbitrator does something, or, in rare cases, the whole committee reviews it. The sticking point is someone needing to do the investigation, and the reason why that's sticky here is because not all arbitrators are proficient sock-hunters and checkusers. Of course, if the results are not very conclusive and it's not going the individual route, then agreeing to a conclusion will take a while.
• COI/UPE reports: as with socks, someone needs to do the investigation. I find these situations are less straightforward than socking reports. Not all arbitrators are experienced with these matters (in a practical context, it's more acute than for sockhunting and checkusering). Expect an even slower response here.
• Functionary issues: in practice, enough arbitrators need to agree that there exists a solid likelihood of an issue, before a more meticulous investigation takes place. This kind of matter can become quite tedious if a thorough investigation is required. It requires expertise with both the usage and the standards of use of the relevant tool. Adjacent to this category are occasional administrator issues, such as Level 1 or Level 2 situations. Functionary appointments follow a fairly fixed format, so that is in essence a separate process altogether.
• Harassment: this is a very broad category, ranging from "someone was mean to me on Discord" to serious cases of off-wiki abuse. It probably requires at least one arbitrator to investigate. Obvious cases get dealt with quickly; something in between will probably stall.
• Long-term improvement: there's a reason it's called "long-term". Oftentimes these are various reform proposals, usually internal, although they can often come out of election platforms.
• Foundation matters: we are in fairly regular communication with the Foundation, mostly Trust and Safety. Sometimes we refer someone or something to them; sometimes it's the other way around. Sometimes it's more that falls under "long-term improvement".

In general, outside of appeals and cases, a given issue or thread probably needs an arbitrator to take the lead on it; however, that is easier said than done. Despite a 15-member committee, it is normal that a few arbitrators are formally inactive, and a few more are somewhat active but maybe don't have enough time or energy to take the lead on something (two years can be a long time). Worm That Turned's taxonomy of arbitrators is also relevant here. The result is somewhat of a limitation of manpower. The committee will handle issues that are the most pressing, in terms of urgency or potential drawbacks of not handling it. As those issues already have arbitrators putting substantial efforts in, it is natural that other issues may fall off to the side. Tracking of the lower-priority issues has much improved (the bolded items in the bulleted list above largely correspond to headings from a tracking page).

Something else that I want to comment on is the paradox of a volunteer committee that, ideally, should review all matters sent to it. Consider an ordinary administrator patrolling AN/I; no one expects this administrator to opine at every thread, as after all, it is a volunteer gig. Yet that framing does not (and perhaps can not) apply to the Arbitration Committee. We all individually have reasons for opining or taking the lead at one issue or another, so I will speak for myself here. For example, I have a better time with appeals (when we had them, but that was a huge time-sink), socking, functionary issues, and arguably administrator issues. I do a certain amount of general administrative work as well. I've drafted a few cases, but I don't think I've ended up being a frequent drafter; oftentimes there are eager volunteers for that already. I've seen arbitrators not do appeals, or not deal with any functionary matters. I think that's completely reasonable, as we are volunteers after all, and for first-time arbs, it's very difficult to get a sense of what exactly comes up until getting subscribing to all the mailing lists.

The short answer to "why did my complaint vanish into a dark abyss" is a mismatch between the nature of the complaint and who is available to investigate it. A slightly fuller answer is that the unavailability depends on what else the committee is working on, but also on the interests, capabilities, and time and energy levels of the arbitrators.