Handling Calls to Remove a Moderator [duplicate]

Question

This discussion informed our process for seven years, but eventually required updates and augmentation.
For discussion regarding the successors to this process, please see: Feedback post: Moderator review and reinstatement processes

---

Stack Exchange has some of the best Moderators around — seriously — and that is due in no small part to the communities who scrutinize and vote in our Moderator elections. We have a rather formal process for electing moderators, but up until now the process for removing a moderator has been quite informal.

On those rare, rare occasion when things just aren't working out, you can see why it becomes such a sticky business to remove someone from a democratically elected position. Our current process is one of "we've got your back." The Community Team routinely monitors moderator activities, so if a moderator violates the Moderator Agreement or fails to uphold the Theory of Moderation, we can and do intervene. If the disputes and problems persist without resolution, eventually we can ask someone to step down.

That's not good enough. We need a much more transparent and prescribed process for when — and more importantly how — to intervene when on-going problems are going unresolved. In urgent situations, we'll step in to protect a site, but ideally the process should be impartial and not subject to accusations of Stack Exchange simply applying undue influence over a community when the whim strikes.

The Question:

Before we start bouncing around ideas internally, I'd like to hear what you think. What process should be in place to add some checks and balances to deal with higher issues like possible impeachment of a moderator? When and how would such a process come into play? How should such situations be resolved? How do we keep it from being abused?

Have at it. This is a round table vetting of ideas.

Answer 1

Moved to: What processes exist to issue formal warnings or remove moderators?
Please refer to that answer, as it may be updated in the future. This answer remains to document the process that was used between 2012 and 2019.

---

Moderator Action Review Process

After much discussion, debate, reflection, consultation, frustration, revelation, constipation and inspiration, we've settled on the following process for allowing a team of moderators to remove one of their own. It is, by necessity, somewhat formal: this process is for those rare situations where communication with one member has completely broken down and the team as a whole feels they cannot continue to work together.

I strongly urge anyone involved in such a situation to do what they can to resolve outstanding issues before resorting to this.

Initiation

The process will be initiated by a formal request from one or more moderators on a site, sent privately via an email to [email protected].

The process may also be initiated by the Community Team at Stack Exchange, Inc. in response to numerous, substantiated complaints from users on the site.

In either case, the complaints will be treated by Stack Exchange as confidential, and their authors will not be named by us at any point in the process.

Proceedings

Once begun, the following steps must be followed, in order, to completion within a reasonable time frame. If this is not possible, all participants will be notified by us that the process has been discontinued and informed of the resolution (if any).

1. All moderators on the site will be contacted by us via email, informed of the situation and asked to meet at their earliest opportunity to discuss the removal of the named moderator. Meeting must be held in a private venue to maintain the confidentiality of those involved — we will provide a chat room on http://chat.meta.stackexchange.com that is inaccessible to anyone not invited. Both the venue and timeframe for the meeting must be accessible to all moderators — we will attempt to coordinate the schedules of individual moderators.

2. At the designated time, a quorum must be present — this shall consist of ⅔rds (rounded up) of the moderators on the site (all those listed as active on the /users?tab=moderators page, whether or not currently active), excluding the moderator to be removed (example: for a team with three moderators, both of those not being considered for removal must be present).

3. A designated person will be selected to record the minutes of the meeting. These should be brief, and suitable for public consumption should the need arise (providing only a broad overview of the process and its outcome, not including any details of what was discussed).

4. At this point, each present member of the moderator team shall be given an opportunity to share their concerns with the group. In the event that this process was initiated by complaints from outside the moderator team, a summary of them will have been provided to the moderators prior to the meeting.

5. Following this, the moderator to be considered for removal shall have a chance to respond. There will be no back-and-forth discussion allowed — this should have been conducted prior to this meeting.

6. Finally, the moderator to be considered for removal shall be asked to leave the room (upon which access to the chat room shall be revoked) and otherwise remain silent through the remainder of the proceedings, and those remaining shall vote on whether or not to revoke the moderator's privileges.

7. If at least ⅔rds of those present vote for removal, this shall be considered a consensus, and recorded in the minutes as the opinion of the moderator team.

8. The meeting shall now be concluded, and the minutes emailed to all members of the moderator team and [email protected].

9. If the consensus was for removal, we will then revoke the privileges of the moderator to be removed, and also remove the moderator's name from the election and /users?tab=moderators pages.

How — or if — the outcome of such a meeting is shared with the broader community will be left to the discretion of the moderator team. However, the details of the meeting must remain confidential — only the minutes can be published if such a need arises. This is done to allow potentially-confidential information to be disclosed without forcing anyone involved to violate the moderator agreement.

In general, the same courtesies should be extended toward removed moderators as to suspended users: no airing of dirty laundry in public, no bringing up issues faced as a moderator in unrelated discussions, questions raised by other members of the public answered with as little detail as possible. Speculation should be discouraged out of respect for those involved.

Answer 2

First, it should never come to this -- as Robert Harvey pointed out, this should be handled via private conversations whenever possible. Moderators are supposed to be exemplars of 99% good community behavior, and part of good community behavior is recognizing problem behaviors -- even your own! -- and helping correct them in a constructive way. If it has come to this, unfortunately, the situation must be quite dire indeed.

I tend to agree with the other answers here indicating that moderators should be primarily subject to review by their other fellow moderators, in tandem with SE, Inc.

I believe this makes the most sense because:

• Most of the data you would be reviewing is private by its very nature, and thus only mods should see it.

• The whole reason we have multiple moderators is to establish a system of checks and balances. Mods do support each other but they can also disagree.

• The closest peer a moderator has is another fellow elected or appointed moderator.

• Moderators are of the community; they are not employed by SE, Inc and thus cannot not be influenced one way or another about any perceived or imagined "we're doing what the company paid us to do" bias.

• SE, Inc. would mostly be acting as a mediator in this case.

If it comes to that, throw together a private chatroom, get all the fellow mods in the room, review the data behind the recall request and decide. The results (but not the underlying data) can be posted on the appropriate meta.

I definitely do not see any need for a special tool for this. If we start building "courtrooms", that's.. scary. I've recently been part of a public jury trial, and the further we can stay away from that level of process, the better off we will be, unless it is a method of absolute last world-ending resort.

The community would have to understand that once convened, the decision of the community mods is final.

Answer 3

but ideally the process should be impartial and not subject to accusations of Stack Exchange simply applying undue influence over a community when the whim strikes.

Frankly, I see this as trying to have our cake and eat it too. We want influence over these communities - the core premise of Stack Exchange 2.0 is that we have some amount of experience here and can help these sites avoid some of the pitfalls they'll encounter along the way.

Of course, the problem is that we don't scale, and can easily become victims of our own success: too many sites with too many users and too many moderators to guide and chide and hand-hold. We're not there yet - but it's looming on the horizon. And no one wants to see us snatch defeat from the jaws of victory by letting a bad moderator run roughshod over a burgeoning community while we're engaged elsewhere...

But back to this "undue influence when the whim strikes" thing. Regardless of how this is conducted, I can't imagine a reasonable system for removing moderators outside of dire emergencies that wasn't foreshadowed by ample amounts of public discussion. We have these systems in place - meta, chat, moderators themselves in part to achieve just this sort of transparency.

Oh yeah... moderators. In well-established communities, moderators are elected, put in place via an open process by which the community itself selects them. We trust them to act as intermediaries in many cases already, handling the bulk of the work of community-building, cleanup, and support on a site. So why not this? If anyone is qualified to say that a moderator should be removed on a site, it would be the other moderators, those who've worked closely with him and are most familiar with his actions or inaction, both public and private.

So how about this for a process: make it clear that, should 2/3rds of the moderators on a site formally request the removal of a moderator, presenting evidence of abuse or misbehavior via email or some other private channel, we'll be obligated to remove him from his post. All sites should have at least three moderators, so this should scale without putting excessive pressure on the moderators of larger sites to play politician.

Yes, it falls apart when there are multiple - perhaps even a majority - of moderators causing problems on a site. But at that point, I think we can safely say we have bigger problems than appearing to exert undue influence.

Answer 4

The central issue here seems to be transparency.

From time to time, I've asked SE if they would tell me about any formal complaints that have been lodged against me. I've always been told no. Here's why:

• Users need the ability to report problems with moderators in private; any public disclosure of such information would have a chilling effect on reporting, since users might fear retribution.

• Moderators already have enough to worry about; they don't need to know about these complaints until they become actionable.

• Meta already provides an adequate feedback mechanism for public review of moderator actions.

For these reasons, I believe the entire process of removing a mod can never be made completely public. However, the community will eventually notice if a mod is surreptitiously removed.

So if a moderator's imminent demise needs to be made public, here are my thoughts on how that process might occur:

1. SE tries to handle the problem privately first. This is a conversation between the mod and SE corporate. We don't want to undermine the authority of good mods who just had a bad day by giving them a public scarlet letter.

2. We escalate it to a semi-private process. The mod gets an annotation on his account summarizing the problem. This surfaces the information to the other mods on the site. A discussion occurs between all site mods and SE corporate.

3. In the event that an actionable offense (or series of offenses) occurs, a notice is posted on the site meta, with a request for feedback and review. SE Corporate gets to decide what constitutes an actionable offense.

4. Details about the user accounts of those who report problems are kept private at all times. It's up to each individual user to decide if they want to "out" themselves on the Meta post notice.

5. Possible actions can include suspension of moderators. Yes, I do mean that. In particular, if a moderator figures out who reported them, and attempts to interact with that user in any way having to do with their report, it should be an immediate 30 day suspension.

6. SE makes a decision on what to do, and takes the action they deem necessary. The final action is posted to the meta notice, and the notice is deleted 30 days later.

I strongly believe that SE corporate (i.e. the Community Team) needs to be the final arbiter in disputes like this, for reasons which should already be obvious.

Answer 5

When this question was first asked, I had a hard time coming up with a response. I've only recently realized why: it's looking at the wrong problem. The discussion veered pretty quickly towards "mods flipping out and abusing their powers," but that's not something we see much of in practice. (According to Shog, it has happened a few times, though; see his comment under this answer.)

It's like the UN got together and put a lot of energy into preventing a hypothetical all-out war between Tonga and the Seychelles* instead of working on [insert any of the numerous existing serious world issues here].

I think it would be more constructive to discuss problems with mods that are more likely to come up in the real world: mods who aren't doing anything blatantly abusive, but also aren't very good at moderating. All of our mods are volunteers who are elected from the community. I believe they all mean well, and it's normal for them to occasionally make mistakes, or decide that an unpopular call is the right one. However, some of them have repeatedly displayed poor judgment — an intentionally subjective term — and others have acted in ways that a reasonable person might consider abrasive, rude or otherwise unbefitting the higher standard that comes with the diamond.

So far, this question, and the official response, haven't addressed this more common case. There's still no mechanism for community members to say "I think this mod is harmful to the site even though he's not actively breaking its rules," or even "this mod is doing a generally okay job, but I have a few specific concerns." These are messages that don't warrant an e-mail to the team, but still ought to be handled in a structured way.

^{*: Tonga and the Seychelles have some of the smallest militaries in the world. They also have no hostility towards each other, as far as I know... or even any regular interaction, for that matter.}

Answer 6

I have a hard time believing that someone who came through our election system would not elect to remove themselves before having to be removed. While our network of moderators is now so large that we all don't know each other as well as we used to, I can't think of a single moderator that would not step down if they found themselves unable to do the job within the guidelines they agreed to when elected.

Outright abuse is clear. If a moderator is abusing their access to information or tools at their disposal, they should be immediately removed. Our actions are verbosely logged, blatant abuse is immediately obvious.

Accidental slips are just that, accidents. If a moderator develops a pattern of disclosing things that should not be disclosed despite warnings, then they do need to be removed no matter how well intentioned they might be.

Our community team can handle this, and we now depart from objective cut and dry territory.

If a moderator is:

• Acting in good faith
• Justifying and explaining their actions, even if many disagree with the justification
• Working for the community that elected them

Then you have a series of details that can (typically) only be worked out between the moderators and the community management team. There are almost always going to be details that matter which can't be disclosed in a meta discussion.

Again, our community team can handle this on a case by case basis. Stack exchange is not in the habit of hiring jerks to handle moderators.

Stack Overflow has another stipulation that other sites don't have, we need to actively moderate while actively using the site. The community team can handle this, in addition to moderators that never show up after being elected or just 'vanish'. That's happened before, and was handled.

To reiterate

I don't think that we'd ever have to catch an exception of an elected moderator realizing that they were more toxic than helpful to the community without changing their behavior or stepping down. Why is leaving this uber rare exception up to our human Pokemon (gotta catch 'em all!) AKA community team handlers not going to scale in the future?

I don't think 'bad moderation' can be sufficiently defined, which needs to happen if some set of procedures is going to be put in place to handle it.

Answer 7

"if a moderator breaches the mod agreement". Of course it's a big assumption and you are begging your case.

The fact is, we initially never know whether this is the case, and the facts of the matter need to be ascertained, first and foremost. The "moderator who is guilty by hypothesis" is not a valid starting point here.

Any user is innocent until proven guilty. I hope there is no discussion here.

Therefore, I demand loudly that no public action is taken against any user, including a moderator, before the facts of the matter are ascertained OR the user requests the discussions to be public.

I don't think it is even remotely acceptable to publicly shame (or put to trial) the action of a unpaid volunteer because someone disagrees with them.

Even if a mod is wrong, please say "thank you for helping us in your free time", not "I will put you on trial in public".

Answer 8

While it's no secret I'm very much in favor of periodic elections¹, others have already voiced that position here so I'll offer something different: the ability to escalate an issue to Stack Exchange, Inc. for arbitration.

Having some experience in the capacity of being a moderator, particularly one who would've been subject to impeachment or violent overthrow if given the chance, one of the most frustrating aspects of acting on or implementing guidelines that happen to be unpopular amongst the more vocal elements of a site's community was having to argue a position to people who did not find anyone who believed such guidelines were useful to be credible.

The end result being that moderators, and people who thought the moderators were doing more or less the right thing, spend an inordinate amount of time trying to "manage" the periodic outbursts about various issues in a way that would never, ever be convincing to those who keep bringing them up.

When it's clear certain community members simply do not believe moderators are acting in the site's best interest or—more benignly—believe the moderators are simply mistaken, it would be extremely helpful if such issues could be escalated to Stack Exchange in a more formal, and transparent, manner than a ping in TL, a member of the community team stopping by on a child meta once in a while, or a private email to [email protected].

Once SE makes a decision on an issue, the matter is closed. If those who have a problem with the moderators don't like the decision, they can formally propose new moderation guidelines to be vetted by the community or, worst case scenario, they can find a place more suited to their idea of how such a site should be run. If a moderator doesn't like the decision and they feel they can't faithfully execute their duties, they can step down.

At any rate, if it takes the form of arbitration or something else, I strongly think there needs to be a step between "people have a problem with a moderator" and "initiate moderator destruct sequence Omega Charlie Alpha One": while it's nice that SE has the back of moderators, if something isn't going great, the quicker moderators find out, the quicker and easier it is to course correct.

^{Note 1: which would benefit moderators just as much as it would benefit users by giving moderators periodic mandates from the users to act in their best judgement, in addition to allowing moderators the ability step down in an orderly transfer of duties, but that's neither here nor there.}

Answer 9

I will participate through votes and comments on other issues, this post is simply to highlight the issue I think is most important to consider.

Don't let squeaky wheels have all the grease! It has to go around. Moderators dealing with these sort of issues already takes an inordinate amount of time compared to their normal contributions. The issue of disgruntled users not liking a moderator based on their enforcement of site policy is quite common, and the number of non-mod users who step in to speak out and counter these views is relatively low. It just isn't worth their time. Who wants to be bothered to put together a defense of a moderator who already has the job of dealing with that sort of mess?

Any system for revoking moderator status should require participation from a larger cross section of the community than the vocal objectors. In order for noise from low quality contributors to become actionable, there should be a minimum threshold of engagement from other users in good standing. Three dissenters out of 10k users shouldn't make waves. If they can rally x% of site users to agree that something needs to be done, fine. On the flip side if enough people step up to say that the objections aren't merited, the whole thing should be put to rest and stop wasting everybody's time.

The road has to end somewhere. The people that have done the most clamoring for a way to impeach moderators have already been through all the channels and gotten lots of attention. They have already meta posted and made their complaints to the community team. They have already been hand reviewed by many many eyeballs. Their meta posts have been soundly rejected by the communities, and neither the jury of peers nor the community team has found fault that requires a revocation of privileges. Yet the clamor continues. There has to be a way to put it to rest. The sort of people that make the most noise in this area won't take no for an answer and no system that doesn't successfully revoke the diamond from their favorite moderator-gone-wild will satisfy.

Any new system for handling these should have clear signage that shows the end of the road. If enough community have voiced their continued confidence, or the actions called into question have been properly peer reviewed or whatever system is put in place -- any votes of no confidence should be marked as not-actionable and everybody gets to move on.

Answer 10

Stack Exchange has some of the best Moderators around — seriously — and that is due in no small part to the communities who scrutinize and vote in our Moderator elections.

Thanks. I'd appreciate if, as some small manner of recompense, Stack Exchange would stop driving wedges into the community (Summer of Love, anyone?) only to immediately realize what a terrible idea it was and make up for it all with a blog post. Because I'm calling it; that's pretty much our best-case outcome here. No, seriously, both this and Summer of Love do little but give good reason for well meaning new users to suddenly feel like the victims of well meaning experienced users; suddenly moderating or voting to close is an offense.

The Community Team routinely monitors moderator activities, so if a moderator violates the Moderator Agreement or fails to uphold the Theory of Moderation, we can and do intervene. If the disputes and problems persist without resolution, eventually we can ask someone to step down.

Yep, that's pretty awesome. Stack Exchange communities have multiple tiers of moderation, giving more and more tools to more trusted and experienced users. This means very rarely do people have to act in any form of unilateral, irreversible private decision; almost every action is already audited publicly, and SE employees manually review the relatively small amount of private-only actions just to make sure everything's on the up-and-up. If all else fails, the Stack Exchange community team is just an email away for edge-cases. The current system is one of the most transparent moderation systems I've seen on the internet and it works very well, you should be proud of it.

...which is why I'm quite puzzled by your following statement:

That's not good enough.

Why not? What's going undone? What havock has been wreaked? Are mods unable to do their jobs, are your community managers drowning in mod abuse emails, is there actual substantial evidence of repeated and significant abuse of moderation powers? Where is the justification? You're suggesting a big change, big changes need justification. At worst even a useless change brings cognitive cost with it; suddenly moderation is more complex for users and moderators.

You don't change stuff just to mix it up, you need data, facts, understanding. I don't see any of those. A comment struck me; saying "if you want real reform"...why do we want real reform? Reform usually happens in the name of violence, oppression, slavery, death. You don't reform your secretary because she wears Crocs and they're stupid looking. You need something big.

We need a much more transparent and prescribed process for when — and more importantly how — to intervene when on-going problems are going unresolved.

This strikes me as an extremely extraordinary and unprecedented suggestion. Looking around the social web, I can't cite a single example of a dedicated "mod abuse" button like this. The standard protocol is "if you have a complaint, email X". Sure, there's a system of escalation (which we have), but if I saw a community website with dedicated features to punish and expunge moderators...well damn, I'd wonder just what in the hell brought that about.

It's freaking scary to imagine a community so terrible you seriously need baked-in features for this. It's like having a Report Child Porn button; features like this don't go in proactively (at least not with anyone sane calling the shots). On that note I find it bitterly ironic we're discussing theoretical moderation issues while a Reddit moderator's personal details have been outed publicly due to a malignant chain of trolling and creepy behavior on his part. Maybe if Reddit was having this conversation I'd understand, but here? Seriously?

Before we start bouncing around ideas internally, I'd like to hear what you think.

It's probably pretty abundantly clear what I think; that this is an ill-advised, unjustified, self destructive and all around stupid idea. It has a massive potential to be toxic to the community/moderator relationship, to be stressful for moderators and awkward for users; again, what the hell community has such bad moderation they need these features baked in? That's scary. It's not communal review; that happens at the level of posts, actions. When you're reviewing people, with generally no reason to suspect them of ill doing in a one-sided way that's called a witch hunt.

I agree with every word of Jeff's Answer. Except for "tandem". That's a silly word and I won't agree to it.

Answer 11

That's not good enough.

I strongly disagree, the current process is good enough:

• Every public action can be contested on Meta,
• Every user can email Stack Exchange and complain about a moderator's actions.

There are several moderator actions that need to be private. Those can only be judged by fellow moderators and Stack Exchange employees, how can you make an informed decision on actions you don't know anything about?

I'm all for more transparency, but changing the process to remove a moderator without radically changing our approach to moderation is not going to work. Right now we have absolutely no idea, for example, if the current process is working or not. Do we really have rogue moderators out there breaking stuff, or are we on a quest for transparency just for the sake of transparency?

An earlier call for increased moderator transparency was initiated by moderators, nothing substantial came out of it and Stack Exchange is visibly absent from that discussion. Yes, the Assembly exists, but 99% of discussions in there have absolutely nothing to do with moderation. Where are all the users that were supposedly interested in more transparency?

I honestly don't care for keeping the f'ing diamond, what I care about is about not having to jump through hoops to keep the site clean and help the community grow. If I'm doing a bad or even average job as a moderator, here I am, tell me all about it. In the few months I've been moderating Programmers I've received zero feedback on my actions. Granted, I never asked, but why would I? Stack Exchange claims to monitor us and respond to every complaint, if there was something going on, you'd tell me about it, right?

If there's actually a problem with rogue moderators, let's talk about it, openly and honestly. People are already assuming there is a problem, just because we are having this discussion, am I the only one who finds that extremely counterproductive?

I'm afraid we are trying to solve a problem that doesn't exist, and our time and energy would be far better spent in finding ways to minimize the gap between regular users and moderators and getting more people involved in policing the sites.

Answer 12

Dealing with Moderator transgression is a tough nut to crack. When you're talking about people who spend their day to day intervening in the site's workings and making decisions that at least one party doesn't like, you're going to have complaints.

If you didn't get complaints about moderators, I'd be really surprised. The question is, what should be considered a valid call to remove a moderator, and what isn't? Once you have something 'valid', then how do you handle it?

I propose something like the following:

1. Moderators should be made aware of the evidence against them in a private setting, preferably with a chat with the community team manager or their surrogate. They should be given a chance to respond to the accusations.

If, after talking to the moderator and viewing all the evidence, the community team believes it warrants further action, the following should occur:

2. The moderator is given a warning (much like we do with normal users). If it's annotated in the moderator's profile, it should be viewable by the Stack Exchange team and that moderator.

3. If the moderator screws up again, they should be suspended.

4. If they mess up again, then they've clearly been given enough rope to hang themselves, and should be ousted by the SE Community Management team. It should be a unanimous vote among the Community management team.

This is not a decision that democracy via the community can make: democracy makes for terrible due process (that's a reason why Juries have such a high burden of proof and the people are selected randomly).

I'm most interested in making sure the following:

1. The Moderator is protected from unfounded accusations.
2. Ousting is a serious as it can be; and shouldn't be the first or second step.
3. The community should know that the community team is looking out for them (a "Who watches the watchers?" sort of thing)

I know that personally as a moderator I've made decisions that were in line with our rules and guidelines but weren't popular at all; and in fact blew up in my face. I also know if I were ousted for that, then we'd have no way for moderators to learn at all; but if we let the mob handle these things, a lynching would be just what we see.

Answer 13

Without some idea of what Moderator problems you're seeing in practice, it's awfully hard to offer coherent suggestions here. It would be glib to suggest a 'recall' election. However, it seems as if there are, roughly, two cases:

1. The moderator violates his or her agreement with you. This is a legal agreement between SEI and the moderator. The community, if you ask me, has nothing to say about it. If he or she has violated the agreement, you do what you need to do. The community can send you email easily enough if they believe this to be the case.

2. The moderator is drifting more or less radically away from the 'theory of moderation.' One might see a recall election model here. Instead of a recall, why not make diamond status be a term of a year or two, stagger the terms, and let people just fail to get reelected?

Answer 14

As Rosinante already says, it's hard to make good suggestions without knowing the realities of moderation. But to put the thoughts from the comment above into an answer:

• Reviewing a mod's actions should continue to be handled internally. Under no circumstances should it become public in any way. The least productive users tend to be the loudest decriers of mod "tyranny", and those would surely occupy the stage if there were a public "recall" process of some kind.

• If the process needs codification, a "jury of your peers" model would sound attractive to me. If serious complaints about a mod come up, there could be a random drawing of, say, three fellow mods who are tasked with reviewing the accused's actions. What these fellow mods recommend, will happen - including removal of the mod if at least two reviewers vote so. A SE team member (with veto rights) could be required to review whatever decision is taken to prevent "corrupt police departments", cliques of moderators behaving destructively, from protecting their peers.

• A variation of this process, providing for even more neutrality, would be involving mods from other sites, but I'm not sure how practical that is - a gardening mod may not be able to truly grasp the culture on Math.SE, and judge the mod's actions properly.

• With the growing number of SE sites, there should be clear instructions how to report mod abuse to SE directly. As mentioned above, that channel will likely be mostly clogged with spurious accusations, but in the long run, something like this is probably necessary anyway.

Answer 15

Let me start by saying that there is one moderator in the system I wish was not a moderator. Decisions are made that I think are wrong occasionally, and the same name crops up over and over. I don't worry though, because they can all be overturned, and even if they aren't, in the end it's all just ones and zeroes.

The point I wanted to bring to the table is that reacting to bad situations generally leads to bad rules. That is, if someone does something you don't like and you make a rule not to do that, you end up (in quite a short time) with a bunch of bad rules that hamper nice people and that nasty people can still game. One of the big successes of the SE network has been the adaptability and generality of the rules.

I'd like to see peer pressure, both intrasite and intersite, control moderator behaviour. Nobody likes to be bad at something and if one mod is making bad decisions, being called out on Meta, seeing their decisions overturned and reversed both publicly and privately, and even being addressed by name (however temporarily - I understand personal attacks will be removed and I agree with that) surely that person would improve or quit. If not, surely the other mods would say something. Our voting mechanisms, even if restricted to 10k or 20k users, are unlikely to achieve anything that a conscience or a little prodding from peers could not. We don't really have the tools to be sure someone is messing up. Once the level of messup reaches the point where most trusted users feel the same way, it stands to reason that most moderators would feel the same way and could take action.

I feel that peer pressure should suffice even though, if a mechanism were invented for a recall, I would get involved in that process. If I am the only one ticked off with a mod it's nutsy to try to remove that mod. If enough 10k users are ticked off that they could "do something" well then I am sure enough mods are. (If not, then perhaps the mods know something the 10ks don't.) And it's easier (in the sense that you don't need to write any code or worry about gaming) to encourage mods to self-regulate than it is to try to implement something to guage the will of the 10k users.

Answer 16

We need a much more transparent and prescribed process for when — and more importantly how — to intervene when on-going problems are going unresolved.

To make the moderator removal process transparent, you will have to hold the investigation and "trial" publicly. This means that the moderator's moderation history must become open to public scrutiny. Otherwise, it's just status reports: "We've moved the process on moderator X to phase Y. Moderator Z was acquitted due to lack of evidence."

The only reason to make the process transparent is to establish faith that the system is bringing about justice. Faith that moderators are being investigated when there is credible evidence of wrongdoing/malfeasance. That the guilty moderators are being punished and that innocent moderators are not. To build that faith, you have to actually show the evidence. You have to let us the community vet the process. See whether the investigators went far enough. Look at the evidence as they saw it.

Otherwise, you're not establishing faith in anything. You're just saying that there is a system in place for something that was handled haphazardly before. Nothing is proven about whether it is a good system.

So the question is this: are the moderators ready to provide the ability for community users to vet them? Are they willing to give us access to their moderation history and various deeds, for good or ill?

Because without that, without the ability to see the evidence and weigh it ourselves, there's no real transparency in the process. If you're going to have a secret process, then keep it secret. A public process where all of the relevant evidence is private sounds rather like a kangaroo court.

Answer 17

I think the process (whether the current one, or a new to-be-determined one) should be transparently defined, but not transparently followed. That is, just as there is a 'Theory of Moderation' post, there should be a 'What Happens When Things Go Wrong' post, but I don't at all think that the 'general public' should be privy to who/what/when is going through that process at any given time.

Yeah, sure, people will end up noticing if a Moderator disappears off the radar, or suddenly loses their diamond. But just as we don't publicly announce suspension reasons, or SE employees being 'let go', I think we should respect privacy here. Except maybe - just maybe - in the case of a particularly personal 'bad thing' happening between a user and a mod. In that case, I think the user is entitled to some feedback about how the 'case' has been resolved.

I also think that if there is a 'user complaint' process, it should be completely separated from any 'moderator impeachment process': that is, there shouldn't be any automatic causal link ("I/we complained about moderator X, when does he get the boot?").

Answer 18

When a meta post is created that calls out a specific user and complains about his behaviour, I either edit the post to make it about a specific behaviour alone, or I shut it down. Focusing on specific behaviour instead of specific users makes those discussions far more constructive. Publicly calling out a specific user gets non-constructive rather quickly; such issues are better dealt with by moderators in private.

We shouldn't forget this lesson when dealing with moderators. Specific instances of moderator behaviour can and should be discussed on meta. But not focused on the user, but on the action he took. This is the part that can and should be discussed in public.

But the decision to dismiss a moderator should be with SE, based on the input of the community. I trust the SE community team to make a good decision. As a user, I trust them to take my complaints seriously when I think a moderator has stepped over the line. And as a moderator I trust them to defend my actions when I have acted appropriately.

Answer 19

The problem that you're facing is that there are few community managers and many many sites, with moderators in the triple digits, and you're not able to keep a watchful eye on everybody. Feedback mechanisms on the sites are currently buried in the footer, and it's not obvious to the casual user or even avid users that those feedback mechanisms are sent only to the Stack Exchange team.

When I browse the network using a browser I'm not logged in on, I noticed that there are some messages below each answer:

Was this post useful to you? [YES] [NO]

My thought is that something similar to this could be used for actions of moderation, worded in a way that's positive, doesn't make the user feel like he/she is saying the moderators all suck or worded in a way to encourage negativity, and is clear that the feedback is only seen by Stack Exchange employees.

What do you think about this action? [Good] [Fair] [Poor]

[Submit Feedback to Stack Exchange]

I suggest strategically placing such feedback links below each and every community or diamond moderator action. Here's why:

• By putting the links under actions performed by both the community and individual moderators, this appears to be more focused on the feedback than it is about pointing the finger at the moderator. It could also result in helpful data about how users feel about certain actions that are taken, regardless of who performed them.

• People get used to seeing the feedback links, which means they're more likely to be used than when they're buried in the Contact Us page.

Furthermore, if a user selects "poor" consider launching a modal window to capture a custom reason, just make sure the user has the option to submit the feedback anonymously, and make sure it's clear that only employees see the feedback.

There are some products on the market, like Hively and Get Satisfaction, that exist to capture and manage user feedback. We use them in our emails when reaching out to customers, and feedback goes to management. It sounds to me like some of the concepts from these tools could be evaluated and customized/tailored for Stack Exchange.

As far as process is concerned, this will help you become a dog with a bone. If enough poor feedback items come in anonymously, then that may indicate a potential problem that you as community managers may want to investigate. This can be done by reaching out to select users via private email, posting meta discussions, or simply spending some time on the site and observing for yourself how the community is doing.

Answer 20

If someone is acting in a way that damages the community of the site, there already is a team of people charged with taking care of it: the moderators of that site. If the "someone" damaging the community is a member of that team, that certainly makes the other moderators' job harder, but it doesn't change the job description, and these human exception handlers can turn to the community team for backup if they think the situation calls for it. Do we need better tools for the mod team to respond to a rogue one of their own? I honestly don't know, and am extremely thankful that the five of us over at Ask Different have had no reason to test that!!!

If a site user thinks that enough is not being done to protect the community from harm, whether that harm comes from spammers, ordinary users, or moderators, and the tools available to that user as an ordinary user are not enough, there is already a mechanism in place: the user can run for moderator in the next election, and if the fellow participants in the site think said user would contribute to improving the site, the user could become a part of the moderator team, and thus have more tools to respond to said threat to the site.

So if a moderator actually goes rogue, users can raise the issue through flags or posts in meta or chat and the other mods reign him/her in. If a moderator goes rogue and a user thinks the team isn't doing enough, the user can run for moderator next cycle and try to fix the problem. If a user is upset with a moderator but the other moderators don't agree with the flags and the community doesn't want to elect the user to the team, the process has rendered an answer (not the one the user wanted to hear, but an answer all the same).

The other possibility is that a user thinks the team of moderators has gone rogue. There isn't a mechanism in place to fix this one internally. Given the power of a team of moderators to shape the site (defining what's on topic through the FAQ, etc), I'm afraid this one has to be solved via exit rather than via voice. If you don't like the direction the entire team of moderators has taken a SE site, perhaps that isn't the SE site for you.

In other words, if any mechanism is needed for moderator removal, it's a mechanism for the rest of the mod team to be able to escalate the problem to the community team. Otherwise, it seems we have the needed tools in place already.

Answer 21

What about terms? Instead of serving forever, moderators serve for a set period and must then run for reelection to continue.

The idea has been brought up before, and the top answer then was that being elected for perpetuity was fine, but that abuse wouldn't be tolerated.

But now we're looking for a way to have a formal process to ask someone to step down.

Terms provide a graceful way to get rid of a moderator without having to ask them to resign. Everyone saves face, there's no problem with bias. It even gives moderators who've lost interest a way to bow out without much extra fuss.

Moderators looking to continue will have to go through re-election, but is that such a bad thing? Participating in the election would give moderators a chance to get feedback from the community about how they've been doing, and I'm sure that they'd have a very different perspective after having actually been on the job awhile.

---

Update:
A hypothetical case I'd encourage everyone to consider is the slightly bad moderator. If the only problem were egregiously bad moderators, it'd be simple enough to just revoke their diamond when they did something awful. The need for a process comes from edge cases like a moderator who generates a lot of complaints, but whose actions are always justifiable. E.g., a mod who deletes a rambling off-topic comment thread, and then leaves a not very tactful note about the deletion that irks the commenters. The sort of mod that leads other mods to say, "What they did was right, but I would have done it differently."

Having moderators stand for re-election handles this gracefully, many other proposals do not.

Answer 22

Have you considered using the upvote/downvote system for moderator actions as well? You already have the nuts and bolts in place. Moderators then garnish their own moderator-reputation, and those whose rep is declining or showing odd patterns automatically get their own feedback in this form, from other moderators I guess. I know this doesn't really answer the question but it at least provides a transparent metric, which addresses part of it.

Answer 23

What about doing periodic moderator renewals (not to be confused with periodic elections) where the community can vote on if their moderators should continue for another term or not?

This would provide the community with a (drama-free?) way of removing a moderator who is not doing their duties, while addressing many of the concerns I've heard about periodic elections, such as unnecessary drama, users taking "popular" stances instead of "correct" ones to complete against other users to hold their seat, and the extra work to put on the election.

It could be something as simple as a page reminding the community of a moderators duties, and letting the community cast a yes or no vote on if they think each moderator should continue for another term. Perhaps it could even include some stats of each moderator's activity for the past term to help the community make an informed vote.

If you wanted, you could also include a "Check all that apply" section if you vote No so users can specify why they think the moderator is not suitable to continue on (inactivity, abuse of powers, neglecting moderator duties, etc)

SE can use this information in addition to the non-public stats they have such as complaints or not-public moderator activity to determine if they should ask that moderator to step down or not. Just be sure to be clear that SE makes the final decision, and that the votes are only there to assist them in that decision and are not the only deciding factor.

There are a number of reasons I think this would be the best solution:

• Moderators will be judged by their performance and actions only

• Unlike something like periodic elections, moderators are not competing with other candidates, or having to spout popular/non-popular opinions to maintain their diamond

• The process would be anonymous on the user level (except to SE), but transparent to the community.

• You're not singling out any single moderator for action, which can lead to a lot of unnecessary drama/tension in a community

• A few very vocal "No" users can easily be outnumbered by the number of "Yes" votes since every user only has a single vote

• You are not waiting for emotional users to come to you with problems, but are instead going to them and asking for feedback. You're more likely to get a rational response this way, and will get the feedback from more users instead of only the users who have enough of a grievance to bother putting it into words and contacting you.

• It handles other reasons why a community may want a moderator to step down instead of just complaints, such as inactivity, or neglecting their duties as a moderator

To address comments about moderator stats, I know they can't be quantified easily by numbers but some numbers to indicate the moderators activity level over the past term would be helpful.

I would prefer something like "Average Monthly X for the past term" since it covers the period being reviewed, and is broad enough to get a good indication of the moderator's activity level without getting too detailed. X can be whatever it is you moderators do - flags handled, meta posts posted, etc. Just something to give the community a rough idea of the moderator's activity level to help them make an informed decision.

Answer 24

I think the best solution would be for this problem if this whole would be automatic.

Every mod could have an upvote/downvote list, and if it hits some limit, automatically a new election could be started (where also the mod could nominate).

In my opinion, it's not the mods who are the worst guys of any hostile site. The bad sites have mainly an "inner circle" of at least of 5 avid close voters. They are the bad guys. Mods I've met tried mostly to mediate between them and the rest of the community.

And those bad guys are there because the community building of the SE has somehow an unintended side-effect, a social selection mechanism, that somehow often results in not really friendly and helpful people getting to the top of the sites. The main problem is that once they've cemented themselves into a site, they can't be easily "de-throned", because they expel new members before they could become strong enough.

From the view of the CMs, it probably seems exactly the opposite: they see a stable site with a stable community, in continuous fights with recalcitrant people. If you prefer always the community consent, or majority opinion, then you ignore the opinion of the - maybe many - worthy members, who aren't already among us, because they were long expelled by this majority.

Users don't know how deeply you track the internal social mechanisms of the communities. Sometimes I see signs that suggest you know about the problem and are trying to fix it, but those signs may be an illusion, too.

In my opinion, investing a lot of effort into the mods and ignoring the rest is not enough, somehow you should find a way to produce a balancing mechanism between the "inner circles" and the normal users.