The new code of conduct contains the following in the enforcement section:

Account Expulsion

For very rare cases, moderators will expel people who display a pattern of harmful destructive behavior toward our community.

Until now, moderator guidance did not include expulsion as a potential method to deal with users. The term expulsion has never even been mentioned before with regards to SE moderation as far as I'm aware.

What changes does this section mean for us moderators? In which cases are we supposed to expel user, and how does it fit into the old 7/30/365 days escalation for suspensions? Is expulsion simply the next step after the 1 year suspension?

What does expulsion actually mean? Are we supposed to simply delete users? And if yes, is that actually reasonable as it complicates handling circumventing their expulsion?

Is account expulsion site-specific or network-wide? Severe misbehavior can result in network-wide consequences, which might be useful to mention in the enforcement section of the code of conduct.

And why are we suddenly introducing new terminology here when we have existing, and in my opinion easier to understand terms like suspension and account deletion?


1 Answer 1


You'll probably want to involve us, is the shortest answer that I can give.

Situations that result in us asking folks that have made positive contributions to leave are incredibly rare, and the dynamics leading to them tend to be pretty unique. You just sort of arrive at a point where you plainly can't envision a future with them participating happily without continuing to create an intolerable amount of disruption.

Expulsion is exactly what it sounds like, informing the user that they're no longer welcome to contribute to the site. This isn't new, in fact, we put a deliberate reason for it in the 'delete' dialog back when we overhauled it:

moderator flag dialog

This user is no longer welcome to participate on the site is probably the least used reason, but it's the one you'd use if someone simply has to go.

It's a good idea to talk to a community manager prior to taking this action.

We're not going to second-guess you; you're the one closest to the situation and you probably know way more about the dynamic than we could learn in a reasonable amount of time. What remains is us helping you figure out the most ideal way to part ways, which often means helping with the communication, or other accommodations we might be able to make in order to help the user accept your decision.

We'll also need to be around in order to help you manage communication on your meta site when questions come up, especially if the removal is felt on multiple sites. As you may or may not know, we'll need to manually process deletion of users with more than 500 rep anyway, so it's nice to know it's coming and we have some context.

To answer your questions directly:

What changes does this section mean for us moderators?

None really. There's always been cases where moderation teams have simply exhausted all means and patience with a particular user, and escalate it to us, at which point we inform the user that they're no longer welcome on a site.

It's not common, but it does happen.

In which cases are we supposed to expel user, and how does it fit into the old 7/30/365 days escalation for suspensions?

The majority of these would be after the first year-long suspension, but there could be cases where things escalate more quickly. For instance, if a user made threats after being suspended for 30 days, we'd delete them and take measures to prevent their return.

What does expulsion actually mean?

Gone; no longer able to participate on a given site. The implementation of this might look like a super-long network-wide suspension, or it might just be the removal of their account - it depends. But they're losing all privileges to participate without hope of regaining them in the future.

Are we supposed to simply delete users? And if yes, is that actually reasonable as it complicates handling circumventing their expulsion?

That reason from the delete dialog that I highlighted does some additional magic (and that's all I'm going to say publicly), but please coordinate it with us if possible. "If they cross the line again then you'll probably have no choice but to expel them" is a good example (if said by us) of when you'd use that option more autonomously.

On Stack Overflow, we might not get consulted as often, as there are quite a few more users on their 'last chance' than anywhere else.

Is account expulsion site-specific or network-wide?

It would depend. Usually (and ideally) site-specific, e.g. cooking just brings out the vinegar in some people as a humorous example. But there might be cases where people just completely come unhinged. It's better to coordinate with us (though we do leave the option available).

So, it's not new, it just hadn't really been alluded to previously, and it's something that you can point to if you need to inform a user that you really need them to demonstrate that a path to them continuing successfully exists. Don't make threats, but at the same time, we're not going to hide the fact that we do have a last resort.

Fortunately, as I said, it's pretty rare. I'll answer any more questions in comments, if anyone has them.

  • Are users allowed "clean starts"? For example, if I do bad things with my account and get myself suspended or expelled, and then a few years later have a constructive question that I want to ask on a site, can I create a new account to ask it? Will I still be disallowed, even if I don't repeat the same behavior that got be removed initially? Aug 7, 2018 at 19:13
  • Well then, could you please write down somewhere in the rules what do you consider "intolerable amount of disruption", using pretty specific metrics? For example, "causing our moderators to spend 1 hour per day dealing with events arising from particular user behavior is intolerable". Because everything is relative and changes with time, and I'm pretty sure there are people out there who will want to consider "leaving comments that caused 5 comment flags to be raised" to be intolerable level of disruption.
    – artem
    Aug 7, 2018 at 20:56
  • 2
    @artem It's so rare and every situation so unique that I don't think we could yet distill out commonalities beyond "Try not to get suspended 9+ times for the same thing, and don't threaten to physically harm staff and moderators" which is so oddly-specific it would only really be useful as morbid decoration. We can update the CoC to be more specific, and we may include more language about threats if it becomes a more consistent problem (in almost 10 years it hasn't). It's not really all that relative, these are truly exceptional cases.
    – Tim Post
    Aug 8, 2018 at 13:14
  • @SonictheInclusiveHedgehog Hard to say. If you returned and didn't exhibit that behavior, nobody would notice or care, so .... maybe you'd get by? Maybe we'd allow them back if it was obvious that they're not a problem any more? I can't say, it hasn't really happened yet.
    – Tim Post
    Aug 8, 2018 at 13:15
  • @Tim when it comes to threats to do harm it's on entirely different level, what I had in mind was non-criminal behavior that people find extremely annoying to the point of being intolerable. If you say it's rare let's hope it will remain so.
    – artem
    Aug 8, 2018 at 15:59
  • @artem If it doesn't remain that way, then we should also have a better idea of what it looks like (and how to describe it) in a general sense, or perhaps we've stumbled on some bit of UX causing folks to do something we didn't anticipate. But we're increasingly (and acutely) aware of that which repeats itself, so I think we'll be okay. We don't get too far into preemptive language because at current length + 1/3, many would consider the CoC onerous to read; that's why we're just watching to see what else absolutely needs to go in.
    – Tim Post
    Aug 8, 2018 at 16:09

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .