RE: Comments on draft-farrresnickel-harassment-01 - A mostly 'NO' view

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



[snup]

> >>The above pretty well meets the definition of "Star Chamber".  Strict,
arbitrary
> >> and secretive.

Part of the reason I am co-authoring this work with Pete is that when the IESG
was making its original statement on harassment I went off the deep end about a
number of pieces of the process some of which coincide with the concerns Mike is
expressing, but I find the reference to Star Chambers is melodramatic and
implies that the ombudsperson has something to gain from producing a specific
result.

> >Generally speaking, there's no appeal chain with a Star Chamber. There's no
> > possibility for the firing of its members. And I'm not at all sure where
you're
> > getting "strict" or "arbitrary" from.
> 
> Strict - this ombudsman can impose the highest penalty available - expulsion,
and
> in fact that may be the only penalty it can impose.

The "highest penalty available" certainly does conjure up heads on gateposts,
but I think it is also seeing "remedy" as punishment (or, at least, so the word
"punishment" implies to me).

Perhaps we can work on the text for this because it is not my intent that the
system invoke punishment. In the same way that a PR Action is not used as a
punishment (it is used to enable the list to continue to function), I see the
remedies applied by an Ombudsperson as necessary to insure the harassment-free
environment for other IETF participants and no more than that.

...but I have now read your closing paragraphs and I see the distinction. I
think a remedy arising from mediation might not need to be "imposed" but would
still be a penalty of sorts. I need to think about this point further.

I also don't quite get why you say this is the only penalty that can be imposed.
Are you saying this because other "penalties" can be refused by the respondent?
I guess that refusing a remedy does escalate things, but there is an implication
that the remedy is handed down without discussion. i don't think this is how
Ombudspersons work.

> Arbitrary - e.g. not predictable.  Not based in an established procedure and
due
> process.    The ombudsman creates their own process and maintains it rather
> than running a process defined for them.  That process - obviously - can
change
> with the whims of the ombudsman.  (section 3 first paragraph)

OK, there is text missing on the process that the Ombudsperson should use.
We got some professional advice that the current text does not capture well, and
we also decided that having the IETF (probably non-expert in this realm) develop
the processes would be an unprofitable use of time. So:

- Ombudsperson should develop general processes
- Ombudsperson should discuss those processes with IESG
- Ombudsperson should publish those processes 
- Discussion is welcome from the IETF community, but these are
   processes that do not need IETF consensus
- Significant objection to the processes that are not addressed
  should result in an RFC 2026 Section 6.5.3 appeal and/or a
  recall petition against the IETF chair (and the rest of the IESG
  if you like) assuming it is the IETF chair who is the hire/fire
  of the Ombudsperson.
- Repeat with changes to Ombudsperson. 

> >Any process we create where we give people authority and allow (or require)
> > them to their keep discussions confidential, whether it's the IAB, the IESG,
design
> > teams, NomComs, etc., there exists the potential for abuse. We set up some
> > guidelines for how the bodies are to behave, try to appoint reasonable
people,
> > put in what reasonable checks and balances we can, and hope for the best. If
> > things go off the rails, we go back and rethink.
> 
> You mean after we damage someone's business reputation unjustly?    
> You keep missing the point that there needs to be protections not only for the
> subject/reporter, but for the respondent.

I don't miss this at all. I think it is critical.
There are four elements of protection...

He-said/she-said cases are notoriously hard to handle and form a key part of
Ombudsperson training. I think it highly unlikely that an Ombudsperson imposes
drastic remedies without more than substance.

Not all remedies are of the sort that would damage someone's reputation, and how
they handle the remedy can make all the difference.

The secrecy that you objected to earlier actually serves to protect the
respondent. They are able to discuss with the Ombudsperson an appropriate
response/remedy without their case or the accusations (true or false) being put
into the public domain.

We have retained text about very rapid appeal handling in cases that arise
during IETF meetings. There was opposition to this speed, but I think it
important that someone who is asked to leave a meeting should be able to have
their appeal looked at while the meeting is still going on.

I would go on to say that a false accusation is itself harassment.

> >>Other issues:
> >>
> >>Can the ombudsman remove anyone without restriction?  E.g. IAB, IESG, IETF
> >> chair, IAOC and IETF Trust?   Or does this only apply at the WG chair level
on
> >> down?  Or only for non-wg chair participants?  I ask this because in every
case
> >> except the general participant, we have defined ways to remove someone and
it
> >> seems problematic to allow a single individual to remove an AD for example.
> >> There are also legal issues with removing a trust member I would expect.

I sincerely hope that the procedures apply equally to ADs. Without wanting to
appear to be making light of this, the next time I sexually harass someone in an
IETF meeting, I hope that the Ombudsperson deals with me as they would deal with
any other person.

You say that we have defined ways to remove people: we don't. We have ways to
recall ADs, that does not stop them from being in a position to harass someone.
"Remove" would appear to be the problematic word. But, I don't want someone I
have sexually harassed to have to explain to a wide community what I did and how
in order to recall me, I want to be treated like any other IETF participant and
held to the same standards.

> Going back to your example of having someone harassed in the hall, you expects
must
> to make an immediate decision (based only on the report?) and pull an AD or
> working group chair?   What happens when - as I expect - you are ignored?
And
> why wouldn't the investigation take time past the end of the current meeting
to
> resolve itself?  (More argument for the word "arbitrary").

Maybe it will take longer to resolve. So? does the text imply that all event be
resolved within a 30 minute episode? If so, we should fix the text.

I do, however, expect that all members of this community would take very
seriously any accusations made against them and would want to talk to the
Ombudsperson as soon as convenient to clarify what happened.

False accusation. Immediate response "He said what? When can I come and talk to
you about this?"

> The term used in the document is exclude - I would expect that to mean that
co-
> incident with removal from the meeting, the person would lose their position
> within the IETF - including the IETF trust.  I would wonder if such removal
would
> conflict with the provisions of the trust.

That is an interesting question. We have not considered removal from IETF
positions. I don't think the Ombudsperson has that power even under this
document. You might argue that an AD will find it hard to continue as AD if they
are excluded from IETF meetings. I might agree, but that is not the same as
removing them from post.

> >> Can the ombudsman get involved with issues of harassment that do not
> >> implicate participation in the IETF?  What's the threshold?
> >
> >I don't understand the question.
>
> Person A and Person B dated.  It ended badly.  It ended a while ago. It had
> nothing to do with the IETF. Person B claims that the mere presence of Person
A
> at the IETF constitutes harassment because it cause them distress and they
can't
> participate in the same working groups.

If someone believes they are being harassed they go to the Ombudsperson.
If the Ombudsperson believes harassment took place within the context of the
IETF they act.

If you were the (untrained) Ombudsperson, how would you react to this situation?
What makes you think that a trained Ombudsperson would react differently?

> >>Will the evidence and the witnesses be made available to the Respondent?
> >
> >The document makes no mention of "evidence" or "witnesses". In the "detailed
> > investigation" section of the document, the Ombudsperson is expected to
> > discuss the circumstances of the situation with all parties. You seem to
presume a
> > formal proceeding that does not exist in this document.
> 
> Goes back to "arbitrary" - if there is no evidence or witnesses, this is a
they
> said/they said argument, and this document seems to imply that all ties go to
the
> subject.

No it doesn't. If you want to make that assertion stick you are going to have to
supply quotes that we can then fix.

And I think your interpretation of process may be inhibiting your view here.
This is closer to an arbitration process than a court of law. 

Sit down with Ombudsperson, talk through what happened, listen to what the
Ombudsperson has to say about what they have heard, say your piece, ask to go
away and think about it, come back discuss some more, etc.

> >>Will there be an advocate assigned to represent the interests of the
> >> Respondent?
> >
> >Again, this seems to presume a formal proceeding that does not exist in the
> > document.
> >
> >>What happens if/when the Respondent declines to participate and instead
> >> brings the issue into the public stream?
> >
> > If/when this happens, it will be unpleasant. However, it has yet to happen
in the
> > current state of affairs, where ADs have been asked to mediate these things.
> 
> Mediation is a lot different from a process where there is the possibility of
> penalties being applied.

We're back to the word penalty, but that aside, I think the possibility of the
respondent refusing to participate is very real. I think there are three
categories (or more):
- Frightened (with or without reason) of the process and outcomes
- Rejecting the authority of the Ombudsperson
- Not wishing to face what happened

We did write...
   The Respondent is not obliged to cooperate, but the Ombudsperson may
   consider failure to cooperate when determining a remedy (Section 5).

> >I don't see why the existence of an ombudsperson changes the possibility that
> >someone might make bring such things to the public. Unless what you're saying
is
> >that once a remedy is imposed that is not to the Respondent's liking, the
> >Respondent might go public. I suppose that's possible. And it will be
unpleasant.
> >But I don't see the Respondent having any particular interest in doing this:
If
> >someone is asked to leave a meeting, with the confidentiality maintained,
they
> >can simply say that they are leaving due to a personal situation. Going
public
> >means that they reveal that they have been asked to leave due to an
accusation
> >of harassment. That seems far less pleasant.
> 
> Ok - now we get further to the problem.  You're going to ask someone to leave
> based on an accusation?  Not even going to investigate?

Now really I don't think Pete said that.

Pete said that the issue can be resolved without going public. They can leave
the meeting without a public statement "Adrian is being asked to leave the
meeting because he did such and such." 

If the whole thing is public then it is harder for me to keep my reputation. But
sure, nothing can stop me from writing a blog, Tweeting, or taking out full page
adverts in the New York Times. Presumably, if I thought I had been unfairly
sanctioned (and my protestations of this to the Ombudsperson had failed to have
an effect), and I didn't fancy appealing or my appeal failed, I could go public
and make sure everyone knew about it.

> >>Or alternately, sues the ombudsman, personally, for defamation and business
> >> interference?
> >
> >As I said earlier, my understanding is that the Ombudsperson will be covered
by
> >our insurance. And again, this assumes that the information about why the
> >person is leaving is public, which the Respondent would have to do themself.
> 
> Not for business interference.

I can see loss of business. But not loss of reputation.

I sincerely hope that people who believe they have been unfairly treated will
sue.
I hate courts and lawyers and people who have friends who play lawyers on TV,
but the court system exists to rectify things that have gone wrong in society.

So, from that perspective, suing is good. It will significantly stress the
confidentiality side of it, but I also believe enough in the Ombudsperson
training and processes that I doubt that any Ombudsperson will find themselves
in court for this.
 
And note that the text does say
   In determining the appropriate remedy, the Ombudsperson may
   communicate with the Reporter, Subject, and Respondent in order to
   assess the impact that the imposition of a remedy might have on any
   of those parties.  However, the Ombudsperson has ultimate
   responsibility for the choice of remedy.
So it is unlikely that the additional "cost" to the Respondent has not already
been aired and considered.

> In any event, AFAIK our insurance covers us as long as we make a good faith
> effort to follow our own procedures, and as long as those procedures are not
> arbitrary and do not lead to capricious results.    Part of the issue with
asking
> questions of lawyers, is making sure you give them all the data.   Provide
this
> document to the lawyers, to the ISOC HR department and to the underwriters
> and have them get together and say they're fine with this specific piece of
prose
> if you really want to ensure protection.

Agree with you there.
But maybe it is better to stabilise the text before spending lawyer dollars.

> >>What's the decision threshold?

[snip]
 
> What I'm pointing to is exactly what you said above  - "left to the discretion
of the
> ombudsman".  There is no bright line (or even fuzzy line) to guide the
> ombudsman as to what constitutes actionable harassment.  There is no due
> process where the Respondent is guaranteed an opportunity to speak and, if
> they choose not to speak its held against them.

I think that the text does not seek to exclude any of this, and I hope that the
establishment of a documented process by the Ombudspersons will resolve these
issues. But I continue to believe that the IETF lacks the expertise to write
such a process itself.

> >>You mentioned good faith complaints.  What about bad faith?  Are the
> >> expulsion penalties applicable to a bad faith reporter?
> >
> >I'm not sure what you're asking: Are you asking whether, if the Ombudsperson
> >determines that a report was made in bad faith, can they impose a remedy on
> >the Reporter for having done so? I don't see how given the description in
section
> >4.
> 
> So what exactly is the downside for making a bad faith report?  Especially one
> that results in the expulsion of the respondent?

Well, it would be harassment, wouldn't it?

> >>You mention that the Ombsudman can consider failure to cooperate on the
> >> part of the Respondent in determining remedy (shouldn't that be
culpability?)
> >
> >Shouldn't *what* be culpability? The Ombudsperson determines a remedy,
> >*not* culpability.
> 
> The Ombusman should determine if there is actually an issue to be remedied and
> the specific actor(s) responsible for the issue (e.g. culpability) before ever
> determining that there needs to be a remedy.

The distinction is that failure to cooperate may influence the remedy not the
assessment of culpability.

> >>but from a legal standpoint it could be stupid for a Respondent to cooperate
as
> >>any discussions in this context could provide fodder for a real world legal
suit
> >>regardless of the facts of the matter.  Is there a fairer way of stating
this?

Cuts both ways.
Failure to participate in early remedy process can go against you in court.

If you are doing things in the IETF that might be subject to legal action, then
the right resolution might be for legal action to be taken against you. So
resolving it through the Ombudsperson might be a good idea.

[snip]

> I'm not saying there isn't a problem.  I'm saying that giving a random person
the
> ability to exclude someone from the IETF without strong guidelines on getting
to
> the exclusion decision, and without strong protections for the person subject
to
> exclusion is JUST PLAIN WRONG and unfair in the extreme.
> 
> Set up the ombusman as a resource for the IETF leadership and participants.
> Make it a counseling and mediation stop.  Make it confidential and covered by
> the legal "counselor" privilege and make the participants sign a
confidentiality
> agreement.  Do this with professionals (work through ISOC to set up something
> like an employee assistance program for this specific issue).  Don't give it
the
> ability to exclude someone from the IETF or craft "remedies" - make it about
> resolving the base problem.
> 
> If the mediation fails, then resolve the issue in the public channels, with a
defined
> and well known due process, protection for the accused, compassion for the
> accuser and a goal that's in keeping with first protecting the standards
process.

Thanks for the time and effort you've put into this.

Adrian





[Index of Archives]     [IETF Annoucements]     [IETF]     [IP Storage]     [Yosemite News]     [Linux SCTP]     [Linux Newbies]     [Fedora Users]