[coc-tf] Balance between between judgement and process (was: Re: (no subject))

Dear TF,

I'd like to encourage you to comment on this updated straw man
proposal. Please share your thoughts on what is right and where we
should make improvements or fill gaps.

Thanks,

Leo

On Wed, Feb 16, 2022 at 2:39 PM Leo Vegoda <leo at vegoda.org> wrote:
>
> Hi,
>
> Thanks for the input on the straw man proposal I shared last week. I
> have an updated straw man proposal below.
>
> I have used capital letters as per BCP 14 to reduce ambiguity.
>
> This updated straw man proposal needs more work.
>
> My updated text mentions reporting requirements but it does not say
> what must be done with the reports to the RIPE Chair and RIPE NCC MD.
> It literally asks what we should recommend about patterns of
> behaviour. Please make suggestions!
>
> Similarly, while we've discussed a process for recruiting volunteers,
> this straw man proposal does not have anything about how to handle
> situations in which the Code of Conduct Team gets things wrong beyond
> correcting them for the person who requests the escalation. What do we
> do if a person on the Code of Conduct Team handles cases poorly?
> Similarly, while an escalation request is always allowed, are we happy
> to allow the next step on the escalation path to refuse to consider
> the escalation?
>
> Once we close in on what we want we'll need to improve the language,
> which I think is a bit hard to parse at the moment. My apologies for
> that.
>
> Thank you for feedback,
>
> Leo
>
> 8<---- START UPDATED STRAW MAN PROPOSAL ----->8
>
> The TF recognises the need to balance goals when evaluating Code of
> Conduct reports. This is similar to how the RIPE community balances
> competing goals for the allocation and registration of Internet Number
> Resources. Sometimes we must rebalance these goals. We need to do the
> same thing when it comes to the implementation of our Code of Conduct
> published as ripe-766. That is why we propose that the implementation
> should be reviewed after one year and then every two years:
>
> Key factors to consider include:
>
> - Can we implement the Code of Conduct in a way that sustains
> participation in the RIPE community?
> - Can we recruit volunteers to serve on the Code of Conduct Team?
> - How confident must the RIPE community be in the systems used to
> report incidents and record investigations?
> - How do we want to link the medium of the incident with the
> consequences? For instance, could the Code of Conduct Team suggest
> barring someone from attending a RIPE meeting for posts to a mailing
> list?
> - How can the RIPE community provide oversight of the Code of Conduct
> Team while maximising confidentiality and enabling an escalation
> process?
> - Given the community's resources, what is the right balance between
> oversight and process?
>
> The balance we suggest in the first year of implementation includes
> the following. Some of these recommendations will need to be adjusted
> after the first year, as the Code of Conduct Team builds experience.
>
> - Team members MUST recuse themselves if they have a conflict of
> interest, such as a business or family relationship with one of the
> people involved in the reported incident.
> - When an incident takes place at an event, the Code of Conduct Team
> CANNOT suggest excluding the subject of the report from participating
> in the RIPE community after the conclusion of the event.[0]
> - When an incident takes place over the Internet, the Code of Conduct
> Team CANNOT suggest excluding the subject of the report from
> participating in the RIPE community for more than one month.
> - The RIPE NCC person supporting the Code of Conduct Team MUST provide
> the Code of Conduct Team, the RIPE Chair, and RIPE NCC MD a
> statistical summary of reports at the end of each calendar quarter and
> shortly after the conclusion of any RIPE event.
> - The statistical report MUST identify patterns of behaviour by an
> individual [AND THEN WHAT?]
> - The RIPE Chair MUST publish a statistical and thematic report as a
> RIPE document each year and MUST publicise the report to the RIPE
> community.
> - The RIPE NCC person supporting the Code of Conduct Team MUST provide
> both the RIPE Chair and RIPE NCC MD a case summary of any report that
> resulted in the exclusion of a person from an Internet service (e.g.
> mailing list posting rights removed) for more than one week, or
> exclusion from the remainder of a RIPE event.
> - People [THIS MEANS ANYONE AND EVERYONE] involved in a reported
> incident MAY escalate the case to the remainder of the Code of Conduct
> Team (barring those who recuse themselves for a conflict of interest).
> - The broader Code of Conduct Team MAY review the facts of the report,
> the suggested consequences, and the process executed by the initial
> team.
> - People involved in a reported incident MAY have that decision
> reviewed by the RIPE Chair Team[1] and the RIPE NCC MD.
> - While people involved in a reported incident MAY escalate a case,
> there is no requirement for the escalation request to be accepted.
>
> Example 1:
> The Code of Conduct Team is asked to investigate an incident at a RIPE
> meeting. The sub-team investigating the report (two trained community
> members, plus one RIPE NCC support person) suggests that the subject
> of the report should be excluded from the rest of the RIPE meeting.
>
> These suggested consequences will be included in a quarterly
> statistical report and a summary of the event will be provided to the
> RIPE Chair and RIPE NCC MD.
>
> Anyone involved in the reported incident MAY escalate the case to the
> rest of the Code of Conduct Team and request a reevaluation of the
> aspect they disagree with.
>
> Question for the TF: should the suggested consequences be implemented
> while an escalation is underway, or should implementation be deferred
> until after the conclusion of the escalation process?
>
> Example 2:
> The Code of Conduct Team is asked to investigate repeated ad hominem
> attacks made on a mailing list. The sub-team investigating the report
> (two trained community members, plus one RIPE NCC support person)
> suggests that A) the postings should be removed from the archive and
> B) the poster should be barred from posting on that list for a
> calendar month.
>
> These suggested consequences will be included in a quarterly statistical report.
>
> Both the poster and the subject of the ad-hominem attack MAY escalate
> the case to the rest of the Code of Conduct Team and request a
> reevaluation of the aspect they disagree with.
>
> [0] Presumably, the RIPE NCC as the legal entity running events and
> providing the community with Internet services could refuse to sell
> meeting tickets or accept e-mail from someone in some situations.
> [1] If this is the RIPE Chair Team the panel can be an odd number and
> the balance can be held by the RIPE community.
>
> 8<---- END UPDATED STRAW MAN PROPOSAL ----->8
>
> On Wed, Feb 16, 2022 at 4:07 AM Niall O'Reilly <niall.oreilly at ucd.ie> wrote:
> >
> > On Fri, Feb 11, 2022 at 10:14 AM Leo Vegoda leo at vegoda.org wrote:
> >
> > Dear TF and RIPE Chair Team,
> > Thank you all for a very interesting discussion earlier today.
> > The action item we took away for us all to think over was the balance
> > between the authority delegated to the team that performs the initial
> > investigation in suggesting a consequence and oversight by people with
> > broader responsibilities. Greater oversight at an earlier stage could
> > reduce the options for escalation. In other words, what is the balance
> > between judgement and process?
> >
> > I think it would be useful to think about the problem along other “axes”
> > too, including:
> >
> > oversight (as mentioned in the action item);
> > context (RIPE meeting, webinar, mailing list, …);
> > documentation (case dossier);
> > communication/notification;
> > limits on the severity of the immediate sanction;
> > possibility of recommending additional sanctions;
> > protection of the appeal process.
> >
> > My suggestion is to avoid oversight, as I feel it is likely both
> > to undermine the authority of the assessment team and
> > to compromise the appeal process. Besides, as Brian has pointed
> > out, it introduces the possibility of cronyism at a higher level
> > of authority.
> >
> > I think the combination of (a) limiting the severity of sanction for
> > which the assessment team has summary competence and (b) allowing them
> > to recommend additional sanctions would allow both effective immediate
> > intervention and a means to deal appropriately with more egregious
> > offences.
> >
> > For example, for an incident at a RIPE meeting, the severity of the
> > immediate sanction might be limited to exclusion from the remainder
> > of the meeting. On the other hand, lifetime exclusion would be a
> > sanction which the assessing team could only recommend.
> >
> > I haven’t thought about appropriate matching between the context
> > of the incident and that of the applied sanction. In case of an
> > incident on a mailing list, should the sanction extend beyond that
> > mailing list? And conversely, for an incident at a RIPE meeting,
> > should mailing-list access be restricted or even denied?
> >
> > In case it may be useful, and encouraged by Brian’s reference to
> > Rugby, here are links to the current edition of the document I
> > mentioned on Friday; the French edition, if you can read it, is
> > clearer on some points.
> >
> > https://static.fie.org/uploads/26/131735-technical%20rules%20ang.pdf
> > https://static.fie.org/uploads/26/131727-Technique%20fra.pdf
> >
> > In particular, article t.137 makes clear the distinction between
> > sanctions which a referee may impose immediately and those which
> > they must recommend to a higher authority.
> >
> > I hope this helps a little.
> >
> > Niall