search

w3c / process

W3C Process Document
https://www.w3.org/policies/process/drafts/
170 stars 120 forks source link

Ambiguity about (super) majority thresholds: of those voting, or of those eligible to vote? #838

Closed frivoal closed 2 months ago

frivoal commented 3 months ago

There are a few couple of places in the Process which define a decision as being made by a majority or a super majority, without being clear about whether we consider the (super) majority of those who did vote, or of those who were eligible to vote regardless of whether they did. Some are also ambiguous about whether we how we count explicit "abstain" votes. We should disambiguate. The appropriate answer might differ case by case, and we might need to split this issue if some of these points turn out to be contentious, but here's a first pass at it.

  1. https://www.w3.org/Consortium/Process/Drafts/#ABParticipation:

    The Chair(s) are subject to ratification by secret ballot by two thirds of the AB upon appointment.

    I think the intent is two thirds or more of those eligible to vote (i.e., AB elected participants). Clarification might not be completely necessary here, but we could go with

    […] two thirds of the ABAB elected participants […]

    to make sure there is absolutely no doubt.

  2. https://www.w3.org/Consortium/Process/Drafts/#TAG-appointments

    The Team's choice of appointee(s) is subject to ratification by secret ballot by both the AB and the TAG, each requiring a two-thirds approval.

    This is the one I am least sure about. I'm tempted to go with ratification failing if 1/3 or more of those eligible to vote explicitly vote against.

  3. https://www.w3.org/Consortium/Process/Drafts/#council-participation

    each dismissal is decided by simple majority of those not abstaining.

    I used to think this was not ambiguous, because I thought that "abstaining" meant "not participating in a vote", but I have been told that native English speakers take it to mean "explicitly decline to take sides". I think we should go with

    each dismissal is enacted if there are more ballots for than against

  4. https://www.w3.org/Consortium/Process/Drafts/#council-deliberations

    However, if despite careful deliberation a W3C Council is unable to reach consensus, the W3C Council Chair may instead resort to voting. In that case, the decision is made by simple majority, with the W3C Council Chair breaking any tie.

    Suggestion:

    […]the decision is made by simple majorityapproved if there are more votes in favor than against[…]

Note: Here are other cases of majority or super majority, which in my view, are already non ambiguous.

  1. https://www.w3.org/Consortium/Process/Drafts/#council-delegation

    A decision to delegate must be supported by a two-thirds supermajority vote (i.e., at least twice as many votes in favor as against).

  2. https://www.w3.org/Consortium/Process/Drafts/#ABParticipation and https://www.w3.org/Consortium/Process/Drafts/#tag-participation

    Chair selection must be run at least at the start of each regular term, as well as when a majority of the participants request it

TallTed commented 3 months ago

For (1), Upon appointment, the Chair(s) are subject to ratification by secret ballot, requiring approval by two thirds of the elected participants.

fantasai commented 3 months ago

For completion, here's the other options:

Upon appointment, the Chair(s) are subject to ratification by secret ballot, requiring approval by two thirds of the elected AB participants.

Upon appointment, the Chair(s) are subject to ratification by secret ballot, requiring approval by two thirds of the elected participants of the AB.

TallTed commented 3 months ago

For (3), each dismissal is decided by simple majority of those voicing a preference

css-meeting-bot commented 3 months ago

The Revising W3C Process CG just discussed Clarifying supermajority votes, and agreed to the following:

The full IRC log of that discussion <fantasai> Subtopic: Clarifying supermajority votes
<fantasai> github: https://github.com/w3c/w3process/issues/838
<fantasai> florian: This relates to what Chris was talking about earlier: when we talk about a majority, what do we mean? Of the people who voted? The people who could vote? how do we count absentees or those who explicitly abstain?
<fantasai> ... I did an audit of the Process
<fantasai> ... 2 are unambiguous
<fantasai> ... 4 are ambiguous
<fantasai> ... The one about TAG appointments is probably most contentious, and might want to deal with later
<fantasai> ... but others not so hard, so let's look
<fantasai> florian: 1st one is borderline, not quite ambiguous
<fantasai> ... about chairs of the AB being subject to ratification by 2/3 "of AB"
<fantasai> ... could say "of elected AB participants" to clarify e.g. is the chair included or whatever
<TallTed> q+
<fantasai> plh: how do we count the CEO?
<plh> ack ta
<fantasai> cwilso: not a member of the AB
<fantasai> TallTed: sentence as written is differently broken, "upon appointment" is in the wrong place
<fantasai> ... would shift and add a comma
<fantasai> fantasai: is "AB elected participants" or can we rely on member?
<fantasai> florian: Chair might not be an elected participant
<fantasai> plh: Proposal to accept the changes we're suggesting here?
<fantasai> fantasai: "AB elected participants" feels weird, reads like "AB-elected participants"
<fantasai> TallTed: can AB be implied, as for chair?
<plh> "Upon appointments, the Chair(s) are subject to ratification by secret ballot by two thirds of the elected participants of the AB."
<fantasai> florian: I think you're right we can go for "elected particpants"
<fantasai> RESOLVED: Accept Ted's proposal in https://github.com/w3c/w3process/issues/838#issuecomment-2022931990
<fantasai> florian: wrt "
<fantasai> each dismissal is decided by simple majority of those not abstaining.
<fantasai> apparently English-speakers think that abstaining doesn't include people who don't vote
<TallTed> `each dismissal is decided by simple majority of those voicing a preference`
<fantasai> which is different from French meaning
<fantasai> cwilso: ????
<fantasai> florian: Quick check of dictionary seems to uphold that point
<fantasai> cwilso: So rather than arguing what it means, then we should make it more explicit
<TallTed> `each dismissal is decided by simple majority of those voicing a preference; a tie fails`
<fantasai> florian: if the number of "for" ballots vs "against" ballots are equal, in my wording it fails, ted's is ambiguous
<plh> "each dismissal is enacted if there are more ballots for than against'
<cwilso> +1
<fantasai> plh: objections?
<fantasai> fantasai: if we're so close that we're tied, probably we should dismiss that person
<fantasai> though that's not fixing an ambiguity :)
<fantasai> florian: Current phrasing is ambiguous, actually, so we need to pick
<TallTed> "each dismissal is enacted if there is at least one more ballot for than against"
<fantasai> "each dismissal is enacted if at least half of the ballots are for than against"
<fantasai> "each dismissal is enacted if at least half of the ballots are for"
<fantasai> there we go :)
<fantasai> s/for/in favor/
<fantasai> florian: You're saying mine is a rephrasing whereas fantasai's is a change?
<TallTed> "each dismissal is enacted if at least half the ballots are in favor."
<cwilso> +1 to Florian's wording
<TallTed> "each dismissal is enacted if more than half the ballots are in favor."
<fantasai> [discussion of whether to address the question of ties]
<fantasai> fantasai: OK with either of Ted's proposals, prefer the 2nd-to-last
<fantasai> A) each dismissal is enacted if at least half the ballots are in favor
<fantasai> B) each dismissal is enacted if more than half the ballots are in favor
<fantasai> florian: Doesn't address the question of explicit abstaining
<fantasai> C) each dismissal is enacted if at least half the non-abstaining ballots are in favor
<florian> C) each dismissal is enacted if there are more ballots for than against
<fantasai> D) each dismissal is enacted if more than half the non-abstaining ballots are in favor
<TallTed> each dismissal is enacted if at least half the ballots expressing a preference are in favor
<fantasai> florian: OK, I'll do a PR with inspiration from these phrasing
<fantasai> ... but want direction: if we reach 50% do we dismiss or not?
<fantasai> fantasai: I think if we're that conflicted about the individual's participation, we should bias to dismiss
<fantasai> ... haven't been dismissing a lot of peope in practice anyway
<fantasai> plh, florian: OK
<fantasai> Next question
<fantasai> florian: Council votes if can't find consensus
<fantasai> ... is it majority of those who could vote? of those who did vote? handlign abstentions?
<fantasai> ... in partice we rarely get participation close to the total number, so basing on that would be bad
<fantasai> ... I think we want phrasing that says "more yes than nos, it passes" and "chair breaks a tie"
<fantasai> plh: wfm, make a PR
<TallTed> +1
<fantasai> Last 2 cases
<fantasai> florian: I listed two which I don't think are ambiguous, if anyone disagrees we can review
TallTed commented 3 months ago

For all votes where abstention (no vote) and/or explicit-neither-yes-nor-no (vote submitted with voiced no preference) is ignored, probably need to be explicit about this ignoring.

frivoal commented 3 months ago

Landed a commit directly for case 1, since we had a resolution with explicit phrasing, and made a PR for case 3 as well as one for case 4.

css-meeting-bot commented 2 months ago

The Revising W3C Process CG just discussed Disambiguate vote thresholds, and agreed to the following:

The full IRC log of that discussion <fantasai> Subtopic: Disambiguate vote thresholds
<fantasai> github: https://github.com/w3c/w3process/issues/838
<fantasai> -> https://github.com/w3c/w3process/pull/841/files
<fantasai> -> https://github.com/w3c/w3process/pull/842/files
<fantasai> florian: Process discusses various votes, passing by majority or supermajority
<fantasai> ... identified 4 ambiguities
<fantasai> ... I landed the first one based on previous call
<fantasai> ... 2nd one is about TAG, come back to it later
<fantasai> ... 3rd and 4th we agreed on what we mean, and I made 2 PRs to address
<fantasai> ... 3rd about Council dismissal, 4th about Council decision votes
<fantasai> <fantasai> +1
<fantasai> plh: objections to 841?
<fantasai> RESOLVED: Merge PR 841
<fantasai> florian: [introduces 842]
<fantasai> plh: objections to merge?
<fantasai> RESOLVED: Merge PR 842
<fantasai> florian: With these merged, the only thing remaining is about the TAG. I suggest we spin out into a separate issue and close.
<fantasai> ... that conversation is complicated, better in a separate plae
<fantasai> plh: +1
<fantasai> <fantasai> +1
<TallTed> +1
<fantasai> RESOLVED: Close issue 838, open separate issue wrt TAG appointment
frivoal commented 2 months ago

Sub-issues 1, 3, and 4 have been addressed. The broader discussion of TAG appointments remains open in https://github.com/w3c/w3process/issues/809, https://github.com/w3c/w3process/issues/810, https://github.com/w3c/w3process/issues/811. Relevant parts of this discussion have been reported into https://github.com/w3c/w3process/issues/809#issuecomment-2076343397.