[CCWG-ACCT] Minority Statements

[Steve Crocker]

On Aug 2, 2015, at 2:37 PM, Alan Greenberg <alan.greenberg at mcgill.ca> wrote:

> Thanks Steve,
> 
> Bernie Turcotte did most of the analyical work on this, but I will offer a few comments and thoughts.
> 
> Regarding your classifications:
> - NP was not relevant, since this is not an in person vote and we are presuming that all parties could exercise their power if they chose.
> - We have not considered Rec, because of the nature of the system. Just as in the GNSO Council, where it is presumed that often, contracted parties have a vested interest in the outcome of PDPs, as long as their interests are declared, they can still vote.
> - We have classes a possible vote that is simply not exercised (through conscious decision or inattention) as an Abs.
> - There is another category in our case, that of an AC/SO that is allowed to vote but prefers to be excluded because they have not opted in to the system. They are simply excluded from the possible pool of voters as if they did not have any rights.

No argument from me.  I was suggesting a possible vocabulary and definitions to facilitate the discussion.  My orientation in these exercises is to provide fine-grained categories that capture all of the distinctions people care about and then build up groupings from there.
> 
> In the CCWG, we have tended to say that most of the rights can be exercised by a super majority (>66%) or a 75% vote in favour. If everyone casts a Yes or No, that seems to be reasonable (one can debate the actual number of course).

Ok, this is where a bit more precision seems useful, at least to me as someone who hasn’t followed every detail.  Certainly the yes votes are included in the numerator and both the yes and no votes are included in the denominator.  What about the abs votes?  What about the AC/SOs that don’t vote?


> We have also tended to say that a single AC/SO should not be able to alone, veto any action (there are some perceived problems with this on the budget veto issue, but let's ignore that for this discussion).
> 
> If you treat an abstention as a No, the above works fine (and this is how most but not all parts of ICANN treat abstentions). But some people feel that an abstention is a message that they do want to be part of the decision.

Yup, this is exactly one of the details it’s important to thrash out and gain agreement on.  Discussion of specific cases, extremes, etc. is also helpful.

> We may have as few as 20 votes in total. If just one person holding a weight=1 vote abstains, then we can simply reduce the denominator and the above criteria can stand. However a problem arises if a substantial part of the community abstains. For instance, if two SOs completely abstain, then the final decision can be made by a very small number of participants, and with sufficient abstentions, high-impact decisions can be made by a single AC/SO. It is true that this happens with the willful agreement of those who have abstained, but it is nevertheless capture by the remaining (and capture due to parts of the community losing interest was one of the things out expert advisors warned about).

I recommend expressing each of these ideas on its own before trying to piece them together into an integrated and complete policy.  Express each of the possible conditions as an explicit statement.

> Which leads to the options that Bernie and I considered for allowing abstentions to be excluded from the count, but nonetheless avoid capture. They essentially say that we cannot have too many abstentions, and that they must be spread around.
> 
> The last version we came up with was the following:
> 
> - To pass, the proposed proposition must have the required percentage of votes - Yes-votes/(Yes-votes + No votes)
> - 75% or more of eligible votes must have been cast as Yes or No (that is, we are limiting the number of abstentions on not-cast votes to less than 25%
> - More than 50% of the participating SO/ACs have to support the proposition - that is, for each of these SO/ACs, they must tally >2.5 Yes votes.

This seems clear enough and not overly complicated.  I am not yet completely sure I understand what it means for an SO or AC to participate (or not participate, but that’s probably because I haven’t read carefully enough yet.  Can you formalize the notion of “participation?”

> Reviewing a number of scenarios, with these requirement, we can avoid capture and still be able to exercise the powers. Essentially it requires a significant part of the community to want to take the action. If too many people opt out and abstain, the power cannot be exercised, but a single SO/AC cannot veto.

It might be helpful to lay out various scenarios you’d consider unacceptable.  For example, can you characterize “capture?”
> 
> However, it was felt that to use this set of rules, the process would simply be TOO obscure and opaque.

Hmm…. I’m not sure why this set of rules is viewed as overly complicated.  In the U.S. we’re quite familiar with notions of “quorum”, i.e. a required number of participants, and various representation schemes, e.g. representatives, senators and electors.  There are advantages and disadvantages in each of these systems, but they are well within the understanding and support of the electorate.

> Hope this helps a bit.

Keep pushing.  This feels like it might help.

Thanks,

Steve

> 
> Alan
> 
> At 01/08/2015 03:48 PM, Steve Crocker wrote:
>> Alan,
>> 
>> I am both curious and willing to help on the technical aspects of the combining multiple requires or desires into an implementable plan.
>> 
>> Is it possible to gather the multiple requirements and/or desires so we can see whether a feasible solution exists?
>> 
>> Let me offer some candidate terminology that may be granular enough to help express the desired results.
>> 
>> Within a given group and with respect to a particular vote, the individuals will usually fall into one of the following five categories:
>> 
>> o (NP) Not present (and hence not participating)
>> 
>> o (Rec) Present but not voting because of recusal, usually because of self-declared conflict or because the the group perceives a conflict.
>> 
>> o (Abs) Present but not voting because the individual doesn't wish to express either a positive or negative vote.
>> 
>> o (No) Present and voting no.
>> 
>> o (Yes) Present and voting yes.
>> 
>> Can you express the desired results in terms of one or more inequalities, ratios or other expressions among these terms?
>> 
>> Thanks,
>> 
>> Steve
>> 
>> 
>> 
>> 
>> 
>> On Aug 1, 2015, at 11:57 AM, Alan Greenberg <alan.greenberg at mcgill.ca> wrote:
>> 
>> > Ed, I believe that I was the first one to make a strong statement that abstentions should be excluded from the vote. As few of us worked VERY hard to come up with a vote-counting methodology that did that and at the same time did not allow a very small part of the community (the ones who "care" about the issue) to make momentous decisions on behalf of ICANN.
>> >
>> > We did arrive at some scenarios that were acceptable from the point of view of outcomes, but that were difficult to implement and perhaps more importantly to explain. In addition to the simple Yes vs No count that you allude to, it involved parallel requirements for a minimum number of SO/ACs to support the proposition (and that support was absolute, only counting Yes votes vs the maximum that could be cast), and perhaps requiring a minimum number of non-sbstention votes to be cast. As I said, it might work, but would be a black-box and completely opaque method to those who did not take the time to thoroughly understand it.
>> >
>> > If you can come up with a simple, clear way of doing it, please propose one.
>> >
>> > Alan
>> >
>> >
>> > At 01/08/2015 08:10 AM, Edward Morris wrote:
>> >> About twenty minutes ago I submitted directly to the Chairs, per the instructions given to us by Thomas in his email of 29 July, two minority statements for (hopefully) inclusion in the report about to be released for public comment. I had not intended to file any minority statement but, upon reflection, two aspects of our proposal caused me concern.
>> >>
>> >> The statements are attached here for community inspection and review.
>> >>
>> >> Kind regards,
>> >>
>> >> Edward Morris
>> >>
>> >>
>> >>
>> >>
>> >>
>> >>
>> >> _______________________________________________
>> >> Accountability-Cross-Community mailing list
>> >> Accountability-Cross-Community at icann.org
>> >> https://mm.icann.org/mailman/listinfo/accountability-cross-community
>> >
>> > _______________________________________________
>> > Accountability-Cross-Community mailing list
>> > Accountability-Cross-Community at icann.org
>> > https://mm.icann.org/mailman/listinfo/accountability-cross-community
>