FW: Follow-Up to NCSG Charter-Structure Questions

[Milton L Mueller]

Hello, all
You will remember that I sent questions to the staff questioning the practicality of certain aspects of their favored model for a NCSG. The good news is that they have taken the inquiry seriously and responded. The bad news is that, as I feared, the only way to make their favored model work requires enormous amounts of organizational overhead - an additional bureaucratic overlay that creates not only complexity but the possibility of top-down manipulation of constituency election results. Interestingly, the more workable approaches start to look a lot like the integrated election process we already proposed. Read for yourself.

________________________________
From: Robert Hoggarth [mailto:robert.hoggarth at icann.org]

Milton:

Thank you for your recent email (below) in which you posed a couple of questions for the Staff concerning the new Stakeholder Group model.  The questions are challenging and we have done our best to provide what we think are reasonable recommendations as to how they might be addressed.

Q1: How does a constituency-based model produced balanced geographic representation in Council seats?

We envision geographic diversity as a representational responsibility of each Stakeholder Group (SG) in fulfilling its role of allocating GNSO Council seats to member Constituencies.  We are currently working with the General Counsel to draft Bylaw amendments consistent with that approach.

Using your example, if the SG has six seats and three Constituencies, it may choose to allocate seats evenly although it would not be required to do so.  In the simplest case, if there are two seats assigned to each Constituency, the SG would alert its members that it needs to have all five geographic regions represented with no more than two Councilors coming from the same one.  In order to accomplish that goal, the SG might ask certain Constituencies to produce its candidates from a limited set of geographic regions.
Another option might be to solicit a larger candidate pool, e.g. three from each Constituency or nine total (voted from within), and choose those six that best satisfy the SG's geographic diversity needs.  In a situation where the number of Council seats to be allocated is not evenly divisible, the SG might decide to designate certain seats to specific geographic regions and candidates from various Constituencies could campaign for those available slots.
MM commentary: note the last line: "designating certain seats to specific geographic regions" and allowing "candidates from various Constituencies [to] campaign for those available slots" sounds suspiciously close to an integrated, Stakeholder Group-wide election, which is what we proposed! In other words, Council candidates would have to appeal for votes from across the entire SG, not just inside their constituency
We believe that the SG, working collaboratively with its member Constituencies, can continue to ensure that its GNSO Councilors reflect a profile consistent with the organization's geographic diversity goals.

Q2: How does a constituency-based model apportion Council seats among Constituencies when they are of different size?

While the decisions may be challenging, we think that one principal role of the SG's leadership team[ 1] <see footnote below> is to establish the very criteria (and methodology) that would allow such apportionment determinations to be made.  A SG could utilize factors other than size, for example, geographic diversity and possibly others.  To take a concrete example, if there were six seats and four constituencies, the SG could end up with a 2-2-1-1 or 3-1-1-1 configuration utilizing whatever decision-making criteria it adopted (and had approved, via its Charter, by the Board).  You raise the possibility of gaming the system and, of course, such behavior is theoretically possible in any proposed model including your own.  In that circumstance, the SG leadership should reexamine its methodology and adjust, as necessary, to minimize any undesirable outcomes.  Once the seats are allocated, if a new constituency is subsequently admitted to the SG by the Board, we recommend that, at its next annual cycle, the SG reallocate seats taking into consideration five members vs. four.  The most likely Council member configuration, given the limited combinations, would be 2-1-1-1.  We do make the tacit assumption that any Constituency approved by the Board would have satisfied ICANN's fundamental stakeholder representational requirements and, thus, would be entitled to at least one seat on the Council.

In terms of oversight, we believe that the Board's role will be not only to ensure that each of the SG Charters is structured in a fair, open, and transparent manner; but, it will also likely monitor SG activities, especially in the period immediately after initial implementation.

We would be happy to continue discussing this matter with you and, of course, we would welcome another set of questions if there are still unresolved issues in your mind.

Regards,
Denise Michel

[1] <#_ftnref> We would envision an Execeutive Committee comprised of one delegate from each recognized Constituency.

MM comment: This proposal creates a potential nightmare. It requires a group of delegates from each constituency to fight among themselves, with no pre-set criteria, to decide who gets how many Council seats. Unacceptable, and unnecessary. In our proposal, size differences among constituencies are automatically reflected in voting totals for Council seats. There is no need for top-down, negotiated allocations. Those negotiations create all the rigidities that the Board Governance Committee was trying to get rid of, and present all kinds of opportunities for abuse. Even when they are not abused, they will consume enormous amounts of time. It is apparent that ICANN's professional staff - which gets paid to do this work - still does not appreciate the way in which imposing additional layers of bureaucracy and a constant need to contend and negotiation for power inside a SG saps the energy of noncommercial groups and prevents them from doing the real work of policy development. I ask for your support to tell the staff that this is not an acceptable option.

MM's Original Email of 6 February, 2009:

Robert, Denise and Ken

Thanks a lot for your valuable feedback on our draft Charter (v4.0). It is clear that we are making progress, although there is a long way to go.

In respect to some of your questions or requests for explanation, let me turn the tables on you a bit. The presumption in many of these exchanges is that there's something complicated or "different" about what we are proposing, and that the "constituency-based SG model" is straightforward and poses no problems. In many ways, however, an integrated SG structure is far simpler, and we have no idea how a constituency model would work even if we thought it desirable to implement it.

Let me give you two examples. I will pose them in the form of questions because it genuinely would like to have answers from you or any other defender of the constituency-based SG model.

Q1: How does a constituency-based model produced balanced geographic representation in Council seats?

Think about this. Let's say there are 3 independent constituencies in a SG, and each of them elects 2 Council seats without reference to the other. So Constituency A elects (in accord with its own geog. representation rules) a person from North American and a person from Latin America; Constituency B elects a person from North America and a person from Latin America; and Constituency C elects a person from North America and another from Latin America. End result: each constituency has, on its own, produced as much geographic diversity as it possibly could, and yet the end result could be that only two world regions are represented on the Council.

I would be very interested to see how you propose to avoid this problem while staying in the constituency model.

An integrated SG model, by contrast, can impose proportions on the six seats as a whole, thereby ensuring that most if not all regions are represented.

Q2: How does a constituency-based model apportion Council seats among Constituencies when they are of different size?

Let's suppose there is an "old constituency" that has 50 members, and a "new" constituency that starts and gets recognized by the Board, and has only 10 initial members (or even less). How many Council seats does each constituency get? Do they inherently get the same number of seats simply by virtue of the fact that they are constituencies? Or does their representation on the Council reflect their relative size? If the latter, who decides what allocation principle is used, when there is no pre-established SG decision-making method? And once Council seats depend on membership size, what is to stop one constituency from extending membership in an overly easy way, regardless of appropriate criteria, to inflate its relative size? Will the Board monitor this?

These questions are not impossible to answer, but they obviously impose a very complex layer of organization, monitoring and procedure that an integrated SG model does not have to worry about.
Frankly, Bob and Denise, I could produce about a dozen more questions like this. But let's see how you do with these two first.

My point is to put this discussion of SG models on a more solid footing with an equal burden of proof. If you can convince us that a constituency-based model handles such basic and obvious issues as well as an integrated model,we'd be more inclined to change our view.
--MM
________________________________
[1] <#_ftnref> We would envision an Execeutive Committee comprised of one delegate from each recognized Constituency.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ncuc.org/pipermail/ncuc-discuss/attachments/20090221/75637668/attachment.html>