Notes from GOSH Governance WG Meeting #9 - 3 Nov 2020

Hi All

Latest notes from a productive governance call below! Next call is on Tue 17 Nov 2020, 14:00 UTC [note new day]

In the meantime, please:

  1. Continue researching, asking peers, and overall looking for equitable election models we can draw from or learn from


Meeting #9
3 Nov 2020, 14:00 UTC

People in the call

  • Jenny
  • Gunner
  • Marina
  • Gayatri
  • Paz

Actions for working group members in advance of this meeting:

  1. Whole WG to review notes from last meeting and add items to the table of what we want and don’t want from a governance group
  2. Consider how we would seat this long-term governance group to best meet our goals for group composition and structure.
  3. SPECIFIC ASK: If you know of existing models for elections or other processes that equitably seat representatives, please come ready to share your knowledge!


What process do we define and implement to get the seated governance group in place?

  • Elections seem to be the way forward

Possible scenarios of how this representation can be made, in terms of equity, inclusion and coverage across the community.

  • Should we strive for equity in gender, race, geography and how?
  • Gets very complicated when these areas of representation intersect. Gender may be “simpler” but still difficult e.g. terminology - e.g. majority non-male.
  • Self-identification could be problematic if there are people who are minded to play with the system.
  • Could have a self nomination form with a set of categories that people identify with. Could be social minority or sector. Give a higher score if this is a priority. Difficult to give subjective scores
  • Gunner: what we land on has to be simple and everyone should understand it. Explaining calculations is going to be divisive.
  • Discussion on 5 vs 7 members. Pro of more members: more representation; con: harder to schedule meetings, less agile. Agreed on 7
  • Need to think about keeping governance structures flexible, keep election process the same but change criteria.
  • Do research, ask our peers for advice and existing models
  • Make a decision on the next call about which model should move forward.
  • Proposal that everyone agreed: one third of the slots will be occupied by the most voted people, two thirds will be occupied by people that had self-identified with social minorities. The list of social minorities can be then changed with time, can be super simple or complex.
  • We should study about this type of election, based on a quota system.

Next steps on engaging broader GOSH community (call) (after first version of proposal is complete and ready for feedback and discussion)

  • Schedule a community call in early Dec where we can share outputs of the WG and open questions.
  • Does the timeframe sound OK? Could be mid-December which would allow three meetings to plan (Gayatri) - +1 from Marine, Paz.
  • How to arrange the call? Could frame as an update on activity with opportunity for input. Will enable us to work out if we are almost done and can get on or if it needs more work.
  • This group should consider what that meeting might look like and who might facilitate parts of it.
  • Aim for call on Wed 16th Dec for 90 min at 14 GMT

For next meeting:

  • Continue discussion nomination and elections model
  • Make a decision on the next call about which model should move forward.
  • Identify remaining steps to assemble governance proposal for community review

Actions for working group members in advance of next meeting

  • Whole WG to continue researching models for equitable elections, in particular of communities that have already put such constructs in place

Ni! Hi there. Just my two cents since elections are things that I read and think somewhat about, even if very informally…

What you’re proposing is not equitable elections, at least in my understanding. What I’d call equitable elections are those where 1 person counts for 1 vote, as opposed to systems where votes are weighted by wealth or status. What is proposed seems to be a system of stratified representation, more precisely of quotas based on strata.

I think therefore that you need to calculate the number of seats as a function of how many strata you want represented and to what proportion, otherwise you may not be able to satisfy your own criteria.

One system that I think corresponds to your choices is score voting, then sequentially pick candidates to satisfy minimum stratification requirements in the order of strata importance, with a chosen candidate counting only towards the stratum that was used to pick them, then proceed normally with remaining candidates.

Note: it may seem counterintuitive to count a pick only towards the criteria used to pick it. It is more complicated than that, and this represents a compromise that favors minorities. Otherwise if picking women we pick enough people from the south we hurt the chances of males from the south.

Following what seems to be your stratification concerns, if you have 7 seats and you want to stratify 5 of them between gender, race and geography, then you could use the following 5 unitary strata in this order:

  • 1 woman of color from the global south
  • 1 woman from the global south
  • 1 person of color from the global south
  • 1 woman
  • 1 person of color

Note: I’ve put “1 woman” down one notch from the “logical order” because in concrete terms it feels like it will be more balanced, the “fairest” order in practice is unknown as it depends on the membership and biases of the community.

Another issue is what to do with unsatisfiable qualifiers. The simplest thing is to treat each part of the qualifier in order and ignore those parts that can’t be satisfied. So for “person of color from the global south” if there are no more eligible people of color we’d pick one from the global south.

Also desirable is some sort of minimal eligibility criteria, like requiring a candidate to prove support of a significant fraction of the community to enter the ballots, so that rogue candidates don’t get elected based only on quotas.

Cheers =)

1 Like

Thanks [solstag], these insights are very much appreciated! In particular, unsatisfiable qualifiers is something we are definitely wanting to design elegantly for.

We will definitely discuss at our next meeting, and keep folks updated on where we are getting to.

1 Like