[members-discuss] Analysis of Voting Registrations at the GM

Hi Denys,

Thanks for the feedback on the voting and registration requirements. It is
useful to get that, and we take note of your proposals.

I've talked to a lot of members who don't vote over the years, and the main
reasoning I have heard is that it takes a lot of time that they don't have
to get involved in RIPE NCC governance; that they don't feel comfortable
voting without fully understanding the issues or knowing the Board
candidates; that they are not authorised to vote for their company; and
several have said that things seem to be run well and they trust their
fellow members who do have time to get involved. But it certainly would be
good to hear more from members who don't vote to better understand their
reasons.

The development of our voting over the years has been done with two-thirds
majority votes from our members to change the Articles of Association.
Similarly, members of other RIRs will have different opinions and needs,
and their voting systems have developed accordingly. Regarding the voting
period, it was a Board decision to open the voting after the deliberations
are completed, so that members can decide how to vote after they have heard
all arguments and discussions.

On the voting method we use, I did an analysis on the voting methods for
the Executive Board a couple of years ago that looked at some of the issues
you raise and to see if the alternatives you mention might be better. You
can see the summary of that in the minutes from the meeting that gives some
of the the reasoning for the current IRV method:

https://www.ripe.net/about-us/executive-board/minutes/2022/minutes-156th-executive-board-meeting/#9-corporate-governance

A more detailed list of the advantages to instant run-off voting are:

- It closely replicates the system that was in place for many years before
electronic voting was possible with the important aspect that it allows an
absolute majority (more than 50%) to be achieved by winning candidates
without the need for several rounds of voting.

- It allows the election of multiple candidates in one ballot – we
currently elect two or three Executive Board candidates per year. The IRV
system allows us to do this at one General Meeting and with one ballot.

- It ensures there is broad support for all elected candidates – the IRV
system rarely elects people on the first round of voting but those elected
need high preferences from a significant proportion of the registered
voters.

- It reduces the “wasted votes” factor – if a voter’s first choice is not
elected, the voter still plays a part in choosing who will be elected.

- It allows voters to give high preferences across various criteria – A
voter can decide to vote for a known and trusted candidate but also give a
high preference to someone who has skills they would like to see on the
Board or who can serve a specific part of the service region.

- It is a relatively simple system – most other forms of preferential
voting have more complex quota requirements depending on the number of
voters.

With the first past the post (FPTP) system, these are some of the factors
that made it unattractive for the RIPE NCC:

- Candidates (especially in a three-seat election) could be elected with a
very small number of votes, so there would be no more majority decisions on
elected candidates. Without broad support for a candidate, they could be
seen as having no great mandate to serve.

- People could be discouraged from voting for certain candidates as a vote
against a likely winner might be seen as a wasted vote.

- We move from a situation where all votes can have impact to one where
only votes for winning candidates have impact.

In the example you mention from 2023, under FPTP the candidate would have
been elected to the Board with 262 votes from a total of 1,841 votes cast.
The instant run-off method ensured that the winning candidates achieved an
absolute majority for their seats.

I hope this helps with understanding why instant run-off voting was chosen,
and feedback on the topic is always welcome.

Best regards,

Fergal

On Sat, Jul 6, 2024 at 6:39 AM Denys Fedoryshchenko <
nuclearcat at nuclearcat.com> wrote:

> Hi!
>
> Thanks for nice article and analysis.
>
> Few comments:
>
> >8.5% might sound low, and in fact, that number is only a little below
> average for a May GM.
>
> Actually, it always has the lowest turnout among RIRs (except for the last
> voting at LACNIC), and IMHO this is something that has to be fixed. And I
> believe some of the reasons are trivial, and it is not lack of your
> community efforts, but more fundamental reasons.
>
> >Throughout, we'll refer to the members that registered to vote in the May
> 2024 GM as *members that voted* and the rest as *members that didn’t vote*.
> In fact, of the members who registered to vote, only 84% went on to cast
> their votes.
> I would like to ask a separate question - why is registration necessary?
> Based on what I see in other RIRs (no registration) and voting history,
> registration was required when voting took place on-site. This was because,
> due to the number of registered people, it was necessary to rent a venue,
> etc.
> In my opinion, it is high time to cancel it and recover the missing 16% of
> the votes.
>
> Additionally, I would like to note that other RIRs conduct voting over a
> week, rather than within 24 hours. It's not clear why RIPE NCC limit people
> to 24 hours (again, on-site voting legacy?). In my experience, its not
> enough.
> Also, subjectively, voting in other RIRs is made much easier; there are
> short instructions right in the voting interface. The fact that the voting
> process in RIPE requires much longer instructions only indicates that it is
> overcomplicated.
>
> So my proposals are quite obvious:
> 1)Cancel registration
> 2)Extend voting duration to 1 week
> 3)Voting should be more intuitive
>
> P.S.
> Not exactly related to the article, but I would like to speak separately
> about IRV.
>
> As mentioned at RIPE-176
> <https://www.ripe.net/publications/docs/ripe-176/>, the initial version
> of the RIPE NCC bylaws used a simple majority voting system, where the
> number of votes depended on the size of the membership. This is similar to
> the systems used by ARIN and APNIC. The voting system was changed to IRV at
> the October 2009 General Meeting in RIPE-487
> <https://www.ripe.net/membership/meetings/gm/meetings/october-2009/general-meeting-minutes/>,
> with the following explanation from the meeting minutes:
>
> *Jochem explained that changes to the Articles of Association would be
> needed to support e-voting and also that this was a good time to introduce
> any housekeeping changes that were necessary.*
>
> It is interesting to highlight that the voting system was changed to
> support e-voting, but it is not clear why IRV was chosen or if there was
> any public discussion or consultation about the voting system. In my
> personal opinion, under the pretext of e-voting, the system was changed to
> IRV, which is HUGE change, to more complex and less transparent to the
> average voter and introduces unfairness into the system.
> I also want to note that other RIRs did not need to change their voting
> systems when switching to online voting.
>
> Moreover, RIPE uses an IRV configuration that allows 'centrists' to gain
> 'amplified voices' and leaves little chance even for moderate reformers.
> While IRV voting may initially foster political stability and moderate
> policies, it ultimately risks stifling innovation and reform, potentially
> leading to long-term stagnation and a failure to address emerging
> challenges.
>
> As good example, in GM 2003 Fahad AlShirawi had change to win seat by
> simple majority, but no chances in IRV.
> I've done my own analysis, but it's not polished enough to publish.
> However, it's not difficult to perform this calculation.
>
> On Thu, 2024-07-04 at 16:56 +0200, Fergal Cunningham wrote:
>
> Dear all,
>
> We just published a RIPE Labs article that takes a closer look at the
> voter registration data from the recent May General Meeting:
>
> https://labs.ripe.net/author/ilke-ilhan/voting-analysis-gm-may-2024-a-first-look-at-voter-make-up/
>
> If you have thoughts on the article, or suggestions on other aspects you'd
> like to see covered in future analyses, we'd love to hear them.
>
> Best regards,
>
> Fergal Cunningham
> Head of Membership Engagement
> RIPE NCC
> _______________________________________________
> members-discuss mailing list
> members-discuss at ripe.net
> https://mailman.ripe.net/
> Unsubscribe:
> https://lists.ripe.net/mailman/options/members-discuss/nuclearcat%40nuclearcat.com
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.ripe.net/ripe/mail/archives/members-discuss/attachments/20240708/c73ddcc7/attachment-0001.html>