[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[council] Some Thoughts on the Elections Process
Council Members,
Please find below a text outlining my thoughts on the elections of
ICANN Directors by the DNSO.
I look forward to a lively discussion
Dennis
---------
The Election of ICANN Board Directors by the Names Council.
Boundary Conditions:
1 Nominations by the DNSO General Assembly.
2 The electorate is the members of the Names Council.
3 Three candidates to be elected.
4 Each successful candidate to have "over 50% of the affirmative votes
of the NC members".
6 "...no more than one-half (1/2) of the total number of Directors, in the
aggregate, serving at any given time pursuant to selection by the
Supporting Organisations shall be citizens of countries located in any
one Geographic Region."
1 Nominations by the DNSO General Assembly
This is the simplest Condition. Some simple mechanism needs to be
found to allow people to nominate. Probably we need a seconding and
support statement mechanism. The suggestion that the person needs
the formal support of 10 other people is fine (but does not really add very
much). I presume that we will have a substantial (4 weeks?) nomination
period.
We should also provide for a period (2 weeks ?) of "campaigning" by
those nominated - essentially to allow those nominated to tell people
who they are and what they stand for. This will be targeted at the Names
Council (and perhaps the Constituencies - see below).
The electorate is the members of the Names Council
This is clear. However, it is also clearly the intent that the Names
Council members consult the various Constituencies to get input on
which candidates the Constituency favours. This suggest that some
formal mechanism of consultation of the Constituencies is required - and
probably it will vary between Constituencies, given the disparate types
of organisation of the Constituencies. The most likely mechanism that
will be and be seen to be transparent is a formal poll of the Constituency
members.
Such polls require time (several weeks), and a polling mechanism that
results in a ranking of the candidates by the Constituency. If required,
should such polls be by secret ballot ?
3 Three candidates to be elected
Since there is no geographic diversity outcome requirement specified
(see below for an explanation of this statement), although clearly one is
implied, there is no need to have "preliminary" type elections to reduce
the field to (say) one from each geographic region.
Therefore, all nominated candidates stand for election.
4 Each successful candidate to have "over 50% of the affirmative votes
of the NC members".
This Boundary Conditions is imprecise, since the meaning of "affirmative
votes" is not specified. I presume that it means that to be elected a
candidate must reach a quota of 50% of the votes cast. This of course
means that the voting must cope with the situation where no candidate
achieves this result on the first ballot, and successive rounds of
balloting are possible (with or without eliminations/standing down), or,
alternatively, that the voting mechanism allow for elimination and re-
distribution of votes.
This needs work to identify suitable mechanisms.
The mechanism I am most familiar with is the Proportional Representation
(PR) election method using the Single Transferable Vote (STV) system
(as was used in the ccTLD Constituency elections to the Names
Council). However, this normally has a quota of the total number of valid
votes cast divided by the number of seats + 1. In this case, with three
seats available, the normal quota under PR/STV would be 25% of the
votes cast - so I don’t know if this mechanism is suitable. It has the
advantage that it includes elimination mechanisms, and also surplus vote
distribution mechanisms, so that every vote counts eventually and no
vote is "wasted".
I do not know enough about any other mechanism to make any
judgement on its suitability to meet the necessary criteria. I'm sure that
the Electoral Reform Society of the UK, who the ccTLD Constituency as
advisers, will have detailed information about all sort of elections
mechanisms.
I presume that the vote will be by secret ballot.
4 Geographical Diversity: "...no more than one-half (1/2) of the total
number of Directors, in the aggregate, serving at any given time
pursuant to selection by the Supporting Organisations shall be citizens
of countries located in any one Geographic Region."
The last of these Boundary Conditions is unfortunately imprecise and
does not lead to a precise constraint on the Directors elected by the
DNSO. This means that whatever the results of the DNSO elections to
the ICANN Board, there exists the possibility that the results of the (set
of all) elections (by the SOs) will be rejected as not meeting this
condition, and new elections required. However, since there is no
precise constraint on the elections, the process could continue
indefinitely without resolution - except by some arrangement between
the various sets of electors.
<This imprecision is (IMHO) ridiculous>.
However, the good thing is that it is so imprecise that it forces no
geographic diversity condition on the outcome of the DNSO elections.
All three directors elected may come from (Oops! - be citizens of) the
same geographic region. (For example, and to illustrate the point, three
Asia/Pacific citizens, residents of the US for most of their adult lives,
could be elected as the three DNSO Directors).
<This citizen rule is (IMHO) ridiculous and should be replaced by a
"permanent residence" rule>.
Next Steps:
The critical next step is to identify a suitable election mechanism.
I recommend that expert advice be obtained from the Electoral Reform
Society of the UK and similar organisations.
Action: By Whom ?
<End>
-------------------------------------------------------
Dennis M. Jennings
Director, Computing Services, University College Dublin.
Address: Daedalus Building, Belfield, Dublin 4, Ireland
E-mail: Dennis.Jennings@ucd.ie
Telephone: +353-(1) 706 7817
Fax: +353-(1) 706 2362