[ga-procedures] nominations process proposal

[Jonathan Weinberg <weinberg@mail.msen.com>]

	The GA list recently has received forged e-mails, and e-mails from persons
posing as a variety of different, fictitious individuals.  Both of these
phenomena pose a threat to any voting process; indeed, some list members
have urged that it is pointless to conduct any vote until we have some form
of membership authentication to address those problems.  Roberto Gaetano
has suggested that, in advance of the election, the NC should inspect the
voting list in order to ensure that each address in the list corresponds to
a unique natural person.  I'm suggesting a somewhat different approach,
intended to minimize the burden of list-checking (given that the DNSO at
this time has only limited tools to determine whether names are authentic).

	Step 1. The DNSO announces a rule that only natural persons can
participate in the nomination process, and that no person can participate
under more than one name.  The NC appoints a small"poll-watching committee"
from within the GA.

	Step 2. Nov 26 - Dec 2: Any member of the GA can propose the name of one
or more potential candidates for the office of GA chair.  Any person whose
name has been proposed has until the end of  Dec 2 to accept the
nomination, and to file a statement explaining how he or she will carry out
the job of chair, and setting out his or her views as to the GA's
relationship with the NC and the constituencies, and its role in the DNSO.
Any person who does not accept the nomination, and file the statement, by
that deadline can't be considered.

	Step 3. Dec 3-Dec. 10: Each member of the GA can cast a vote for up to N
people (see below) on the list of potential candidates.

	Step 4. The poll-watching committee counts the votes, and assembles a list
of the top N vote-getters.  It also examines the votes cast, and makes a
judgment whether there is a substantial possibility that any person on that
list made it on only because of fraudulent votes.  If so, the committee
adds additional names to the list (starting with the person who received
the next largest number of votes, and so on) until it appears that the
expanded list includes each of the N people who would have received the
most votes absent fraudulent voting.  The committee then sends the expanded
list to the NC, and send a statement to the NC and the GA explaining their
conclusions and the bases for their actions.

Comments:

	N should be from three to five.  I think it would be most appropriate for
the initial list to contain three names (the expanded list might contain
more, if fraudulent voting were a problem).  The NC may conclude, however,
that the number should be larger.

	A difficulty here is that the evidence available to the poll-watching
committee may not always allow the committee to draw firm conclusions as to
whether votes are fraudulent.  On the other hand, once the actual votes are
cast, it may turn out that there's no practical problem -- for example, it
might happen that N candidates get sufficient landslide support from
obviously legitimate votes that any fraudulent voting couldn't have
affected their standing.  My goal in setting up the poll-watching committee
is to allow enough fudge to deal adequately with the problem *if* it
arises.  The problem is a short-term one, since, once the ICANN membership
is established, that verification mechanism will be available to the DNSO.

	Step 5. The NC selects a chair from the expanded list.


Jonathan Weinberg
weinberg@msen.com