explanation of geographical diversity election process

[Kent Crispin <kent@songbird.com>]

Here's further explanation of the somewhat complex geographical
diversity provisions in the DNSO proposal.


    Example Illustrating Geographical Diversity Election Process

Assume the five geographical regions are labeled A-E.  Assume the
Registry Constituency ("RC") has 6 representatives on the Names
Council ("NC").  Each year, two of those positions become vacant, and
are up for reelection.  The 4 representative that remain on the NC
have the the following regional designations: A, A, B, C.  [Say A is
"North America"]. 

Suppose that members of the RC nominate 7 nominees for 2 vacant slots
on the NC, nominees 1-7.  Suppose that the nominees are identified
with regions as follows: 1(A), 2(B), 3(B), 4(C), 5(D), 6(E), 7(E) --
nominee 1 is from region A, etc.  The vote is conducted; the 
nominees get the following

    1(A) -  201 votes
    2(B) -   65 votes
    3(B) -  536 votes
    4(C) -   25 votes
    5(D) -   55 votes
    6(E) -  135 votes
    7(E) -  135 votes

We consider the "regions with the least number of current
representatives for the Constituency".  They are regions D and E,
which have zero representatives from the RC.  The nominees that come
from regions in this group are 5, 6, and 7.  The nominee with the 
most votes from (5,6,7) is selected.

In this case there is a tie between 6 and 7; a coin is tossed; 6
wins, so 6 is elected. 

We now have 5 representatives selected for the RC -- the 4 
incumbents, and the newly elected #6 (who may be now considered as an 
incumbent with a remaining term of 3 years).  They have regional 
representation as follows:

    I(A)
    I(A)
    I(B)
    I(C)
    6(E)

The same rules are followed again, except that 6 is no longer a 
nominee, but an incumbent.

The "regions with the least number of current representatives" are 
now just region D, with zero representatives.  Nominee 5 is from D; 
there are no other nominees to consider, so Nominee 5 is selected.  
The end result is 

    I(A)
    I(A)
    I(B)
    I(C)
    5(D)
    6(E)


Suppose there had been no nominee from region D? In that case the
rule is that the election would stop with the selection of nominee
#6, and that the RC would for the next year have only 5
representatives on the NC, instead of their allowed 6.  The effect is
the same as if they had provided only one nominee.  This is an
incentive for the RC to find qualified nominees from unrepresented
regions. 

The nomination of nominees 1, 2, and 3 was something I put in the
example for illustrative purposes: since there were 2 unrepresented
regions, and only 2 slots available, the only viable nominees would
be from those two regions, and the RC was wasting peoples time
nominating the first three nominees. 

If a constituency has fewer representatives than regions, the rule 
simplifies down to one where every representative must be from a 
different region -- the rule is designed to handle any number of 
representatives and regions.

However, the scheme does depend on there being several
representatives from each constituency -- if every constituency had
one representative the rule would be useless.  Currently, because of
the 3 year terms and staggered elections, representation should be in
multiples of 3.  If every constituency had 3 representatives, some
regions could be left out of the Names Council -- for example, every
constituency could elect representatives from regions A, B, and C,
and regions D and E would have no representation.  This problem was
brought up to the many international participants at Monterrey, and 
they were not concerned -- with there being at least 6 groups of 3 
the odds of one region being left out were considered to be minimal, 
and would likely only last for one year, in any case.


-- 
Kent Crispin, PAB Chair				"Do good, and you'll be
kent@songbird.com				lonesome." -- Mark Twain