[EM] Rob Lanphier's hierarchical scheme

Warren Smith wds at math.temple.edu
Sat Aug 26 10:42:45 PDT 2006


Yes, it does bear some resemblance to A.Lomax's proxy ideas.
I too have devoted some thought to these ideas.

However, I suspect Lomax's ideas are better and Lanphier's worse.  Or I understand neither.

Specifically, as far as I understand it, with Lomax's proxies, you can
select anybody on the planet to be your proxy.
With Lanphier's, you are a member of a group of 6.  The membership of that 6-set
was not selected by you.  It was selected in some undescribed way by the government,
perhaps randomly.

I may have just misrepresented Lomax or Lanphier or both.   We shall continue on
blithlely anyway.  In fact, to avoid calling them "Lanphier's" and "Lomax's" schemes
from now on we shall call them the "Govt's" vs the "People's" selection schemes.

Now either way, the scheme continues hierarchically, reducing the population
by a factor of 6 (or whatever) each stage, until at the end we have some manaegably small
legislature which makes decisions.

Now.   The problem (I think) with Govt-selection, is (a) it is highly manipulable
by some faction that gets control then amplifies their control.  This is kind of
like gerrymandering.  And (b) if the govt-selection is done randomly, then
there is no gerrymandering, but it may lead to a horribly effective form of
massive-conformity, wherein ideas that are not "mainstream" are systematically
reduced each stage, resulting in exponential decrease by the time the top of the
hierarchy is reached.   It is kind of like the median voter is selected each time to get
promoted one hierarchy level, and that causes, in a big hierachy, the extremes of the
idea-spectrum to be totally filtered out extremely effectively.
(Try a computer sim if you do not believe me.  The high levels of the hierarchy will
have vastly reduced fraction of extremes.)

With, however, people-selection, a bunch of Wackos can get together and promote
one of them to the next level, then they try to do so again next level, etc, thus getting
Wacko representation even at high levels, and without
filtering out Wacko ideas.

That is good.  On the other hand it might go too far and lead to artificial
extremeness-amplification as we go up the hierarchy.
Also, note that if each 7-set (here assuming 7X reduction each level)
is 4 Wackos and 3 normal people,
then we get vastly more Wackos at high levels in the Hierarchy than if each
7-set had been 7 Wackos.  This is the effect of "gerrymandering" I was talking about.
Such gerrymandering could vastly manipulate things if externally applied.  If however
people-selection is employed, then the factions that "trick" their members
into arranging just the right gerrymandering (i.e. 4 Wackos
each stage, and 3 dumb normals who don't understand what is being done to them
thanks to a con job) will get exponentially tremendously more power at the
high levels.  So there is a premium on recruiting dumb people.
So people-selection may also lead to problems.
But I suspect fewer problems.

Warren D Smith
http://Rangevoting.org



More information about the Election-Methods mailing list