[EM]

LAYTON Craig Craig.LAYTON at add.nsw.gov.au
Tue Oct 10 18:36:36 PDT 2000


Okay, this is my last tirade (maybe) on PR vs single winner systems before I
retire gracefully;

A prerequisite of political systems/electoral systems, before considering
whether they're 'good', 'fair' or meet any of the various criteria, is that
they're legitimate.  That is, they have some legitimacy to claim a monopoly
over the use of force.  While this legitimacy contains a number of elements
(and there is no agreement about what they all are), I believe that a
fundamental element is the consent of the people to the existance of this
coercive institution.

As the government has absolute coercive power over all people permanently
residing in that country, and not just those who vote / are eligible to
vote, I think that the minimum number of people having some say in the
makeup of government in order for it to be legitimate is half of the total
population.  I define 'having some say in the makeup of government' as
helping to elect a member to the legislature, either by voting for that
member (in non-rank ballot elections), voting for that member's party (in PR
party list systems) or voting for that candidate above some other candidate
who didn't get elected, and that vote is actually counted to help elect that
candidate (in preferential systems).

Generally, IRV, FPTP, Condorcet, Tideman, Approval and most of the single
winner methods would fail.  I say generally because they wouldn't always
fail, simply because they're largely inclined to.  It's not a criteria test.
(Borda & some other points systems may be an exception, but not a very good
one).

PR passes most of the time (under my criteria, all of the governments over
the past 30 years in Australia have been legitimate to some extent.  I don't
believe any of the US governments have been - ever).  I'm attempting to
compare different PR systems, to see which one produces the most legitimate
governments.  I suspect STV PR is the best, but I may be biased.  At any
rate, single winner methods likely to fail are not acceptable, because they
elect a government that has no right to govern.



More information about the Election-Methods mailing list