[EM] Confirmed!: Condorcet efficiency of IRV > 2-stage runoff

[Steve Barney]

I've discovered a professional article on voting theory which seems
to confirm my argument (see 
<http://groups.yahoo.com/group/election-methods-list/message/9057>)
that the IRV is more likely to elect a Condorcet candidate when one
exists (with sincere votes) than the two-stage runoff procedure
(assuming that the preferences remain fixed from one stage to the
next). Here is an excerpt from Table 1, page 6 (in this article, IRV
is called the "Hare" method):

Table 1: Condorcet efficiencies for a random profile with 25 voters
by Merrill (1984)
procedure \ # alternatives 2 3 4 5 7 10
RUNOFF 100,0 96,2 90,1 83,6 73,5 61,3
HARE (TIES) 100,0 96,2 92,7 89,1 84,8 77,9

See pg 6 of the article:

"Analysis of voting procedures in one-seat elections:
Condorcet efficiency and Borda efficiency"
DIMITRI VANDERCRUYSSEN (KUL)1
March 1999
<http://citeseer.nj.nec.com/vandercruyssen99analysis.html>

That article notes that this "far-famed table" comes from an earlier
article:
Merrill, S., III (1984) “A Comparison of Efficiency of
Multialternative Electoral Systems”, American Journal of Political
Science, Vol 28, Issue 1, pp. 23-48.
(available in the "JSTOR" academic database)


Notice that, just as I surmised, the two methods start out equivalent
with 2 or 3 candidates, and then the IRV does better and better than
the 2-stage runoff as the number of candidates grows. 

Also, on a lighter note, this article cites this list in footnote 9
on page 6:

9 "We use the name “Baldwin” rule in order to stress the difference
with the Nanson rule. John Taplin mentioned that name on the Election
Methods Internet Site. The procedure may be better known as e.g.
Nanson’s modification of his own rule (cf. McLean and Urken (1995)),
but this terminology is confusing."



=====
Richard M. Hare, 1919 - 2002, In Memoriam, http://www.petersingerlinks.com/hare.htm

__________________________________________________
Do you Yahoo!?
New DSL Internet Access from SBC & Yahoo!
http://sbc.yahoo.com

----
For more information about this list (subscribe, unsubscribe, FAQ, etc), 
please see http://www.eskimo.com/~robla/em