[EM] First Condorcet cycle ever spotted in a national presidential election (!?! apparently)

[seppley at alumni.caltech.edu]

An analysis of the 1992 U.S. presidential election between Clinton, Bush &
Perot suggested the voters' preferences were a majority cycle: Bush over
Clinton, Perot over Bush, and Clinton over Perot.  If true, the recent
Romanian election was not the first majority cycle in a national
presidential election.

The analysis was by a professor of political science at UC San Diego. I
don't recall his name... maybe Gary Jacobson?  His poll numbers indicated
the smallest of the three majorities was Bush over Clinton. That means
Clinton would presumably have won given a voting method like MAM (which
gives precedence to larger majorities when constructing the order of
finish). (Some pundits claimed Perot was a spoiler and that Bush should
have won. They either didn't understand that in a majority cycle *each*
candidate is in a sense a spoiler, or understood but chose to mislead the
public in order to undermine Clinton's presidency.)

Regards,
Steve Eppley
-----------------------
Warren Smith wrote:
> Hello.
>
> It appears the Romanian 2009 presidential election, which just (allegedly)
> ended with Basescu winning re-election, involved two different Condorcet
> top-cycles involving 4 candidates.
> The winner Basescu (elected via plurality+top2runoff) unfortunately
> probably was the worst choice among the four (based on both pairwise
> table based on numerous pairwise polls and the official results; and
> also on an approval-style poll; I am unaware of any range-voting-style
> poll).
> Runoff-style voting (and plurality voting too -- Basescu won the 1st
> round) both severely distorted democracy in this election and also
> both elected probably the worst winner among the 4.
>
> More details will be found at this post:
> http://groups.yahoo.com/group/RangeVoting/message/12804
> which summarizes other recent posts on that same bulletin board
> and one on the election science foundation bulletin board.
>
> I hopefully will write this observation up in the form of yet another
> CRV web page... you may want to inform me of more data and news.  The
> cycles are not "strong" ones
> (i.e. they failed to have large margins in each pairwise election)
> but they look strong enough to have a high probability they genuinely
> existed.
> That will have to be evaluated, but since there actually are TWO cycles it
> seems the confidence at least one was real, is pretty good.
>
> Need to sleep now :)
> --
> Warren D. Smith
> http://RangeVoting.org  <-- add your endorsement (by clicking
> "endorse" as 1st step)
> and
> math.temple.edu/~wds/homepage/works.html