[EM] To Bill Lewis Clark re: stepping-stone

[Bart Ingles]

Bill Lewis Clark wrote:
> 
> So, when (if ever) does IRV perform worse than Plurality?

The recent 2002 and 2003 gubernatorial elections in California might be
an example.  Although not an IRV election, the 2002 elections were held
in two rounds, and had three candidates who were clearly more viable
than the others.  Thus the results should have been the same if IRV or
two-round runoff were used rather than a party primary and general
election.

Moderate Republican Riordan was believed by many to have been a sincere
Condorcet winner, defeating both Davis and Simon in head-to-head polls. 
But Riordan was eliminated in the primary, leaving Simon and Davis in
the general election.  While there was some talk of the Davis camp
supporting Simon in the primary, most voters weren't able to grasp the
idea of the party's front-runner being a likely spoiler.

In a near replay of 2002, but using FPP this time, Arnold took Riordan's
place as the moderate republican, and McClintock replaced Simon in the
recall election of 2003.  With the simpler FPP system, more voters
understood the need to support Schwarzenegger in order to defeat Davis. 
If the two-round system hadn't confounded this need for strategy in
2002, there probably wouldn't have been a recall.

Far from being an anomaly, in situations where (1) the Dems and Reps are
competitive, enough so that the race can be tipped by centrist swing
voters, and (2) the non-incumbent party is split between a moderate and
an extreme faction, with the extreme faction slightly larger than the
moderate group, then IRV will fairly reliably throw the race to the
incumbent.  In other words, the final result will be the same as with
Plurality.  But at least with Plurality, the extreme voters won't be
under the illusion that winning the primary necessarily means that their
candidate is the best-equipped for the final round.

Other advantages of Plurality over IRV:

1) simplicity
2) cost
3) relative ease of conducting recounts
4) higher worst-case social utility

As for Dean and Kucinich being national candidates in favor if IRV, I
have a hunch they won't be for much longer.   YEEHAW!!  ROFL

Bart