[EM] Real IRV Election, Disputable Result

[Jonathan Lundell]

At 9:06 PM -0700 3/12/06, Jan Kok wrote:
>At any rate, I think the incentives for strategic voting occur much
>more often with IRV than with Condorcet.  If Condorcet gets a
>different winner than IRV when applied to ballots from an IRV
>election, I think the burden of proof would fall on the IRV advocate
>to show that the CW determined from the ballots was not the true CW,
>or that there was no CW.

I'll respond in more detail to the rest of your message, but my 
underlying point is not whether voters would be more or less inclined 
to vote strategically under one method or the other, but that the 
strategic considerations would be *different*.

Likewise, the detailed voting instructions (not necessarily on the 
ballot, but the associated "how does this system work" instructions). 
When I'm instructing voters in an STV election (IRV being of course 
the single-seat version of STV), part of my instructions are that 
they can freely rank as many candidates as they like because later 
rankings can have no detrimental effect on their prior choices. If 
the ballots are to be counted with a Condorcet method, I can't say 
that (or rather it would be untrue if I did).

And yet it's a helpful and reassuring instruction, and I'm fairly 
sure it makes a difference to some voters. It makes a difference to 
me.

I apologize for harping on an advantage of STV; most of my real-world 
ranked-voting experience is with Green Party (US and California) 
internal elections, and we use STV there. There are of course other 
assurances that I could make to Condorcet voters that I could not 
make to IRV voters.

I don't want to minimize the problems of STV. Its failure to find 
"everybody's second choice" when the electorate is badly 
factionalized is only the most obvious of a whole class of 
premature-exclusion issues (David Hill and Simon Gazeley try to 
address this problem with Sequential STV, but it's not a perfect 
solution--and of course there can't be a perfect solution).

But to repeat myself: we should expect a different election profile 
depending on whether we run an election with STV or Condorcet rules. 
No doubt we'll always be examining IRV elections for Condorcet 
upsets, but the interpretation of such an event (which we do not 
appear to have in Burlington) is non-trivial.
-- 
/Jonathan Lundell.