[EM] IRV vs Plurality

[Jonathan Lundell]

On Jan 14, 2010, at 11:00 AM, robert bristow-johnson wrote:

> simply, if a Condorcet winner exists, and your election authority elevates to office someone else, that elected person is rejected by a majority of the electorate.  what other democratic value papers over that flaw?  LNH?  monotonicity?

Terry covered pretty much what I have to say on the subject.

> 
> like the popular vote is the gold standard we use to judge how well the Electoral College does, it seems to me that the Condocet criterion is the gold standard to use to judge how well some other method works.  in both cases it seems logical to ditch the "experimental" method and just use the gold standard.

But it's not (the gold standard) as things stand now. 

Sure, we should ditch the EC and move to national IRV ;-). But with the EC in place, a candidate will (justifiably) campaign for electoral votes. In 2008, it was a waste of funds for McCain to campaign in (say) New York, Massachusetts or California, so he didn't bother to compete for a bigger share of the popular vote, which presumably he could have gotten by spending some campaign cash. Obama, on the other hand, was motivated to spend at least *some* time campaigning in those states, if for no other reason but that they were a good *source* of funds for him.

Regardless, you'll recall that the big fuss in 2000 wasn't really over the popular vs electoral vote, mainly, I think, because most people understood (from campaign coverage) that the electoral vote was what counted.

To be clear, I repeat: I don't think that's the way it should be. But we must interpret voter and candidate behavior in the context of the existing rules.