[EM] Voting by selecting a published ordering (was sidetracked to EC)

[Abd ul-Rahman Lomax]

At 04:22 PM 4/28/2006, Simmons, Forest wrote:

>In your scenario of
>
>35M  Bush>McCain>Gore
>20M  McCain>Bush>Gore
>45M  Gore>McCain>Bush

actually, what was given by Mr. Eppley were vote totals for candidates:

     Bush     selected by 35,000,000 voters
     McCain   selected by 20,000,000 voters
     Gore     selected by 45,000,000 voters

Mr. Simmons apparently assumes that votes will be independent of the 
lists published by the candidates. It is highly unlikely that 
candidate choices, published in advance of the election, would have 
*no* effect on votes.

>Suppose Rove saw this coming, and not relying completely on 
>slandering McCain this time, decided to publish the Bush ranking as
>
>Bush>Gore>McCain .
>
>Wouldn't it be to McCain's advantage to fail to rank the other two candidates?

Do we assume that McCain knows Rove's choice before making his own? 
Suppose somehow he does. What would he do then?

Unstated here is the election method being used. But if we assume 
that it is a Condorcet-compliant method, and McCain does as suggested 
he might by Mr. Simmons, and there is no effect on votes from the 
published rankings (if this would be the case, why would they be 
published?) we'd have

35: Bush>Gore>McCain
20: McCain
45: Gore>McCain>Bush

the races:
Bush vs. Gore 35:45 Gore wins
Bush vs. McCain 35:65 McCain wins
McCain vs. Gore 20:80 Gore wins

Condorcet winner: Gore

Suppose McCain didn't truncate his rankings. If he voted second for 
Gore, the outcome would still be Gore as the Condorcet winner, with a 
vote of 65:35. If he voted second for Bush, Bush would win the 
pairwise contest with Gore, 55:45.

McCain's abstention would be a fitting response, actually, to the 
betrayal by Bush (and Bush will be the one signing off on that vote, 
not Rove). He'd be handing the election to Gore without having 
actually voted for him.

What would have happened if Bush had ranked as party affiliation 
would have predicted, and likewise McCain?

35: Bush>McCain>Gore
20: McCain>Bush>Gore
45: Gore>McCain>Bush

McCain wins. Now, tell me this: why would Rove trash Bush's career, 
as well as his own, by recommending that Bush vote as described, 
simply to take the election away from McCain? And why would Bush take 
the advice?

It does appear that McCain is the proper winner of this election, if 
we assume that the voters supporting the candidates, on average, 
would vote the same way as described in this last scenario.

But something has been neglected here, which is the effect on the 
voters of the Bush betrayal. I fail to see why the Bush position 
would be preserved through that betrayal. What would be more likely, 
I think, is that McCain would gain votes from those who would 
otherwise have voted for Bush. Would that be enough to turn the 
election? I don't know. But it looks to me that this scenario shows 
that Published Ordering is an improvement over present methods.

What has been overlooked is that Published Ordering, by candidates, 
turns the candidates into electors. Votes for a "losing candidate" 
are not wasted. So I'd expect there to be *many* candidates, the 
situation would be nothing like the simplistic scenario presented. 
Personally, I'd prefer that the candidates become, simply, electors, 
*not* bound by a published ordering or pledge. If I trust candidate X 
to be President, surely it is reasonable for me to trust him or her 
to pass on my vote in selecting the President, if it is not going to 
be him or her. If I don't trust the candidate to do such a relatively 
simple thing, why should I trust the candidate to actually function 
well in the office, if elected?

Note that if candidates become electors, and if there are many of 
them, it becomes quite possible that someone not running for the 
office, not on the ballot, and not getting any direct votes in the 
election could be elected President. This is the kind of effect we 
could get if we had a *real* electoral college: someone as President 
who did not run for the office, but who was trusted by a majority of 
the electors. I think this is likely to be much better than what we 
routinely get under current practice!

Anyway, this is the reason why this discussion devolving into a 
discussion of the Electoral College was not so much of a sidetrack. 
The two topics are closely related.