[EM] Why I think IRV isn't a serious alternative 2

[Abd ul-Rahman Lomax]

At 05:18 AM 12/22/2008, James Gilmour wrote:
>But, of course, if it were possible to elect a "no first 
>preferences" candidate as the Condorcet winner, such a result
>would completely unacceptable politically and the consequences would 
>be disastrous.

No example is known to me. It's easy to see examples in small direct 
democratic situations where this compromise could clearly be the best 
result. We know that sometimes the best candidate, for example, 
doesn't make it to the ballot, even. Suppose there is a majority 
requirement, two candidates on the ballot. But so many voters would 
prefer candidate C, that some of them write it in, causing majority 
failure. If it's top two runoff, and write-ins are allowed, and 
*especially* if the runoff method doesn't cause a serious spoiler 
effect, the write-in can win the runoff *with a majority*. Or, in 
spite of the obstacles, with a plurality. Given the obstacle of not 
being on the ballot, it's quite likely in that case, that a majority 
would ratify the election.

How would this be "disastrous?"

>The two situations I had in mind were:
>Democrat candidate D;  Republican candidate R;  "centrist" candidate M

Centrist candidate M, let's say, was a Republican who didn't get the 
party's nomination because he didn't please the right "core" of the 
Republican Party. He's popular with many Republicans, maybe just 
short of a majority.... and he's popular with many Democrats, maybe 
even most of them. He runs as an independent in the election, or as a 
"Reform Party" candidate.


>Election 1
>35% D>M;  33% R>M;  32% M
>
>Election 2
>48% D>M;  47% R>M;  5% M
>
>M is the Condorcet winner in both elections, but the political 
>consequences of the two results would be very different.

Note that in both elections there is Majority failure. Thus in a 
primary-majority required situation, there would be a runoff. Given 
the Condorcet principle, and the same electorate and votes, M, if 
allowed to be on the ballot, would win the runoff against either of 
the other candidates.

If not allowed to be on the ballot, it would not escape the notice of 
the supporters of M that M is the Condorcet winner, a runoff write-in 
candidacy makes sense, as long as it doesn't spoil it.

The election of either the R or the D produces a result which is 
unsatisfactory to the majority. Majority rule requires something 
different. Majority rule requires a disaster? Minority -- plurality 
-- rule is better?

Bucklin in the runoff handles this situation with ease -- even if a 
write-in candidacy is necessary. The situation probably would not 
exist in the first place -- the need for a runoff -- with Bucklin or 
a Condorcet-compliant method. Note that in both cases, ballot 
truncation shows significant preference gap of M over other 
candidates, and minor preference gap between the D and R candidate.

How in the world would the election of M be a poor result? This is 
the second preference candidate of *everyone*. And that doesn't mean 
"lesser evil"? With poor core support in the second election, M is 
nevertheless considered a good alternative, a good compromise.

You are standing in a relatively isolated position, James. Robert's 
Rules of Order considers this failure to find a compromise winner a 
serious argument against sequential elimination ranked methods.

>   My own view
>is that the result of the first election would be acceptable, but 
>the result of the second election would be unacceptable to the
>electorate as well as to the partisan politicians (who cannot be 
>ignored completely!).

Actually, partisan politicians voiced strong objections to 
preferential voting systems when they "won" the first preference 
vote, but lost when voluntary additional preferences were added in 
(Bucklin) or were substituted in (IRV).

The electorate, however, was undisturbed, except for minorities 
supporting those politicians. Thus in Ann Arbor, MI, the Republicans 
arranged a repeal of IRV, scheduled when many of the students who 
supported the Human Rights Party and Democratic candidate were out of 
town. They won, with low participation in the repeal.

There is no substitute for the majority being organized! Which 
organization must reach across party lines.

>   If such an outcome is possible with a
>particular voting system (as it is with Condorcet), that voting 
>system will not be adopted for public elections.

Bucklin, which makes the result possible, was adopted and wasn't 
rejected by the electorate because of this. It was rejected, often 
not by the electorate per se, for other reasons; the idea that the 
first preference winner should win was used as an argument as part of 
this. Want to stand on that side, the side that favors party power 
over public power? It's your choice!