[EM] Burlington dumps IRV; Immunity from Majority Complaints (IMC) criterion

[Abd ul-Rahman Lomax]

At 08:13 AM 7/5/2013, Jameson Quinn wrote:
>IMC seems to me to be too narrow to be a general criterion, if only 
>one custom-built voting system passes it. WIMC is an interesting 
>refinement of Condorcet and Smith. But neither belongs on Wikipedia 
>without a "reliable" citation.
>
>2013/7/5 <<mailto:seppley at alumni.caltech.edu>seppley at alumni.caltech.edu>
>Should IMC and WIMC be added to Wikipedia?

The obvious first: Many Wikipedia articles violate Wikipedia policy, 
it's an issue of what one can get away with. If anyone objects, and 
knows how to pursue the process, such articles would quickly be 
deleted. No wiki is considered a reliable source, for starters, nor 
are mailing lists. Neologisms will generally be rejected even if 
there is a single "Reliable Source." But a redirect might be allowed 
for a neologism, if it has widespread usage.

Many *existing* articles are vulnerable to deletion.

As to IMC and WIMC, I'm not going to go into detail, but I have many 
times indicated that there is a problem with resolving an election 
contrary to the wish of a majority. In classical deliberative 
process, *no decision is made -- ever! -- without the express consent 
of a majority of those voting.* That includes, in the basis for 
majority, spoiled ballots, it only excluses *completely blank ballots 
with no mark.* "Scrap paper," Robert's Rules calls them. No majority, 
election *fails.*

Yet we know that it is possible, with a practically-ideal voting 
system (Range), that the optimal choice is not the first preference 
of a majority. So ... we can't just assert as some sort of absolute, 
the Majority Criterion. Rather, a full resolution of this would 
involve another poll. It is possible that if the voting system 
indicates a very weak preference on the part of the majority, under 
some conditions, it might be possible to complete the election. 
Defining those conditions would, however, be relatively complex. It's 
more common that there is no majority preference, only a plurality, 
and, under those conditions, it is well understood that a normal 
solution is a runoff election. So the problem reduces, then, to 
making sure that the nominations for that runoff are likely to truly 
represent the will of the majority, as might be inferred, i.e., that 
they will include a majority preference.

It is commonly assumed that runoffs will have two candidates only. 
That's artificial, for sure, and it sets up finding a majority as if 
it were, itself, an absolute. A good voting system in the runoff can 
handle three candidates, and, remember, a runoff election is a very 
different animal than a primary. The voters will be much more informed.

The point about violating the wishes of a majority is clear: it can 
easily lead to dumping the voting system and other responses, and the 
only way around that would be to *disempower* the majority. Kind of a 
Bad Idea in a democracy, eh? Exactly who does this job, who holds the gun?