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

[Abd ul-Rahman Lomax]

At 07:54 PM 12/29/2008, Kathy Dopp wrote:
>There are so many examples of provably incorrect and misleading
>statements being made by Fair Vote and other IRV/STV proponents, even
>after these proponents were amply informed of the falsity of their
>statements, that the only conclusion one can reasonably draw is that
>these IRV/STV proponents are deliberately trying to mislead the
>public, in which case, the avowed publicly stated goals of IRV/STV
>proponents must also be treated as suspect.

Indeed.

What happened was that a political cynic and spin doctor was given 
leadership of FairVote. What matters to such people is winning, and 
truth doesn't matter. Sound bites, brief, reasonable-sounding 
arguments matter. Such a one will appeal to ignorance, use the 
ignorance of people -- an ignorance which is natural when facing a 
topic without study -- and manipulate it to the effect he desires.

However, information about voting systems is spreading. FairVote is 
starting to run into obstacles, people who are *informed* about 
voting systems. There were classic arguments against IRV, many of 
them quite ignorant.

However, it's fascinating to read debates back in the 1920s about 
American Preferential Voting vs. English Preferential Voting, i.e., 
Bucklin vs. IRV.

We see many of the same arguments then. What was missing, though, was 
the kind of understanding of voting systems that arose when 
economists started studying voting. Arrow's work was seminal, but 
it's almost as if Arrow wasn't an economist; he missed utility, 
discarded it as impractical to use. It was a strange lacuna. Since 
then, though, it has been discovered that there is a unique solution 
to the problem of determining an overall social order from a set of 
individual preference orders while satisfying general Arrovian 
conditions: and it involves using, not merely preferences, but 
utilities, a particular kind of utility that factors in 
probabilities. Popularly, an approximation to this solution is Range Voting.

And Bucklin, American Preferential Voting, is a tweak on Range Voting 
that continues to satisfy the Majority Criterion. It's 
"instant-runoff Approval," and it was used in over fifty U.S. cities 
for a time, beginning in roughly 1915. It's simple to canvass, it is 
a sum-of-votes method, all the votes are counted, all the votes count 
(at least, if any vote in a rank is counted and used, all the votes 
in that rank are counted and used. Bucklin terminates and does not 
count lower ranks when it finds a majority, which is why it respects 
the Majority Criterion. And, unlike Approval Voting, one can specify 
one's unique favorite, while still participating in the rest of the election.)

Contrary to what is sometimes implied, Bucklin wasn't generally found 
unconstitutional; the reverse was true. I've encountered only two 
cases where Bucklin implementations were tossed: Brown v. Smallwood, 
where, contrary to FairVote propaganda, the decision was clearly 
against *all kinds* of alternative votes, not just the particular 
Bucklin method, and Dove v. Oglesby, where Bucklin itself wasn't 
rejected, but an additional requirement that voters rank additional 
candidates or their first preference votes wouldn't be counted.

(I.e., what they do in Australia!)