[EM] some July 4 comments

[Adam Tarr]

John B. Hodges wrote:

>(JBH) It occurs to me that the advocates of Condorcet/Approval etc. are in 
>the same position as advocates of socialism in the years before 1917. 
>(See, for example, LOOKING BACKWARD by Edward Bellamy.) They are comparing 
>a theoretical system as they imagine it would work, to a real system as it 
>has been found to work in practice. Naturally the imaginary system comes 
>off best.

I don't see anyone here attempting to gloss over the "bad" scenarios for 
Condorcet.  We've seen the "hated middle" or "turkey" scenario written up a 
dozen times in the last month.  Just to show how sporting I am, I'll 
re-print the most disturbing Condorcet scenario I've ever seen:

36% George
9% Al>George
18% Al>Ralph
18% Ralph>Al
19% Ralph

I call this the most disturbing I've seen, because it is one of the 
exceedingly rare cases in winning-votes based Condorcet where voting a 
second preference can hurt you.

>>John Hodges wrote:
>>>Whether IRV is the best method possible is
>>>open to debate; but it ain't all THAT bad. C'mon, people.
>>
>>I think it is comparably bad to plurality, because it still encourages 
>>two-party
>>dominance.  An electoral method that can't even avoid that, in my view, isn't
>>worth the effort to implement.
>
>(JBH) Oh, really? I'd love to hear your reasoning on that. Even better 
>would be some evidence. How many significant parties are there in 
>Australia and Ireland? In ELECTORAL SYSTEMS, A Comparative Introduction, 
>by David M. Farrell (2001) I find mention of seven different parties in 
>Australia.

But how many of them win significant numbers of seats in Australia's upper 
house, which is the only one that uses IRV?  I imagine lots of parties get 
seats in the lower house, which uses a form of proportional representation 
(STV).  So, saying that there are lots of parties is sort of skirting the 
issue of whether IRV allows for more than two strong factions.

>[...] But I think any honest person would grant that IRV drastically 
>reduces the "spoiler effect" relative to Plurality.

Reduces?  Sure.  But as Kevin wrote, this is largely due to third parties 
getting too few votes to be significant.  As soon as a third party gets 
near a third of the votes, then suddenly there's tons of possibilities of 
manipulation, spoiler effects, and seemingly undemocratic results.  The 
fact that there has generally only been two strong parties in Australia's 
IRV elections is what has kept this from happening.  The same is true in Malta.

>Another point: Assume Condorcet-plus-some-completion-method; instead of 
>two-party dominance, should we fear one-party dominance? Condorcet methods 
>strongly tend to pick centrists. Could one party stake out the center and 
>sit on it, unassailable? Or perhaps, might we see a competition of several 
>"center" parties each vying to be indistinguishable from the others,

This doesn't seem too likely.  Parties have always attempted to attract a 
base of voters and appeal to them on a range of issues.  Picking the median 
position on everything doesn't tend to work well.

If this sort of fear were well-founded, then it would apply to plurality as 
well, and yet the Democrats and the Republicans are quite distinct on a 
wide range of issues.

>the Society for Social Choice and Welfare elected their president by AV, 
>and ALSO asked each voter for a complete preference ranking of the 
>candidates, as a "test of voting theory". There were three candidates, all 
>highly regarded, so it was a competitive election. It turned out that the 
>AV winner was not the Condorcet winner. It also turned out that only one 
>voter listed a second choice in the AV election; since everyone knew the 
>election was close, most chose to "bullet-vote" for their first choice. 
>Saari speculates that AV will reduce to Plurality in close races.

As an aside, Saari may be a brilliant geometric mathematician, but he's no 
better than the rest of us in predicting real-world outcomes.  At any rate, 
this test seems to be a poor paralell to political elections for several 
reasons.  The most obvious ones are:

- the election of the Society for Social Choice and Welfare probably 
doesn't have a definable "issue spectrum" like political elections do; this 
implies that there won't be a clear second choice for most voters.

- the voters apparently had little prior knowledge of the relative position 
of the candidates.  Knowing whether your candidate is likely to win is 
crucial in deciding whether to vote your second choice as well.

In the end, if the race is actually very close, and most people don't have 
a strong second choice, then bullet voting does make sense, and the result 
isn't bad at all from a social utility perspective.

>Saari makes the point that, to get a definite predicted result for AV, you 
>not only have to specify a profile of voter preferences, you ALSO have to 
>specify some assumptions about voter behavior in choosing how many 
>candidates to "approve". I.E in an election with five candidates, in AV 
>the voter may vote for one, two, three, or four; which will they choose to 
>do and why? With ONE assumed profile of voter preferences, AV may produce 
>just about ANY result, depending on how the voters choose to set their 
>individual thresholds of approval.

I agree that different people will vote different ways in approval; that said:

- People just voting for who they like better than average, AND people 
voting in a highly strategic game-theoretic fashion, tend to produce good 
results from what I have seen.  The former give "high utility" (i.e. 
generally well-liked) candidates, and the latter tends to elect Condorcet 
winners.

- Much like currently in Australia, you can be sure that parties will 
advise their followers on the optimal strategy for their party.  And no, 
this would not always be bullet voting.  Parties on the same side of the 
spectrum would most likely cut deals with each other.

>One implication of this is that ANY quirky (undesirable) result possible 
>under Plurality, Runoff, Borda, or any other "positional" method may 
>theoretically be reproduced under AV,

It takes truly tortured reasoning to come up with a voting approach for AV 
that would reproduce my IRV "nightmare scenario" where Right wins a narrow 
victory over Left while Center sits out.  There are also several "clone 
candidate disasters" I could dream up involving the Borda count which would 
be extremely hard to reproduce with IRV.  Of course, neither of us support 
Borda anyway so it's beside the point.

>>>Oddly enough, even Condorcet-completion methods cannot be relied on
>>>to select the Condorcet winner, once you take into account the
>>>incentives for and effects of insincere voting.

Those incentives are varnishing small in winning-votes based Condorcet.  I 
furnished one such example at the top of this message, but I defy you to 
come up with another one from scratch.

>(JBH) Plain Condorcet is not an option, because it does not reliably pick 
>a winner. There may not BE a candidate or party that beats all others in 
>pairwise comparisons. All Condorcet methods are Condorcet-Plus, some 
>additional procedure for picking a winner when plain Condorcet does not.

Many "Condorcet methods" make no explicit mention of picking the Condorcet 
winner.  Ranked Pairs, for example, is a positional method that satisfies 
the Condorcet criterion, and yet does not read, "pick the Condorcet winner, 
and if none exists then..."  Picking the Condorcet winner, if one exists, 
is just a natural extension of ranking and locking pairwise defeats.

>So, basically what you have shown with this example is that pure center 
>parties can get squeezed out under IRV. To remain viable, a pure center 
>party has to stay larger than at least one wing, taken as a coalition. It 
>has to poach voters from at least one wing; it cannot allow itself 
>passively to be poached upon.

In other words, being the largest and most popular party, as well as being 
the compromise choice, is not good enough.  I see this as completely absurd.

>probably the equilibrium solution involves two parties glaring at each 
>other over the 50-yard line, each fighting also to guard their rear from 
>additional parties.

Which is, of course, the same predicted equilibrium of plurality.  Does 
Kevin's criticism make sense now?

>My own conclusion is that one-seat elections should be avoided wherever 
>possible. Legislatures should be elected by proportional representation, 
>with a "district magnitude" of at least five seats. Single-seat methods 
>should be reserved for electing executives, such as Mayor, Governor, 
>President.

Again, this is a laudable goal the I agree with completely.  However, if 
this actually becomes reality, then doesn't it argue all the more for using 
Condorcet in the single-seat elections?

Proportional representation will tend to produce a result that is both 
highly representational of the electorate and (for that exact reason) very 
stable.  In fact, I believe there is a running joke in Cambridge that PR 
stands for "perpetual re-election".  Given this, it makes sense to have an 
executive who is both moderate (i.e. in the center relative to the 
distribution of the legislature) and fairly stable.  Both of these 
properties could be expected of a Condorcet winner.

Food for thought, at any rate.  Hope everyone enjoyed the 4th.

-Adam