[EM] Correction of false statements by Ossipoff & Schudy about range voting.

[Abd ul-Rahman Lomax]

At 04:24 PM 7/22/2007, Steve Eppley wrote:
>I partially agree with Chris Benham (see below).  Warren Smith's
>example, in which a voter has total knowledge of all other votes before
>casting her own vote, is implausible in the elections we're interested
>in reforming.

Implausible but possible. Secret ballot is *not* an assumption for 
voting methods, unless stated. That is, total knowledge is among the 
possibilities an election method faces. Since Warren wrote what he 
wrote only as an off-the-cuff and trivial example, to refute a 
categorical statement not qualified to limit it to certain classes of 
elections, he was able, as election theorists often do, to pick *any* 
example showing different.

Now that we are starting to see qualifications appearing in the 
assertions about Range strategy, more generally applicable examples 
need to be designed. And that's what I'm doing. I *started* with a 
two-voter case, zero knowledge, three candidates, Range 2. I have not 
presented a generalization to N voters. Warren's simulations do cover 
such larger elections, however, it would be worth looking at his 
writings on the subject.

What Warren's work shows is that, as knowledge of the presenting vote 
distribution excluding the voter's vote grows, the optimal strategy 
may shift away from sincere. But writers such as Ossipoff and Benham 
have claimed that sincere voting is never optimal, without presenting proof.

(It's certainly possible that proof has presented, but it has not in 
the present discussion, nor has proof that is accessible been referenced.)

>Recall the example someone posted here several weeks ago intending to
>undermine the value of the majority rule criterion. Three friends, say
>X, Y and Z, are ordering a pizza. (It doesn't matter if they're friends;
>in the worst case they're not and they expect to never hear from each
>other again.)

Yes, that's the worst case. And with a poor election method and the 
insistence of the victors on implementing their preference, it is 
practically guaranteed as an outcome. I'm the author of the "pizza 
election" example.

>   Z is terribly allergic to mushrooms so he strongly
>prefers pepperoni, but a mushroom pizza is slightly preferred by X and
>Y.  There was no time to deliberate--which of course is implausible in
>the elections we're interested in reforming--so majority rule picks the
>mushroom pizza.

That's not the precise example, but we can accept it as a substitute.

>Or does it?  When X or Y proposes mushroom pizza, what if Z responds by
>proposing "pepperoni pizza plus the transfer of $1 from Z to X."

Deliberative process. Not election method. Sorry, I expect better 
from Eppley. In deliberative process, the preferences and preference 
strengths of the factions become exposed, which is why deliberative 
process is the ideal "election method." But the subject here is 
generally aggregative, not deliberative.

By noting that engaging in deliberation the voters will, by majority 
rule, choose the correct pizza -- in a functional society -- Eppley 
is making a point that I've made many, many times, to the derision of 
people like Ossipoff -- and of one of the dedicated Range supporters 
(Shentrup), but not of Smith and Kok; Kok clearly gets my argument on 
this, and Smith is the inventor of Asset Voting, which is really a 
form of deliberative process.


>   When Y
>hears this proposal, he thinks to himself that he'd prefer "pepperoni
>plus 50 cents" over mushroom, so Y proposes "pepperoni pizza plus a
>transfer of 50 cents from Z to Y."  Suppose X is even more indifferent
>between pepperoni and mushroom than Y is, and would prefer mushroom over
>pepperoni for just a dime.  X is clever, though, and bids 49 cents
>instead of a dime.  There was no deliberation; Z never admitted the
>allergy.  X and Z, a majority, both prefer X's final proposal over
>mushroom pizza.

There is negotiation. Some political scientists consider negotiation 
a separate method than deliberation, I don't. But negotiation is 
*not* aggregation, unless you incorporate it into the method. For 
example, if your Range Votes were the amount you would pay to get the outcome.

However, in functional societies, a real person with Pepperoni allegy 
would simply disclose it. Negotiation is used, typically, when 
deliberation without offers of logrolling fails to come to sufficient 
agreement. And negotiation is not used by the majority, in 
majoritarian systems; it is, rather, initiated by a minority which 
wishes to induce the majority to change its preferences.

What Eppley is doing is setting up a negotiation process which is 
equivalent to Range Voting, by assigning values to the choices.

This is tantamount to sincere Range voting, thus making my point!

>Economists and political scientists call the transfer of 49 cents from Z
>to X a "side payment."  Side payments are a specific case of the more
>general solution: proposals that bundle alternatives.  That can also be
>called vote trading. in the case where the bundling is accomplished by
>trading votes on otherwise unlinked issues.

Negotiation is a general term that covers it.

>[... interesting stuff deleted....]

>Many people in this maillist make the same
>mistake--treating the set of alternatives and the voters' preferences as
>constant when comparing wildly different voting methods--and it's a huge
>mistake.)

Indeed.

>If the good preference order method also permits each voter on election
>day to begin by selecting a ranking published before election day and
>modifying it if desired--perhaps by drag & drop; see the feature in the
>new NetFlix user interface for an example--before submitting it as her
>vote, then we won't have to worry, when there are many candidates, about
>the possibility that voters will fail to rank a compromise candidate
>needed to defeat a "greater evil."  This could also sharply reduce the
>amount of campaign money needed for good candidates to win, since it
>doesn't cost much for a candidate to contact the people publishing
>rankings (presumably other candidates, and some NGOs).  For more
>information on Voting by Selecting a Published Ranking see my message
>posted in April 2006.

Selecting a Published Ranking is a variation on Asset Voting that 
preselects the assignments. I prefer, however, a system which allows 
the candidates to alter their vote reassignments intelligently, based 
on negotiation and deliberation. The reduces the voter's decision to 
a simple one: whom do I trust most? And, generally, with Asset 
Voting, the voter is not restricted more than very minimally to whom 
the voter can choose. The voter can, himself or herself, become an 
elector, with subsequent votes being public. Should the voter be 
required to follow a published list?

Note that we do not require elected candidates to follow any 
preselected pattern of voting. And this is not only permissible, it 
is actually necessary for true deliberative process.

>Some people here believe Approval is best, or like Mike Ossipoff seem
>willing to settle for Approval.  One of Mike's arguments is that it
>would be good for society to minimize the number of voters who
>(strongly) disapprove of the winner.  He seems to define disapproval as
>some strong negative emotional response rather than as a preference for
>leaving the office vacant (which I believe--based on a conversation with
>Matt Jackson, professor of economics and political science at
>Caltech--is the standard definition in the rational choice model).

I have in the past, in many other fields, discovered that my 
intuitions and ad-hoc analyses, innocent of formal training and 
study, often coincide with expert opinion, but *not* with the 
opinions of pseudo-experts. It appears that it is happening here. I 
follow, typically, common-law arguments, based on a large body of 
practice and thought. I can sometimes err drastically when straying 
into specialized fields, I'm certainly not asserting that I have some 
special authority, but only that my intuitions are sometimes cut 
through the clutter.

Perhaps I'll be able to visit Caltech again. I was at Page House for 
two years, 1961-3, and lived off campus later. If I visit, I'll 
certainly look up those in the political science field!

>Those negative emotions may or may not be something society should
>minimize.  Even if they are, it does not follow that voting methods that
>ask voters to express approval or disapproval will do this as well as
>voting methods that ask voters only to rank the candidates in order of
>preference.

It's particularly difficult to minimize "strong disapproval" if you 
don't distinguish, in the method, between strong and weak 
disapproval! Only Range does this, of the methods on the table, 
though ranked methods or analysts do sometimes assume that the 
presence of many intermediate candidates in a ranking is a function 
of strong disapproval. It's usually true, but not universally true.

>   I believe the best way to minimize those negative emotions
>is with a voting method that encourages candidates to compete to be the
>best compromise and doesn't deter candidates who don't care about
>winning from taking "uncompromising" positions.  It would enable voters
>to rank the less corrupt compromise candidates over the more corrupt
>compromise candidates, since their platforms will be similar.  Election
>day is for making a collective decision, involving compromising as far
>as is necessary to defeat what's worse; the rest of the time is
>available for deliberating over what's really best.

You can go to great lengths to use rankings, but ratings deal 
directly with the problem. And I have come to the conclusion that 
Range (and Approval) should include a preference analysis runoff; 
preference analysis (i.e., ranking) is used, based on the Range 
ballot and possibly some additional information, to determine if 
there is, in fact, someone else who might prevail in a pairwise 
election against the Range winner.

Thus the majority is actually asked if it is willing to give up a 
small benefit in order to provide a larger net benefit to a minority. 
My opinion is that if the Range result has not been distorted in some 
way, it will normally prevail. But it's up to the majority to decide!

(And I've elsewhere presented arguments as to why I think this: the 
initial majority is less motivated to vote in the runoff, whereas the 
minority is strongly motivated. Unless the Range result is wonky. 
Some members of the previous majority may have such a weak preference 
that they, quite rationally, don't bother to vote, it isn't worth the 
effort, they are already guaranteed a good outcome.)