[EM] Student government - what voting system to recommend?

[Abd ul-Rahman Lomax]

At 06:35 AM 4/25/2007, Juho wrote:
>As you said, Condorcet and IRV don't see the ratings. I give another 
>example of IRV. C=Centrist, ER=ExtremeRight, EL=ExtremeLeft. (I use 
>the slightly exaggerating word "extremist" (although I just 
>complained about use of "PW" above) to make it clear that there is 
>no exact match to the main parties of current two-party systems.)

PW, of course, was a reference to two different real politicians: 
Arnold Schwartzenegger, and if I spelled his name correctly, I should 
get a medal, and Jess Ventura, both of whom were professional 
wrestlers and both of whom were elected state governors in the U.S. 
Both elections were rather unusual. Ventura was a Reform Party 
candidate In Minnesota and unexpectedly beat the two major-party 
candidates, which is fairly rare in the U.S. Apparently he was also 
very popular during his term. Arnold, who is so popular in the U.S. 
that all you need to say is his first name, was elected governor as 
part of the recall election of the prior Governor of California, 
which had a huge number of candidates on the ballot and which was, as 
usual, a plurality election. While he did not gain a majority, his 
48.6% of the vote was amazingly high to me. The next runner-up, the 
Democratic candidate Bustamante, only received 30%.

>34% - 10 ER 8 C 0 EL
>34% - 10 EL 8 C 0 ER
>32% - 10 C 5 EL 0 ER
>
>IRV first drops C and then elects EL. The centrist candidate would 
>have been a quite good end result, also from sincere ratings point of view.
>
>One more example on Range (that takes the ratings into account but 
>is vulnerable to some strategies). The example is as above but the 
>numbers are a bit different to make it clear who the "font runners" are.
>
>Sincere opinions:
>40% - 10 ER 8 C 0 EL
>40% - 10 EL 8 C 0 ER
>20% - 10 C 5 EL 0 ER
>
>Actual ballots when voters exaggerate their votes and apply Approval 
>style strategy after observing that EL and ER are the "major 
>candidates" / "front runners" (and after assuming that the final 
>decision will be between these two):
>40% - 10 ER 0 C 0 EL
>40% - 10 EL 0 C 0 ER
>20% - 10 C 10 EL 0 ER
>
>Range picks EL as the winner. The sincere opinions would have given 
>C the best score but strategies changed the situation.

Let's look at this example closely:

Average sincere ratings:
EL - 50
ER - 40
C  - 74

Range, sincere votes, C is not only the winner, C is the winner *by 
far*. In order to move this to EL as the winner, somehow the 
electorate must collectively decide to go for party affiliation 
rather than sincere ratings. Range critics, quite simply, assume that 
this will happen. Why isn't stated.

That the center voters would unanimously prefer the left candidate to 
the right, quite simply, makes no sense. However, if they rate ER 
higher, but still not as high as EL, EL still wins. What really 
doesn't makes sense is that the EL/ER voters unanimously decide to 
derate C. Given that they sincerely rate C nearly as high as their 
faovrite, this strikes me as so unlikely as to be worth calling 
"impossible" in a real election with more than a few voters.

Further, while this example is shown as if it's a defective result, 
it is defective only from the point of view or Range analysis of 
sincere votes!  EL is the Condorcet winner in both cases! What would 
be objectionable would be if strategic voting turned the result into 
a victory for ER!

What happens if the ER voters, alone, conspire to vote strategically? 
Let's assume that the EL voters are more honest! Are they harmed by 
their honesty?

EL - 50
ER - 40
C - 42

By voting insincerely, by derating C, they converted the election 
from a victory for C, 80% satisfactory to them, to one for EL, 0% 
satisfactory for them.

Far from rewarding insincere voting, Range severely punished it! One 
experience of an election like this, those voters would never again 
be tempted to lie on a Range ballot.

What is really odd to me is that examples like this are proposed, and 
asserted as examples of Range rewarding insincere voting. Insincere 
voting harms the electorate as a whole, but it most specifically 
risks harm for the segment that votes insincerely. I haven't proven 
this, it would be interesting to see an analysis. Warren or someone 
might like to look at the simulations.... This particular example is 
quite warped, so all it proves is that it is possible to completely 
overlook the implications of an election scenario by just focusing on 
one narrow aspect of it.

By voting insincerely, the ER voters, quite simply, abstained from 
the election. They might as well have! They could have, collectively, 
moved the election from disaster for them to almost complete success! 
Simply by voting sincerely....

>In summary, all methods have some problems. One needs to estimate 
>which problems are lesser and which worse.

Indeed. The example given shows, however, that Range -- at least in 
this example -- reduces to Approval if all voters exaggerate their 
ratings. Which is actually a decent outcome.

Average satisfaction with sincere: 74%
Average satisfaction with Approval: 50%

The loss of satisfaction, however, is heavily concentrated in the ER 
faction. EL improved their outcome from 74% to 100% by voting 
strategically, so one might say that they had been rewarded. But they 
are left with a polarized electorate. Would I rather see a unified 
electorate, unified on C, and thus a government which can function 
effectively and fairly, or a government that is really the chosen 
government of barely more than half the people, but which favors, in 
theory, my political positions? In the long run, such a victory tends 
to create a backlash. The Republicans, by not attempting to seek 
broad social consensus before proceeding with partisan "reforms," 
have thoroughly damaged their ability to maintain power. At this 
point, with Bush effectively pulling the rest of the party into a 
very unpopular position, hoping that arguments which failed in 2006 
to maintain hold on power will somehow magically become popular 
again, is probably setting the stage for a more thorough Democratic 
takeover of not only the Presidency in 2008, but also the gaining of 
clear majorities in both the House and Senate.

If I were a Republican, I'd be horrified! There are some Republicans 
who have claimed to see a silver lining in the cloud of 2006, that 
giving the Democrats some power will cause them to hang themselves. 
Dangerous thinking. They might, true. That is, they might follow the 
Republicans in the same error, the error of arrogance, of assuming 
that a narrow electoral victory (or even a loss converted into 
victory by legal and illegal tricks) is a mandate.

The present U.S. system tends to push toward narrow margins. It does 
not seek consensus winners, at all. Range, at least, would make it 
possible for consensus winners to emerge. Or it wouldn't. If it 
fails, it will not have made things *worse*!

>  Use of ratings would maybe be nice, but in competitive elections 
> (where giving sincere ratings means losing power) methods like 
> Condorcet seem to perform better.

The example shows what I think the simulations confirm: Range tends 
to pick the Condorcet winner. Where it does not pick the Condorcet 
winner, it picks a *better* one!

Again, I have certainly not proven this, but this is my impression 
and intuition. While I suspect that scenarios can be manufactured 
which would show strategic (in Range, this only means exaggerated) 
voting as converting a Range win into a Condorcet loser win, I doubt 
that these scenarios would be at all realistic, and thus unlikely to 
show up in randomly-generated voting patterns where the patterns are 
based on issue or approval space. (I.e., are relatively realistic.)

>  Condorcet allows the majority to rule and make decisions where the 
> sum of utilities is actually smaller that what electing the 
> proposal of the minority would have given. Out of the ranking based 
> methods Condorcet seems to me to perform better than IRV. (But if 
> you need IRV to make the case easy to explain, that is maybe not a 
> catastrophic move. :-)

The explanations of IRV that purport to make it seem batter than 
Condorcet seem to me to be thoroughly specious, essentially 
manufactured by people who have already decided to back IRV for 
reasons other than the superiority of the method. As has been noted, 
IRV is pretty bad in single-winner elections, but the real agenda of 
the IRV supporters is multiwinner PR, and the equivalent of IRV, 
multiwinner, is STV, which, because it is multiwinner, behaves much 
better than IRV.

"Core support" is a totally bogus argument.

Range *also* allows the majority to rule. Essentially, a majority 
which decides to express its preference as a strict one, will always 
prevail in Range. A majority, however, which decides, in part, to 
allow a candidate, perhaps a centrist, to be rated higher than zero, 
to that degree opens the door for a selection which is not th first 
preference of the majority, so Range technically fails the Majority 
Criterion. Which is not the same thing as failing Majority Rule, 
though the confusion is common.

The truth would out and it would become explicit if there is a runoff 
when a Majority preference and the Range winner differ. Would enough 
voters switch from their first preference to the second in order to 
effectively ratify the Range winner? In any case, the Majority 
Criterion, for the overall process, would be satisfied!

Many jurisdictions already provide for runoffs under some conditions, 
most notably majority failure. Range improves the situation from 
top-two, which, like IRV, can pass over a widely acceptable centrist 
(as it would in the example given above), to 
Range/MajorityPreference, when they differ. Which would probably be 
fairly rare. When it *does* happen, the runoff would be worth the effort.

You know, many of the supporters of EL and ER, above, would simply 
stay home.... After all, they are happy with the outcome either way! 
C supporters, though, would be highly motivated to turn out!

Really, there is a paucity of imagination among Range critics. The 
common opinion that Range would devolve to Approval is an example. 
It's really a bizarre position, when examined in detail. Too often, 
though, the critics come up with a scenario like that above, see an 
alleged reward for strategic voting, and stop there.