[EM] Reply to Randy

[MIKE OSSIPOFF]

Randy Kunkee posted some comments on statements regarding IRV at
the website at http://www.electionmethods.org
I'm not the website owner, but I participate in the website. I'll
reply here to some of Randy's statements. I'll also forward Randy's
letter to the website's owner, who may reply when he has time, if
he considers the comments worth replying to. I tend to have lower standards 
regarding what deserves a reply, hence my reply to this
message. Actually I don't waste time on replying fully to long
letters by IRVies, but I'll make an exception this time. But if
Randy posts a long reply to this, I won't reply, and that _won't_\
mean that Randy has said something irrefutable. It will just mean
that I don't waste time on IRVies.

Randy Kunkee wrote:

The page referenced above makes several vague, unsubstantiated arguments.
The claim that IRV votes cannot be summed is flat out wrong.

I reply:

Perhaps Randy didn't read Russ's definition of summability. Maybe
Randy's talking here about precincts sending in sums of how many
people voted each of the more than N! possible rankings. I answer
about that lower in the letter when Randy  brings that subject up.

Randy continues:

The phraseology is deceptive (it is your web page, you are welcome to do
that).  For example, take the sentence


     IRV therefore seems to allow supporters of minor parties to cast
     protest votes without ``wasting'' their votes.

In my view, there is no "seeming" about it

I reply:

Thank you. But stating that, in your view, things are different
isn't what we call a refuting argument.

In IRV, sincere voters will often regret that they didn't
insincerely vote a lower choice over their favorite. That will happen
when someone even worse wins because the lower choice compromise
is eliminated before your traveling vote reaches him/her.

Randy continues:

You go on to state:


     This advantage is illusory, however, because it applies only as
     long as those minor parties are sure to lose

The advantage is not illusory.  If an Al Gore is campaigning and knows that
getting the Nader votes in round 2 will win him the election, then he will
try to appeal to them.  Likewise, a George Bush would realize that he has to
moderate his position because he knows he's not really in the majority to
start with

I reply:

We said that in IRV, the voter's confidence that it's safe to vote
sincerely is illusory unless Favorite is a sure loser. It isn't quite
clear why you think that your comments above refute our statement.

If you vote 1Nader, 2Gore, and Gore gets eliminated first, and
Bush wins, then any confidence that you had in the safety of sincere
voting in IRV was misguided.

IRV is not perfect.  But I will argue that at its worst, it is no worse than
our current plurality system

Oh that's reassuring :-)

Aside from that, IRV is notorious for its failure of the
Monotonicity Criterion. Even Plurality doesn't fail that one.

Randy continues:

, and at its best, it is a great improvement.

I reply:

But not in the ways that are considered worthwhile by people who
study voting systems.

Randy continues:

In particular, I agree that IRV can have a certain kind of plurality-like
affect in regards to eliminating the bottom candidate, when Condorcet might
actually keep such a candidate in.

I reply:

That's been amply discussed here. Your designation "bottom candidate"
for the candidate who has the fewest 1st choice positions suggests
that 1st choice vote totals is your standard. Isn't the candidate
with the most 1st choice vote positions the "top candidate"? Then
isn't that whom you'd want to elect?

Randy continues:

However, in a country that currently
accepts plurailty and a President who got less than 50% of the votes cast, I
doubt that's really a big problem.

I reply:

We accept that so well that half of us refuse to vote.

Randy continues:

Your consideration of what happens if a 3rd party actually gets strong
enough to win is an interesting and valuable inquiry:
[...]


So far so good.  You go on:


     But what if the
     Republican is eliminated before the Libertarian? Unless all the
     Republican votes transfer to
     the Libertarian, which is extremely unlikely, the Democrat might
     then beat the Libertarian.
     If so, I will have helped the Democrat win by not strategically
     ranking the Republican first.
     But that's the same situation I'm in now if I vote my true
     preference for the Libertarian!


No. The Republican lost.  That's all.


I reply:

Sure, just like if you vote for Nader, the Democrat might lose to
the Republican, that's all. That may be all, but many consider it
undesirable enough so that they dump their favorite in order to
support a disliked 2nd choice.

Randy continues:

If the people who vote Republican
don't cast 2nd rank votes for Libertarian, then they don't want a
Libertarian.  By inference they either don't care, or might prefer a
Democrat to a Libertarian.

I reply. Sure. So what? The Republicans & Democrats are so similar
that it would be surprising if they didn't recommend eachother as 2nd
choice.

Randy continues:

The Republican lost.    If those Libs. don't get
enough votes transfered in to win, the Democrat wins, and that is a correct
result.

I reply:

Certainly, in terms of IRV's rules. Sure, you IRVies tend to use
your method as the standard by which you justify your method.

Randy continues:

I agree that you could consider that you must strategicly vote for
the Republicans, lest they not make it to the final round.  But I would
argue still that it not your fault they lost, but the fault of those who
ranked Republicans #1 to not put a #2 vote to the Libertarians, which they
have the power to do.

I reply:

Why should the Republicans be at fault for not ranking the Libs 2nd
if they don't consider the Libs their 2nd choice?? It's true that
if Compromise loses to Worst in Plurality or IRV, we can blame that
on the people who voted for Worst--they made Worst win. But that
reassurance doesn't keep people from abandoning their favorite to
help a lesser-evil. You say it isn't their fault if their greater evil
wins, but that doesn't relieve their perceived need to insincerely
vote a lesser-evil in 1st place.

Randy continues:

You assume, however, that Republica would not vote Libertarian as their 2nd
choice, and there are two problems I have with that.  I say first of all
that if they don't, it is their right and a conscious choice.

I reply:

Why do you have a problem with the fact that the Republicans have
a right to not vote the Libertarian in 2nd choice? You should have
a problem with a system that _didn't_ allow people to vote as they
want to.

Randy continues:

Secondly, you
are applying old thinking to an environment where a 3rd party is actually
competetive with the major parties.  If a 3rd party is really that
competetive, then everybody's going to know it, and it will make that #2
rank and #3 rank etc. all the more important to mark.

I reply:

Why is it a problem to you that people will consider it important to
mark their 2nd & 3rd choices?

Randy continues:

(As an aside, there
is no reason to believe that a conservative Republican would vote #2 for a
Libertarian

I reply:

Thank you, Randy, we knew that. If they all did, then you'd have
the fortuitous "mutual majority" situation, and IRV wouldn't embarrass
you.

Randy continues:

You refer to Australia as still having a 2 party system, and in the same
paragraph suggest that IRV will "wreak havoc" on our political landscape.

I reply:

Are you suggesting that thoses statements are mutually contradictory?

People who talk about IRV's "track record" should know that IRV,
like Plurality, has maintained a 2-party system. And it was initially
hoped that IRV would encourage parties to run more than 1 candidate.
No such luck.

Probably the havoc referred to was a disastrous election of an
extremist, something that will of course often happen with IRV,
under sincere voting. When that happens, people will know to not
repeat that sincere voting, and the 2-party system will be as safe
as it is now.


You go on later:

     In other words, IRV can have either of two completely opposite
     effects, depending on whether
     a third party is truly competitive or not. Before a third party is
     competitive, the effect of IRV is
     equivalent to a plurality system in which all supporters of minor
     parties are somehow
     convinced to abandon their principles and vote for the ``lesser of
     two evils.''

What are you talking about here?   Are you suggesting that voting for, eg.
Nader#1, Gore#2, is a lessor of two evils vote because I'm voting for Gore
over Bush in my #2 vote?

I reply:

I doubt that that's what Russ meant. The lesser-of-2-evils voting
to which Russ was referring is to vote Gore in 1st place so that
Nader won't eliminate him and then lose to Bush. That's a risk
when Nader is big enough to threaten to eliminate Gore, but not
big enough to be a sure winner against Bush.

Randy continues:

That's exactly what I want to be able to do.  No,
I suspect you are assuming you've made your case about the #1 rank somehow
not coming through, and the #2 rank doesn't either.  It's very simple: your
#1 lost and your #2 lost.  We'll live with it.

I reply:

And most of us will vote #2 in 1st place to avoid it under those
conditions, just as nearly all Nader-preferers did in 2000.

Randy's relentless ramble continues:

You go on:

     IRV has other serious problems too, which are explained in more
     detail elsewhere at the
     website. It is an erratic voting system because ranking a
     candidate higher can actually cause
     the candidate to lose, and ranking a candidate lower can cause the
     candidate to win. As if that
     weren't bad enough, it can also fail to elect a candidate who is
     preferred over each of the other
     candidates by a majority of the voters. It is also much more
     difficult to implement with
     security and integrity because the votes cannot be summed as in
     most other election methods.

You have given no evidence to argue that the higher ranked candidate will
lose unfairly, or that a lower ranked candidate will win unfairly.


I reply:

Randy, IRV's nonmonotonicity is common knowledge to everyone who
discusses voting systems. You can easily find examples in voting
system literature. Or maybe someone here might look up an example
and post it, or write an example. I believe it was Brams who wrote
an example in which moving someone from 1st place to last place
caused him to win, where he'd have otherwise lost. I'm sorry, Randy,
but that isn't acceptable.

No doubt you can find IRV nonmonotonicity examples in the EM
archives.

Randy coninues:

It is no
more difficult than regular elections to turn in counts that can be summed,
and I have demonstrated that in other postings to this mailing list, and
will gladly and easily demonstrate it to any who ask.  Briefly, all you have
to do is turn in counts for each unique set of rankings.  Eg. there will be
a certain number of L-D-R, L-R-D, R-L-D votes, etc.  So the actual number of
counts turned in might go up a lot, but are certainly manageable.

I reply:

Not really, Randy. The number of possible complete rankings varies
factorially with the number of candidates. And that's not even counting
the number of possible incomplete rankings.

Randy continues:

I consider IRVs similarities to the current plurality system an advantage --
in the people understand it and will accept it.

I reply:

IRVies endlessly repeat their unsupported claim that IRV is all
that people will understand & accept. I admit that IRV shares
Plurality's problems, notably its lesser-of-2-evils problem, due
to giving people only 1 vote to put on 1 candidate. But IRV is
a completely new voting system that Randy & his Irvie friends are
asking people to accept.

In comparison, Approval is nothing but Plurality done right. Plurality
with the added voter freedom that it should have had all along.
Approval is so similar to Plurality that it doesn't even require
new ballots. The only change needed is: Where the ballot now says
"Vote for 1", it would instead say "Vote for 1 or more". Two new
words printed on the ballot. Cost of changing to Approval? Zero.

I want to acknowlede that CR, aks Range Voting, is strategically
equivalent to Approval, and is already familiar to people, and
seems less subject to misunderstanding. Maybe start with CR, and
then point out that we get the same benefit more easily with Approval.

Randy continues:

It also does not have to be
the final solution, but will be an improvement over what we already have.

I reply:

Sure, if you want to make someone win by moving him from 1st choice
to last choice, and if you don't want to get rid of the
lesser-of-2-evils problem.

Mike Ossipoff




_________________________________________________________________
Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp