[EM] Making a Bad Thing Worse

[Greg Nisbet]

On Sun, Oct 19, 2008 at 6:30 AM, Raph Frank <raphfrk at gmail.com> wrote:
> On Sun, Oct 19, 2008 at 6:55 AM, Greg Nisbet <gregory.nisbet at gmail.com> wrote:
>> My thoughts on primaries were challenged. Let me explain:
>>
>> Primaries may be the rational response to FPTP. It doesn't matter.
>> Without Draconian sore loser, candidate oppression laws the parties
>> would have no way of stopping popular primary rejects from running.
>
> Well, it depends on how popular the candidate is.  There would be some
> candidates who can disregard primary results and some who can't.  It
> only works for very popular candidates.  A reasonable number of
> candidates wouldn't be able to pull it off.

Just because they can't pull it off won't stop them from trying. I
think the only reason candidate accept the results of primaries is
because they are forced to. The obstacles facing an independent
candidate are formidable. They aren't prevented from running for want
of trying.

>
> You have to convince all the supporters of the party that you are the
> one who is going to win and not the one with your (old) party's label.
>  You also have to convince them that setting aside the primary
> (democratic) result is acceptable and also deflect accusation that you
> will end up splitting the vote.
>
> If 80% of candidates have to accept primary results, then they serve
> some function.
>
>> At the point where they are
>> strictly voluntary vote pooling agreements, I argue they break so much
>> continuity with the current system as not to be regarded as the same
>> thing.
>
> You mean, it would be a completely different system?  I am not sure it
> would be that much different if they were voluntary ... except that
> they possibly wouldn't happen.

What I am saying is, without legal force, primaries would be very
different. I was trying to say that my earlier criticism of primaries
does not apply to this because it candidates are not coerced into
participating in the primary. They can run without participating at
all in the world absent the add-ons to FPTP.
>
>> The Electoral College:
>>
>> Asset voting as a single winner voting
>> method makes no sense.
>
> Sure it does, think of it like IRV but with intelligent vote
> transfers.  This helps solve some of the defects.

I said this because I don't see it accomplishing anything. First of
all the current system does not allow transfers, so that is pretty
much out of the question. Second I don't think it's going to give so
much power to people who weren't elected by name in the first place.
>
> I am not entirely in favour of asset voting in this case, but it isn't
> completely unreasonable.
>
> My problem is that there are conflicts of interest.  For example,
> let's say there are 3 candidates and the supporters have utilities of:
>
> 45: A(100)>B(70)>C(0)
> 10: B>A=C
> 45: C(100)>B(70)>A(0)
>
> B is the condorcet winner.
>
> Both A's and C's supporters would rather have B elected than a 50%
> chance of their favourite being elected (70 utility vs (50/50 chance
> between 100 and 0) ).
>
> However, since the electors are likely to be much more partisan, the
> makeup of the electors is likely to be something like
>
> 45: A(100)>B(10)>C(0)
> 10: B>A=C
> 45: C(100)>B(10)>A(0)
>
> In this instance, both A's and C's electors would be willing to hold
> out.  The end result is that B's supporters must pick one or other of
> them.
>
> One possible tactic for B's supporters would be to flick a coin in
> public and say that they will 100% support the winner of the coin
> toss, unless the loser agrees to support B.
>
> Ofc, that can be countered by A and C committing to their candidate
> publicly too.  Who would break first :p.
>

That is one of the many problems I have with asset voting. In the
single winner case its faults are more obvious.

I'm not sure that any modfication to asset voting is sufficient to
solve your problem. I think the faults that plague IRV plague Asset
Voting as well (albeit to a lesser extent because of the restrictions
placed on who you can vote for).

>> First of all, this violates unrestricted domain. Voters
>> should not have arbitrary limits placed on what they are able to vote
>> for.
>
> What limits?  Surely, the same applies to Congress, you are picking a
> group of people to act on your behalf.

That is correct. I was comparing single winner asset voting to a
competitor like IRV or some Condorcet method. It is definitely more
restrictive than either IRV or Condorcet.

Compared to FPTP, it is about as expressive except you haven't a clue
how your vote will transfer in the future.

>
>> The Senate:
>>
>> The United States' heritage as a federation has no impact whatsoever
>> on the legitimacy of bending the will of the people. See You Can't
>> Have it Both Ways.
>
> The States agreed to it and that only makes sense as part of a federation.
>
>> Two Parties:
>>
>> I think we pretty much agree that the Democrats and Republicans
>> actively prevent competition through silly laws and their perpetual
>> monopoly on power. I say monopoly because they are both relatively
>> centrist. At least that is the impression I get.
>
> There is an argument that this is a natural result of the 2 party
> system.  If a party keeps losing, it moves towards the centre to pick
> up more votes.  The other party then starts to lose, so it moves too.
> In the end, they are very close to each other.
>

It's still better than one party domination though. If the median
opinion shifts then the two main parties will scramble to change their
platform. In a one party system that probably won't happen as
frequently.