[EM] Methods

[Kevin Venzke]

Hi Matt,

Writing very quickly, apologies in advance.

--- En date de : Dim 16.10.11, matt welland <matt at kiatoa.com> a écrit :
> > Approval's weakness is that it has to decide where the
> main contest is
> > prior to the vote. If there are few good options (i.e.
> any pair of
> > frontrunners leaves a large percentage of voters
> approving neither) or 
> > too many good options (i.e. several likely candidates
> for sincere CW) a
> > rank method, with its "higher resolution," may be able
> to fish out a
> > better result.
> 
> Hmmm... It seems to me that both those scenarios actually
> say something
> useful and even possibly important about the election
> results that would
> be lost in a ranked election.

I'm not sure I would dispute that, but even if they say something
useful that doesn't mean it helps to pick the best winner.

> Assuming that a) decent information about the candidates
> has been
> available via news, web and debates and b) reasonable
> quality approval
> polls have been conducted prior to the election then:
> 
> In the case where there are too few good options then
> clearly the
> candidates do not represent a good cross section of the
> values and
> criteria considered important to the people or the people
> are are too
> diverse to be easily represented. This is not a problem
> that can be
> solved by an election system. 

True, but...

> All a ranked system would do
> is hide the
> issue and choose some candidate that clearly a large
> portion of the
> population would not be happy with.

Yes, but it has a better chance of being the sincere CW than if you
use Approval.

If people cannot rally around two candidates under Approval, there is
a serious danger of having an almost arbitrary outcome.

I believe this could happen not just if none of the candidates are
very good, but also if sincere preferences are cyclic so that the 
polls cannot stabilize anywhere.

> In the case where there are many good options then approval
> is exposing
> that fact. 

That might be true when you look at the results, sure, but it's not
clear anything can be learned from it or that the failure to elect the
actual CW on election day could have been avoided.

> It is true that this scenario makes strategic
> voting more
> important but since we are assuming that decent information
> and prior
> polling is available I think voters can apply a pretty
> simple strategy
> to decide if it is safe to not vote for the front runner
> they don't
> really like. Assuming a party or conservative/liberal
> philosophical
> split then if the candidate they do like is ahead of the
> leading
> candidate in the opposing camp then they can safely not
> vote for the
> front runner in their camp they don't like. Hard to explain
> but trivial
> once understood. 

Yes, if there's left/right/center and Right looks relatively weak,
the votes you end up with will most likely give right no exclusive
approvals. Right supporters, seeing the same polls as everyone else,
will add support for Center. Left supporters won't. The result is that
whatever odds Right had of being the sincere CW get translated into
Center wins.

> Again, I think it is very, very important to note that the
> ranked
> systems actually lose or hide information relative to
> approval in both
> these cases.

In the second case I wouldn't agree with that. But see lower as I don't
think we're talking about the same situation.

> Note that in the first case the results and impact of a
> ranked system
> are actually worse than the results of approval.

As far as the results of a single election, I wouldn't agree with that.
This is Approval's worst case scenario in my view. As far as the ongoing
effects of Approval, you may be right. I am not sure. Off-center
candidates are less likely to be viable under Approval, but this is
true even when on election day they would be the actual CW.

It might be something like a "do you want to win each battle" vs "do
you want to win the war" type of choice. I can see that it is possible
to argue that it could be better to fail to elect sincere CWs sometimes
if in the long run the candidates are superior.

> The
> political pressure
> to converge and appeal to a broad spectrum is greater under
> approval
> than the ranked systems. The evaluation of a voting system
> only makes
> sense in the context of all the other things going on in a
> society. The
> pressure on politicians to actually meet the needs of the
> people is a
> massively important factor and ranked systems appear to
> wash out some of
> that force which is a very bad thing IMHO.

Understood...

> In the second case a ranked system *might* select a "more
> preferred"
> candidate but if you have several candidates all getting
> 75% approval
> then really, do you (pragmatically speaking) care which one
> gets chosen?

Well, I'm not envisioning the scenario where everybody gets 75%
approval. It's the scenario where several candidates could be the 
sincere CW on election day. That's more like having two candidates with
majority+ approval.

> I think we'd all be thrilled to have that problem.

This might still be true though. It would be a nice surprise if it
happened that this many candidates were viable (at least on sincere
preferences).

Kevin Venzke