[EM] [Election-Methods] [english 94%] PRfavoringracialminorities

[Raph Frank]

On Sun, Aug 17, 2008 at 10:28 PM, Juho <juho4880 at yahoo.co.uk> wrote:
> On Aug 17, 2008, at 20:05 , Raph Frank wrote:
>> Voting the 'party ticket' in this context is just voting for all
>> candidates that your party puts forward before giving any
>> rankings to any other candidate.
>
> It makes sense to me to allow such generic names and not force the voters to
> list all the candidates of the party to be sure that the vote will be
> counted for the right party (assuming that the number of candidates is
> large).

In Ireland, it is rare that parties run more than 2+ candidates in a given
constituency and if then, only the 2 main parties.

>> The problem for parties is that the surplus doesn't remain within
>> the party and leads to a vote management strategy.  (If none
>> of their candidates have a large surplus, then they get to keep
>> most of the personal votes for any of their candidates).
>
> This is a very interesting real life example on how such "horizontal"
> preference orders may impact the elections and strategies in them.
>
> Do you have a list of the strategies/tricks that are used?

The main one is 'vote management'.  This is where you split up
the constituency and only allow certain candidates to campaign in
those areas.

A very popular candidate mightn't be allow campaign at all.  In
practice, this doesn't always work out.  For example, in Limerick
one of the FF candidates takes great pride in getting lots of first
choice votes.

Also, sometimes it might backfire and the very popular candidate
might fail to get elected as they don't campaign in any specific regions.

> (What I was thinking was basically that if there is one quota of voters that
> have opinion X then the representative body could have one representative
> that has opinion X. This could apply to parties but also to smaller
> groupings and individuals as well as other criteria like regions (=>
> regional proportionality) (and even representation of other orthogonal
> groups like women, age groups, religions, races if we want to make the
> system more complex).)

This is the party centric viewpoint.  PR-STV is more based on the
candidate-centric viewpoint.  You vote for someone because you think
they would make a good representative.

>> I think PR-STV at the national level where the voter votes for a list but
>> can override the initial votes is the best compromise between maximum
>> expression and reasonably low complexity.
>
> I mentioned earlier also the possibility to use trees here. A bullet vote to
> a candidate would by default be inherited by the (small) group that the
> candidate belongs to, then the party ('party ticket') and so on.

Right, however, a tree can kinda be considered a list.  The candidate would
rank themselves first and then all the members of their leaf and work back
along all the branches.

> Yes this adds complexity. But this would not be too far from the use of the
> 'party ticket'. Right?

Well, party ticket is just manually voting for all members of your current
party.  It is no more complex than a standard ranked ballot.

Ranked ballots are inherently more complex than just voting for a list/tree etc.

>> Another option is to allow a voter vote for local candidates and then
>> as their last choice, vote for a national list.
>
> This is maybe yet one step more complex since now candidates can belong to
> different orthogonal groupings (several local parties; one party covers all
> local regions). Or maybe you meant to allow voting only individuals locally,
> not to support all local candidates of all parties as a group.

Right, there would be a party list at the top/national level.

Basically, it would work like PR-STV just with a fixed quota (same in all
districts).

You would rank local candidates as per normal and then specify which
national list you want the remainder of your vote to go to.

If your first choice gets elected with 2 quotas, you would have half a vote
remaining to pass to the national list.

>> The local count would be standard PR-STV, but with the same quota
>> nationwide (and a rule that you must reach the quota to get elected).
>
> Ok. National level proportionality could influence the election of the last
> candidates in the districts.

There would be a mix of nationally elected and locally elected candidates.
If everyone in a consitutency just voted for a national list as first choice,
then that constituency would have no local candidate elected.

> Maybe all districts would be guaranteed their fixed number of seats
> (typically based on the number of citizens of each district). The extra
> seats would be first allocated to parties and then to districts (using some
> appropriate algorithm).

Right.  However, that is actually more complex than it sounds :).

A party might be entitled to one more seat, but the only remaining
candidate it has left is in a constituency that has already received its
max number of seats.

Also, constituencies don't need to be assigned integer numbers of seats.

> Due to the involved rounding errors I'm not sure that this style of sending
> the remaining votes to national level would make the results better (more
> proportional?) than just allocating the remaining vote fractions to the
> local candidates of the party ('local party ticket'?). (I'm however not sure
> that I even understood the intention correctly.)

Well, the point is that a party might get 1.5 quotas worth of votes.

This would mean that once of the party's candidates gets elected locally and
then 0.5 quotas gets passed to their national list.

This assumed that everyone voter votes for the party's local candidates and then
for the party's national list.

Anyway, the intention is to allow voters pick a local that they like
and only get to
transfer to the national lists whatever vote strength they have remaining.