[EM] (no subject)

[Juho]

On Nov 1, 2009, at 5:59 PM, James Gilmour wrote:

> Anthony O'Neal  > Sent: Sunday, November 01, 2009 7:12 AM
>> I don't necessarily think that STV is better than an open party list
>> system.
>
> I think STV-PR is better than open-list party-list PR in three ways.
>
> Firstly, STV-PR can be used in all public elections, including those  
> that are non-partisan.

Yes. Non-partisan multi-winner elections are however rare in politics.  
They may be more common e.g. when electing only a small number of  
representatives within a small community.

> Secondly, STV-PR can deliver proportionality within individual  
> political parties, where most open-list party-list systems will not.

Yes. List based methods typically don't do this. Tree method (see  
short description at the end of this mail) should maybe be classified  
as a variant of the open list method. It provides party internal  
proportionality on topics that are included in the tree hierarchy.

> Thirdly, and rather more politically, STV-PR can shift the balance  
> of power away from the parties to the voters, IF the voters
> decide to make than happen.

In this question STV and open lists are quite close to each others  
while closed lists leave more power to the party internal decisions.  
(Also in closed lists parties should nominate candidates that voters  
want to elect, but this process is one step more indirect than in STV  
and open lists, and therefore leaves more power to the party internal  
decisions.)

I think there is a general need to empower the voters (that is the key  
idea of democracy) but that doesn't necessarily mean that we should  
take the power away from the parties. Having formal parties and other  
opinion groupings is a practical tool. Maybe parties should just be  
such that they reflect and drive topics that the voters want them to  
represent. (Non-working parties that have become tools of incumbent  
party officials and other limited interest groups could be classified  
as one form of corruption.) Party-less systems might be also one form  
of democracy but I'm not aware of any. Having parties simplifies  
(maybe oversimplifies) things a lot, but to some extent that may be  
necessary to make the politics understandable and easier to handle to  
the voters that are expected to rule (despite of not being experts in  
all the details). The ideal situation could thus be described also as  
empowering the voters, both directly and via parties.



>> STV is proportional if people vote by party.
>> It is also proportional if people vote by eye color.
>
> Yes, and need not be "either or"  -  it can be "both and".  The  
> voters can rank by party and then by eye colour.  Or the voters can
> rank by eye colour and then by party.  With STV-PR the voters are  
> free to base their rankings of the candidates on as many
> dimensions as each voter wishes.

Yes. STV allows any criteria and related proportionality while for  
example the tree method supports proportionality only on topics that  
have been described in the tree hierarchy. That means simplification  
(both good and bad). That also makes the decisions more explicit  
(good). If there are many representatives with green eyes that may  
lead to something, or may be forgotten. If there are subgroups that  
explicitly are named to drive "green eyed policy" then that might have  
an impact and the developments and the role of the representatives  
that were elected as "green eyed" could be monitored by the voters  
easily.


>> It's main problem is that it's complicated as hell to explain, and  
>> the
>> opposition at the BC-STV referendum exploited this mercilessly.
>
> Yes, a great deal can be made of this, and was by the opponents of  
> reform in BC, but it need not be so.
...

Most election methods are complicated enough so that regular voters  
are not able to describe them in detail. What the voters need is a  
rough understanding on how the method works and some trust that the  
system works as planned. The complexity and whatever properties of the  
methods may be used in campaigns that promote or oppose different  
methods but I think that has often more to do with the art of  
campaigning than with the real benefits and problems of the methods in  
question. STV should thus not be doomed for this reason (although this  
may cause some problems to it, and the vote counting process is a bit  
heavier than in some other methods).


Here are also some more possible comparison points between STV and  
open lists (and trees).

- Open lists are summable and therefore easy to count and verify  
locally. STV requires centralized counting or centralized control of  
the local counting process.
- Open lists have simple votes, which makes it less vulnerable to vote  
buying and coercion and maintains privacy better.
- SImple counting process also allows the results to be counted quickly.
- Open lists / party based systems (and subgroups in trees) tie  
candidates and representatives to some announced policy. STV allows  
each candidate and representative to form his/her own space. STV is  
richer and more flexible but that also opens more possibilities to  
escape from what the voters wanted them to represent and may keep the  
political understanding of the voters more obscure. (In principle STV  
candidates may tell to both green and red voter groups privately that  
they support them, and after the election the representatives may also  
not know why their voters supported them.) Lists and trees may force  
candidates and representatives to simplify the political space and  
their opinions (this may be more good than bad from some points of  
view).
- Open lists allow larger number of candidates (and parties and seats)  
than STV. There are no theoretical limits in STV but the size of the  
ballot and the need of the voters to rank (and evaluate the opinions  
of) numerous candidates set some practical limits. Large number of  
candidates offers the voters more power to make the choices (instead  
of leaving that to the party officials).
- Large number of candidates and seats also supports better  
proportionality / representation of smaller groups.

Juho


P.S. Here is a very short description of the tree method
- candidates are ordered as leaves in a tree
- the tree structure will be fixed before the election
- parties may contain subgroups and they may contain subsubgroups etc.  
(also party alliances are possible, and independent groupings)
- there may be additional rules that encourage a good tree structure  
to be formed (e.g. limitations on the number of candidates/subgroups  
that can form a flat group)
- voters vote for one individual candidate
- each branch of the tree will get seats in proportion to the number  
of votes of all candidates within that branch
- tree method can be seen as an extension of open lists, to support  
also party internal proportionality and to allow the voters to better  
steer the direction and evolution of the country / parties / political  
groupings