[Election-Methods] RE : Best electoral system under real circumstances

[Juho]

On Nov 20, 2007, at 2:30 , Diego Santos wrote:

> 2007/11/19, Kevin Venzke <stepjak at yahoo.fr>:
>
> I don't remember that it is possible for surplus transfers to go to
> different parties.
>
> According Brazilian law, parties of same coalition are counted as a  
> single party. After elections, is not rare these parties to  
> separate to opposite political sides.

This problem exists in the Finnish system too. The system slightly  
favours large parties. Especially in the smallest districts the only  
way for small parties to get seats is to form coalitions. The  
election method allows coalitions and sees these coalitions as single  
parties. The allocation of the seats within the coalition is based on  
which individuals get the highest number of votes. This breaks the  
proportionality by allocating the seats in a rather random fashion.  
Small parties try to guarantee the seat(s) to themselves by naming  
only a small number of candidates (e.g. 1) and thereby focusing all  
their votes to this individual.

One solution to solve this problem would be simply to count the votes  
hierarchically per party also within the coalition. The seats would  
thus be given proportionally to different parties within the  
coalition. This could work slightly against the smallest parties  
since if the coalition would get one seat it would more typically go  
to the largest member of the coalition.

In order to increase the level of proportionality from this the  
number of seats per party could be counted proportionally based on  
the votes that the party gets within the whole country. Seat  
allocation within the districts would be counted only after this.  
This kind of methods would lead to some "rounding errors" at the  
district level seat allocation. But it seems that people often  
(typically?) value absolute proportionality between parties highly.  
It is thus possible to guarantee very exact country wide  
proportionality between parties and between districts, and push the  
"rounding errors" to district level seat allocations. (I'll skip the  
algorithms here.)

One more rather simple technique to solve this problem is simply to  
ban the coalitions (this is under discussion in Finland). This change  
could be accompanied by increasing the size of the (smallest)  
districts in order to keep also the current smallest parties alive in  
those areas.

Juho


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.electorama.com/pipermail/election-methods-electorama.com/attachments/20071120/63bb91d8/attachment-0003.htm>