[EM] connection between multiwinner voting systems & districting problems

[Juho Laatu]

On 28.7.2011, at 16.50, Warren Smith wrote:

> I daresay this has been pointed out before, but I do not think much analysis has
> been done before, of this idea:
> 
> If you want to have V voters elect W winners, that can be considered as the
> same, or anyhow a highly related, problem as the problem of drawing W
> equipopulous districts on a map.
...

I just sent another mail with a proposed method that makes the proposals of my earlier mail on local representation and geographical proportionality more concrete. It has a somewhat unusual approach to the districting problem. It does not aim at forming fixed districts but rather creates them dynamically in each election for each party depending on the received votes. Political proportionality and geographic proportionality are calculated at the same time. There can be also other proportionality criteria (e.g. based on age groups, sex, ethnicity,...), but these two are of course the most common ones and present in most elections.

Juho