[EM] Census re-districting instead of PR for allocating seats to districts.

[Juho Laatu]

On 2.7.2012, at 1.05, Michael Ossipoff wrote:

> There's no reason why, STV, voting for and electing candidates, couldn't be used in each multimember district, with the parties afterwards topped-up according to a national at-large list-PR allocation.

STV ballots may rank candidates of multiple parties. Would the national party vote maybe go to the party of the first candidate, or would the candidate maybe indicate his favourite party separately?

You probably also assume that (most) candidates are associated with some country wide party. Many STV proponents like also the idea that candidates could be totally free of party connections. They could be so also in this model, but they would not get (easily) any top-up seats.

> STV systems usually use small districts, tending to defeat the purpose of PR.

One hybrid PR oriented approach would be to use STV within the parties. The voters would be able to rank candidates of one party only (a simplification to keep the method manageable). Seats would be allocated to the parties using some list style method. STV would be used to allocate seats within each party. The voter's could safely rank few candidates only since also an exhausted vote would go to the correct party. This approach would allow also large district sizes and large nuber of candidates (thanks to easy inheritance).

Juho