3 ms·
Use a computer to optimise for getting as close to the same number of people in each district with the shortest borders (probably roughly hexagonal districts).
by erpellan 9y ago
Use a computer to optimise for getting as close to the same number of people in each district with the shortest borders (probably roughly hexagonal districts). Carving up the populace on any other basis than 'number of voters' is sheer sophistry. No amount of weasel words about 'communities' will change that.
- vidarh 9y agoThe much simpler solution is proportional representation. It takes away the benefit of gerrymandering, while designing a computer solution for this just changes the arguments people must use to justify changes to the algorithms employed.
- int_19h 9y agoThe problem with this is that it defeats the purpose of having per-district representation in the first place, which is supposed to be about shared geographic interests. If your randomly cut districts end up including parts with vastly different interests, one of them ends up denying representation to the other. You can say that it averages out across the country, but again, the whole point of such districts is to have representation that is unique to their needs. If we're okay with averaging things out, we might as well just ditch the districts, and count party votes on the national level.