Multi-Member Congressional Districts

[muon2]

I think the thread is fine here, and it's interesting that you picked IL as an example. Up through 1980 IL had multimember districts for the state House. The elections involved modified cumulative voting. Each voter had three votes and could cast them for three candidates with 1 vote each, two candidates with 1.5 votes each, or one candidate with 3 votes.

[muon2]

One part of your plan that would be difficult is the goal of creating districts that are self-similar to the state. In IL there is a sharp difference between Chicago, its suburbs, and Downstate. To create self-similar districts would entail a number of pie slices out of Chicago. If there were 6 such slices (3 reps each) they would look pretty ugly, and it's likely they would violate the VRA by cracking the minority populations. To guarantee that the blacks and Latinos could each elect the candidates of their choice they would have to be concentrated in certain districts, which would skew the other districts.

[muon2]

Yeah, that's the sketchiest and least thought-out part of it, to be sure. But my vague notion was that something like an STV system would change the dynamics of what it would take to achieve minority representation in accordance with the VRA.

I don't think that STV works here. Consider that the black population is about 15% of IL, and they are sufficiently concentrated to get 3 of 18 seats when the state is gerrymandered, and certainly should have the opportunity to elect at least 2 of 18 reps. Cook plus a couple of adjacent counties is 9 reps (eg. Lake and Kane) and it would be easy to divide that into three districts of three reps each with one at 51% BVAP and one at 36% HVAP. Then an STV could work in those districts, but a district with 3 reps and 15% BVAP would be unlikely to elect a minority candidate of choice with STV.