US House Redistricting: Connecticut

[muon2]

Here's my version for the GOP.

No town is split and the deviations are within 0.2% (-545, -226, 1269, 320, -816).

[muon2]

Putting Bridgeport in with New Haven would give a non white a chance at winning a New England Congressional district.

The district is still 62.3% WVAP. BVAP is 16.3% and HVAP is 15.7%. The Hartford district is actually less white (barely). It has 62.2% WVAP, 14.8% BVAP, and 16.9% HVAP.

[muon2]

Quite natural behavior on Republican's part. They have 0-5 plan now, and they are basically offered to prolonge the same plan for next 10 years. Naturally - they refuse. What can they lose? Nothing, it can't get worse then 0-5. So they "gamble" hoping that Court plan (especially if Court is "activist" enough and will try to create "minority-heavy" district) will give them at least 1 winnable district. All is very natural. Democrats would do the same if situation would be reversed.

Minority heavy congressional districts are hard to produce if town lines are generally followed. The minority populations aren't as dense as in other states and the towns they are in are separated by towns with low minority populations. Using whole towns a string that connects New Haven, Waterbury, Meriden, New Brittain, and Hartford is still just over 50% WVAP, and no court would find the justification to do that.

I think my map is about as far as anyone could push on the grounds of creating minority opportunities. Waterbury is the only town with significant minority population, but connecting it to either the Hartford or New Haven districts creates an ugly map with district 5 stretching to the Sound or wrapping east over the top of Hartford.
 

Putting Bridgeport in with New Haven would give a non white a chance at winning a New England Congressional district.

The district is still 62.3% WVAP. BVAP is 16.3% and HVAP is 15.7%. The Hartford district is actually less white (barely). It has 62.2% WVAP, 14.8% BVAP, and 16.9% HVAP.

[muon2]

So what would a fair court draw, that is the question. Then we can measure what they do draw against that. What really interests me is just hard it is for both sides of the ledger to agree on what is a fair map, even if presumably acting in good faith. That has been my experience on this very site in fact, which is kind of sobering. If we can't do it here, where can it be done?

I'll offer my fair solution, though I'm sure there will be objectors.

The goals for my fair map are as follows.
Maintain the core area of the current districts.
Have no more than one town split between any two districts.
Improve compactness in the boundary between CD 1 and 5.

The following map does the above, with shading to show the areas changed.
As drawn the maximum deviation is 49 persons, and exact equality would only require shifts within the existing split towns. The shift would allow the small tentacle of CD 3 in Southbury to be smoothed.

[muon2]

So what would a fair court draw, that is the question. Then we can measure what they do draw against that. What really interests me is just hard it is for both sides of the ledger to agree on what is a fair map, even if presumably acting in good faith. That has been my experience on this very site in fact, which is kind of sobering. If we can't do it here, where can it be done?

I'll offer my fair solution, though I'm sure there will be objectors.

The goals for my fair map are as follows.
Maintain the core area of the current districts.
Have no more than one town split between any two districts.
Improve compactness in the boundary between CD 1 and 5.

I agree that you have the right approach, although existing lines to the extent a previous gerrymander should be ignored. You have one little spike there of CN-03 towards NE towards Hartford impinging a new county, that was there before, true. But it will need to be justified on its merits. A state court is not constrained by the "least change" rule of federal courts, when stepping in to draw a map a dysfunctional state legislature, or commission, could not draw. To me, shape and such are more important than avoiding one municipal split. As to counties in Connecticut, do they really matter?  Is there such a thing as county government?

The projection to the NE from CD-3 makes more sense geographically than one to the NW. That runs along the I-91 corridor. In fact if the judges were less partisan, they would move Meriden into CD-3 and Naugatuck into CD-5. However, I was concerned that a panel of Dem judges might not want to take the two big Dem nodes of New Brittain and Meriden both out of CD-5. But I can draw it anyway, and here it is with a maximum deviation of 75 persons.

Counties are meaningless.

Indeed counties have no governmental purpose in Connecticut any more. They are used for some state judicial districts and electoral tabulation. If for some reason they were used for redistricting purposes, then a map that minimized county splits and preserved towns would look like the following. The maximum deviation is 0.3% and it would be easy to pick four towns to split to achieve exact equality.