US with Indian constituencies

[traininthedistance]

Bama.



Max deviation 518.  Metro areas are almost kept whole, except for Bibb County which is part of the Birmingham metro but is in District 1 anyway.

District 1: SOUTH- is 41.6% Obama, 44.9% Dem, and 31% Black VAP.  Mobile, Montgomery, Tuscaloosa, the Black Belt, the Wiregrass, etc.  Safe R.
District 2: NORTH- is 36.1% O, 43.7% Dem, and 19% Black VAP.  Birmingham, Huntsville, the Tennessee Valley and Appalachian parts.  Even safer R.

A black-plurality district isn't really possible, alas, even with the Birmingham cutout.  Playing around with it, I could push the southern district up to 37-38%, and roughly even on the partisan numbers, pretty easily (as shown below), then it gets hard.  I'd expect a black district to be possible once you get to three, and necessary at four, but at two districts it just ain't going to happen. 

[traininthedistance]

Jokelahoma.



Deviation 1756.  No metros split.

District 1: TULSA-EAST is 34.6% Obama, 48.6% Dem (har har har), 11% Native, and supa-safe R.
District 2: OKC-WEST is 34.1% O, 45.5% D (yeah right), and even safer R.

Whee.

Workin' on California, which is taking a while on account of CA being too big for DRA to really handle, bah.

[traininthedistance]

Kentucky. 



Deviations 476.  Metro areas and CSAs are not hewn to exactly; the southern bit of the Louisville-Elizabethtown CSA is split off, as well as one of the micropolitan areas that gets appended to Lexington, but I wanted compact lines with low deviation that kept 1 within the Bluegrass area, and KY is relatively rural, so that sort of natural geography split seemed better.

District 1- BLUEGRASS- is 45.8% Obama.  It's clearly the more urban of the districts, with Louisville, Lexington, and the Cincy burbs.  Despite said Cincy burbs, it's a good deal more Democratic, and this district is probably friendly enough to a Blue Dog that it's merely Lean R.
District 2- SOUTH AND EAST- is 35.9% Obama, and Safe R.

[traininthedistance]

Cali.  Blurgh.



Max deviation 2506 on account of Districts 1 and 2, which are underpopulated and whole-county.  And, incredibly, I haven't found any other sensible whole-county groups- with the exception that 3/4/5 could all be around 6K overpopulated if I gave it the rest of San Benito, but whatever.  I can't say I'm really satisfied with this, but I also can't say I'm going to struggle with DRA busting at its seams any more.

There are a lot of min-maj districts, and five of them have Hispanic pluralities, but only actually one has a Hispanic VAP majority.  VRA gerrymandering is obviously possible here to up that number, and if this wasn't goddamn California I'd do it as an alternate.



District 1: NORTHERN.  O 49.3%, D 46.6%.  75W/13H by VAP.  So, rural northern CA, the suburbs north and east of Sacramento, and some Sierra Nevada counties to the south of that.  Once can, in fact, make a district that keeps the Sacramento CSA whole, and does so in whole counties, but that would basically force this district into San Francisco, which I am going to say is worse than splitting the Sacramento area.  Lean R.

District 2: SACRAMENTO-STOCKTON-MID-CENTRAL VALLEY.  O 58.8%, D 56.7%.  45W/31H by VAP (40/35 by total population).  Basically, what it says here.  One of those districts that might need a VRA finagling; one could probably make a Hispanic-majority district between this and 7 (and would probably want to change some lines with 1 too).  Safe D.

District 3: SAN FRAN-BAY AREA PACIFIC. O 79.3%, D 73.3%.  55W/19H/21A.  Sonoma, Marin, San Fran, Santa Cruz, almost all of San Mateo. The Bay Area is almost perfectly three districts, and West Bay/East Bay/South Bay is the obvious thing to do.  You might want to put Napa here instead of Santa Cruz, but SC doesn't really fit with 5 either.  Safe D.

District 4: OAKLAND-EAST BAY AND NORTH.  O 73.5%, D 68.8%.  49W/12B/20H/18A.  Napa, Solano, Contra Costa, most of Alameda.  Safe D.

District 5: SAN JOSE- SOUTH BAY.  O 72.1%, D 65.4%.  35W/25H/33A.  San Jose, most of San Benito, parts of San Mateo and Alameda.  The decision to put Santa Cruz in 3 instead of here, while forcing this district further into the East Bay, allows for the wonderful and unique possibility of a district that is plurality-Asian by total population (32.4% to 31.8%).  Safe D.

District 6: CENTRAL COAST. O 58.0%, D 50.5%.  52W/35H.  Also includes the Antelope Valley for lack of a better place to put it (and a desire to not split the San Fernando Valley, or have this district encroach into the Central Valley).  Monterey, Santa Barbara, Ventura/Oxnard, all that jazz.  Lean D.

District 7: CENTRAL VALLEY SOUTH. O 45.9%, D 42.3%.  40W/47H.  Fresno, Bakersfield, farms.  Kern is the only split.  Yes, Safe R and Hispanic plurality (majority by total population, even).  CVAP is probably another story.



District 8: INLAND EMPIRE. O 57.3%, D 52.6%.  33W/48H.  Again, majority-Hispanic by total population.  Splitting both San Bernardino and Riverside with 14, and also not reaching to break the 50 percent Hispanic barrier, are both things that I'd consider justified due to CA's mountainous geography- you want to leave those peaks as barriers as much as possible.  I tried to hew to town lines, but the precincts don't always cooperate here. Lean D.

District 9: SAN FERNANDO VALLEY-HOLLYWOOD. O 69.5%, D 63.6%,  50W/32H/12A.  Again, going with natural geography here in the LA area.  First of three districts entirely within LA; only Whittier and LA are split within the county at least.  Safe D.

District 10: LONG BEACH-SOUTH BAY.  O 74.4%, D 69.4%.  31W/19B/35H/13A.  Hispanic-plurality again; this district is CA's best chance at electing a black representative.  Safe D.

District 11: CENTRAL LA-GATEWAY. O 76.8%, D 76.5% 13W/68H/12A.  Going with natural regions kind of makes this a Hispanic pack to rival South Texas, you could easily swap land with 12 at the very least to push that from plurality to majority as well.  Safe D, in the running with 3 for safest depending on which numbers you prefer.

District 12: SAN GABRIEL.  O 60.5%, D 55.8%.  28W/41H/27A.  The rest of LA County, a bit of San Bernardino, and whatever OC that 13 can't take.  Another heavily-Asian district, another Hispanic-plurality.  Safe D.

District 13: THE O.C.  O 48.2%, D 38.7%.  51W/27H/18A.  Only 47% white by total population.  Safe R, of course.

District 14: MOJAVE-COACHELLA-IMPERIAL-OCEANSIDE.  O 46.2%, D 40.8%.  55W/34H.  The desert leftovers, basically.  Death Valley, Coachella, El Centro, some San Diego outskirts, a bit of Kern for road and population.  Safe R.



District 15: SAN DIEGO AREA.  O 57.7%, D 49.5%.  50W/29H/13A.  Another one of those districts that is min-maj by total population.  As the only one to show a split between the Obama and the 2010 Governor numbers (which is the D average here)… erm, let's say Tossup.  Probably tilts D by a couple points, though.

[Хahar 🤔]

The Salinas-Palmdale district is unfortunate, but beyond that it's a very good map. I was playing around with this earlier (without making a full map because, you know, California) and I opted to split the far north along the Coast Ranges, since coastal and inland areas there are very different culturally (and politically) and are also not well-connected by road. The result was a district stretching from the Oregon border down to the Bay Area; I had no problem with this, since Arcata and San Francisco instinctively feel like they could be in the same district, but others might. The other northern district would take in some more of the Sacramento area and probably still remain Lean R, since the areas it took in couldn't be more D than the northern coast.

[muon2]

The Salinas-Palmdale district is unfortunate, but beyond that it's a very good map. I was playing around with this earlier (without making a full map because, you know, California) and I opted to split the far north along the Coast Ranges, since coastal and inland areas there are very different culturally (and politically) and are also not well-connected by road. The result was a district stretching from the Oregon border down to the Bay Area; I had no problem with this, since Arcata and San Francisco instinctively feel like they could be in the same district, but others might. The other northern district would take in some more of the Sacramento area and probably still remain Lean R, since the areas it took in couldn't be more D than the northern coast.

I looked at CA with 15 districts, and decided to see how well one could stick to whole counties using a maximum 5% deviation as would be applicable from SCOTUS rulings for state legislative seats. Other than SoCal things fit remarkably well with a couple of exceptions. I haven't had a chance to fill in the SoCal detail with DRA yet. but here's my conceptual plan with the population deviations in parentheses.



District 1 - Sacramento Valley (+1.5%): The northern part of the state east of the Coast Range. One can't fit all of the Sacto metro in a northern district, and Sacto county was too much to move so I moved the eastern burbs out.

District 2 - Golden Gate (-3.1%): This works out much like Xahar describes. SF city plus the North Coast fit quite well together.

District 3 - East Bay (+3.0%): Alameda and Contra Costa together fit perfectly within the 5% limit.

District 4 - Silicon Valley (+0.7%): Santa Clara and San Mateo fit even better together.

District 5 - Sierra Nevada (-0.1%): The other districts in central CA and the Valley leave the Sacto burbs and the northern SJ valley about 563K short of a district. This is almost exactly the population in and beyond the San Bernardino mountains in that county. So to make the best of the situation the high Sierras ad Mono Valley link Roseville to Victorville.

District 6 - San Joaquin Valley (+1.3%): The southern end of the Valley from Fresno to Bakersfield is a nice fit for a district. It's majority Latino, but only plurality HVAP.

District 7 - Central Coast (+0.9%): From Santa Cruz to Ventura is about 234K people short of a district. LA county is a natural stopping point as you'll see shortly. One of the better highways across the Coast Range in this stretch connects Merced to Salinas and Monterrey, albeit through an uninhabited corner of Santa Clara county. Since Merced has 255K it's the best choice that doesn't crack the clean districts to the north and south of this one.

Districts 8-11 - Los Angeles County (4 districts, -4.7%): With only an average of -1.2% for each district there no good reason to break up LAC. District 9 is 59.3% HVAP, district 10 is 58.4% HVAP, and district 11 is plurality white, but with only 33.8% WVAP.



Districts 12-15 - SoCal (4 districts, +0.4%): Leaving SanB beyond the mountains to district 5 gives a nice fit for four districts. Riverside and Imperial make one district within 5% and an HVAP of 42.6%. SanB linked to Santa Ana is the second and is 50.0% HVAP. The HVAP could be increased by swapping areas in San B county for areas in Riverside, but with the whole county match for district 15, I decided to leave it at 50%. That leaves a district for the rest of OC plus NW SD county, and then a final district entirely in SD county.



One additional edit - I assume that the addition of Merced to balance the population of the Central Coast district is generally unattractive. By splitting up the counties of the Silicon Valley and East Bay districts I can go to this map instead.

[muon2]

IL already has 5 districts created after the 1970 Census for the judicial appellate districts. At the time they were defined as Cook (3/7) and the rest of the state divided into equal parts (1/7 each).



The populations as a fraction of the state total in 2010 were:
JD 1 - 40.5%
JD 2 - 24.9%
JD 3 - 14.1%
JD 4 - 10.3%
JD 5 - 10.3%

To make this into equal constituencies it would only require that JD 5 be merged into JD 4 to form a district with 20.6% of the state's population. The city of Chicago is 21.0% of the population and can be split from Cook leaving the rest with 19.5% of the state's population. Keeping DeKalb, DuPage, Kane, Lake, and McHenry from JD 2 is 19.9% of the state's population. The rest of JD 2 would combine with JD 3 to form a district with 19.1% of the state's population.

[minionofmidas]

That sounds more like an Indian way of drawing constituencies.

[muon2]

The OH constituencies could also be based on their judicial districts that elect appellate judges. There are 12 districts with the fraction of state population shown.



JD 1 - 7.0%
JD 2 - 8.9%
JD 3 - 6.8%
JD 4 - 5.5%
JD 5 - 12.9%
JD 6 - 7.7 %
JD 7 - 4.9%
JD 8 - 11.1%
JD 9 - 9.8%
JD 10 - 10.1%
JD 11 - 6.9%
JD 12 - 8.5%

Together JD 1, 12, 2, 3, and 6 are 38.9% of the state. Putting 1 and 12 together plus Montgomery in a Cinci constituency leaves the remainder of 2 with 3 and 6 more than 5% short of the ideal. Shifting Madison and Fayette back to the Toledo district creates two constituencies at 19.5% and 19.4%

JD 8 and 9 together make a Cleveland constituency at 20.9% of the state population.

The remainder could be divided with JD 5 and 11 at 19.8% of the population and then JD 4, 7 and 10 would be at 20.5% of the population. However, that's quite ugly with a Delaware-Ashtabula constituency and another from Columbus to Youngstown. Probably the better split is to group JD 4 and 10 along with Delaware, Licking, Fairfield and Perry in a Columbus constituency at 20.1% of the population. The reminder of JD 5 along with JD 7 and 11 form a Youngstown constituency also at 20.1%.

[they don't love you like i love you]

What would be the apportionment of these with the new Census numbers?

[President Punxsutawney Phil]

What would be the apportionment of these with the new Census numbers?

Eyeballing it (assuming 2.5 million is the quota):
+1 VA, +1 NJ, +1 NC, +1 GA, +1 FL, +2 TX, +1 CA

[they don't love you like i love you]

The OH constituencies could also be based on their judicial districts that elect appellate judges. There are 12 districts with the fraction of state population shown.



JD 1 - 7.0%
JD 2 - 8.9%
JD 3 - 6.8%
JD 4 - 5.5%
JD 5 - 12.9%
JD 6 - 7.7 %
JD 7 - 4.9%
JD 8 - 11.1%
JD 9 - 9.8%
JD 10 - 10.1%
JD 11 - 6.9%
JD 12 - 8.5%

Together JD 1, 12, 2, 3, and 6 are 38.9% of the state. Putting 1 and 12 together plus Montgomery in a Cinci constituency leaves the remainder of 2 with 3 and 6 more than 5% short of the ideal. Shifting Madison and Fayette back to the Toledo district creates two constituencies at 19.5% and 19.4%

JD 8 and 9 together make a Cleveland constituency at 20.9% of the state population.

The remainder could be divided with JD 5 and 11 at 19.8% of the population and then JD 4, 7 and 10 would be at 20.5% of the population. However, that's quite ugly with a Delaware-Ashtabula constituency and another from Columbus to Youngstown. Probably the better split is to group JD 4 and 10 along with Delaware, Licking, Fairfield and Perry in a Columbus constituency at 20.1% of the population. The reminder of JD 5 along with JD 7 and 11 form a Youngstown constituency also at 20.1%.

What the hey. I tried this with Minnesota with two different maps.

Judicial districts:


First one actually splits Minneapolis and St. Paul:


The blue district combines the 1st, 3rd, 4th, and 5th judicial circuits, the green is obviously the 2nd, 6th, 7th, 8th, 9th, and 10th.

This is actually a much cleaner looking map than you'd expect from such a thing. The blue is obviously effectively a Hennepin County-anchored district that is Safe D, Biden 57.46 to Trump 40.2. The green meanwhile actually voted for Trump 50.59-47.12 so it'd be a nasty district probably at least Lean R where the DFL tries to get through moderate candidates from outstate against St. Paul liberals and probably fails, while the Republicans quarrel over relative moderates vs. far right candidates. Who represents the blue district is an interesting question though, obviously Ilhan Omar isn't winning a primary there and we've already seen how Dean Phillips would do...Angie Craig lives there and is a good fit though.

Next map is an attempt at a more metro vs. outstate thing:

Blue combines 2nd, 4th, and 10th. The green combines 1st, 3rd, 5th, 6th, 7th, 8th, and 9th.

Blue is Biden 61-Trump 36.63. Green is 54 Trump-43.63 Biden.

Obviously blue is Safe D, although another Ilhan Omar isn't winning the primary situation. So either Dean Phillips unless he also makes a fool of himself or Betty McCollum. Green is of course Safe R with lots of candidates.

[President Punxsutawney Phil]

Virginia:
King George+Stafford+Fauquier+ Loudoun+Fredericksburg+PW County and everything inside it+Fairfax County and everything inside it
2,834,981
67% Biden 60% McAulliffe

Everything west of Caroline, Hanover, Goochland, Powhatan, Amelia, Dinwiddie, and Greensville counties
2,982,562
57% Biden 52% McAulliffe

The rest of the the Commonwealth of Virginia
2,813,850
60% Trump 67% Youngkin

[President Punxsutawney Phil]

NJ:
Bergen, Hudson, Essex
2,545,508
67% Biden 63% Murphy

Union, Middlesex, Monmouth
2,083,525
58% Biden 51% Murphy

Morris, Somerset, Mercer, Hunterdon, Warren, Sussex, Passaic
2,148,168
55% Biden 48% Murphy

Burlington, Ocean, Camden, Atlantic, Gloucester, Cape May, Salem, Cumberland
2,505,815
51% Biden 45% Murphy