US with Indian constituencies

[traininthedistance]

Joisey.

While the intra-state apportionment needs to be done on 2001 numbers, the seats per state should be based on the 1961 (or 71, I forget) census.

Now draw the districts. They should be built of segments (state assembly districts) but mildly gerrymandered anyways.

Whee, giant districts are easy-peasy.  Not bothering with using state assembly districts here tho since 3 doesn't divide into 40.



District 1- South (and part of Central, if you believe it exists) Jersey- is 56.4% Obama, 53.9% Dem.  Lean Dem.
District 2- Central and Northwest- is 52.4% Obama, and 49.4% Dem.  Tossup, maybe even a Pub tilt.
District 3- the urban Northeast- is 64.6% O, 63.6% Dem.  45% White VAP.  Safe D, of course.

District 1 is 111 folks overpopulated, and the others are even closer.

[traininthedistance]

Virginia, cause these districts are just that quick to draw.  



Compact and whole counties (max deviation is +904), but at the expense of separating Chesterfield from the rest of the Richmond metro.  (Which is a metro that must be split, given that NoVA and Hampton Roads are both larger, and in corners.)

Sensible min-maj districts are of course impossible at this size.  1 and 2 are both around 60 percent white, 'bout as good as you can get without heavy gerrymandering.

1: NoVA. 58.7% O, 53.8% D.  Lean D.
2: Hampton Roads-Tidewater-Richmond.  55.8% O, 51.2% D.  I guess still Lean D?  Closer to Tossup though.  30 percent black, not a maxpack but pretty much the best you can do with whole counties.
3: Western.  43.6% O, 42.7% D.  Safe R.

[traininthedistance]

Once you pop you can't stop. Colorado.



Deviations are plus and minus 1,851. 

1: Denver Metro. 57.8% Obama/51.7% Dem.  Entirely within the metro, some outlying and CSA area goes to 2.  Lean D.
2:  Rest o' the state.  49.6% O/45.0% D.  Lean R.

[traininthedistance]

Arizona.  For once, towns are split, because, hey, a Hispanic VAP plurality district is possible.  And given the craziness that is AZ, they deserve one.





Max deviation 102.

District 1: NON-MARICOPA.  Yeah.  44.4% O, 43.2% D.  In most states these numbers are Safe R, but of course a crazy vs. the doggiest of Blue Dogs could put it into play.
District 2: EL NORTE.  55.4% O, 55.5% D.  Hispanic-majoirty by total population, but merely plurality by VAP (44.1 to 42.7).  Safe(ish) D.
District 3: ARPAIOLAND. 39.7% O, 37.4% D.  Safe R for even the craziest.

[traininthedistance]

Tennessee.



Max deviation 463.  Obviously one starts with the Grand Divisions and then shifts the lines east, since the West needs people and the East has too many. 

1: East TN.  33.5% O, 39.2% D.  The most consistently R area of the nation, safe since Reconstruction.
2. Middle TN. 42.0% O, 47.6% D.  I assume that the ancestral Blue Dog tendencies in the rural parts are just gone, and this is probably Safe R going forward.  I mean, even Scott desJarlais got re-elected.
3. West TN. 49.9% O, 51.8% D.  West TN itself is still legitimately Dem-leaning, b/c of Memphis.  But you have to go pretty far into Middle TN.  30% Black VAP, Tossup.

[traininthedistance]

There are two basic ways to draw Maryland.  You can either split it into a Washington district and a Baltimore district, or you can try and pack blacks.



Deviations 901.

District 1: Washington Metro, South, and West.  68.2% O, 68.1% D.  Min-maj by total population but just barely white-majority by VAP.  Safe D of course.
District 2: Baltimore and Eastern Shore. 56.3% O, 57.7% D.  Lean D.

Or...

 

Deviations 250. 

District 1: South. 70.6% O, 70.5% D. Black-plurality- 44.7% VAP compared to 42.5% White VAP.  You can finesse things more if you get real ugly, but I don't think 50% is actually possible.
District 2. North.  54.0% O, 55.9% D.  Lean D, but getting closer to even. 

Obviously, I prefer the first option. 

[traininthedistance]

You knew this was coming.  New Yawk.





Max deviation 1,065.

District 1: LAWN GUYLAND.  O 53.0%, D 55.0%.  Oyster Bay town is split, something has to be in Nassau.  Lean D.

District 2: JAMAICA BAY.  O 87.6%, D 88.7%.  Before drawing this, I was thinking "there are probably only three places in the country where districts of this size can make compact black-majority districts- here in NYC, South Side Chicago, and the southern half of the Atlanta metro".  Now I'm convinced that only the Atlanta one has any chance of actually working.  This is 46.1% Black by VAP and will certainly elect an AA rep anyway (whites and Hispanics are about even at 21 percent each).  Black-majority is possible if you get real ugly... but you have to basically make a block-long snake up into Central Harlem, and I was not about to do that.  Rather keep this district strictly Brooklyn-Queens, and 4 out of Brooklyn.  Safe D obv.

District 3: NEW YORK HARBOR. O 69.4%, D 69.5%.  There are three all-NYC districts, each named after bodies of water for lack of a better idea, and this is obviously the "white" one.  Most of Manhattan, all of Staten Island, and the white parts of Brooklyn (minus Williamsburg/Greenpoint).  Sizable Hispanic and Asian minorities here, each about 15 percent.  Safe D.

District 4: HARLEM RIVER.  O 87.1%, D 87.8%. And this is the Hispanic district.  51.0% by VAP.  Also the smallest district in the nation, no doubt.  Safe D again.

District 5: LONG ISLAND SOUND-TAPPAN ZEE.  O 62.7%, D 64.1%.  By connecting the Nassau North Shore to (most of) Westchester/Rockland via north Queens and Bronx, we keep three all-NYC districts.  Very diverse: 52.8% VAP white, 13.1 black, 17.0 Hispanic, 15.5 Asian.  I decided I'd rather split both Westchester and Rockland than split a town.  Safe D.

District 6: HUDSON VALLEY-CATSKILLS-CAPITAL.  O 54.0%, D 55.5%.  As is says on the tin.  Schoharie County (part of 7) should be here since it's part of the Albany Metro, but hey, it keeps the 6-7 line to whole counties.  Lean D.

District 7: UPSTATE NORTH AND CENTRAL.  O 52.6% D 54.1%.  Lots o' area.  North Country, Syracuse, Rome/Utica, Binghamton, Ithaca, most of the Finger Lakes and Southern Tier, even encroaching into the Rochester fringe.  The most Republican district in the state, so let's say Tossup, though all of the non-NYC districts are really best described as competitive but with a Dem tilt.

District 8: WEST NY.  O 54.5%, D 52.1%.  Buffalo, Rochester, Holland Purchase, some Southern Tier.  The lower D average is largely due to the regional effect of the Carl Paladino candidacy, where he was uniquely strong here and uniquely dreadful everywhere else, so as with the other districts I'm more inclined to believe the Obama numbers.  Lean D.

[traininthedistance]

Georgia on my mind. 



Max deviation 999.

District 1: SOUTH GEORGIA.  O 46.4%, D 49.2%.  34.7% Black VAP; Georgia affords a pretty unique opportunity to have two districts with a substantial AA percentage, and it only seemed right to take it.  Lean R anyway, though of course there's an opening for a Blue Dog.
District 2: METRO ATLANTA SOUTH-MACON.  O 67.8%, D 66.3%.  54% Black VAP, without a doubt this is the most heavily-black constituency to be found in America.  Is it the only black-majority district to be made at all?  We'll see what happens in Illinois.  Cobb, Fulton, and DeKalb are all split to facilitate this, as well as the arm down to Macon.  Safe D.
District 3: METRO ATLANTA NORTH.  O 43.0%, D 39.1%.  The Atlanta metro extends out to ginormous exurban lengths, but this district has the core of it, or at least its non-black parts.  Probably trending D, but still Safe R for awhile.
District 4: NORTH GEORGIA. O 30.1%, D 34.7%.  And the rest.  Safe R.

[traininthedistance]

Lookie here, it's South Carolina in whole counties and a whopping deviation of fifteen!



District 1: LOWCOUNTRY.  O 52.6%, D 49.0%.  33.6% Black VAP.  I wasn't even trying for an Obama district, honest officer!  Tossup.
District 2: UPCOUNTRY.  O 37.1%, D 35.5%.  Safe R, duh.

Of course I don't see the Republicans going for this (not that they'd go for a lot of the other fair maps, but they might even be able to convince an independent panel on this one), and it does have the legitimate disadvantage of slicing right through the heart of the Columbia metro.  So here's an alternate:



Deviations 2,437.

Lowcountry is now 48.6% Obama/44.8% Dem, 30% black VAP, and let's say lean R.  Upcountry is 41.1% O/39.6% D, less black, and safe R still.  The Columbia metro is still technically split, but the central urbanized area is whole at least.

[traininthedistance]

Ha, an appropriate use of LSU colors!

There isn't any way to get a compact(ish) black-plurality district in LA with districts this size, is there?

[traininthedistance]

Flo Rida.



Max deviation 779.  The northern 4 districts are one group, the southern 4 another.  No cities are split.  The black population is too spread out to get a district here, but you can bet there's a Hispanic district.

District 1: NORTHWEST-PANHANDLE.  O 40.6%.  Pensacola, Tallahassee, Gulf Coast some areas north of Tampa Bay, etc.  The northernmost, and therefore, most "Southern" district here.  Safe R.
District 2: JACKSONVILLE-NORTH CENTRAL.  O 46.3%.  Jax, St. Augustine, Gainesville, Ocala, Daytona Beach, etc.  A lot closer (and 17% Black VAP FWIW, second-highest in the state), but still Safe R.
District 3: ORLANDO METRO.  O 53.6%.  All of the Orlando MSA, and entirely within the Orlando CSA.  Basically just takes all of Deltona from Volusia for its one county split.  Only 59% White VAP, with 13% Black and 22% Hispanic.  Tossup.
District 4: TAMPA BAY.  O 53.6%.  Tampa, St. Pete, and the most urban part of Pasco, entirely within the metro.  70% White, 11% Black, 15% Hispanic.  Another Tossup, fitting for the I-4 Corridor, the state's prime swing region.
District 5: LAKE-SUNCOAST.  O 46.0%.  Lakeland/Winter Haven, Sarasota, Cape Coral, Ft. Myers, Naples.  Safe R.
District 6: SPACE COAST-TREASURE COAST-PALM BEACH.  O 54.1%.  72% White, 12% Black, 13% Hispanic.  Down the developed coast.  Sort of on the border between Tossup and Lean D.
District 7: BROWARD-LAKE OKECHOBEE.  O 65.5%.  Only 45% White VAP, with 24% black and 26% Hispanic.  The decision to put the Florida Heartland here instead of in 6, and to split Palm Beach east-west rather than north-south, was out of a desire to get the Okechobee area in one district, and a realization that the inevitable Collier split fit better here, and also the fact that there is actually quite a large minority population inland in places like Belle Glade and Immokalee, so best to put them in this min-maj district.  Anyway, Safe D, the only one in the state.
District 8: MIAMI-DADE-KEYS.  O 55.5%.  Monroe and the vast majority of Miami-Dade.  67% Hispanic VAP!  (18% White and 13% Black.)  Of course, a lot of that is Cubans, and it will be beaten in South Texas anyway.  I wonder about possibly splitting 7 and 8 so that 7 takes Monroe and western Dade, while 8 stays entirely within the county; there are multiple possibilities.  Anyway, Lean D.

[traininthedistance]

Several alternate versions of South Florida are possible (I'm pretty satisfied with Districts 1 through 5).  Here's three of them.

First, you can keep 7 within the Miami metro, and give the rural parts to 6.  This is the Pub-favoring alternative.



6: SPACE COAST-TREASURE COAST-OKECHOBEE is now 51.8% Obama, 69W/12B/16H by VAP.  Still Tossup, but more Pub than before.
7: BROWARD-BOCA is 67.2% Obama, and still min-maj at 48W/23B/24H.  Super safe D.
8: MIAMI-DADE-KEYS takes in some empty land but remains basically unchanged, heavily-Hispanic and Lean D.

The Dems would prefer a change that kept 8 entirely within Miami-Dade instead:



6 is the same as the original, an arguably Dem-tilting Tossup.
7: BROWARD-OKECHOBEE-EVERGLADES is down to 63.2% Obama, and 48W/21B/26H. Still Safe D of course.
8: MIAMI-DADE is now 58.1% Obama, and is now 15W/16B/67H.  Still plenty Hispanic, and still Lean D, but a very strong lean now, closer to safe.


Or, you can keep Okechobee in 6 and split Palm Beach north-south, but also have a Miami-Dade district.  In this case it probably makes more sense for the remainder of Collier to go in 7 to unite the Everglades along I-75 and the Tamiami Trail.



6: SPACE COAST-TREASURE COAST-OKECHOBEE is 52.9% Obama, 70W/13B/15H.  Still a tossup.
7: BROWARD-BOCA-EVERGLADES-KEYS is now 64.1% Obama, 50W/20B/25H.  Still min-maj by total population, but no longer by VAP.  Safe D as always.
8: MIAMI-DADE remains 58.1% O, 15W/16B/67H.  Strong Lean D.

…

All four of these alternatives are reasonable.  I think I still like the original version best, though the last one has some appeal as well.

[traininthedistance]

Penn's Woods.



Max deviation 802.

District 1: PHILLY-DELCO-CHESTER.  O 72.3%, D 70.6%.  55% White VAP, 29% Black.  The Philly-Reading CSA plus the Lehigh Valley is pretty close to two districts, and is definitely the logical way to combine SEPA.  The most compact and county-conserving way to split them is north-south, so here you have it.  Safe D.
District 2: MONTCO-BUCKS-READING-LEHIGH.  O 56.4%, D 53.7%.  The rest of the Delaware Valley, etc.  Carbon is the best split (and it's a small split, less than 10K) on the grounds that it's part of the Allentown-Bethlehem metro, but it's also outlying and of a piece with NE Pennsylvania as well.  Lean D.
District 3: SOUTH CENTRAL.  O 42.3%, D 39.6%.  Harrisburg-York-Lancaster, Altoona, Johnstown, the southern bits of the T.  Safe R of course.
District 4: NORTHERN PA.  O 49.9%, D 48.7%.  Northeast PA, State College, Northern Tier, Erie.  Obama won this narrowly in 2008, certainly not in 2012, Lean R sounds right.
District 5: PITTSBURGH-SOUTHWEST.  O 50.7%, D 54.6%.  Romney probably narrowly won this in 2012, it's an interesting question whether the ancestral Dem strength here can bounce back or not.  Tossup.

[traininthedistance]

Penn's Woods.

How much would it change things if you moved Cambria and Somerset Counties into CD5?

As far as partisan impact, I would assume that 5 would become more Dem at the expense of pushing 4 further out of reach (since 3 would then likely have to reach up and take State College, while 4 takes Republican areas along the north of 5).  Probably not *that* much change, though, since these are monster districts. 

[traininthedistance]

For SC, how bout putting Columbia with the rest of the upstate area?

Yeah, that's much better than both of my maps.  Unless of course you really need a deviation of only 15.

[traininthedistance]

Tixas.



Texas is, incredibly, min-maj by VAP.  But the SSVR number, which is a better measure of the Hispanic voter base than the 33-plus percent Hispanic VAP population, is a mere 21.6 percent.  So two SSVR-maj districts seems like a good ideal here, and that is exactly what we have.  Max deviation is merely 560.

District 1: EL PASO-BIG BEND-LAREDO-BEXAR WEST.  O 58.5%, D 55.9%.  71% Hispanic, 60% SSVR.  Districts 1 and 2 together make up Hispanic South Texas, and are both SSVR-majority.  Splitting San Antonio between them, while somewhat unfortunate, balances the numbers and keeps the lines clean.  Anyway, this is sure to be the most Hispanic district in the nation.  Safe(ish) D.

District 2: RIO GRANDE VALLEY-CORPUS CHRISTI-BEXAR EAST. O 55.2%, D 53.6%.  68% Hispanic, 57% SSVR.  A little less secure, vulnerable to a Blake Farenthold-type upset in waves.  Lean D.

District 3: AUSTIN-CENTRAL TEXAS.  O 50.4%,  D 44.9%.  The rest of the San Antonio metro, all of the Austin and Killeen-Temple metros, a couple rural counties for population.  Yes, Obama won this in '08, but even as blue as Austin is this district is still Lean R because it remains surrounded by Texas.  It's on the verge of competitiveness, though, which is good because there should be something up for grabs in a state this large (well, District 2 is about as competitive, also).

District 4: WEST TEXAS.  O 24.8%, D 28.0%.  Yeah, about that.  Encroaches on the fringes of the Metroplex, too.  Safe R.



District 5: INNER METROPLEX.  O 61.7%, D 56.6%.  Only 35% White VAP; 24% Black and 35% Hispanic (plurality, but only 16% SSVR).  There are a zillion ways to draw the lines in DFW, and if you want to keep two districts entirely within the four core counties then two splits are necessary.  Making 5 Hispanic-plurality may not actually be necessary, and many alternatives would decrease the Obama margin here (I might come back and map a couple of them at some point), but it's hard to imagine a map that passes VRA muster where the DFW core- or at least Dallas County and remnants- district isn't Safe D or at least with a Dem lean.

District 6: COLLIN-DENTON-OUTER TARRANT.  O 37.1%, D 30.6%.  Safe R of course for the Metroplex burbs.

District 7: NORTHEAST-PINEYWOODS-WACO-REST OF METROPLEX.  O 33.4%, D 34.9%.  The (too-wordy) name says it all.  Safe R.



District 8: SOUTHEAST-BEAUMONT-HOUSTON FRINGE.  O 31.6%, D 34.9%.  Another giant and super-safe R district.  Like 7, it has to go into a major metro area that is too big for two whole districts, so it takes the whitest parts.

District 9: GALVESTON-BRAZORIA-FORT BEND-VILLAGES.  O 40.0%, D 36.1%.  A bit of Harris, bust mostly the southern bits of the Houston metro.  Amazingly, this district is min-maj by total population (only 52% White VAP) despite still being Safe R.

District 10: HOUSTON.  O 64.1%, D 61.7%.  Entirely within Harris, the Houston core and minority-heavy suburbs.  Majority-Hispanic by total population and 46% VAP (25% White and 23% Black).  Safe D.

With districts these size, and hewing to an ideal of "otherwise non-partisan map with VRA considerations", Texas comes out as… rather favorable for the Dems, actually. 

[traininthedistance]

Indiana.



Max deviation 533.

District 1- NORTH INDIANA.  O 52.3%.  The Obama-in-2008 numbers are hilarious, of course- Indiana is a Safe R state despite them.  With Gary and South Bend, this district should be competitive, but it's a tossup at best and probably actually deserves to be called Lean R. 

District 2- INDIANAPOLIS-CENTRAL.  O 50.1%.  I'd have loved to put the Indy metro all in one district, but the lines would be pretty awkward that way (too much population north of the metro, not enough south of it); getting the entire ring of counties around Marion should be good enough (and as long as I'm drawing the lines there will be an Indy-plus-ring district in the center).  I hesitate to call an Obama district Safe R... but I think this one is pretty close.

District 3- WABASH-EASTCENTRAL-SOUTH.  O 47.6%.  And the rest.  Safe R.

Or, alternatively, if you really hate the U-shape of 3, there's this one, with max deviation skyrocketing up to 773:



District 1 (NORTHWEST-WABASH) is now 55.5% Obama and might even legitimately be ranked Tossup.  2 remains 50.1% O yet solid for the Pubs, and 3 (EAST AND SOUTH) is now 44.5% Obama, even safer Pub than before. 

[traininthedistance]

Oregon.



Obviously you need to split it into a Safe D Portland Metro district (1) and a Lean R everywhere else (2).  I couldn't find a reasonable whole-county grouping with low deviation, so I just split Marion.  Better than crossing the Cascades.  No, I'm not going to go hunting for an even split that throws Portland in with Eastern Oregon, hilarious as that would undoubtedly be.

[traininthedistance]

Wisconsin.



Deviations 452.  Both the 2008 Obama and the "partisan average" numbers strike me as pretty far away from what a 50/50 nationwide election would bring.  They're closer to outer bounds than anything else.

District 1: Milwaukee-Madison.  O 58.1%, D 50.8%.  A pretty strong Lean D, getting close to safe despite Waukesha and friends.  Basically the more urban and less elastic southeast of the state; the Madison metro is technically split but virtually all the built-up area is in Dane anyway, and whole counties don't really work well if the Milwaukee district goes north instead, so I'm not going to sweat it that much.
District 2: Rest o'state.  O 54.3%, D 43.6%.  I can't tell whether this is a Tossup or a Lean R; in any case it is very swingy.

[traininthedistance]

Michigan.



Max deviation is 1780, for District 3.

District 1: UP-NORTH-THUMB-LANSING-SAGINAW.  O 52.7%, D 45.2%.  The northern district, has to encroach south to take the entirety of the Lansing-Owosso CSA and some far Detroit fringe.  I guess it's more Tossup than Lean R, but the Pubs probably have an advantage here. 
District 2: GRAND RAPIDS-SOUTH AND WEST.  O 50.1%, D 39.8%.  GR, Muskegon, Kalamazoo, Battle Creek, all of those sorts of places.  While the Obama-in-'08 number is not much less than in District 1, the baseline expectation is far more Republican (you don't have much in the way of ancestral Dem areas like the UP, or solid liberal parts like Lansing) and Obama's win should be considered quite the fluke.  Sort of like the Indianapolis district, probably deserves Safe R.
District 3: OAKLAND-MACOMB-GENESSEE.  O 57.1%, D 48.0%.  Yep, those three huge Detroit CSA counties, exactly.  Pretty nice to have two entirely Detroit-area districts, only one county cut, and preserve 8 Mile as a boundary.  Lean D I guess.
District 4: DETROIT-ANN ARBOR-DOWRIVER.  O 68.1%, D 58.2%.  No, there are not enough blacks in Michigan to even get a min-maj district; this district is 30% black.  Safe D.

[traininthedistance]

Come on feel the Illinoise.  





Max deviation 807.

District 1: SOUTH & WEST SIDE CHICAGO-JOLIET-KANAKEE.  O 78.5%.  42% White VAP, 44% Black VAP, 10% Hispanic.  No, a black-majority district is not possible in Illinois.  But black-plurality is, and I assume is necessary.  Since majority is impossible, I decided to make a map that splits no townships outside of Chicago rather than maxing out the numbers.  (The voting precincts line up with [square] townships rather than cities and villages, so I obviously used them as building blocks instead.)  Safe D.
District 2: DEKALB-WEST COOK.  O 63.2%.  55% White VAP, 31% Hispanic.  (36% by total population).  Yeah, a Hispanic-influence district.  These lines are a bit Dem-favoring, OK, but Cook needs to be split again anyway, it's just that big, and given the shape of 2 and 3 together this is just about as compact as you can get.  Safe D.
District 3: LAKE-NORTH SHORE.  O 64.8%.  70% White VAP, 13% Hispanic, 10% Asian.  And the last of our inner Chicagoland districts.  I suspect that all three of them are mostly in the suburbs rather than in Chicago proper; 1 might be majority-Chicago though.  Gotta split something else too, McHenry seemed easiest.  (I tried giving Kanakee to 5, to see if that would obviate the need for this county split, but no dice.)  Safe D.
District 4: AURORA-ROCKFORD-PEORIA-ROCK ISLAND-NORTHWEST.  O 53.7%.  82% White VAP, 10% Hispanic.  The rest of Chicagoland and the northern non-Chicago parts.  A couple outlying counties in the Peoria and Springfield MSAs are in the wrong district, fixing that would up the deviation to about 2K and not really change the partisan numbers much.  Anyway, despite the Obama numbers this is clearly Lean R.  
District 5: DOWNSTATE.  O 49.3%.  Obama won this by the thinnest of margins in 2008, which means it's obviously Safe R.

[traininthedistance]

Show me...Missouri.



Deviations 246.  No metros are split, and only the Columbia-Jeff City TV market is split.  Yes, I know that splitting north-south allows one to make a "Missouri" and a "Missourah" district... but a St.Louis and a KC district is obviously better anyway.

District 1- ST. LOUIS-COLUMBIA-SOUTHEAST is 53.6% Obama and 57.1% Dem.  Lean D I guess?  Not by much tho.
District 2- KANSAS CITY-SPRINGFIELD-JOPLIN-NORTH is 44.7% Obama, 48.4% Dem, and presumably pretty much Safe R, unless you get another Todd Akin or something.

[traininthedistance]

Washington.





Max deviation 584.  The geography of Washington does not work well with three districts; you don't want to cross the Cascades up north, and Eastern WA plus the Vancouver area is just a bit too small, and of course there's the Stevens Pass bit in King that doesn't connect to the rest of the county, and all those islands and ferries... and screw it, I'm just tri-chopping King and throwing the empty bits of Pierce in with 1 because the mountains ought to be together.  Can't say I'm happy about any of it, though.  Alternative maps welcome.

District 1- EASTERN WASHINGTON-VANCOUVER is 46.1% Obama, 40.8% Dem (the Gregoire-Rossi race in 2010), and Safe R.
District 2- TACOMA-OLYMPI(C/A)-SOUTH KING is 57.9% O, 52.9% Dem, and a very strong Lean D.
District 3- SEATTLE-BELLEVUE-EVERETT-NORTH is 67.2% O, 61.7% Dem, and Safe D.

[traininthedistance]

Massachusetts.



Max deviation 878.  Much like Washington, three districts is not a great number for a sensible division- the MetroWest area gets sliced and diced, and Boston is at the edge of its district, and I'm not sure there's anything I can do about either of those things.  Counties don't really matter in MA, but I tried to mostly keep them together anyway- basically Middlesex County is home to all of the choppage.

District 1: WESTERN AND CENTRAL.  O 60.5%, D 48.8%.  The town of Blackstone is taken from Worcester County and given to 2 to keep MA's portion of the Providence NECTA together.  Other than that... it's all of the west and central bits, and heading pretty deep into Greater Boston.  Safe D.
District 2: SOUTH.  O 58.6%, D 46.7%.  South Coast, South Shore, southern burbs, etc.  The least-safe district in the state, for what that's worth.
District 3: BOSTON AND NORTH.  O 67.7%, D 57.2%.  Boston, Essex County (the North Shore), and the inner suburbs in between.  Safe D.

[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.