The 1,000 Districts Series

[traininthedistance]

I think Miles likely used total population for each state rather than the apportionment population, and then did a division and rounded down.

Yeah, I tried; I figured someone else would put something more accurate out though.

This would be a good topic for a series.

Why the heck not.

I'm not going to do all of the states, but I did (surprise surprise) New Jersey, so let's kick it off there.

300K is a really really reasonable number for the most part; I allowed a little more deviation than I usually do in the name of clean lines and keeping municipalities, counties, CoIs together; every district is no further than 1 percent away from the ideal (maximum allowed deviation 3.03K, maximum actual deviation 2.53K) and only four municipalities are split, all of them in or around VRA districts.  VAP is given for all races with over 10% in a district, if no numbers are given then it's supermajority white.  I'll do this in two posts b/c of word limits.

The entire state:



DISTRICT 1- CAPE MAY-ATLANTIC CITY.  Pop 303,357.  O(bama) 54.0%, D(em) 52.5%.  69W/12B/11H.  The southern Jersey Shore, all of Cape May and most of Atlantic County.  Toss-up.

DISTRICT 2- VINELAND-DELAWARE BAY.  Pop 302,907.  O 56.4%, D 54.3%.  66W/16B/16H.  All of Cumberland and Salem counties, and adjoining rural portions of Gloucester and Atlantic.  Probably the most agricultural district in the state.  Lean D.

Close up on south and central NJ:



DISTRICT 3- GLOUCESTER-WINSLOW.  Pop 301,535.  O 58.2%, D 56.9%.  80W/12B.  Most of Gloucester County, and the southernmost townships in Camden (mostly Winslow, hence the name).  Suburbs and exurbs mostly.  Lean D.

DISTRICT 4- MID-CAMDEN.  Pop 301,541.  O 63.4%, D 60.7%.  77W/11B.  Entirely within suburban areas of Camden, dominated by Cherry Hill, Gloucester, and the older small towns in between.  Safe D.

DISTRICT 5- CAMDEN-BURLINGTON RIVERBANK.  Pop 303,048.  O 73.2%, D 69.2.  49.6W/28B/17.2H.  A line of cities and towns along the Delaware River from Camden City to Burlington.  The first majority-minority district, and the only one possible in southern NJ.  Safe D.

DISTRICT 6- BURLINGTON COUNTY.  Pop 303,890.  O 55.1%, D 49.9%.  78W/10B.  Almost all of the rest of Burlington County, except for the very southern tip.  Ergo, a broad mix of suburbs, farms, and the Pines (which it mostly splits with 7).  Toss-up.

DISTRICT 7- LONG BEACH-BARNEGAT-PINES.  Pop 304,111.  O 42.3%, D 43.1%.  This district takes in the rest of Atlantic County (primarily the Piney hub of Hammonton), cuts off the bottom Piney tip of Burlington, and then takes up the bulk of Ocean County, sweeping in all of the less-dense southern and western portions.  Safe R.

DISTRICT 8- SUBURBAN OCEAN.  Pop 301,752. O 41.3%, D 41.7%.  The northeastern corner of Ocean County, which is much more densely populated and clearly within NYC's orbit.  Brick, Lakewood, Toms River, and several smaller towns.  Safe R.

DISTRICT 9- TRENTON-HAMILTON.  Pop 303,365.  O 63.9%, D 59.4%.  60W/21B/14H.  All of Mercer County is here except for Princeton and the two Windsors, and then it takes in much of southern Hunterdon.  I tried to avoid so much of Hunterdon going into the more urban Trenton district (having this take more of Mercer and 12 take more of Somerset instead), but I couldn't make the math work out to do that and not split towns.  Anyway, the part of Mercer in 12 is more NYC-oriented, whereas the part here still has more of a Trenton-focused separate identity.  Safe D.

DISTRICT 10- MONMOUTH INLAND.  Pop 301,125.  O 43.5%, D 42.8%.  Population math unfortunately didn't let me take 11 all the way down the coast, so it has Manalapan and Brielle.  Safe R.

DISTRICT 11- MONMOUTH SHORE.  Pop 302,235.  O 51.6%, D 49.9%.  Pretty self-explanatory.  Lean R.

DISTRICT 12- PRINCETON-FRANKLIN-SOUTH MIDDLESEX.  Pop 303,975.  O 64.3%, D 58.5% 59W/21A.  This one's a hard one to name, but it's also a cohesive district that epitomizes Central Jersey.  Princeton and the Windsors, and then the southernmost reaches of Middlesex and Somerset.  As mentioned before, I'd rather Hillsborough be here, and have it withdraw from Middlesex so 9 can take a Windsor or two, but the math doesn't quite work out.  Safe D.

DISTRICT 13- CENTRAL MIDDLESEX.  Pop 305,699.  O 58.4%, D 56.9%.  57W/20H/13A. Takes in Aberdeen/Matawan from Monmouth (to finish that county), but the rest is Middlesex, mostly south of the Raritan but with Perth Amboy rather than New Brunswick for population purposes.  Lean D.

DISTRICT 14- EDISON-WOODBRIDGE-NEW BRUNSWICK.   Pop 305,133.  O 63.2%, D 62.5%. 47W/18H/25A.  Mostly those three large towns, with a couple adjoining smaller ones (Metuchen, Carteret, Highland Park).  The only district entirely within Middlesex (though 13 is almost as well), the second minority-majority district, and the district with the highest Asian population in the state.  Safe D.

DISTRICT 15- HUNTERDON-SOMERSET.  Pop 301,781.  O 44.0%, D 38.6%.  The rest of Hunterdon (most of it), and almost all of the rest of Somerset (excluding only Green Brook, Watchung, and North Plainfield, which are on the wrong side of Watchung Mountain).  I don't know for sure, but likely the richest district in the state.  Safe R.

More to come in the next post!

[traininthedistance]

Closeup on the urban northeast:



DISTRICT 16- ELIZABETH-NEWARK NORTH AND MEADOWS.  Pop 301,352.  O 75.2%, D 75.2%.  26W/15B/53H.  The first VRA district, it takes all of Elizabeth and the portion of Hudson County west of the Hackensack River (Kearney, Harrison, E. Newark), and splits Newark to take in the Hispanic portions (Ironbound and most of the North Ward).  Safe D.

DISTRICT 17- NEWARK SOUTH AND CENTRAL-UNION EAST.  Pop 302,014.  O 81.9%, D 80.1%.  25W/52B/18H.  Our second VRA district, and first of two AA districts.  The rest of Newark, and then a string down the eastern third of Union County, taking in towns with sizable AA populations like Roselle, Rahway, and Linden.  Clark doesn't really fit but it needed to be added for population, and Union is mostly in 18 but a little bit is here for connection and population.  Safe D.

DISTRICT 18- ORANGE-IRVINGTON-UNION CENTRAL.  Pop 302,856.  O 79.3%, D 73.4%.  33W/53B/10H.  The second AA district, it is centered on the black, non-Newark areas of Essex County (Orange, East Orange, Irvington), including Essex towns with sizable AA minorities (Maplewood/S. Orange/southern half of Montclair), and then dips into Union to make room for District 20 and unpack things a bit.  Montclair and Union are split here.  Safe D.

DISTRICT 19- PLAINFIELD-PISCATAWAY-UNION WEST.  Pop 303,905. O 60.5%, D 55.8%.  55W/16B/16H/12A.  The rest of Union, Middlesex, and Somerset are all here, centered on the two named towns and a series of mostly-swingy, uniformly-affluent towns in Union.  Lean D.

DISTRICT 20- SUBURBAN ESSEX.  Pop 302,283.  O 56.7%, D 56.2%.  67W/13H.  The rest of Essex County, including super-liberal areas like Upper Montclair and West Orange, super-conservative areas like the Caldwells, and everything in between.   Lean D.

DISTRICT 21- HUDSON SOUTH.  Pop 300,855.  O 74.9%, D 74.8%.  40W/19B/20H/18A.  Bayonne, Hoboken, most of Jersey City.  Much like JC itself, very very diverse.  Safe D.

DISTRICT 22- HUDSON NORTH.  Pop 301,568.  O 71.0%, D 71.4%.  25W/60H/11A.  The rest of Hudson County, and two towns in Bergen.  Very heavily Hispanic, obviously our fourth VRA district.  Safe D.

DISTRICT 23- PALISADES.  Pop 305,143.  O 62.4%, D 63.4%.  50W/17H/24A.  Englewood and Teaneck are the largest of many many towns in this all-Bergen district that takes the Hackensack River as its western border.  It's minority-majority by total population, and has the second-highest Asian population.  Safe D.

DISTRICT 24- MEADOWLANDS-HACKENSACK-FAIR LAWN.  Pop 305,467.  O 56.1%, D 57.8%.  65W/19H.  Another all-Bergen district, this one in between the Hackensack and Passaic rivers.  Lean D.

DISTRICT 25- NORTH BERGEN-HALEDON-HAWTHORNE.  Pop 305,165.  O 45.6%, D 43.7%.  The rest of Bergen , and the aformentioned towns north of Paterson to fill out population.  Safe R.

DISTRICT 26- URBAN PASSAIC.  Pop 305,981.  O 76.5%, D 74.4%.  26W/17B/51H.  Paterson, Passaic, Clifton, Prospect Park, and our last VRA district. Safe D.

DISTRICT 27- POMPTON-MORRIS NORTH.  Pop 301,460.  O 43.5%, D 42.5%.  The rest of Passaic, and a large swath of northern Morris County.  The part of Passaic north of the wasp-waist was almost its own county called "Pompton", hence the name.  Safe R.

DISTRICT 28- MORRIS CENTRAL AND SOUTH.  Pop 303,885.  O 47.1%, D 43.0%.  72W/13H/10A.  Most of Morris County, including all the well-known towns of note (Madison, Boonton, Morristown, Dover).  Probably the least secure of all Safe R districts on this map, if only because they don't take too kindly to Palin-types here.

DISTRICT 29- SKYLANDS.  Pop 304,607.  O 40.8%, D 37.1%.  Sussex, Warren, and the western tip of Morris.  The most Republican district on the map, Safe R.

...

Final tally includes:
3 Hispanic VRA districts (16,22,26)
2 Black VRA districts (17,18)
3 more districts which are majority-minority by VAP (5,14,21) and one by total pop only (23)

12 Safe D
6 Lean D
2 Tossup
1 Lean R
8 Safe R

[traininthedistance]

I'm glad you took me up on that!

I may add in a few southern states if you don't mind

Oh, of course.  I don't have the time or inclination to do everything, so contributions are obviously welcome.

[traininthedistance]

here is the state of Arkansas

I'm guessing district 9 is majority black?

Anyway, here's an easy little one.  Delaware!  Max deviation is only 746 this time.



DISTRICT 1- WILMINGTON AREA.  Pop 299,113. 70.5% Obama. 64W/23B.  Not much to say here, safe D.

DISTRICT 2- NEWARK-DOVER.  Pop 298,764. 67.5% O.  65W/23B.  The border between 2 and 3 tries to exactly mirror the city limits of Smyrna and Dover in Kent County, it's not possible to be exact with the lines we're given but this is pretty close.  Note that it is technically possible to do a Dover-Wilmington district and a Newark-suburbs district such that the former is below 50% white by total population, but this seems obviously better.  Safe D.

DISTRICT 3- SLOWER LOWER. Pop 300,057.  46.6% O.  78W/13B.  Safe R, of course.

[traininthedistance]

Oh, btw, the apportionment for 1000 districts:

Actually it would be exactly this, with DC getting 4 EV's  (Green indicates where I have a state getting one more rep than Miles did.  These are mostly the states that have enough representatives that they have an average population per district below than the national average, but not all of them. I think Miles likely used total population for each state rather than the apportionment population, and then did a division and rounded down.

BTW, Georgia got the 1000th seat, with Ohio in line to get the 1001th seat.

Alabama   16
Alaska   2
Arizona   21
Arkansas   9
California   121
Colorado   16
Connecticut   12
Delaware   3
Florida   61
Georgia   32
Hawaii   4
Idaho   5
Illinois   42
Indiana   21
Iowa   10
Kansas   9
Kentucky   14
Louisiana   15
Maine   4
Maryland   19
Massachusetts   21
Michigan   32
Minnesota   17
Mississippi   10
Missouri   19
Montana   3
Nebraska   6
Nevada   9
New Hampshire   4
New Jersey   29
New Mexico   7
New York   63
North Carolina   31
North Dakota   2
Ohio   37
Oklahoma   12
Oregon   12
Pennsylvania   41
Rhode Island   3
South Carolina   15
South Dakota   3
Tennessee   21
Texas   82
Utah   9
Vermont   2
Virginia   26
Washington   22
West Virginia   6
Wisconsin   18
Wyoming   2
TOTAL   1000

Workin' on New York, which is going to take a nice long time (not just because it's a large state, but DRA is so sloooooow whenever I load NY).  In the meantime, let's get through New England.  Starting with Maine:



All districts are super-white, of course, and no towns are split.  DRA combines a lot of towns in Maine into single voting districts (especially in the sparsely populated North Woods); if I had finer control I'd  be able to clean up the boundary between 2 and 4. 

DISTRICT 1: PORTLAND-SOUTH PORTLAND-BIDDEFORD.  Pop 333,755.  63.9% Obama.  York County, the city of Portland, and a couple of urbanized towns to connect the two.  This being New England, I'd rather name things after towns than counties (though there will be a couple exceptions); I still try to keep counties together when possible though.  Anyway, Safe D.

DISTRICT 2: AUBURN-LEWISTON-CUMBERLAND NORTH.  Pop 330,512.  57.6% Obama.  Working up the coast, this district is centered on Auburn/Lewiston and the northern suburbs of Portland, and then shares Saghadoc with 3 and Oxford with 4.  Lean D.

DISTRICT 3: AUGUSTA-DOWN EAST.  Pop 331,765.  56.2% Obama.  Kennebec County and the rest of the coastal counties.  Lean D.

DISTRICT 4: BANGOR-AROOSTOOK-NORTH WOODS.   Pop 332,329. 52.7% Obama.  As mentioned in the intro, I don't really like the hook around Rumford in Oxford County, and finer gradations there (where there are a lot of uninhabited and nearly-uninhabited townships) would help clean it up.  This may very well be the least-densely populated district east of the Mississippi- a lot of those woods are empty.  Tossup.

[traininthedistance]

Vermont!



These two districts are about as even as you can get.  District 1 (VERMONT EAST) is 312,867; District 2 (VERMONT WEST) is 312,874; both are 67.8% Obama and Safe D. 

While I'm sure perfect equality is possible (max deviation was 4 here, oh no!), I chose instead to get as close as I could while also hewing to the main division in the state- the Green Mountains, whose ridgeline conveniently tends to mirror county boundaries.  The western half is, while smaller in area, more populous due to Burlington, ergo I had to cross the mountains at some point.  From a topographic perspective crossing up by St. Albans might have been better than putting the bulk of Bennington County in the east district, but I also wanted to keep the relatively dense Burlington hinterlands together.

…



New Hampshire.

Trying to figure out the best lines for NH was surprisingly difficult for such a small and homogeneous state.  Something like District 4 was obvious, and even then, I toggled back and forth trying to hew closer to county lines or the state's official tourist regions (eventually settling on county lines).  But how to divide the Merrimack Valley and surroundings was a lot less clear.  Max deviation is 595, not great but it's going to be hard to get perfection if you want sensible districts, especially the large towns in the south of the state.

DISTRICT 1: MERRIMACK EAST (ROCKINGHAM).  Pop 328,995.  47.5% Obama. Eventually what I decided to do was to make District 2 an all-Hillsborough district and work out from there.  So, 1 is a thoroughly exurban district that is mostly Rockingham County, but takes in the rest of Hillsborough and excludes the older city of Portsmouth.  Lotsa Taxachusetts exiles in what is going to be the most Republican district in all of New England.  Lean R.

DISTRICT 2: MANCHESTER-NASHUA-HILLSBOROUGH.  Pop 329,592. 52.6% Obama.  The vast majority of Hillsborough is here, including the two main cities of Manchester and Nashua.  Tossup.

DISTRICT 3: PORTSMOUTH-CONCORD-LACONIA.  Pop 329,360.  57.5% Obama.  Less exurban and more small-town than District 1 and 2, it takes in all of Belknap and Strafford, the remainder of Rockingham (mostly Portsmouth), and most of Merrimack.  Early drafts tried to push it both further north and/or down the coast in concordance with the tourist regions, but that led to deviation problems when dividing 1 and 2 among the large exurban towns, and I'm not sure it was actually any better on the CoI front.  Lean D.

DISTRICT 4: WHITE MOUNTAINS-NORTH AND WEST.  Pop 328,523.  60.0% Obama.  Splits Merrimack and takes the entirety of all five other counties.  Touristy, rural, basically an extension of Vermont.  Safe D.

…

Triple feature!  Li'l Rhody!  Rhode Island has three districts to apportion among 39 towns; if you make the following assumptions:

1) towns should not be split
2) deviation should be as low as possible
3) lines should be reasonably sensible, and not split obvious distinct CoIs like the urban Providence core, Newport County, or Washington (aka South County).

then I think this might in fact be the only possible map.



RI does not have partisan statistics.  Max deviation is fairly high (3,175), and while that might be improvable, it would only be at the expense of a much worse map.  Unlike VT and NH, there is one district with a sizable minority population.

DISTRICT 1: PROVIDENCE-PAWTUCKET.  Pop 347,681.  57W/10B/24H.  Providence, Pawtucket, Central Falls, N. Prov, E. Prov. So, yeah, the urban core.  Safe D.

DISTRICT 2: WOONSOCKET-CRANSTON-WARWICK.  Pop 351,329.  The rest of Providence County, minus Scituate, plus Warwick.  You could probably get a cleaner-looking map by giving Scituate to 2 and microchopping Warwick, but that's the coward's way out.  No partisan stats, but this is RI so I'll assume Safe D.

DISTRICT 3: NARRAGANSETT BAY.  Pop 353,557.  South County, Newport, Bristol, and as much of Kent and Scituate as is left.  All two townships that voted for Romney are here, but let's be serious.  Safe D.

[traininthedistance]

My (very quick) attempt at what Republicans might draw in NC:



This would be 22-9.

Democratic sinks:
- 2 majority-black seats where Butterfield's district is currently located.
- 3 in the Triangle
- 2 in near Fayetteville; the Lumberton-Wilmington CD is less than 60% Obama, but is very heavily D downballot.
- 1 in Charlotte, which is 85% Obama. Black VAP majority.
- 1 in the Triad, which is also (barely) over 80% Obama. Black VAP majority.

Very nice; I wonder if the southern Charlotte district is one that would be in danger of flipping over the next decade.  I may try my hand at a neutral map at some point, but that won't be for quite awhile, if ever.

[traininthedistance]

Next up, Massachusetts and its 21 districts.  Only Boston is split; there is one all-Boston district, one almost-all-Boston district, and one district with a tiny sliver of Bunker Hill.  County lines are split all over the place, which is OK because most MA counties have literally no function.  Max deviation is 2,776, which is a little higher than I'd like, but I've worked on this enough.  Also, it's worth noting that the "Dem Average" is hilariously R-skewed, seeing as it's basically the Brown v. Coakley race.  You'll be seeing a lot of Lean D districts with R averages here, but I'm not actually going to call something safe unless it has a D average- with one exception informed by the extraordinary circumstances of an actual 2012 race.



DISTRICT 1: BERKSHIRE-MOHAWK-QUABBIN NORTH.  Pop 313,619.  O(bama) 67.4%, D(em) 60.4%.  All of the Berkshires and Franklin County (whose Mohawk Trail gives the district part of its name), the northwestern rural portion of Worcester near Quabbin Reservoir, and some more rural towns in Hampden and Hampshire.  Basically the most rural Western MA district, and also the most Democratic outside of the inner Boston area; you could easily call this "Baja Vermont".  Safe D.

Closeup on Springfield and Worcester area:



DISTRICT 2: SPRINGFIELD METRO.  Pop 311,332.  O 64.1%, D 51.7%. 67W/10B/19H.  Pretty self-explanatory.  Safe D.

DISTRICT 3: PIONEER VALLEY-SOUTH QUABBIN.  Pop 313,391. O 64.5%, D 54.3%.  Intermediate between the urban District 2 and the rural 1, this district is centered on the very socially liberal "Five Colleges" area of Amherst, Northampton, Holyoke, etc., which has a somewhat distinct identity from Springfield but is within its orbit.  Then it sweeps up some rural areas to the south and east, because  Western Mass. is underpopulated.  Safe D.

DISTRICT 4: WORCESTER METRO.  Pop 310,497.  O 61.1%, D 49.2%.  74W/11H.  You could probably do slightly better on the CoI front by making this district more spherical, rather than extending east-west, but I couldn't really make the population math work out in the MetroWest area with that configuration, le sigh.  The first of many strong Obama districts with Republican averages.  Lean D on principle, though it's probably safe in practice.

DISTRICT 5: BLACKSTONE VALLEY.  Pop 312,163. O 51.0%, D 35.7%. A swath of exurban towns in southern Worcester and Norfolk counties, and the second-most Republican district in the state.  Tossup.

DISTRICT 6: FITCHBURG-LEOMINSTER AREA.  Pop 314,568.  O 55.2%, D 42.1%  The "North County" designation doesn't really work, because the western edge of North County has to be eaten up by 1, and this district takes in a fair amount of northwestern Middlesex as well.  Lean D, and unlike 4 this one actually could fall in a wave.

Greater Boston.



DISTRICT 7: METROWEST. Pop 312,192.  O 61.8%, D 49.3%.  Except for Wellesley, which I wanted to fit in 14 but couldn't do in the end, every one of these towns is actually part of some MetroWest organization or another.  Mostly Middlesex but a couple Worcester County towns as well; like 4 this is functionally safe but Lean D on principle.

DISTRICT 8: LAWRENCE- MERRIMACK VALLEY EAST.  Pop 309,763.  O 58.7%, D 43.9%. 73W/22H.  Entirely within Essex County.  Lean D.

DISTRICT 9: NORTH SHORE.  Pop 310,720. O 62.4%, D 48.9%. 81W/11H.  Lynn is the largest city here.  Despite the R average, I will in fact call this one Safe D, on the grounds that John freaking Tierney managed to beat Richard Tisei here despite giving up, because the national Republican brand is just so toxic here.

DISTRICT 10: LOWELL- MERRIMACK VALLEY WEST.  Pop 309,400. O 54.3%, D 39.9% 79W/10A.  Let's be generous to the Republicans and call this Tossup.

DISTRICT 11: WONDERLAND.  Pop 312,419.  O 60.5%, D 48.4%.  70W/15H.  It's a cohesive district- the near-north areas of Greater Boston centered on Revere and Malden, transitioning between the urban core and the North Shore, but damned if I have a good name for it.  Perhaps "Revere/Malden" would be better than what I chose, which is the northern terminus of the MBTA's Blue Line, but whatever.  This district does dip a tiny bit into Boston proper, splitting Bunker Hill with 15.  "Lean D on principle", yet again.

DISTRICT 12: GREATER HAHVAHD.  Pop 311,848.  O 78.3%, D 73.2%.  73W/10A.  Camberville, Medford, Arlington, Watertown.  More or less urban, and college-clogged.  Super-safe D.

DISTRICT 13: ROUTE 128 CORRIDOR.  Pop 314,065.  O 61.5%, D 51.0%.  More suburban than 12, but like it entirely within Middlesex, and generally high-tech and socially liberal.  Lexington and Concord are here.  Safe D.

DISTRICT 14: NEWTON-BROOKLINE AND SOUTH.  Pop 311,204.  O 67.0%, D 57.8%.  This is kind of the "leftovers" district, as it were, centered on Newton and Brookline and then taking in some of the near-southern suburbs of Boston, down to the heavily Jewish town of Sharon.  The most socially liberal of the "southern" districts, which is I guess its unifying factor.  Wellesley would be a better fit here than 14, but I couldn't get the population math to work out before giving up.  Safe D.

DISTRICT 15: BOSTON NORTH.  Pop 313,873. O 73.0%, D 62.3%.  68W/13H/12A.  Splits Bunker Hill with 11, and takes in Winthrop as well (an earlier draft split Winthrop instead, but this lowers the  deviations for 11, 15, and 16).  Confusingly, it contains Southie.  Safe D.

DISTRICT 16:  BOSTON SOUTH.  Pop 313,306.  O 83.7%, D 79.6%.  35W/37B/17H.  Roxbury, Dorchester, etc.  Black VAP majority, and I suspect the only district in all of New England where whites aren't the largest ethnic group (though I have yet to see what's possible around Hartford).  I don't know if it technically counts as a VRA district or not, but it seems sporting to make it anyway.  Safe D.

DISTRICT 17: QUINCY-BRAINTREE.  Pop 309,789.  O 56.9%, D 47.2%.  78W/10A.  The inner areas of the South Shore, mostly Norfolk but a couple towns in Plymouth.  Basically, Stephen Lynch territory.  Lean D.

And closeup on southern Massachusetts.



DISTRICT 18: SOUTH SHORE-PLYMOUTH.  Pop 311,213.  O 49.2%, D 36.0%.  I tried my damndest to create an honest-to-god McCain district here, and still fell just short- 49.2 to 49.1.  Mostly in Plymouth, but a couple towns in Bristol, since 17 and 21 have to take bites out of the county, and Brockton really belongs in 19.  A Republican will probably win here more often than not, but Tossup anyway.

DISTRICT 19: BROCKTON-TAUNTON-ATTLEBORO.  Pop 309,302.  O 58.2%, D 43.1%. 76W/10B.  This Bristol County-based district takes in Brockton (for CoI) and Plainville (for population) as well, to create a district centered on those older industrial towns.  Lean D.

DISTRICT 20: NEW BEDFORD-FALL RIVER.  Pop 312,393.  O 66.6%, D 56.7%.  Entirely within Bedford, the South Coast region.  Safe D.

DISTRICT 21: CAPE COD AND ISLANDS.  Pop 310,572.  O 57.0%, D 47.4%.  Barnstable, Nantucket, Dukes, and the towns of Plymouth and Carver to fill population.  An obvious district.  Lean D.

...

One district is black VAP plurality.

10 Safe D
8 Lean D (three of which are functionally safe, seeing as they're over 60% Obama)
3 Tossup

[traininthedistance]

Given the shenanigans going on in Virginia, I decided to skip ahead and finish it next.  The goal, as in all of the maps I'm doing here, is a clean and fair map, but the techniques required to make one is very different here than in New England- first off, counties are much more important (as many areas are not part of incorporated towns), and second off there is the VRA to deal with.  While Virginia's 18% black population would theoretically be enough for 4 AA-majority seats, I was only able to make three of them without resorting to horrible gerrymandering.  Three of 26 is still better than one of 11, and a couple of the NoVA districts are under 50% white, as well.  A total of thirteen counties/independent cities are split, and the vast majority of those are unavoidable due to either being larger than a district, VRA concerns, or both.  Really, the splits of Spotsylvania and Roanoke are the only ones which aren't in these categories, and I'm not sure its possible to do better than that.

The whole state:



Closeup on Hampton Roads and the Richmond metro:



DISTRICT 1: EASTERN SHORE-VIRGINIA BEACH NORTH.  Pop 308,259.  O(bama) 51.5%, D(em) 47.6%.  67W/20B.  VA Beach is too large for a district, and the Eastern Shore only has a road connection to VA Beach.  Ergo, the best thing to do is split it north-south, with two districts mostly in Hampton Roads' largest (but not most central) city.  Tossup.

DISTRICT 2: PORTSMOUTH-INNER SOUTH HAMPTON ROADS.  Pop 308,744.  O 72.7%, D 67.3%.  39W/52B.  All of Portsmouth, and the blacker parts of Norfolk and Chesapeake = our first VRA district.  I'd talk about all the Naval stuff here, but it's just as true for all the surrounding districts.  Safe D.

DISTRICT 3: VIRGINIA BEACH-CHESAPEAKE SOUTH.  Pop 305,922. O 43.0%, D 41.3%.  70W/18B.  Self-explanatory.  Safe R.

DISTRICT 4: NORTH HAMPTON ROADS.  Pop 307,459.  O 57.8%, D 52.9%.  61W/27B.  The northern and whiter halves of Hampton and Norfolk, the middle of Newport News.  Basically, the more heavily urban areas that need to be outside of a VRA district.  Lean D.

DISTRICT 5: SOUTHSIDE EAST-INNER NORTH HAMPTON ROADS.  Pop 309,083.  O 64.3%, D 59.4%.  45W/50B.  Combining inner-city Newport News and Hampton with the Southside isn't great CoI, but it is by far the best way to get our second VRA district.  Safe D.

DISTRICT 6: TIDEWATER.  Pop 307,030.  O 44.0%, D 42.0%.  75W/17B.  The only split county here is Newport News, since it has to take the northern edge.  Takes in the northern reaches of Hampton Roads, Williamsburg, and the rural Tidewater.  Safe R.

DISTRICT 7: RICHMOND AND HENRICO EAST-PETERSBURG AREA.  Pop 305,963.  O 73.6%, D 67.5%.  36W/57B.  Richmond and Henrico are split with 8 to make our third and final black-majority district, and the Petersburg area is an obvious fit for this district.  Safe D.

DISTRICT 8: RICHMOND AND HENRICO WEST.  Pop 304,839.  O 57.3%, D 51.7%.  66W/21B.  The rest of Richmond, almost all of the rest of Henrico (a tiny bit is given to 10 for population, as much along community boundaries as I could).  This district has the highest deviation, at -2893; all the Richmond-area districts are underpopulated so that 7/8/9/10 can be a cohesive block of whole counties.  Lean D.

DISTRICT 9:  CHESTERFIELD/COLONIAL HEIGHTS.  Pop 306,669.  O 46.4%, D 41.8%.  67W/22B.  Almost all of Chesterfield, and Colonial Heights (it makes more sense IMO for the smaller indy cities to just be treated as part of the counties they're surrounded by).  Lean R.

DISTRICT 10: RICHMOND COUNTRYSIDE.  Pop 305,813.  O 37.1%, D 35.7%.  80W/15B.  Slivers of Henrico and Chesterfield, and then eight exurban and rural counties all of which are part of the Richmond metro area, Hanover being the largest.  Safe R.

DISTRICT 11: CHARLOTTESVILLE-LYNCHBURG.  Pop 308,223.  O 55.5%, D 52.7%.  74W/18B.  Charlottesville/Albemarle, the city of Lynchburg, and four counties in between.  Western VA was able to neatly subdivide into whole counties for the most part.  Lean D.

DISTRICT 12: DANVILLE-SOUTHSIDE WEST.  Pop 310,496.  O 44.0%, D 42.2%.  68W/29B.  Nine counties and Danville; largely rural, conservative, with a sizable black minority.  Safe R.

DISTRICT 13: SOUTHWEST-ALLEGHENY.  Pop 306,832.  O 36.6%, D 42.6%.  Eleven whole counties, Bristol, and Norton, not much to say here.  Safe R.

DISTRICT 14: SOUTHWEST-BLUE RIDGE.  Pop 306,832.  O 39.0%, D 41.7%.  Nine counties, Galax, and Martinsburg.  Safe R.

DISTRICT 15: ROANOKE-BLACKSBURG.  Pop 307,810.  O 50.0%, D 48.5%.  81W/11B.  Technically, you could avoid the split of Roanoke County here by giving Radford to 14 instead, but that would force multiple splits elsewhere, so better to do it here.  Tossup.

DISTRICT 16: SHENANDOAH SOUTH.  Pop 308,579.  O 36.9%, D 37.8%.  The rest of Roanoke, seven counties, Staunton and Waynsboro.  Bedford County is on the wrong side of the Blue Ridge Mountains, but even so I'm surprised I was able to get the Shenandoah Valley to work out as well as it did.  Safe R.

DISTRICT 17: SHENANDOAH NORTH.  Pop 309,805.  O 40.2%, D 36.4%.  Five counties, Winchester, Harrisonborg, blah blah blah.  Safe R.

DISTRICT 18: PIEDMONT.  Pop 309,012.  O 44.0%, D 40.9%.  81W/11B.  Rural and rural-ish areas north of Richmond/Charlottesville and south of Fake Virginia; seven whole counties and most of Spotsylvania.  Safe R.

And, finally, closeup on NoVA.  It's impossible to keep entirely to CDP boundaries, but as usual I tried to get close.



DISTRICT 19: FREDERICKSBURG-QUANTICO.  Pop 307,283.  O 50.3%, D 45.9%.  67W/19B.  Centered on Fredericksburg, Stafford, and the Quantico Marine Corps Reservation, it also take in a little bit of the Northern Neck (King George & Westmoreland), and parts of Spotsylvania and Prince William.  Most maps get stuck with an ugly duckling leftovers district, and this is that district.  Lean R.

DISTRICT 20: LOUDON. Pop 305,971.  O 53.9%, D 47.7%.  64W/12H/14A.  All of Loudon save the Dulles Airport district (for population).  The quintessential Tossup.

DISTRICT 21. MANASSAS-PRINCE WILLIAM. Pop 309,445.  O 60.6%, D 52.3%.  47W/20B/23H.  Manassas, Manassas Park, and most of Prince William, especially minority-heavy portions like Dale City.  One of two minority-majority districts I was able to make in NoVA.  Safe D.

DISTRICT 22: DULLES-CENTREVILLE-CHANTILLY.  Pop 306,706.  O 55.1%, D 48.9%.  59W/10H/21A.  The Dulles district of Loudon, and the further-out, more conservative portions of Fairfax and Prince William combine to create a district that still voted for Obama by over ten points.  Tossup.

DISTRICT 23: TYSON'S CORNERS-RESTON-FAIRFAX.  Pop 307,238.  O 60.5%, D 56.1%.  63W/12H/17A.  The City of Fairfax and surroundings, including the major employment center of Tyson's Corners.  This is probably what most people think of when they think of Fairfax County.   Safe D.

DISTRICT 24: ANNANDALE-SPRINGFIELD.  Pop 307,394. O 60.9%, D 56.1%.  49W/21H/19A.  There's probably a better name for this district, but I can't think of one.  It's the only one that's technically entirely within Fairfax County (though 23 functionally is as well), and the other minority-majority NoVA district.  Safe D.

DISTRICT 25: ALEXANDRIA-MOUNT VERNON.  Pop 309,405.  O 67.9%, D 63.7%.  55W/19B/16W.  Alexandria and Arlington are too big to be a district together, so I had to split them up.  Safe D.

DISTRICT 26: ARLINGTON-FALLS CHURCH-MCLEAN. Pop 306,883. O 67.8%, D 64.5%.  68W/12H/11A.  The one district here where CDP lines and voting districts lines up perfectly.  Safe D.

Final tally:  three AA-majority districts (2, 5, 7) and two more min-maj (21, 24).

Safe D: 8
Lean D: 3
Tossup: 4
Lean R: 2
Safe R: 9

Pretty perfect for one of our swingiest swing states, I think.

[traininthedistance]

Pennsylvania in 41 districts.  This will be in two parts.

PA, like NJ, is entirely made up of incorporated municipalities, but the counties also mean something.  However, also like NJ, you have quite a few large counties which necessitate splits, so county lines were something I had to take more as a suggestion than as an iron-clad rule.  There are three AA-majority districts, which may not be the theoretical maximum (I was not willing to get very ugly in my search to see what was possible) but is, again, better than 1 out of 18.  

Only two cities are split: Philly, obviously... and Bethlehem, which is itself split between two counties.  I ended up further subdividing the smaller portion of Bethlehem in Northampton.  I suppose there may be a couple other towns which straddle county lines that were broken along said county lines, but none come to mind immediately.  Within Philly, I tried to hew exactly to ward boundaries, but gave up and had to split one ward.

I'm not very happy with the northeastern quarter of the state, and may try different things up there at some point.

The state as a whole:  



DISTRICT 1: ERIE.  Pop 309,166.  O(bama) 58.0%, D(em) 53.7%.  All of Erie, and enough of Crawford for population.  Lean D.

DISTRICT 2: MAHONING VALLEY.  Pop 309,444.  O 47.3%, D 49.4%.  Centered on Lawrence (New Castle) and Mercer (Sharon) Counties, it also takes in most of Crawford and part of Beaver.  While on the fringe of both the Erie and Pittsburgh areas, the real center of gravity in this district is just across the state line- Youngstown, Ohio.  Lean R.

SWPA zoom:



DISTRICT 3: SOUTH MONOGAHELA.  Pop 310,042.  O 51.0%, D 58.2%.  The most SWPA of all SWPA districts, it's all of Fayette and Greene, most of Washington, and the old river town of Monessen in Westmoreland County.  Ancestrally super-Dem, but for how long?  Lean D.

DISTRICT 4: WESTMORELAND.  Pop 308,443. O 39.2%, D 45.0%.  Almost all of Westmoreland County; save the small bites taken out for 3 and 11.  By now, probably Safe R.

DISTRICT 5: NORTH MONOGAHELA-PENN HILLS.  Pop 310,881.  O 59.3%, D 62.9%.  78W/18B.  The westernmost slice of Allegheny County; centered on the Monogahela up to towns west of Pittsburgh and south of the Allegheny like Penn Hills and Plum.  Safe D.

DISTRICT 6:  METRO PITTSBURGH-EAST AND SOUTH HILLS.  Pop 309,558.  O 51.0%, D 54.2%.  Mostly the inner-ring Pittsburgh suburbs not in the Mon Valley; Ross and Bethel Park are the largest towns here, of which there are many.  I thought Lean D for a moment, but given trends in SWPA it's probably Tossup.

DISTRICT 7: PITTSBURGH CITY.  Pop 309,107.  O 75.5%, D 76.4%.  69W/23B.  Just Pittsburgh!  Well, and tiny Mt. Oliver, which is surrounded.  Along with 5 and 6, one of three districts entirely within Allegheny (FWIW, the county is too large for four).  A convenient Safe D pack for the natural Pub gerrymander that is PA.

DISTRICT 8: AIRPORT-BEAVER-NORTH WASHINGTON.  Pop 310,256.  O 44.6%, D 49.4%.  Most of Beaver, the rest of Washington, and the further out portions of the West and South Hills in Allegheny, including the airport.  Really, just the exurban counterpart to 6; splitting them south/west rather than inner/outer would have made for easier names, but either a much uglier Washington County tentacle or another county cut.  Probably still Lean R, for now.

DISTRICT 9: NORTH ALLEGHENY-SOUTH BEAVER.  Pop 310,213.  O 40.8%, D 42.5%.  Pretty self-explanatory.  These numbers probably ought to indicate a safe Republican seat, but this also was Jason Altmire's turf.  Eh, was.  Safe R.

DISTRICT 10: NORTH BUTLER-OIL-ALLEGHENY FOREST.  Pop 307,456.  O 38.0%, D 40.4%.  The rest of Butler, the northern two-thirds of Armstrong, and all of five rural counties  (Warren, Forest, Venango, Clarion, Jefferson).  Most of Allegheny National Forest is here, along with the birthplace of the petroleum industry and the Marcellus Shale.  Safe R.

DISTRICT 11: JOHNSTOWN-INDIANA-CONEMAUGH.  Pop 308,054.  O 48.0%, D 53.1%.  All of Cambria and Indiana, filling out the rest of Armstrong and Westmoreland along the Conemaugh and Allegheny, and a tiny bit of Somerset for population.  As is the case with the other Western PA districts dominated by smaller old Rust Belt towns (2, 3, 8 in particular) a lot of ancestral Dem strength, but may not be a Tossup for long.

DISTRICT 12: NITTANY.  Pop 311,901.  O 50.9%, D 50.8%.  Five whole counties (Centre, Clearfield, Clinton, Cameron, Elk).  A nice rectangular shape.  State College and some forests.  Clearly the best-looking district around.  Tossup.

DISTRICT 13: ALTOONA-SOUTHERN ALLEGHENIES.  Pop 307,041.  O 34.1%, D 36.6%.  The rest of Somerset (most of it) and four whole counties (Somerset, Bedford, Blair, Huntingdon).  This very rural, very red district seals off the 12-district whole county group that is Western PA.  Also the highest deviation on the map, at -2,773 (which is a result of the whole-county grouping).  Safe R.

DISTRICT 14: WILLIAMSPORT-NORTHERN.  Pop 310,076.  O 37.5%, D 35.8%.  Another whole-county grouping (seven of them) in the deepest reaches of the T, so I had to make it, but a good deal uglier than 12.  Recombining this with 24 to have a northern and southern district, rather than eastern and western, might actually be worth it despite the additional county cut.  Safe R.

Closeup on South-Central PA.  I was able to get one more whole-county grouping out of districts 15 through 19.



DISTRICT 15: CARLISLE-WEST SUSQUEHANNA.  Pop 312,232. O 39.8%, D 36.7%.  Three rural ridge-and-valley counties (Juniata, Mifflin, and Perry), and the bulk of Cumberland, including Carlisle and Harrisburg suburbs along the Susquehanna.  Safe R.

DISTRICT 16: YORK COUNTRYSIDE.  Pop 311,580.  O 39.2%, D 35.7%.  The bulk of York County, with a notch given to 19 for population and the urban areas cut out for 17.  Safe R.

DISTRICT 17: URBAN LANCASTER-YORK.  Pop 311,568.  O 55.6%, D 48.5%.  76W/13H.  The cities of Lancaster and York, connected by their inner suburbs and the Lincoln Highway.  Yes, Virginia, that is a 55% Obama district mostly in Lancaster County, and yes it makes all the sense in the world to draw.  Tossup.

DISTRICT 18: LANCASTER COUNTRYSIDE.  Pop 312,297.  O 35.0%, D 29.2%.  The rest of Lancaster.  I half want to give this district the joking name of "Amish Paradise".  Safe R.

DISTRICT 19: FOUR SCORE-SEVEN YEARS AGO.  Pop 310,458.  O 36.0%, D 34.2%.  All of Franklin and Adams, and the leftover bits of Cumberland and York.  Chambersburg is the largest town here, but most people will recognize Gettysburg instead.  Safe R.

DISTRICT 20: HARRISBURG-LEBANON.  Pop 310,673.  O 53.3%, D 48.1%.  75W/14B.  All of Dauphin, and an arm out to Lebanon in its namesake county.  A little ugly, but I tried many iterations here and in the northeast and Lebanon seemed to be the best partner to a Harrisburg district.  Tossup.

[traininthedistance]

Northeastern PA (Scranton-Wilkes Barre, Lehigh Valley, Poconos)



DISTRICT 21: SCRANTON-NORTHEAST.  Pop 310,019.  O 57.3%, D 54.9%.  All of Lackawanna (Scranton) and Wayne; parts of Susquehanna and Pike.  I'm not a huge fan of the two county splits, or how Scranton gets thrown in with a lot of rural area, but I also wanted a Poconos district and an all-Luzerne district, and this was what was left.  Lean D.

DISTRICT 22: POCONOS.  Pop 310,006.  O 54.1%, D 51.8%.  Centered on Monroe and Carbon, takes the rest of Pike and part of Schuylkill.  There used to be coal, now there's tourism.  Tossup; Monroe is trending D but the rest isn't.

DISTRICT 23: LUZERNE.  Pop 309,394.  O 54.2%, D 55.9%.  Luzerne is large enough for its own district, so it gets one.  The decision of "what to chop" for population was necessitated by 24 needing a bridge.  Lean D.

DISTRICT 24: ANTHRACITE-ENDLESS MOUNTAINS.  Pop 310,305.  O 43.0%, D 41.1%.  This is basically two distinct areas which are too small for their own district, strung together by the remainder of Luzerne.  To the south, there's the old coal counties of Northumberland, Montour, and Columbia.  To the north, you have the rest of the Northern Tier (most is in 14): Bradford, Wyoming, most of Susquehanna.  Safe R.

DISTRICT 25: SCHUYLKILLL-BLUE MOUNTAIN.  Pop 308,618.  O 41.9%, D 39.9%.  Roughly a rectangle around the Schuylkill River and Blue Mountain, this district is most of Schuylkill, the heavily agricultural and/or Appalachian portions of Lebanon and Berks, and a couple towns in Lehigh so that the Lehigh Valley can have two full districts otherwise.  Pottstown is the largest city here, for what that's worth.  Safe R.

DISTRICT 26: NORTHAMPTON.  Pop 307,754.  O 55.8%, D 55.2%.  All of Northampton, and a sliver of Bethlehem's Lehigh portion to fill population.  Lean D in a vacuum, though presumably this district would keep sending Charlie Dent to congress.

DISTRICT 27: LEHIGH.  Pop 310,263.  O 58.0%, D 55.2%.  75W/16H.  Entirely within Lehigh, Allentown is obviously the anchor.  Lean D.

DISTRICT 28: GREATER READING.  Pop 308,299.  O 57.2%, D 53.6%.  77W/17H.  Entirely within Berks.  Lean D.

And finally, closeup on the Delaware Valley.



DISTRICT 29: LOWER BUCKS.  Pop 312,067.  O 58.1%, D 57.6%.  Bucks almost perfectly fits into two districts, and upper/lower is the obvious way to divide.  Levittown is here, and it's still almost as white as Arthur Levitt dreamed it would stay forever.  Lean D pretty strongly.

DISTRICT 30: UPPER BUCKS.  Pop 310,975.  O 49.7%, D 46.8%.  Doylestown, Quakertown, rich exurbs, yadda yadda.  I think there might be a few farms left.  Obama did barely win this in '08 but it's a pretty strong Lean R anyway.

DISTRICT 31: MONTGOMERY EAST-ABINGTON-LANSDALE.  Pop 312,295.  O 61.3%, D 57.9%.  Abington is the largest town in the district (and second-largest in the county), Lansdale is the county seat.  Thoroughly suburban, and Safe D.

DISTRICT 32:  MONTGOMERY CENTRAL-NORRISTOWN-PERKIOMEN.  Pop 310,577.  O 56.7%, D 52.5%.  This one goes further into the exurbs, running from Conshy, Plymouth Meeting, and Norristown up the Schuylkill and Perkiomen Creek to hit MontCo's northernmost reaches.  It also takes a tiny sliver of Bucks, since those districts are overpopulated and Telford crosses the county line anyway.  Lean D.

DISTRICT 33: POTTSTOWN-CHESTER AND MONTGOMERY WEST.  Pop 310,351.  O 53.6%, D 48.9%.  A couple towns in Berks for population and the westernmost edge of Montgomery, but this is mostly a Chester County district that encapsulates the fringes of the metro.  Lean R, because we all know that Jim Gerlach has won here even in a bad year.

DISTRICT 34: MAIN LINE.  Pop 309,832.  O 58.5%, D 54.1%.  The Main Line is enough of its own thing that it's worth another county slice to have its own district.  The core is Lancaster Ave and the SEPTA line from Lower Merion in Montgomery to Downingtown in Chester, and then surrounding and intermediate towns (including a bit of Delaware) are taken in.  Famously old money, but also quite socially progressive these days.  Lean D.

DISTRICT 35: WEST CHESTER-KENNETT SQUARE-WEST DELCO.  Pop 308,746.  O 59.0%, D 54.6%.  77W/14B.  The rest of Chester (mostly the inner portions not part of the Main Line) and the western half of Delco, mostly to the west of Crum Creek but I had to add Swarthmore for population.  Should be Lean D but you never know with the GOP's historic downballot strength here.

DISTRICT 36: INNER DELCO.  Pop 308,346. O 59.3%, D 57.0%.  75W/16B.  Upper Darby is the largest town in this district which is mostly prewar suburbia (and that's not something you can make very often!).  Lean D much like the last one.

DISTRICT 37: WEST PHILLY.  Pop 309,050.  O 92.0%, D 90.7%.  23W/66B.  66 percent black is rather more packed than I'd like, but it's also hard to argue with a boundary that is almost perfect West Philly, with just a couple Delco towns (the AA-majority Yeadon and Colwyn, the airport-majority Tinicum) and one South Philly ward added on.  And it doesn't prevent me from making two more VRA districts.  Safe D.

DISTRICT 38: PHILLY RIVERFRONT AND NEAR NORTHEAST.  Pop 308,541.  O 78.6%, D 79.6%.  49.6W/15B/26H.  This district is more defined by what it is not- i.e. neither AA-majority nor Far Northeast.  It takes in the eastern,  whiter half of South Philly, and runs up the Delaware through Center City, NoLibs, up to Frankford.  It also tries to get as much of the Hispanic community in North Philly as it can.  (In fact, my desire to do this, and do it in a district separate from the black districts, was one of the main reasons I settled on only three AA districts rather than four.)  The one split ward is between this and 39- most of Ward 5 (Center City east) is here, but Chinatown is lopped off for population.  Safe D.

DISTRICT 39:  MID-PHILLY.  Pop 308,326.  O 93.4%, D 91.7%.  29W/54B.  Hard to name this one concisely.  It runs from Point Breeze through Center City West up to (most of) North Philly, and has some of the city's richest (Rittenhouse Square) and poorest neighborhoods.  Safe D.

DISTRICT 40:  NORTHWEST AND FAR NORTH PHILLY.  Pop 308,495.  O 88.5%, D 86.7%. 31W/53B.  The Northwest (Manayunk/Roxborough, Chestnut Hill, Germantown); the middle-class black neighborhood of Oak Lane, and some of the more diverse Northeast neighborhoods north of Roosevelt Boulevard.  Our final VRA district.  Safe D.

DISTRICT 41: FAR NORTHEAST.  Pop 309,674.  O 58.5%, D 61.3%.  73W/12B.  Self-explanatory.  Safe D.

...

Three AA-majority districts (37, 39, 40) and one more min-maj (38), all in Philly of course.

Safe D  8
Lean D  12
Tossup  6
Lean R  4
Safe R  11

This basically ended up being an inadvertent "gerrymander for competitiveness".  Which I think I'm okay with.

[traininthedistance]

Wouldn't 8 and 11 be pretty favorable to Altmire and Critz though? I'd put them in the D column.

I've been trying to not take incumbents or specific candidates into account too much (otherwise one of those Lehigh Valley districts would be Lean R for Dent).  But I think 8 would be sufficiently fertile ground for Tim Bishop as well (it has his home), and Altmire is out of poltics now.  Critz might have a better chance in 11, but again that whole area is trending pretty hard R and he no longer has the benefit of incumbency.

[traininthedistance]

Don't you mean Westmoreland County for District 4, not Washington?

Thanks, corrected.

[traininthedistance]

And now for something completely different.



Deviations are 56, no partisan figures given. 

DISTRICT 1: LAST FRONTIER.  Pop 354,341. 66W/20N(ative).  Fairbanks, Juneau, the Aleutians, North Slope, the Bush, basically everything that isn't Anchorage or Mat-Su.  Unfortunately, that's still less than half the state's population, so it has to encroach on Mat-Su, taking much of Palmer and the empty areas of that borough.  This is Alaska, so it's obviously not contiguous by road.  Almost all of the state's heavily Dem areas (primarily Juneau and the Bush) are here, so it's definitely winnable by the right Dem in the right year.  But it also has Fairbanks and enough of Mat-Su to probably keep it Lean R at the moment.

DISTRICT 2: ANCHORAGE-MAT-SU.  Pop 354,229.  Only 70% white, but no other group breaks ten percent.  No, you can't see Russia from here.  Safe R.

[traininthedistance]

It would be neat to see a map of Alaska where you try to divide it evenly by land area as well. You'd have to divide up Anchorage of course.

I bet the splitline method gets you close, seeing as it'll likely give you a straight north-south chop.  (This would, of course, be yet another illustration of just how horrible splitline redistricting actually is.)

[traininthedistance]

I just did Alaska, why not Hawaii?



DISTRICT 1: MAUI-MOLOKAI-LANAI-BIG ISLAND.  Pop 340,003.  76.3% Obama.  36W/37A/18O (Other- presumably some native Pacific Islanders are choosing this, and/or there are a lot of people with multiple races).  Hawaii and Maui counties (and the leper colony of Kalawao, if that even still counts); the four named islands plus smaller Kahoolawe.  Almost perfect, with a deviation of only 72 (unfortunately, large precinct sizes in Oahu make it hard to get the others as close).  This is the whitest district in the state, by the way.  Safe D.

All of the other three districts are Oahu-based.



DISTRICT 2: HONOLULU CITY.  Pop 339,552.  72.4% Obama. 19W/63A/12O.  As close to the CDP boundaries as I could get (there are no actual government functions in Hawaii below the county level).  Probably the most Asian district in the country.  Safe D.

DISTRICT 3: PEARL HARBOR-CENTRAL OAHU.  Pop 341,423.  18W/56A/15O.  Given the heavy military presence here, no surprise this is the state's most Republican district at a blistering 68.1% Obama.  Safe D.

DISTRICT 4: OUTER OAHU-KAUAI.  Pop 339,323.  70.2% Obama.  30W/42A/20O.  Mostly the areas separated from the "town" side of Oahu by the Wainae and Koolau; also the North Shore and the area east of Honolulu.  Then Kauai (island and county), Nihau, and all the tiny outlying atolls that stretch deep into the Pacific.  Safe D.

[traininthedistance]

I went ahead and tried to make a reasonable 9-district Utah map. It's primarily interesting to see what isn't safely Republican. I don't really like the way CD-01 and CD-02 turned out, but they were basically boxed in after the others were drawn. The districts in Salt Lake County only look somewhat odd because of the large and mostly empty voting districts within the county. The largest deviation of any district was 619.

District 1: UTAH NORTHWEST: McCain 66.4-30.9 Despite being geographically large, the vast majority of the population of this district is in Weber and Box Elder Counties. Safe R.
District 2: UTAH NORTHEAST: McCain 65.7-31.0 I don't like how this district really turnout out overall. Like CD-01, the population is heavily concentrated along the Wasatch Front. Safe R.
District 3: DAVIS COUNTY: McCain 69.7-27.5 Davis County is virtually a perfect fit for one district. Romney won this 80-18. Safe R.
District 4: SALT LAKE CITY: Obama 62.5-34.8 This is SLC plus some of its immediate southeastern suburbs. The eastern part of this is mostly empty space that didn't really fit anywhere else. This district should have no problem electing a fairly progressive Democrat. Safe D.
District 5: WEST VALLEY CITY-TAYLORSVILLE-MURRAY: Obama 48.6-48.3 Yes, that's actually a second Obama district in Utah. CD-04 probably has too many Democrats for Matheson to win a primary, so this could be a possibility for him. This is also the least white district I drew in Utah. Toss-up/Lean D (Likely/Safe D with Matheson?).
District 6: WEST JORDAN-SOUTH JORDAN-SANDY: McCain 57.0-40.4 The third and final district entirely contained within Salt Lake County. It's certainly not inconceivable that Matheson could win this one as well. He represents the majority of this district now. Likely R (Toss-up/Lean D with Matheson?).
District 7: RIVERTON-DRAPER-UTAH LAKE: McCain 75.5-21.4 I couldn't think of good name for the Utah County portion of this. It's about 2/3 Utah County and 1/3 Salt Lake County. Not much else to really say about this district. Safe R.
District 8: PROVO-OREM: McCain 76.3-20.2 This district competes with CD-07 as probably the most pro-Romney district that can possibly be drawn in the entire country. It's quite possible Obama didn't even break 10% here in 2012. Safe R.
District 9: UTAH SOUTH: McCain 73.7-22.9 I wanted to draw a Southern Utah seat, which turned out quite nice. Close to half the population comes from Washington County. I think it's interesting to see a district with 1/9 of the population take up over half the area of the state. Safe R.

This map should give Democrats at least one seat, possibly three on a very good night. I think the most likely result is 7R-2D.

I think District 1 might actually be non-contiguous by road, seeing as the Bonneville Salt Flats don't really have any north-south connections.  It's a pretty hard barrier, in any case.  So I imagine that would get the map thrown out.  

[traininthedistance]

Continuing to procrastinate on New York (which remains about half done) with... Arizona.

County lines (such as they are in a state where over half the districts enter Maricopa) were mostly respected, except for Districts 1 and 6, which required a fairly high level of erosity in the name of trying to create minority opportunity districts (especially important given Arizona's... unique political culture).  Town lines are mostly kept to when possible, but town lines out west are often crazy and impossible, with numerous unincorporated enclaves.  Also, the VRA districts tend to slice towns in several directions even when they try to be compact.

The whole state:



DISTRICT 1: NAVAJO-HOPI-FLAGSTAFF.  Pop 301,364.  O 58.1%, D 58.9%.  43.6W/44.1N.  Underpopulated by just about as much as I'm willing to go (3,018) in the search for a Native-plurality district.  And it just makes it.  All of Coconino, most of Navajo and Apache, smaller bits of Mojave, Graham, and Gila.  More to the point, it takes in all the northern reservations (Navajo, Hopi, Fort Apache, and a couple smaller ones).  Safe D.

Closeup on Tucson and Pinal:



DISTRICT 2:  COCHISE-SAFFORD-SOUTHWEST TUCSON.  Pop 303,685.  O 37.1%, D 37.5%.  69W/23H.  The southwestern corner of the state, centered on Cochise County and reaching north to fill in white areas behind District 1, then taking in parts of Pima to the western edge of Tucson.  Very much the map's leftovers district.  Safe R.

DISTRICT 3: SOUTH TUCSON-SANTA CRUZ.  Pop 304,504.  O 57.1%, D 57.1%.  37W/56H.  All of Santa Cruz County and the southern part of the Tucson area, reaching into the heavily Hispanic portions of Tucson proper.  Our first of four Hispanic-majority districts.  Safe D.

DISTRICT 4: TUCSON CITY.  Pop 303,088.  O 61.7%, D 61.6%.  62W/26H.  Entirely within city limits.  Safe D.

DISTRICT 5: TUCSON NORTH AND WEST-FOOTHILLS.  Pop 303,823.  O 47.5%, D 46.3%.  72W/21H.  Tucson west of I-10, and all its northern and western suburbs within Pima County, including Catalina Foothills, Casas Adobes, and Oro Valley.  The western border of this district with 6 looks strange, but it's the boundary for the Tohono O'odham rez.  Lean R.

DISTRICT 6: SONORAN DESERT-BORDER-MARICOPA MOUNTAINS.  Pop 305,465.  O 51.6%, D 52.0%.  35W/50H.  Probably the ugliest district I will draw in this entire series, but I'm drawing it anyway.  The heart of this district is a wide and mostly empty swath along the Mexican border from the Tohono O'odham reservation to Yuma.  But then it gets kind of ugly in an effort to accommodate county boundaries for districts 7 (Yuma) and 9 (Pinal), as well as take in as many Native and Hispanic voters as it can.  To that end, it goes north into Maricopa, grabbing the Gila River reservation and a small slice of the Phoenix metro.  By VAP this should be a VRA district, but by CVAP and partisan stats I'm not sure it quite passes muster.  Still, ground zero for the SB 1070 controversy, so appropriately a Tossup.  

DISTRICT 7: COLORADO RIVER.  Pop 303,661.  O 33.3%, D 30.4%.  74W/20H.  The western edge of the state- almost all of Mohave, all of La Paz, the whiter areas of Yuma.  The Grand Canyon is here, and so is the London Bridge.  Safe R, the most Republican district in the state.

DISTRICT 8: PRESCOTT-MARICOPA FAR WEST.  Pop 303,841.  O 36.6%, D 34.0%. 80W/15H.  Self-explanatory.  Goes in to the fringes of the Phoenix metro area, but most of the Maricopa space here is just that, space.  Safe R.

DISTRICT 9: PINAL-APACHE JUNCTION.  Pop 304,369.  O 41.6%, D 39.9%.  68W/23H.  All of Pinal that isn't in 6.  The jagged boundary is mostly the fault of the Gila River and Ak-Chin reservations.  Safe R.

And, of course, Maricopa.



DISTRICT 10: SCOTTSDALE-GILA-FOUNTAIN HILLS.  Pop 303,477.  O 40.2%, D 36.9%.  Takes in the rest of Gila County (our final non-Maricopa portion), as well as the Ft. MacDowell Yavapai and Salt River Pima reservations, but the majority of population can be found in Scottsdale.  As legendarily Republican as Scottsdale is supposed to be, there are totally a bunch of Democrats in south Scottsdale, which is not true for some other districts here.  Safe R anyway.

DISTRICT 11: MESA EAST.  Pop 303,413.  O 35.1%, D 34.1%.  79W/15H.  Mesa is pretty big, so it has to get split.  This is functionally an all-Mesa district (there are a couple unincorporated enclaves and the edge of Apache Junction), so most of the city is here.  Safe R.

DISTRICT 12: GILBERT.  Pop 305,255.  O 35.8%, D 32.8%.  76W/13H.  All of Gilbert.  Also Queen Creek and what look like the southeastern outskirts of Mesa and Chandler.  Safe R.

DISTRICT 13: CHANDLER- AHWAHTUKEE FOOTHILLS.  O 45.8%, D 42.8%.  6W/17H.  Virtually all of Chandler, and the Ahwatukee Foothills of Phoenix, which have some mountains in between them and the rest of the city, and therefor fit here better (also, this allows us to better follow city boundaries in this part of Maricopa.)  Lean R pretty strongly.

DISTRICT 14: TEMPE-MESA WEST.  Pop 304,888.  O 53.1%, D 51.7%.  61W/25H.  Exactly as it says on the tin.  Eh, Lean D though not by much.  

DISTRICT 15: CENTER CITY AND SOUTH PHOENIX.  Pop 305,447.  O 70.8%, D72.4%.  25W/13B/56H.  The most Democratic district by far, Safe D.

DISTRICT 16: PHOENIX NEAR NORTH-ALHAMBRA-CAMELBACK EAST.  Pop 304,661.  O 54.5%, D 54.9%.  48W/38H.  Older suburbs just north of downtown, names taken from Phoenix's "urban villages". This district is actually Hispanic-plurality by total population.  Lean D.

DISTRICT 17: PHOENIX NORTHWEST-DEER VALLEY.  Pop 306,021.  O 44.7%, D 44.0%.  71W/19H.  More burbs in the city.  Whee.  Lean R but pretty close to safe.

DISTRICT 18: PHOENIX NORTHEAST AND FAR NORTH-PARADISE VALLEY.  Pop 304,690.  O 38.5%, D 35.7%.  Again, pretty self-explanatory.  Safe R.

DISTRICT 19:  PEORIA-OUTER GLENDALE.  Pop 304,919.  O 38.8%, D 37.9%.  72W/18H.  Just those two cities (well, most of Glendale), give or take some weirdness with city boundaries etc.  Safe R.

DISTRICT 20: PHOENIX MARYVALE-INNER GLENDALE- NORTH AVONDALE.   Pop 304,977.  O 57.8%, D 60.1%.  28W/60H.  Not the least white district in the state, but yes the most heavily Hispanic.  The last district in any part of Phoenix proper.  (South Avondale was given to 6.)  Safe D.

DISTRICT 21: SUN CITY-SURPRISE-NORTH GOODYEAR.  Pop 304,641.  O 39.7%, D 37.5%.  75W/16H.  Lotsa retirees.  Of course, the same can be said for most of these districts, but Sun City makes it especially true here.  Safe R.

...

Four Hispanic-majority districts (3, 6, 15, 21), and two min-maj (1, which is also Native plurality; and 16).  

Safe D  5
Lean D  2
Tossup  1
Lean R  3
Safe R  10

[traininthedistance]

Is anybody else going to be working on Massachusetts?

I did one, but feel free to do a different one.  The more the merrier.

[traininthedistance]

Colorado in sixteen.

Although the state is 17.5% VAP Hispanic, there's really only one place where you can put a compact Hispanic-majority seat, so I made that district.  In general, I tried to follow town lines as best I could with the districts in or abutting the Front Range, but town lines are *exceptionally* ridiculous here so of course it's not perfect.  The rural areas and inner Denver-area seats basically drew themselves, but I struggled mightily with the north Range (Ft. Collins, Greeley, Boulder), and still am not completely happy with the splits necessary up there.  

The whole state:



DISTRICT 1: GRAND JUNCTION-MONTROSE.  Pop 314,734.  O 38.6%, D 37.4%.  83W/13H.  Eight whole counties along the western border of the state; Grand Junction is obviously the primary center here.  I bet Telluride feels pretty blue here, but no avoiding that.  Safe R.

DISTRICT 2: PUEBLO-SAN LUIS VALLEY-DURANGO.  Pop 315,354.  O 54.6%, D 51.9%.  64W/32H.  Pueblo County and fourteen more, mostly rural, counties in Southern Colorado.  There's a sizable Hispanic community here, but not quite enough for its own district (and I wasn't going to gerrymander south Colorado Springs and Vail in here in an attempt to get 49.9% white).  Lean D.

DISTRICT 3: ROCKY MOUNTAIN HIGH.  Pop 314,046.  O 57.2%, D 51.0%.  87W/10H.  Twelve whole counties, and rugged western portions of Boulder and Jefferson, going as far east as Golden.  Most of this district is obvious- the rest of Western Colorado, heavily mountainous and besotted with ski areas.  But I struggled mightily with the final 100K that eventually went into Boulder and Jefferson- earlier drafts split Larimer, or El Paso, or some other combination of those four counties.  Eventually I decided I valued keeping Laramie whole sufficiently to do this instead.  Lean D.

Closeup on Colorado Springs and environs:



DISTRICT 4: EL PASO WEST-CAÑON CITY.  Pop 314,868.  O 35.8%, D 31.5%.  Fremont, Teller, and parts of El Paso to the west, north, and south of Colorado Springs proper.  (Well, the fringes of CS proper are here too).  For all that Colorado Springs is a hotbed of evangelical Christianity, this district is more a hotbed of government employees in uniform, with the Air Force Academy and Cañon City's nine (9!) prisons.  Safe R.

DISTRICT 5:  COLORADO SPRINGS.  Pop 314,158.  O 44.6%, D 38.7%.  72W/15H.  Entirely within the city.  Less lopsided, but Safe R all the same.

DISTRICT 6: DOUGLAS-SOUTH JEFFERSON.  Pop 314,058.  O 41.5%, D 34.9%.  Douglas County is almost enough for a whole district, and the mostly-empty southern bit of Jeffco fills things out to make a squarish district that at first blush looks like it might bridge Denver and Colorado Springs, but is really all just south Denver exurbs.  Safe R.

Closeup on the other end of the Front Range, Ft. Collins and Greeley and Boulder.



DISTRICT 7: EASTERN PLAINS.  Pop 314,999.  O 33.7%, D 30.0%.  77W/16H.  Fifteen whole counties, and then four more are split.  The eastern half of El Paso, including largely-undeveloped parts of Colorado Springs proper, is first.  The suburban counties of Adams and Arapahoe have their rural eastern ends chopped off, in such a way as to best accommodate county lines at the other end.  (Cutting these counties is good CoI anyway.)  And most of the land area (though not most of the people) of Weld is here too.  While it does dip into the Front Range's fringes, this district should be dominated by the rural east.  Safest R in the state.

DISTRICT 8: FORT COLLINS-LOVELAND.  Pop 314,125.  O 53.5%, D 47.3%.  All of Larimer and a tiny sliver of Weld for population; I tried as best as I could to line it up with the town of Windsor, since it's partially in Larimer.  The first draft of this had a Ft. Collins-Greeley urban district and gave most of Laramie to 3, but I eventually decided that splitting Loveland (both internally and from Fort Collins) was not good.  First of many Obama-Republican districts in Colorado; in general I'm going to assume that the truth lies in the middle and go Tossup.

DISTRICT 9: GREELEY-LONGMONT-LAFAYETTE.  Pop 314,194.  O 55.7%, D 49.4%.  74W/21H.  A northern Front Range strip from Greeley (in Weld) down to Longmont, Lafayette, and Louisville- basically the Boulder County cities that are not actually Boulder.  A little ugly but ultimately has a sort of logic to it, I hope.  I really don't like 9 and 10 very much at all but everything else on the map was so obvious to draw that I felt boxed in, and the lines here at least prevent a split down the middle of Boulder itself.  Tossup.

Finally, Denver and close surroundings:



DISTRICT 10: BOULDER-BROOMFIELD-ARVADA.  Pop 314,442.  O 66.3%, D 60.8%.   Still not entirely happy with this decidedly leftovers district, which splits Boulder and Jeffco (and even cuts off the bottom of Arvada, at least it does so along a major road) but does have Broomfield all to itself, somewhat of a saving grace.  Safe D.

DISTRCT 11: INNER JEFFERSON-LAKEWOOD.  Pop 313,260.  O 54.3%, D 48.7%.  80W/15H.  For all that I'm not thrilled about the four-way split of Jefferson, I'm happy that like Adams and Arapahoe it has a coherent district all to itself, which is nice symmetry.  Tossup.  

DISTRICT 12:  ADAMS-THORNTON.  Pop 313,405.  O 56.5%, D 48.8%.  66W/27H. The all-Adams district, it makes room for the Hispanic and Aurora districts and takes what's left.  Tossup.

DISTRICT 13: ARAPAHOE-CENTENNIAL.  Pop 313,000.  O 52.2%, D 46.1%.  And it's mirror, the all-Arapahoe district to the south.  Has to take in some of Aurora, even though 14 is the Aurora district, or else split Denver four ways.  This district has the highest deviation, at -1,325; with town lines so bad, I could give up on sticking to them exactly and drive the deviation down.  If 12 was a tossup with a Democratic tilt, this is a Tossup with an R tilt.

DISTRICT 14: AURORA-AIRPORT.  Pop 315,522.  O 69.4%, D 23.5%.  45W/20B/27H.  Mostly Aurora (so Arapahoe and a bit of Adams), but has to retreat from Aurora's southern edge for 13, and also has to take in two separate portions of Denver to enable the Hispanic district.  One of two minority-majority districts, and Safe D.

DISTRICT 15: DENVER SOUTH, EAST, AND CENTER.  Pop 314,270.  O 74.5%, D 70.9%.  73W/12H.  Entirely within Denver, except for a couple Arapahoe exclaves.  Basically, the white parts of the city, with the South Platte an obvious dividing line between it and 16.  Safe D.

DISTRICT 16: DENVER WEST AND STAPLETON-ADAMS SOUTH.  Pop 314,761.  O 73.3%, D 69.7%.  40W/51H.  This is basically the only way to draw a compact Hispanic district in a state with enough Hispanics for multiple districts, so it's more or less forced and everything else is built around it.  Stapleton is that piece that divides 14's two Denver portions; it was the site of the old airport and has now been developed.  Safe D.

…

One Hispanic district (16), one min-maj (14).

Safe D: 4
Lean D: 2
Tossup: 5
Lean R: 0
Safe R: 5

There's a lot of variation in those tossups, but Colorado is sufficiently elastic that I think a mild R or D tilt isn't enough to take it out of that category.  (And it's sufficiently elastic that so many tossups is a Good Thing, too.)

[traininthedistance]

I think District 1 might actually be non-contiguous by road, seeing as the Bonneville Salt Flats don't really have any north-south connections.  It's a pretty hard barrier, in any case.  So I imagine that would get the map thrown out.

I don't think that's ever made a map get thrown out, but you're probably right that it's not a good idea. I could have simply shifted the three outer districts, but I went for some bigger revisions. This map keeps a Davis County district, three entirely in Salt Lake County, one entirely in Utah County, and a Southern Utah seat (and leaves a nice Northern Utah seat).



District 1: UTAH NORTH: McCain 70.9-25.8 Safe R.
District 2: UTAH NORTHEAST: McCain 56.2-41.2 Likely R.*
District 3: DAVIS COUNTY: McCain 69.7-27.5 Safe R.
District 4: SALT LAKE CITY: Obama 62.2-35.0 Safe D.
District 5: WEST VALLEY CITY-TAYLORSVILLE-MURRAY: Obama 48.6-48.4 Toss-up/Lean D*
District 6: WEST JORDAN-SOUTH JORDAN-RIVERTON-DRAPER McCain 63.0-34.3 Safe R.
District 7: UTAH LAKE-TOOELE: McCain 74.1-22.2 Safe R.
District 8: PROVO-OREM: McCain 77.6-19.1 Safe R.
District 9: UTAH SOUTH: McCain 72.2-24.6 Safe R.

The changes are pretty self-explanatory (the only significant changes being made to 1-2 and 6-7). CD-06 loses any chance of being competitive, but CD-02 takes that role instead. The asterisks account for the possibility of Matheson. In fact, I'd say this CD-02 (including Ogden, Eastern Salt Lake County, and Summit County) is far more likely to go his way than the previous CD-06 if that's the district he were to choose. Also, I don't think any other state can produce a more pro-Romney district than CD-08. As before, it's probably 6R-3D on a good night, but more likely 7R-2D.

(I'll admit I couldn't really think of good names for CD-02 and CD-07, as the population is somewhat more dispersed.)

Yeah, this is much better.  I like it a lot.  (It could very easily 8-1 on a bad night, mind.  Which is more than reasonable for such a lopsided state like Utah.)

[traininthedistance]

DISTRICT 1: GRAND JUNCTION-MONTROSE.  Pop 314,734.  O 38.6%, D 37.4%.  83W/13H.  Eight whole counties along the western border of the state; Grand Junction is obviously the primary center here.  I bet Telluride feels pretty blue here, but no avoiding that.  Safe R.

DISTRICT 2: PUEBLO-SAN LUIS VALLEY-DURANGO.  Pop 315,354.  O 54.6%, D 51.9%.  64W/32H.  Pueblo County and fourteen more, mostly rural, counties in Southern Colorado.  There's a sizable Hispanic community here, but not quite enough for its own district (and I wasn't going to gerrymander south Colorado Springs and Vail in here in an attempt to get 49.9% white).  Lean D.

Durango belongs on the Western Slope.    The current configuration of CO-3 is because the Western Slope didn't have enough population for a district and Pueblo was small enough to make up the difference.  As the Western Slope grew, they didn't want to split Pueblo, so they extended across in the north.  This will also keep more of the Ute reservations together.

Go ahead and extend the green district out into the Lower Arkansas.  This may even bump the Hispanic population up a point or two.

I originally tried to put Durango in the Grand Junction district, actually; this configuration of 1 and 2 was largely a function of trying to get low deviations and no county splits, especially for 1 (which is pretty constrained).  It's going to be hard to change these two.

You must be logged in to read this quote.

Laramie is in Wyoming.  Larimer is in Colorado.

Boulder wants to be its own district.  Otherwise it wouldn't have 300,000.

After adding Garfield, add Fremont.  If you just barely come into western Boulder and Jefferson counties, you aren't picking up enough people to make splitting worthwhile.  If you do come further, you're taking in Denver suburbs in a district that stretches to Utah.

Thanks for the Laramie/Larimer correction, will fix.

An earlier draft tried to send 3 south instead of cutting Boulder/Jefferson, and you need to microchop El Paso which is no good either.  As 3 stands right now, the line is just about exactly where the suburbs end and the mountains start, which seems to have a geographic logic to it, but I'll try some other things.

DISTRICT 6: DOUGLAS-SOUTH JEFFERSON.  Pop 314,058.  O 41.5%, D 34.9%.  Douglas County is almost enough for a whole district, and the mostly-empty southern bit of Jeffco fills things out to make a squarish district that at first blush looks like it might bridge Denver and Colorado Springs, but is really all just south Denver exurbs.  Safe R.

Elbert should go with Douglas.  Almost all the population is barely into the western edge of the county.

Elbert plus Douglas is still underpopulated by the standards I'm using, and all the options to fill it out are uglier than what's up there already.

I really don't think your Eastern Plains district works as well as it might appear.  You're cutting too deep into Weld, Adams, Arapahoe, Elbert, and El Paso, so you've got big chunks of farm land linking suburbs.  Better to put the northeast with Greeley and the southeast with Pueblo.

It's the reality of settlement patterns in America today that most rural districts are going to be underpopulated unless they're willing to take in some exurban fringes, and it's also the case that many of these exurbanites are living where they are because they want to feel like they're living the rural lifestyle.  And, FWLIW, the actual Colorado map does indeed split Adams and Arapahoe in this exact manner (not that I agree with everything on that map, but it's better than most this cycle).  Looking at the actual population numbers, the cuts into suburban areas of these counties (and I'm skeptical that Elbert is much of a "cut") are actually pretty shallow, population-wise, and this district would in fact stand a good chance of having real rural representation.

But I'll try a few things and maybe update this.

[traininthedistance]

This seems like a lot of unneeded county splits and it ties chunks of Albuquerque to parts way out into the rural areas. The combination of Bernalillo, Valencia, Sandoval, and Los Alamos is less than 1% over the population to make 3 districts and keeps the Albuquerque area within those same 3. The western edge plus Luna make 1 district, Dona Ana+Otero+Lincoln make 1 district, and the eastern counties of Little Texas can be formed into one district all within 1%. The last district isn't as pretty but keeps Santa Fe, Taos and Las Vegas together with no county splits.

So, there's actually a very good reason to split Bernalillo, Valencia, Sandoval, and Santa Fe: namely, to put all the reservations (besides Mescalero Apache, which is on the other side of the state) in one district- and while I do think that trying to keep counties together is, if anything, more important out West (since they are the smallest unit of government for many many people), New Mexico affords the rare possibility for a Native-plurality district, and making that with as few missing reservations as possible is, I think, an overriding concern.

I actually had a NM map I was playing with before BRTD posted his, so might as well add it.  Many of the ideas are the same but if I may be immodest I think mine are slightly cleaner and more cohesive.

The outer four districts mostly drew themselves (at least once I made the decision that uniting Native voters was a priority), the inner three are more awkward (actually, really its just 5 that's awkward).  New Mexico is one of the very few states that's easier to split with its current number of districts than this larger number (since you can easily split North/South/ABQ.)  New Mexico is Hispanic plurality, actually, so it's fitting that along with Hawaii it's one of two states where most districts are going to be min-maj.



DISTRICT 1: FARMINGTON-NAVAJO-PUEBLO.  Pop 294,191.  O 55.8%, D 50.9%.  34W/20H/44N. Roughly the northwestern quarter of the state, this district is willing to get a little ugly in order to unite the vast majority of New Mexico's reservations in one district, which is counterbalanced by the super-Republican area of Farmington.  All of San Juan, McKinley, Cibola, and Los Alamos (which is arguably a better fit with Santa Fe, but has necessary road connections for several smaller reservations).  Most of the land area of Sandoval, but none of the parts which are in Albuquerque's urbanized area.  And the most heavily-Native portions of Rio Arriba, Santa Fe, Bernalillo, Valencia, and Socorro.  Taking Isleta forces awkward constructions in the Albuquerque area, but it's the right thing to do from a CoI perspective. Probably has a D lean in Presidential years, but overall I'm going to say Tossup.

DISTRICT 2: SANTA FE-NORTHEAST.  Pop 294,292.  O 73.2%, D 66.6%.  41W/53H.  Most of Santa Fe County is here, and then it takes areas which have been Hispanic for hundreds of years.  Eight whole counties, and then three are split- Rio Arriba with 1, Curry with 4 (taking the most Hispanic precincts for VRA purposes), and Santa Fe with 1 and 5 (for purposes of connectivity and CoI; Edgewood is a world apart from Santa Fe).  Safe D.

DISTRICT 3: LAS CRUCES-SOUTHWEST.  Pop 293,192.  O 57.1%, D 51.1%.  40W/56H.  Our second Hispanic district.  Mostly keeps to county lines- Doña Ana and five others- but that would not be quite enough population.  It thus grabs  the portion of Chaparral on the other side of the county line in Otero.  (An obvious choice.)  Lean D.

DISTRICT 4: LITTLE TEXAS.  Pop 294,568.  O 35.1%, D 31.8%.  57W/37H.  Six whole counties, almost all of Otero (with 3) and half of Curry (with 2).  Safe R.

And closeup on the Albuquerque area:



DISTRICT 5: RIO RANCHO-VALENCIA-SANDIA.  Pop 295,163.  O 51.4%, D 43.2%.  57W/36H.  All of Torrance and nearly all of Socorro, but most people here are in Sandoval and Valencia, with connecting areas in Edgewood, the western edge of Bernalillo.  Some ugliness was forced here when I put Isleta in 1, preventing a South Valley/Valencia district- but this district still manages a mostly cohesive identity, as the exurbs and satellite cities of Santa Fe.  At +995 it has the highest deviation.  Lean R.

DISTRICT 6: ALBUQUERQUE OLD TOWN-EAST SIDE.  Pop 294,245.  O 61.9%, D 54.7%.  53W/35H.  Entirely within city limits and east of the Rio Grande.  Eh, close enough to call it Safe D.

DISTRICT 7: ALBUQUERQUE WEST-SOUTH VALLEY-NORTH VALLEY.  Pop 293,528.  O 64.0%, D 56.5%.  33W/59H.  Except for the heavily-Native portions of 1, all of Bernalillo west of the Rio Grande, and the North and South Valley areas make a district that is our third Hispanic-majority district and lets 6 stay entirely in the city proper.  Safe D.

Three Hispanic-majority districts and one Native-plurality district; 3 Safe D and one each of the other four ratings.

[traininthedistance]

Wisconsin in eighteen.

I split a lot more towns in Wisconsin than I'm used to, but with the exception of Milwaukee, which is too large for one district, all the town splits were along county lines- as is the case in much of the country, towns often straddle county lines here.  Within counties, towns and townships are always unsplit save for Milwaukee.  In fact, Wisconsin is probably the state I've done so far where I paid the most deference to county lines, trying to make as many whole-county groupings as I could (that didn't totally infringe upon CoI concerns).  Eventually I was able to get it up to six.  

There's a *lot* of area here with an Obama average but also a Rep average, and unlike Massachusetts the Rep average needs to be taken seriously because Obama in '08 was very much a high-water mark for the Dems here.  A lot of these areas are swingy, or more Dem for Prez than for local races, so the truth is in the middle but perhaps more towards the generic partisan average.

The whole state:



And closeup on the Milwaukee area:



DISTRICT 1: KENOSHA-RACINE.  Pop 317,475.  O 57.9%, D 48.2%.  78W/10H.  All of Kenosha and most of Racine, of course.  Um, Tossup I guess.

DISTRICT 2: WATERTOWN-BEAVER DAM-WEST MILWAUKEE COUNTRYSIDE.  Pop 314,902.  O 42.1%, D 31.8%.  Um… this is the ugly duckling leftovers district.  All of Dodge and Jefferson, in the no-man's-land between Milwaukee and Madison, the outer edge of Waukesha (whatever didn't fit in the all-Waukesha 6), the leftovers in Racine, and then some area in Walworth.  Splitting Walworth north-south would have looked nicer, except that there's a weird noncontiguous exclave in Troy Twp.  Safe R.

DISTRICT 3: MILWAUKEE SOUTH.  Pop 315,890.  O 50.6%, D 43.8%.  Milwaukee County is the perfect size for three districts, so I did that.  This is mostly the southern and near-western suburbs, but part of the city has to be taken in as well. Lean R.

DISTRICT 4: MILWAUKEE CENTRAL.  Pop 316,345. O 72.5%, D 68.4%.  56W/16B/23H.  The only district entirely within Milwaukee, since 5 needs to take a couple northern suburbs, it takes in all of the city's Hispanic community, the downtown, and the area along Lake Michigan.  Safe D.

DISTRICT 5: MILWAUKEE NORTH.  Pop 315,500.  O 80.1%, D 76.5%.  38W/54B.  The state's only possible VRA district, this black-majority Dem super-pack is unfortunately unavoidable.  Safe D.

DISTRICT 6: WAUKESHA.  Pop 315,211.  O 37.2%, D 28.9%.  Of course, to be fair, you do have a couple obvious super-Republican packs in the Milwaukee suburbs, too.  The Milwaukee area is bizarrely hyperpolarized.  Safe R.

DISTRICT 7: WASHINGTON-OZAUKEE-SHEBOYGAN.  O 40.0%, D 30.0%.  Milwaukee's northern suburbs, and most of Sheboygan County as well.  The northern suburbs ended up in the county-grouping with the two Fox River districts.  Safe R.

Closeup on Madison (and the western edge of the Milwaukee area) now.



DISTRICT 8: JANESVILLE- SOUTH CENTRAL.  O 60.8%, D 51.2%.  All of Rock and Green, most of Walworth, and the southern edge of Dane.  Along with 2, another intermediate part of the largest county-grouping I had to make (1,2,6,8,9,11).  Lean D.

DISTRICT 9:  MADISON.  O 77.4%, D 75.6%.  Madison and a few surrounding towns, entirely within Dane.  Safe D of course.

DISTRICT 10: LA CROSSE-SOUTH DRIFTLESS AREA. O 60.3%, D 48.9%.  Seven whole counties, self-explanatory.  Lean D, I guess- the numbers in DRA aren't much different than 1, but my impression is that the general political culture is.

DISTRICT 11: DELLS- NORTH DANE.  O 59.9%, D 51.3%.  Sealing off the giant south-central county grouping, this takes the rest of Dane (i.e. most of the outlying areas), and five counties to the north, centered on the resort area of Wisconsin Dells (which straddles four of them).  Lean D.

Two closeups now, on the northwest and the Fox River area:





DISTRICT 12: EAU CLAIRE- NORTH DRIFTLESS AREA.  O 57.5%, D 46.9%.  Eight counties, splits Pierce with 13.  Not strictly Driftless, but that's the basic idea here. Tossup.

DISTRICT 13: EAST MINNESOTA.  O 52.5%, D 43.9%.  Again, eight whole counties and splits Pierce.  The bulk of the population is in Twin Cities exurbs, but it has to go fairly far afield into the North Woods and Superior (another satellite of a Minnesota city) to fill out population.  Lean R.

DISTRICT 14: MANITOWOC-FOND DU LAC-FOX RIVER SOUTH.  O 48.1%, D 36.9%.  Splits Sheboygan with 7 and Winnebago with 15, and then five whole counties.  Safe R.

DISTRICT 15: FOX CITIES.  O 55.9%, D 45.7%.  All of Outagamie, and a finger down Winnebago to unite the Fox Cities (Appleton, Oshkosh, etc.)  I tried to get the portion of Appleton in Calumet County here too, but it didn't work without splitting more things elsewhere, alas.  Tossup.

DISTRICT 16: GREEN BAY-DOOR.  Pop 314,239.  O 54.4%, D 43.3%.  All of Brown, the Door Peninusla, and splits Oconto with 18.  Eh, Lean R.

DISTRICT 17: WAUSAU-STEVENS POINT.  Pop 316,577.  O 55.9%, D 45.1%.  Marathon, Portage, Wood, and the southern half of Waupaca.  Basically the last sizable urbanized areas in the northern reaches of the Wisconsin River.  Tossup.

DISTRICT 18: NORTH WOODS AND LAKES.  Pop 314,191.  O 53.8%, D 43.8%.  Splits the aforementioned Oconto and Waupaca with 16 and 17, and fourteen whole counties.  Thoroughly rural, includes many of the state's Native population, has to split its region with 13.  Lean R.

…

One VRA district, 5.

3 Safe D
3 Lean D
4 Tossup
4 Lean R
4 Safe R

Either the state is a natural R gerrymander or my race ratings are too conservative... really it's probably a bit of both.

[traininthedistance]

Nebraska in six.

With a zillion counties and only six districts, it's possible to make something Iowa-style that keeps the deviation low and doesn't split any counties, aside from Omaha's Douglas (which is too large for a district).  

The state as a whole:



And closeup on Omaha:



DISTRICT 1: OMAHA CITY.  Pop 304,828.  O 60.1%, D 53.8%.  68W/15B/12H.  The lines are drawn here so as to ensure every precinct is within city limits, and that the rest of Douglas can be within 2.  The only district with a non-negligible percentage of minorities.  As close to a safely Democratic district as you'll get in Nebraska.

DISTRICT 2: OUTER DOUGLAS-FREMONT-SOUTH SIOUX CITY.  Pop 304,011.  O 42.0%, D 36.9%.  Keeping to whole counties means that in addition to the inner Omaha district, the rest of the metro area is going to have be split among two districts that take in rural areas.  This one goes north along the Missouri River, taking in the rest of Douglas and five further counties; very little of the population is actually rural, but most is in Douglas and most of the rest is in Fremont, Blair, and the Sioux City area.  Safe R.

DISTRICT 3: SARPY-SOUTHEAST.  Pop 304,433.  O 40.0%, D 37.6%.  Sarpy and twelve further counties in the southeast corner of the state, wrapping around the Lincoln district.  This one is more rural, but even so roughly half the population is in Sarpy.  Safe R.

DISTRICT 4: LINCOLN AREA.  Pop 305,497.  O 51.3%, D 45.2%.  Lancaster, Saline, and Fillmore counties.  The Lincoln area is the hardest to do in whole counties since Lancaster is pretty close to a full district, and this version gives us the lowest possible deviation (1,107).  Obama won it, but Lean R anyway.

DISTRICT 5: TRI-CITIES-COLUMBUS-NORFOLK.  Pop 303,733. O 31.4%, D 33.2%.  Twenty counties, the main population centers (such as they are) all in the district name (Tri-Cities being Kearny, Grand Island, and Hastings of course).  Safe R.

DISTRICT 6: PRAIRIE.  Pop 303,799.  O 27.0%, D 30.8%.  I'm not going to count how many counties are in this giant expanse of emptiness.  Safe R.