Congressional Districts - 2016 ACS estimates.

[jimrtex]

These are the Census estimated population based on 2016, 1 year estimates. These are the sum of 12 monthly surveys, and thus correspond roughly to July 2016, assuming linear change throughout a year.

The largest districts (10% greater than national averages of 741K)

AL Montana         1043  Montana
AL Delaware         952  Delaware
22 Texas            881  Houston SW suburbs.
AL South Dakota     865  South Dakota
1 Idaho             865  Boise to Panhandle
12 North Carolina   849  Charlotte
4 North Carolina    847  Raleigh
26 Texas            845  Denton
3 Texas             843  Collin
3 Oregon            835  Portland
9 Florida           833  Orlando, Kissimmee
31 Texas            831  Williamson+Bell
1 Oregon            831  Portland Metro West
10 Virginia         827  Washington Metro West
3 Iowa              824  Des Moines, Southwest Iowa
10 Texas            823  Austin to NW Harris
35 Texas            823  Austin to San Antonio (I-35)
5 Oregon            822  Salem, Portland Metro South, Northern Coast
2 North Carolina    820  Raleigh Metro
7 Arizona           818  Phoenix
2 Idaho             818  Boise to East
5 Oklahoma          816  OKC, Oklahoma Pottawatomie, Seminole counties

[jimrtex]

What about the smallest?

These districts have less than 90% of the ideal population.

7 Alabama           666  Black Belt - Birmingham
2 New Hampshire     663  West, Connecticut Valley, Nashua, Concord
8 Minnesota         662  Northeast
2 Nebraska          661  Omaha
13 Michigan         658  Detroit, West and North Wayne
2 Maine             655  North
1 Nebraska          639  East, Lincoln
AL Vermont          625  Vermont
2 West Virginia     623  Middle, Charleston
1 West Virginia     617  Northern, Wheeling
3 Nebraska          607  West, Central
3 West Virginia     591  South, Huntington
AL Wyoming          586  Wyoming
1 Rhode Island      537  East
2 Rhode Island      520  West

[jimrtex]

These are projected to 2020 assuming a linear (not exponential) increase. For the 2010 base, I used the average population for the state, this will only make a trifling difference for West Virginia and Arkansas.  Districts 10% greater than ideal of 761K.

AL Montana         1074  Montana
22 Texas            991  Houston SW suburbs.
AL Delaware         985  Delaware
26 Texas            934  Denton
3 Texas             929  Collins
12 North Carolina   919  Charlotte
4 North Carolina    915  Raleigh
9 Florida           915  Orlando, Kississimee
1 Idaho             913  Boise to Panhandle
31 Texas            910  Williamson, Bell
10 Texas            898  Austin to NW Harris
35 Texas            897  Austin to San Antonio I-35
AL South Dakota     896  South Dakota
10 Virginia         887  Washington Metro West
7 Arizona           883  Phoenix
8 Texas             883  Montgomery and Northward
3 Oregon            876  Portland
2 North Carolina    872  Raleigh Metro
6 Colorado          871  Aurora, Centennial, Highland Ranch, Brighton
13 New York         870  Manhattan (Harlem)
1 Oregon            870  Portland Metro West
1 Colorado          869  Denver
21 Texas            868  Northern Bexar to Travis, Hill Country
16 Florida          868  Southern Hillsborough, Manatee (Bradenton), Sarasota
3 Iowa              862  Des Moines, Southwest Iowa
19 Florida          861  Lee, Collier (Cape Coral, Ft. Myers, Naples)
5 Oregon            856  Salem, Portland Metro South, Northern Coast
5 Oklahoma          855  OKC, Oklahoma, Pottawatomie, Seminole Counties
2 Colorado          854  Larimer, Boulder, North Jeffco
4 Utah              854  South Valley (West Jordan to Logan)
20 Florida          854  Palm Beach, Broward, Three Fingers
9 Texas             851  Southwest Houston
7 Georgia           849  Northeast Atlanta Metro (Norcross)
10 Florida          849  Western Orange (Orlando)
11 Virginia         848  Southwestern Washington Metro
42 California       847  Southwestern Riverside (Corona, Lake Elsinore, Murietta)
20 Texas            847  Central San Antonio
24 Texas            845  Northern Tarrant
26 Florida          844  Southern Miami-Dade, Florida Keys
3 Nevada            843  Southern Clark, Henderson
4 Colorado          842  Weld, Douglas, Eastern Plains
18 Texas            841  South, Northwest, Northeast Houston
5 New York          840  Queens (Jamaica)
2 Idaho             839  Idaho Boise to East

[jimrtex]

2020 Projections for districts 90% or less of national average.

4 Illinois          684  Chicago - Ear Muffs
2 Mississippi       684  West - Delta, Jackson
4 Alabama           684  North Central
1 Maine             683  South
12 Illinois         682  Southwest - St.Louis Metro East
2 Alabama           681  Southeast
9 Pennsylvania      680  Southwest
1 New Hampshire     680  East - Manchester, Portsmouth
10 Pennsylvania     679  North Central
1 Illinois          679  Chicago
17 Illinois         678  Northwest - Quad Cities
1 Minnesota         678  South
14 New York         676  Queens (Astoria) - Bronx (Southeast)
11 Ohio             675  Cleveland, Akron
16 Illinois         673  Chicagoland Fringe - Rockford
7 Minnesota         671  West
5 Michigan          666  Flint, Tri-Cities, Saginaw Bay
2 New Hampshire     666  West, Connecticut Valley, Concord, Nashua
8 Minnesota         662  Northeast
2 Georgia           660  Southwest, Albany
1 Nebraska          657  East, Lincoln
7 Alabama           655  Black Belt - Birmingham
2 Maine             650  North
13 Michigan         630  Detroit, North and West Wayne
2 West Virginia     626  Middle, Charleston
AL Vermont          624  Vermont
1 West Virginia     617  North, Wheeling
3 Nebraska          606  West, Central
AL Wyoming          599  Wyoming
3 West Virginia     575  South, Huntington
1 Rhode Island      543  East
2 Rhode Island      516  West

Noteworthy: NE-2 (Omaha) is on 2016 list, but not 2020. Nebraska is close to losing a third seat, and so dividing in three equal parts makes all three undersized, but Omaha is increasing faster than USA as whole. In fact, Omaha growth is causing state to grow faster (barely) than USA, and maintaining third seat.

New York (City) has both large districts (NY-13 and NY-5) and small (NY-14). NY-13 and NY-14 are adjacent.

Georgia also has large and small districts (GA-7 and GA-2).

[jimrtex]

These are the 2020 projections relative to the intrastate quota:

Maine

1 Maine             1.025  South
2 Maine             0.975  North

To equalize populations, about 17K persons will be shifted from South to North.

New Hampshire

1 New Hampshire     1.010  East - Manchester, Portsmouth
2 New Hampshire     0.990  West - Connecticut Valley, Concord, Nashua

To equalize population, about 7K persons need to be shifted.

Vermont

AL Vermont          1.000  Vermont

After 2020, VT-AL will be the second smallest district, after West Virginia and Rhode Island lose a district, and Nebraska rebalances its districts. Northern New England will be home to five of the six smallest districts in the country.

Massachusetts

1 Massachusetts     0.943  Berkshires, Connecticut Valley - Springfield
2 Massachusetts     0.984  Worcester - Worcester
3 Massachusetts     1.010  Far North Boston Metro - Fitchburg, Lowell, Lawrence
4 Massachusetts     0.992  West Boston Metro, Bristol - Taunton, Fall River
5 Massachusetts     1.005  Near North Boston Metro - Framingham
6 Massachusetts     1.020  Northeast Boston Metro - North Shore, Essex, Cape Anne
7 Massachusetts     1.046  Boston
8 Massachusetts     1.035  South Boston Metro - Taunton
9 Massachusetts     0.965  Plymouth, Cape Cod, Islands, Bristol - New Bedford

50K need to be shifted to MA-1 and MA-2, and 27K to MA-9. To equalize all districts requires a minimum of only 179K in shifts (about 2.6% of the state population), though I would go a bit farther. MA-1, MA-2, and MA-9 can be kept out of the Boston area, shifting MA-9 to include Fall River.

MA-4 can be pushed further east and MA-8 a bit south. The reduction to 9 districts greatly improved the Massachusetts map, and this would further undo the map that was drawn to gerrymander Margaret Heckler out of a district 40 years ago.

Rhode Island

1 Rhode Island      1.026  East
2 Rhode Island      0.974  West

About 14,000 persons would need to be shifted to equalize the districts, but instead they will be combined into a single at-large seat. Rhode Island will go from having the two smallest districts, to the second largest after Montana, and there is a possibility of Montana of gaining a seat.

Connecticut

1 Connecticut       0.994  North Central - Hartford
2 Connecticut       0.977  East - New London, Norwich
3 Connecticut       1.004  South Central  - New Haven
4 Connecticut       1.038  Southwest - Bridgeport, Norwalk, Stamford
5 Connecticut       0.987  Northwest - Waterbury, Torrington, New Britain, Danbury

To equalize all districts requires a shift of about 48K persons (1.3% of total population). It will be increasingly difficult to keep CT-2 from encroaching on Hartford and New Haven. There could also be some swaps to make the boundary between CT-1 and CT-5 more regular.

[jimrtex]

Feel free to suggest alternative names.

New York

1 New York          0.972  Suffolk Central, East
2 New York          0.963  Suffolk-Nassau, South Shore
3 New York          1.003  Nassau-Suffolk, North Shore
4 New York          0.983  Nassau, South Shore
5 New York          1.136  Queens, Jamaica
6 New York          1.049  Queens, Flushing, Forest Hill
7 New York          1.070  Brooklyn, Williamsburg, Bushwick, Sunset Park
8 New York          1.087  Brooklyn, Bedford-Stuyvesant, East Flatbush, East New York
9 New York          1.016  Brooklyn, Kensington, Park Slope
10 New York         0.952  Manhattan-Brooklyn (Ugh), West Side, Borough Park
11 New York         1.009  Staten Island-Brooklyn, Staten Island, Bensonhurst
12 New York         0.966  Manhattan-Queens, East Side
13 New York         1.176  Manhattan, Harlem
14 New York         0.914  Queens-Bronx, Astoria
15 New York         1.081  Bronx, South and West
16 New York         1.012  Westchester-Bronx, Yonkers, Rye, North Bronx
17 New York         1.028  Westchester-Rockland, White Plains, Rockland
18 New York         0.981  Exurban New York, Newburgh, Poughkeepsie, Middletown
19 New York         0.928  Mid Hudson Valley
20 New York         0.983  Albany, Schenectady, Troy
21 New York         0.958  North Country, Plattsburgh, Jamestown
22 New York         0.940  Near West New York, Utica, Binghamton
23 New York         0.940  Southern Tier, Ithaca, Elmira, Jamestown
24 New York         0.948  Syracuse, Oswego
25 New York         0.978  Rochester
26 New York         0.952  Western New York, Buffalo Metro
27 New York         0.976  Buffalo

The ten districts in Upstate New York (NY-18 to NY-27) have a population equivalent to 9.541 districts, which would result in considerable sliding of districts to the east and south, as NY-18 would need to slide southward by about 308K persons, and the other districts in decreasing amounts.

The thirteen mostly insular districts (NY-1 to NY-12, and NY-14) have a population equivalent to 13.118 districts, which permits the districts to be largely maintained with source districts in Brooklyn and Queens (NY-5 to NY-8) flowing into districts fairly nearby (NY-1, NY-2, NY-4, NY-10, NY-12, and NY-14).

Four districts in between (NY-13 to NY-15, and NY-17) have a population equivalent to 4.298 districts, which provides most of the population that must flow upstate.

The minimum shift needed to equalize districts is 2.268M (about 11.8% of the state population). But this all presumes that New York will not lose a congressional district. This is increasingly unlikely as movement into city centers abates.

The following are relative to 26 districts (a loss of one).

1 New York          0.936
2 New York          0.927
3 New York          0.966
4 New York          0.946
5 New York          1.094
6 New York          1.010
7 New York          1.030
8 New York          1.047
9 New York          0.978
10 New York         0.916
11 New York         0.972
12 New York         0.930
13 New York         1.133
14 New York         0.880
15 New York         1.041
16 New York         0.974
17 New York         0.990
18 New York         0.945
19 New York         0.894
20 New York         0.947
21 New York         0.922
22 New York         0.905
23 New York         0.905
24 New York         0.913
25 New York         0.941
26 New York         0.917
27 New York         0.940

The twelve upstate districts (NY-17 to NY-28) have a population equivalent to 11.194 districts (based on 26 total), so that one district can be eliminated, and produce a modest surplus of about 148K to be shifted south.

Since all the districts are underpopulated, a least shift strategy would eliminate one in a more central location, and dividing its population outward among adjacent districts. A large multi-county district is preferred, since it can be divided, without cutting into a larger metropolitan area. NY-19, NY-21, and NY-22 are possibilities. NY-22 may be the best choice since it can be divided among NY-23 and NY-24, with further distributions westward, NY-21, and NY-19 with further distributions eastward. In addition, Utica and Binghamton are not large metropolitan areas.

NY-22 may be distributed among: NY-24 (and NY-25) 118K; NY-23 (and NY-26, NY-27) 177K; NY-21 (and NY-20) 101K; NY-19 (and points south) 300K. Because of the large share of the population shifted to NY-19, and alternate interpretation is that the two districts are merged, and the excess population distributed. If the intent was to pair incumbents, this might be more equitable.

The seven districts in Manhattan, Brooklyn, and Staten Island (NY-7 to NY-13) have a population equivalent to 7.006 districts, which permits shifts within the area from the districts with a surplus (NY-7, NY-8, and  NY-13) to those with a deficit (NY-9 to NY-12). Only 178K (3.3%) need to be shifted to equalize the seven districts.

The four Long Island districts in Nassau and Suffolk have a population equivalent to 3.775 districts, or around 173K persons, the surplus from upstate (after the loss of a seat), NY-15, and NY-5 can be used to fill in these four districts and NY-14. NY-14 which currently has about 2/3 in Queens will become a majority Bronx district.

To equalize among 26 districts requires shifting 1.908M persons, about 9.6% of the state population. This counts the population that remains in the merged NY-19/22 as not being shifted.

[jimrtex]

New York

1 New York          0.972
2 New York          0.963
3 New York          1.003
4 New York          0.983
5 New York          1.136
6 New York          1.049
7 New York          1.070
8 New York          1.087
9 New York          1.016
10 New York         0.952
11 New York         1.009
12 New York         0.966
13 New York         1.176
14 New York         0.914
15 New York         1.081
16 New York         1.012
17 New York         1.028
18 New York         0.981
19 New York         0.928
20 New York         0.983
21 New York         0.958
22 New York         0.940
23 New York         0.940
24 New York         0.948
25 New York         0.978
26 New York         0.952
27 New York         0.976

With these numbers, its obvious Upstate will have to take the hit. The obvious district to cut is the 22nd, since it lacks a geographic base and is instead squished between those that do. It getting cut however probaby means a general resuffling upstate that will lock down most seats: Ithica to the 24th, Binghamton to the 23th, Rome/utica to 21st, etc.

An alternative interpretation is that NY-19 and NY-22 are merged, with the collective surplus shifted westward and southward. Kenney (New Hartford, near Utica) and Faso (Kinderhook) would be paired in a district that wraps around the southern edge of Albany.

See modified message, above.

[jimrtex]

New Jersey

1 New Jersey        0.971
2 New Jersey        0.957
3 New Jersey        0.988
4 New Jersey        0.972
5 New Jersey        0.995
6 New Jersey        1.011
7 New Jersey        0.972
8 New Jersey        1.070
9 New Jersey        1.031
10 New Jersey       1.027
11 New Jersey       0.975
12 New Jersey       1.031

New Jersey districts don't have much deviation, with only two outside 5% deviation, but the smaller districts are concentrated in the south (NJ-1 to NJ-4) or in the western suburbs (NY-5, NY-7, and NY-11). The most populous district is NY-8, opposite New York City.

To equalize population, 371K persons (4.1% of the state need to be shifted). This is about three times that which would be needed if the adjustments could be made between adjacent district.

[jimrtex]

With these numbers, its obvious Upstate will have to take the hit. The obvious district to cut is the 22nd, since it lacks a geographic base and is instead squished between those that do. It getting cut however probaby means a general resuffling upstate that will lock down most seats: Ithica to the 24th, Binghamton to the 23th, Rome/utica to 21st, etc.

The two questions I would have is 1.  Is it possible to make a realistic district that combines Ithaca and Syracuse?

2.  Is NY-26 going to expand into the suburbs enough to make it a competitive district?

If Democrats control redistricting they would probably make #1 happen somehow, but #2 would probably be unavoidable.

1. Not possible. NY-23 will have to already take Broome (Binghamton) and losing Tompkins would force NY-23 into Delaware, or wrapping around Tompkins. Simpler alternatives, remainder of Oswego plus Cortland exist for NY-24.

2. No. It only needs to add about 60K. Even if this were 100%R it wouldn't flip.

3. New York has a redistricting commission (approved in 2014). The legislature has to approve the plan, and could conceivably sabotage the effort - but this might be challenged successfully in state court (favoring a party or a candidate).

[jimrtex]

Feel free to suggest alternative names.

New York

1 New York          0.972  Suffolk Central, East
2 New York          0.963  Suffolk-Nassau, South Shore
3 New York          1.003  Nassau-Suffolk, North Shore
4 New York          0.983  Nassau, South Shore
5 New York          1.136  Queens, Jamaica
6 New York          1.049  Queens, Flushing, Forest Hill
7 New York          1.070  Brooklyn, Williamsburg, Bushwick, Sunset Park
8 New York          1.087  Brooklyn, Bedford-Stuyvesant, East Flatbush, East New York
9 New York          1.016  Brooklyn, Kensington, Park Slope
10 New York         0.952  Manhattan-Brooklyn (Ugh), West Side, Borough Park
11 New York         1.009  Staten Island-Brooklyn, Staten Island, Bensonhurst
12 New York         0.966  Manhattan-Queens, East Side
13 New York         1.176  Manhattan, Harlem
14 New York         0.914  Queens-Bronx, Astoria
15 New York         1.081  Bronx, South and West
16 New York         1.012  Westchester-Bronx, Yonkers, Rye, North Bronx
17 New York         1.028  Westchester-Rockland, White Plains, Rockland
18 New York         0.981  Exurban New York, Newburgh, Poughkeepsie, Middletown
19 New York         0.928  Mid Hudson Valley
20 New York         0.983  Albany, Schenectady, Troy
21 New York         0.958  North Country, Plattsburgh, Jamestown
22 New York         0.940  Near West New York, Utica, Binghamton
23 New York         0.940  Southern Tier, Ithaca, Elmira, Jamestown
24 New York         0.948  Syracuse, Oswego
25 New York         0.978  Rochester
26 New York         0.952  Western New York, Buffalo Metro
27 New York         0.976  Buffalo

These estimates seem like they may not be great for estimating within counties. NY-10 and NY-12 contain some really fast-growing areas, like the Financial District/Tribeca and Hell's Kitchen/Chelsea (NY-10) and Long Island City/Astoria and Williamsburg/Greenpoint (NY-12) that are among the fastest-growing areas in all of NYC (only other genuinely fast-growing areas are the South Bronx (NY-15), Flushing (NY-6) and Downtown Brooklyn (mostly in NY-8)). It seems deeply unlikely to me that they are growing slower than the state as a whole.

The American Community Survey is a monthly census conducted on a sample of households. Housing units are assigned to one of five sample frames, sorting based on census tract, block number, and street address, so essentially every 5th address (including apartment and unit numbers) is in the same sample frame. These five sample frames are used in a five-year cycle, to avoid surveying the same household unit more than once in a five-year period.

Each year a sample of housing units is selected, again using geographical order. In NYC it appears that about 1/100 housing units are sampled each year. Higher sampling rates are used in areas that have less population. The ACS goal is to produce statistically significant data for counties, places with organized governments, school districts, and minor civil divisions in 12 states (ME, NH, VT, MA, RI, CT, NY, NJ, PA, MI, WI, and MN), plus census tracts, and block groups. If a town has 100 households, you need a relatively large sample. If you only did 1/100, you might get the family with 8 kids, a grandmother, son-in-law, and three grandchildren; or the widow. Neither is representative.

But for NYC, the only governmental units are the city as a whole, and the five boroughs (counties), so sampling is based on the census tract sizes. In NY-21 (North Country) there were roughly twice as many households sampled because of all the microscopic rural towns.

NYC said that Tribeca had 18,000 persons in 2010. If we guesstimate this represents about 9,000 housing units, then the census uses about 90 housing units for the 2016 ACS. In any case, the estimate is not based on a subestimate of county populations, but directly from a census sample from the area.

Manhattan Community Board 1 Population Change Update (PDF)

This gives a population for the entire area (Tribeca, Battery Park City, Financial District, and Seaport/Civic Center as 61K, which is less than 10% of the NY-10. Even if that area has increased by 20% since then, that would be only a 1.7% gain for the district as a whole.

Your perception of growth in the area might be based on past growth. For example, in 1970, TriBeCa had a population of 243, when it was entirely commercial. But remember that the 18K in Tribeca in 2010, including the 6K growth between 2000 and 2010 was already included in the population base for NY-10.

There are three ways to increase population:
(1) Add housing units;
(2) Increasing occupancy rates;
(3) Increase household size.

The areas are so expensive and in demand that occupancy is probably close to 100%. People who can afford to live there, might not be able to afford children. So the only way left is to increase housing units. Are there more buildings left for conversion to lofts in Tribeca? The growth in the Financial District (245% between 2000 and 2010) appears to be more the result of new construction. But with the crash in 2008, development of new buildings may have stopped. Those under construction at the time might have been completed, but perhaps after a pause as developers worked out their financing, or realized they might not be able to rent apartments immediately. Building must be a multi-year project, replatting, design, deep foundations, etc. With banks less willing to lend, it may only in the last couple of years that new buildings have begun to come on line.

There are ACS estimates at the Census Tract level, but to be statistically significant, they need accumulate data over a 5-year period. The data for 2012-2016 has not been released yet, so the latest available is for 2011-2015, which would on average be representative of mid-year 2013.

[jimrtex]

Armory Houghton is still alive at age 91, so I guess carving up NY-23 is still forbidden.

In 2020, the area between Lake Ontario will have just enough population for five districts. The cities are in the northern area, near the Erie Canal. With two districts for Buffalo, and one each for Rochester, and Syracuse, that leaves enough for another district, though it will experience a eastward shift as it takes Binghamton as a new largest city, replacing Elmira, Ithaca, Jamestown, and Corning.

By 2040 or 2050, there will be four districts, and the southern area will be divided among the city-based districts.

p.s. I thought you might have misspelled Amo's first name, but double-checked with a Google Search.

The first entry was for a Armory Houghton, and I thought you had made a mistake about his being alive, but it turned out that that Armory Houghton had died in the 19th Century.

There was an entry about the national guard armory in Houghton, Michigan.

And a FOIA release from the CIA that was written to Amo's dad Amory Houghton, Sr. from 1959. It was short note from SOS Allen Dulles to Houghton, who was ambassador to France at the time. Inexplicably, the transcription from a carbon copy of a typewritten dispatch said it was from Alien Dulles. I knew about Kissinger and Albright, but had not known about Dulles.

Corning Glass was founded by Amo's g-g-grandfather, Amory Houghton, Sr. There have been four Amory Houghton's over 5 generations. With Amo's grandfather, Alanson Houghton being the exception. Alanson was however a US representative for two terms (1919-1922) before becoming Ambassador to Germany, and then Great Britain.

[jimrtex]

Pennsylvania

1 Pennsylvania      0.969  SE
2 Pennsylvania      0.947  SE
3 Pennsylvania      0.910  W
4 Pennsylvania      0.979  SE
5 Pennsylvania      0.924  W
6 Pennsylvania      1.002  SE
7 Pennsylvania      0.949  SE
8 Pennsylvania      0.943  SE
9 Pennsylvania      0.901  W
10 Pennsylvania     0.900  NE
11 Pennsylvania     0.938  NE
12 Pennsylvania     0.918  W
13 Pennsylvania     0.990  SE
14 Pennsylvania     0.915  W
15 Pennsylvania     0.977  SE
16 Pennsylvania     0.978  SE
17 Pennsylvania     0.923  NE
18 Pennsylvania     0.938  W

Population shares are based on 17 districts (loss of one district).

The southeast region (Philadelphia metro, plus York, Lancaster, Reading, Allentown, and Bethlehem) has nine districts, and is entitled to 8.734 districts, or a deficit of about 201K, which it could easily pick up in the Harrisburg or Allentown areas. The western half of the state has six districts, and is entitled to 5.506 districts. The northeast has three districts and is entitled to to 2.775 districts.

While PA-7 deserves to die, there is no apparent way to do this in a least change regime. PA-11 is about the median for the lost district, but it is not simple to shift population to the west and south from the western end of the district. So instead PA-10 and PA-11 are merged. Collectively, the counties from Harrisburg north are sufficient to shift to PA-9, and on to the three Pittsburgh area districts (PA-12, PA-14, PA-18). The population to shift to the southeast can be from Dauphin into PA-15. None of this improves the shape of the districts.

Overall, 1.280M or 10% of the state's population would be assigned to new districts. This does not count the area in the northeast corner that is assigned to the remnant PA-10/PA-11 district.

[jimrtex]

An alternate approach is to assign whole counties to the district which contains the largest share.



Population shares are based on 2020 projections with 17 districts. Since I am using whole counties, I used the 2016 Census estimates, projected linearly to 2020, rather than the ACS.

This results in the elimination of PA-2, and making PA-1 (Philadelphia) and PA-14 (Allegheny) two-member districts. Four counties, Beaver, Indiana, Fayette, and Lebanon are isolated from the main part of their current districts. Rather than trying to reconnect these counties, which is due to connections by narrow corridors they could be considered proto-districts.

There are 23 protodistricts (18 + 4 isolated counties + 2 two-member districts - 1 eliminated district (PA-2).

We first eliminate the four isolated county districts, by assigning them to an adjacent district to better balance population.



This reduces to 19 proto-districts, two more than our goal. We eliminate the smallest district, PA-9, distributing its population among its neighbors to produce population balance, and maintain general compactness.



We still need to eliminate one more district. PA-7 is the smallest, but it consists of Delaware, and can't be divided. The next smallest is PA-10, which is distributed among its neighbors.



Finally we shift some counties to balance the population of the districts. This might have served as a plausible map prior to Wesberry v Sanders

[jimrtex]

Why on earth have I thought all this time that Berks was in the UCC? 

Because there is a small yet significant exurban region in the south of the country - in don't know.

Reading forms a MSA separate from Philly and consists only of Berks. It has enough urbanized area population to be a single county UCC. Berks is part of the larger Philadelphia CSA.

Pottstown is a Urbanized Area (more than 50,000) persons that was delineated prior to 2000 (back when the Census Bureau constructed Urbanized Areas from large places and adjacent territories). These have been grandfathered in under the new definition. When urbanized areas ran together, the Census Bureau divided them, either at an isthmus, if they just touch, or at a county line if they run together. An urbanized area can still cross into another county as long as it avoids contact. For example, Philadelphia UA goes into New Jersey northwest of Trenton (in Hunterdon). Philadelphia UA reaches the Berks and Lehigh county border, but would be blocked from much further advancement by Allentown and Reading UAs. The Allentown UA also extends into northern Bucks. The Pottstown UA does extend into Berks.

Urbanized areas can absorb Urban Clusters (urban areas < 50,000 person) so it may depend on whether areas can be established separate from the main area before they get overrun.

There wasn't a separate Wilmington, DE urbanized area so, the Philadelhia, UA (and MSA) now extends into Maryland. Philadelphia is now blocked by Lancaster, Baltimore, Dover, and Vineland UA. Proto-CBSA (core-based statistical areas) are defined by the dominant urban area in each county.

So Montco is part of the Philadelphia MSA, while there is no MSA associated with Pottstown. It is possible for a proto-CBSA to be captured by another MSA by commuting patterns. Reading is far enough out to have its own job centers, plus the ordinary local workers (grocery stores, schools, doctors, etc.) plus some of the commuting will be from Berks into Lehigh, which reduces the percentage that could commute into the Philadephia Proto-MSA whether Montco, Bucks, and Chester or Philadelphia.

[jimrtex]

Has anyone in the Wisconsin case via an amicus brief or otherwise pointed out that the efficiency metric pushed by the Dems takes no cognizance of chops? In other words, any metric which focuses on chops and erosity (including - to deal with the Michigan law fail - intra-county erosity where a county is chopped), should not be characterized as an illicit gerrymander even if the result is "inefficient?"

The plaintiffs in Wisconsin had an automated program that claimed it took counties and municipalities into account in drawing boundaries (they generated 2000 maps), but their metric appears to be number of small counties that were chopped. Since Milwaukee can't fit in a single district, it doesn't matter how many times you cross the border.

And even then they did a mediocre job, with around 20 small counties kept whole (vs. my 40+). But they did do more than the legislature did.

A telling example of how bad there program was, was that they did not have one map that kept Columbia as a single district (0 for 2000), when they should have gone 2000 for 2000.

[jimrtex]

The Census Bureau has recalculated the 2010 Census results for the 115th Congress (2017-2019) for Virgnia, Florida, North Carolina and Minnesota.

Did something not-so-nice happen in Minnesota?

[jimrtex]

The Census Bureau has recalculated the 2010 Census results for the 115th Congress (2017-2019) for Virginia, Florida, North Carolina and Minnesota.

Did something not-so-nice happen in Minnesota?

What did they find?

I'm not sure what you mean by "they" and "find".

Anyhow, I compared the 113th data and the 115th data, and the 115th data moved 3 people from MN-6 to MN-7 (the CD populations are no longer within 1, with MN-7 and MN-6 now having a deviation of +/- 3.

The change occurred in Munson Township of Stearns County, which is the only subdivision in the county divided between CD's. The city of Richmond is mostly in the PLSS township for Munson, and in MN-6,

The area that was moved is on the western edge of Richmond, and the only way to get to it from the remainder of MN-6 in Munson, is through Richmond. It appears that the census blocks don't actually conform to the city limits. When the map drawers were dividing Munson, it would be natural to put that area in MN-7. It is likely that the area with the three persons was actually in Richmond, but not recognized by the Census Bureau. This would then have required Richmond to be divided between the two CD's, with 3 persons being in MN-7.



For the 115th Congress, MN-6 is the dark green portion of Munson Township, along with the city of Richmond, and townships to the northeast, east, southeast, and south. MN-7 is the pale yellow portion of Munson Township, along with townships to the north, northwest, west, and southwest, and the city of Roscoe.

The triangular exclave in the southeastern part of Richmond was in MN-6 for the 113th Congress, and is simply an unincorporated island in the city. The area that was added for the 115th Congress is the dark green blob on the northwestern edge of Richmond.

The current congressional districts were drawn by a court, and don't appear to have been enacted into statute. It is conceivable that everyone simply assumed that all of Richmond was in MN-6 (in 2012 and 2014 and 2016 it had one precinct and was in MN-6. Perhaps some looked at the actual map, and if the court retained jurisdiction changed the map, and the Census Bureau was informed.