Using Urban County Clusters To Guide Redistricting

[jimrtex]

To extend the example of the pastures, suppose there is a deep ravine cutting across part of the pastureland that the cattle cannot cross. One could put up a fence, but there is no reason to do so since the cattle will not cross.

One significant difference in my proxy to a direct perimeter calculation is that I don't charge for unneeded fencing. For example, In AL there is no crossing of the Alabama river between Autauga and Lowndes. I assign no erosity when that segment is on the border. This has the effect of recognizing that significant natural water and mountain barriers are acceptable as borders even though they might elongate a district. It also solves the issue of when a body of water gets large enough to become a barrier. For instance, how does one distinguish between the Alabama river and Chesapeake bay in MD when state county boundaries extend to the middle of both bodies of water. A connection-based decision resolves that.

Extending my analogy (for the 9th grade):

Certain pastures have only a limited connection to their neighbors, they connect at a point or nearly do so.  Herds can not easily pass through the narrow opening.  You may not give the Lawrence and Cullman pastures to the same child, unless you also include Morgan or Winston.
The same restriction applies between Bibb and Jefferson, and Baldwin and Washington.  Note that you must still provide fencing across these narrow gaps.

In other locations it might appear that the connection is too narrow, but well established pathways permit easy movement of cattle.  These include Greene-Marengo and Marengo-Dallas.

In some other states there are significant bodies of water, which the cattle can not swim across.   In places there are bridges or cow ferries which can be used to transport cattle between pastures.  You may use these ferries or bridges to connect the pastures of one of the children.  If the body of water is used as a boundary between pastures, you don't need to provide fencing.

Extra credit: Discuss the role of navigable rivers of Alabama as highways rather than barriers.

Note: Autauga and Lowndes were in a congressional district which included neither Montgomery or Dallas County.

In certain other states mountains prevent the cattle from moving from one pasture to another, and there are not even roads connecting the pastures.  These pastures may not be connected, and no fencing is required.

Extra credit: What is the highest point in Alabama.

[muon2]

County borders are simple to measure.



Highway connectivity is harder to measure, and reduces connection to a true/false value.  The Census Bureau does not connect Randolph to Cleburne by highway.  St.Clair County, the green county to the west has two county seats.  And a road that nicks the corner of a county, by as little as a block, separates the county.  It is not simple.



Perimeter length comports well with simple common sense notions of compactness.

I would note that you don't actually measure county boundaries either. I like your reduction because it eliminates the distinction between boundaries based on natural wiggly lines and those by a surveyor's straightedge. The distinction causes well known biases in a perimeter system of measure. Though I like your simplification, I know of no standard redistricting package that provides that measure. Sure it can be built by marking corners on a gis map of counties, but at that point it's just as much effort to mark segments as 1 or 0.

Your last picture seems to support my proxy. I see three county pairs with similar border and erosity measures, and one that is notably more erose. My score clearly isolates the least desirable of those shapes.

[jimrtex]

County borders are simple to measure.



I would note that you don't actually measure county boundaries either. I like your reduction because it eliminates the distinction between boundaries based on natural wiggly lines and those by a surveyor's straightedge. The distinction causes well known biases in a perimeter system of measure. Though I like your simplification, I know of no standard redistricting package that provides that measure. Sure it can be built by marking corners on a gis map of counties, but at that point it's just as much effort to mark segments as 1 or 0.

Your last picture seems to support my proxy. I see three county pairs with similar border and erosity measures, and one that is notably more erose. My score clearly isolates the least desirable of those shapes.

DistrictBuilder is the Open Source redistricting software used for the second Ohio competition.  You might recall that they added Roeck's circle measurement, but had incorrectly calculated the parameters of the circumscribing circle (I think they assumed that the two most distant vertices defined the diameter of the circle).

Redistricting packages have to be very good at merging shapes, which involves taking the union of the two perimeters, and removing the intersection (the common border).  The distance between the end points of the common border is what is being measured, and it is trivial to calculate distance between two points.

The administrators would necessarily have to provide files with records like:

Calhoun,Talladega,21.86,TRUE
Calhoun,Cleburne,6.107,TRUE

or

Calhoun,Talladega,TRUE
Calhoun,Cleburne,FALSE

This information would be integrated into the public entry applications (both the interactive software, and the validation/scoring software).  Distance is easy to visualize on a map, but an information mode could let the user click on a border and get its distance.

It would be pretty hard for a user to evaluate a map for road connectivity (vs. an eyeball estimate of distance).  So they would have to turn on a layer that displays which borders are good and which are bad.  I would be concerned that would lead to gaming - trying to get a high score, whereas just picking counties leads to a pretty reasonable compactness.  Remember, I draw my maps with Paint.

Distance can also be measured with Google Earth.



Of course, we might expect satellite imagery to be integrated with our redistricting application by 2020.

ps There is a paper referenced from Some Lessons from Crowd-sourced Mapping for Open Government, September 5, 2013 blog entry which is kind of interesting, but superficial, more of a survey.

[jimrtex]

In the improbable case where the courts did not accept the modest deviations of this whole county map:



This map would replace it (based on VTDs).



Covington County: Nominal shift 5,441; Actual shift 5,597.  This shifts Florala and others just to the north from Jaguar to Tiger.  An alternative would move the Opp precinct, but it a bit too large, and Tiger is very close to 0.5% threshold.

Bibb County: Nominal shift 7,011, Actual shift 6,948.  Three precincts from the north off I-59 from Tiger to Lion, which should have a closer connection to Birmingham.   There appear to be 3 precincts at one polling location, that would be split, but this shouldn't be unmanageable, even if other locations need to be found.

Walker County: Nominal shift 24,353, Actual shift 24,219.   From Lion to Blazer.  This shift can be from Blount, Walker, or Tuscaloosa.  Another map will show the Blount option.   Tuscaloosa does not appear to be particularly viable, because it would cut into the city of Tuscaloosa, and also tend to cut off the Bibb County area that was added to Lion.   There is not much suburban spillover into Walker County, just people commuting from smaller towns and rural areas into Birmingham (the urbanized area is toward the southeast edge of Jefferson County).   The area selected is generally along the border, avoiding Jasper (which is the notch in the northeast edge).  Blazer was carried by Obama in 2008.  Blount is Vermont-like white, while Walker is about 5% black.

This is an alternative that moves the shift from Lion to Blazer, from Walker to Blount.



Blount County: Nominal shift 24,353, Actual shift 25,605.   From Lion to Blazer.   There is a bit more spillover from Jefferson County, as Blount aligns with the general southwest-northeast orientation of Birmingham.  The last precinct pushed the shift a little bit higher than in Walker County.   There is an alternative that takes a couple of more rural precincts on the southeast edge, but this would tend to infringe on Oneonta.  US 278 provides good connectivity between the remainder of Blount County and Cullman County to the west.

[jimrtex]

Edit: Correct placement of Carbondale-Marion, IL

This is an updated map (no changes to definitions).   It shows metropolitan statistical areas, with rural counties indicated with a lighter shade.



Changes include:

Champaign-Urbana, IL: Douglas and Ford swapped.
South Bend, IN-MI: Cass, MI shown as dominated by South Bend urbanized area.
Syracuse, NY: Madison and Oswego indicated as dominated by Oneida and Oswego urban clusters, respectively.
Bismarck, ND: Sioux and Oliver indicated as rural parts of MSA.
Gettysburg, PA: Adams indicated as dominated by Hanover urbanized area.
Bloomsburg, PA: Montour indicated as dominated by Bloomsburg urbanized area.
Washington, DC-VA-MD-WV: Jefferson, WV indicated as dominated by Charles Town-Ranson Urban Cluster.

[jimrtex]

Edit: Correct placement of Carbondale-Marion, IL.

Northeast UCC.    Map shows Metropolitan Statistical Areas with rural counties removed.   Rural counties are considered to be those counties with either:

(1) No population in urban areas;
(2) Urban area with largest share of county population has less than 10,000 persons (ie not large enough to be core of core-based statistical area (CBSA); or
(3) Urban area with largest share of county population has less than 5,000 persons in county and less than 50% of county population (ie the urban area is not a core for the county).

Counties in solid colors have 25K or 25% of county population in urbanized areas.   These form the Urban County Clusters under the 25K/25% threshold.  Counties below the 25K/25% threshold are indicated with a lighter shade, and are not in the UCC.



The text shows included counties in black; and excluded counties in red.   Counties whose urbanized population is above 25K/25% but below 50K/50% are shown in green text.  These counties might be excluded under a more restrictive definition.

For each metropolitan statistical area (portion within the state), the population for UCC counties and in urban counties is shown, along with the number of UCC counties, and urban counties.  For each county, the population in each urbanized area in the county is shown, along with the total county population, the county population in urbanized areas, and the percentage of the county population in urbanized areas.  Populations are in thousands.


Connecticut

HARTFORD-WEST HARTFORD-EAST HARTFORD, CT 1212/1212 (3/3): Hartford County (Hartford, CT 765; Springfield, MA--CT 81)  894  846  95% SUA;  Middlesex County (Hartford, CT 68; New Haven, CT 45)  166  113  68% SUA;  Tolland County (Hartford, CT 72; Springfield, MA--CT 9)  153  81  53% SUA.

BRIDGEPORT-STAMFORD-NORWALK, CT 917/917 (1/1): Fairfield County (Bridgeport--Stamford, CT--NY 735; Danbury, CT--NY 140; New York--Newark, NY--NJ--CT 0)  917  875  95% SUA.

NEW HAVEN-MILFORD, CT 862/862 (1/1): New Haven County (New Haven, CT 518; Waterbury, CT 174; Bridgeport--Stamford, CT--NY 139)  862  831  96% SUA.

NORWICH-NEW LONDON, CT 274/274 (1/1): New London County (Norwich--New London, CT--RI 188)  274  188  69% SUA.

WORCESTER, MA-CT 118/118 (1/1): Windham County (Worcester, MA--CT 33)  118  33  28% SUA.


Delaware

PHILADELPHIA-CAMDEN-WILMINGTON, PA-NJ-DE-MD 538/538 (1/1): New Castle County (Philadelphia, PA--NJ--DE--MD 482; Dover, DE 3)  538  484  90% SUA.

SALISBURY, MD-DE 0/197 (0/1): Sussex County (Salisbury, MD--DE 25)  197  25  12% NSUA.

DOVER, DE 162/162 (1/1): Kent County (Dover, DE 108)  162  108  67% SUA.


District of Columbia

WASHINGTON-ARLINGTON-ALEXANDRIA, DC-VA-MD-WV 602/602 (1/1): District of Columbia (Washington, DC--VA--MD 602)  602  602  100% SUA.


Illinois

CHICAGO-NAPERVILLE-ELGIN, IL-IN-WI 8537/8587 (8/9): Cook County (Chicago, IL--IN 5192)  5195  5192  100% SUA;  DuPage County (Chicago, IL--IN 916)  917  916  100% SUA;  Lake County (Chicago, IL--IN 498; Round Lake Beach--McHenry--Grayslake, IL--WI 197; Kenosha, WI--IL 0)  703  695  99% SUA;  Will County (Chicago, IL--IN 630)  678  630  93% SUA;  Kane County (Chicago, IL--IN 492)  515  492  96% SUA;  McHenry County (Chicago, IL--IN 168; Round Lake Beach--McHenry--Grayslake, IL--WI 63)  309  231  75% SUA;  Kendall County (Chicago, IL--IN 103)  115  103  90% SUA;  DeKalb County (DeKalb, IL 69; Chicago, IL--IN 7)  105  76  72% SUA;  Grundy County (Chicago, IL--IN 12)  50  12  25% NSUA.

ST. LOUIS, MO-IL 572/572 (3/3): St. Clair County (St. Louis, MO--IL 229)  270  229  85% SUA;  Madison County (St. Louis, MO--IL 134; Alton, IL--MO 84)  269  218  81% SUA;  Monroe County (St. Louis, MO--IL 10)  33  10  29% SUA.

PEORIA, IL 322/322 (2/2): Peoria County (Peoria, IL 157)  186  157  84% SUA;  Tazewell County (Peoria, IL 104)  135  104  77% SUA.

ROCKFORD, IL 349/349 (2/2): Winnebago County (Rockford, IL 253; Beloit, WI--IL 19)  295  272  92% SUA;  Boone County (Rockford, IL 44)  54  44  81% SUA.

CHAMPAIGN-URBANA, IL 201/201 (1/1): Champaign County (Champaign, IL 145)  201  145  72% SUA.

DAVENPORT-MOLINE-ROCK ISLAND, IA-IL 148/198 (1/2): Rock Island County (Davenport, IA--IL 131)  148  131  89% SUA;  Henry County (Davenport, IA--IL 6)  50  6  11% NSUA.

SPRINGFIELD, IL 197/197 (1/1): Sangamon County (Springfield, IL 161)  197  161  82% SUA.

BLOOMINGTON, IL 170/170 (1/1): McLean County (Bloomington--Normal, IL 133)  170  133  78% SUA.

CARBONDALE-MARION, IL 127/127 (2/2): Williamson County (Carbondale, IL 40)  66  40  60% SUA;  Jackson County (Carbondale, IL 28)  60  28  47% SUA.

KANKAKEE, IL 113/113 (1/1): Kankakee County (Kankakee, IL 82)  113  82  72% SUA.

DECATUR, IL 111/111 (1/1): Macon County (Decatur, IL 94)  111  94  85% SUA.

DANVILLE, IL 82/82 (1/1): Vermilion County (Danville, IL 51)  82  51  62% SUA.

CAPE GIRARDEAU, MO-IL 0/0 (0/0): None.


Indiana

INDIANAPOLIS-CARMEL-ANDERSON, IN 1790/1873 (8/10): Marion County (Indianapolis, IN 898)  903  898  99% SUA;  Hamilton County (Indianapolis, IN 256)  275  256  93% SUA;  Hendricks County (Indianapolis, IN 119)  145  119  82% SUA;  Johnson County (Indianapolis, IN 116; Columbus, IN 4)  140  120  86% SUA;  Madison County (Anderson, IN 87; Indianapolis, IN 5)  132  92  70% SUA;  Hancock County (Indianapolis, IN 49)  70  49  70% SUA;  Morgan County (Indianapolis, IN 22)  69  22  32% SUA;  Boone County (Indianapolis, IN 22)  57  22  38% SUA;  Shelby County (Indianapolis, IN 0; Columbus, IN 0)  44  1  1% NSUA;  Putnam County ()  38  0  0% NSUA.

CHICAGO-NAPERVILLE-ELGIN, IL-IN-WI 660/660 (2/2): Lake County (Chicago, IL--IN 466)  496  466  94% SUA;  Porter County (Chicago, IL--IN 123; Michigan City--La Porte, IN--MI 1)  164  124  75% SUA.

FORT WAYNE, IN 355/355 (1/1): Allen County (Fort Wayne, IN 313)  355  313  88% SUA.

LOUISVILLE/JEFFERSON COUNTY, KY-IN 185/209 (2/3): Clark County (Louisville/Jefferson County, KY--IN 81)  110  81  73% SUA;  Floyd County (Louisville/Jefferson County, KY--IN 59)  75  59  80% SUA;  Scott County ()  24  0  0% NSUA.

SOUTH BEND-MISHAWAKA, IN-MI 267/267 (1/1): St. Joseph County (South Bend, IN--MI 238)  267  238  89% SUA.

EVANSVILLE, IN-KY 239/239 (2/2): Vanderburgh County (Evansville, IN--KY 163)  180  163  91% SUA;  Warrick County (Evansville, IN--KY 36)  60  36  60% SUA.

LAFAYETTE-WEST LAFAYETTE, IN 173/173 (1/1): Tippecanoe County (Lafayette, IN 148)  173  148  85% SUA.

ELKHART-GOSHEN, IN 198/198 (1/1): Elkhart County (Elkhart, IN--MI 143; South Bend, IN--MI 4)  198  146  74% SUA.

TERRE HAUTE, IN 135/135 (2/2): Vigo County (Terre Haute, IN 82)  108  82  76% SUA;  Clay County (Terre Haute, IN 11)  27  11  39% SUA.

BLOOMINGTON, IN 138/138 (1/1): Monroe County (Bloomington, IN 109)  138  109  79% SUA.

MUNCIE, IN 118/118 (1/1): Delaware County (Muncie, IN 89; Anderson, IN 2)  118  91  77% SUA.

MICHIGAN CITY-LA PORTE, IN 111/111 (1/1): LaPorte County (Michigan City--La Porte, IN--MI 65)  111  65  58% SUA.

KOKOMO, IN 83/83 (1/1): Howard County (Kokomo, IN 62)  83  62  75% SUA.

COLUMBUS, IN 77/77 (1/1): Bartholomew County (Columbus, IN 51)  77  51  66% SUA.

CINCINNATI, OH-KY-IN 0/50 (0/1): Dearborn County (Cincinnati, OH--KY--IN 10)  50  10  20% NSUA.


Iowa

DES MOINES-WEST DES MOINES, IA 497/543 (2/3): Polk County (Des Moines, IA 406)  431  406  94% SUA;  Dallas County (Des Moines, IA 35)  66  35  53% SUA;  Warren County (Des Moines, IA 9)  46  9  19% NSUA.

CEDAR RAPIDS, IA 211/211 (1/1): Linn County (Cedar Rapids, IA 178)  211  178  84% SUA.

WATERLOO-CEDAR FALLS, IA 131/131 (1/1): Black Hawk County (Waterloo, IA 113)  131  113  87% SUA.

DAVENPORT-MOLINE-ROCK ISLAND, IA-IL 165/165 (1/1): Scott County (Davenport, IA--IL 143)  165  143  86% SUA.

IOWA CITY, IA 131/131 (1/1): Johnson County (Iowa City, IA 107)  131  107  81% SUA.

SIOUX CITY, IA-NE-SD 102/102 (1/1): Woodbury County (Sioux City, IA--NE--SD 84)  102  84  83% SUA.

OMAHA-COUNCIL BLUFFS, NE-IA 93/93 (1/1): Pottawattamie County (Omaha, NE--IA 69)  93  69  74% SUA.

DUBUQUE, IA 94/94 (1/1): Dubuque County (Dubuque, IA--IL 65)  94  65  69% SUA.

AMES, IA 90/90 (1/1): Story County (Ames, IA 60)  90  60  67% SUA.


Kansas

KANSAS CITY, MO-KS 702/778 (2/3): Johnson County (Kansas City, MO--KS 516)  544  516  95% SUA;  Wyandotte County (Kansas City, MO--KS 148)  158  148  94% SUA;  Leavenworth County ()  76  0  0% NSUA.

WICHITA, KS 498/599 (1/3): Sedgwick County (Wichita, KS 460)  498  460  92% SUA;  Butler County (Wichita, KS 11)  66  11  17% NSUA;  Harvey County ()  35  0  0% NSUA.

TOPEKA, KS 178/178 (1/1): Shawnee County (Topeka, KS 150)  178  150  84% SUA.

LAWRENCE, KS 111/111 (1/1): Douglas County (Lawrence, KS 88)  111  88  79% SUA.

MANHATTAN, KS 71/71 (1/1): Riley County (Manhattan, KS 53)  71  53  74% SUA.

ST. JOSEPH, MO-KS 0/0 (0/0): None.

[jimrtex]

Edit: Corrected placement of Carbondale-Marion, IL.



Kentucky

LOUISVILLE/JEFFERSON COUNTY, KY-IN 876/918 (3/4): Jefferson County (Louisville/Jefferson County, KY--IN 731)  741  731  99% SUA;  Bullitt County (Louisville/Jefferson County, KY--IN 52)  74  52  70% SUA;  Oldham County (Louisville/Jefferson County, KY--IN 48)  60  48  80% SUA;  Shelby County (Louisville/Jefferson County, KY--IN 2)  42  2  4% NSUA.

LEXINGTON-FAYETTE, KY 296/472 (1/6): Fayette County (Lexington-Fayette, KY 287)  296  287  97% SUA;  Jessamine County (Lexington-Fayette, KY 2)  49  2  5% NSUA;  Scott County (Lexington-Fayette, KY 1)  47  1  2% NSUA;  Clark County ()  36  0  0% NSUA;  Woodford County ()  25  0  0% NSUA;  Bourbon County ()  20  0  0% NSUA.

CINCINNATI, OH-KY-IN 369/369 (3/3): Kenton County (Cincinnati, OH--KY--IN 149)  160  149  93% SUA;  Boone County (Cincinnati, OH--KY--IN 103)  119  103  87% SUA;  Campbell County (Cincinnati, OH--KY--IN 77)  90  77  85% SUA.

BOWLING GREEN, KY 114/114 (1/1): Warren County (Bowling Green, KY 78)  114  78  69% SUA.

ELIZABETHTOWN-FORT KNOX, KY 106/106 (1/1): Hardin County (Elizabethtown--Radcliff, KY 69)  106  69  66% SUA.

OWENSBORO, KY 97/97 (1/1): Daviess County (Owensboro, KY 71)  97  71  73% SUA.

CLARKSVILLE, TN-KY 74/74 (1/1): Christian County (Clarksville, TN--KY 20)  74  20  28% SUA.

HUNTINGTON-ASHLAND, WV-KY-OH 86/86 (2/2): Boyd County (Huntington, WV--KY--OH 37)  50  37  75% SUA;  Greenup County (Huntington, WV--KY--OH 20)  37  20  53% SUA.

EVANSVILLE, IN-KY 46/46 (1/1): Henderson County (Evansville, IN--KY 29)  46  29  62% SUA.


Maine

PORTLAND-SOUTH PORTLAND, ME 479/514 (2/3): Cumberland County (Portland, ME 164; Lewiston, ME 0)  282  164  58% SUA;  York County (Portland, ME 40; Portsmouth, NH--ME 16; Dover--Rochester, NH--ME 8)  197  63  32% SUA;  Sagadahoc County ()  35  0  0% NSUA.

BANGOR, ME 154/154 (1/1): Penobscot County (Bangor, ME 61)  154  61  40% SUA.

LEWISTON-AUBURN, ME 108/108 (1/1): Androscoggin County (Lewiston, ME 59)  108  59  55% SUA.


Maryland

BALTIMORE-COLUMBIA-TOWSON, MD 2710/2710 (7/7): Baltimore County (Baltimore, MD 750; Aberdeen--Bel Air South--Bel Air North, MD 3)  805  752  93% SUA;  Baltimore city (Baltimore, MD 621)  621  621  100% SUA;  Anne Arundel County (Baltimore, MD 496; Washington, DC--VA--MD 2)  538  498  93% SUA;  Howard County (Baltimore, MD 257; Westminster--Eldersburg, MD 0)  287  257  89% SUA;  Harford County (Aberdeen--Bel Air South--Bel Air North, MD 201)  245  201  82% SUA;  Carroll County (Westminster--Eldersburg, MD 73; Baltimore, MD 9; Washington, DC--VA--MD 7)  167  89  53% SUA;  Queen Anne's County (Baltimore, MD 12)  48  12  26% SUA.

WASHINGTON-ARLINGTON-ALEXANDRIA, DC-VA-MD-WV 2215/2304 (4/5): Montgomery County (Washington, DC--VA--MD 944; Baltimore, MD 0)  972  944  97% SUA;  Prince George's County (Washington, DC--VA--MD 781; Baltimore, MD 58; Waldorf, MD 7)  863  846  98% SUA;  Frederick County (Frederick, MD 142; Washington, DC--VA--MD 15)  233  157  67% SUA;  Charles County (Waldorf, MD 103)  147  103  70% SUA;  Calvert County (Lexington Park--California--Chesapeake Ranch Estates, MD 19)  89  19  21% NSUA.

SALISBURY, MD-DE 99/177 (1/3): Wicomico County (Salisbury, MD--DE 73)  99  73  74% SUA;  Worcester County ()  51  0  0% NSUA;  Somerset County (Salisbury, MD--DE 0)  26  0  1% NSUA.

HAGERSTOWN-MARTINSBURG, MD-WV 147/147 (1/1): Washington County (Hagerstown, MD--WV--PA 101)  147  101  69% SUA.

CALIFORNIA-LEXINGTON PARK, MD 105/105 (1/1): St. Mary's County (Lexington Park--California--Chesapeake Ranch Estates, MD 40)  105  40  38% SUA.

PHILADELPHIA-CAMDEN-WILMINGTON, PA-NJ-DE-MD 101/101 (1/1): Cecil County (Philadelphia, PA--NJ--DE--MD 49; Aberdeen--Bel Air South--Bel Air North, MD 10)  101  59  58% SUA.

CUMBERLAND, MD-WV 75/75 (1/1): Allegany County (Cumberland, MD--WV--PA 50)  75  50  66% SUA.


Massachusetts

BOSTON-CAMBRIDGE-NEWTON, MA-NH 4134/4134 (5/5): Middlesex County (Boston, MA--NH--RI 1449; Nashua, NH--MA 7; Worcester, MA--CT 1; Leominster--Fitchburg, MA 0)  1503  1458  97% SUA;  Essex County (Boston, MA--NH--RI 712)  743  712  96% SUA;  Suffolk County (Boston, MA--NH--RI 721)  722  721  100% SUA;  Norfolk County (Boston, MA--NH--RI 643; Providence, RI--MA 19)  671  663  99% SUA;  Plymouth County (Boston, MA--NH--RI 388; Barnstable Town, MA 50; New Bedford, MA 6)  495  444  90% SUA.

WORCESTER, MA-CT 799/799 (1/1): Worcester County (Worcester, MA--CT 452; Leominster--Fitchburg, MA 117; Boston, MA--NH--RI 50; Providence, RI--MA 14)  799  634  79% SUA.

SPRINGFIELD, MA 622/622 (2/2): Hampden County (Springfield, MA--CT 424)  463  424  91% SUA;  Hampshire County (Springfield, MA--CT 108)  158  108  68% SUA.

PROVIDENCE-WARWICK, RI-MA 548/548 (1/1): Bristol County (Providence, RI--MA 226; New Bedford, MA 143; Boston, MA--NH--RI 124)  548  494  90% SUA.

BARNSTABLE TOWN, MA 216/216 (1/1): Barnstable County (Barnstable Town, MA 197)  216  197  91% SUA.

PITTSFIELD, MA 131/131 (1/1): Berkshire County (Pittsfield, MA 59)  131  59  45% SUA.


Michigan

DETROIT-WARREN-DEARBORN, MI 4208/4296 (5/6): Wayne County (Detroit, MI 1772; Ann Arbor, MI 36)  1821  1808  99% SUA;  Oakland County (Detroit, MI 1117; South Lyon--Howell, MI 16; Flint, MI 1)  1202  1134  94% SUA;  Macomb County (Detroit, MI 811)  841  811  96% SUA;  Livingston County (South Lyon--Howell, MI 97; Detroit, MI 10; Flint, MI 4)  181  110  61% SUA;  St. Clair County (Port Huron, MI 86; Detroit, MI 14)  163  100  61% SUA;  Lapeer County (Detroit, MI 0)  88  0  0% NSUA.

GRAND RAPIDS-WYOMING, MI 866/866 (2/2): Kent County (Grand Rapids, MI 494)  603  494  82% SUA;  Ottawa County (Holland, MI 90; Grand Rapids, MI 76; Muskegon, MI 40)  264  206  78% SUA.

LANSING-EAST LANSING, MI 464/464 (3/3): Ingham County (Lansing, MI 237)  281  237  85% SUA;  Eaton County (Lansing, MI 49)  108  49  45% SUA;  Clinton County (Lansing, MI 27)  75  27  36% SUA.

FLINT, MI 426/426 (1/1): Genesee County (Flint, MI 352)  426  352  83% SUA.

ANN ARBOR, MI 345/345 (1/1): Washtenaw County (Ann Arbor, MI 270; South Lyon--Howell, MI 7; Detroit, MI 0)  345  277  80% SUA.

KALAMAZOO-PORTAGE, MI 250/250 (1/1): Kalamazoo County (Kalamazoo, MI 205; Battle Creek, MI 1)  250  206  82% SUA.

SAGINAW, MI 200/200 (1/1): Saginaw County (Saginaw, MI 126; Midland, MI 7)  200  133  66% SUA.

MUSKEGON, MI 172/172 (1/1): Muskegon County (Muskegon, MI 122)  172  122  71% SUA.

JACKSON, MI 160/160 (1/1): Jackson County (Jackson, MI 90)  160  90  56% SUA.

NILES-BENTON HARBOR, MI 157/157 (1/1): Berrien County (Benton Harbor--St. Joseph--Fair Plain, MI 61; South Bend, IN--MI 28; Michigan City--La Porte, IN--MI 1)  157  90  57% SUA.

MONROE, MI 152/152 (1/1): Monroe County (Monroe, MI 51; Toledo, OH--MI 28; Detroit, MI 9)  152  89  59% SUA.

BATTLE CREEK, MI 136/136 (1/1): Calhoun County (Battle Creek, MI 77)  136  77  57% SUA.

BAY CITY, MI 108/108 (1/1): Bay County (Bay City, MI 71; Midland, MI 4; Saginaw, MI 0)  108  75  70% SUA.

MIDLAND, MI 84/84 (1/1): Midland County (Midland, MI 48)  84  48  57% SUA.

SOUTH BEND-MISHAWAKA, IN-MI 0/52 (0/1): Cass County (South Bend, IN--MI 8; Elkhart, IN--MI 1)  52  9  17% NSUA.


Minnesota

MINNEAPOLIS-ST. PAUL-BLOOMINGTON, MN-WI 3063/3101 (9/10): Hennepin County (Minneapolis--St. Paul, MN--WI 1126)  1152  1126  98% SUA;  Ramsey County (Minneapolis--St. Paul, MN--WI 508)  509  508  100% SUA;  Dakota County (Minneapolis--St. Paul, MN--WI 356)  399  356  89% SUA;  Anoka County (Minneapolis--St. Paul, MN--WI 280)  331  280  85% SUA;  Washington County (Minneapolis--St. Paul, MN--WI 185)  238  185  78% SUA;  Scott County (Minneapolis--St. Paul, MN--WI 88)  130  88  68% SUA;  Wright County (Minneapolis--St. Paul, MN--WI 34)  125  34  27% SUA;  Carver County (Minneapolis--St. Paul, MN--WI 55)  91  55  61% SUA;  Sherburne County (Minneapolis--St. Paul, MN--WI 19; St. Cloud, MN 7)  88  26  29% SUA;  Isanti County ()  38  0  0% NSUA.

DULUTH, MN-WI 200/236 (1/2): St. Louis County (Duluth, MN--WI 93)  200  93  47% SUA;  Carlton County ()  35  0  0% NSUA.

ROCHESTER, MN 144/144 (1/1): Olmsted County (Rochester, MN 108)  144  108  75% SUA.

ST. CLOUD, MN 189/189 (2/2): Stearns County (St. Cloud, MN 81)  151  81  54% SUA;  Benton County (St. Cloud, MN 23)  38  23  60% SUA.

MANKATO-NORTH MANKATO, MN 97/97 (2/2): Blue Earth County (Mankato, MN 44)  64  44  69% SUA;  Nicollet County (Mankato, MN 13)  33  13  41% SUA.

FARGO, ND-MN 59/59 (1/1): Clay County (Fargo, ND--MN 43)  59  43  72% SUA.

GRAND FORKS, ND-MN 32/32 (1/1): Polk County (Grand Forks, ND--MN 8)  32  8  26% SUA.

LA CROSSE-ONALASKA, WI-MN 19/19 (1/1): Houston County (La Crosse, WI--MN 5)  19  5  28% SUA.

[jimrtex]

Edit: Corrected placement of Carbondale-Marion, IL.




Missouri

ST. LOUIS, MO-IL 1897/2052 (4/6): St. Louis County (St. Louis, MO--IL 976)  999  976  98% SUA;  St. Charles County (St. Louis, MO--IL 339; Alton, IL--MO 0)  360  339  94% SUA;  St. Louis city (St. Louis, MO--IL 319)  319  319  100% SUA;  Jefferson County (St. Louis, MO--IL 143)  219  143  65% SUA;  Franklin County ()  101  0  0% NSUA;  Lincoln County ()  53  0  0% NSUA.

KANSAS CITY, MO-KS 1085/1085 (4/4): Jackson County (Kansas City, MO--KS 564; Lee's Summit, MO 81)  674  645  96% SUA;  Clay County (Kansas City, MO--KS 180)  222  180  81% SUA;  Cass County (Kansas City, MO--KS 46; Lee's Summit, MO 4)  99  50  50% SUA;  Platte County (Kansas City, MO--KS 66)  89  66  74% SUA.

SPRINGFIELD, MO 353/353 (2/2): Greene County (Springfield, MO 231)  275  231  84% SUA;  Christian County (Springfield, MO 43)  77  43  55% SUA.

JOPLIN, MO 117/176 (1/2): Jasper County (Joplin, MO 74)  117  74  63% SUA;  Newton County (Joplin, MO 9)  58  9  15% NSUA.

COLUMBIA, MO 163/163 (1/1): Boone County (Columbia, MO 125)  163  125  77% SUA.

JEFFERSON CITY, MO 76/120 (1/2): Cole County (Jefferson City, MO 54)  76  54  71% SUA;  Callaway County (Jefferson City, MO 5)  44  5  10% NSUA.

ST. JOSEPH, MO-KS 89/89 (1/1): Buchanan County (St. Joseph, MO--KS 77)  89  77  87% SUA.

CAPE GIRARDEAU, MO-IL 76/76 (1/1): Cape Girardeau County (Cape Girardeau, MO--IL 53)  76  53  69% SUA.

FAYETTEVILLE-SPRINGDALE-ROGERS, AR-MO 0/0 (0/0): None.


Nebraska

OMAHA-COUNCIL BLUFFS, NE-IA 676/676 (2/2): Douglas County (Omaha, NE--IA 506)  517  506  98% SUA;  Sarpy County (Omaha, NE--IA 150)  159  150  95% SUA.

LINCOLN, NE 285/285 (1/1): Lancaster County (Lincoln, NE 259)  285  259  91% SUA.

GRAND ISLAND, NE 59/59 (1/1): Hall County (Grand Island, NE 50)  59  50  85% SUA.

SIOUX CITY, IA-NE-SD 21/21 (1/1): Dakota County (Sioux City, IA--NE--SD 17)  21  17  79% SUA.


New Hampshire

BOSTON-CAMBRIDGE-NEWTON, MA-NH 418/418 (2/2): Rockingham County (Boston, MA--NH--RI 84; Portsmouth, NH--ME 72; Nashua, NH--MA 61; Manchester, NH 1)  295  218  74% SUA;  Strafford County (Dover--Rochester, NH--ME 80; Portsmouth, NH--ME 0)  123  80  65% SUA.

MANCHESTER-NASHUA, NH 401/401 (1/1): Hillsborough County (Nashua, NH--MA 158; Manchester, NH 141; Boston, MA--NH--RI 9)  401  309  77% SUA.


New Jersey

NEW YORK-NEWARK-JERSEY CITY, NY-NJ-PA 6471/6471 (12/12): Bergen County (New York--Newark, NY--NJ--CT 904)  905  904  100% SUA;  Middlesex County (New York--Newark, NY--NJ--CT 799; Twin Rivers--Hightstown, NJ 5)  810  804  99% SUA;  Essex County (New York--Newark, NY--NJ--CT 784)  784  784  100% SUA;  Hudson County (New York--Newark, NY--NJ--CT 634)  634  634  100% SUA;  Monmouth County (New York--Newark, NY--NJ--CT 603; Trenton, NJ 3; Twin Rivers--Hightstown, NJ 1)  630  607  96% SUA;  Ocean County (New York--Newark, NY--NJ--CT 532)  577  532  92% SUA;  Union County (New York--Newark, NY--NJ--CT 536)  536  536  100% SUA;  Passaic County (New York--Newark, NY--NJ--CT 473; Poughkeepsie--Newburgh, NY--NJ 11)  501  484  97% SUA;  Morris County (New York--Newark, NY--NJ--CT 449)  492  449  91% SUA;  Somerset County (New York--Newark, NY--NJ--CT 305)  323  305  94% SUA;  Sussex County (New York--Newark, NY--NJ--CT 37)  149  37  25% SUA;  Hunterdon County (New York--Newark, NY--NJ--CT 52; Philadelphia, PA--NJ--DE--MD 6; Allentown, PA--NJ 1)  128  58  46% SUA.

PHILADELPHIA-CAMDEN-WILMINGTON, PA-NJ-DE-MD 1317/1317 (4/4): Camden County (Philadelphia, PA--NJ--DE--MD 505)  514  505  98% SUA;  Burlington County (Philadelphia, PA--NJ--DE--MD 356; Trenton, NJ 28)  449  384  86% SUA;  Gloucester County (Philadelphia, PA--NJ--DE--MD 258; Vineland, NJ 6)  288  264  92% SUA;  Salem County (Philadelphia, PA--NJ--DE--MD 24; Vineland, NJ 3)  66  26  40% SUA.

TRENTON, NJ 367/367 (1/1): Mercer County (Trenton, NJ 265; Twin Rivers--Hightstown, NJ 58; New York--Newark, NY--NJ--CT 30)  367  354  97% SUA.

ATLANTIC CITY-HAMMONTON, NJ 275/275 (1/1): Atlantic County (Atlantic City, NJ 219; Vineland, NJ 4; Philadelphia, PA--NJ--DE--MD 3)  275  227  83% SUA.

VINELAND-BRIDGETON, NJ 157/157 (1/1): Cumberland County (Vineland, NJ 82)  157  82  52% SUA.

ALLENTOWN-BETHLEHEM-EASTON, PA-NJ 109/109 (1/1): Warren County (Allentown, PA--NJ 32; New York--Newark, NY--NJ--CT 21; East Stroudsburg, PA--NJ 0)  109  53  48% SUA.

OCEAN CITY, NJ 97/97 (1/1): Cape May County (Villas, NJ 51; Atlantic City, NJ 29)  97  80  83% SUA.


New York

NEW YORK-NEWARK-JERSEY CITY, NY-NJ-PA 13039/13039 (12/12): Kings County (New York--Newark, NY--NJ--CT 2505)  2505  2505  100% SUA;  Queens County (New York--Newark, NY--NJ--CT 2231)  2231  2231  100% SUA;  New York County (New York--Newark, NY--NJ--CT 1586)  1586  1586  100% SUA;  Suffolk County (New York--Newark, NY--NJ--CT 1429)  1493  1429  96% SUA;  Bronx County (New York--Newark, NY--NJ--CT 1385)  1385  1385  100% SUA;  Nassau County (New York--Newark, NY--NJ--CT 1337)  1340  1337  100% SUA;  Westchester County (New York--Newark, NY--NJ--CT 870; Bridgeport--Stamford, CT--NY 46)  949  915  96% SUA;  Richmond County (New York--Newark, NY--NJ--CT 469)  469  469  100% SUA;  Orange County (Poughkeepsie--Newburgh, NY--NJ 170; Middletown, NY 58)  373  229  61% SUA;  Rockland County (New York--Newark, NY--NJ--CT 309)  312  309  99% SUA;  Dutchess County (Poughkeepsie--Newburgh, NY--NJ 210; New York--Newark, NY--NJ--CT 2)  297  212  71% SUA;  Putnam County (New York--Newark, NY--NJ--CT 69; Danbury, CT--NY 7)  100  76  76% SUA.

BUFFALO-CHEEKTOWAGA-NIAGARA FALLS, NY 1136/1136 (2/2): Erie County (Buffalo, NY 812)  919  812  88% SUA;  Niagara County (Buffalo, NY 124)  216  124  57% SUA.

ROCHESTER, NY 852/946 (2/3): Monroe County (Rochester, NY 673)  744  673  90% SUA;  Ontario County (Rochester, NY 35)  108  35  32% SUA;  Wayne County (Rochester, NY 12)  94  12  13% NSUA.

ALBANY-SCHENECTADY-TROY, NY 838/838 (4/4): Albany County (Albany--Schenectady, NY 270)  304  270  89% SUA;  Saratoga County (Albany--Schenectady, NY 76; Saratoga Springs, NY 64; Glens Falls, NY 11)  220  151  69% SUA;  Rensselaer County (Albany--Schenectady, NY 107)  159  107  67% SUA;  Schenectady County (Albany--Schenectady, NY 142)  155  142  92% SUA.

SYRACUSE, NY 467/663 (1/3): Onondaga County (Syracuse, NY 405)  467  405  87% SUA;  Oswego County (Syracuse, NY 7)  122  7  6% NSUA;  Madison County (Syracuse, NY 0)  73  0  0% NSUA.

UTICA-ROME, NY 235/299 (1/2): Oneida County (Utica, NY 116)  235  116  49% SUA;  Herkimer County (Utica, NY 1)  65  1  2% NSUA.

BINGHAMTON, NY 201/252 (1/2): Broome County (Binghamton, NY--PA 148)  201  148  74% SUA;  Tioga County (Binghamton, NY--PA 7)  51  7  14% NSUA.

KINGSTON, NY 182/182 (1/1): Ulster County (Kingston, NY 57; Poughkeepsie--Newburgh, NY--NJ 32)  182  89  49% SUA.

GLENS FALLS, NY 66/129 (1/2): Warren County (Glens Falls, NY 40)  66  40  61% SUA;  Washington County (Glens Falls, NY 15)  63  15  24% NSUA.

WATERTOWN-FORT DRUM, NY 116/116 (1/1): Jefferson County (Watertown, NY 58)  116  58  50% SUA.

ITHACA, NY 102/102 (1/1): Tompkins County (Ithaca, NY 54)  102  54  53% SUA.

ELMIRA, NY 89/89 (1/1): Chemung County (Elmira, NY 67)  89  67  76% SUA.


North Dakota

FARGO, ND-MN 150/150 (1/1): Cass County (Fargo, ND--MN 134)  150  134  90% SUA.

BISMARCK, ND 109/109 (2/2): Burleigh County (Bismarck, ND 63)  81  63  78% SUA;  Morton County (Bismarck, ND 19)  27  19  68% SUA.

GRAND FORKS, ND-MN 67/67 (1/1): Grand Forks County (Grand Forks, ND--MN 53)  67  53  79% SUA.

[jimrtex]

Edit: Corrected placement of Carbondale-Marion, IL.




Ohio

CLEVELAND-ELYRIA, OH 1984/2077 (4/5): Cuyahoga County (Cleveland, OH 1273)  1280  1273  99% SUA;  Lorain County (Lorain--Elyria, OH 175; Cleveland, OH 75)  301  250  83% SUA;  Lake County (Cleveland, OH 215)  230  215  93% SUA;  Medina County (Cleveland, OH 89; Akron, OH 24)  172  113  65% SUA;  Geauga County (Cleveland, OH 21)  93  21  22% NSUA.

COLUMBUS, OH 1650/1802 (4/7): Franklin County (Columbus, OH 1146)  1163  1146  98% SUA;  Delaware County (Columbus, OH 141)  174  141  81% SUA;  Licking County (Newark, OH 74; Columbus, OH 28)  166  103  62% SUA;  Fairfield County (Columbus, OH 48; Newark, OH 2)  146  49  34% SUA;  Pickaway County (Columbus, OH 3)  56  3  5% NSUA;  Union County (Columbus, OH 3)  52  3  6% NSUA;  Madison County (Columbus, OH 0)  43  0  0% NSUA.

CINCINNATI, OH-KY-IN 1581/1581 (4/4): Hamilton County (Cincinnati, OH--KY--IN 772)  802  772  96% SUA;  Butler County (Cincinnati, OH--KY--IN 250; Middletown, OH 62)  368  312  85% SUA;  Warren County (Cincinnati, OH--KY--IN 118; Middletown, OH 29; Dayton, OH 29)  213  176  83% SUA;  Clermont County (Cincinnati, OH--KY--IN 146)  197  146  74% SUA.

DAYTON, OH 799/799 (3/3): Montgomery County (Dayton, OH 502; Middletown, OH 6)  535  508  95% SUA;  Greene County (Dayton, OH 126)  162  126  78% SUA;  Miami County (Dayton, OH 47)  103  47  46% SUA.

AKRON, OH 703/703 (2/2): Summit County (Akron, OH 441; Cleveland, OH 79; Canton, OH 1)  542  521  96% SUA;  Portage County (Akron, OH 73; Cleveland, OH 29; Youngstown, OH--PA 0)  161  102  63% SUA.

TOLEDO, OH 567/567 (2/2): Lucas County (Toledo, OH--MI 421)  442  421  95% SUA;  Wood County (Toledo, OH--MI 53)  125  53  42% SUA.

YOUNGSTOWN-WARREN-BOARDMAN, OH-PA 449/449 (2/2): Mahoning County (Youngstown, OH--PA 195)  239  195  82% SUA;  Trumbull County (Youngstown, OH--PA 153)  210  153  73% SUA.

CANTON-MASSILLON, OH 376/376 (1/1): Stark County (Canton, OH 278; Akron, OH 19)  376  297  79% SUA.

SPRINGFIELD, OH 138/138 (1/1): Clark County (Springfield, OH 85; Dayton, OH 20)  138  106  76% SUA.

MANSFIELD, OH 124/124 (1/1): Richland County (Mansfield, OH 75)  124  75  60% SUA.

LIMA, OH 106/106 (1/1): Allen County (Lima, OH 71)  106  71  67% SUA.

WHEELING, WV-OH 70/70 (1/1): Belmont County (Wheeling, WV--OH 28)  70  28  39% SUA.

WEIRTON-STEUBENVILLE, WV-OH 70/70 (1/1): Jefferson County (Weirton--Steubenville, WV--OH--PA 40; Wheeling, WV--OH 3)  70  43  61% SUA.

HUNTINGTON-ASHLAND, WV-KY-OH 62/62 (1/1): Lawrence County (Huntington, WV--KY--OH 34)  62  34  54% SUA.


Pennsylvania

PHILADELPHIA-CAMDEN-WILMINGTON, PA-NJ-DE-MD 4009/4009 (5/5): Philadelphia County (Philadelphia, PA--NJ--DE--MD 1526)  1526  1526  100% SUA;  Montgomery County (Philadelphia, PA--NJ--DE--MD 716; Pottstown, PA 60)  800  776  97% SUA;  Bucks County (Philadelphia, PA--NJ--DE--MD 541; Allentown, PA--NJ 29)  625  570  91% SUA;  Delaware County (Philadelphia, PA--NJ--DE--MD 556)  559  556  100% SUA;  Chester County (Philadelphia, PA--NJ--DE--MD 416; Pottstown, PA 15)  499  431  86% SUA.

PITTSBURGH, PA 2287/2356 (6/7): Allegheny County (Pittsburgh, PA 1192; Monessen--California, PA 1)  1223  1193  98% SUA;  Westmoreland County (Pittsburgh, PA 252; Monessen--California, PA 17)  365  269  74% SUA;  Washington County (Pittsburgh, PA 101; Monessen--California, PA 37; Weirton--Steubenville, WV--OH--PA 0)  208  139  67% SUA;  Butler County (Pittsburgh, PA 59)  184  59  32% SUA;  Beaver County (Pittsburgh, PA 120)  171  120  70% SUA;  Fayette County (Uniontown--Connellsville, PA 51; Monessen--California, PA 11; Pittsburgh, PA 2)  137  65  47% SUA;  Armstrong County (Pittsburgh, PA 8)  69  8  12% NSUA.

ALLENTOWN-BETHLEHEM-EASTON, PA-NJ 712/712 (3/3): Lehigh County (Allentown, PA--NJ 322)  349  322  92% SUA;  Northampton County (Allentown, PA--NJ 260)  298  260  87% SUA;  Carbon County (Allentown, PA--NJ 16; Hazleton, PA 2)  65  18  27% SUA.

SCRANTON--WILKES-BARRE--HAZLETON, PA 535/535 (2/2): Luzerne County (Scranton, PA 201; Hazleton, PA 52; Bloomsburg--Berwick, PA 4)  321  257  80% SUA;  Lackawanna County (Scranton, PA 177)  214  177  82% SUA.

HARRISBURG-CARLISLE, PA 504/504 (2/2): Dauphin County (Harrisburg, PA 222; Lancaster, PA 1)  268  223  83% SUA;  Cumberland County (Harrisburg, PA 166)  235  166  70% SUA.

LANCASTER, PA 519/519 (1/1): Lancaster County (Lancaster, PA 397; Philadelphia, PA--NJ--DE--MD 5)  519  402  77% SUA.

YORK-HANOVER, PA 435/435 (1/1): York County (York, PA 232; Hanover, PA 35; Harrisburg, PA 35; Lancaster, PA 4)  435  306  70% SUA.

READING, PA 411/411 (1/1): Berks County (Reading, PA 264; Pottstown, PA 32; Allentown, PA--NJ 5; Lebanon, PA 0; Lancaster, PA 0)  411  301  73% SUA.

ERIE, PA 281/281 (1/1): Erie County (Erie, PA 197)  281  197  70% SUA.

EAST STROUDSBURG, PA 170/170 (1/1): Monroe County (East Stroudsburg, PA--NJ 54; Allentown, PA--NJ 1)  170  55  33% SUA.

STATE COLLEGE, PA 154/154 (1/1): Centre County (State College, PA 87)  154  87  57% SUA.

CHAMBERSBURG-WAYNESBORO, PA 150/150 (1/1): Franklin County (Chambersburg, PA 51; Hagerstown, MD--WV--PA 10)  150  60  40% SUA.

JOHNSTOWN, PA 144/144 (1/1): Cambria County (Johnstown, PA 61)  144  61  43% SUA.

LEBANON, PA 134/134 (1/1): Lebanon County (Lebanon, PA 77; Harrisburg, PA 19; Reading, PA 2; Lancaster, PA 0)  134  98  73% SUA.

ALTOONA, PA 127/127 (1/1): Blair County (Altoona, PA 80)  127  80  63% SUA.

YOUNGSTOWN-WARREN-BOARDMAN, OH-PA 117/117 (1/1): Mercer County (Youngstown, OH--PA 39)  117  39  33% SUA.

WILLIAMSPORT, PA 116/116 (1/1): Lycoming County (Williamsport, PA 56)  116  56  48% SUA.

GETTYSBURG, PA 101/101 (1/1): Adams County (Hanover, PA 31)  101  31  31% SUA.

BLOOMSBURG-BERWICK, PA 86/86 (2/2): Columbia County (Bloomsburg--Berwick, PA 40)  67  40  59% SUA;  Montour County (Bloomsburg--Berwick, PA 8)  18  8  46% SUA.

NEW YORK-NEWARK-JERSEY CITY, NY-NJ-PA 0/0 (0/0): None.


Rhode Island

PROVIDENCE-WARWICK, RI-MA 1053/1053 (5/5): Providence County (Providence, RI--MA 592; Boston, MA--NH--RI 0)  627  592  94% SUA;  Kent County (Providence, RI--MA 153)  166  153  92% SUA;  Washington County (Providence, RI--MA 64; Norwich--New London, CT--RI 21)  127  85  67% SUA;  Newport County (Providence, RI--MA 73)  83  73  88% SUA;  Bristol County (Providence, RI--MA 49)  50  49  99% SUA.


South Dakota

SIOUX FALLS, SD 214/214 (2/2): Minnehaha County (Sioux Falls, SD 135)  169  135  79% SUA;  Lincoln County (Sioux Falls, SD 22)  45  22  49% SUA.

RAPID CITY, SD 101/101 (1/1): Pennington County (Rapid City, SD 75)  101  75  74% SUA.

SIOUX CITY, IA-NE-SD 14/14 (1/1): Union County (Sioux City, IA--NE--SD 6)  14  6  39% SUA.


Vermont

BURLINGTON-SOUTH BURLINGTON, VT 157/157 (1/1): Chittenden County (Burlington, VT 109)  157  109  69% SUA.

[jimrtex]

Edit: Corrected placement of Carbondale-Marion, IL.




Virginia

WASHINGTON-ARLINGTON-ALEXANDRIA, DC-VA-MD-WV 2571/2656 (7/9): Fairfax County (Fairfax city, Falls Church city) (Washington, DC--VA--MD 1102)  1117  1102  99% SUA;  Prince William County (Manassas city, Manassas Park city) (Washington, DC--VA--MD 437)  454  437  96% SUA;  Arlington County (Alexandria city) (Washington, DC--VA--MD 348)  348  348  100% SUA;  Loudoun County (Washington, DC--VA--MD 259)  312  259  83% SUA;  Spotsylvania County (Fredericksburg city) (Fredericksburg, VA 107)  147  107  73% SUA;  Stafford County (Washington, DC--VA--MD 69; Fredericksburg, VA 34)  129  103  80% SUA;  Fauquier County (Washington, DC--VA--MD 21)  65  21  32% SUA;  Culpeper County ()  47  0  0% NSUA;  Warren County ()  38  0  0% NSUA.

VIRGINIA BEACH-NORFOLK-NEWPORT NEWS, VA-NC 1560/1632 (7/9): Norfolk city (Chesapeake city, Portsmouth city) (Virginia Beach, VA 544)  561  544  97% SUA;  Virginia Beach city (Virginia Beach, VA 431)  438  431  98% SUA;  Newport News city (Virginia Beach, VA 181)  181  181  100% SUA;  Hampton city (Virginia Beach, VA 137)  137  137  100% SUA;  Suffolk city (Virginia Beach, VA 66)  85  66  78% SUA;  James City County (Williamsburg city) (Williamsburg, VA 67; Virginia Beach, VA 4)  81  70  87% SUA;  York County (Poquoson city) (Virginia Beach, VA 64; Williamsburg, VA 9)  78  73  94% SUA;  Gloucester County (Virginia Beach, VA 9)  37  9  24% NSUA;  Isle of Wight County (Virginia Beach, VA 4)  35  4  12% NSUA.

RICHMOND, VA 1063/1063 (5/5): Henrico County (Richmond city) (Richmond, VA 498)  511  498  97% SUA;  Chesterfield County (Colonial Heights city) (Richmond, VA 315)  334  315  94% SUA;  Hanover County (Richmond, VA 61)  100  61  61% SUA;  Dinwiddie County (Petersburg city) (Richmond, VA 40)  60  40  66% SUA;  Prince George County (Hopewell city) (Richmond, VA 39)  58  39  67% SUA.

ROANOKE, VA 247/247 (2/2): Roanoke County (Roanoke city, Salem city) (Roanoke, VA 197)  214  197  92% SUA;  Botetourt County (Roanoke, VA 12)  33  12  36% SUA.

LYNCHBURG, VA 163/238 (2/3): Campbell County (Lynchburg city) (Lynchburg, VA 91)  130  91  70% SUA;  Bedford County (Bedford city) (Lynchburg, VA 14; Roanoke, VA 0)  75  14  19% NSUA;  Amherst County (Lynchburg, VA 12)  32  12  36% SUA.

CHARLOTTESVILLE, VA 142/142 (1/1): Albemarle County (Charlottesville city) (Charlottesville, VA 92)  142  92  65% SUA.

BLACKSBURG-CHRISTIANSBURG-RADFORD, VA 111/146 (1/2): Montgomery County (Radford city) (Blacksburg, VA 86; Roanoke, VA 1)  111  87  78% SUA;  Pulaski County (Blacksburg, VA 2)  35  2  7% NSUA.

HARRISONBURG, VA 125/125 (1/1): Rockingham County (Harrisonburg city) (Harrisonburg, VA 67)  125  67  53% SUA.

STAUNTON-WAYNESBORO, VA 119/119 (1/1): Augusta County (Staunton city, Waynesboro city) (Staunton--Waynesboro, VA 57)  119  57  48% SUA.

WINCHESTER, VA-WV 105/105 (1/1): Frederick County (Winchester city) (Winchester, VA 69)  105  69  66% SUA.

KINGSPORT-BRISTOL-BRISTOL, TN-VA 73/73 (1/1): Washington County (Bristol city) (Bristol--Bristol, TN--VA 33)  73  33  46% SUA.


West Virginia

CHARLESTON, WV 193/193 (1/1): Kanawha County (Charleston, WV 140)  193  140  73% SUA.

HUNTINGTON-ASHLAND, WV-KY-OH 194/194 (3/3): Cabell County (Huntington, WV--KY--OH 76)  96  76  78% SUA;  Putnam County (Huntington, WV--KY--OH 23; Charleston, WV 13)  55  36  64% SUA;  Wayne County (Huntington, WV--KY--OH 14)  42  14  33% SUA.

MORGANTOWN, WV 96/96 (1/1): Monongalia County (Morgantown, WV 70)  96  70  73% SUA.

BECKLEY, WV 125/125 (2/2): Raleigh County (Beckley, WV 48)  79  48  61% SUA;  Fayette County (Beckley, WV 16)  46  16  35% SUA.

HAGERSTOWN-MARTINSBURG, MD-WV 104/104 (1/1): Berkeley County (Hagerstown, MD--WV--PA 71)  104  71  68% SUA.

PARKERSBURG-VIENNA, WV 87/87 (1/1): Wood County (Parkersburg, WV--OH 60)  87  60  69% SUA.

WHEELING, WV-OH 78/78 (2/2): Ohio County (Wheeling, WV--OH 34)  44  34  77% SUA;  Marshall County (Wheeling, WV--OH 17)  33  17  51% SUA.

WEIRTON-STEUBENVILLE, WV-OH 55/55 (2/2): Hancock County (Weirton--Steubenville, WV--OH--PA 17)  31  17  55% SUA;  Brooke County (Weirton--Steubenville, WV--OH--PA 14)  24  14  58% SUA.

WASHINGTON-ARLINGTON-ALEXANDRIA, DC-VA-MD-WV 0/53 (0/1): Jefferson County (Hagerstown, MD--WV--PA 1)  53  1  1% NSUA.

CUMBERLAND, MD-WV 0/0 (0/0): None.

WINCHESTER, VA-WV 0/0 (0/0): None.


Wisconsin

MILWAUKEE-WAUKESHA-WEST ALLIS, WI 1556/1556 (4/4): Milwaukee County (Milwaukee, WI 946)  948  946  100% SUA;  Waukesha County (Milwaukee, WI 333)  390  333  85% SUA;  Washington County (West Bend, WI 68; Milwaukee, WI 23)  132  91  69% SUA;  Ozaukee County (Milwaukee, WI 65)  86  65  75% SUA.

MADISON, WI 488/582 (1/3): Dane County (Madison, WI 402)  488  402  82% SUA;  Columbia County ()  57  0  0% NSUA;  Green County ()  37  0  0% NSUA.

GREEN BAY, WI 248/248 (1/1): Brown County (Green Bay, WI 206)  248  206  83% SUA.

APPLETON, WI 226/226 (2/2): Outagamie County (Appleton, WI 128)  177  128  72% SUA;  Calumet County (Appleton, WI 25)  49  25  52% SUA.

RACINE, WI 195/195 (1/1): Racine County (Racine, WI 133; Milwaukee, WI 8)  195  141  72% SUA.

OSHKOSH-NEENAH, WI 167/167 (1/1): Winnebago County (Oshkosh, WI 74; Appleton, WI 63; Fond du Lac, WI 0)  167  138  82% SUA.

CHICAGO-NAPERVILLE-ELGIN, IL-IN-WI 166/166 (1/1): Kenosha County (Kenosha, WI--IL 124; Round Lake Beach--McHenry--Grayslake, IL--WI 24; Racine, WI 1)  166  149  89% SUA.

EAU CLAIRE, WI 161/161 (2/2): Eau Claire County (Eau Claire, WI 76)  99  76  77% SUA;  Chippewa County (Eau Claire, WI 27)  62  27  43% SUA.

JANESVILLE-BELOIT, WI 160/160 (1/1): Rock County (Janesville, WI 70; Beloit, WI--IL 45)  160  115  72% SUA.

WAUSAU, WI 134/134 (1/1): Marathon County (Wausau, WI 75)  134  75  56% SUA.

MINNEAPOLIS-ST. PAUL-BLOOMINGTON, MN-WI 0/125 (0/2): St. Croix County (Minneapolis--St. Paul, MN--WI 0)  84  0  0% NSUA;  Pierce County ()  41  0  0% NSUA.

SHEBOYGAN, WI 116/116 (1/1): Sheboygan County (Sheboygan, WI 71)  116  71  62% SUA.

LA CROSSE-ONALASKA, WI-MN 115/115 (1/1): La Crosse County (La Crosse, WI--MN 95)  115  95  83% SUA.

FOND DU LAC, WI 102/102 (1/1): Fond du Lac County (Fond du Lac, WI 55)  102  55  54% SUA.

DULUTH, MN-WI 44/44 (1/1): Douglas County (Duluth, MN--WI 27)  44  27  61% SUA.

[jimrtex]

Continuing on with edge cases in the Northeastern US.

Summary:

437 metropolitan counties in 11 southeastern states.
133 lack an urban core.

Of the 304 remaining, 232 would pass a 50K/50% threshold, 49 would fail a 25K/25% threshold, and 23 are in between.

Worcester, MA-CT/Windham, CT 33K/28%.  The Worcester Urbanized Area strings south along I-395.  The only real connection to Worcester is that it is the closest city along an interstate.  The second largest urban area in the county is Willimantic urban cluster in the extreme southeast.

St.Louis, MO-IL/Monroe, IL 10K/29%.   I-255 cuts through the northern tip of the otherwise rural county southeast of St.Louis.  Rather than providing access to the city, it is a bypass through Illinois for those travelling northward to Chicago.

Carbondale-Marion, IL/Jackson, IL 28K/47%.   Carbondale is in Jackson County, but the larger share of the urbanized area is in Williamson County, based on a combination of Marion, Cartersville, and Herrin.

Indianapolis, IN/Morgan 22K/32% and Boone 22K/38%.  Morgan and Boone are to the southwest and northwest of Marion County so are more peripheral to the metropolitan area.  Martinsviile and Lebanon urban clusters also hold down the urbanized population share in the two counties.

Terre Haute, IN/Clay 11K/39%.  The urbanized area extends along US 40 into Clay County.

Clarksville, TN-KY/Christian KY 20K/28%.  The Clarksville urbanized area takes in Fort Campbell, but the Hopkinsville urban cluster is separate, and holds the percentage down.

Baltimore, MD/Queen Anne's 12K/26%.  The Census Bureau's rules for delineating urban areas permits them to jump across undevelopable territory, in this case across Chesapeake Bay on the bridge.  The proximity of northern Anne Arundel to Baltimore means that it was part of the Baltimore metropolitan area, and prevented development of a separate Annapolis urban area.

California-Lexington Park, MD/St.Mary's 40K/38%. The Lexington Park-California-Chesapeake Ranch Estates Urbanized Area, straddles the St.Mary's-Calvert county line, which keeps the percentage down in St.Mary's which makes it one of the few metropolitan areas that would not have an associated urban county cluster.   The urbanized area is not the largest urban area in Calvert County (Cheasapeake Beach urban cluster is).   This results in Calvert and St.Mary's being treated separately, with Calvert becoming part of the Washington metropolitan area.

Lansing, MI/Eaton 49K/45% and Clinton 27K/36%.  The county seats of Charlotte and St. Johns hold the percentage down from the spillover of Lansing from Ingham into the adjacent counties.

Minneapolis-St.Paul, MN/Wright 34K/27% and Sherburne 26/29%.  Separate urban clusters for Monticello and Buffalo hold the percentage down as the urbanized area stretches along the Mississippi River on I-94 and US 20.

Mankato-North Mankato, MN/Nicollet 13K/41%.  St. Peter keeps the percentage down, and challenges the urbanized area for dominance.

Grand Forks, ND-MN/Polk, MN 8K/26%.  Crookston challenges East Grand Forks for county dominance.

La Crosse-Onalaska/WI-MN, Houston,MN 5K/28%.  A tiny bit of spillover across the river into an otherwise rural area.

New York-Newark-Jersey City, NY-NJ-PA/Sussex, NJ 37K/25%.  Sussex is the extreme northwestern corner of the state nearly 50 miles from Manhattan, and the urbanized area enters the southern part of the county along I-80.  Franklin-Highland Lake urban cluster keeps the percentage down.

Philadelphia-Camden-Wilmington, PA-NJ-DE-MD/Salem NJ 26K/40%.  The urbanized area includes Wilmington, DE and enters the county across the Delaware River on I-295 rather than southward from Camden.

Rochester, NY/Ontario 35K/32%.  The urbanized area strings down to Canandaiuga, but Geneva in the east edge of the county is a separate urban cluster.

Columbus, OH/Fairfield 49K/34%.  Lancaster urban cluster holds the percentage down - contrast with Licking, where it is the Newark urbanized area that qualifies the county.

Dayton, OH/Miami 47K/46%.  The urbanized area extends along I-75 to Tipp City and Troy, but not quite to Piqua, where the urban cluster drops the percentage a bit.

Wheeling, WV/Belmont, OH 28K/39%.  This is a bit of spillover across the Ohio River, with the urbanized area extending along US 40 to St.Clairsville.

Allenton, PA/Carbon 18K/27%.  The urbanized area extends along I-476 for a bit, and Hazelton urbanized area gives a bit of a boost to barely reach 25%.

Youngstown-Warren-Boardman, OH-PA/Mercer, PA 39K/33%.  Sharon and Hermitage, just inside the Pennsylvania line are about the same distance from Younstown as Warren is.

Gettysburg, PA/Adams 31K/31%.   Hanover urbanized area spills over into Adams County, qualifying it as central county of a metropolitan statistical area.   Meanwhile, Hanover urbanized area is secondary to York urbanized area in neighboring York County.  Metropolitan areas are named based on cities, rather than urban areas, so the metropolitan areas are York-Hanover and Gettysburg (which has 7600 persons).

Bloomsburg-Berwick, PA/Montour 8K/46% a bit of spillover goes a long way in a small county.

Sioux Falls, SD/Lincoln 22K/49% spillover into a previously rural county.

Sioux City, IA-NE-SD/Union, SD 6K/39%.  The urbanized area crosses the Big Sioux River to include the southern tip of the county.

Washington-Arlington-Alexandria, DC-VA-MD-WV/Fauquier, VA 21K/32%.  The urbanized area just reaches the county along US 15/US 29.

Roanoke, VA/Botetourt 12K/36%.  The utbanized area just spills over into the county.

Lynchburg, VA/Amherst 12K/36%.  Spillover into the county.

Kingsport-Bristol-Bristol, TN-VA/Washington + Bristol IC, VA 33K/46%.  Bristol, TN is sligntly larger than Bristol, VA.

Huntington-Ashland, WV-KY-OH, Wayne 14K/33%.  Wayne links Ashland along the south bank of the Ohio.

Beckley, WV/Fayette 16K/35%.  Urbanized area strings along US 19.

Eau Claire, WI/Chippewa 27K/43%.  Eau Claire is just south of the Chippewa line.

Summary through southeastern and northeastern United States (37 states plus DC).

1004 metropolitan counties.
282 rural.
564 pass 50K/50% test.
58 are edge cases.
100 fail 25K/25% test.

[muon2]

On the IL map you have the Marion-Carbondale area in the wrong counties. Jackson is one further south than you have colored.

[jimrtex]

On the IL map you have the Marion-Carbondale area in the wrong counties. Jackson is one further south than you have colored.

You know what would be handy.   When you corrected a map, if the references to the map in messages were automatically updated.

Thanks.  Both Marion and Jackson are edge cases, and so I had looked at the map both individually, and had not noticed.

Musing: If I had the lat/long, I could use Excel to produce a scatter chart.  I wonder how hard it would be get projections and scales to match up with a map.

[jimrtex]

Does this map comply with the US Constitution?  Yes.

In Wesberry v Sanders the SCOTUS held that "construed in its historical context, the command of Art. I, § 2 that Representatives be chosen 'by the People of the several States' means that, as nearly as is practicable, one man's vote in a congressional election is to be worth as much as another's."  "as nearly as is practicable" apparently was borrowed from an earlier federal statute.  As Justice Harlan noted in his dissent, it was as if the court had determined that the repealed law was not necessary in the first place.

In any event, the court did not define the meaning of the phrase.  Justice Harlan suggested that the court's decision might mean that all but 37 of the representative were illegitimate - 22 elected at large and 15 elected by district where the range was less than 100,000 (which was more than 20% at that time).

Of the remaining 5 districted states, all but North Dakota had a greater numeric range than the Alabama whole-county plan, and the relative range was much greater (North Dakota had fewer persons than any of the Alabama districts, with a relative range of 10.8%).

It is quite unlikely that the Alabama whole-county map would have been overturned in 1963.

In Kirkpatrick v Preisler, the SCOTUS considered a Missouri congressional district plan.  The District Court had found "that the state legislature had not relied on the census reports, but used less accurate data, that it had rejected a plan with smaller variances, and that, by simply switching some counties from one district to another, it would have produced a plan with markedly reduced variances".

The SCOTUS held that Art. I, § 2 required that "only the limited population variances which are unavoidable despite a good faith effort. to achieve absolute equality, or for which justification is shown", and also that there could be no de minimis threshold for equality, but that a State must justify any variance.

Since Missouri was doing a post-Wesberry redistricting, they had attempted to use population estimates, had rejected an alternative plan with less population variance, and that simply swapping counties would have reduced variance.

The premise behind the Alabama plan is that alternatives would have been proposed.  If they had less variance they might have been accepted.  If they had less variance and been rejected, it would have been for justifiable reasons, such as significantly less compactness, non-compliance with the VRA, or popular preference in the case of comparable plans.

The SCOTUS held that the "extent to which equality may practicably be achieved may differ from State to State and from district to district."   That is to say, just because Iowa may achieve a certain level of equality, does not mean that West Virginia or Alabama may be able to.  And in any case, Iowa or West Virginia or Alabama must justify any variance.

The Alabama plan has  less variance than that in Missouri.   Unlike in Missouri, there was a systematic effort to find an alternative map with less variance.  The actual census figures were used, rather than the post hoc rationalization presented by Missouri.  No consideration of politics was used beyond that used in checking the Black Belt district complied with the VRA.  Incumbents or potential candidates were not considered.  The legislature was completely removed from the redistricting process.  Consideration of demographics was limited to treating urban county clusters as a unit.   The UCC were constructed based on an objective measure, and weren't subjective interests that Missouri used.

Variance among districts in Alabama is quite small.  The following chart shows the population of each district as a circle with proportional area.



Alabama is justified in using whole counties because:

(1) It simplifies election administration.
(2) It permitted ordinary citizens to propose maps.
(3) It permitted ordinary citizens to evaluate their district, in determining the best plan.  If the Constitution commands that the people of a State elect their representatives, then surely it permits the people of the State to actively and effectively participate in defining the districts from which their representatives are elected.
(4) It acts as a constraint on gerrymandering.
(5) It permits more effective representation for voters, with less likelihood of confusion, and more influence over the election.  Dividing a county is not that dissimilar from cracking a racial minority among multiple districts.

Alabama is justified in recognizing urban county clusters as part of the redistricting process because:

(1) It makes it easier for voters to follow their race when there are fewer races competing for media attention.
(2) It may reduce campaign expenses.
(3) It makes it easier for representatives to represent there constituency, as opposed to requiring multiple offices in different cities which only serve part of each city.
(4) It reduces the risk of a metropolitan area being overrepresented or underrepresented, which is contrary to the whole purpose of having districts in the first place - so that people in various parts of the State have representation.

At the same time as Kirkpatrick v. Preisler, the SCOTUS in Wells v Rockefeller upheld a district court decision blocking use of congressional districts in New York.  New York had divided the state into seven multi-district regions, plus 10 multi-county regions electing a single member.  It then divided the multi-district regions into districts of almost equal population.  There was thus a high level of equality within regions, but not statewide, because there had been no attempt to equalize regional populations.

The regions and their populations were:

Nassau-Suffolk 4.81 (-0.19)
Queens 4.25 (+.25)
Kings 7.13 (+.13)
New York-Bronx 7.63 (-0.37)
Westchester-Putnam 2.05 (+.05+)
Monroe+others 2.00 (0.00).
Erie-Niagara 3.19 (+0.19).
Single member regions: 1.00, 0.97, 1.01. 1.04, 1.04, 0.94, 1.03, 1.01, 0.94, 0.93.

Only 9 districts were within 5% of the ideal.  As one might expect there was not only a large range of 13.095%, but a high standard deviation of 4.23%.   Because of the use of unequal regions, there was major systemic bias in the plan.

The Alabama plan is readily distinguishable, because the 5% limit for regions is simply a gross starting point for excluding plans.  Plans that simply tried to get regions within 5% would be badly beaten by alternatives.

In Karcher v Daggett the SCOTUS rejected a New Jersey plan that had just a 0.69% population range.  In doing so, they said that the legislature had not made a good faith effort to achieve population equality, or to justify their variation.  Like in the earlier cases, New Jersey made rationalizations such as the deviation was less than the error in the census.  Like in Missouri, political subdivisions were used (townships in this case), and simply swapping townships would have produced a more equal map.

The court offered this question, regarding whether a de minimis deviation could be established:  "Furthermore, choosing a different standard would import a high degree of arbitrariness into the process of reviewing apportionment plans. Ibid. In this case, appellants argue that a maximum deviation of approximately 0.7% should be considered de minimis. If we accept that argument, how are we to regard deviations of 0.8%, 0.95%, 1%, or 1.1%?"

New Jersey was looking for a safe harbor, like the 10% range used for legislative districts.

The SCOTUS did not say that the deviation in New Jersey was too large in absolute terms.  Quite to the contrary, they said that it was impossible to set such a limit.  But they did require justification.   New Jersey was able to offer only excuses, such as their deviation was less than the error in the census, or they had VRA concerns.  But the availability of plans with less deviation exposed their "justification" as mere rationalizations or excuses.

Subsequent to Karcher, lawyers began to advise states to have no deviation.  But that simply meant that they did not have to justify any deviation.  And besides, population equality provides greater gerrymandering opportunity.  Vieth v Bandemer demonstrated this effect.  An initial map had a deviation of 12.  But this was corrected, and the plan slid through the courts.

In Tennant v Jefferson County, the SCOTUS reiterated its test from Kirkpatrick and Karcher.   There are the two prongs to the test.

First, the plaintiff must demonstrate that there are population that “could practicably be avoided".  In West Virginia there were whole-county plans with less deviation than the State plan.  It is unclear whether the SCOTUS considered the "perfect plan" which had no deviation, but split counties as being "practicable".

The premise in Alabama is that there were no alternative whole-county plans that had significantly less deviation or greater compactness, or were preferred by voters.  But assume that there were.

Then the second part of the Karcher test comes into play.  The State must  "show with some specificity that the population differences were necessary to achieve some legitimate state objective.   This burden is a flexible one, which depends on the size of the deviations, the importance of the State’s interests, the consistency with which the plan as a whole reflects those interests, and the availability of alternatives that might substantially vindicate those interests yet approximate population equality more closely."

In West Virginia, the legislature justified its plan based on using whole counties, and making the minimum amount of changes from the existing plan.  The motivation for this was partly political - it would maintain the status quo, and could be passed by the legislature.  The district court faulted the legislature for not building a contemporaneous record of their justification.

In Alabama, the standards were established well in advance, political considerations other than the VRA were kept out of the process, and there was the opportunity for more equal alternatives to be considered.  If they had been significantly more equal, they would have replaced the plan that was adopted.  If they were about as equal, then the voters selected the final plan.

The Alabama plan vindicates the state interests of maintaining county boundaries, unnecessary splitting of urban concentrations of population, reasonable compactness, non-consideration of political results or incumbents, adherence to the voting rights act, and active participation of the voters in the process.

[muon2]

In justifying the Jefferson county district it seems that you would rely on the court to treat the language used for congressional districts in Wesberry, "as nearly as practicable," in the same way that the court views the language governing state legislative districts in Reynolds v Sims "substantial equality of population." The Reynolds phrase has come to be interpreted as allowing a range of up to 10%, and the range in the above AL congressional plan is 6.1%.

In contrast, SCOTUS has made clear that the two standards are not the same. In no subsequent case has the court ever allowed a range in excess of 1% under the Wesberry language, though ranges in excess of 10% using the type of argument made for the AL plan have succeeded under the Reynolds language. Even the exception phrases used in the two cases are different - "legitimate state objective" compared to "rational state policy." If the above AL argument were to succeed there would be no practical difference between the two standards in range or exception. Unless SCOTUS is willing to declare Wesberry void and replaced by Reynolds for congressional as well as state legislative districts I don't think the argument will stand.

[jimrtex]

In justifying the Jefferson county district it seems that you would rely on the court to treat the language used for congressional districts in Wesberry, "as nearly as practicable," in the same way that the court views the language governing state legislative districts in Reynolds v Sims "substantial equality of population." The Reynolds phrase has come to be interpreted as allowing a range of up to 10%, and the range in the above AL congressional plan is 6.1%.

In contrast, SCOTUS has made clear that the two standards are not the same. In no subsequent case has the court ever allowed a range in excess of 1% under the Wesberry language, though ranges in excess of 10% using the type of argument made for the AL plan have succeeded under the Reynolds language. Even the exception phrases used in the two cases are different - "legitimate state objective" compared to "rational state policy." If the above AL argument were to succeed there would be no practical difference between the two standards in range or exception. Unless SCOTUS is willing to declare Wesberry void and replaced by Reynolds for congressional as well as state legislative districts I don't think the argument will stand.

I deliberately only cited congressional cases.

Practicable means able to be put into practice.  The SCOTUS based on Tennant v Jefferson County apparently agrees that "practicable" does not necessarily encompass perfect-equality plans, when they rejected the District Court's assertion that because other State's didn't respect county lines, that West Virginia might not either.

In Kirkpatrick v Preisler and Karcher v Daggett, the SCOTUS has stressed that there is no de minimis standard for congressional districts.   Assuming no means no, that also means that there can be greater deviation in some cases, that might have been rejected in previous cases in other States.  Just as Iowa doesn't have to be as equal as most States, West Virginia does not have to be as equal as Iowa, and Alabama need not be as equal as West Virginia.

Legitimate State Objectives in Alabama:

Alabama is justified in using whole counties because:

(1) It simplifies election administration.
(2) It permitted ordinary citizens to propose maps.
(3) It permitted ordinary citizens to evaluate their district, in determining the best plan.  If the Constitution commands that the people of a State elect their representatives, then surely it permits the people of the State to actively and effectively participate in defining the districts from which their representatives are elected.
(4) It acts as a constraint on gerrymandering.
(5) It permits more effective representation for voters, with less likelihood of confusion, and more influence over the election.  Dividing a county is not that dissimilar from cracking a racial minority among multiple districts.

Not splitting counties is a legitimate objective.

Alabama is justified in recognizing urban county clusters as part of the redistricting process because:

(1) It makes it easier for voters to follow their race when there are fewer races competing for media attention.
(2) It may reduce campaign expenses.
(3) It makes it easier for representatives to represent there constituency, as opposed to requiring multiple offices in different cities which only serve part of each city.
(4) It reduces the risk of a metropolitan area being overrepresented or underrepresented, which is contrary to the whole purpose of having districts in the first place - so that people in various parts of the State have representation.

Not unnecessarily splitting urban county clusters is a legitimate state objective.

Alabama is justified in using standard deviation rather than range for measuring equality because it is a better measure of overall equality.  As the SCOTUS noted in Kirkpatrick, setting a de minimis range standard might encourage plans that sought to get just within the range limit.  Standard deviation is the best measure for determining whether there was a good faith effort to achieve equality among all districts in a State.

Alabama is justified in using a mild measure of compactness, and using simplified county boundaries for doing so.  The objective is not to avoid use of rivers or other sinuous boundaries, but to arrange whole counties in a relatively compact form that also achieves practicable equality.

Compliance with the Voting Rights Act is a legitimate state objective.

Ignoring political considerations other than the Voting Rights Act is a legitimate state objective, because it provides a disincentive to gerrymandering.

Providing for real and effective active participation of voters in proposing and selecting plans is a legitimate state objective, since it eliminates self-interested politicians from the process.  Basing plans on whole counties is essential to the participation model used in Alabama.

Which if any of these objectives do you believe to be illegitimate?

The Karcher opinion said, "[t]he showing required to justify population deviations is flexible, depending on the size of the deviations, the importance of the State's interests, the consistency with which the plan as a whole reflects those interests, and the availability of alternatives that might substantially vindicate those interests yet approximate population equality more closely. By necessity, whether deviations are justified requires case-by-case attention to these factors."

You certainly can't claim the standards weren't consistently applied.  And there were no alternatives which might substantially vindicate the legitimate state objectives, but better approximate equality.  If there had been, they would have been adopted.

[muon2]

But that doesn't change my central claim. If a court accepts your argument then they are functionally agreeing to standards no different than those that currently exist for state legislative districts. SCOTUS has said in the past that the standards are different, so to effectively make them the same is a reversal of that position. I think that the operative phrase would be "depending on the size of the deviations" and they would rest on that to require more rigorous justification. If you can show me how accepting your argument preserves separate standards for congressional and legislative redistricting I'm happy to revisit my statement.

I was happy to see them loosen the standards in Tennant, because I think that maintains the distinction between the congressional and legislative redistricting while recognizing the case-by-case nature presented in WV. WV successfully argued that whole counties and keeping voters in the same district were legitimate objectives. By your analysis they could have made the same claims with their 2000's map moving no people at all, but WV expected that such a plan would not be upheld with its 9.6% range. Why wouldn't WV be able to keep its old map using your argument?

[jimrtex]

But that doesn't change my central claim. If a court accepts your argument then they are functionally agreeing to standards no different than those that currently exist for state legislative districts. SCOTUS has said in the past that the standards are different, so to effectively make them the same is a reversal of that position. I think that the operative phrase would be "depending on the size of the deviations" and they would rest on that to require more rigorous justification. If you can show me how accepting your argument preserves separate standards for congressional and legislative redistricting I'm happy to revisit my statement.

The standard for congressional redistricting, is a relative one based on alternative plans.   For legislative plans the courts have accepted that there is an absolute safe harbor (though they have implied that may not always be true).

As a plaintiff you would have to demonstrate that the population differences could be practicably avoided.  You have not satisfied the first prong of the Karcher test.

I was happy to see them loosen the standards in Tennant, because I think that maintains the distinction between the congressional and legislative redistricting while recognizing the case-by-case nature presented in WV. WV successfully argued that whole counties and keeping voters in the same district were legitimate objectives. By your analysis they could have made the same claims with their 2000's map moving no people at all, but WV expected that such a plan would not be upheld with its 9.6% range. Why wouldn't WV be able to keep its old map using your argument?

If West Virginia had presented such a plan, I could produce a plan that shifted only Mason County and reduced the standard deviation from 3.93% to 0.37% an almost 11-fold improvement.

Not making any changes is NOT a legitimate state objective.  Making minimal changes is a legitimate state objective.

The SCOTUS did NOT change their standard for congressional districts.  They clarified them.  The SCOTUS did not even hear oral arguments in Tennant v Jefferson County

[muon2]

But that doesn't change my central claim. If a court accepts your argument then they are functionally agreeing to standards no different than those that currently exist for state legislative districts. SCOTUS has said in the past that the standards are different, so to effectively make them the same is a reversal of that position. I think that the operative phrase would be "depending on the size of the deviations" and they would rest on that to require more rigorous justification. If you can show me how accepting your argument preserves separate standards for congressional and legislative redistricting I'm happy to revisit my statement.

The standard for congressional redistricting, is a relative one based on alternative plans.   For legislative plans the courts have accepted that there is an absolute safe harbor (though they have implied that may not always be true).

As a plaintiff you would have to demonstrate that the population differences could be practicably avoided.  You have not satisfied the first prong of the Karcher test.

I was happy to see them loosen the standards in Tennant, because I think that maintains the distinction between the congressional and legislative redistricting while recognizing the case-by-case nature presented in WV. WV successfully argued that whole counties and keeping voters in the same district were legitimate objectives. By your analysis they could have made the same claims with their 2000's map moving no people at all, but WV expected that such a plan would not be upheld with its 9.6% range. Why wouldn't WV be able to keep its old map using your argument?

If West Virginia had presented such a plan, I could produce a plan that shifted only Mason County and reduced the standard deviation from 3.93% to 0.37% an almost 11-fold improvement.

Not making any changes is NOT a legitimate state objective.  Making minimal changes is a legitimate state objective.

The SCOTUS did NOT change their standard for congressional districts.  They clarified them.  The SCOTUS did not even hear oral arguments in Tennant v Jefferson County

Minimizing the shifted population is a legitimate state objective. The Cooper 3 plan that was presented were presented that made additional population shifts with a 21-fold reduction in range (27-fold in SD). If that reduction was not sufficient to justify the extra population shift, why would the initial reduction be required?

[jimrtex]

But that doesn't change my central claim. If a court accepts your argument then they are functionally agreeing to standards no different than those that currently exist for state legislative districts. SCOTUS has said in the past that the standards are different, so to effectively make them the same is a reversal of that position. I think that the operative phrase would be "depending on the size of the deviations" and they would rest on that to require more rigorous justification. If you can show me how accepting your argument preserves separate standards for congressional and legislative redistricting I'm happy to revisit my statement.

The standard for congressional redistricting, is a relative one based on alternative plans.   For legislative plans the courts have accepted that there is an absolute safe harbor (though they have implied that may not always be true).

As a plaintiff you would have to demonstrate that the population differences could be practicably avoided.  You have not satisfied the first prong of the Karcher test.

I was happy to see them loosen the standards in Tennant, because I think that maintains the distinction between the congressional and legislative redistricting while recognizing the case-by-case nature presented in WV. WV successfully argued that whole counties and keeping voters in the same district were legitimate objectives. By your analysis they could have made the same claims with their 2000's map moving no people at all, but WV expected that such a plan would not be upheld with its 9.6% range. Why wouldn't WV be able to keep its old map using your argument?

If West Virginia had presented such a plan, I could produce a plan that shifted only Mason County and reduced the standard deviation from 3.93% to 0.37% an almost 11-fold improvement.

Not making any changes is NOT a legitimate state objective.  Making minimal changes is a legitimate state objective.

The SCOTUS did NOT change their standard for congressional districts.  They clarified them.  The SCOTUS did not even hear oral arguments in Tennant v Jefferson County

Minimizing the shifted population is a legitimate state objective. The Cooper 3 plan that was presented were presented that made additional population shifts with a 21-fold reduction in range (27-fold in SD). If that reduction was not sufficient to justify the extra population shift, why would the initial reduction be required?

Not making any changes is not a good faith effort to achieve practicable equality.

If we were basing the decision on relative improvement, then we wouldn't bother with Cooper 3, we'd go right to "Perfect".  But the SCOTUS didn't really spend much time on "Perfect", even though the District Court said it was their inclination to impose Perfect or Cooper 4.

[muon2]

And I contend that have a range in excess of 1% is also not a good faith effort to achieve practicable equality, though it is to achieve substantial equality.

[jimrtex]

And I contend that have a range in excess of 1% is also not a good faith effort to achieve practicable equality, though it is to achieve substantial equality.

If you were the plaintiff, under the first prong of the Karcher test, you would "bear the burden of proving the existence of population differences that could practicably be avoided."

Also, the SCOTUS has said there is no de minimis standard for congressional equality.  That is, you can't claim that 1% is unacceptable on a universal basis.

[muon2]

And I contend that have a range in excess of 1% is also not a good faith effort to achieve practicable equality, though it is to achieve substantial equality.

If you were the plaintiff, under the first prong of the Karcher test, you would "bear the burden of proving the existence of population differences that could practicably be avoided."

Also, the SCOTUS has said there is no de minimis standard for congressional equality.  That is, you can't claim that 1% is unacceptable on a universal basis.

I think we will have to disagree on this and leave debate for another topic. I believe that SCOTUS will want to maintain a distinction between congressional and state plans that your argument would effectively negate. You could be right that SCOTUS wants to walk back Wesberry all the way to Baker v Carr, but we'll just have to wait to see who prevails.

[jimrtex]

And I contend that have a range in excess of 1% is also not a good faith effort to achieve practicable equality, though it is to achieve substantial equality.

If you were the plaintiff, under the first prong of the Karcher test, you would "bear the burden of proving the existence of population differences that could practicably be avoided."

Also, the SCOTUS has said there is no de minimis standard for congressional equality.  That is, you can't claim that 1% is unacceptable on a universal basis.

I think we will have to disagree on this and leave debate for another topic. I believe that SCOTUS will want to maintain a distinction between congressional and state plans that your argument would effectively negate. You could be right that SCOTUS wants to walk back Wesberry all the way to Baker v Carr, but we'll just have to wait to see who prevails.

You are misrepresenting my position entirely.   I have NOT asserted that the SCOTUS "wants to walk Wesberry all the way to 'Baker v Carr'".   There is clearly an effective difference between "are all districts within 5% deviation", and is "is this the most equal plan that is practicable, consistent with legitimate State interests."

Your view may reflect advice given to the Illinois legislature.  The lawyers might advise that there was no way that Illinois could justify drawing unequal districts so motivated by politics, or they might simply advise the legislature that congressional districts must have equal population, since this avoids a record of political intent, and they know that the legislature wants to do a political gerrymander.  In addition, the Illinois Supreme Court has imposed a much stricter standard for the legislature than the SCOTUS has generally.  It is quite unlikely that a legislative plan in Texas would be close to as equal as the Alabama plan.

State and Federal Law Governing Redistricting in Texas

As Justice Harlan pointed out in his Wesberry dissent, the SCOTUS did not define the term "as nearly as is practicable", which they borrowed from earlier congressional legislation, and then later had been deliberately excised.  As Harlan noted, it was as if the SCOTUS was ruling that the legislation was superfluous in the first place, and it didn't matter in that it was removed, or for that matter what Congress had meant by the phrase.  Such an interpretation essentially strips Congress of any authority to regulate districting.  If the only authority Congress has is to decide whether Iowa can have whole-county districts, then it really does not have any meaningful authority at all.

Since Wesberry did not determine the meaning of "as equal as practicable" other than Georgia was not, we have to look at other cases where the SCOTUS has given an interpretation, and set precedents.

I do believe that the SCOTUS wants to get out of the business of redistricting litigation, where the courts are used as a weapon.  In particular they have given increasing deference to the State legislatures.  Before a plaintiff would simply go to court to crash a plan, and get the court to draw a replacement plan from scratch.  But the SCOTUS has told the federal courts to respect the legislatures.  So even if West Virginia had lost, they would have let the legislature shift 3,196 persons out of District 2 into District 1 and 2.

And I suspect that the SCOTUS would attempt to refine their precedents.   Their decision in Tennant v Jefferson County was a clear rebuke of the District Court.  The District Court opinion was so sure that exact equality was required, and the SCOTUS didn't even bother to hear the case.  The essence of the decision was the District Court did not understand Karcher v Daggett.

In the case of Kirkpatrick v Preisler, the Missouri legislature had been playing games with the District Court (see the Fortas concurring opinion).  Much of their "justification" came after they were sued, and they couldn't show that they had attempted to account for population growth or non-residential population in any consistent manner.

In the case of Karcher v Daggett, the New Jersey legislature had slapped together a second map to assuage black Democrats, but had done so carelessly.  They couldn't show that their particular choice of townships had anything to do with the VRA, because it really didn't.  Citing census error was totally an excuse.  Even if you are working with estimates, you should still try to equalize the estimates, to ensure that you weren't increasing the error.

The important point to understand about Kirkpatrick and Karcher was NOT that the deviation in Missouri and New Jersey was too large.  It was that Missouri and New Jersey had failed to justify it, and that were other plans that satisfied the state's interests with less deviation.

In Karcher, the court said that that New Jersey "must show with specificity that a particular objective required the specific deviations in its plan".  

If no county is to be split, then there will be one district with a 3.6% deviation in any Alabama plan.  It is a requirement of a no-split-county plan that the Jefferson district have a 3.6% deviation.

The SCOTUS is not going to accept "politics as usual" as justification for much deviation at all.

But Alabama would not have been politics as usual.  The voters preferred my proposed plan.  That gets Kennedy's vote right there.

If the Alabama plan went before the SCOTUS there are several possible outcomes:

(1) The SCOTUS approves the map while writing glowingly of the process that produced the map.  They will quote Karcher

"The showing required to justify population deviations is flexible, depending on the size of the deviations, the importance of the State's interests, the consistency with which the plan as a whole reflects those interests, and the availability of alternatives that might substantially vindicate those interests yet approximate population equality more closely. By necessity, whether deviations are justified requires case-by-case attention to these factors."

They will note while the deviation in Alabama was somewhat large, that it was justified by the consistency that was applied by the State and the importance of real citizen involvement.

(2) The SCOTUS approves the Alabama map, but requires that population be added to the Jefferson district.  Alabama is no worse off than if it

Of course, what the SCOTUS should do is admit that the legal basis for Wesberry was flawed (read Justice Clark's concurring opinion).  The reason why the rationale for congressional redistricting standards is so hard to comprehend, is because there is no justification in the Constitution.

[muon2]

Of course, what the SCOTUS should do is admit that the legal basis for Wesberry was flawed (read Justice Clark's concurring opinion).  The reason why the rationale for congressional redistricting standards is so hard to comprehend, is because there is no justification in the Constitution.

And then presuming that SCOTUS followed Clark's opinion, they would by necessity walk it back to Baker v. Carr. As Clark put it "The trial court, however, did not pass upon the merits of the case, ...  I believe that the court erred in so doing. In my view, we should therefore vacate this judgment and remand the case for a hearing on the merits. At that hearing, the court should apply the standards laid down in Baker v. Carr."