Weighted Voting For Congress

[muon2]

And the districts with their populations:

CA (10)
   CA-Shasta 3368K
   CA-San Pablo Bay 3845K
   CA-Santa Cruz 4308K
   CA-San Joaquin Valley 4129K
   CA-San Emigdio 2538K (including LAC areas north and west of LA including pockets within LA)
   CA-San Gabriel 4735K (the area east and south of LA including the Torrance pocket)
   CA-Los Angeles 3793K (just the city)
   CA-Santa Ana 3010K
   CA-San Bernardino 4258K
   CA-Palomar 3270K

Does San Emigdio swing around to include Claremont and Pomona?

How far north does the Torrance pocket go?  Inglewood?  Culver City?

I used the Census CCDs in LAC to define the districts. San Emigdio has the following CCDs less the city of LA:
North Antelope Valley
South Antelope Valley
Newhall
San Fernando Valley
Agoura Hills-Malibu
Los Angeles (which includes Culver City among other communities)
Santa Monica

San Gabriel is the following CCDs:
Pasadena
Upper San Gabriel Valley
East San Gabriel Valley
Southwest San Gabriel Valley
South Gate-East Los Angeles
Whittier
Downey-Norwalk
Long Beach-Lakewood
Compton
Inglewood
South Bay Cities
Torrance
Palos Verdes

San Ferndando Valley CCD includes Burbank and Glendale.  That is the population I couldn't find.

I tried to do a two-way split of Los Angeles County in order to get a 2nd district in Northern California.   But that ended up having to include Marin and San Joaquin in the northern districts.  There just aren't enough people for two districts.   The "Northern Coast" has to take in everything in the Central Valley north of Sacramento, leaving the other district as Greater Sacramento (Sacramento, Stockton, and Davis).

I think my alternative will be 3 districts wholly in Los Angeles County.

I didn't want to pull Shasta into the Bay Area through Napa and Sonoma, and that left me with too much for Santa Cruz. Certainly one can make that shift if the priority is to keep LAC from combining with anything adjacent. It does make balancing a bit harder since LAC minus the city leaves two districts of about 0.8 of the quota, and Orange is about the same. CA already has an above average district size.

To get to 3 districts in LAC I would shift Napa, Sonoma, and Solano to Shasta (4401K). The complete SF UCC would be one district by shifting SF and San Mateo to San Pablo Bay (4336K). Then Santa Cruz can extend down to Santa Barbara and Ventura (4031K). I think that does the least damage to the CoIs.

[Хahar 🤔]

Having a district that connects San Francisco with San Luis Obispo County and not Alameda County or Marin County so obviously flies in the face of actual settlement patterns that it defeats the purpose of the whole assignment.

[muon2]

Having a district that connects San Francisco with San Luis Obispo County and not Alameda County or Marin County so obviously flies in the face of actual settlement patterns that it defeats the purpose of the whole assignment.

As I noted in the post above, the Bay Area is a challenge since the extended area is too large for one district, but when combined with SV and the Central Coast its big enough for two. There are basically two splits that balance population within the required range. One keeps the wine counties with the Bay Area, but that forces the SF peninsula to go with SV and the Central Coast, which you note concerns about. The other puts the wine counties with Sacto and northern CA and keeps the traditional Bay Area together, but then links SV with the Coast, potentially down to Ventura so that LAC can be a stand alone entity. A district with just SV and the Central Coast is only about 2784K, and while it would be permitted, it is quite small in population. Would the rest of the state be ok with that?

[Хahar 🤔]

When you say SV, do you mean Santa Clara County? If so, it's an inseparable part of the Bay Area; despite the Census Bureau's definition of the San Francisco metropolitan area as separate from the San Jose metropolitan area, there is no definition of the Bay Area that excludes Santa Clara County.

If shoehorning San Jose into a non-Bay Area district is off-limits (as it should be) there are then only three potential options:

1) Draw a district consisting only of the Central Coast
2) Connect the Central Coast to the San Joaquin Valley
3) Split the Central Coast between the north and the south, connecting the northern part to the Bay Area and the southern part to greater Los Angeles

The first solution is obviously unfeasible from a population standpoint. The second solution might keep the Bay Area and the Los Angeles area intact, but it connects two areas that are not connected either culturally or by major roads. That leaves the third option, which involves connecting the parts of the Central Coast that are in Northern California (Monterey County and points northward) with the Bay Area, while connecting the parts that are in Southern California (San Luis Obispo County and points southward) with greater Los Angeles. This line is basically non-negotiable; lumping in Salinas with a Southern California district is straightforwardly wrong, as is grouping San Luis Obispo with a Northern California district.

I haven't looked to see whether the numbers work by splitting the Central Coast like this. If they do not, then the only conclusion is that California simply does not work with this number of districts and this level of acceptable deviation.

[traininthedistance]

When you say SV, do you mean Santa Clara County? If so, it's an inseparable part of the Bay Area; despite the Census Bureau's definition of the San Francisco metropolitan area as separate from the San Jose metropolitan area, there is no definition of the Bay Area that excludes Santa Clara County.

If shoehorning San Jose into a non-Bay Area district is off-limits (as it should be) there are then only three potential options:

1) Draw a district consisting only of the Central Coast
2) Connect the Central Coast to the San Joaquin Valley
3) Split the Central Coast between the north and the south, connecting the northern part to the Bay Area and the southern part to greater Los Angeles

The first solution is obviously unfeasible from a population standpoint. The second solution might keep the Bay Area and the Los Angeles area intact, but it connects two areas that are not connected either culturally or by major roads. That leaves the third option, which involves connecting the parts of the Central Coast that are in Northern California (Monterey County and points northward) with the Bay Area, while connecting the parts that are in Southern California (San Luis Obispo County and points southward) with greater Los Angeles. This line is basically non-negotiable; lumping in Salinas with a Southern California district is straightforwardly wrong, as is grouping San Luis Obispo with a Northern California district.

I haven't looked to see whether the numbers work by splitting the Central Coast like this. If they do not, then the only conclusion is that California simply does not work with this number of districts and this level of acceptable deviation.

A ten-second look makes it pretty clear that, to comply with your Option 3, all muon would need to do is shift San Luis Obispo south- and such a move would actually decrease variance anyway.  

EDIT: I misread where San Francisco itself was, which complicates matters.  But presumably removing San Luis Obispo would render that split of the Bay Area less objectionable?

(Yes, muon, I still owe you my thoughts on Philly/Atlanta/Detroit.  Comin' soon.)

[jimrtex]

Here's my draft of the plan for all 100 districts.



CA (10)
   CA-Shasta 3368K
   CA-San Pablo Bay 3845K
   CA-Santa Cruz 4308K
   CA-San Joaquin Valley 4129K
   CA-San Emigdio 2538K (including LAC areas north and west of LA including pockets within LA)
   CA-San Gabriel 4735K (the area east and south of LA including the Torrance pocket)
   CA-Los Angeles 3793K (just the city)
   CA-Santa Ana 3010K
   CA-San Bernardino 4258K
   CA-Palomar 3270K

This is my alternative.





I wanted to have a district centered on the San Francisco Bay area, and wanted to keep the Los Angeles districts in the county.

Northern California 4402K
San Francisco Bay 4335K
Silicon Valley-Central Coast 4032K
San Joaquin Valley 4129K
Inland Empire 4432K
San Gabriel&Antelope Valleys 2796K
City of Los Angeles 4134K
Los Angeles South 2889K
Orange County 3010K
San Diego 3095K

Alternative names include:

Shasta;
Golden Gate;
Central Coast-Silicon Valley; Central Coast;

Local option for:

Imperial;
Alpine;
Mono and Inyo;
Sonoma, Napa, and Solano;
Amador;

Los Angeles enclaves;
Western Los Angles County (Malibu, Calabasas, etc.) as a block.
City on a boundary (may be subject to population limits).

[Хahar 🤔]

When you say SV, do you mean Santa Clara County? If so, it's an inseparable part of the Bay Area; despite the Census Bureau's definition of the San Francisco metropolitan area as separate from the San Jose metropolitan area, there is no definition of the Bay Area that excludes Santa Clara County.

If shoehorning San Jose into a non-Bay Area district is off-limits (as it should be) there are then only three potential options:

1) Draw a district consisting only of the Central Coast
2) Connect the Central Coast to the San Joaquin Valley
3) Split the Central Coast between the north and the south, connecting the northern part to the Bay Area and the southern part to greater Los Angeles

The first solution is obviously unfeasible from a population standpoint. The second solution might keep the Bay Area and the Los Angeles area intact, but it connects two areas that are not connected either culturally or by major roads. That leaves the third option, which involves connecting the parts of the Central Coast that are in Northern California (Monterey County and points northward) with the Bay Area, while connecting the parts that are in Southern California (San Luis Obispo County and points southward) with greater Los Angeles. This line is basically non-negotiable; lumping in Salinas with a Southern California district is straightforwardly wrong, as is grouping San Luis Obispo with a Northern California district.

I haven't looked to see whether the numbers work by splitting the Central Coast like this. If they do not, then the only conclusion is that California simply does not work with this number of districts and this level of acceptable deviation.

A ten-second look makes it pretty clear that, to comply with your Option 3, all muon would need to do is shift San Luis Obispo south- and such a move would actually decrease variance anyway. 

EDIT: I misread where San Francisco itself was, which complicates matters.  But presumably removing San Luis Obispo would render that split of the Bay Area less objectionable?

(Yes, muon, I still owe you my thoughts on Philly/Atlanta/Detroit.  Comin' soon.)

If the numbers check out, then moving San Luis Obispo out would work, yes. Having Marin in the same district as Alameda County but not San Francisco is not ideal, but it's acceptable.

This map doesn't work at all for reasons that have already been discussed. There's no possible configuration in which Palo Alto and Thousand Oaks can belong to the same district.

[muon2]

When you say SV, do you mean Santa Clara County? If so, it's an inseparable part of the Bay Area; despite the Census Bureau's definition of the San Francisco metropolitan area as separate from the San Jose metropolitan area, there is no definition of the Bay Area that excludes Santa Clara County.

If shoehorning San Jose into a non-Bay Area district is off-limits (as it should be) there are then only three potential options:

1) Draw a district consisting only of the Central Coast
2) Connect the Central Coast to the San Joaquin Valley
3) Split the Central Coast between the north and the south, connecting the northern part to the Bay Area and the southern part to greater Los Angeles

The first solution is obviously unfeasible from a population standpoint. The second solution might keep the Bay Area and the Los Angeles area intact, but it connects two areas that are not connected either culturally or by major roads. That leaves the third option, which involves connecting the parts of the Central Coast that are in Northern California (Monterey County and points northward) with the Bay Area, while connecting the parts that are in Southern California (San Luis Obispo County and points southward) with greater Los Angeles. This line is basically non-negotiable; lumping in Salinas with a Southern California district is straightforwardly wrong, as is grouping San Luis Obispo with a Northern California district.

I haven't looked to see whether the numbers work by splitting the Central Coast like this. If they do not, then the only conclusion is that California simply does not work with this number of districts and this level of acceptable deviation.

A ten-second look makes it pretty clear that, to comply with your Option 3, all muon would need to do is shift San Luis Obispo south- and such a move would actually decrease variance anyway.  

EDIT: I misread where San Francisco itself was, which complicates matters.  But presumably removing San Luis Obispo would render that split of the Bay Area less objectionable?

(Yes, muon, I still owe you my thoughts on Philly/Atlanta/Detroit.  Comin' soon.)

If the numbers check out, then moving San Luis Obispo out would work, yes. Having Marin in the same district as Alameda County but not San Francisco is not ideal, but it's acceptable.

Moving SLO south shouldn't be a problem. The underlying issue in the south is Ventura which for large areas must be in the south, but even with SB and SLO is too small to stand on its own and connecting Ventura to Kern doesn't make any more sense than connecting it to SV. That's the main reason I opted to have one LAC district spread west along the coast. The CoI needs were more important than maintaining LAC with districts unique to it.

Edit - Part of my reason for my original split vs Xahar's comments is that the LA orbit keeps shifting north. When I was in college in the 70's my SoCal friends said that Santa Barbara was more Central Coast than LA. By the 90's colleagues from the area said that the natural division was at Point Conception. When I did some of my maps on the forum a few years ago, I was told that it was better to split SB from SLO, than SLO from Monterey. Now it seems like the line has shifted again.

[jimrtex]

This map doesn't work at all for reasons that have already been discussed. There's no possible configuration in which Palo Alto and Thousand Oaks can belong to the same district.

If Ventura was given its choice of district:

(1) Silicon Valley-Central Coast;
(2) San Joaquin Valley;
(3) Orange County;
(4,5) One of the suburban LA districts;

which would they choose?

[jimrtex]

CA (10)
I didn't want to pull Shasta into the Bay Area through Napa and Sonoma, and that left me with too much for Santa Cruz. Certainly one can make that shift if the priority is to keep LAC from combining with anything adjacent. It does make balancing a bit harder since LAC minus the city leaves two districts of about 0.8 of the quota, and Orange is about the same. CA already has an above average district size.

My emphasis.

California had 11 districts in 1990, but lost one when burgeoning populations in mid-sized Georgia, North Carolina, and Arizona caused it to miss the last seat.  It had barely added the 11th seat in 1990, when it added two, so that was somewhat of an overshoot.

It is conceivable that the apportionment rule could grandfather in the number of districts from a previous apportionment, so long as the individual districts remain within some national range.

In 1789, Congress apportioned twice as many representatives as there were states (13x2 = 26).   When a new state joined the Union, they were given one representative.  After the next census, additional representatives were apportioned.

California had one representative from statehood through 1880, and two representatives form 1890 through 1910, at which time it began to add representatives almost every decade.

1920: 3
1930: 4
1940: 4
1950: 6
1960: 7
1970: 8
1980: 9
1990: 11
2000: 10
2010: 10

Other states:

Alabama had one representative from statehood in 1819 through 1830; then two representatives from 1840 through 1990; it has had one representative since 2000, and it has been 101st in line for an additional representative.

Arizona gained its 2nd representative in 2000.

Arkansas had one representative from statehood in 1836 through 1890; it had two representatives in 1900 and 1910; and then dropped back to one.

[jimrtex]

Here's my draft of the plan for all 100 districts.

CO (2)
   CO-Red Rocks 2490K
   CO-Front Range 2540K

Your map doesn't include Broomfield, but you put in the Denver district.  Your names are not clearly associated with either district.

Very few people live in the Front Range, but they live along the Front Range.  It is conventionally applied from Wyoming to Colorado Springs, though the Front Range does not extend that far south (Pikes Peak is an outlier).  It sounds better than I-25 corridor.  Most of the people in the area live in the Denver district.

A particular instance of the Fountain Formation is near Denver, the most significant instance is near Colorado Springs.

Alternative names:

Denver; Mile High;

Colorado; Centennial State; Mountains&Plains.

History:

Colorado gained its 2nd representative in 2010.

Connecticut had two representatives in 1800.  It was the 8th most populous state in 1790, but had dropped to 31st by 1910.  It slowly climbed the ladder, reaching 24th in 1970 as a result of New York exurban growth into western Fairfield County.  At that point it was only 4 places from regaining a 2nd district.   Since then it has faded to 29th, and has averaged 0.4% annual growth over the 40 years from 1970-2010.

Delaware has always had one representative.

Congress continued to include the population of the District of Columbia in the Maryland (and Virginia) populations.  At first it was a tiny portion of the population, but later became critical to Maryland maintaining two districts.

[muon2]

Here's my draft of the plan for all 100 districts.

CO (2)
   CO-Red Rocks 2490K
   CO-Front Range 2540K

Your map doesn't include Broomfield, but you put in the Denver district.  Your names are not clearly associated with either district.

Very few people live in the Front Range, but they live along the Front Range.  It is conventionally applied from Wyoming to Colorado Springs, though the Front Range does not extend that far south (Pikes Peak is an outlier).  It sounds better than I-25 corridor.  Most of the people in the area live in the Denver district.

A particular instance of the Fountain Formation is near Denver, the most significant instance is near Colorado Springs.

Alternative names:

Denver; Mile High;

Colorado; Centennial State; Mountains&Plains.

History:

For CO I felt that names should reflect mountain features. I own property in CO, and in my experience most people in the state would recognize Red Rocks as the jewel of the Denver Mountain Park system with its world famous amphitheater. Though the Front Range extends into WY it is most associated with CO. It's visible from the eastern plains, the home to most of the population even without the Denver metro, and dominates transportation from the western part of the state to the population centers.

[jimrtex]

Here's my draft of the plan for all 100 districts.

CO (2)
   CO-Red Rocks 2490K
   CO-Front Range 2540K

Your map doesn't include Broomfield, but you put in the Denver district.  Your names are not clearly associated with either district.

Very few people live in the Front Range, but they live along the Front Range.  It is conventionally applied from Wyoming to Colorado Springs, though the Front Range does not extend that far south (Pikes Peak is an outlier).  It sounds better than I-25 corridor.  Most of the people in the area live in the Denver district.

A particular instance of the Fountain Formation is near Denver, the most significant instance is near Colorado Springs.

Alternative names:

Denver; Mile High;

Colorado; Centennial State; Mountains&Plains.

History:

For CO I felt that names should reflect mountain features. I own property in CO, and in my experience most people in the state would recognize Red Rocks as the jewel of the Denver Mountain Park system with its world famous amphitheater. Though the Front Range extends into WY it is most associated with CO. It's visible from the eastern plains, the home to most of the population even without the Denver metro, and dominates transportation from the western part of the state to the population centers.

Why wouldn't you consider the parks on Mount Evans, the most prominent peak on the Front Range, the jewels of the Denver Mountain Park system?

Summit Lake is the jewel of Denver's century-old mountain-park system

About half of the schools in the Front Range league are in the Denver area.  "Front Range" is either too exclusive, representing the area north of Denver (Longmont, Loveland, Fort Collins, etc.), an area that represents perhaps 1/3 of the district; or too inclusive including Denver, as a way of saying "Greater Denver" without really saying so, which puts about 2/3 of the population of the area in another district.

[jimrtex]

Edit: Moved Hernando and Citrus to North Florida to keep Tampa Bay within limits.

FL (5)
   FL-Apalachicola 3326K
   FL-Tampa Bay 3634K
   FL-Cape Canaveral 4747K
   FL-Miami-Dade 2496K
   FL-Everglades 4598K

I think you shifted the population from Polk, Hernando, and Citrus from Tampa Bay into Cape Canaveral.

In any event, I think that all of the west coast belongs with Tampa Bay, and Key West with Miami.  To make that possible, I did move Polk to the east.  It is in between Orlando and Tampa.  Though the southeastdistrict is really Fort Lauderdale and Palm Beach, I included all of the counties that share Lake Okeechobee, as well as Highlands because it makes a neater border.  I also shifted Marion.  It is getting a little distant from Orlando, and it makes for a bit more population balance, particularly with Polk moved east.



North Florida 3971
Tampa Bay 4476
Central Florida 4102
Southeast Florida 3683
Miami 2570

Names

Apalachicola is too exclusive for a district that stretches from Pensacola to Jacksonville to Ocala.

Gulf Coast might also be OK.

Cape Canaveral is maybe, but excludes the largest city.

Fort Lauderdale-Palm Beach, or Okeechobee is also possible.  Everglades might also refer to a couple of other districts.

Miami-Dade is too specific to local government structures.  "Miami-Dade" might just mean no one had heard of Dade County, but everyone knows where Miami is.

History

Florida was long a backwater, and was 28th ranked as late as 1940.

1950: It jumped to 21st and gained its 2nd representative.
1960: It jumped to 10th, skipping over 6 Southern states.
1970: To 9th and gaining a 3rd representative.
1980: To 7th and adding a 4th representative.
1990: To 4th and adding a 5th representative.
2000: No change.
2010: No change (between 1990 and 2010, Florida cut the gap with New York from 5 million to 500,000.

[jimrtex]

Georgia is, for all its zillions of counties, really easy to do.  And even with the mandate for coarse equality it all ends up being well within plus or minus 10 percent.



District 1: SOUTH GEORGIA.  Population 3,013,994 (deviation -215,224).  Obama 45.7%, Dem 48.0%.  36% Black (34% Black VAP).  I used the Atlanta media market as the dividing line here: everything south of it is in 1, everything within it (plus the few peripheral northern counties in other markets such as Rome) in 2 and 3.  Likely R, there's an opening for a Blue Dog in a good year here, but it's probably a narrow one.

District 2: ATLANTA.  Population 3,365,297 (deviation +136,079).  Obama 61.4%, Dem 58.2%.  40W/38B/13H (44W/37B/11H VAP), so min-maj.  These five counties were the original Atlanta metro area in 1950 and I imagine they're still considered to be the core of it today.  Obviously the exurbs spill far out into District 3 by now.  Safe D.

District 3: NORTH GEORGIA.  Population 3,308,362 (deviation +79,144).  Obama 36.6%, Dem 35.9%.  Pretty self-explanatory.  Safe R.

Here's my draft of the plan for all 100 districts.



GA (3)
   GA-Blue Ridge 1688K
   GA-Kennesaw 4901K
   GA-Okefenoke 3099K

This is somewhat of compromise between the two plans.  To achieve more population balance,  Train closely limited the Atlanta district.   To be more inclusive, Muon left the northern district severely underpopulated, and also ended up with a narrow southern and western wisp of a doughnut.

I started out with Train's map, adding up the population for a couple of tiers of northern counties.   I then took his Atlanta core, and began adding suburban counties that contained over 100,000 persons and had multiplied in population over the past few decades (5X over 30 years is not uncommon).   I realized I was quickly eating into the population needed for the northern district.

There are too many people in the Atlanta area for a single district; but if you tried to put them in two districts, you end up with a district with Dalton and Valdosta together.  So I instead did an asymmetric split, with northern counties such as Cherokee, Forsyth, and Gwinnett going into the northern district, and letting the Atlanta district to extend down to meet the South Georgia district.

If you count Paulding, Bartow, and Walton along with the other 3, about half the population of the North Georgia district is in the Atlanta metropolitan area, but it also has a regional core arcing from Rome and Dalton to Hallsville and Athens.



North Georgia 2982
Atlanta 3539
South Georgia 3167

History

Georgia had one representative from 1790 through 1810 when it gained its second.  Because of its large area (it is the largest state in the cis-Mississippi) it was able to maintain a population share as a rural area.   The booming Atlanta area resulted in a 3rd district in 2000.

It is likely that an initial split would have been Northwest/Southeast with the upland district extending down the Alabama border.  Over time it might have shifted to more of a north-south configuration.   As time went on the border may have moved northward, particularly after Atlanta began to expand (in 1950 Cobb had only reached 61K) so as to keep 1/3 of the population in the south.   So it would probably have made sense to carve an Atlanta district out of the much more populous northern district.

Hawaii and Idaho have always had one representative.

[muon2]

There seems to be a bit of a shift on the criteria for the districts. The initial rules said nothing about shape and provided for very loose conditions for connectivity. Discontiguous districts were even suggested if justified. OTOH, community of interest was made a strong requirement. Population equality is expected to take a back seat to CoI as well.

You must be logged in to read this quote.

The GA plan may be a compromise, but it doesn't particularly follow the rules. If the split between the Atlanta UCC and the rest of north GA is too unequal, then there should be a CoI justification. I would suggest using minority representation based on BVAP. This map shows counties shaded for 25-33.3%, 33.3-40%, 40-50%, and 50%+.



If one selects the Atlanta UCC counties that exceed 33.3% BVAP, those seven counties have a population of 2393K. That is 0.74 of the quota size for GA so it exceeds the 2/3 rule. The remainder in the north is 4253K and is 1.32 of the GA quota which is less than 4/3. This is a compromise in the spirit of the rules.

Since I abhor dull names for this exercise, I prefer geographic names that would be identifiable by residents of the state even if they are not inclusive of the entire area. Naming the district after a famous person from the district would beat a dull name here. Directional names should only be used when they are part of a specific geographic feature, and city or county names should be reserved for districts that comprise only the city or county in question. My revised GA plan based on the above CoI would be as follows.



Kennesaw Mt: 4253K, Obama 34.7%, McCain 64.4%
King: 2393K, 46.1% BVAP, Obama 68.6%, McCain 30.8%
Ocmulgee: 3041K, Obama 45.6%, McCain 53.8%

[jimrtex]

There seems to be a bit of a shift on the criteria for the districts. The initial rules said nothing about shape and provided for very loose conditions for connectivity. Discontiguous districts were even suggested if justified. OTOH, community of interest was made a strong requirement. Population equality is expected to take a back seat to CoI as well.

You must be logged in to read this quote.

The example of a discontiguous district was when another (central) district spanned the state cutting another district into parts.  Think Massachusetts and Boston.   A disconnected district might link Whatcom and Skagit with Okanogan, if the other district was very Seattle-Tacomia centric.

Conventional sensibility would be try to make the districts quite similar in population; but then you risk a situation like in Cortland County.  I was trying to discourage that.

If the districts are within a range of 2/3 to 1-1/3 of the quota, it ensures that the largest district is not more than twice the size that of the smallest district.  That would be at the extreme end of "coarse equality", particularly since it is somewhat unlikely that you would be at both extremes.

This is also consistent with the overall apportionment which used the harmonic mean as the divisor, rather than the geometric or arithmetic mean.  If I were implementing this scheme for a state legislature, I would expect that districts would be limited to the range of 2/3 to 1-1/3 of the statewide quota, with districts for the most part being unchanged so long as they remained within those limits.

I don't want to say never ever go outside those limits, but I don't think that Georgia provides a compelling enough case.

One particular county that resulted in my switch was Carroll County, which has double in population to over 110,000 in the 3 decades from 1980-2010, with the growth concentrated on the panhandle crossed by I-20.  It is thus similar to Sherburne County in Minnesota - where interstate access makes longer distance commuting possible.  So you have cut the doughnut along the Alabama line.

I don't see the COI that the southern fringe of your district has with Rome and Dalton.  Were they "North Georgia" before someone dropped the A(tlanta)-Bomb on the area?

You must be logged in to read this quote.

Does race form a geographic COI?

You must be logged in to read this quote.

Vice President King was from Alabama.

The founders were strict republicans.  I suspect they would have established a pattern of dull names.

And why give preference to the Ocmulgee over the Savannah or Chattahoochee?

[jimrtex]

Here's my draft of the plan for all 100 districts.
IL (4)
   IL-Chicago 2696K
   IL-Cook 2499K (all of the county except the city)
   IL-Fox and Kankakee 3505K
   IL-Great Rivers 4131K

Alternative names:

Chicago;
Cook County;
Chicagoland;
Illinois, Prairie State; Land of Lincoln.

History

Illinois underwent Floridian growth, jumping from 20th to 4th between 1830 and 1860.

Illinois gained its 2nd representative in 1850, after narrowly missing in 1840.  It gained a 3rd in 1860, and a 4th in 1870.  It flirted with a 5th, before finally reaching that level in 1900.  It moved somewhat closer to a 6th before fading, and dropping down to 4 districts in 1980.

Presumably it was a 3:2 split before 1980, with the two non-Chicago districts being somewhat smaller.  It would make sense to merge them, and perhaps shed some counties such as DeKalb, Grundy, and Kankakee.

[muon2]

There seems to be a bit of a shift on the criteria for the districts. The initial rules said nothing about shape and provided for very loose conditions for connectivity. Discontiguous districts were even suggested if justified. OTOH, community of interest was made a strong requirement. Population equality is expected to take a back seat to CoI as well.

You must be logged in to read this quote.

The example of a discontiguous district was when another (central) district spanned the state cutting another district into parts.  Think Massachusetts and Boston.   A disconnected district might link Whatcom and Skagit with Okanogan, if the other district was very Seattle-Tacomia centric.

Conventional sensibility would be try to make the districts quite similar in population; but then you risk a situation like in Cortland County.  I was trying to discourage that.

If the districts are within a range of 2/3 to 1-1/3 of the quota, it ensures that the largest district is not more than twice the size that of the smallest district.  That would be at the extreme end of "coarse equality", particularly since it is somewhat unlikely that you would be at both extremes.

This is also consistent with the overall apportionment which used the harmonic mean as the divisor, rather than the geometric or arithmetic mean.  If I were implementing this scheme for a state legislature, I would expect that districts would be limited to the range of 2/3 to 1-1/3 of the statewide quota, with districts for the most part being unchanged so long as they remained within those limits.

I don't want to say never ever go outside those limits, but I don't think that Georgia provides a compelling enough case.

My confusion was with the text is your guidelines that I have bolded. My initial split had all the Atlanta UCC in one district, but that went beyond 4/3. Even though it seems to me that a whole UCC would fall within the meaning of a CoI and I set a hard upper limit of the population of the smallest state that was apportioned two seats, I can accept that the 4/3 of the quota is a much harder limit than I read it. If you can give an example of where you would accept a district in excess of 4/3 the quota it would help.

That then leaves the question as to how the Atlanta metro should split. The metro split you suggest is not based on any CoI, so I'm think some effort should go to see where there might be a CoI that keeps the pieces within the 2/3 - 4/3 quota range.

You must be logged in to read this quote.

The north-south boundary in GA is pretty well accepted and runs along the fall line that separates the Piedmont from the Coastal Plain. The mountains of the north form a much smaller division. The fall line runs from Columbus through Macon to Augusta. I used the metro areas of the fall line as the northern edge of the south, which also kept the Black Belt intact in the southern district.

The remainder becomes the northern two districts. Merriwether. Pike and Lamar are on the southern edge of the Atlanta metro as much as Pickens and Dawson are on the northern edge. They are all semi-rural counties in the metro with commuting populations. If the population split were acceptable, I would have put Rome, Dalton, Athens, and everything else north and east of the Atlanta metro in one district and the metro in another, but I can't.

You must be logged in to read this quote.

Does race form a geographic COI?

[/quote]

This was the question I posed a couple of weeks ago to train. He was using media markets while I was using census-based metro areas. Atlanta metro is large so the question of internal CoI comes up when looking for a split. Areas of a high proportion of a demographic group are very much what the states have used and the courts supported for CoI. Race is one of those demographic groups that has meaning in Atlanta. I used a rational basis that could get public support, and why not name it after the most prominent resident recognized with a national holiday.

[muon2]

The example of a discontiguous district was when another (central) district spanned the state cutting another district into parts.  Think Massachusetts and Boston.   A disconnected district might link Whatcom and Skagit with Okanogan, if the other district was very Seattle-Tacomia centric.

Isn't this the same case for south FL? Key West is less a part of Miami than other nearby areas. The natural CoI is Miami-Dade/Broward/Palm Beach which is all a dense metro area, but that's too big. The Keys have more in common with the other more vacation-oriented areas of south FL, and they aren't even in the Miami CSA. I expect that in a plebiscite Monroe would rather not be in a district completely dominated by Miami-Dade.

[jimrtex]

There seems to be a bit of a shift on the criteria for the districts. The initial rules said nothing about shape and provided for very loose conditions for connectivity. Discontiguous districts were even suggested if justified. OTOH, community of interest was made a strong requirement. Population equality is expected to take a back seat to CoI as well.

You must be logged in to read this quote.

The example of a discontiguous district was when another (central) district spanned the state cutting another district into parts.  Think Massachusetts and Boston.   A disconnected district might link Whatcom and Skagit with Okanogan, if the other district was very Seattle-Tacomia centric.

Conventional sensibility would be try to make the districts quite similar in population; but then you risk a situation like in Cortland County.  I was trying to discourage that.

If the districts are within a range of 2/3 to 1-1/3 of the quota, it ensures that the largest district is not more than twice the size that of the smallest district.  That would be at the extreme end of "coarse equality", particularly since it is somewhat unlikely that you would be at both extremes.

This is also consistent with the overall apportionment which used the harmonic mean as the divisor, rather than the geometric or arithmetic mean.  If I were implementing this scheme for a state legislature, I would expect that districts would be limited to the range of 2/3 to 1-1/3 of the statewide quota, with districts for the most part being unchanged so long as they remained within those limits.

I don't want to say never ever go outside those limits, but I don't think that Georgia provides a compelling enough case.

My confusion was with the text is your guidelines that I have bolded. My initial split had all the Atlanta UCC in one district, but that went beyond 4/3. Even though it seems to me that a whole UCC would fall within the meaning of a CoI and I set a hard upper limit of the population of the smallest state that was apportioned two seats, I can accept that the 4/3 of the quota is a much harder limit than I read it. If you can give an example of where you would accept a district in excess of 4/3 the quota it would help.

I've been rethinking a more formal standard.

Besides voting on formal motions, representatives make speeches, persuade colleagues, serve on committees, provide constituent services, are elected, and represent their communities.  In these roles they have equal weight, and so should represent districts of roughly comparable size.   The 2/3 to 4/3 standard ensures that no representative serves more than twice as many constituents/voters, and the deviation above and below the mean is equal, and is unlikely to have remarkably greater persuasive capacity based on his voting strength, nor that he has grossly greater difficulty providing constituency services, etc.

But the 2/3 to 4/3 standard should be applied to the whole body, rather than within individual areas.

Within this band, it is desirable to have a large degree of random variation.  Weighted voting breaks down when there are fewer members, or fewer weights.   Hopefully districts from 50 states, divided on the basis of community of interest (and whole counties) will mitigate this problem.

The mean population for 100 districts is 3,087,455 (I include DC with MD), and the 2/3 to 4/3 limits are 2,058,304 to 4,116,607.

14 small states (WV, NE, ID, HI, ME, NH, RI, MT, DE, SD, AK, ND, VT, WY) are below the limit, and
4 large undivided states are above the limit (AL, SC, LA, KY).  These are acceptable exceptions based on the constitutional requirement to provide representation to each state. 

The smaller states are not necessarily without influence, but it does reduce the effective size of the body for weighting purposes.   On the other hand, had weighting been the practice, the constitution might have been amended to provide representation for territories.

Virginia with two districts is barely above twice the upper maximum.  To keep both districts above the lower threshold requires at worst a 51.5%:48.5% split (as it turns out, your proposal actually complies with a 50.5%:49.5% split).   Over the decades, it might require continued fidgeting with the boundaries.  It might also produce a bright spot of districts right at the limits.  On the other hand, if the statewide mean were used, a largest district could  have about 16/9, or 1.78% of the national mean.

So there might be a relaxed standard for states on the edges:

min( mean_state∙⁹/₁₀, mean_nation∙²/₃ )

max( mean_state∙¹¹/₁₀, mean_nation∙⁴/₃ )

Were this a state legislature we might also combine counties:

HI-AK
ID-MT-WY
ME-NH-VT
NE-SD-ND
MD-DE (retaining 2 districts)
CT-RI (adding district, splitting CT with eastern part added to RI)
KY-WV (adding district, splitting KY with eastern part added to WV)
Adding districts to CA(11 total), TX(8), NY(6), FL(6), OH(4), AL(2), SC(2), LA(2)

This would leave 13 single member districts:

OR, OK, NE-SD-ND, ME-NH-VT, ID-MT-WY, IA, NS, AR, KS, UT, NV, HI-AK, NM.

All would be within the 2/3 to 4/3 of the national average.

The mean for the 87 districts in the multi-district areas is 3,123,004; for the 13 single-district areas 2,974,121; vs 3,087,455.  Similar results should be achievable for a state legislature.

LA and VA would be subject to the relaxed limits that would permit a district slightly outside the national limits.

For our purposes, and to avoid redistricting of 10 states, we can use the mean of the multi-district states, which is 3,394,813 (the mean for the single district states is 2,297,107).

The 4/3 and 2/3 limits are 4,526,417 to 2,263,209

For your proposal this would put the following out of range:

CA San Gabriel above.  This could be fixed by moving Pasadena and at least its northern neighbors to San Emigdio.

WA Columbia is underpopulated.  Moving northward to include Lewis and the coast could put it into range, as would an alternative of a very Seattle-centric district (4 counties).

TX Western Gulf below and Permian Basin above.  I would have created a Central Texas district (San Antonio and Austin), with a Borders district from El Paso to Brownsville and up to Corpus Christi, with the remainder of the Western Gulf being moved to East Texas.  This might leave the remainder of West Texas underpopulated.

MN Itasca is underpopulated.  Moving Sherburne and perhaps Wright fixes that.

MO Ozarks is underpopulated.  Moving Vernon and Bates makes that within range.  Is there something distinctively "northern" about them?  Coming across the north of the state to connect St.Louis and Kansas City is pretty radical as it is, is there a reason to extend the finger further south?  I'd like to see an east-west split, and perhaps a St.Louis-Kansas City district (two unconnected metro areas), or a Missouri River district with northern and southern rural areas separated.   The Missouri River district would include Columbia, Jefferson City, and probably extend north to St.Joseph.

MI Wayne is underpopulated.  Adding Oakland and Macomb, and shifting the Lansing 3 counties should balance that.

FL Everglades is overpopulated.  My proposal splits off the west coast, but my Tampa Bay is overpopulated.  Possibly shifting Citrus and Hernando north, since I've dropped down to Ft.Myers.  Your Cape Canaveral shows as being overpopulated, but that is because you included Lake, Hernando, and Citrus in its population (or your map is wrong).

GA Blue Ridge is underpopulated and Kennesaw overpopulated, but we are already discussing that.

NC Blue Ridge is underpopulated (isn't the Blue Ridge most associated with VA?).  Adding Charlotte or Greensboro would correct that, but would mess up the other districts.  Does Asheville-Charlotte, Greensboro-Raleigh, and Wilmington-Fayetteville and whatever we can stuff into it work?

NJ Pinelands is too small, but I like Train's map better.

NY Ontario is overpopulated, but counties like Jefferson, Delaware, Otsego, etc. would work OK.

NY Brooklyn is overpopulated but may be acceptable as an exception that avoids splitting counties.  Queens alone is barely below the minimum, but that would make 3 LI districts, and force the two non-LI districts close to 6 million each.

PA Philadelphia is underpopulated.   Adding in the 3 adjacent counties is right near the national mean.   This would leave the remnant of Delaware underpopulated, so it would have to include Wilkes-Barre and Scranton and the rest of NE PA.  This would make Susquehanna as Harrisburg-Lancaster-York-Reading district in the south central part of the state.  State College and Altoona may need to be shifted as well.

MA Berkshires is underpopulated.  I would add the southern part of the state (Bristol-Plymouth-Barnstable, Duke and Nantucket.  Splitting Middlesex doesn't really solve the problem.  If you are using towns, then you could argue that the Boros belong with Boston area.

[muon2]

Just so I understand, I'm prepared to take the following as the revised criteria:

I want to test on a body that has a size comparable to a legislature and which the districts are are of coarsely comparable sizes.  For my model, I have chosen a 100-member House of Representatives, the apportionment of which is shown in the map above.

What I need are districting plans for the 22 states that have more than one representative.

Guidelines:

(1) Don't split counties, with the possible exceptions of Los Angeles, CA and Cook, IL.  New York City may simply be treated as 5 counties, though of course they likely form communities of interest. Split counties should only be combined with other counties to avoid violations of population range.

(2) Strong community of interest.

(3) Coarse equality.  Precise equality is undesirable.  Even equality within 10% of the average for the state is not so good, unless it just happens to match a community of interest.  As a general guideline, try to kKeep districts in the range of 2/3 to 1-1/3 of the quota for the state average quota for states with more than one district (3,394,813).  You may go outside with justification. Examples of justification include keeping counties intact, avoiding out-of range districts elsewhere in the state, and significant unavoidable violations of communities of interest.

(4) Connectivity is not a requirement, at this scale.  Contiguity might be waived in instances where there is a central district that spans across a state. For example, Cape Cod could be linked to western MA to keep the Boston metro intact within population range.

(5) Assume there is some mechanism in place to act as a check of excessively partisan plans.

(6) Plans may be subject to state plebiscites, so be prepared to advocate to the state voters that your plan should be adopted.

(7) Provide 2010 Census populations for the districts.  These will be adjusted based on the apportionment populations which include certain overseas Americans.

(7a) Provide electoral data for the 2008 presidential election for each district.

(7b) Provide each district with should have a name, with the state name as the prefix part of the name. For example, CA San Joaquin Valley.

[Хahar 🤔]

If Ventura was given its choice of district:

(1) Silicon Valley-Central Coast;
(2) San Joaquin Valley;
(3) Orange County;
(4,5) One of the suburban LA districts;

which would they choose?

Suburban Los Angeles, absolutely. There's not even any question in that regard. Ventura County is suburban Los Angeles. Orange County would be awkward because of non-contiguity but not terrible. The San Joaquin Valley would make very little sense but it would still be better than a connection with San Jose.

[muon2]

If Ventura was given its choice of district:

(1) Silicon Valley-Central Coast;
(2) San Joaquin Valley;
(3) Orange County;
(4,5) One of the suburban LA districts;

which would they choose?

Suburban Los Angeles, absolutely. There's not even any question in that regard. Ventura County is suburban Los Angeles. Orange County would be awkward because of non-contiguity but not terrible. The San Joaquin Valley would make very little sense but it would still be better than a connection with San Jose.

Which of these two options would make more sense to you for the LA area:

Option A (keep LA county whole)
1) SLO, SB, Ventura, Orange (about 1K over population but justifiable)
2) Antelope Valley, San Gabriel Valley, Agoura Hills, and communities included inside LA city
3) City of LA
4) Torrance pocket, Los Angeles Valley (East LA to Long Beach)

Option B (keep Orange separate)
1) SLO, SB, Ventura, Antelope Valley, Agoura Hills, Torrance pocket, and communities in LAC
2) San Gabriel Valley, Los Angeles Valley
3) City of LA
4) Orange

[jimrtex]

Just so I understand, I'm prepared to take the following as the revised criteria:

I want to test on a body that has a size comparable to a legislature and which the districts are are of coarsely comparable sizes.  For my model, I have chosen a 100-member House of Representatives, the apportionment of which is shown in the map above.

What I need are districting plans for the 22 states that have more than one representative.

Guidelines:

(1) Don't split counties, with the possible exceptions of Los Angeles, CA and Cook, IL.  New York City may simply be treated as 5 counties, though of course they likely form communities of interest. Split counties should only be combined with other counties to avoid violations of population range.

(2) Strong community of interest.

(3) Coarse equality.  Precise equality is undesirable.  Even equality within 10% of the average for the state is not so good, unless it just happens to match a community of interest.  As a general guideline, try to kKeep districts in the range of 2/3 to 1-1/3 of the quota for the state average quota for states with more than one district (3,394,813).  You may go outside with justification. Examples of justification include keeping counties intact, avoiding out-of range districts elsewhere in the state, and significant unavoidable violations of communities of interest.

(4) Connectivity is not a requirement, at this scale.  Contiguity might be waived in instances where there is a central district that spans across a state. For example, Cape Cod could be linked to western MA to keep the Boston metro intact within population range.

(5) Assume there is some mechanism in place to act as a check of excessively partisan plans.

(6) Plans may be subject to state plebiscites, so be prepared to advocate to the state voters that your plan should be adopted.

(7) Provide 2010 Census populations for the districts.  These will be adjusted based on the apportionment populations which include certain overseas Americans.

(7a) Provide electoral data for the 2008 presidential election for each district.

(7b) Provide each district with should have a name, with the state name as the prefix part of the name. For example, CA San Joaquin Valley.

My comments:

(1) I'm not totally averse to adding Ventura County to part of Los Angeles.  I didn't really like the San Emigdio district.   That part of the city of Los Angeles is immediately adjacent to Simi Valley and Thousand Oaks is somewhat of a problem.  I suspect that whenever Los Angeles County became a single district, that Ventura would have been placed with either Orange County, or portions of Central California.

Proposed new wording:

(3) Districts should be within 2/3 to 1-1/3 of the average quota for states with more than one district (3,994,813)  Range: 2,263,209 to 4,526,417.

Districts in two-district states may range from between 90% and 110% of the state average, if such limits exceed the national limits.

The limits may be exceeded in exceptional cases, with justification.  Examples of justification include keeping counties intact, avoiding out-of range districts elsewhere in the state, and significant unavoidable violations of communities of interest.

Guidance: Districts should be analogous to globs of clay, similar in size, rather than blocks of clay that are carefully carved with a scalpel and weighed on a scale.  Districts should represent communities of interest, when possible.  But population can not be totally ignored.

(7) I don't think I will adjust the population to the apportionment population (overseas population).

(7b) Provide each district with a name.  If the name does not include the name of the state, affix the postal abbreviation of its state.   Example: San Joaquin Valley (CA).

Note: I am prepared to accept names like North Jersey, but will likely include the (NJ) just to provide a consistency with a name such as East Carolina (NC).