North Carolina 2020 Redistricting

[jimrtex]

North Carolina has 16 regional council of governments. These are voluntary organizations of cities and counties that provide cooperation on issues of regional interest. They reflect meaningful and officially recognized communities of interest. They are the starting point for drawing a congressional district map independent of past gerrymanders.

This table show the populations of the regions relative to the quota for a congressional district.

RCOG	City	Population
Albermarle	Hertford	0.234
Cape Fear	Wilmington	0.573
Centralina	Charlotte	2.684
Eastern Carolina	New Bern	0.863
High Country	Boone	0.286
Isothermal	Rutherfordtown	0.315
Kerr-Tar	Henderson	0.309
Land-of-Sky	Asheville	0.544
Lumber River	Pembroke	0.408
Mid-Carolina	Fayetteville	0.678
Mid-East	Washington	0.390
Piedmont Triad	Greensboro	2.237
Southwestern	Sylva	0.265
Triangle J	Raleigh	2.292
Upper Coastal Plain	Wilson	0.423
Western Piedmont	Hickory	0.498

Three central regions of Centralina (Charlotte), Piedmont Triad (Greensboro-Winston Salem), and Triangle J (Raleigh-Durham) are all entitled to two plus districts each and are fairly extensive, extending beyond what we would consider their metropolitan areas. To the extent possible we will avoid drawing areas outside these major cities, and create separate districts for secondary cities such as Fayetteville, Wilmington, Asheville, Rocky Mount, Greenville, Jacksonville, and Hickory(Unifour)

We can groups the RCOG's into West, Central, and Eastern regions. The central region is the three RCOG's based on Charlotte, Winston-Salem, Greensboro, and Durham. The west region is the five RCOG's to the west, and the east region is 8 RCOG's to the west.

Area	Population
West	1.909
Central	7.212
East	3.879

Based on these regions, we should be able to draw 2 districts largely to the west of the Charlotte, and Greensboro-Winston-Salem metro areas, and 4 districts to the east.
This map groups RCOG's in areas of roughly equal population.



The western area is divided in a way that produces the best population balance, but extending along the Appalachian spine is not an unreasonable community of interest.

Mecklenburg, Wake, and Guilford+Forsyth were split out in separate areas, each with a bit more than the quota. Mecklenburg and Wake will necessarily be divided, and Guilford would be unless we split Forsyth adn Guilford.

Southern suburban counties around Charlotte were split out, which along with Mecklenburg has a population equivalent to two quotas. Stanly was included to hit the quota target. The remainder of the Centralina RCOG was used to form the 13th district.

I could not find a reasonable way to combine Kerr-Tar (north of Raleigh) and Mid-Carolina (Fayetteville) with the other eastern RCOG's so I instead attached them to the Triangle J RCOG, and split that area three ways: Wake (Raleigh); Durham-Orange(Durham-Chapel Hill) and the areas to the north, and areas to the south. Moore and Lee are somewhat remote from Raleigh, while Harnett has spillover from Wake.

The relative population of the areas.

Area	Population
Appalachian	1.095
Western	0.814
Charlotte	1.254
Charlotte Metro	0.745
I-85	0.686
Greensboro-Winston-Salem	1.144
Central	1.093
Raleigh	1.228
Durham-North	0.856
Raleigh Metro	1.194
Northeast	1.048
Eastern	0.863
Southeast	0.981

This iteration refines the districts to get the districts more in population balance, particularly in the east and the west.



Surry is added to the western area to get the population up to two quotas. The county has the right size, and is remote from Greensboro. Some counties were swapped to better balance the populations of the Appalachian and Western districts.

In the east, Cumberland(Fayetteville) was moved into the southeastern district, in exchange for New Hanover(Wilmington) being moved into the eastern district. Wilmington fits better with a coastal district, and Fayetteville is better in a district with Robeson than a district that pushes into the Raleigh suburbs.

Anson, Richmond, Duplin, Wayne, Greene, Beaufort, and Hyde were moved to balance population.

The I-85 district was expanded to the north to get sufficient population. This produces a compact district, with a mix of suburban Charlotte, suburban Greensboro, and exurban areas, as well as smaller mill towns along the Fall Line.

The central area including the big question mark around Greensboro needs more work.

Updated population:

Area	Population
Appalachian	1.009
Western	1.000
Charlotte	1.254
Charlotte Metro	0.745
I-85	0.979
Greensboro-Winston Salem	1.144
Central	0.762
Raleigh	1.228
Durham-North	0.856
Raleigh Metro	1.006
Northeast	1.004
Eastern	1.007
Southeast	1.005

This is a refinement that produces whole-county districts, except for splits of Mecklenburg, Wake, and Guilford counties. The splits of Mecklenburg and Wake are necessary. The split of of Guilford permits Winston-Salem and Greensboro to be maintained in a single district.



It is expected that the cut of Mecklenburg will permit the western and eastern suburbs of Charlotte to be linked, and the cut of Wake will provide a link between Johnston and Franklin counties. This introduces district numbers which are added in a snake-like fashion from east to west.

Whole-county districts in bold.

District	Population
1-Appalachian	1.009
2-Western	1.000
3-Charlotte	1.254
4-Charlotte Metro	0.745
5-I-85	0.979
6-Greensboro-Winston Salem	1.336
7-Central	0.617
9-Raleigh	1.228
8-Durham-North	1.012
10-Raleigh Metro	0.803
13-Northeast	1.004
12-Eastern	1.007
11-Southeast	1.005

Mecklenburg Split:

4-Charlotte	0.999
3-Charlotte Metro	0.255

Guilford Split:

5-I-85	0.013
6-Greensboro-Winston Salem	0.322
7-Central	0.330

Wake Split:

7-Central	0.045
9-Raleigh	0.993
10-Raleigh Metro	0.190

This is the division of Mecklenburg, and the surrounding area.



CD-4 is wholly in Mecklenburg and includes the center and north of the county. About 80% of Mecklenburg is in the district. The boundary line generally follows I-485, coming inside to include Mint Hill, Mathews, Pineville, and unincorporated areas, while cutting off the areas of Charlotte that extend out past the loop.

Roughly 1/4 of the district is in each of Gaston, Union, and Mecklenburg counties.

The initial attempt at dividing Guilford did not work well. While including Rockingham and Stokes with Forsyth and Guilford produced a compact whole-county district, it required excising a larger portion of Guilford to reach population equality. An island with Greensboro would have to be connected by an isthmus to either Rockingham or Forsyth, which seems to be contrary of the attempt to have a district centered on Greensboro and Winston-Salem.

This alternative arrangement shifts Rockingham and Stokes to the Durham-Chapel Hill district, and Alamance to the Central district. This results in a much larger share of Guilford remaining in the Greensboro-Winston-Salem district.



As a secondary benefit this improves overall equality of the map.

This shows the division of Guilford.



Greensboro and (most of) High Point are included in NC-6 with Forsyth and Winston-Salem. The district is split 48-52 between the two counties, with the slight majority in Guilford.

The southern and eastern part of the county outside Greensboro and High Point is in NC-7. This is about 20% of the county. The hook on the west end is unincorporated territory between Greensboro and High Point. The hook on the north is for population reasons. This area fits well with Randolph and Alamance counties. Burlington is on the western edge of Alamance and may be growing towards Greensboro.

A small nibble of High Point is included in NC-4 for overall population equality. High Point does lap into Davidson County at that point. But Guilford was chosen in otder to limit county splits to three counties. An alternative would be to take a nibble out northern Mecklenburg, at a slight decline in overall equality (standard deviation 0.48% to 0.52%). In a rational world, a single 2.1% deviation would be acceptable.

[jimrtex]

That really just leaves Ohio and Texas as the last two surviving "major" R Gerrymanders left in the country.   (Not saying the NC Map is perfect now or anything)

We'll see how well the Texas one holds up next year though.

Important to note that Texas uses the same statewide partisan judicial election system as PA and NC for its state supreme court (except with normal primaries, so it is always 1 D and 1 R running per court seat in the GE).  This is all enshrined in the state constitution, which would require a 2/3rds majority vote of both state legislative chambers (which Republicans have never held in both chambers simultaneously) and then a majority statewide referendum vote to change.  Currently, all of the state supreme court seats are held by Republicans, but if Democrats reach the point where they are often able to win statewide elections, PA/NC style map reform is likely coming to Texas during the 2020's. 


Normally 3 of the 9 seats are up every 2 years.

There is no requirement in the Texas constitution for partisan judicial elections.

[jimrtex]

If you honestly think that is why the Democrats in then states do that. Then you are fooling yourself.

I’m not saying it’s the only reason. But it was definitely republicans that started the modern gerrymandering war with the old Texas, Louisiana, Florida maps of the 2000s & 2010s. Democrats also are playing in this, but we’re the ones trying to end it.

I think you are forgetting what Democrats did in NC for years. Republicans in NC at least are making maps that are too crazy

...back when the Democratic party in states like NC and TX was still led by white conservatives. You may not be aware that a decade ago, Republicans saw gerrymandering as a necessary tool to achieve power, while the people advocating for commissions and independent redistricting were Democrats. Good government and independent redistricting are liberal priorities. After Republicans used gerrymandering to lock up state legislatures and Congress, many liberals said Inks this, let's go for the jugular, too.

Democrats are not liberals. They are leftists.

Have you never heard of Vera vs. Richards?

[jimrtex]

North Carolina Supreme Court has agreed to delay the primaries.

https://twitter.com/SFalmlen/status/1468697706994544640?s=20

GOOD

If the NCGOP wants war, they'll get war. They might still win it, but they'll have to fight it out and pay a price.

All out war ends in impeachment, the NC court should be careful about pushing it too far especially with regards to legislative districts. You tell me what you would strike down on the above senate maps.

https://sites.duke.edu/quantifyinggerrymandering/files/2021/08/countyClusters2020.pdf

Use these if you are confused about any weird districts. For example 19/21 look weird but are 100% justifiable.  Democrats wanted to split Fayetteville in 2 to get 2 Dem seats. The GOP merely said no to that. In exchange they were however consistent on the senate map and did not split Wilmington. If the NC court is going to say you can't split Wilmington but must split Fayetteville they really are pushing their luck.

It is possible that the clustering algorithm, which was devised by the NC Supreme Court, is not working. It requires smallest clusters (i.e. single counties, two counties, three counties) be drawn first. The remnant area or areas must also be able to divided into districts within 5% limits. The initial clusters may push these remnant areas closer to the 5% threshold which makes it harder to split on counties. I bet that 7-county, 4-district cluster that snakes from north of Charlotte, to east of Greensboro has a significant deviation. It is surrounded by two-county and three-county clusters, and instead gets chopped.

Better would be to simply ignore the five percent limts and divide the state into (1) multi-county, single-member districts, or (2) single-county multi-member districts, or (3) single-county single-member districts. Alternative plans can be compared based on standard deviation. Then all that has to be done is to divide the single-county multi-member districts.

This complies with the NC Constitution, the VRA, and equal protection.

Alternatively do the same and adopt weighted voting.

[jimrtex]

This is pointless. The SCONC will flip anyway and republicans will draw another gerrymander mid decade

There’s no downside for us; why not fight it in court and see what we can get?

This.  It's also unconstitutional in NC to redraw the state legislative districts mid-decade, so if either of those maps get thrown out, the changes will stick.  Dems could plausibly flip a chamber over the course of the decade.

Did they explicitly forbid it? If not, this will be up to the likely GOP SC in 2022.

It says the process is done at the first legislative session after the census is released. 

Have you checked TX, GA, NV constitutions? I believe such language should be there as well. I am curious which state does not have this type of language.

TX and NV have this language for the legislative districts.  Georgia’s language is a bit more open.

TX did this back in 2000s. NV threatened to do this in 2010s. Hence I don't think such languages are major obstacles, especially with a friendly court.

The legislature failed to do legislative redistricting in 2001, so it was done by the Legislative Redistricting Board under terms of the Texas Constitution.

The legislature failed to redistrict Congress in 2001 because of a deadlock between the Senate and House. The federal court intervened and imposed a map largely based on the 1990's Democrat gerrymander.

In 2003, the legislature was able to draw a congressional map and do so, radically undoing the past Democrat gerrymanders.

[jimrtex]

The word "fair" should probably be banned from the English language.

Should we switch "fair" to "middling"?