Virginia 2020 County Projection Map

[muon2]

Nice work. I use a slightly different method to project the 2020 populations. I take the 2016 estimate (Fv) and compare it to the 2010 census (Pv) using the rate function to get the annual percentage rate based on 6.25 years (N). The extra quarter year is because the census is on Apr 1 and the estimate is from Jul 1. The rate function is mathematically [(Fv/Pv)^(1/N) - 1].

To get the 2020 projection I use a future value based on the 2010 census (Pv), annual percentage rate from above (R) and 10 years (P). The future value is P[(1+R)^N - 1]/R. This method assumes that the average rate of growth from 2010 to 2016 continues until 2020 and it is based on a compounding, not linear formula.

For VA Bedford county needs special treatment. In the 2010 census Bedford city was independent but was dissolved in 2013 and returned to the county. For the 2010 census I use the sum of Bedford county and Bedford city.

If this method is applied to the state as a whole the 2020 population would be 8668.3 K. If it is applied separately to each county and independent city then added up, the population would be 8681.9 K. The difference is due to the fact that the counties are growing at different rates, and this introduces variances compared to the state overall. For redistricting I use the total from the counties so with 11 CDs the population of each would be 789.3 K.

Here is the my list of all counties/independent cities, growth rates and their 2020 projected populations (in thousands):
Accomack County   -0.1%   32.8
Albemarle County   1.2%   111.9
Alleghany County   -0.7%   15.2
Amelia County   0.3%   13.0
Amherst County   -0.4%   31.2
Appomattox County   0.5%   15.8
Arlington County   1.7%   244.7
Augusta County   0.3%   75.8
Bath County   -0.9%   4.3
Bedford County   0.6%   79.9
Bland County   -0.7%   6.3
Botetourt County   0.0%   33.3
Brunswick County   -1.1%   15.6
Buchanan County   -1.3%   21.1
Buckingham County   -0.1%   17.0
Campbell County   0.0%   55.0
Caroline County   0.9%   31.2
Carroll County   -0.3%   29.2
Charles City County   -0.4%   7.0
Charlotte County   -0.6%   11.9
Chesterfield County   1.1%   353.5
Clarke County   0.4%   14.6
Craig County   -0.1%   5.1
Culpeper County   1.1%   52.2
Cumberland County   -0.6%   9.4
Dickenson County   -1.0%   14.4
Dinwiddie County   0.1%   28.2
Essex County   -0.0%   11.1
Fairfax County   0.8%   1174.2
Fauquier County   0.9%   71.5
Floyd County   0.5%   16.0
Fluvanna County   0.4%   26.6
Franklin County   -0.0%   56.0
Frederick County   1.2%   88.3
Giles County   -0.4%   16.6
Gloucester County   0.2%   37.4
Goochland County   0.7%   23.3
Grayson County   -0.4%   14.9
Greene County   0.8%   20.0
Greensville County   -0.7%   11.4
Halifax County   -0.6%   34.3
Hanover County   0.7%   107.2
Henrico County   1.0%   338.8
Henry County   -0.8%   49.9
Highland County   -0.7%   2.2
Isle of Wight County   0.6%   37.4
James City County   1.7%   79.2
King and Queen County   0.5%   7.3
King George County   1.6%   27.5
King William County   0.4%   16.6
Lancaster County   -0.6%   10.7
Lee County   -0.9%   23.4
Loudoun County   3.4%   438.2
Louisa County   1.0%   36.5
Lunenburg County   -0.8%   11.9
Madison County   -0.3%   12.9
Mathews County   -0.4%   8.7
Mecklenburg County   -0.9%   29.8
Middlesex County   -0.3%   10.7
Montgomery County   0.7%   101.2
Nelson County   -0.2%   14.8
New Kent County   2.2%   23.0
Northampton County   -0.3%   12.0
Northumberland County   -0.1%   12.2
Nottoway County   -0.3%   15.4
Orange County   1.0%   36.8
Page County   -0.3%   23.4
Patrick County   -0.5%   17.6
Pittsylvania County   -0.5%   60.6
Powhatan County   0.2%   28.7
Prince Edward County   -0.2%   23.0
Prince George County   0.9%   39.2
Prince William County   2.0%   490.5
Pulaski County   -0.3%   33.8
Rappahannock County   0.0%   7.4
Richmond County   -0.8%   8.5
Roanoke County   0.3%   95.0
Rockbridge County   0.1%   22.4
Rockingham County   0.7%   81.9
Russell County   -0.9%   26.5
Scott County   -0.9%   21.2
Shenandoah County   0.4%   43.9
Smyth County   -0.6%   30.4
Southampton County   -0.4%   17.8
Spotsylvania County   1.2%   138.1
Stafford County   1.8%   154.5
Surry County   -1.2%   6.3
Sussex County   -0.8%   11.2
Tazewell County   -1.1%   40.5
Warren County   0.7%   40.1
Washington County   -0.2%   53.8
Westmoreland County   0.1%   17.7
Wise County   -0.9%   38.0
Wythe County   -0.1%   28.9
York County   0.6%   69.5

Alexandria city 1.7%   166.2
Bristol city   -0.8%   16.5
Buena Vista city   -0.5%   6.3
Charlottesville city   1.2%   49.1
Chesapeake city   1.1%   247.9
Colonial Heights city   0.3%   18.0
Covington city   -1.2%   5.3
Danville city   -0.4%   41.2
Emporia city   -1.8%   5.0
Fairfax city   1.1%   25.2
Falls Church city   2.1%   15.1
Franklin city   -0.5%   8.1
Fredericksburg city   2.5%   31.0
Galax city   -0.6%   6.6
Hampton city   -0.2%   134.2
Harrisonburg city   1.3%   55.7
Hopewell city   0.1%   22.8
Lexington city   0.0%   7.0
Lynchburg city   1.0%   83.1
Manassas city   1.5%   43.8
Manassas Park city   1.8%   17.0
Martinsville city   -0.4%   13.2
Newport News city   0.1%   182.5
Norfolk city   0.2%   246.5
Norton city   -0.4%   3.8
Petersburg city   -0.3%   31.6
Poquoson city   -0.2%   11.9
Portsmouth city   -0.0%   95.1
Radford city   1.0%   18.2
Richmond city   1.4%   235.4
Roanoke city   0.4%   101.3
Salem city   0.5%   26.0
Staunton city   0.4%   24.7
Suffolk city   0.9%   92.2
Virginia Beach city   0.5%   461.6
Waynesboro city   0.7%   22.4
Williamsburg city   1.3%   15.9
Winchester city   0.8%   28.3

[muon2]

The courts will probably not look favorably on a Dem gerrymander that cracks the black population of SE VA. As the recent congressional case and subsequent election show, a district does not have to be 50% to elect the black candidate of choice, but it can't be too low that it would likely elect a white Dem. A rough guide is probably to check that the black population is a substantial majority of the Dem vote in the primary.

To help, here is a map showing the 2010 BVAP colored as follows:
yellow 25.0-33.3%
lime 33.4-39.9%
green 40.0%-49.9%
dark green > 50.0%



This is the full table of counties with the 2010 BVAP% and the 2015 5-year estimate of black%.

.Accomack County   26.9%   28.7%
.Albemarle County   9.4%   9.4%
.Alleghany County   4.5%   5.3%
.Amelia County   23.4%   24.3%
.Amherst County   18.9%   18.4%
.Appomattox County   19.1%   20.3%
.Arlington County   8.0%   8.5%
.Augusta County   4.3%   4.0%
.Bath County   4.7%   3.4%
.Bedford County   7.0%   7.1%
.Bland County   3.8%   5.0%
.Botetourt County   3.0%   3.5%
.Brunswick County   55.7%   55.8%
.Buchanan County   2.7%   2.7%
.Buckingham County   35.5%   34.5%
.Campbell County   14.1%   14.0%
.Caroline County   29.6%   29.3%
.Carroll County   0.6%   0.9%
.Charles City County   48.5%   48.2%
.Charlotte County   28.8%   31.5%
.Chesterfield County   20.9%   22.7%
.Clarke County   5.6%   5.5%
.Craig County   0.0%   0.2%
.Culpeper County   15.9%   14.4%
.Cumberland County   31.9%   33.1%
.Dickenson County   0.3%   0.9%
.Dinwiddie County   32.5%   33.0%
.Essex County   36.3%   41.0%
.Fairfax County   8.6%   9.4%
.Fauquier County   8.2%   7.4%
.Floyd County   1.9%   2.5%
.Fluvanna County   15.7%   14.4%
.Franklin County   8.2%   8.5%
.Frederick County   3.7%   4.4%
.Giles County   1.4%   1.6%
.Gloucester County   8.8%   8.3%
.Goochland County   20.2%   17.8%
.Grayson County   1.9%   2.7%
.Greene County   6.1%   6.6%
.Greensville County   58.9%   58.0%
.Halifax County   35.6%   36.9%
.Hanover County   9.3%   9.3%
.Henrico County   27.9%   29.6%
.Henry County   21.7%   22.3%
.Highland County   0.2%   0.0%
.Isle of Wight County   24.1%   22.8%
.James City County   12.5%   13.5%
.King and Queen County   28.6%   26.5%
.King George County   17.6%   18.0%
.King William County   18.4%   19.5%
.Lancaster County   25.0%   28.0%
.Lee County   4.3%   3.8%
.Loudoun County   7.3%   7.4%
.Louisa County   17.9%   16.7%
.Lunenburg County   35.4%   34.0%
.Madison County   10.1%   8.3%
.Mathews County   9.1%   10.3%
.Mecklenburg County   35.5%   35.1%
.Middlesex County   17.8%   19.3%
.Montgomery County   3.7%   4.3%
.Nelson County   13.0%   13.4%
.New Kent County   13.9%   12.3%
.Northampton County   35.5%   36.4%
.Northumberland County   22.9%   28.7%
.Nottoway County   39.5%   39.6%
.Orange County   12.4%   12.7%
.Page County   1.9%   1.6%
.Patrick County   5.7%   6.4%
.Pittsylvania County   22.1%   21.4%
.Powhatan County   14.8%   12.3%
.Prince Edward County   30.6%   32.9%
.Prince George County   31.4%   32.1%
.Prince William County   19.3%   20.5%
.Pulaski County   5.0%   5.8%
.Rappahannock County   4.5%   4.2%
.Richmond County   31.2%   28.8%
.Roanoke County   4.8%   5.8%
.Rockbridge County   2.6%   2.9%
.Rockingham County   1.6%   1.8%
.Russell County   0.8%   1.0%
.Scott County   0.6%   0.8%
.Shenandoah County   1.6%   2.4%
.Smyth County   2.0%   2.1%
.Southampton County   37.8%   35.9%
.Spotsylvania County   14.6%   16.0%
.Stafford County   16.1%   17.1%
.Surry County   45.4%   45.1%
.Sussex County   58.2%   57.1%
.Tazewell County   3.2%   2.9%
.Warren County   4.6%   4.0%
.Washington County   1.3%   1.5%
.Westmoreland County   26.5%   28.2%
.Wise County   6.1%   5.4%
.Wythe County   2.7%   3.6%
.York County   13.2%   13.5%
.Alexandria city   20.4%   21.4%
.Bedford city   19.2%
.Bristol city   5.2%   5.7%
.Buena Vista city   4.8%   1.1%
.Charlottesville city   16.9%   19.2%
.Chesapeake city   28.8%   29.7%
.Colonial Heights city   8.7%   13.1%
.Covington city   12.7%   13.9%
.Danville city   44.7%   48.7%
.Emporia city   59.5%   55.1%
.Fairfax city   4.4%   4.8%
.Falls Church city   4.2%   2.1%
.Franklin city   54.1%   58.6%
.Fredericksburg city   19.8%   23.3%
.Galax city   5.3%   5.6%
.Hampton city   47.4%   50.1%
.Harrisonburg city   5.4%   7.1%
.Hopewell city   33.2%   38.1%
.Lexington city   9.3%   10.9%
.Lynchburg city   26.4%   28.4%
.Manassas city   12.9%   14.3%
.Manassas Park city   12.8%   12.4%
.Martinsville city   42.6%   45.9%
.Newport News city   37.5%   40.4%
.Norfolk city   39.7%   42.1%
.Norton city   5.9%   3.4%
.Petersburg city   77.0%   77.1%
.Poquoson city   0.7%   1.0%
.Portsmouth city   50.3%   52.6%
.Radford city   7.4%   9.2%
.Richmond city   46.4%   48.6%
.Roanoke city   26.0%   27.8%
.Salem city   6.2%   7.8%
.Staunton city   11.3%   12.2%
.Suffolk city   40.9%   41.8%
.Virginia Beach city   18.1%   19.3%
.Waynesboro city   9.9%   12.7%
.Williamsburg city   12.4%   15.3%
.Winchester city   10.2%   10.6%

[muon2]

Would you project the BVAP's forward to 2020 in the same way you project population figures forward?

It would be hard since the BVAP is not the same as the black estimate, and the estimate is based on a 5 year span. What I would say is that there isn't a great deal of change in percent in most of the SE counties, so using the 2010 DRA numbers should be a decent estimate of the BVAP.

[muon2]

Would you project the BVAP's forward to 2020 in the same way you project population figures forward?

It would be hard since the BVAP is not the same as the black estimate, and the estimate is based on a 5 year span. What I would say is that there isn't a great deal of change in percent in most of the SE counties, so using the 2010 DRA numbers should be a decent estimate of the BVAP.

It might be relevant as to whether a performing black district can be drawn that is Richmond based.

In 2010 the total black population was about 1.5% higher than the BVAP in SE VA. Over half of those excess black children will be voting age in 2020, so using the total black percentage from 2010 should be close to the BVAP in 2020.

I find that I can make two whole county CDs in SE VA that both have about 40% black pop. One is Richmond, Petersburg and the southern rural areas. The other is in Hampton Roads. Those should be pretty close to performing. With a couple of chops they would clearly be performing.

[muon2]

Here's a baseline to think about. I took my county projections for 2020 and divided them into whole counties, such that only Fairfax is chopped and all the CDs are within 0.5% of the quota. I preserved two 40% BVAP (38%+ in 2010) CDs in SE VA. Where I had choices I chose the less erose option, and one could certainly swap a chop in Louisa to lower the erosity between my CD 5 and 7. CD 2 was kept with whole counties by using the Tangier Island ferry. The ferry's only seasonal so it only counts as a local connection and costs erosity, but avoids a chop.



CD 1: D+5
CD 2: D+3
CD 3: D+8
CD 4: D+12
CD 5: R+5
CD 6: R+15
CD 7: R+12
CD 8: D+21
CD 9: R+17
CD 10: R+5
CD 11: D+12

The Dems have a 6-5 advantage, which is what one would hope for in a neutral map given the slight D+1 lean for the state overall.

[muon2]

The courts will probably not look favorably on a Dem gerrymander that cracks the black population of SE VA. As the recent congressional case and subsequent election show, a district does not have to be 50% to elect the black candidate of choice, but it can't be too low that it would likely elect a white Dem. A rough guide is probably to check that the black population is a substantial majority of the Dem vote in the primary.

Gingles test requires 50% BVAP in a compact area. The reason the court intervened was because VA-3 was not compact. But a district that stretches from Chesapeake to Richmond, while skipping through Chesterfield is not compact. A court would also look askance at the splitting of the current VA-3. And trying to create 3 Democratic districts in SE Virginia looks like a dummymander.

The court has found that though 50% in a compact area is necessary to trigger Gingles, the district that is drawn does not need 50%. The district only has to be able to elect the candidate of choice of the minority. Under that interpretation, there's a lot more wiggle room for mappers, though the burden is on a mapper to show that a sub 50% district can perform to elect a candidate of choice. In VA that depends on how well the minority can control the Dem primary and have enough crossover white Dem votes to prevail in the general election.

[muon2]

The courts will probably not look favorably on a Dem gerrymander that cracks the black population of SE VA. As the recent congressional case and subsequent election show, a district does not have to be 50% to elect the black candidate of choice, but it can't be too low that it would likely elect a white Dem. A rough guide is probably to check that the black population is a substantial majority of the Dem vote in the primary.

Gingles test requires 50% BVAP in a compact area. The reason the court intervened was because VA-3 was not compact. But a district that stretches from Chesapeake to Richmond, while skipping through Chesterfield is not compact. A court would also look askance at the splitting of the current VA-3. And trying to create 3 Democratic districts in SE Virginia looks like a dummymander.

The court has found that though 50% in a compact area is necessary to trigger Gingles, the district that is drawn does not need 50%. The district only has to be able to elect the candidate of choice of the minority. Under that interpretation, there's a lot more wiggle room for mappers, though the burden is on a mapper to show that a sub 50% district can perform to elect a candidate of choice. In VA that depends on how well the minority can control the Dem primary and have enough crossover white Dem votes to prevail in the general election.

You can't get to 50% in a compact area in SE Virginia, and if Dems can win in a 40% BVAP district, you don't have racially polarized voting.

We don't know how SCOTUS views this since there was no standing when Personhubullah came before them. However, the district court clearly expected the newly drawn VA-3 to maintain the ability to elect the candidate of the black minority. This is from their Jan 2016 order to proceed with the map of the special master.

You must be logged in to read this quote.

The phrase used by the district court clearly indicates that they believed that section 2 of the VRA could potentially be applied to VA-3. Later in the same opinion they recognize that they were not asked to take up a section 2 challenge to the map, but they did state that the adopted plan of the special master was selected and approved in order to withstand a potential section 2 challenge following the logic in McDaniel v Sanchez (1981).

[muon2]

Whats the 2012/16 PVIs int hose districts? I still think the BVAP is to low, but eh.

If you track down the opinion I cite above, you'll see that the special master found that both his drawn VA-3 and VA-4 could be expected to elect the candidate of choice of the black minority. The current VA-4 was just over 40% BVAP in 2010.

[muon2]

The courts will probably not look favorably on a Dem gerrymander that cracks the black population of SE VA. As the recent congressional case and subsequent election show, a district does not have to be 50% to elect the black candidate of choice, but it can't be too low that it would likely elect a white Dem. A rough guide is probably to check that the black population is a substantial majority of the Dem vote in the primary.

Gingles test requires 50% BVAP in a compact area. The reason the court intervened was because VA-3 was not compact. But a district that stretches from Chesapeake to Richmond, while skipping through Chesterfield is not compact. A court would also look askance at the splitting of the current VA-3. And trying to create 3 Democratic districts in SE Virginia looks like a dummymander.

The court has found that though 50% in a compact area is necessary to trigger Gingles, the district that is drawn does not need 50%. The district only has to be able to elect the candidate of choice of the minority. Under that interpretation, there's a lot more wiggle room for mappers, though the burden is on a mapper to show that a sub 50% district can perform to elect a candidate of choice. In VA that depends on how well the minority can control the Dem primary and have enough crossover white Dem votes to prevail in the general election.

You can't get to 50% in a compact area in SE Virginia, and if Dems can win in a 40% BVAP district, you don't have racially polarized voting.

Actually using 2010 data one can make a 51% BVAP district that is at least as compact as the current NC-1 and doesn't go into Richmond at all. Link the Hampton Roads to Petersburg through the rural counties south of the James.

Racially polarized voting can still occur even when there is a fraction of the white majority that votes for the candidate preferred by the minority population. As long as the white majority would usually defeat the black candidate of choice then Gingles is satisfied. Witness on both sides tend to run detailed statistical analyses (such as ecological inference) to show the existence or non-existence of bloc voting. My observation is that in areas with a significant black majority these analyses tend to be equivalent to testing whether the whites in the area under consideration are majority Pub or Dem. If they are majority Pub, then the conclusion is in favor of polarized voting.

[muon2]

Whats the 2012/16 PVIs int hose districts? I still think the BVAP is to low, but eh.

If you track down the opinion I cite above, you'll see that the special master found that both his drawn VA-3 and VA-4 could be expected to elect the candidate of choice of the black minority. The current VA-4 was just over 40% BVAP in 2010.

You must be logged in to read this quote.

My emphasis. The court did not consider whether the first condition could be met.

Correct. But my own analysis is that it is possible to meet the first condition. Also, I was not suggesting that VA-04 is required, only that VA-03 is required under the VRA.

[muon2]

I'm inspired to draw a 9-2 Virginia now. Don't think it's possible? I'm going to draw 6 NOVA bacon strips and keep VA-2, 3 and 4. 7-1-3 is way too generous to Republicans.

And I could draw a 27-0 map of NY with bacon strips from the city. Doesn't make it all that practical. Something like that belongs i. Kamilas thread.

Not just is it impractical, the public hates it, too. In IL there has been wide bipartisan support for neutral redistricting. It has only been officials with the party in power and party hacks who want to continue to bet that they can win the cycle before redistricting and exercise maximum party control. Even in IL it's not just Dems, but there are plenty of county-level Pubs who enjoy the majority in their jurisdictions. Neither side wants to give up that edge once they have it.

[muon2]

It looks like you are using 2010 population numbers. VA-2 with just VA Beach, Norfolk, and Delmarva is too low in population. That's why I posted the table of 2020 projections. My baseline crosses the Chesapeake up north to pick up the necessary extra counties, and that way avoids a chop of one of the other Hampton Roads communities.

[muon2]

If you can link me to this "erosity metric" that'd be much appreciated. In any case, such a map would probably still hold up until (READ: if) the courts strike it down, so it could still be of use for 2-3 cycles in all likelihood.

The complete set of metrics I use are in the stickied Muon Rules thread on this board. The full details for erosity begin at this post. It is not a currently recognized standard for compactness like Polsby-Popper or Reock, but there is academic research investigating it with perhaps some papers and a student's thesis to follow.

For Torie: I revised the rules in 2016 to permit intercounty connections by local roads, albeit with a penalty, and that included seasonal ferries. All season ferries are still regional connections without penalty.

[muon2]

I must say, Torie, I've outdone myself here. Oh, @muon, these districts are all roughly 790K in their projected 2020 populations, give or take a few thousand for each district.



CD-1: The blue PWC-based seat. 2012/2016 PVI of D+3.58
CD-2: successor to current one. 2016 PVI D+2.88
CD-3: successor to current one (and compact!). 40% black. D+7.75
CD-4: successor to current one (and compact!). 41% black. D+8.96
CD-5: Yellow seat. 2016 PVI of R+15.85
CD-6: Staunton-Danville seat. 2012/2016 PVI of R+14.86
CD-7: Compact-ish Albemarle-Chesterfield County seat with Henrico portion. R+0.12
CD-8: successor to current one. D+19.73
CD-9: successor to current one. R+16.69
CD-10: successor to current one. R+0.75 and trending blue.
CD-11: Fairfax County one. D+12.12

So 6 likely/safe D, 2 swing seats (though both are trending D) and 3 safe R seats. Doesn't appear to violate VRA. Only splits two counties (aside from the necessary Fairfax): Henrico and Augusta. It's worth pointing out that Clinton carried CD-07 and CD-10 pretty handily, so those R+ figures may be somewhat misleading. Romney did much better than Trump did in Albemarle/Henrico/Chesterfield/Loudoun/Fairfax. Indeed, CD-07 and CD-10 are both Obama '08-Romney '12-Clinton '16 seats.

CD-10 kind of concerns me, but if Jennifer Wexton or whichever Democrat is representing the seat come 2020 has any say, I'm sure they can shift the lines in Fairfax County to secure CD-10 to make it around D+5 or so. After all, they do have a bit of wiggle room to secure CD-01 and CD-10 with the many extra Democrats from CD-08 and CD-11 if they choose to do so. This map is just to show you that a D-gerrymander is, in fact, pretty easy, and looks much better than the current Republican gerrymander that appears to resemble an abortion that didn't take.

That a much nicer gerrymander and I hope you don't mind that I made your map full size. Just cut the _thumb part out of the name in the img and you can do it next time yourself.

I do have a quibble with the plan. Though it is legal under VA law it is unpleasant that you can't get from Chesterfield to the rest of CD 7 by road without going through CD 4. That type of adjacency would violate the muon rules as contiguity without connections are usually signs of partisan gerrymandering.

[muon2]

If you have the 2020 population numbers for this plan (deviations are fine, too), I can do a comparison with my baseline plan on the muon metrics. For your chopped counties, I need to know the expected population in each fragment. You can assume that they are split in proportion to the fragments' populations in 2010. I'll ignore Fairfax since that requires knowledge of the county subdivisions populations, which I haven't projected. If we were doing this right we would have been using those lines as if they were county lines.

[muon2]

Yes, I agree that PNM's map is about the best one can do, to achieve his nefarious goals.

There does not appear to be a way to avoid the yellow CD traveling the entire length of the state, given the assorted and sundry "blockages."

I might revise my map to avoid a macrochop in Chesterfield County and in VA-02 to play Muon2's game (I will ignore Muon2's rule to the extent it exists of banning bridge chops, because I don't agree with it, except as a preference). It raises a public policy issue actually. In the real world, it is better when creating a black performing CD to take in immediately adjacent black hoods in an adjacent county, rather than have the CD take in disparate and far away black rural counties. Indeed, it raises the Kennedy issue when one does so, although it would probably pass muster based on the rationale of avoiding a county chop, particularly a macro-chop, and the rural counties are adjacent, and where there is not a nefarious purpose of gerrymandering away a white seat of a given party. If in fact avoiding a chop of Chesterfield served the purpose of creating another white CD for the party of the map drawers, the map would be more vulnerable. Which is what PNM's map does (well a swing CD), but I digress.  My revised map will not be doing that, so it should be less vulnerable.

How many times do I have to say that bridge chops are no longer banned and haven't been for over a year? There may be an erosity penalty depending on the type of connection, but that applies to whole counties, too.

[muon2]

If you have the 2020 population numbers for this plan (deviations are fine, too), I can do a comparison with my baseline plan on the muon metrics. For your chopped counties, I need to know the expected population in each fragment. You can assume that they are split in proportion to the fragments' populations in 2010. I'll ignore Fairfax since that requires knowledge of the county subdivisions populations, which I haven't projected. If we were doing this right we would have been using those lines as if they were county lines.

Unfortunately I don't think the voting wards exactly match up with that map.

They don't always, but we looked at this and few years ago and concluded that VTDs that span the line could count as keeping communities whole. I just assume that they a split in such a way as to conform to the community boundaries.