NC redistricting revisited

[krazen1211]

It is truly a shame that it is rather difficult to draw an effective clean GOP map. I suppose something could be done from that to mesh 4 and 13 into 1 Dem superfortress while at the same time cracking Greensboro in a manner that is not as ugly as the official map.


Probably it has to do with the size of the various cities, and the spread nature of the black population. Eventually I suspect that will change and CD-1 will become a far more urban district.

[Miles]

It is truly a shame that it is rather difficult to draw an effective clean GOP map. I suppose something could be done from that to mesh 4 and 13 into 1 Dem superfortress while at the same time cracking Greensboro in a manner that is not as ugly as the official map.


Probably it has to do with the size of the various cities, and the spread nature of the black population. Eventually I suspect that will change and CD-1 will become a far more urban district.

NC seems to be naturally favorable to the Ds, like Florida and Illinois are for the Rs.

I'll work on a clean GOP map next...I may need your input

[muon2]

Here's my map that embeds the CD 1 I created above using neutral redistricting principles. I sought minimal county splits while maintaining some degree of compactness. The only splits are for CD 1 to comply with section 2 (50.04% BVAP), for the two large counties, and for two other counties to get all CDs within 1500 of the ideal size. Microchops of counties smaller than a precinct would be used to get exact population equality. Population deviation and 2008 results are in parentheses. I look forward to comments.



CD 1 (-799, 70.3% Obama) Section 2 compliant, it is about 1/3 from Raleigh-Durham.

CD 2 (+286, 54.0% Obama) Very compact, competitive CD centered on Fayetteville with only one split. Laurinburg in Scotland is in this CD allowing Ft Bragg to act as a natural border.

CD 3 (-962, 51.5% McCain) Competitive CD and more compact than most for this region. Whole counties only with half the population in the Greenville-Goldsboro area and the rest along the northern sounds.

CD 4 (+42, 51.6% McCain) Competitive CD with county splits to accommodate CD 1. Wake is divided so that it separates Raleigh in this CD from Cary in CD 13. Wake is the only county with a 3-way split.

CD 5 (+1029, 55.5% McCain) Compact CD centered on Winston-Salem. Only Wilkes is split with only about 7K shifted to CD 9.

CD 6 (-319, 54.1% Obama) Very compact, competitive CD centered on Greensboro with no county splits. The population of Triad and its suburbs is going to be in 2 CDs. This is the cleanest split I could construct.

CD 7 (-187, 56.2% McCain) Dedicated district for the southern coastal region with no county splits and almost no population deviation.

CD 8 (+948, 58.9% McCain) I 73 and US 74 provide the natural corridors linking this district. There are two partial counties to equalize population with CD 2 and 12.

CD 9 (+1225, 62.6% McCain) Reasonably compact CD with only one county split to equalize population with CD 5. Roughly equal population is in western Charlotte metro and in the Blue Ridge foothills.

CD 10 (-1440, 61.9% McCain) Very compact CD with whole counties only. The CD is entirely in the central Piedmont.

CD 11 (+224, 52.0% McCain) Potentially competitive CD with whole counties only. Anchored by Asheville it connects all the counties west of the Blue Ridge.

CD 12 (+580, 67.1% Obama) Entirely within Mecklenburg county it is 34.2% BVAP.

CD 13 (-631, 58.7% Obama) Reasonably compact CD with county splits due to CD 1 and the division of Wake. Cary is entirely in this CD.

[krazen1211]

It is truly a shame that it is rather difficult to draw an effective clean GOP map. I suppose something could be done from that to mesh 4 and 13 into 1 Dem superfortress while at the same time cracking Greensboro in a manner that is not as ugly as the official map.


Probably it has to do with the size of the various cities, and the spread nature of the black population. Eventually I suspect that will change and CD-1 will become a far more urban district.

NC seems to be naturally favorable to the Ds, like Florida and Illinois are for the Rs.

I'll work on a clean GOP map next...I may need your input

Well, that is mainly due to the fact that while a 50/50 district in Virginia is at the minimum lean GOP, that is not so in North Carolina.

This is where I get stuck. Out east you have 1 black and 1 white district.




The Triad districts work nicely enough. The problem is there is no place to go with the 8th and 7th. The 7th ends up being an extremely inefficient Dem vote sink and you are forced to yield a 4th district. Fundamentally there seems to be no way to crack Fayetteville, and if you turn it into a 60% vote sink, you run into problems with cracking Greensboro.

[muon2]

It is truly a shame that it is rather difficult to draw an effective clean GOP map. I suppose something could be done from that to mesh 4 and 13 into 1 Dem superfortress while at the same time cracking Greensboro in a manner that is not as ugly as the official map.


Probably it has to do with the size of the various cities, and the spread nature of the black population. Eventually I suspect that will change and CD-1 will become a far more urban district.

NC seems to be naturally favorable to the Ds, like Florida and Illinois are for the Rs.

I'll work on a clean GOP map next...I may need your input

Well, that is mainly due to the fact that while a 50/50 district in Virginia is at the minimum lean GOP, that is not so in North Carolina.

This is where I get stuck. Out east you have 1 black and 1 white district.




The Triad districts work nicely enough. The problem is there is no place to go with the 8th and 7th. The 7th ends up being an extremely inefficient Dem vote sink and you are forced to yield a 4th district. Fundamentally there seems to be no way to crack Fayetteville, and if you turn it into a 60% vote sink, you run into problems with cracking Greensboro.

I think you will also be stuck making a BVAP majority district in the east without either Raleigh or Durham.

[Miles]

krazen, my Triad districts will probably look like yours; there's adequate GOP territory around W-S/Greensboro to drown out the Democratic votes there.

I think I'll end up sinking McIntyre...there's a lot of Democratic strength permeating out from the Lumberton area.

[krazen1211]

I will use a county bridge snake courtesy of Muon2. This will properly vote sink the 7th district while not causing excessive problems with the 6th, with the cost of making the 8th district a very poorly drawn district (but a GOP hold).




Mathematically there are enough blacks left to make 1 district, although truthfully given that North Carolina Democrats have shown themselves to be happy with mid 40s vap perhaps that would be sufficient to avoid county butchery. Section 5 is perhaps not long for this world anyway.

[muon2]

I will use a county bridge snake courtesy of Muon2. This will properly vote sink the 7th district while not causing excessive problems with the 6th, with the cost of making the 8th district a very poorly drawn district (but a GOP hold).




Mathematically there are enough blacks left to make 1 district, although truthfully given that North Carolina Democrats have shown themselves to be happy with mid 40s vap perhaps that would be sufficient to avoid county butchery. Section 5 is perhaps not long for this world anyway.

Section 5 may not be long for the world, but section 2 still applies. The blacks don't complain about a 45% BVAP in a Dem map, but they sue if it is in a map to favor the GOP.

[krazen1211]

I will use a county bridge snake courtesy of Muon2. This will properly vote sink the 7th district while not causing excessive problems with the 6th, with the cost of making the 8th district a very poorly drawn district (but a GOP hold).




Mathematically there are enough blacks left to make 1 district, although truthfully given that North Carolina Democrats have shown themselves to be happy with mid 40s vap perhaps that would be sufficient to avoid county butchery. Section 5 is perhaps not long for this world anyway.

Section 5 may not be long for the world, but section 2 still applies. The blacks don't complain about a 45% BVAP in a Dem map, but they sue if it is in a map to favor the GOP.

That's true. And in the end easily corrected-just not with country integrity goals. Tentacles and snakes seem to require other tentacles and snakes and then you get tri-split counties all over the place. If one is going to siphon Durham with the 1st district, one might as well run the 4th to Greensboro; of course, High Point ends up being on the western edge of Guilford rather than east. Real pain in the neck.

Mecklenberg and Wake Counties are just all wrong. Lousy population figures and lousy layout. In Georgia one can plop a nice 80% circle or two in Dekalb County and a 60% white district in South Georgia can carry the day. Perhaps in a decade that can be done in Mecklenberg/East North Carolina.

[Vazdul (Formerly Chairman of the Communist Party of Ontario)]

Making a "clean" Republican gerrymander is an interesting challenge, and one that I may take up when I'm done with "Plan B."

[Miles]

I can't say I'm particularly happy with it:




Things to work on for next draft:

-That CD9 stretching from Gastonia to High Point bugs me the most.
-CD2 turned out to be even more of a GOP pack than it was on my nonpartisan map.
-CD6 could fall in a neutral year; in a D wave year 5 and 13 could also be possible pickups.

[traininthedistance]

Here's my clean North Carolina:



The maximum deviation is +1,618 for District 5; 5 and 6 are over 1K while 10 and 12 are under (normally I use 1K as my maximum deviation).  Splitting a county between 5 and 10 would fix that, but I wanted to show how close you could get with whole counties.

1: 66.3% Obama, 50.3% BVAP.  Durham is a lot more accessible to the rural black counties than the black areas of Raleigh, and besides I want to preserve the all-Wake district (and keep in the Section 5 counties, though it's possible I withdrew from one or two).  I think it's one of the best-looking iterations of this district out there, despite the eight county splits.  I did tinker with the boundaries to make the counties work out for 2, 3, 4, 7.

2: 38.9% Obama.  Still Johnston and Harnett-based, one split with 8 and two with 1.

3: 41.1% Obama.  Mostly just cleaner lines; three splits all with 1.

4: 57.4% Obama.  Two splits with 1, and all of Wake that's not in 13.  Bridging through northern Wake to Franklin saves at least one split elsewhere.

5: 37.1% Obama.  The district lines don't quite match the borders of Winston-Salem, but the intent is that areas within the city limits are in 6.  The only split is with 6, which is nice.

6: 60.1% Obama.  The Triad district.  Only splits Forsyth with 5.

7: 46.1% Obama. One split with district 1.  I tried to make something work like Muon's Lumberton-Fayetteville district and the southern coast instead, but 2 and 3 just ended up getting too squeezed.

8: 50.6% Obama.  One split with district 2.  

9: 39.8% Obama.  Splits Mecklenburg with 12 and Lincoln with 10.  Looks a bit ugly, but "Charlotte burbs" is one of those things which not only has lots of precedent, but is a coherent CoI.  So I'm keeping it.

10: 35.1% Obama.  Only splits Lincoln.  I don't think you can make this district any more compact.

11: 45.5% Obama.  You've seen this one before.

12: 66.0% Obama, 48.9% W/33.7%BVAP.  Any fair map has to have something like this district.

13: 58.8% Obama.  And Wake gets its own district too.

13 county splits, adding one between 5 and 10 would almost halve the max deviation from 1.6K to +844.

[jimrtex]

My best shot at a court-drawn/nonpartisan NC. I'm open to suggestions

The one thing I don't like about this map is that the Triad is split three ways. I tend to believe that fair maps should try and keep metro areas together as much as possible. In fact, I find metro area integrity to more important than county integrity, to be honest.  Urban and suburban Greensboro-Winston Salem is an area that ought to have its own congressional district, but has been cracked by both parties. 

Myrick's district is ugly, but it's a perfect CoI.

But it's a dangerous ugliness from the perspective of neutral redistricting principles. There was a bit of discussion of this on other threads, and the conclusion, which I now share, is that using a split county to bridge two whole counties in a district opens the door to gerrymandering mischief.

I would outlaw multi-spanning where two or more counties are split between a pair of districts.

[muon2]

My best shot at a court-drawn/nonpartisan NC. I'm open to suggestions

The one thing I don't like about this map is that the Triad is split three ways. I tend to believe that fair maps should try and keep metro areas together as much as possible. In fact, I find metro area integrity to more important than county integrity, to be honest.  Urban and suburban Greensboro-Winston Salem is an area that ought to have its own congressional district, but has been cracked by both parties. 

Myrick's district is ugly, but it's a perfect CoI.

But it's a dangerous ugliness from the perspective of neutral redistricting principles. There was a bit of discussion of this on other threads, and the conclusion, which I now share, is that using a split county to bridge two whole counties in a district opens the door to gerrymandering mischief.

I would outlaw multi-spanning where two or more counties are split between a pair of districts.

That's essentially part of the MI rules, though the VRA creates an exception for minority populations that span county lines. In NC I used multi-spanning for CD 1 as the only method to comply with section 2. I avoided it everywhere else.

[Miles]

traininthedistance, I really like your map. Seems to make sense to me...your 2nd reminds me of Vazdul's, but a bit more compact.

'Take two at a (reasonably) clean Republican map:




-From my last map, 5, 6 and 13 are now off the table for the Demcocrats.

-I punted on the Triad and drew a D vote sink there.

-For Fayetteville, I tried cracking it between 2 and 3. The 2nd would favor the GOP (they'd be best off here with Rouzer instead of Ellmers though); Jones would still hold down the 3rd easily, but it has less than 52% R average, so a Blue Dog could win there after he retires.

-The 9th uses Mecklenburg to connect Gaston to everything else. While this is ugly, it enables the 10th to use all whole counties (other than the Asheville crack) and is good from a CoI perspective.

-The 7th is similar to the version that McIntyre got under the original plan. Though he could lose in a wave year, I think he could still hold it otherwise.

[timothyinMD]

Your percents for districts 5 and 6 are reversed I believe

[Miles]

Your percents for districts 5 and 6 are reversed I believe

Thanks.

I'll fix that when I post my next version.

[Miles]

Changes from my last version:

-The McCain % in 2 and 3 is up slightly; Democrats will be hard-pressed to win 2, but they could still potentially pick up 3 when Jones retires. 3 becomes somewhat more compact.

-The 8th is a point or so more R...its now even worse for Kissell than the one in the actual Republican map.

- The 9th becomes more Charlotte-centric and is the only district to border the 12th.

[timothyinMD]

McCain districts
9 - 62.7
10 - 62.0
3 - 58.9
6 - 57.8
8 - 55.3
7 - 54.3
5 - 54.1
11 - 53.1
2 - 50.6

Obama districts
4 - 73.8 (34.5% VAP black, 50.2% VAP minorities)
1 - 68.1 (50.3% VAP black)
12 - 68.5 (35% VAP black, 53.9% VAP minorities)
13 - 52.4

[Vazdul (Formerly Chairman of the Communist Party of Ontario)]

McCain districts
9 - 62.7
10 - 62.0
3 - 58.9
6 - 57.8
8 - 55.3
7 - 54.3
5 - 54.1
11 - 53.1
2 - 50.6

Obama districts
4 - 73.8 (34.5% VAP black, 50.2% VAP minorities)
1 - 68.1 (50.3% VAP black)
12 - 68.5 (35% VAP black, 53.9% VAP minorities)
13 - 52.4

Not bad.

[krazen1211]

Changes from my last version:

-The McCain % in 2 and 3 is up slightly; Democrats will be hard-pressed to win 2, but they could still potentially pick up 3 when Jones retires. 3 becomes somewhat more compact.

-The 8th is a point or so more R...its now even worse for Kissell than the one in the actual Republican map.

- The 9th becomes more Charlotte-centric and is the only district to border the 12th.

Very effective milesmander.

What is interesting is that yielding a 4th district doesn't really improve the GOP's changes of winning the remaining 9 in any material fashion compared to the legislative map. It just puts extra Republicans in 8, 9, 5, 10.

[Miles]

Changes from my last version:

-The McCain % in 2 and 3 is up slightly; Democrats will be hard-pressed to win 2, but they could still potentially pick up 3 when Jones retires. 3 becomes somewhat more compact.

-The 8th is a point or so more R...its now even worse for Kissell than the one in the actual Republican map.

- The 9th becomes more Charlotte-centric and is the only district to border the 12th.

Very effective milesmander.

What is interesting is that yielding a 4th district doesn't really improve the GOP's changes of winning the remaining 9 in any material fashion compared to the legislative map. It just puts extra Republicans in 8, 9, 5, 10.

Yes, its kinda frustrating. Even with the Triad vote sink, its hard to contain the Democratic votes in the southeastern region. The Republicans would still have, at best, a 50/50 shot at this CD7 in normal circumstances. All the Triad sink did here was shore-up those districts that you mentioned.

I'll have one more variation of this map that will be a bit better for the GOP.

[Miles]

Ok, I'm going to say that this will be the final version of my "clean" Republican map in this series:




CD1:
51.2% Black VAP...similar to the ones in the previous maps.

CD2:
Still a hodgepodge of Raleigh/Fayetteville exurbs. Centered in Johnston/Harnett, cracks Fayetteville with the 3rd and on the east, takes whatever's left of the counties that CD1 reaches into.

CD3:
Putting half of Fayetteville helps to deprive CDs 2/7/8 of D votes. Safe for Jones but only has a 52% R average. Still could be a tossup when he retires; Perdue probably won handily here.

CD4:
The Chapel Hill/Durham/Wake county vote pack. I would have kept Orange county whole, but that would have meant splitting Durham-proper 3 ways (between 1/4/5)...I think keeping municipalities together is more important than keeping counties whole.

CD5:
Still recalls the 5th of the 1990's, though reaches down to take Iredell and parts of Rowan. Foxx could be vulnerable, as mant VA border counties would be new to her; Senate Majority Leader Phil Berge represents Rockingham and part of Guilford, so he could launch a credible challenge.

CD6:
The Triad vote sink. If I dismantle this, it would be swingy and CD5 and CD13 could buckle in a D wave year. Probably best for the GOP to concede this while keeping 5 and 13.

CD7:
A few minor line tweaks in Robeson county and this is up a bit to 56.0% McCain. McIntyre would still be slightly favored, but when he retires, the presence of Camp LeJuene here would give Ilario Pantano a boost.

CD8:
Stopping just barely short of 58% McCain, this would be a likely GOP pickup. The actual 57% NC-08 is no better than a tossup for Kissell, and this just puts him further in the hole. The 8th is a close second, after CD13, for the swing away from Obama.

CD9:
Similar to the current 9th, just substitutes most of southern Charlotte for Cabarrus county.

CD10:
Still moves northward. All whole counties except for the Asheville crack.

CD11:
If a 58.2% McCain district was enough to make Shuler fold, a 57.7% McCain seat should also prompt a retirement announcement. I always thought dragging the 11th along the SC border was the best for the Rs; that way, the Democrat running to replace Shuler will have to appeal to mountain voters and those in exurban Charlotte.

CD12:
Still the Charlotte sink. 34.3% Black VAP.

CD13:
Becomes more compact than my previous map. I didn't like that hook around Greensboro, so I replaced it with Alamance county. Its down to 53.5% McCain, from 55.3%, but it has a 56.2% R average. It may be competitive by 2016 or later, but if a R wins there in 2012, he'd have ample time to entrench himself even as the district trends D.

[muon2]

No comments?

Here's my map that embeds the CD 1 I created above using neutral redistricting principles. I sought minimal county splits while maintaining some degree of compactness. The only splits are for CD 1 to comply with section 2 (50.04% BVAP), for the two large counties, and for two other counties to get all CDs within 1500 of the ideal size. Microchops of counties smaller than a precinct would be used to get exact population equality. Population deviation and 2008 results are in parentheses. I look forward to comments.



CD 1 (-799, 70.3% Obama) Section 2 compliant, it is about 1/3 from Raleigh-Durham.

CD 2 (+286, 54.0% Obama) Very compact, competitive CD centered on Fayetteville with only one split. Laurinburg in Scotland is in this CD allowing Ft Bragg to act as a natural border.

CD 3 (-962, 51.5% McCain) Competitive CD and more compact than most for this region. Whole counties only with half the population in the Greenville-Goldsboro area and the rest along the northern sounds.

CD 4 (+42, 51.6% McCain) Competitive CD with county splits to accommodate CD 1. Wake is divided so that it separates Raleigh in this CD from Cary in CD 13. Wake is the only county with a 3-way split.

CD 5 (+1029, 55.5% McCain) Compact CD centered on Winston-Salem. Only Wilkes is split with only about 7K shifted to CD 9.

CD 6 (-319, 54.1% Obama) Very compact, competitive CD centered on Greensboro with no county splits. The population of Triad and its suburbs is going to be in 2 CDs. This is the cleanest split I could construct.

CD 7 (-187, 56.2% McCain) Dedicated district for the southern coastal region with no county splits and almost no population deviation.

CD 8 (+948, 58.9% McCain) I 73 and US 74 provide the natural corridors linking this district. There are two partial counties to equalize population with CD 2 and 12.

CD 9 (+1225, 62.6% McCain) Reasonably compact CD with only one county split to equalize population with CD 5. Roughly equal population is in western Charlotte metro and in the Blue Ridge foothills.

CD 10 (-1440, 61.9% McCain) Very compact CD with whole counties only. The CD is entirely in the central Piedmont.

CD 11 (+224, 52.0% McCain) Potentially competitive CD with whole counties only. Anchored by Asheville it connects all the counties west of the Blue Ridge.

CD 12 (+580, 67.1% Obama) Entirely within Mecklenburg county it is 34.2% BVAP.

CD 13 (-631, 58.7% Obama) Reasonably compact CD with county splits due to CD 1 and the division of Wake. Cary is entirely in this CD.

In the meantime I've calculated the PVIs for the districts. For whole county CDs I used the actual 04 and 08 votes. I've approximated the other CDs using 08 and a weighting factor based on the 08 votes in the CDs that I can directly determine. It shifts the PVI between 0.5 to 1.0 in favor of the GOP compared to the 08 numbers alone. Of course there's a history of Dems holding seats a few PVI to the GOP.

CD 1: D+16
CD 2: D+0
CD 3: R+6
CD 4: R+6
CD 5: R+11
CD 6: R+0
CD 7: R+10
CD 8: R+15
CD 9: R+16
CD 10: R+16
CD 11: R+6
CD 12: D+13
CD 13: D+5

[Miles]

muon, even in your neutral map, it would probably be 7 Democrats, possibly 8 without Jones.

I like that you've got 5 whole-county CDs while still keeping CD1 legal.