Flo's and ElectionGuy's Maps

[muon2]

Democratic sweep

1 (blue): 52-46 Obama
2 (green): 51-46 Obama
3 (purple): 59-39 Obama
4 (red): 60-38 Obama

Are these are the 2008 numbers from DRA? If so then the PVIs of the CDs are 4 points more Pub than the 2-party vote in the presidential race. For example the list above would be R+2, R+3, D+5, D+6 abd would favor a 2-2 delegation. To be more precise the PVI can be estimated from the 2008 presidential vote by taking Obama's share of the two party vote and subtracting 3.7%.

[muon2]

Flo's Maps: Now with Labels!



I wanted to make a map with a majority of districts voting 54%+ Obama so it would go to him in the 2012 election, with the remaining Republican districts grossly and unnecessarily gerrymandered.

And I think I accomplished that.

Missouri 1:
54.5% Obama
44.3% McCain

Missouri 2:
54.7% Obama
44.4% McCain

Missouri 3:
55.4% Obama
43.5% McCain

Missouri 4:
53.6% Obama
44.9% McCain

Missouri 5:
54.8% McCain
43.5% Obama

Missouri 6:
60.0% Obama
38.9% McCain

Missouri 7:
64.6% McCain
33.9% Obama

Missouri 8:
64.0% McCain
34.5% Obama

Since 5, 7, and 8 are all McCain CDs, why the need to gerrymander them? Your others are all reasonably shaped. As a map for the Dems, there's no reason to give the Pubs grounds for a challenge no matter how slim the chance, and the gerrymanders really only affect the Pubs.

[muon2]

The statement seemed so strange that I just failed to read it literally. I didn't anticipate that someone would want to mix two so unrelated and somewhat antithetical goals.

[muon2]

Nothing too extreme here, but a Colorado republican gerrymander 4-3.



Denver:



Red: 51.6% McCain, 46.6% Obama
Blue: 74.8% Obama, 23.6% McCain
Yellow: 64.9% Obama, 33.3% McCain
Green: 58.3% Obama, 40.0% McCain
Orange: 50.4% McCain, 48.2% Obama
Purple: 59.0% McCain, 39.3% Obama
Lime Green: 53.5% McCain, 45.0% Obama

It looks like you could regroup the green and purple districts to make it 5-2.

[muon2]

Here's a legitimate whole state plan I put together for KS in our redistricting thread a couple of years ago. It was designed to keep all the urban areas intact and compact. CD 1 has no urbanized area at all, and only Harvey county is even in a metro area as a commuter county.

The Manhattan metro is split (as also under the current map - Junction City and Manhattan would be better off together. If someone believes they ought to be in the 2nd together, fine with me. Except it does require taking the 1st all the way to either the northeast or the southeast corner.) The Wichita metro is also split in that proposal, and that is new. Though it's not a very consequential split, slicing off one satellite-town-dominated county of 20k people.

Honestly, if you want to keep Manhattan out of a reasonably designed first district... go repopulate the High Plains.

In my first map I kept Manhattan and Junction City together, but in CD 1. I can put them into CD 2 with KC and Topeka to make that district more compact. If I also make CD 4 more of a compact block, CD 1 runs to the SE corner. This version has a maximum deviation of 512 using only whole counties.

[muon2]

The figures and location for the limegreen district strongly suggest 11 is doable.

12 is not, in fact, hard to do.



1 50.1 - 49.2
2 49.9 - 49.5
3 50.2 - 49.1
4 53.9 - 45.0
5 55.8 - 43.1
6 51.3 - 44.7
7 50.1 - 49.1
8 51.2 - 47.9
9 49.8 - 49.4
10 50.1 - 49.1
11 49.8 - 49.2
12 49.9 - 48.8
13 32.5 - 66.3

If you want over 50%, D2 is not hard to do (just move some more Raleigh Blacks across to the 3rd and mess with the 2nd-3rd line), 9 and 11 clearly are doable as well. 12 probably is not. If I missed any trick there then it's not in the book.

With that huge surplus in the fifth, ie fairly far west... one can't help wondering if 13 contiguous Obama districts might not be just about theoretically possible somehow.

Mathematically it could be possible to go 13-0. The R margin in 13 is 33.8% so you need to shift at least half of that (16.9%) into CD 13 to flip it as well. The total D excess in your CDs 1-12 is 38.0%, so that leaves only 4.2% that doesn't need to be shifted. If you left 1,2,3 and 7 alone that takes up 3.2% of your excess. All other districts would have to slim down to a 0.1% margin and somehow shift enough Dems into 13.

[muon2]

Arizona Congressional:



Phoenix/Tucson Closeup:



Blue (Tuscon): 58.1% Obama, 40.7% McCain. 55.1% White, 34.8% Hispanic
Green (Inner Phoenix): 65.9% Obama, 32.9% McCain. 55.5% Hispanic, 29.7% White.
Purple (Chandler/Gilbert): 56.6% McCain, 42.4% Obama. 71.8% White, 15.2% Hispanic.
Red (Mesa/Scottsdale): 57.3% McCain, 41.5% Obama. 74.2% White, 16.9% Hispanic.
Yellow (Northern Phoenix): 55.7% McCain, 43.1% Obama. 72.6% White, 17.7% Hispanic.
Teal (Glendale/Peoria): 58.0% McCain, 41.0% Obama. 68.1% White, 21.8% Hispanic.
Gray (SW Arizona): 58.2% McCain, 40.8% Obama. 59.9% White, 32.8% Hispanic.
Slate Blue (SE Arizona): 58.9% McCain, 39.9% Obama. 64.9% White, 23.0% Hispanic.
Lime Green (Northern Arizona): 55.5% McCain, 43.1% Obama. 70.6% White, 15.7% Native, 10.6% Hispanic.

This is a neat little republican gerrymander. And the thing I like about it is that it looks like it was neatly constructed and not purposefully gerrymandered.

(Racial demographics are Voter Age Percentage, so they're much whiter than they actually are).

It's a nice gerrymander, but I think it loses to the VRA. The actual map has two CDs at over 60% Hispanic (not VAP).

[muon2]

Montana:



This is the one I used in a earlier post for states with 3 electoral votes currently anticipating 4 electoral votes.

MT-1 (blue): Obama 2008, Romney 2012. Lean R
MT-2 (green): McCain 2008, Romney 2012. Safe R

This is my take on MT with 2 CDs. I used the 2012 county estimates and projected them forward to 2020. That allowed me to include Bozeman instead of Helena in the west CD.

[muon2]

Illinois:



IL-10: 53.3% Obama, 45.1% McCain = Lean D
IL-11: 50.4% Obama, 47.9% McCain = Lean R
IL-12: 49.3% McCain, 49.0% Obama = Likely R
IL-13: 49.2% McCain, 49.1% Obama = Likely R



IL-1: 82.6% Obama, 16.8% McCain = Safe D
IL-2: 88.1% Obama, 11.1% McCain = Safe D
IL-3: 84.5% Obama, 14.4% McCain = Safe D
IL-4: 75.3% Obama, 23.4% McCain = Safe D
IL-5: 58.9% Obama, 40.0% McCain = Likely D
IL-6: 55.2% Obama, 43.4% McCain = Lean D
IL-7: 54.2% Obama, 44.6% McCain = Lean D
IL-8: 57.1% Obama, 41.7% McCain = Lean D
IL-9: 55.2% Obama, 43.2% McCain = Lean D

I'm probably being a bit generous with my ratings as well. Obama numbers in the Midwest in general (except Minnesota and Iowa) are very misleading.

You need to adjust a lot to compensate for the 2008 numbers in IL. IL 6 in your map would perform better than Roskam's pre-2012 district, DuPage minus the northeast corner is safe R. Will plus Kendall in your IL-7 would be likely R (think Weller from the 2000's). With so much of McHenry. your IL-6 is probably toss-up to lean R. Your IL-9 would be well suited for someone like Manzullo and is likely R. Schock would be a lock in your IL-10, so it too would be likely R. The other downstate districts in that map (11-13) are all safe R.

[muon2]

Pennsylvania:



1 (blue): 89% Obama, 11% McCain = Safe D.
2 (green): 69% Obama, 30% McCain = Safe D.
3 (purple): 56% Obama, 43% McCain = Lean D.
4 (red): 58% Obama, 41% McCain = Lean/Likely D.
5 (yellow): 56% Obama, 43% McCain = Lean D.
6 (teal): 55% Obama, 44% McCain = Lean D.
7 (grey): 56% McCain, 43% Obama = Safe R.
8 (light purple?): 54% McCain, 45% Obama = Likely R.
9 (light blue/turquoise): 55% McCain, 44% Obama = Likely R.
10 (pink): 50% McCain, 48% Obama = Toss-up.
11 (lime): 61% Obama, 38% McCain = Safe D.
12 (some blue which is hard to explain): 52% McCain, 47% Obama = Toss-up.
13 (brown-ish): 60% McCain, 39% Obama = Safe R.

Pretty competitive map.

Any district McCain won is by 5 points or more is going to be safe R. Any district McCain won at all is going to be likely R. Districts where Obama was held under 53% will still be lean R. Remember that 2008 was the largest percentage win for a Dem since LBJ and shouldn't be viewed as typical for determining congressional outcomes.

[muon2]

Any district McCain won is by 5 points or more is going to be safe R. Any district McCain won at all is going to be likely R. Districts where Obama was held under 53% will still be lean R. Remember that 2008 was the largest percentage win for a Dem since LBJ and shouldn't be viewed as typical for determining congressional outcomes.

Would that really apply to western PA though? e.g, Critz barely lost a McCain +9 seat in 2012 and Altmire held a McCain +11 seat pre-redistricting.

That's about the power of incumbency. I was referring to ratings like the PVI that best apply to open seats. With only the benefit of 2008 data, an approximate PVI can be found by subtracting 53.7% from the Dem share of the two-party vote. Positive results are D-leaning and negative results indicate an R-leaning district.

[muon2]

NJ-1 (Blue) = Safe D
56.2% Obama
43.8% McCain

NJ-2 (Green) = Safe D
54.4% Obama
45.6% McCain

NJ-3 (Purple) = Likely D
53.7% Obama
46.3% McCain

NJ-4 (Red) = Safe D
56.8% Obama
43.2% McCain

NJ-5 (Yellow) = Safe D
55.2 Obama
44.8% McCain

NJ-6 (Teal) = Safe D
55.6% Obama
44.4% McCain

NJ-7 (Grey) = Safe D
65% Obama
35% McCain

NJ-8 (Slate Blue) = Safe D
57.4% Obama
42.6% McCain

NJ-9 (Cyan) = Safe D
56.3% Obama
43.7% McCain

That's actually a map with a number of competitive districts. Christie probably won from 4 to 6 of those 9 in 2009 and 8 of 9 in 2013. They certainly aren't safe D, mostly lean to likely D.

[muon2]

CD1 (blue): 60.8% McCain, 38.6% Obama
CD2 (green): 52.0% Obama, 47.6% McCain
CD3 (purple): 60.6% McCain, 38.9% Obama
CD4 (red): 51.4% Obama, 47.9% McCain
CD5 (yellow): 72.0% McCain, 26.8% Obama
CD6 (teal): 64.2% McCain, 34.8% Obama
CD7 (gray): 72.4% McCain, 26.7% Obama

5 Republicans
2 Democrats

This look like you should have been part of our marathon discussion on AL last year.

[muon2]

A neater and better version of an Iowa republican gerrymander (My other one is one the first page). Numbers are:

blue: 60.6-37.9 Obama
green: 58.9-39.4 Obama
purple: 51.1-47.3 McCain
red: 49.4-48.7 McCain

If you take the blue district and add Louisa and all of Dubuque counties while dropping Delaware county you get a compact whole county CD still at 60.6% Obama. A CD at 52-46 Obama is R+1, so it seems that the GOP could gamble on splitting the rest of the state into 3 CDs that are all R+1 or R+2. All four CDs went for Obama, but A 3R - 1D split wouldn't be unusual given the competitive districts.

[muon2]

I'm sorry, is there some rule of yours that's a pet peeve?? As you can see, I almost used all whole counties, only splitting two between the 1st and 4th.

I would suggest two rules that are frequently invoked here. The first is to avoid splitting two counties between the same two districts. The second is to avoid splitting counties except when needed to have sufficient population equality or to reduce district erosity.

[muon2]

Chicago Area:

Your orange CD is quite bizarre, but if it's going to be bizarre why not accommodate another minority district. My guess is one could make a 50% HVAP district linking Aurora, Elgin, Addison, and Melrose Park.

[muon2]

I started to work on a series of cuberoot apportionment maps, but Alabama defeated me right out of the bat.  When you need districts to be 28,451 in size average and you have some 20K+ precincts, it's just not doable with any sort of reasonable deviation.  Why the heck does Alabama have such large precincts?

I've been told that it has to do with VRA preclearance. While Section 5 was in effect for AL any precinct changes had to go through the DOJ. It was easier just to leave the precinct boundaries in place rather than open up the process to VRA scrutiny.