New Jersey Gerrymander

[Verily]

I tried to gerrymander New Jersey for both parties. The Republican gerrymander failed miserably; it couldn't get any better than the current 7D-6R for the Republicans. The Democrats, however, could make New Jersey 11D-2R with the proper gerrymandering:

Target population: 647,258 per district, 2000 Census
Safe means D/R+13 or greater
Likely means D/R+8 to +12
Lean means D/R+0 to +7

CD1
100% of Sussex
100% of Warren
100% of Hunterdon
59% of Morris
Population: 646,017
2004 Results: R+22
Safe Republican

CD2
100% of Passaic
18% of Bergen
Population: 648,190
2004 Results: D+10
Likely Democrat

CD3
73% of Bergen (probably my district)
Population: 645,406
2004 Results: D+5
Lean Democrat

CD4
100% of Hudson
4% of Bergen
Population: 644,340
2004 Results: D+34
Safe Democrat

CD5
65% of Essex
19% of Morris
5% of Bergen (poor Bergen got split 4 ways; it's what happens to counties that are too big and in a corner)
Population: 649,408
2004 Results: D+31
Safe Democrat

CD6
51% of Union
35% of Essex
22% of Morris
Population: 647,714
2004 Results: D+23
Safe Democrat

CD7
100% of Somerset
49% of Union
12% of Middlesex
Population: 643,555 (smallest district)
2004 Results: D+7
Lean Democrat

CD8
64% of Middlesex
27% of Monmouth
Population: 646,235
2004 Results: D+8
Likely Democrat

CD9
51% of Monmouth
44% of Mercer
24% of Middlesex
Population: 648,177
2004 Result: D+5
Lean Democrat

CD10
100% of Ocean
22% of Monmouth
Population: 646,282
2004 Result: R+19
Safe Republican

CD11
100% of Burlington
56% of Mercer
6% of Camden
Population: 650,356 (largest district)
2004 Result: D+13
Safe Democrat

CD12
100% of Atlantic
100% of Cape May
58% of Camden
Population: 650,059
2004 Result: D+12
Likely Democrat

CD13
100% of Salem
100% of Cumberland
100% of Gloucester
36% of Camden
Population: 648,612
2004 Result: D+10
Likely Democrat

[○∙◄☻¥tπ[╪AV┼cVê└]

Yes, it's pretty clear that "bipartisan commission" drew a Republican gerrymander.

[Verily]

The map is inelegant, and the areas on the map don't correspond to the actual areas that would be in the CDs (I just went by area and eyeballed it).

[○∙◄☻¥tπ[╪AV┼cVê└]

Would the districts be 8D - 5R under a "fair" map?

[Verily]

Probably, or maybe 9D-4R. Frelinghuysen (moderate Republican) could probably hold on to a D+5 district or so, though, so 8D-5R would be more likely in the short term.

Obviously, I went to some pains to dilute the largest Republican counties (Monmouth and Morris), both of which are split three ways with two chunks in Democratic districts and one chunk in a Republican district. They would both probably have their own districts in a more fair map, and the others would be more safely Democratic, making 9D-4R (which would also have the convenience of putting Frelinghuysen in a Republican district anyway; he's in Morris).

[Verily]

Here's what the partisan breakdown looks like visually; very reminiscent of the 2004 Presidential election map with Dem strength on the northeast-southwest axis of the Turnpike and GOP strength in the central east and northwest.

[Conan]

Great job. However theres no sense to make such safe dem districts. They need to share the wealth. As it stands now here are the 2004 results for president in each district:

NJ-1: Bush- 39 Kerry- 60
NJ-2: Bush- 50 Kerry- 49
NJ-3: Bush- 51 Kerry- 48
NJ-4: Bush- 56 Kerry- 43
NJ-5: Bush- 57 Kerry- 42
NJ-6: Bush- 56 Kerry- 43
NJ-7: Bush- 53 Kerry- 46
NJ-8: Bush- 41 Kerry- 58
NJ-9: Bush- 41 Kerry- 58
NJ-10: Bush- 18 Kerry- 81
NJ-11: Bush- 57 Kerry- 42
NJ-12: Bush- 45 Kerry- 54
NJ-13: Bush- 31 Kerry- 68

NJ-2,3,7 can all go dem but 2 and 3 will have to be open.
Payne in NJ-10 and Sires in NJ-13 need to share the wealthy and we can easily shutout the reps.

[Verily]

The difficulty is that Hudson County is surrounded by Democratic strongholds and the gigantic Bergen County, and then Essex and Union Counties border only marginally Republican Somerset County and moderately Republican Morris County. I tried, with that gerrymander, to distribute the strength of Hudson, Essex, Union and Camden as far outward as I could. Splitting Camden three ways worked well and easily. Unfortunately, it's not so easy to do so in the north.

Theoretically, one could have two districts crossing Hudson and Bergen that were both safely Democratic instead of one safe Democratic one mostly in Hudson and one less safe but still fairly safe Democratic one entirely in Bergen, but I preferred to make the geography work out. Doing so might also be considered an attempt to dilute the Hispanic bloc in Hudson and the Korean and Filipino blocs in Bergen.

[Verily]

With the intention of comparing a generic, nonpartisan map to the gerrymander, I've created a nonpartisan version of a NJ CD map. It strives first and foremost to keep districts about as wide as they are high. The first 4 CDs remain unchanged, as does the 7th. This results in a 9-4 split, as I predicted.

CD1
100% of Sussex
100% of Warren
100% of Hunterdon
59% of Morris
Population: 646,017
2004 Results: R+22
Safe Republican

CD2
100% of Passaic
18% of Bergen
Population: 648,190
2004 Results: D+10
Likely Democrat

CD3
73% of Bergen (probably my district)
Population: 645,406
2004 Results: D+5
Lean Democrat

CD4
100% of Hudson
4% of Bergen
Population: 644,340
2004 Results: D+34
Safe Democrat

CD5
76% of Essex
5% of Bergen
Population: 647,367
2004 Results: D+40
Safe Democrat

CD6
51% of Union
41% of Morris
24% of Essex
Population: 649,755
2004 Results: D+15
Safe Democrat

CD7
100% of Somerset
49% of Union
12% of Middlesex
Population: 643,555 (smallest district)
2004 Results: D+7
Lean Democrat

CD8
86% of Middlesex
Population: 645,139
2004 Results: D+14
Safe Democrat

CD9
100% of Monmouth
5% of Mercer
2% of Middlesex
Population: 647,842
2004 Result: R+9
Likely Republican

CD10
100% of Ocean
32% of Burlington
Population: 646,402
2004 Result: R+15
Safe Republican

CD11
95% of Mercer
68% of Burlington
5% of Camden
Population: 646,577
2004 Result: D+18
Safe Democrat

CD12
95% of Camden
64% of Gloucester
Population: 646,476
2004 Result: D+21
Safe Democrat

CD13
100% of Atlantic
100% of Cape May
100% of Salem
100% of Cumberland
36% of Gloucester
Population: 657,283 (almost every other district was closest at just below, so this district is a too big; you'd need to split percentages across the state to get it smaller, but the partisan makeup wouldn't change significantly)
2004 Result: R+1
Marginal Republican

[muon2]

I did a version in 2005 that broke the state down to the town level. Here is a relevant passage:

You must be logged in to read this quote.

And this was the map (with NY, PA and DE)