Fair redistricting: Illinois

[muon2]

That was why I came up with a plan B back then.

Yes, your Plan B is much more acceptable   To reiterate Muon's point, counties do matter quite a bit in Minnesota, including around the Cities.  Although I know the objective metrics rightly can't force it, folks around the Cities would breathe a lot easier if (1) Hennepin only had two districts (Minneapolis vs. suburbs + random other suburban county bits) and (2) Ramsay was kept whole... although AustralianSwingVoter's Map 1 split isn't a bad split if you must go three ways (save for the random Savage jaunt in District 2).

It's interesting to note that when Ramsey stays together it packs the Dems and makes the remaining suburban seats more Pub. It shows that the county can be more important than partisan fairness. The Hennepin-Ramsey rivalry reflected the suburban counterpart of the Minneapolis-St Paul rivalry, which is a real thing, not to mention the library and park services provided by the counties to areas outside of Mpls and St Paul proper (at least when I was living there).

[muon2]

Muh CounTIes

If you lived in one of those MN counties, particularly the smaller ones, I think you would have a different attitude. On Solid's list, I've lived in two, plus one not on the list. They mattered to the residents, sometimes more than the municipality.

Do counties matter so little in FL? Are there no shared services at the county level?

I don't think I would. You say they "matter" over and over again, but what does that mean? In what sense? And how does that translate to "these people CANNOT be in different congressional districts"
I don't live in Florida either...

My bad, I was going off of your avatar.

I'm saying that people identify with their counties, and those shared services create a community of interest. I'm not saying that you can't split them, but that you should split them as little as possible absent other compelling reasons. States with neutral redistricting laws also recognize counties as entities that should be preserved intact to the extent possible.

[muon2]

It's not against the rules here, but IA does not permit county splits for CDs. It's interesting that IA, WV, and AR (no splits before 2010) are all up at once, all states that avoided county splits can do so reasonably.

Polk+Dallas make up the only UCC in IA.

My dad's family is from Cedar Rapids, and as someone with roots there, Cedar Rapids fits better with Waterloo than with the Quad Cities on the Mississippi.

[muon2]

Since IA is generally considered to be the gold standard for fair redistricting, it seems only right that one of the submissions should be the official plan. I'll make that my submission. There are no split counties, the UCC is intact, and the population range is a mere 76 or 0.01%.



CD 1: (-41) D+1.0
CD 2: (+35) D+0.7
CD 3: (+23) R+1.4
CD 4: (-18) R+11

[muon2]

Hmm, we have 2 plans that respect the integrity of County boundaries, and 2 plans that ignore the integrity of County boundaries.

the world ain't gonna end if a county's in two different congressional districts

There is no reason not to have a whole County plan if it can be done without high deviation, or really ugly shapes.

Such is in fact possible to do in Nebraska, so it should be required.

This debate is why we started looking at a scoring UCC integrity. The Omaha UCC is Douglas and Sarpy and is larger than one CD. If the UCC is kept whole there is a county chop. If counties are kept whole there is a penalty for the UCC chop. The chop score is the same either way. By recognizing the metro area as a community of interest like counties it equalizes the two plans.

BTW, if you really wanted a whole county plan with low deviations (like IA), this one has deviations of -34, +59 and -24. It does chop the UCC of course.



If you didn't care about compactness you can get the deviations down to -8, +12 and -3. The shape of the CDs in this plan show why erosity is an important measure to consider beyond the population with whole counties.



However, the advantage of maintaining the UCC is that it tends to produce less erose plans as seen in the two submissions, especially singletxguyforfun. What I can't tell is whether either of the UCC plans preserves the municipal boundaries, since there is no detail. If they don't then they would have more chops than the whole county plans.

[muon2]

1. R+11 (+10)
2. R+3 (+109)
3. R+28 (-118)

I think there is a misplaced precinct. I get that the deviation for your CD 3 is +1962.

[muon2]

Nebraska Non-Partisan plan.

My nonpartisan redistricting plan for Nebraska splits only one county.

District 1 R+10.71 - 45.1 - 53.3
District 2 R+02.96 - 50.5 - 48.2
District 3 R+28.23 - 29.7 - 68.6

If I'm guessing correctly at your Sarpy chop, you preserve city lines, but chop two townships.

I would note that if you wanted to reduce erosity by two points with a slight increase in inequality (though no change in the inequality score) you could swap the following counties.

CD 1 to CD 3: Dixon, Dakota, Thurston, Stanton, Colfax, Cuming, Burt
CD 3 to CD 1: Saline, Jefferson, York, Fillmore, Thayer, Hamilton, Clay, Nuckolls

The deviations would be +149, +87, -235

[muon2]

This is my busy work season, so I had to drop off. However, I put IL together during my spring break and posted it elsewhere, so I'll put it in as my submission.

Here is the Chicagoland detail.

Here's a preview of my division of Chicago/Cook that I will use when cvparty gets to IL. 10 CDs fit into Cook+Dupage+Lake+McHenry, and all CDs are within 0.5% of the quota. The idea is that south and west Cook are used to build minority districts and then the remaining area is treated as if the townships are unchoppable units.



CD 1 is BVAP 53.9% and is 79% outside of Chicago and follows township lines in Cook.
CD 2 is BVAP 52.3% and is over 60% in Chicago.
CD 3 is HVAP 59.6% and is over 70% in Chicago.
CD 4 is minority-majority with BVAP 42.5% and is about half in Chicago.
CD 5 is HVAP 36.8% and is over 60% in Chicago.
CD 6 is entirely within Chicago.
CD 7-10 follow township lines for boundaries.

And here is how it fits into the whole state.

Here's the rest of my version of IL, with only Cook DuPage, Lake and Will chopped. Obama won 17 of 18 in 2008 and in 15 he exceeded his national percentage. Today the PVI is 9 D, 6 R, 3 even.

Here are the deets by CD:

CD 1: (-2611); D+30; BVAP 53.9%
CD 2: (-2880); D+28; BVAP 52.3%
CD 3: (-2433); D+27; HVAP 59.6% (exceeds the standards set by the 7th circuit for the VRA)
CD 4: (-1606); D+33; BVAP 42.5%
CD 5: (-264); D+22; HVAP 36.8%
CD 6: (+706); D+27
CD 7: (+236); D+15
CD 8: (+455); D+3.4
CD 9: (+2684); D+2.8
CD 10: (+1404); R+0.7
CD 11: (-2490); D+1.0
CD 12: (-2122); D+0.3
CD 13: (+1299); R+3.7
CD 14: (+2267); R+5.8
CD 15: (+3331); R+9.2
CD 16: (+1098); R+6.9
CD 17: (+478); R+5.5
CD 18: (+446); R+20

[muon2]

ASV - are your racial/ethnic numbers population or voting age population (VAP)? They look like the former which are not used for VRA purposes. The VRA relies on the citizen VAP, which is close to the VAP for the black population. For Hispanics the non-citizen population is too high to use VAP directly. The courts have established 59.2% HVAP as sufficient to meet the VRA.


I know that the exercise does not require VRA compliance, but we should probably quote the right stats.

[muon2]

ASV - are your racial/ethnic numbers population or voting age population (VAP)? They look like the former which are not used for VRA purposes. The VRA relies on the citizen VAP, which is close to the VAP for the black population. For Hispanics the non-citizen population is too high to use VAP directly. The courts have established 59.2% HVAP as sufficient to meet the VRA.


I know that the exercise does not require VRA compliance, but we should probably quote the right stats.

I don't use the VAP solely because if I use it it takes even longer to load a state, and given that I use a 2012 MacBook Air you can probably imagine how glacial the loading already is.

Here's what the comparison between pop and VAP is for my Cook CDs:

Here is the Chicagoland detail.

Here's a preview of my division of Chicago/Cook that I will use when cvparty gets to IL. 10 CDs fit into Cook+Dupage+Lake+McHenry, and all CDs are within 0.5% of the quota. The idea is that south and west Cook are used to build minority districts and then the remaining area is treated as if the townships are unchoppable units.



CD 1 is BVAP 53.9% and is 79% outside of Chicago and follows township lines in Cook. (Bpop 55.8%)
CD 2 is BVAP 52.3% and is over 60% in Chicago. (Bpop 53.5%)
CD 3 is HVAP 59.6% and is over 70% in Chicago. Hpop 65.2%
CD 4 is minority-majority with BVAP 42.5% and is about half in Chicago. (Bpop 45.5%)
CD 5 is HVAP 36.8% and is over 60% in Chicago. (Hpop 41.9%)
CD 6 is entirely within Chicago.
CD 7-10 follow township lines for boundaries.

Since your CDs largely overlap mine, it's probably a good assumption that the differences between the pop and VAP will be similar in your plan.

Illinois Non-Partisan plan.

My non-partisan redistricting plan for Illinois doubles the number Hispanic VRA districts, while maintaining the number of African American VRA districts. Only 4 counties are split, of which only 2 were not required.

District 1 D+28.70 - 81.6 - 17.8 - 51.3 African American (est 50.1% BVAP)
District 2 D+28.03 - 79.7 - 19.6 - 51.0 African American (est 49.1% BVAP)
District 3 D+21.30 - 70.1 - 28.6 - 60.6 Hispanic (est 55.0% HVAP)
District 4 D+29.76 - 78.9 - 19.7 - 50.4 Hispanic (est 45.3% HVAP)
District 5 D+26.90 - 76.3 - 22.5
District 6 D+04.90 - 56.5 - 42.2
District 7 D+36.71 - 88.2 - 11.0 - 50.7 African American (est 47.7% BVAP)
District 8 D+06.20 - 59.1 - 39.7
District 9 D+03.72 - 56.2 - 42.5
District 10 D+06.26 - 59.5 - 39.4
District 11 D+01.36 - 55.7 - 43.1
District 12 R+05.47 - 54.7 - 43.8
District 13 R+11.85 - 48.2 - 50.2
District 14 R+05.23 - 51.4 - 47.2
District 15 R+19.12 - 44.6 - 53.6
District 16 R+04.60 - 53.1 - 45.1
District 17 R+03.52 - 54.0 - 44.5
District 18 R+06.94 - 50.1 - 48.2

Only one of the three black CDs is probably over 50% BVAP, though given Chicago's politics all three would probably elect the choice of the black population. However, if the Dems didn't like the map they would probably challenge the lack of at least 2 CDs over 50% in a VRA case.

The 59.2% HVAP threshold for Cook Hispanic CDs is based on the performance of the current CD-4, considering the lack of HCVAP numbers. Since it successfully has elected Gutierrez over the last 20 years it is considered to have demonstrated the ability to elect the candidate of choice. However, Chicago has a lot of recent history of white candidates defeating Latinos in state and local districts with even higher HVAP. Both of your Hispanic CDs would likely be won by white candidates, though the Latinos would certainly be an important voting bloc.