Blumenauer Congressional Plan

[Oryxslayer]

Yeah as far as Alabama goes, you can keep two white seats south of Montgomery, allowing you to stretch the Black belt seat all the way across. You can obviously do it neater than this, but the maps in service of this hypothetical I have already drawn were created tried to replicate the political decisions and outcomes of the 2021/2 redistricting process just with these alternative seat totals.

[Oryxslayer]

I don't like this. It benefits the Republicans here!

snip

In most cases the state's geographic bias should not shift by adding more seats - generally you just got to detach yourself from the existing districts geographic cores and create new pairings that allign better with the new population totals. Here's such a case in MN where you now use Washington as the building block and aim to nest 3 core seats in Hennepin. See my comments above about the framework I tried to draw these maps in, so MN comes to 5. Obviously if you want more competitiveness there are things you can do with districts 1, 8, and 10 like include Fargo in 8.  One other thing about MN is that the Washington - Duluth county linkage is only slightly overpopulated, if one was to desire such a gerrymander.





One state where the geographic bias does shift is Michigan. It all comes down to 3 things: Macomb now being large enough that the south of the county needs to be paired with other blue areas OR you nest the seat in the southern towns and it overall leans Dem, Oakland being a deal larger than two seats so you'll likely end up sinking some GOP voters in 2 dem leaning seats based on how the pop works out, and Kalamazoo now having enough demographic weight to not be drowned out by her neighbors so easily.

[Oryxslayer]

Reach over to the Fargo area. Technically Moorhead.

You know what I mean

[Oryxslayer]

Nah, the court would have imposed a 9-9 map, just like how the one presently in place is 7-7. Their Dems after all.



Bonus points for this map are the questionable decisions in regards to COIs and regions I made, since the court's map made so many! Chapel Hill - North Greensboro so as to allow for a minority seat in the rest of the metro, Keeping the Gastonia arm even though Mecklenburg is now close to perfect population, cutting Wake East/West cause Raleigh has to be split anyway so why not create an access seat, and whatever happens to Lejune since Jacksonville is yanked out of Onslow. Kinda like what they did to Fayetteville, W-S, Raleigh and Wayne IRL.

Also, its now an open question of what happens in NC-11. Chuck Edwards won the Primary cause of Henderson and Buncombe overlapped with his Senate seat. Would he even have run if Henderson was split unfavorably? Would he have still won the primary? Could one of the other candidates become the consensus vs Cawthorn and win? Cause if Cawthorn ends up as the hypothetical 2022 GOP candidate, he loses based on the underperformances displayed by other members of his young extremist clique like Boebert.

[Oryxslayer]

I will note that I'm playing around with New York right now and surprisingly, once again the Haredim get screwed. You would think that with smaller districts the sotuh Brooklyn Whites would get their own seat. However, the smaller seats mean that there should in most circumstances, given the legal and community guidelines, a new Asian access seat already sliding in there. With much of the Bensonhurst and Gravesend area already getting used up, there just isn't enough precincts favorable to the GOP left and throwing Park slope in there defeats the whole purpose of such a seat. The other option is the ugly Rockaway gerrymander, but beyond it's obvious downsides, the Rockaways already fit perfectly with Long Island based on population deviation.

Which brings me to my second comment on NY: it really likes the 10 extra seats from County and community line purposes. Upstate just...works into nest county groups in most cases. Manhattan is almost perfectly 3 seats, and Staten Island + Bay Ridge/Ft Hamilton is only slightly overpopulated. Long Island + Rockaways and minus North Hempstead town is 5 seats. Its all a little too perfect.

[Oryxslayer]

Gave AZ a go. Pretty happy with this.

How many hispanic majority seats are there?

4 are majority/plurality hispanic, 2 others are majority-minority as well.

2020 CVAP please

[Oryxslayer]

My attempt at NY. Since the state was remapped by a court master, this map is intended to respect COIs unlike some of the previous ones I posted here. And like I described above, things kinda fall into place a little too well with 10 extra seats.





Starting with upstate you can kinda see how things nest almost perfectly into county groupings. Only in the west does this break down a bit but it still is in a neat multi-county grouping of ~5 seats. Politically, it is also competitive, though once again I have to express my dislike with DRA's partisanship color categories. 54-44 NY-18 around Poughkeepsie is the same color as 57-41 NY-24 in Syracuse and 58-39 NY-20 in Albany. Anyway Trump won the three seats less Blue than NY-18 and barely lost NY-18 in 2016, so there's plenty of opportunity here - GOP would probably win all three in the 2022 environment.





Long Island also basically draws itself. Dropping North Hempstead and adding in the Rockaways is perfectly 5 seats. Islip and Babylon are themselves almost exactly 1 seat, so shuffling a few of the suburban 'villages' around the two for population equity while maintaining communities means that seat is done. Which means NY-01 and NY-03 draw themselves through the geography of the region. Throw in making NY-04 a majority-minority seat by CVAP and everything's done. And every seat except the interlocking majority-minority could be competitive.



And like I mentioned above, if you don't have something like this Asian (Chinese') access seat - which itself is just the 17th Senate seat plus a bit - you are ignoring a large and growing COI. And like I said above, this kinda screws over the Orthodox and Eastern Europeans who live next door. Without any politically  'marginal' areas to throw in with them, or the Rockaways to do the ugly linkage, their only potential partner is the White Progressive areas. This defeats the whole point of a Orthodox seat since the added areas would outvote the supposed 'core' of the seat. Add in the possibility of a 4th AA access seat and things don't get any better.

Of course the new Chinese seat isn't as democratic as the rest of NYC - it probably goes GOP in the 2022 environment.



And NYC, which itself is a maze of access seats in most areas. Manhatten is almost exactly 3 seats, so you can restore the appropriate divide between the east and west sides. The existing minority seat reaches from Harlem and northwards.

The Bronx has it's two Hispanic seats and a new Majority African American seat. Though outside of Mt Vernon and Wakefield I'm sure most of the people measured as "African American" are better identified as Caribbean-American and more similar to their Hispanic neighbors.

Four African American seats in Brooklyn and Jamaica, which necessitates some peculiar linkages to balance out the various populations. District 6 covers most of the Chinese population centers in Queens, with the new 29 grabbing what remains - multiethnic or Whiter areas. 39 is the Hispanic seat, which means the existing 7th is used as a White Pack for the Progressive neighborhoods.

All seats described here are Safe D, though I'm sure the one labeled 29 would be close in 2022.

[Oryxslayer]

If you're increasing the number of congressional districts by 50%, I think you need to make an effort to increase the number of VRA seats by the same percentage. You can't draw three black-majority districts in Greater Detroit by CVAP without squiggly lines, but you can easily draw three fairly compact districts where black voters would be a majority in the Democratic primary so I think you should.

It's doable just using Wayne and Oakland counties (with an arm reaching up to Pontiac) but much cleaner if you use southern Macomb (where Eastpointe is now black-plurality and Warren and Roseville are much less white than a decade ago). In either case, you don't need to split any cities.

See, I agree and disagree with you here.

In this hypothetical project, you must create new and fair access seats in many states. I already showed of NY where you can create new urban AA and Chinese access seats for example. Every southern state needs at least one more majority-minority seat by VAP/CVAP. One Columbus seat in a Ohio map should be plurality African American.

The issue is that I don't think this is the case with Michigan. MI has a 13% Black alone VAP. That technically is below two African American seats under our present 13 district arrangement, to give you some indication of where things are going. Of course, the concentration of African American population in the Detroit area means that you can get two seats. Under 19 seats, you are not required to draw a third seat.

Which then begs the two big questions: What do you define to be Black access under plurality, and is this infringing on the access of another group? The former question gets at the heart of the Detroit area seats under the 2022 commission legislative maps, and one has to do primary Racially-polarized-voting analysis outside of DRA to find the appropriate percentage based on turnout, candidate choice, and the relevant elections. Based on last year, 42% is probably not enough.

The second question is also crucial, because the commission legislatively did not treat Detroit area Arabs as Whites, which the census does. They drew Access seats for Arab voters in the Dearborn area. And so we must ask ourselves, is a third African American seat going to come at the expense of their interests? Cause Dearborn and Dearborn are now large enough under the smaller districts to carry a lot of weight in the right western Wayne seat. And even if we don't want to create an access seat, we should also be aware that drawing these three hypothetical AA seats in an ideal fashion my force the communities to be only put in the black seats, which would only serve to deny access to one of the groups, as the 2018 elections show.

[Oryxslayer]

It's really not hard to draw three Black-influence districts:

It keeps heavily Arab areas in Dearborn and Dearborn Heights in one district, and unifies them with Hamtramck.

Sometimes it feels like Atlas redistricters just don't want to draw minority-influence seats.

I feel like I have to respond to this: literally go upthread ands you'll find me arguing how a 4th AA access seat and a Chinese opportunity seat override the interests of White Hasidim in NYC. I'll provide another one here. But if you really want more info, DM me.


So I have two big issues with all these 3-AA seats maps. One, the linkage with Pontiac. The commission made a clear choice to destroy the linkage between it and the two AA seats. They did not see why the two cities should be together, and instead viewed it as a gerrymander designed to pack and crack Oakland. This is despite the fact that it's removal and Detroit African American population loss meant that both AA seats would be starting at the mid-40s on VAP. This is also despite the fact they created a senate seat very similar to the old 14th's Oakland arm that goes from Pontiac to Southfield on the border. This was viewed to now expand African American access, and it did not force unequal gerrymanders on the surrounding seats because of the overall seat population size.

It also bears reminding that the old Detroit seats lost ~4% BVAP between 2010 and 2020. So far there is no indication this will stop, and it's something I think both the commission and their RPV expect should have accounted for.

Next, as we saw in the 2018 primary that produced Tlaib, putting the the Dearborns and African American areas together if you have the option of separating them should be avoided. One will come out on top and defeat the candidate(s) of choice of the other. Of course the commission didn't have the option to not pair the two, since they were nesting seats in Wayne, so both got paired in the district already with an Arab congresswoman who by now had taken steps for African American outreach. In a similar vein to many Los Angeles districts on various maps, you shouldn't zero-sum discriminate against one group if you have the option not to.


Which brings me to the map I have had for a while.





9-9 in 2016, 11-7 in 2020. Like I commented upthread, the reduction of district sizes means that a split of Macomb and a comparatively larger Kalamazoo works to reverse democratic geographic disadvantage.

District 10 may seem unorthodox, but from the Michigander's I have conferred with in the past online and in person they are adamant that the rural thumb has more in common with the other rurals to the north along the lakeshore than anything southwards. Putting St. Clair with the Detroit seats afforded me such a luxury this time.

Interestingly all 4 of the new seats - since I started assigning seat number locations from the old 14-district 2010 map - end up in and around Detroit. Guess that's what you get from a GOP gerrymander that gave all seats some rural profile.

The real interesting thing though is that the 4 districts I would define as competitive on this map all have very different profiles.


MI-06 is a classic swing seat that's more conservative than the nation. Obama wins it by a decent percentage in 2012, then Trump in 2016 by a decent percentage, then Biden in 2020 by a tiny amount. Probably would keep electing Upton.

MI-03 goes from Romney in 2012 to a tiny Clinton margin to such a strong Biden result that some might see it as off the field. Probably would have elected the same people as IRL until 2022, then maybe a different dem.

MI-09 still lurches right. It goes from a strong Obama win to a marginal Clinton one, then to Biden by a margin only slightly more Dem than the nation. Probably still elects Levin. The northern border of the seat is designed to try and follow the river.

MI-11 goes from a comfortable Romney win to a marginal trump win, then a Biden win equivalent to MI-09. Probably still has similar electoral history, but 2018 and 2020 elections under similar lines might have been more competitive.

 



Finally we must touch on Detroit, which features this new phallic district nested inside the county. Downloading ACS data and analyzing the seat in GIS gives me a district that is between 25 and 30% Arab + adjacent groups. Putting in the Hispanic neighborhoods and adding in the fact that the outer towns with small Arab populations also have some GOP primary voting whites and you end up with a access seat that will elect a candidate of choice. Though maybe specifically not Tlaib...

[Oryxslayer]

It's really not hard to draw three Black-influence districts:



It keeps heavily Arab areas in Dearborn and Dearborn Heights in one district, and unifies them with Hamtramck.

Sometimes it feels like Atlas redistricters just don't want to draw minority-influence seats.

There's currently zero Black representatives from Detroit. It's not a hypothetical where a 40ish Black VAP could be risky, both current seats have failed to elect Black representatives. As the population decline continues in Detroit, it's safer to make two seats with higher Black CVAP.

It's not good to conflate the race of the representatives with whether or not they are the candidates of choice of the Black community.

2 and 3 are majority Black on CVAP; 1 is plurality white but should easily elect the preferred candidates since a large chunk of those whites are Republicans. It's of course not impossible for the Black candidate of choice to lose a primary, but that's very difficult to guarantee and the non-preferred candidate winning would require substantial support from Black voters in most cases (and you can cook up similar scenarios for higher CVAP numbers too).

Yeah one needs to remember that candidate of choice =/= candidate of same ethnic group. The classic example here is Steve Cohen.

Thanedar is the plurality candidate of choice for example based on the 2022 primary results in AA precincts. Now we can debate whether it is solely cause he had substantial goodwill within his former legislative district, or whether he would have won a 1 v 1 with an African American candidate, but we don't have that information.

Now Tlaib was not the candidate of choice fer her seat really until recently. She won 2018 off just a concentrated base of support, and then in 2020 won landslide victories in all the non-Detroit parts of her seat. Which is why I prefer separating and not zero-summing these ethnic groups.

[Oryxslayer]

Next, as we saw in the 2018 primary that produced Tlaib, putting the the Dearborns and African American areas together if you have the option of separating them should be avoided. One will come out on top and defeat the candidate(s) of choice of the other. Of course the commission didn't have the option to not pair the two, since they were nesting seats in Wayne, so both got paired in the district already with an Arab congresswoman who by now had taken steps for African American outreach. In a similar vein to many Los Angeles districts on various maps, you shouldn't zero-sum discriminate against one group if you have the option not to.

I'm not sure how the current outcome is zero-sum discrimination--Tlaib is now the candidate of choice for both communities and does a good job representing both groups.

Frankly, I don't think trying to maximize the Arab percentage in a district, even at this size, is worth it; Arab communities in Detroit are pretty diverse in both religion and background and highly diffuse. Even in your map, it's only 25-30%; certainly not worth throwing out fair redistricting principles for.

I'm also curious what you're considering as "adjacent groups;" Hamtramck has a lot of Yemenis but also a lot of Bangladeshis who are of course not Arab.

The current map is not. See how I said if you have the option of separating them. Since the commission was nesting seats effectively inside Wayne, there is no option of separation. Therefore, the best option is to give Tlaib as much Arab areas as possible and keep them together.

The criteria I deemed for adjacent groups: would a member of said community hypothetically go before a commission and be asked or argue for pairing with specific other ethnic groups they consider part of their greater social fabric. If a member would consider themselves closer through neighborhood, social, or economic ties to the Arab access seat than European Whites or African Americans, then they went into my percentage total. This is also the ACS data, which is both very much a lagging indicator that gets progressively more inaccurate the further we get from a census year, and it comes with a MOE, so that all needs to be mentioned.

And finally, I guess agree to disagree on the point of separation. Maybe it's just my background in city politics, so I am acutely aware of when the point of distinction stops being D v R and it starts being access for a specific community - not just an ethnic one. Of maybe I have just spent to long analyzing LA's districts after last year. and seeing how zero-sum redistricting is used to silence one ethnic group at the expense of another. But the 2018 election results are a fairly strong point towards what I am getting at.

[Oryxslayer]

I randomly selected a state and got hawaii which also works very well with these numbers. It allows for the not-oahua islands to have a seat (blue) unto their own and a seat for ubran Oahu(green) and one for rural Oahu(purple).

The rural Oahu is interesting as it has seen some dramtic rightward trends in the last decade and could potentially have been a surprise pick up for republicans in 2022 with the right candidate, still it would have been a very longshot.




The new third would have been Hanabusa's seat. Patrick Branco might have it now.

Maybe I'm doing something wrong but: aren't the non-Oahu islands 47K+ underpopulated and the Hawaii law that okay's greater deviations for island-based nesting only applies to the legislature?