Politico: Problems for Harlem and Watts

[muon2]

Politico posted a story this week about a problem for Rangel and Waters unrelated to their ethics charges. The demographic changes in their districts are replacing blacks with Hispanics, and neither live in majority-black districts anymore.

You must be logged in to read this quote.

Back in January I looked at majority-minority districts in NYC and LA. I see the same changes reported by Politico, though both districts could be preserved at the expense of other areas.

In Manhattan, I found a district similar to Rangel's would be about 51% hispanic. To make that district black either requires cutting into Long Island, or going into the Bronx. I did the latter, and had to include Mount Vernon in Westchester as well to get a district that is 54% black (CD 15). It was easy to make two hispanic districts with over 60% (CD 7, 14) with the remaining area in those two boroughs less a small part to connect white districts from upstate to NYC. The yellow district at the bottom of the picture is a majority-asian district.



In LA I found that no more than one majority-black district (CD 35) could be created. The article notes that there are currently three black districts, but since the one has only 52% and no neighboring district has more than 13% black population, it's hard to see how even two districts could be created. The result of that reduction is that the hispanic seats jump from 4 to 7 (CD 28, 31, 32, 33, 34, 37, and 39), plus a an asian seat for Chu (CD 29).





Map modified to match post below. Second edit to Asian district with expanded view to Riverside

[Verily]

Interesting, although I think we knew this already. None of the three black districts in LA were majority black in 2000, though, and nor was Rangel's, so there's no need to draw majority black seats.

Out of curiosity, what's the partisan split on CA-46 on your map?

I'm also curious how you drew the Asian districts; I've tried both but never managed either since Dave rejigged the maps when adding partisan data.

[muon2]

Interesting, although I think we knew this already. None of the three black districts in LA were majority black in 2000, though, and nor was Rangel's, so there's no need to draw majority black seats.

I think the question is whether it is appropriate to deny the creation of a Hispanic seat to have two or three  black-influence seats. I suspect that Hispanics would argue that they are entitled to districts where they can elect representatives of their choice. The result is that a black district is created with the remainder.

You must be logged in to read this quote.

The racial split is White 46%, Black 6%, Asian 20%, Hispanic 25%. I don't have political data for my CA map on Dave's Redistricting App since I'm using block groups.

You must be logged in to read this quote.

I've used block groups in CA which helps create the snaking district from Arcadia to Artesia. I didn't redo them since January.

[Torie]

Poor Hispanics don't vote much in LA, so blacks can easily win a district that is say 30%, black, 10% Asian, 50% Hispanic, and 10% Anglo. And in fact they do - like clockwork. Heck, you can even be someone like Maxine Waters, and win.

[Sbane]

Out of curiosity, what's the partisan split on CA-46 on your map?

My guess would be about a D+1 district. The only really democratic areas are Gardena and a few parts of San Pedro and Long Beach. Lakewood is probably right around the national average. Torrance is lean Republican although Obama won it. And then it has some Republican territory on the Palos verdes peninsula.

[Torie]

Here is a precinct map of the Bush-Kerry race in CA. The data takes a long time to load over the google earth template, but hey, the maps are pretty!  

[jimrtex]

Politico posted a story this week about a problem for Rangel and Watts unrelated to their ethics charges.

Rangel and Waters?

[muon2]

Politico posted a story this week about a problem for Rangel and Watts unrelated to their ethics charges.

Rangel and Waters?

Fixed.

[muon2]

Poor Hispanics don't vote much in LA, so blacks can easily win a district that is say 30%, black, 10% Asian, 50% Hispanic, and 10% Anglo. And in fact they do - like clockwork. Heck, you can even be someone like Maxine Waters, and win.

Good point. I've rearranged the Hispanic districts so that none have less than 63%. Could the Hispanics reasonably elect their candidates?



Here's the details for the 7 LA county Hispanic seats:
CD 28 - W 23%, B 4%, A 7%, H 65%
CD 31 - W 15%, B 4%, A 16%, H 64%
CD 32 - W 18%, B 3%, A 12%, H 66%
CD 33 - W 8%, B 10%, A 9%, H 72%
CD 34 - W 14%, B 5%, A 10%, H 70%
CD 37 - W 11%, B 14%, A 11%, H 63%
CD 39 - W 20%, B 1%, A 6%, H 72%

Edited to improve geographic shapes. Numbers reflect new map.

[Verily]

Well, Judy Chu right now holds a seat that is 18% Asian and 62% Hispanic (15% white, 3% black), and the primary was not particularly close, so... maybe. I think an Asian would probably win CA-33 (Koreatown has outsized political influence). In CA-28, most of the whites are Republicans, so the seat should be Hispanic-held.

[I spent the winter writing songs about getting better]

Of course Chu replaced a Hispanic.

[Torie]

You must be logged in to read this quote.

Er, no (e.g., lots of Jews, and other assorted and sundry Hollyweird types trying to climb the industry ladder into fame and fortune, and in the meantime wait tables).

You must be logged in to read this quote.

A lot of the Asians in this district are not citizens, so no to that too. It is possible of course, but highly unlikely. And some of the Koreans vote GOP, along with some of the Persians in Los Feliz (are Persians Asian for census purposes?). Check out the precincts.

[tpfkaw]

are Persians Asian for census purposes?

They're considered white.

[Verily]

Out of curiosity, what's the partisan split on CA-46 on your map?

My guess would be about a D+1 district. The only really democratic areas are Gardena and a few parts of San Pedro and Long Beach. Lakewood is probably right around the national average. Torrance is lean Republican although Obama won it. And then it has some Republican territory on the Palos verdes peninsula.

Answered my own question: 58% Obama, so D+5 in 2008. It was easy to do a close approximation of the district with partisan data.

Also, Torie's completely right on the demographic info.

[Torie]

Muon2, I tend to doubt that your CD 38 would be legal if the new California redistricting law is extended to cover the drawing of Congressional lines via passing the initiative this November that does that. It is very erose, cuts through three counties, does not hew to municipal boundaries, and is probably not necessary to draw in such a fashion to create some legally mandated minority majority district. Are you taking that into consideration when you draw your lines?

[muon2]

Muon2, I tend to doubt that your CD 38 would be legal if the new California redistricting law is extended to cover the drawing of Congressional lines via passing the initiative this November that does that. It is very erose, cuts through three counties, does not hew to municipal boundaries, and is probably not necessary to draw in such a fashion to create some legally mandated minority majority district. Are you taking that into consideration when you draw your lines?

The districts in the area were drawn in part to maximize the minority-majority districts, followed by county and municipal integrity. All the districts were within 100 persons of the ideal, where I presume exact equality could be achieved by drawing at the block level.

The controlling feature in the area you mention is the Asian-majority district 29. It jumps into OC along the boundary between Brea and La Habra to reach northern Fullerton and the Artesia/Cerritos/La Palma area. I kept those communities intact given the coarseness of the block groups. I then drew the rest of OC to include four districts entirely within the county, one of which is Hispanic-majority (CD 47 from Anaheim to Santa Ana at 69%). La Habra could easily go with a Hispanic district in LAC, and that left Brea as excess population. The only other alternative in OC would be to link Anaheim Hills/Yorba Linda to Corona or San Clemente to Oceanside and Temecula.

One other constraint I worked with was creating four Hispanic-majority districts between Riverside and San Bernardino counties. One of which is the CD 38 you are concerned about. However, three of those four districts are only 51-52% Hispanic and might not have enough population to elect a Hispanic as we discussed earlier in the thread. It might be more reasonable to go for three more solid Hispanic districts and use the aforementioned Corona - Anaheim Hills link.

[Torie]

I don't think there is any legal requirement to create additional majority minority districts with lines that erose; indeed it might be illegal. Either way, not legal, or not legally required, if the new law passes, the districts must be more compact than CD 38, and respect community lines more (some of the metrics about that are fairly detailed iirc), unless federal law dictates otherwise. Make sense?

[muon2]

I don't think there is any legal requirement to create additional majority minority districts with lines that erose; indeed it might be illegal. Either way, not legal, or not legally required, if the new law passes, the districts must be more compact than CD 38, and respect community lines more (some of the metrics about that are fairly detailed iirc), unless federal law dictates otherwise. Make sense?

With your comments in mind, I've revised the map. I linked the Asian district (29) to Chinatown instead of Cerritos to get a majority. That caused a number of shifts in the LA Hispanic districts, but they all stay over 63%.

I then used both my alternative connections to strengthen the Chino/San Bernardino/Riverside districts, and avoided a 3 county district in that area. Hopefully the percentages would be high enough to meet the federal standard for the Hispanics to elect their preferred candidate. I've widened the view to include that area. At the bottom are the demographics for those 3 districts, plus the OC Hispanic district.

Off the map I had to rearrange the other two Hispanic districts to accomdate the changes in OC. The Palm Springs area is now linked to all of Imperial County and the southern San Diego/Chula Vista district stands on its own. They are 60% and 58% Hispanic respectively.



Here's the details for the 7 LA county Hispanic seats:
CD 28 - W 23%, B 4%, A 7%, H 65%
CD 31 - W 15%, B 4%, A 16%, H 64%
CD 32 - W 18%, B 3%, A 12%, H 66%
CD 33 - W 8%, B 10%, A 9%, H 72%
CD 34 - W 14%, B 5%, A 10%, H 70%
CD 37 - W 11%, B 14%, A 11%, H 63%
CD 39 - W 20%, B 1%, A 6%, H 72%


Other Hispanic districts on the map:
CD 38 - W 23%, B 7%, A 7%, H 61%
CD 43 - W 21%, B 13%, A 4%, H 60%
CD 44 - W 27%, B 5%, A 9%, H 57%
CD 47 - W 15%, B 2%, A 12%, H 71%

[muon2]

Torie PM'ed me some additional comments, and I was able to utilize them to create more compact districts along the beach by linking Long Beach to Huntington Beach. The minority districts remained largely the same except for the Riverside district which now is only 55% Hispanic. There were some number changes as well to match up with some of the current distrcit numbers. In particular CD 38 is now the Asian-majority district (instead of CD 29), and CD 42 is now a Hispanic-majority district, replacing 38. CD 26 is somewhat ugly now, but I didn't like putting the Glendora-Upland area in CD 25 since there is no road connection.

[Sbane]

I wonder what the PVI of CD-36 is in your second map. It depends on how much into democratic territory the eastern part of the district goes. What % is anglo in that district?

[muon2]

I wonder what the PVI of CD-36 is in your second map. It depends on how much into democratic territory the eastern part of the district goes. What % is anglo in that district?

I used block groups so the partisan data was not directly available. I put together an estimate from the test data version and I get 59% Obama, 39% McCain for CD 36 in the latest version. The district is about 42% Anglo, 17% Asian, and 29% Hispanic.