California Redistricting!

We’ve already had a lively discussion with regard to California redistricting in this diary below, but now that I have diary rights I want to finally post my own projected map of the new California districts. Needless to mention, redistricting California is a daunting task, particularly in light of the newly approved ‘nonpartisan’ commission. So, a few notes and caveats are in order.

1) My map is only as good as the data provided by Dave’s Redistricting App. Clearly then, to whatever extent that data is invalid, my output will be likewise.

2) In this installment I’ll mainly just outline the process that I followed in placing the districts. I might go over the political results or the VRA ramifications in a subsequent installment if there seems to be enough interest. In short, I welcome any feedback or criticism regarding my decisions. I want to be reasonably confident about the validity of my mapping scheme before I go into much detail about how it affects individual districts.

3) I did not take any account of the current districts when putting together my map. However, I went back and numbered them based on the closest current district merely so as to facilitate discussion. I have added the incumbent that holds that currently numbered district, though I generally have no idea whether they’d still live in said districts. Also, when I doublechecked this afternoon I realized that several of the LA County and OC County districts were not optimally labelled in the map I posted to the other thread. More on that when I get to it.

4) The purpose of this exercise is in part to counter the popular notion that we can’t project some reasonably accurate version of the new California district configuration. My basic premise is that: If the commission simply follows the rule of crossing county and city boundaries only when necessary, then in most cases it’s obvious which way to go, so long as you have a starting point.

5) None of my districts deviate by more than +/- 600 people. The majority deviate by less than +/- 200. Again, this is based on Dave’s app of course. And, speaking of Dave’s app: It’s glitchy at spots. I’ve edited these glitches out of the maps below.

So, without further ado, here we go:

To begin with, here’s my statewide map:

Now I chose San Francisco as a natural starting point. The simple reason for this being that it’s bounded on three sides by water boundaries. It’s also one of the state’s leading cities and a classic tenet of purely geographic districting is to minimize the subdivision of major population centers. Below is my final map of the Bay Area. This is the process by which I arrived at these boundaries:

1) I began with CA-08 in San Francisco County, then simply added voting blocks horizontally until I reach the correct population. Now, there’s been some discussion about the idea of instead dividing the county roughly in half and joining the northern district with Marin County across the Bay. For the record, I tried that just to see what happens. In short, all of these compact districts around the Bay are heavily Democratic districts, and taking CA-08 north into Marin simply rotates them clockwise, still leaving you with a set of heavily Democratic districts. The only meaningful difference is that population centers are harder to keep intact.

CA-08 (Pelosi – D): 85% Obama / 13% McCain

2) I started CA-12 with the remainder of San Francisco County and finished it with San Mateo County except for Redwood City.

CA-12 (Speier – D): 74% Obama / 24% McCain

3) I started CA-14 with Redwood City and added Santa Clara to complete the district. I then put San Jose in its own CA-15 district, and then started CA-16 with the rest of Santa Clara County.

CA-14 (Eshoo – D): 73% Obama / 25% McCain

CA-15 (Honda – D): 70% Obama / 28% McCain

4) I wasn’t sure which way to go with CA-16, so I switched gears to start CA-17 with Santa Cruz County. I then added Monterey County to CA-17. It then became clear enough that I should add San Benito County to CA-16, because if I added it to CA-17 it would not take up the whole county.

5) At this point it became clear enough that I should complete CA-16 with the eastern part of Alameda County.

CA-16 (Lofgren – D): 64% Obama / 34% McCain

6) All of central Alameda County can then become CA-13 (Fremont, Union City, Pleasonton, Hayword) and then Oakland can take up most of CA-09.

CA-13 (Stark – D): 72% Obama / 26% McCain

7) I then finished up CA-09 with the western tip of Contra Costa County, then added CA-10 to take up the central county through Concord and Danville.

CA-09 (Lee – D): 89% Obama / 9% McCain

CA-10 (Garamendi – D): 69% Obama / 30% McCain

8) I wasn’t sure where the rest of Contra Costa belonged, so I started CA-07 with Solano County. I wasn’t sure where to go for the rest of CA-07 so I switched to CA-06 in Marin County, then finished it up with most of Sonoma County.

CA-06 (Woolsey – D): 76% Obama / 22% McCain

OK, so now what? More after the map…

Below I have my maps of Northern California and the Sacramento area.

9) As I pondered CA-07, I realized that I could add all of Napa County and all of Yolo County except for West Sacramento. The alternatives would either create a weirdly shaped district or unnecessarily divide Sacramento.

CA-07 (Miller – D): 65% Obama / 33% McCain

10) I then started CA-01 with Mendocino, Trinity, Humboldt, Del Norte, and Lake counties. It also became clear that the rest of Sonoma belonged in CA-01. The CA-01 district still needed 365,000 people.

11) I went ahead and added CA-05 in West Sacramento (Yolo County) and Sacramento proper. Sacramento should clearly anchor 2 districts, so I placed CA-03 fully in the eastern part of the county.

CA-05 (Matsui – D): 69% Obama / 29% McCain

CA-03 (Lungren – R): 50% Obama / 48% McCain

12) I still had 55,000 people left in the southwestern salient of Sacramento County. It now became clear enough that CA-11 should take those, the remaining half of Contra Costa County, and San Joaquin including half of Stockton.

CA-11 (McNerney – D): 59% Obama / 39% McCain

13) I now added Colusa, Glenn, Tehama, Shasta, and Siskiyou counties to CA-01. I still needed 23,000 more people. The most efficient way to finish up CA-01 was with Modoc County and part of Lassen County.

CA-01 (Thompson – D): 50% Obama / 47% McCain

14) I could now start CA-02 with the rest of Lassen. I then added Plumas, Butte, Sierra, Yuba, Nevada, and Sutter. This left me needing 170,000 people. The obvious place to get them was Placer County except for the Rocklin/Roseville corner.

CA-02 (Herger – R): 47% Obama / 51% McCain

15) It now made sense to start CA-04 with this corner of Placer, and to add the sparsely populated eastern counties of El Dorado, Amador, Calaveras, Alpine, Mono, Tuolomne, and Mariposa. But, where to go now? More after the maps…

Below I’ve added my maps of the Central Valley and Southern California. The next set of decisions involving the Central Valley are the ones that I think are most debatable, as I’ve stated previously. In any case, here’s my reasoning

16) It made sense to start CA-18 with the rest of San Joaquin County and then finish it off in Stanislaus County, including the city of Modesto.

CA-18 (Cardoza – D): 54% Obama / 44% McCain

17) I then started CA-19 with the remainder of Stanislaus County and all of Merced County. At this point, the next population center was Fresno, but I had to decide what to do with Madera County that was in the way. I played with several alternatives and realized that my options were to either split Madera County or split the city of Fresno or end up with several weirdly shaped districts. I chose to split Madera County and then finish off CD-19 with western half of Fresno County.

CA-19 (Denham – R): 49% Obama / 49% McCain

18) It then made sense to give Fresno it’s own CA-21 district, and to place the sparsely populated eastern remainder of Madera & Fresno counties in CA-04.

CA-21 (Nunes – R): 51% Obama / 47% McCain

19) At this juncture, it seemed fairly obvious to start CA-20 with Kings and Tulare counties. Once I did that, it was obvious that CA-04 should finally be completed with Inyo and the sparsely populated east of San Bernardino County.

CA-04 (McClintock – R): 42% Obama / 56% McCain

20) I now decided it was time to switch back to the coast. It was clear that I should finish CA-17 with the Cambria corner of San Luis Obispo County. I could then start CA-23 with the rest of San Luis Obispo and add all of Santa Barbera County, which left me needing 67,000 people.

CA-17 (Farr – D): 71% Obama / 27% McCain

21) I then switched back to CA-20, finishing it in Kern County. Then I added CA-22 fully contained in Kern County, which left 12,000 people in one corner. I decided to add these to CA-04, swapping them out for 12,000 in San Bernardino (which didn’t change the partisan breakdown of CA-04).

CA-20 (Costa – D): 43% Obama / 56% McCain

CA-22 (McCarthy – R): 39% Obama / 59% McCain

22) I then added CA-41 in central San Bernardino County, finished CA-23 in northern Ventura County, and placed CA-24 in southern Ventura County. This left me with 34,000 people in Ventura County and I was ready to start on LA – after the maps!

CA-41 (Lewis – R): 44% Obama / 53% McCain

CA-23 (Capps – D): 58% Obama / 40% McCain

CA-24 (Gallegly – R): 55% Obama / 44% McCain

Below is my LA County map. Note that the district numbering has changed from the map that I posted in the other thread, because when I went back over it I realized that the current Dreier district (CA-26) had been dismantled and that the one which I had labeled as CA-26 should’ve been Chu’s CA-32, while the one that I originally labelled as CA-32 should’ve clearly been CA-39 (Linda Sanchez).

23) Anyhow, here is how I proceeded with LA County (with the above amendments):

CA-25: I started with northern LA County, and added the San Fernando Valley.

CA-30: I finished Ventura County, and added the Westside cities.

CA-27: I took the rest of San Fernando, Burbank, and Glendale.

CA-33: I started with Culver City, added Santa Monica, and Beverly Hills.

CA-28: Centered on Hollywood.

CA-29: Centered on Pasadena.

CA-32: I started in Glendora and took in the northern suburbs.

CA-35: Centered on Inglewood.

CA-36: Centered on Rancho Palos Verdes – with Redondo Beach and Manhattan Beach.

CA-38: Centered on Pomona and Covina.

CA-31: Centered on downtown LA.

CA-34: Centered on Huntington Park.

CA-37: Centered on Compton.

CA-39: East LA, leaving the southeast waterfront corner of Los Angeles County.

CA-25 (McKeon – R): 53% Obama / 45% McCain

CA-27 (Sherman – D): 71% Obama / 27% McCain

CA-28 (Berman – D): 80% Obama / 18% McCain

CA-29 (Schiff – D): 69% Obama / 29% McCain

CA-30 (Waxman – D): 63% Obama / 35% McCain

CA-31 (Becerra – D): 80% Obama / 17% McCain

CA-32 (Chu – D): 58% Obama / 40% McCain I think Dreier actually lives here.

CA-33 (Bass – D): 77% Obama / 22% McCain

CA-34 (Roybal-Allard – D): 76% Obama / 22% McCain

CA-35 (Waters – D): 88% Obama / 11% McCain

CA-36 (Harman – D): 59% Obama / 39% McCain

CA-37 (Richardson – D): 84% Obama / 15% McCain

CA-38 (Napolitano – D): 64% Obama / 34% McCain

CA-39 (Linda Sanchez – D): 60% Obama / 38% McCain

With LA out of the way, it’s time to wrap up SoCal after the map.

Below I have added my final maps. The first covers Orange County & the Inland Empire; the second covers the San Diego area. Here is how I proceeded to map these districts.

24) I started with Orange County by taking the last bit of LA and joining it with Huntington Beach to make CA-46. I then put CA-47 in Fullerton, Anaheim, and Buena Park and Irvine/Newport Beach in CA-48.

CA-46 (Rohrabacher – R): 48% Obama / 50% McCain

CA-47 (Loretta Sanchez – D): 52% Obama / 46% McCain

CA-48 (Campbell – R): 55% Obama / 43% McCain

25) I then put the city of San Bernardino in CA-43, and finished off San Bernardino County with CA-42, which still needed about 90,000 people. However, I wasn’t sure whether these should come from Orange County or Riverside County.

CA-43 (Baca – D): Obama 61% / McCain 37%

26) I now switched to Riverside County by placing CA-45 in the eastern 2/3 anchored with  Palm Springs, and then centered CA-44 on the city of Riverside and Moreno Valley. I think CA-44 might actually be the vacant seat, so maybe I should’ve labeled it CA-26..

CA-45 (Bono Mack – R): 51% Obama / 48% McCain

CA-44 (Calvert – R, or maybe vacant): 58% Obama / 40% McCain

27) I then started CA-51 with Imperial County and added eastern San Diego County basically up to the coastal strip. That still left me needing 440,000 people, and neatest way to add them was to take the South Bay area (Chula Vista & Imperial Beach).

CA-51 (Filner – D): 51% Obama / 47% McCain

28) The city of San Diego can clearly anchor two compact districts, so I just split it down the middle with CA-52 and CA-53. I then added CA-50 along the coast, and started CA-49 in Oceanside.

CA-50 (Bilbray – R): 50% Obama / 48% McCain

CA-52 (Hunter – R): 55% Obama / 44% McCain

CA-53 (Davis – D): 66% Obama / 32% McCain

29) At this point it’s clear that if CA-49 goes into Riverside County, either Riverside or Orange will be subdivided once more than necessary. So, I finish off Orange County with CA-49 and CA-40, leaving 15,000 people in the southeast corner.

CA-40 (Royce – R): 43% Obama / 55% McCain

CA-49 (Issa – R): 46% Obama / 52% McCain

30) This leaves only Riverside County, where I wrap up CA-42 with Norco, and create what is essentially a new Inland Empire seat from Temecula to Corona (along with those 15,000 people from the corner of O.C.) I’ve numbered it CA-26, but it doesn’t actually overlap Dreier’s current district, and I think Calvert lives here in Corona, which would make the CA-44 district the vacant one.

CA-42 (Miller – R): 53% Obama / 45% McCain

CA-26 (Dreier – R; but actually either vacant or maybe Calvert – R): 44% Obama / 55% McCain

Whatever the case, I think that more than covers it! Please let me know what you think of my maps. Am I on track or way off base??

111 thoughts on “California Redistricting!”

  1. Really liking the way compact districts fuck over so many Republicans here. How I hope Reps. Dreier and Calvert end up in a primary battle; I would savor the schadenfreude of such a sure-to-be-filthy deathmatch between two hypocritical scumbags. Mmm.

    And I’ll trade Rep. Costa for a crack at knocking off Reps. Calvert, Dreier, Gallegly, Hunter, and Campbell and Rep.-elect Denham (whom I already loathe) any day of the week and twice on Sundays, so I like it.

  2. Miller lives in Martinez and would run in the 10th.

    Garamendi lives in Walnut Grove (Sac County) and would run in the 11th.

    Mike Thompson lives in St. Helena and would run in the 7th. This would leave the 1st an open seat.  With its near even partisan split it would certainly be a marquee race of 2012.

    McNerney is the odd man out. His old district is split so many ways he doesn’t have a clear place to run.  He lives in Pleasanton, which appears to be split between the 16th and 13th.

  3. beat a dead horse or kick a sleeping dog but here goes.

    1st Thanks for posting the whole map.  I appreciate your work on it and it took time.

    2nd as to VRA issues that can be discussed by others but suffice it to say that hispanic community gets sliced and diced up in  this map. Same old culprits-San Diego county  San Jose area-Central Valley and East LA county.  To an extent it appears that West LA has as well.  

    3rd your map is all very logical but at every step of the way the logic has evolved around dividing up GOP areas while spreading democratic areas around to various districts. For instance what is the community of interest that connects Imperial county with downtown San Diego? In addition since 1960 CA1 plunged southward and now after neatly dividing heavily democratic Sonoma county it moves eastward.  Every step of the way-like in Silicon Valley and in the Oakland Basin you make very logical decisions but they all the democrats.  Instead of keeping the communities in the hills (more republican oriented) in a community of interest district you connect the folks in the valley with the hills.  Why?  The same pattern in LA county is followed.  You neatly connect democratic areas in the vallies with the republican voters in the hills.  

    This has the effect of dividing communities of interest -reducing the minority vote and electing more democrats.  

    I have no problem starting with SF but dividing the city in half is the classic gerrymander and is the type of political decision this process was meant to stop.  You want to keep counties together when they benefit one party but divide them when they help you.  

    who can be sure what we will see but I think it will not be like this.

  4. But again, there is probably some VRA regression here, particularly in L.A. and the Central Valley, that would prompt a credible lawsuit. I wish it wasn’t so, because it’s nice to see how Dem-friendly one of these non-gerrymandered maps can be, but the commission absolutely will not pass a map that could even possibly run afoul of the VRA.  

  5. Thanks for taking the time to do all of this. I know it can’t be easy and takes a lot of time. From looking at this map and other maps around it is pretty obvious that a lot of Republicans are going to be in trouble. Only a select few will likely not have to sweat. Some Dems will likely have their margins trimmed a bit but whatever Steve Israel does he needs to make darn well sure that he is ahead of the recruiting game in CA. The GOP House isn’t likely going to just sit back and play nice and many of the CA Republicans aren’t going to buck the party so it is time to roll up the sleeves and tie them to whatever non-sense comes out of the House. From this map and other maps it clearly looks like the following are going to be in trouble to some extent:

    1) Lungren

    2) Denham

    3) Gallegy

    4) McKeon

    5) Calvert

    6) Billbray

    7) Bono-Mack

    It would be nice to rid ourselves of Darrell Issa and Duncan Hunter but without partisan control that would likely not happen. Plus Issa is ridiculously loaded so that would likely scare off a top tier challenger.

    Still a huge effort must be made to support Congressional candidates in CA as the state isn’t likely going to be contested by Obama (if it is we have major problems) and the top tier race may be an open Senate seat if DiFi retires which I think that she will.  

  6. What I don’t get is why did Republicans so heavily push this Redistricting by NP Commission when they are likely to lose a bunch of seats still?  Is it because its better than Democrats having total control (Obviously).  But why have Republicans been so excited about it?  Are we maybe misreading what the commission will do?

  7. Costa is from Fresno and Nunes is from Tulare. Costa’s district is meant to connect Hispanics in the Central Valley, so it may end up running from Fresno to Bakersfield once again.

    Loretta Sanchez may or may not end up retaining a district that looks like her current one. It’s currently over 60% Hispanic and the VRA would probably require it to stay close to that.

  8. Let me say though, the downside for Dems is probably worse than we think.

    PVI is calculated by how far presidental votes differ from the mean.  For example, nationwide, Obama won by roughly 6%, so a district which Obama wins by 5% actually has a PVI of R+1.  Of course, Kerry votes also have to be taken into account, but since Dave’s Redistricting App doesn’t allow this, we’ll have to make due without it.  

    Nationwide, Republicans essentially cannot hold a seat with a PVI of greater than D+4.  Therefore, their certain losses are:

    CA-44 (Calvert) Obama PVI D+12

    CA-48 (Campbell) D+6

    CA-24 (Gallegly) D+5

    CA-42 (Hunter) D+5

    In California, however, the Republicans haven’t been able to hold onto any seats with a D PVI of any sort.  Indeed, McNerney has held on in an R+1 district.  Thus they’d also probably lose…

    CA-25 (McKeon) D+2

    CA-42 (Miller) D+2

    As districts change, and if the Dems stage a national comeback, Nunes and Bono Mack could lose their seats.  

    On the Democratic side, they would have one certain loss

    CA-20 (Costa) R+19

    And three probable losses

    CA-1 (Thompson) R+3

    CA-51 (Filner) R+2

    CA-47 (Loretta Sanchez) TIE

    In a worst case scenario, one more Dem would be threatened

    CA-18 (Cardoza) D+4

    In the Democrats best case scenario, they would lose one seat, but pick up six, giving them a net of five.  In the worst-case scenario, they would pick up four seats, but lose five, giving them a net loss of one seat.  Most likely, they would gain three seats I think, grabbing all vulnerable Republicans, and losing Costa plus two of the three probable losses.  

    Although the racial demographics are old for Dave’s redistricting app, you might want to try a map where you take the VRA into account first, then completeness.  EG, aim to make ten Latino-majority districts which are reasonably compact and do not cross county lines excessively.  Then create minority-majority seats, then the rest.  

  9. CA-02 (Herger – R): 47% Obama / 51% McCain

    “15) It now made sense to start CA-04 with this corner of Placer, and to add the sparsely populated eastern counties of El Dorado, Amador, Calaveras, Alpine, Mono, Tuolomne, and Mariposa. But, where to go now? More after the maps… ”

    There is a reason that the far northeastern counties are not presently in CD02: it is called the Sierras.  Those counties are in the Great Basin, with a major mountain chain between them and the balance of CD02.  They are best connected with the other counties which have a presence on Lake Tahoe.

    Even the post office services those counties from Reno, NV rather than from Redding CA, in Shasta county.  

  10. and let me refine my VRA/minority majority view.

    In CA you will see-like CA47-seats at 75% minority threshold to elect an hispanic.  With the low voting numbers 50% lower then whitesyou need that higher number.  That’s why CA47 stacks and packs them now.

    My other point is GOP is at 19 plus these seats now are at a 1 million population surplus.  That’s 20 1/2 seats plus four they nearly won in 2010.  The only way to get them at 18 or lower is to massively gerrymander them.  As noted every decision you make gives the democrats an edge.  Instead of losing 1/2 seat in Bay area the democratic seats expand to take in rural area.  

    In the case of CD4 you have no problem expanding Miller’s seat into rural area north of it.  Yet its not logical to attach the Northern part of CD4 to Central valley counties that are nearby?  Somehow its easy for you to attach democratic areas to republican areas but republican areas do not?  

    I don’t mean this as an attack but rather to show you that I believe the commission will follow a very different pattern.  

  11. Good job on this.  Having worked on redistricting in the past, one humble suggestion.  Instead of drawing a state (particularly a larger state such as CA) district by district, you might consider creating 3, 4 or 5 “footprints” within the state and then dividing up those footprints.  

    So in other words, create a 10-district footprint for the Bay Area, a 7-district region for Northern California and Central Valley, another LA region, an Inland Empire and a San Diego/Orange county region.  That way you can work out the boundaries between the regions on the front end and one region isn’t disadvantaged by any of the others.  

Comments are closed.