California with 37 VRA seats

Here we go again! Another California map…

I dunno how much demand is actually left for California redistricting maps, but here’s my latest version anyhow. Since this’ll be the last map I post (of California or any other state) until actual Census data is available, I’ve decided to go all out.

When I posted my original California map that basically disregarded the VRA, and again when I posted my VRA maps of Southern states, I was strongly encouraged to create a new California map that takes the VRA as its starting point. While initially somewhat resistant, I was persuaded that I should do exactly that mainly when I looked over the new California redistricting guidelines and saw that they prioritize the VRA above all other criteria.

Moreover, whereas my initial map used the 2008 population estimates, I wanted to make a map using the more accurate American Community Survey projections. So, I’ve created the following maps and analysis by this methodology:

1) I first addressed Section 5 preclearance concerns to ensure that none of the four covered counties (Kings, Merced, Monterrey, Yuba) would be subject to a “retrogression” challenge.

2) I then addressed Section 2 requirements that majority-minority districts be drawn wherever compact minority communities permit the drawing of such districts without substantially ignoring traditional redistricting criteria.

3) I then filled in the rest of the map based on the non-partisan criteria specified by the California guidelines (contiguity, geographic integrity, communities of interest, compactness). As such, I altogether disregarded partisan data and incumbent residency as required by law (especially easy to do on the ACS version at Dave’s app which doesn’t feature partisan data).

4) However, I wanted to know what the partisan effects would be, so I then had to translate my ACS-VRA map into partisan figures – which was easily the most time-consuming part of this exercise.

5) Finally, I was curious to see how the maps I drew would match up with the current incumbents, so I pinpointed where each one lives and identified in which district on my map they would end up.

The resulting maps below feature 37 majority-minority seats of which 15 are majority-Latino and 1 is majority-Asian.

So, I’m finally ready to present all that after the jump. Needless to mention, analyzing California is a daunting task, so I’ve decided to organize my presentation as follows:

For the purposes of my discussion, I’ve divided the state into the following regions: (1) Northern California (including Sacramento); (2) the Bay Area; (3) the Central Valley & Central Coast; (4) Los Angeles; (5) Orange County & the Inland Empire; (6) the San Diego Area.

For each region, I’ve posted some general comments, then listed the 2008 Obama/McCain figures for each district including the incumbent of that currently numbered district, and then posted the relevant maps. I then have three sections of commentary: VRA Implications, Partisan Impact, and Incumbents.

A quick note on the latter: It goes without saying that where the incumbents end up in the new maps is highly speculative. The reason why I decided to look at this anyhow is twofold: (1) at the very least, it gives some sense of how the new maps are likely to scramble the incumbents, even if the actual final arrangement is very different; (2) the places where incumbents are likely to end up together in the same district or where districts are likely to initially end up without a resident incumbent obviously correlate with the most highly gerrymandered parts of the state, and so therefore it’s safe to say that the incumbents that are most affected on my map will be the ones at highest risk under the actual maps. Of course, residency within a district is not a requirement to run for Congress, but Congress critters do generally prefer to run where they live – or to move if they have to.

One last note before I get underway: I’ve played with the California maps enough at this point that I have a rather good sense of what decisions are likely to help Democrats or Republicans. All else being equal, I have generally made those decisions which are most helpful to the GOP. On one hand, I’ve done this to minimize unfounded accusations of partisan mapping bias on my part. On the other hand, I’ve done this because I think similar decisions will maximize the likelihood of a map being approved by the Redistricting Commission which requires three GOP votes to pass the maps. On the third hand (?!), I’ve done this because I would rather see a ‘worst-case scenario’ for Democrats, with the awareness that the actual maps are likely to be somewhat more favorable, than to promote a rose-colored glasses view inconsistent with the probable outcome. In short, if you are a Republican and you don’t like my maps, you probably need to reflect on how consistent your perception of California is with reality.

So, without further ado, here goes!

Northern California & Sacramento

Below I’ve started with two maps. The first is a broader view of NorCal while the second is a closer view of Sacramento so that it’s clear where I’ve drawn the lines and how the Sacramento districts intersect with the surrounding districts.

In my view, the NorCal mapping scheme is very likely to resemble this arrangement. (I guess that’s stating the obvious..) On one hand, there’s a certain symmetry: two districts anchored by Sacramento, two compact districts extending toward the northwest and the northeast respectively, and one district covering the sparsely populated northern end of the state.

I’ve seen several maps that further subdivide Sacramento County or that create more elongated districts stretching as far as the Oregon border. I personally don’t think the such maps are consistent with the current rules that call for minimizing county/city subdivisions and that generally prohibit bypassing centers of population to reach more distant populations.

In any case, here are the partisan figures for these five districts. I’ll post further comments after the maps.

CA-01 (D-Thompson): 61% Obama – 37% McCain

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

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

CA-04 (R-McClintock): 44% Obama – 54% McCain

CA-05 (D-Matsui): 68% Obama – 30% McCain (61% Minority – 18% Asian; 25% Latino)

VRA Implications: There’s not much in the way of VRA implications with concern to mapping Northern California. Yuba County is a Section 5 preclearance county but it’s currently in a 76% white district. On my map, it goes from its present location in CA-02 represented by Herger to the CA-04 district represented by McClintock. The new CA-04 district has virtually identical racial/partisan stats as the current CA-02, so there’s no “retrogression” concern. The only other district of VRA interest is CA-05 and my map takes it from 56% minority to 61% minority. There’s no prospect of creating any other majority-minority districts, nor is there a prospect of creating a district with a majority from a single minority group.

Partisan Impact: Herger’s CA-02 district goes from 55% McCain to 51% McCain and Lungren’s CA-03 goes from an even 49%-49% split to 50% Obama/48% McCain. In the case of CA-02, the district should be safe enough for Herger, but he’d probably have to actually put a bit of effort into his reelection campaigns and there might be an opening for the right Democrat in a good year for Dems, particularly if the seat comes open.

Lungren’s CA-03 might be far more problematic from his standpoint. On paper, it’s a ‘tossup’ at worse, but Lungren’s been facing stiff challenges in recent elections, and he doesn’t strike me as a particularly good fit for this district that’s now fully contained in Sacramento County. While the district should be quite suitable for a Sacramento Republican, I’m not too sure that Lungren is that Republican, and the area is gradually trending leftward (which will be a problem for the GOP no matter how the ‘second’ Sacramento district gets drawn).

The other three districts are not meaningfully affected, though Thompson’s CA-01 does drop from 66% Obama to 61% Obama.

Incumbents: My map doesn’t displace any of the current incumbents in these five districts and it’s unlikely that any mapping scheme would. That includes Tom McClintock: A lot of sources still list his residence as Elk Grove in Sacramento County, but he’s officially moved to Roseville in Placer County.

The Bay Area

My latest map of the Bay Area is below.

The most consequential effect of using the ACS figures has been that CA-11 is clearly pushed out of the Bay Area and into the Central Valley (which is where I’ll deal with it below). Otherwise, my main concerns here were maximizing the number of majority-minority seats and carving out a majority-Asian seat in particular.

In my experimentation with mapping the Bay Area, it seems fairly clear to me that there are only two districts with significant ambiguity as to how the lines will end up: On my map below, they’re CA-10 and CA-16. What does seem clear enough, however, is that most of the districts will be arranged along the Bay coast, with a ‘hinterland’ district centered in Contra Costa/Alameda, and then leaving what I think of as a ‘mop-up’ district piecing together the ‘leftovers’ from the districts with more obvious placement (e.g., CA-10 on this map).

It’s also worth noting that the contours of this mapping scheme were partially determined by what has to happen in the Central Valley to maintain a VRA-compliant CA-18 district. To do that, you have to split off the heavily Latino parts of Stanislaus and San Joaquin counties, which means you have to situate a district on the remainder which still leaves a part left over on one end or the other. The most coherent mapping scheme I could identify involved attaching that leftover part to CA-10 rather than, say, CA-19. Either way, it doesn’t make much difference: CA-10 will still be heavily Dem and CA-19 will still be heavily GOP.

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

CA-07 (D-Miller): 73% Obama – 25% McCain (52% Minority – 15% Asian; 24% Latino)

CA-08 (D-Pelosi): 85% Obama – 13% McCain (52% Minority – 29% Asian)

CA-09 (D-Lee): 88% Obama – 10% McCain (64% Minority – 20% AA; 20% Asian; 21% Latino)

CA-10 (D-Garamendi): 61% Obama – 38% McCain (55% Minority – 24% Latino)

CA-12 (D-Speier): 74% Obama – 25% McCain(57% Minority – 32% Asian; 19% Latino)

CA-13 (D-Stark): 70% Obama – 28% McCain (74% Minority – 51% Asian)

CA-14 (D-Eshoo): 71% Obama – 27% McCain

CA-15 (D-Honda): 71% Obama – 27% McCain (71% Minority – 26% Asian; 39% Latino)

CA-16 (D-Lofgren): 65% Obama – 34% McCain (52% Minority – 17% Asian; 25% Latino)

VRA Implications: My map takes the Bay Area from seven majority-minority districts out of 11 to eight majority-minority districts out of 10. The CA-13 district is also majority-Asian. In some cases, I’ve swapped proportions (i.e., CA-08 goes from 58% to 52% minority while CA-12 goes from 52% to 57% minority; CA-16 goes from 69% to 52% minority while CA-15 goes from 53% to 71% minority). This is augmented by the fact that CA-11 goes from the current 63% white district to a 52% minority district in the Central Valley and CA-17 on the Central Coast becomes a majority-minority district.

For the record, you evidently cannot create a majority-Latino district in San Jose even if you ‘give up’ the majority-Asian district. The Latino population is too intermixed with the Asian population in much of the area.

Partisan Impact: The Bay Area seats will all be heavily Democratic seats no matter how they’re mapped out.

It’s been suggested elsewhere that ‘a foothills seat in Contra Costa / Alameda’ could be a tossup seat. This is obviously not true – mainly because you don’t have anywhere near enough people there as should be self-evident in these maps. The most Republican seat that I was able to create following that guideline was a 59% Obama seat. Once you take in all the most Republican areas in the eastern part of the Bay counties, you must necessarily go into heavily Democratic areas in order to fill out the necessary population.

There’s also been a suggestion that: “The outer areas of current CD14-CD15-CD16 could end up being a tossup seat.” That makes no sense whatsoever. The “outer areas” of those districts are basically Santa Cruz County, which was a 77% Obama county. To make the long story short, anyone who sees a “tossup” district in the Bay Area (much less a Republican district) is hallucinating.

Incumbents: My mapping scheme would inconvenience several Bay Area incumbents, with at least one of them being out of a job (unless he or she moves to the Central Valley). McNerney lives in Pleasanton (central Alameda County) which ends up in CA-16 on this map. Since this district is most similar to his current district, I think he’d be very likely to run there. Mike Honda lives in the Campbell area of San Jose, which ends up in CA-14 on this map. This is essentially Anna Eshoo’s district who lives in Atherton which is still in CA-14. Meanwhile, Zoe Lofgren lives in San Jose and would end up in CA-15.

As I see it, the most obvious place for Mike Honda to run would actually be CA-13 since it becomes majority-Asian and takes in substantial parts of his current district, but if Pete Stark doesn’t retire that’d mean a primary battle – aside from the residency issue. Ultimately, I don’t know how this would be sorted out. Perhaps some Californians could comment on it.

Whatever the case, tidying up the lines in the San Jose area makes it highly probable that at least two Democrats will end up in the same district – even leaving aside the fact that the Bay Area will lose a seat.

It’s also worth noting that Garamendi lives in Walnut Grove (Sacramento County) which just barely ends up in this CA-10 district. That said, I think Garamendi will end up running in whatever district most resembles his current district regardless, whether or not he has to move.

Central Coast & Central Valley

The Central Coast and the Central Valley are really two separate regions, but I’ve decided to cover them together for the sake of expedience. By and large, the mapping scheme that I settled on was dictated by Section 5 preclearance concerns, since three of these counties are covered: Kings, Merced, and Monterey. In short, I mapped the relevant districts first (CA-20, CA-18, and CA-17) to satisfy ‘retrogression’ concerns, and then mapped the rest accordingly.

A couple of general points are worth noting. San Benito County only has about 55,000 people and doesn’t really make a significant difference regardless how it’s mapped. I basically chose to add just enough of it to CA-17 so that the boundaries with CA-12 to the north and CA-23 to the south could be the coastal county lines. The remainder then obviously belonged with CA-20 (but again, it makes no meaningful difference to either the partisan or ethnic breakdown).

As for the Great Basin, it’s even more sparsely populated and even less consequential than San Benito. I’m sure there’s some Republican out there who imagines that if the Great Basin were instead divided and appended to the more central districts it’d somehow benefit the GOP. Any such Republicans are misguided. About 30,000 people live in the Great Basin and chopping it up does little besides create a less coherent map. But, hey, if it makes the Commission feel better for some reason, I say go for it.

Besides reiterating that CA-11 isn’t really ‘McNerney’s district’ on this map – having been displaced from the Bay Area to the Central Valley due to population shifts – I’ll save the rest of my commentary for below.

Here’s the Obama/McCain breakdown for these districts:

Central Coast

CA-17 (D-Farr): 73% Obama – 25% McCain (54% Minority – 44% Latino)

CA-23 (D-Capps): 57% Obama – 41% McCain

CA-24 (R-Gallegly): 56% Obama – 43% McCain (52% Minority – 41% Latino)

Central Valley

CA-11 (D-McNerney): 51% Obama – 47% McCain (52% Minority – 33% Latino)

CA-18 (D-Cardoza): 57% Obama – 41% McCain (69% Minority – 54% Latino)

CA-19 (R-Denham): 41% Obama – 56% McCain

CA-20 (D-Costa): 50% Obama – 48% McCain (75% Minority – 66% Latino)

CA-21 (R-Nunes): 51% Obama – 47% McCain (62% Minority – 42% Latino)

CA-22 (R-McCarthy): 39% Obama – 59% McCain (54% Minority – 42% Latino)

VRA Implications: The VRA implications are quite significant in that three of the four California counties that require Section 5 preclearance are in this central region. Here’s how I’ve handled each in turn:

Merced County goes from a 42% Latino plurality in the current CA-18 district to a 54% Latino majority on my map. Clearly this more than satisfies Section 5 requirements.

Monterey County goes from the current 50% white/39% Latino CA-17 district to a 46% white/44% Latino district on my map. This is about as Latino as this district will get absent extreme gerrymandering, and it certainly cannot be challenged on ‘retrogression’ grounds.

I’m far more ambivalent about the CA-20 district that I’ve drawn which includes Kings County. On the one hand, it goes from 63% Latino to 66% Latino according to the ACS figures. On the other hand, it drops from a 60% Obama district to a 50% Obama district. This is mainly because it trades heavily Democratic parts of Fresno and Bakersfield with settled Latino communities for rural areas where the Latino population is more comprised of often-transient agricultural workers. In short, the CA-20 district that I’ve drawn may not be deemed as providing Latinos with a true opportunity to elect a “candidate of choice” despite the high Latino percentage. Even more importantly, it may be the case that Latino registration figures are so out of sync with the Census count that this district would still be considered ‘retrogressive’ under Section 5.

To make the long story short, if this district fails to satisfy Section 5 requirements, then obviously the solution is to restore the current dip down to Bakersfield, and the district will obviously become much more Democratic. On the flip-side, I decided to roll with the district I mapped out because that may well be the kind of district that will satisfy the GOP enough to garner three Republican votes for passage by the Commission.

Aside from these Section 5 issues, my map features seven majority-minority districts where there are currently four, and two of these are majority-Latino versus just one on the current map.

Partisan Impact: For the Central Coast the partisan impact is virtually assured: CA-17 remains heavily Democratic, CA-23 becomes somewhat more competitve but still likely Democratic, and CA-24 gets a clear Democratic lean. I’ve seen it mentioned elsewhere that – in maps similar to mine – CA-30 somehow crosses over to take in the more Republican parts of Ventura County as a “political” move to favor Democrats, but this is total nonsense. The parts of Ventura County closer to the LA County border are more Democratic, not less Democratic, and if you instead map CA-24 starting at the LA County line northward you end up with a CA-24 district that’s between 58% to 60% Obama.

For the Central Valley, the partisan effect is even more consequential. My map creates a ‘toss-up’ CA-11 district in San Joaquin & Stanislaus, with perhaps a slight GOP lean, in place of a strongly Democratic seat in the Bay Area. Otherwise, CA-20 and CA-21 also become ‘toss-ups’ with a slight GOP lean. The consequences for 2012 are probably more contingent on where the incumbents decide to run, so I’ll comment further below.

Incumbents: None of the Central Coast incumbents are displaced, though Gallegly might as well be.

As for the Central Valley, we can safely say that McCarthy will end up in whatever district covers the bulk of Kern County. Since it strikes me as implausible that one of the Bay Area Democrats or one of the dispossesed SoCal Republicans would move to the Central Valley, that leaves four incumbents for five districts. The question then is where will they run.

In my mapping scheme above, Cardoza would obviously run in CA-18. Denham would also be in CA-18 (both Cardoza and Denham live in Atwater in Merced County), but that’s the case already and it obviously didn’t stop him from running in the current CA-19 district. That said, on my map Nunes (Visalia) is actually in CA-19 and he would represent more of the new district’s population than would Denham. So, I imagine Nunes would likely end up running there. As for Costa, he lives in Fresno and I think he’d almost surely run in CA-21 unless the new CA-20 district ends up being more like the current CA-20 district.

So, that leaves Denham to run in one of the slight-GOP ‘toss-ups’ (CA-11 or CA-20) leaving the other one vacant.

Los Angeles

My latest map of Los Angeles is below. There isn’t really that much to say about Los Angeles in general, other than to note that Dreier’s CA-26 district is effectively gone to the Inland Empire (and I’ll discuss it there). Besides that, my goal was, yet again, to create as many majority-minority seats as possible, and to make as many of them as reasonable majority-Latino.

It’s also worth explaining why I chose to have three districts cross over from neighboring counties.

Bringing the CA-42 district across from San Bernardino is the best way to maximize the Latino population of that district while also maintaining the Latino proportion of CA-43. If the goal is to maximize majority-Latino districts, then this is obviously the way to go.

As for CA-40 and CA-46, bringing them into LA County to take in the more white parts of Long Beach (CA-46) and the Whittier/Diamond Bar area (CA-40) is the most effective way to maximize the minority percentage of the neighboring LA County seats, while also maximizing the minority percentages of CA-47 and CA-48 in Orange County. So, from a VRA standpoint, this is again the obvious way to go. From a partisan standpoint it makes no difference of consequence.

Anyhow, here are the districts:

CA-25 (R-McKeon): 50% Obama – 48% McCain (53% Minority – 34% Latino)

CA-27 (D-Sherman): 69% Obama – 29% McCain (76% Minority – 60% Latino)

CA-28 (D-Berman): 71% Obama – 27% McCain (53% Minority – 38% Latino)

CA-29 (D-Schiff): 68% Obama – 30% McCain (69% Minority – 22% Asian; 39% Latino)

CA-30 (D-Waxman): 67% Obama – 32% McCain

CA-31 (D-Becerra): 81% Obama – 17% McCain (82% Minority – 62% Latino)

CA-32 (D-Chu): 63% Obama – 35% McCain (82% Minority – 28% Asian; 52% Latino)

CA-33 (D-Bass): 84% Obama – 14% McCain (79% Minority – 54% Latino)

CA-34 (D-Roybal-Allard): 70% Obama – 28% McCain (82% Minority – 67% Latino)

CA-35 (D-Waters): 88% Obama – 10% McCain (92% Minority – 40% AA; 45% Latino)

CA-36 (D-Harman): 63% Obama – 35% McCain

CA-37 (D-Richardson): 80% Obama – 19% McCain (89% Minority – 56% Latino)

CA-38 (D-Napolitano): 64% Obama – 34% McCain (71% Minority – 60% Latino)

CA-39 (D-Sanchez): 64% Obama – 34% McCain (84% Minority – 67% Latino)

VRA Implications: My map features 12 majority-minority seats in LA County which is the same as the current map. However, the Dreier seat which is currently majority-white has been pushed entirely out of LA County to become a majority-Latino Inland Empire seat. On the flip-side, CA-36 (Harmon) has dropped from 54% minority to 43% minority – which is an unavoidable consequence of eliminating the CA-46 coastal strip to take in Palos Verdes. Meanwhile, McKeon’s CA-25 seat has gone from 43% minority to 53% majority-minority. So, to make the long story short, this represents a net gain of one majority-minority seat.

The other significant VRA development is that my map features 8 majority-Latino seats versus 6 in the current map. The current majority-Latino seats have these Latino percentages: 77%, 70%, 69%, 62%, 61%, 55%. My Latino-majority seats have these percentages: 67%, 67%, 62%, 60%, 60%, 56%, 54%, 52%. In short, I’ve chosen the best balance I could come up with between maintaining the population of the current majority-Latino seats versus creating new ones. If the Commission would prefer one or two more seats above 65% then they’ll just create choppier lines. If so, more power to them; I don’t see the need to do so.

Partisan Impact: There’s really not much to say here. LA County will feature 13 heavily Dem seats and one McKeon seat. The Dreier seat, of which a majority lives outside LA County anyhow, will surely be displaced. The CA-25 McKeon seat that I’ve drawn here is basically the most Republican leaning seat that can be drawn in the less-populated northern half of LA County. If it dips south anywhere besides the places where I’ve chosen then the Obama percentage will rise.

Beyond that, there is of course no possibility of a second GOP seat in LA County. However, I do want to comment on what seems to be a widespread misconception that a more GOP leaning CA-36 might be drawn if it were oriented eastward (from Palos Verdes to Long Beach) rather than northward (from Palos Verdes to Santa Monica). In reality, this is simply not true and the only reason why I can figure this mistaken belief has seemingly become common on SSP is because it keeps getting repeated in every California thread despite having no support whatsoever. In actuality, the CA-36 district that I’ve drawn here is the most Republican seat that can viably be drawn in southwest LA, and everything around it is heavily Democratic. The more you shift it toward Long Beach, the more Democratic it becomes.

Incumbents: Perhaps the most interesting part of my LA map is in how it scrambles the incumbents, which I imagine is very likely to be the case with any LA mapping scheme that is less gerrymandered than the current one. Needless to mention, this is only representative of what might happen, but here’s the list of LA districts with who currently represents the district of that number and with where the incumbents end up on my map:

CA-25 (R-McKeon): McKeon (Santa Clarita)

CA-27 (D-Sherman): Vacant

CA-28 (D-Berman): Sherman (Sherman Oaks); Berman (Valley Village); Schiff (Burbank)

CA-29 (D-Schiff): Becerra (Eagle Rock)

CA-30 (D-Waxman): Waxman (Beverly Hills)

CA-31 (D-Becerra): Roybal-Allard (Boyle Heights)

CA-32 (D-Chu): Chu (Monterey Park)

CA-33 (D-Bass): Bass (Culver City); Waters (Hancock Park)

CA-34 (D-Roybal-Allard): Linda Sanchez (Lakewood)

CA-35 (D-Waters): Vacant

CA-36 (D-Harman): Harman (Venice)

CA-37 (D-Richardson): Richardson (Long Beach)

CA-38 (D-Napolitano): Dreier (San Dimas)

CA-39 (D-Sanchez): Napolitano (Norwalk)

How they’d sort it all out is obviously speculative at best, but I do want to make a few observations.

Waters does not live in the current CA-35 district anyhow, so I doubt this makes any difference in that regard. I don’t know if she maintains a property or whatever in Watts, but her current arrangement would obviously work just as well no matter how the new maps are drawn.

Berman and Sherman live very close together and it took some creative map-drawing to keep them in separate districts in 2001. The likelihood of them ending up in the same district is high, especially if the San Fernando Valley is consolidated into a 60%+ majority-Latino district, as seems quite likely.

It doesn’t really matter where Dreier ends up, since he won’t be able to win an LA County district regardless (unless he moves to CA-25), so I presume that if he does end up as the only resident incumbent of a district then one of the Democrats will move over to help sort things out.

Finally, Harman doesn’t actually matter anymore for these purposes. So, it’s worth noting that of the two contenders for the CA-36 seat, Debra Bowen lives in Marina del Rey and would still be in CA-36 on my map. Janice Hahn lives in San Pedro near the Port of Los Angeles which ends up in CA-37 on this map. Since the lower part of CA-36 will almost surely have to shift westward to take in Palos Verdes, it’ll be a very close call which side of the line San Pedro ends up on.

If you’re a Winograd supporter, get real. (She lives in Marina del Rey.)

Orange County & the Inland Empire

The next two maps cover Orange County and the Inland Empire. In my mapping scheme, this also includes CA-26, which is now a Riverside County seat. It’s worth noting that, although CA-26 is more commonly thought of as an LA County seat, a majority of the current CA-26 population actually lives outside LA County. Since the seat anchored by Riverside city is the least congruent with any of the current districts, it makes sense to label that CA-26 for comparative purposes.

My mapping scheme is very straightforward, so there’s not really much to add that I don’t cover below. Since the numbering ends up out of order, here’s how these seats break down by county:

San Bernardino

CA-41 (R-Lewis): 44% Obama – 54% McCain

CA-42 (R-Miller): 55% Obama – 43% McCain (71% Minority – 54% Latino)

CA-43 (D-Baca): 65% Obama – 33% McCain (79% Minority – 61% Latino)


CA-26 (R-Dreier): 59% Obama – 39% McCain (70% Minority – 53% Latino)

CA-44 (R-Calvert): 41% Obama – 57% McCain

CA-45 (R-Bono Mack): 50% Obama – 48% McCain (52% Minority – 43% Latino)


CA-40 (R-Royce): 44% Obama – 54% McCain

CA-46 (R-Rohrabacher): 50% Obama – 48% McCain (52% Minority – 22% Asian; 25% Latino)

CA-47 (D-Sanchez): 56% Obama – 42% McCain (77% Minority – 61% Latino)

CA-48 (R-Campbell): 53% Obama – 45% McCain (59% Minority – 37% Latino)

VRA Implications: On the current map, these 10 seats include only 3 minority-majority seats (CA-40, CA-43, CA-47) two of which are also majority-Latino (CA-43 & CA-47). My map has 7 out of 10 seats as majority-minority and 4 of these are majority-Latino. It’s difficult to envision a more drastic improvement from a VRA standpoint.

Of the current majority-Latino seats, I’ve nudged Baca’s CA-43 up from 58% Latino to 61% Latino. Loretta Sanchez’s CA-47 has dipped from 65% Latino to 61% Latino. I should say that I was somewhat ambivalent about this latter move, but it’s a necessary tradeoff if CA-48 is to become a solid majority-minority seat in its own right. More importantly, I suspect that the actual Census figures will make it quite easy to turn CA-48 into a majority-minority Irvine-based seat alongside an Anaheim-based CA-47 district with a Latino percentage about where it’s at now.

Otherwise, there seems little doubt that three compact majority-Latino seats can be anchored in San Bernardino, Chino, and Riverside. Beyond that, pulling CA-46 out of Palos Verdes virtually assures that it becomes majority-minority – even if only slightly so – and CA-45 is probably even more Latino than the ACS figures project.

Finally, it’s also worth noting that CA-40 and CA-41 are each 51% white on this ACS map, so they may actually be majority-minority in the Census count.

Partisan Impact: The partisan impact for the Inland Empire is perhaps the most substantial aside from the Central Valley. The new Riverside seat is virtually assured to be strongly Dem – and basically replaces the current heavily-gerrymandered Dreier seat. Meanwhile, Gary Miller’s seat goes from a 53% McCain district to a 55% Obama district.

In Orange County, the most significant impact is on Campbell who goes from a 49%O-49%M district to a 53% Obama district. Rohrabacher’s seat becomes just slightly less secure on my map (going from 48%O/50%M to 50%O/48%M) but it’s worth noting that if CA-46 were to take in any more of Long Beach, or if it took in the part of LA County running more along the county border rather than along the Long Beach coastline, then it would become steadily more Democratic.

On the flip-side, CA-40 and CA-44 both become more solidly GOP as white voters get packed here to make way for the majority-minority seats. CA-44 in particular goes from a 50% Obama district to a 57% McCain district.

Incumbents: Several incumbents get displaced by this mapping scheme. Gary Miller lives in Diamond Bar which ends up in the CA-40 district. This would probably be just as well because if something like these maps were the case then it makes far more sense for Miller to challenge Royce (Fullerton) in a primary than to run in the CA-42 district. If CA-40 doesn’t enter LA County, then I think the most likely alternative is for Miller to end up in one of the ultra-Dem LA County districts, so my map is probably about the best-case scenario for him.

This map also draws Lewis (Redlands) and Baca (Rialto) into the same CA-43 district, but I’m sure Lewis could easily run in CA-41 regardless (and he’s reportedly considering retirement anyhow). Finally, Calvert lives in Corona which ends up in CA-42 here, but I’m sure he’d run in the CA-44 district (or its equivalent) no matter where he ends up.

San Diego

Below I’ve included two maps. The first is a broader view of SoCal while the second is a closer view of San Diego so that it’s clear exactly where I’ve drawn the lines.

In my view, this mapping scheme is the most likely arrangement for San Diego County. The city and suburbs of San Diego can clearly anchor two compact seats, and most people seem to agree that Imperial County is most ‘naturally’ attached to San Diego County versus Riverside County. Once those decisions are made, then the rest seems obvious: maximize Latino percentage in CA-51, turn CA-52 into a majority-minority district, and then place CA-50 & CA-49 northward along the coast.

That said, I realize that there’s a contingent that favors an Imperial to Riverside mapping scheme instead, so I’ve actually decided to create an alternative map on that basis. For the sake of keeping things clear, and since the main body of this diary is long enough as it is, I’ll post that separately in the comments.

For now, here are the partisan figures for the five San Diego County districts. Once again, I’ll post further comments after the maps.

CA-49 (R-Issa): 47% Obama – 51% McCain

CA-50 (R-Bilbray): 50% Obama – 49% McCain

CA-51 (D-Filner): 57% Obama – 42% McCain (76% Minority – 61% Latino)

CA-52 (R-Hunter): 54% Obama – 44% McCain (55% Minority – 32% Latino)

CA-53 (D-Davis): 63% Obama – 36% McCain

VRA Implications: The main VRA implications are that I’ve pushed the Latino percentage of CA-51 from 53% to above 60% and I’ve turned CA-52 into a majority-minority district. I chose CA-52 instead of CA-53 because, much to my surprise, when I looked at what remained after I pushed CA-51 to 60% Latino it was clear that CA-52 could much more easily become majority-minority than could CA-53. When I tried doing it with CA-53 instead, I ended up with far more weirdly shaped districts. That said, we can’t rule out the possibility that the actual Census figures will make CA-53 the more viable majority-minority district.

It’s also worth noting that CA-52 was already a 52% Obama district when I initially mapped it based on geographic cohesion without regard to minority percentage. The ‘VRA modification’ just bumped it up, obviously.

Partisan Impact: The main partisan consequence of this mapping scheme is to turn CA-52 from a 53% McCain district to a 54% Obama district.

This is what it comes down to for San Diego, as I see things: CA-51 and CA-53 will be safe Dem no matter what. (I’ve seen a fantasy floating around that pushing up the minority percentage of CA-51 might somehow endanger Susan Davis. This is clearly false. The heavily Dem western parts of San Diego city will anchor a Democratic district no matter what happens with CA-51.) Now, it might well turn out that Filner is subject to a strong Latino primary challenge (assuming he doesn’t run for mayor of San Diego), but that won’t matter in partisan terms.

On the flip-side, CA-49 and CA-50 are likely to remain GOP leaning much as they are now (they each become a bit less GOP on my map, but should still be expected to elect Republicans for the time-being, and certainly to re-elect the current incumbents).

The only real question is what happens with CA-52. In short, the more compact CA-52 becomes the more Democratic it becomes. Since the California guidelines mandate compactness except where VRA requirements suggest otherwise, and especially since a more compact CA-52 district in the eastern suburbs of San Diego city can be made majority-minority (which is not the case if a district similar to the current CA-52 in the San Diego County hinterland were maintained) then I think this is the more likely outcome.

Incumbents: Issa and Bilbray live just 8 miles away from each other, in Vista and Carlsbad respectively, so it’s quite likely they’ll end up mapped in the same district. My map happens to put them both in CA-49 (while my alternative Imperial-to-Riverside map happens to put them both in CA-50). The short of it is that one or the other will very likely have to move or else run in a district where he no longer resides. The other three incumbents are very unlikely to be displaced.

63 thoughts on “California with 37 VRA seats”

  1. I love the neatness of the lines.

    Some comments:

    I’m fairly sure you mixed up the Obama-McCain numbers in CA-22; there’s no way Obama got 59% of the vote in Kern County.

    You kind of screw over the Asians in Los Angeles. I think a VRA map would try to create an Asian-plurality seat there, which is hard but doable. But the Asian community in Monterey Park is placed into CA-29, while the Asian community in Diamond Bar is placed into a CA-40 (a white majority, strongly Republican-voting district).

  2. Well reasoned and well written. It also particularly well formatted. How did you center the section headings? (One minor correction: You have Miller, Pelosi, and Lee listed as Republicans in your district summaries).

    I’m interested in seeing the alternative map. I’m working on doing the ACS to Voting District transition myself on a CA map with Imperial-Riverside. The commission’s standards are somewhat in conflict here. Either way results can result in a Hispanic-majority seat. The connection to San Diego County better reflects the commmunity of interests standard (transportation and media links, and also the Mexican border issue). The connection to Riverside better fits the compactness and the not-bypassing-nearer-population-centers standards (at least in my view).

    One thing I’m curious about is the extent to which you took into consideration locality boundaries. My impression of the maps makes me think that you didn’t try to avoid city-splitting as hard as you did county-splitting. But that’s just an impression.

  3. drawn to be favorable to Republicans, the picture for them is still pretty bleak on this map. Their number of “sure thing” seats doesn’t seem to make it into the double digits.  

  4. As a SF resident, I agree with your take that Stark is the odd man out in the bay area; Eshoo certainly will stay in CD14 here, so Honda goes to CD13, Barbara Lee stays in CD9, and thus Pete Stark gets to retire.  He could try to run in CD16, but my guess is that McNerney would win the primary there.  I don’t know how much of Fremont and Hayward end up in 16 though.

  5. A good analysis generally of California.

    It is clear that how the commission views the community of interest in the far north of California will be critical to CD02.

    If, as you do, it considers the northerly position of Humboldt County to be the defining fact, CD is lean or likely R.  If it regards Humboldt county to be part of a coastal community of interest, CD is safe R.  The net effect on other districts in the area would not be affected.

    The major transportation network for the county is oriented north and south, not to the east.

    It has more in common economically, in the type of industry along highway 101, south to Mendocino County and Santa Rosa, or, to a lesser extent, to the sea rather than jumping the mountains.

    The type of agriculture is to the north and south rather than to the factory farms of the Sacramento Valley.

    It has no relation to the water distribution concerns of the Sacramento Valley.

    Politically, of course, as demonstrated by your map, the Democratic Party would be better off putting it into CD02

  6. Imperial to Riverside Alternative

    Below I’ve included my alternative mapping scheme for SoCal that attaches Imperial County to Riverside County instead of San Diego County.

    I don’t have a strong view one way or the other, but as I said to begin with, I think this is the less likely alternative. The main reason is because it seems to me that San Diego itself is the most important consideration and this alternative would reduce the number of districts clearly anchored by the city. The other reason is because you lose a majority-minority seat, though that’s mitigated somewhat because you gain a majority-Latino seat (that would otherwise be a ‘coalition’ majority-minority). Whatever the case may be, I agree that reasonable people can differ on this, so here’s the alternative.

    Although I make slight changes to a few other districts, the ones below are where significant changes take place:

    CA-44 (R-Calvert): 42% Obama – 56% McCain

    CA-45 (R-Bono Mack): 55% Obama – 44% McCain (61% Minority – 52% Latino)

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

    CA-50 (R-Bilbray): 50% Obama – 49% McCain

    CA-51 (D-Filner): 66% Obama – 33% McCain (84% Minority – 58% Latino)

    CA-52 (R-Hunter): 44% Obama – 54% McCain

    CA-53 (D-Davis): 63% Obama – 35% McCain

    VRA Implications: From a VRA standpoint, the Imperial to Riverside maps result in two majority-minority districts (CA-45 & CA-51) where there were three in the Imperial to San Diego version (CA-45, CA-51, CA-52). On the other hand, both of these districts are majority-Latino whereas only one of the three would be majority-Latino on the first alternative.

    Partisan Impact: The partisan consequences are essentially a wash – at least on paper. CA-45 goes from a 50% Obama district to a 55% Obama district in the Imperial to Riverside version. CA-52 goes from a 54% Obama to a 54% McCain district. The only other district with a significant partisan shift is CA-51, which goes from 57% Obama to 66% Obama.

    That said, Mary Bono Mack certainly has more crossover appeal than Duncan Hunter, so the odds of her holding on to a Dem-leaning district are obviously much higher.

    Incumbents: As mentioned above, the only difference between the two alternatives so far as incumbents are concerned is that Issa and Bilbray end up in CA-50 rather than CA-49.

  7. An Asian majority district.  It would be interesting to see how that district votes and the candidates it puts through the primaries for both D and R.

  8. might have our disagreements on VRA & politics but I stand in awe of a map like this.  I know that there is a great deal of work that goes into this. Let me say that this is by far the best map done by anyone here for CA.

    This map does take care one of my pet peeves about CA’s maps.  Monterey county is not split. There is only one hybrid district that I see (some sort of urban, suburban & rural combo) that is clearly a choice.  That would be CD10.  I would perfer CD10 drop down into Santa Clara and then move CD12(eventually) into CD17.  That’s a choice and I don’t quibble with that too much.  In a map with 53 CD’s people will have different perferences.  So I am not bent of out shape screaming about that.  I might add, however, that the arrangement of coastal counties would do away with the West LA to Ventura county move.  Every map here still have that arrangement.  

    I see a few different moves being made by the redistricting commission but overall this is not an unreasonable map.  Reasonable people can see 53 congressional districts being drawn different ways.

    I am eager to see updated partisan numbers for CA.  Just eyeballing this map I see 24 districts that were won by Carly F in 2010.  That includes Capps & Galleghy’s seats on the coast plus most of the Inland valley seats.  In 2008 McCain only won 11 CD’s in CA.  There was a huge ticket split in 2008.  In this map alone, which I think favors the Ds a bit, I still see +20 seats the GOP could win.  

    Thanks for posting it.  

  9. This is a question I posed in a Texas redistricting diary and I’ll ask again.  What are the paramters on creating Latino majority/plurality districts where there is a large segment who can’t in fact vote?

    I ask because the census counts all residents and not just citizens.  And despite the obvious Latino bias (towards Democrats which is more manifest in California than elswhere) their percentage at times lags far behind the Democratic vote telling me there is a large pool of Latinos that are not residents and/or undocumented.

    Would these three central valley districts really qualify as VRA districts?

    CA-20: 50% Obama – 48% McCain (75% Minority – 66% Latino)

    CA-21: 51% Obama – 47% McCain (62% Minority – 42% Latino)

    CA-22: 39% Obama – 59% McCain (54% Minority – 42% Latino)

    And what are the rules about packing in people who can’t in fact vote?  A district with a high illegal alien population seems to be the perfect trojan horse to get a white conservative republican elected in a technically “minority” district.

  10. diary on CA yet …

    I could see the map very close to this in reality (the only “problem” here I had may be w/ CA-19 as it seems to encompass two rather disparate areas … but then again, as everything else seems to make sense in a logical way, then one district may necessarily have to be a “left-over” one like this) …

    what’s interesting here are the coastal districts of southern CA (CA-46, 48, 49, 50) … they all come out basically 50/50 in partisan terms and for the next few elections may continue to elect Republicans to Congress, but if I had to bet on outcomes in the latter part of the decade, I would put my money on Dem. reps here … this coastal area has trended very Dem. over last few decades; here’s pres. vote for each district (using 1992 map equivalents, which were very similar to existing districts):

    Dem. vote for Pres.: 1992/1996/2000/2004/2008

    CA-46: 32/38/42/42/48

    CA-48: 31/36/39/40/49

    CA-49: 29/34/39/36/45

    CA-50: 32/38/43/44/51

    in this light, Obama’s percentage does not appear as an anomaly, but just part of a trend in this area …

    what’s interesting from above is that even between 2000 and 2004 the Dem. vote went down in only 1 out of 4 of these districts — DESPITE the fact that in general Kerry did worse than Gore in CA and nationally and DESPITE the fact that the 2002-version CA districts were made more favorable to the GOP vs. 1992-version districts …

    the clock will eventually run out for Rohrabacher, Campbell, Bilbray and even Issa …

  11. dude. I have to say I have great admiration and appreciation for people who make the time to create quality maps for the rest us to view!

  12. As I said in your first California diary, everything won by Obama by a 9% margin or more is Democratic, everything won by a 6% margin or less, or lost, is held by a Republican.  

    Using the same rationale for these seats…

    Democratic pick-ups: CA-24, CA-26, CA-42, CA-52

    Republican pick-ups: CA-11, CA-20

    Toss-up: CA-48

    So the baseline performance for the Democrats would be 1-3 net pickups.  Plausibly they could lose a seat or two, as nationally seats with PVIs as high as D+4 are often held by Democrats – although I doubt the current incumbents are in all cases the right sort of moderate Republicans to hold onto these seats.  

  13. I especially love the smacking down of the Rethug talking points.

    BTW, Garamendi already lives in CA-03, so unless his house is really far from whatever district he wants to run in, I doubt he’ll move.

  14. Imperial County being attached to CA-51 looks much better in this form than it does in the current map. Duncan Hunter Jr. would obviously face a climb, he may not even run. Former Assemblywoman Shirley Horton would be the GOP’s best bet.

    Orange County will likely get another majority-minority district like on this map. There is a lot of minority voting strength spread out there, so it’s probably time for another one.

    Costa would likely run in the Fresno district, rather than the pink one, as he resides in Fresno. Nunes is in the pink district.

  15. You’re right. I fixed CA-22 and looked back over all the rest to make sure there weren’t any more typos in the Obama/McCain figures. I didn’t find any more.

  16. I tried to make an Asian plurality seat in eastern LA County (and also in Orange County) but the Asian concentrations are too scattered. Maybe someone could do a better job than I did, but wherever I picked an Asian community to anchor a district, I had to add far too many non-Asians before I could get the population up to where it needed to be.

    In short, I could not make a coherent district with a substantially higher Asian percentage than in the districts I’ve mapped out. I settled for bumping CA-32 from 19% Asian to 28% Asian. At best I could get it up into the low 30s, and I didn’t think that was worth losing a majority-Latino district in the exchange.

  17. I had Schiff labelled as a Republican too. They’re all fixed now.

    It’s tough to double-check everything when there’s so much text and I was doing a lot of cutting & pasting to expedite things. I’m going to go back over the minority figures too and make sure those are correct, then I’ll get back to the comments.

  18. For centering, I just tested to see whether the [center][/center] tag would work, and it did! Just replace the brackets with tags.

    I did my best to maintain city/town boundaries, but it’s not entirely easy when they’re not demarcated on the latest versions of Dave’s app. What I did when in doubt was use the road maps and concentrations of census blocks to guess where the lines were at. I figured that the margin of error would not be significant, by which I mean that any adjustments to better match city limits should be slight enough that it would not result in a discernible change in the partisan or ethnic composition of my mapping scheme.

    I should also add that I basically disregarded city boundaries in the places where it was necessary to do so for VRA purposes (mainly the Central Valley). The rules set out in the California statutes dictate that VRA concerns take priority.

  19. over VRA issues as RMapper and I have starkly different views on this issue.  There are two issues in my opinion.

    1. Retrogression in districts.  If a seat is 70% hispanic after the 2010 census anything less then that is diluting the hispanic population for purposes of splitting the vote. I might add, however, a seat that is 50% hispanic is clearly not minority majority hispanic. That is  because the actual voting population of hispanics will be nowhere near a majority.

    2. So what % should you shot for a VRA hispanic seat. A good test for minority/majority hispanic seat is 65/70%. That will usually give you a seat with the majority of actual voters being hispanic. Not all hispanic populations are “created equal” in voting matters.  In settled urban areas the % is higher while in agricultural migrant areas its much lower.  There are some real unanswered questions in VRA hispanic matters.  We may see some court guidance this year and we may not.  

    There are several +60% hispanic districts that have white congressman from either party. This will be a big question for redistricting in 2012. I am not sure what the DOJ will say.  

  20. Here are my thoughts on these three districts.

    CA-11: This one’s dicey. I had to manipulate the lines quite a bit before I could nudge it above 50% without reducing the CA-18 Latino percentage. What it comes down to is this: The district that takes up the eastern parts of Stanislaus and San Joaquin will be something of an afterthought. It’ll basically be whatever’s left over once CA-18 is designed to meet Section 5 requirements.

    However, it was clear enough to me that if I weren’t constrained by the census blocks on Dave’s app – meaning if I could subdivide census blocks as the Commission has free rein to do – then it’s very likely that CA-11 can be a slight majority-minority district or very close to it. From a partisan standpoint, it didn’t make any difference. The Obama-McCain numbers didn’t change when I tweaked the boundaries to take it from 47% minority to 52% minority. I just did it out of principle, and there’s no reason to think the Commission wouldn’t do likewise if faced with similar stats. It just makes things ‘sound’ better to have the added majority-minority district, even if it’s just barely above 50%

    CA-21: There are two ways I look at this one. The city of Fresno should quite clearly anchor its own district and so I think that will be the case whether or not it’s a majority-minority district. Secondly, since both CA-18 and CA-20 use parts of Fresno in their current configuration, my deduction is that the Fresno area Latino population is fairly well-settled and reliable. Beyond that, it’s clear enough to me that CA-18 does not need to enter Fresno County to create a VRA compliant district. The issue then is whether CA-20 needs to take more of Fresno.

    This is basically a coin-flip in my view. As I see things, it makes more sense to give Fresno its own district and then wrap CA-20 around the southwest of Bakersfield if necessary (but I don’t mean with the narrow salient of the current map; I mean take southwest Kern County as a whole). The reason is because, if my proposed CA-20 district is not adequate for VRA purposes, then you have to decide whether to preserve the integrity of Fresno or of Bakersfield or of neither. Since I think Fresno is the ‘more important’ city in that regard, and since Bakersfield cannot anchor a district on its own regardless, I think that’d be the way to go. But one could just as easily choose otherwise. It’s ultimately a value-judgment that I can’t predict.

    CA-22: I think it’s clear that the Kern County Latino population is mostly stable and reliable or else the current CA-20 configuration would make no sense. If you do not put the Latino parts of Bakersfield in CA-20, then it’s virtually assured that the resulting Kern County district will be majority-minority. The fact that the very conservative white population will cancel out their vote won’t change that. To put it differently, my CA-22 district isn’t “designed” to be a VRA district per se, it just becomes majority-minority by default – so long as CA-20 doesn’t carve out the Latino areas.

    As to your general question about illegals, I think what it basically comes down to is voter registration. If the registration numbers for Latinos in CA-20 were to drop from, say, 55% to 40% then it would almost surely be deemed retrogressive. Clearly then, it would have to be reconfigured more like the current CA-20 district. In my view, this is actually quite likely, but since I can only work with the data that I actually have available, I’ve created the district that I think meets the requirements based on the ACS figures at Dave’s app.

    In part, that’s for the simple reason that I think so-called “Trojan Horse” districts will be exactly what the Republicans on the Commission will advocate. So, as I noted above, this may be what it takes to get three GOP votes for approval. That requirement has been my guiding principle in trying to make decisions that favor the GOP all else being equal.

    Whether such districts fly with the courts, with the DOJ, or with the non-GOP members of the Commission is a whole ‘nother matter..

  21. San Diego county is one of those spots where I find fault with this map.  That’s not a hit on Rmapper as its common here for all the CA maps.

    Filner & Davis’s democratic % decrease neatly but without endangering while those democratic voters are neatly put in Hunters & Bilbray’s seat which makes them less democratic.  What a nifty move.

    Filner’s seat becames 5% or so less minority by my calculation.  The % of white voters goes from 19 to 24% under this map.  Thus the GOP % improves but is swamped in Filner’s seat.  In a nifty move Davis’s seat which is less then 50% white gains white voters and becomes a bit more GOP while Hunter’s seat becames minority majority.  That’s clear retrogression as minority majority voters are moved from one seat to another.  

    That’s just a minor move and does not affect the map per se just how voters are arranged in one county.  That’s why I think the map overall is pretty good its just how you arrange particular voters. San Diego and Orange county plus the Inland imperial counties are places where I would arrange things differently but I still respect how RM did it.  

  22. and I like the term “trojan horse”.  I struggled with a word to describe what the GOP may very well do in TX23 & TX27.  

    I believe you will see hispanic seats there with similar % to the current seats but the GOP will pay very particular attention to what hispanics are included and what white voters are included.  Even in TX not all white voters are the same.  In high income white suburbs you see see very high GOP % while in gentrified or blue collar white precincts you see higher democratic %.  In addition you can match up census blocs to hispanic voter precincts and determine where you get the higher turnout.  You can look in CA and see that 70% hispanic in Kings county is pointless when very few actually hispanics there vote.

    All very intersting stuff

  23. I know this is surprising to you, but when I made CA-53 a majority-minority district instead of CA-52, Duncan actually ends up with slightly worse numbers (55% Obama). That’s because once you push CA-51 up to 60%+ Latino, then in order to put enough of a minority population in CA-53, you have to wrap it around the northwest boundary of CA-51. That means that the CA-52 seat has to push into northern San Diego. In short, you end up with much the same outcome but with more distorted boundaries.

    The alternative, of course, is to push CA-52 out into San Diego County will pulling CA-50 down into San Diego, which then puts that seat at high risk for the GOP. The problem that the GOP faces in San Diego is much the same as the problem it faces in Los Angeles and the Bay Area: A lot of white liberals that have to go somewhere. The basic assumption that holds in the South, by contrast, where there’s a direct correlation between packing minorities and GOP performance doesn’t hold. If you let go of that notion, then I think a lot of your apparent misconceptions about these parts of California will go with it.

  24. I think roguemapper, if anything, creates SD districts that are as favorable to the GOP as can be created (like was postulated at beginning of diary — “All else being equal, I have generally made those decisions which are most helpful to the GOP.”) …

    you can plainly see this by just looking at how SD City and SD are configured:  using the most “neutral” criteria possible — city and county lines — according to the ACS, San Diego City has almost the exact number of people to equal 2 congressional districts … while the part of SD Co. outside City limits has enough for 2 districts (w/ a little bit left over)

    … the SD City part voted 63 Obama – 36 McCain — and that percentage is relatively evenly spread out throughout the city (no huge differences in voting pattern like you would find in, for ex. Jacksonville FL between white and black areas) … so, if using the most “neutral” of criteria (city limits in this case) all one can do is create 2 SOLID Dem. seats for San Diego City …

    … now, on to part 2 — the part of SD Co. outside of City limits — that area voted 48 Obama – 50 McCain — so, you can either create 2 toss-ups or 1 Dem. seat/1 GOP seat …

    so, in total, “neutral” plan for SD should encompass either:

    2 Dem. seats & 2 toss-ups OR 3 Dem. seats & 1 GOP seat — rouguemapper creates 2 versions for SD, in first it’s 1 solid Dem.; 1 lean Dem.; 2 toss-ups (and “left-over” area/dist. 49 that’s lean GOP)– this is perfectly in accordance w/ what a “neutral” map would do …

    in alternative version, it’s 2 solid Dem.; 1 toss-up; 1 solid GOP (and “left-over” area/dist. 49 that’s lean GOP) — if anything, this version is slightly MORE favorable to the GOP than “should” result under “neutral” criteria …

    any idea that one can create some sort of map with 1 “super-minority-majority; super-Democratic” district in SD surrounded by 3 GOP-leaning districts is not based in reality — it’s technically not possible using any fair measure of compactness, and even if you could draw 1 “super-Democratic” dist. there using extreme forms of pro-GOP gerrymandering, you’d still have a hard to impossible time to get ALL the surrounding districts to be GOP — the math just doesn’t add up in a county that voted 54 Obama – 44 McCain !

  25. Again I think your map is great.  I just disagree with some of the particulars within certain counties.  You could be right.  I will attempt a CA map when all the data comes in. I see SD just a little different then you do.  For instance I don’t see Filner’s seat getting Imperial county (classical Trojan as its high in hispanic % and low in turnout)or really any of San Diego  county east of say Chula Vista.  I see Filner seat as being centered on Chula Vista and the hispanic cities and area around there.

    That’s just my guess.  I don’t see Filner and Davis’s districts each becomine more white and less democratic.

    I think its likely Imperial county will be attached to San Diego just not to  downtown district but rather a district with the Eastern desert area and other suburbs.  I see that as more logical as there are plenty of hispanics, Asians and AA  in San Diego proper for  two minorit majority seats.  Why stretch out to Imperial county?  That’s my logic and who knows I could be wrong.  

  26. The issue with San Diego in this regard is that when you push the CA-51 district up from Chula Vista into the National City area to take in enough Latinos to get to 60%, you end up with a significant Asian population (mainly Filipino) to the east side of the district. If you then divide the San Diego area evenly into two compact districts (CA-53 & CA-52) then it’s far easier to make the eastern district majority-minority by simply taking in the North Park area. In order to make CA-53 majority-minority instead, then you have to wrap it around CA-51 to get at those Asian neighborhoods.

  27. and let me respond here to both SS & RM here. I take in both of your points and I don’t even feel ganged up here as its 2-1 against.  Both of your points are along the same and make sense.  That’s why the bipartisan commission will have tons of thorny issues to deal with.  

    I will only concede that my local knowledge of the San Diego suburbs is inferior to you guys.  Wow.  I am very impressed.  You guys know your Chula Vista from your Cucamonga.

    That being said I am convinced that as CD53 & CD51 can be configured in San Diego county to nearly match their current %.  We do not have the census tract data yet but CD51 looks to be about ideal and CD53 looks to need to add 50K or so.  Just a guess.  You could leave the seats as they are and swap area between almost to maintain the same partisan % and the same population variances. The current map is not that twisted and I might the map above has some twists too in SD.

    So my basic point remains its just odd that CD51 adds GOP area to it while it sheds democratic area.  While CD53 adds some GOP area and loses D area.  Stronger and weaker GOP areas would be accurate.  I am assuming that if I get my map making skills up that I do compact seats to duplicate the current ones.

    I enjoyed your points and I do see them.  I hope you see mine.  

  28. So my basic point remains its just odd that CD51 adds GOP area to it while it sheds democratic area.  While CD53 adds some GOP area and loses D area.  Stronger and weaker GOP areas would be accurate.  I am assuming that if I get my map making skills up that I do compact seats to duplicate the current ones.

    I’m not sure I understand what you’re saying here. I think there may be a typo that’s making it difficult for me to decipher (not to mention I have mild dyslexia which is why I so often have typos of my own).

  29. If I gather what you’re saying correctly, you think that it would be adequate to simply shift minority population from CA-53 to CA-51, and then add 50K to CA-53. If so, then here’s the problem with that: The current CA-51 is far from a compact district.

    As such, I have taken it as a given that if CA-51 is an Imperial to San Diego district, it needs to take in the San Diego County backcountry in order to get eliminate that thin border strip. That also serves the dual purpose of making CA-52 a compact seat in eastern San Diego.

    Then, if you keep all of Imperial, and if you take in the backcountry, you obviously have to withdraw CA-51 from parts of San Diego, which again serves a dual purpose because you can thereby shed the less Latino areas.

    So, if you then give the white areas along the coast south of San Diego to CA-53, and the Asian/white areas east of National City to CA-52, you have the outcome that you see here, which maximizes the Latino percentage of CA-51.

    Meanwhile, CA-53 is also not a compact district:

    My intent then, in keeping with the new rules, was to compact CA-53 along the coast, eliminating that salient into Lemon Grove, and then placing a compact CA-52 alongside it. CA-53 therefore adds GOP area while shedding Dem area for the simple reason that it is no longer gerrymandered along partisan lines, but rather mapped according to geographic principles.

    It’s also worth noting that I did not look at the current district configurations when drawing up new districts in California, since that is not part of the guidelines. I only look at how the districts match up afterward so that I can number them accordingly for the sake of conversation.

  30. I think that San Diego is one of the areas where the current districts are “cleverly” drawn to benefit the GOP and it shows in the lines (as was the purpose of the “bipartisan protection” plan of 2002) … for ex., that triangular area that is northern-most part of the existing CA-53 is a relatively very Dem. area around UCSD that was attached to CA-53 to make it more Dem. while relatively more GOP areas in inner La Jolla were “passed over” and attached to CA-50 — making that district more GOP than it would otherwise be … (granted, in SD the differences between the most Dem. areas and the most GOP areas are not as big as in many other cities) …

    the biggest “pro-GOP” factor in the 2002 plan in SD was that the northern part of the city was basically split in half and each half was attached to two otherwise GOP-leaning districts, CA-50 and CA-52 … if the two halves were together in one district, they would basically be enough population for one district contained almost entirely within the city itself that would be approx. 55 Obama – 44 McCain — which would have most probably made it elect a Dem. rep around 2006 (I’m extrapolating from how well someone like Busby performed against Bilbray in a much more GOP existing district) …

    a lot of this may turn out to be a moot point, as San Diego Co. overall has rapidly moved towards the Dems. over the last 3 decades … the Dem. candidate has performed progressively better in every single election since 1980 (*save for 1992, but that was the Perot factor):

    Dem. vote for Pres., San Diego Co.:

    1980 – 27%

    1984 – 33

    1988 – 38

    1992 – 37 (*Bush got 36)

    1996 – 44

    2000 – 45.7

    2004 – 46.4

    2008 – 54

    2012 – ?

  31. and I need to walk in this 60 degree heat as I go through the mail.

    Here’s my point

    CD 50 was 51Obama 47McCain you have it 50 obama 49McCain

    Cd 51 63 Obama 35 McCain  now its 57Obama 42 McCain

    CD52 45 Obmaa 53 MCain  now its 54 Obama 44 McCain

    CD53 68 Obama 30 McCain you have it 63Obama and 36 McCain

    These four districts are nearly perfectly balance population wise.  Yet when you redrew the map you swapped out areas to weaken the D seats but not enough to change the balance but the GOP seats were shifted to the D side.  You moved CD52 westward and CD53 eastward.  Why?  Its clearly a partisan move in my opinion.  

    Here’s what I suggest.  Give Imperial & the desert precincts to CD52.  CD51 then picks up hispanic voters in the  Lemon grove area of CD53 and in Downtown SD. CD53 can pick up some parts of western CD52 closer to the coast.

    You have too much movements in your map in SD that changes the nature of each seat.  CD 53 picks up GOP area east of its current location while CD51 picks up  GOP area the desert area of Northern san Diego county.  In return each seat gives up democratic area.  The obvious solution for instance is to have CD51 go northward to CD53 and not to Riverside county area. Likewise was does CD53 bypass area that CD52 has in its westward area? You move areas around when its more logical not to IMO.

    Basically you swapping areas and moving seats eastward and westward plus northward when its more logical to take the nearest area available to fill out CD’s.

  32. Here’s what I suggest.  Give Imperial & the desert precincts to CD52.  CD51 then picks up hispanic voters in the  Lemon grove area of CD53 and in Downtown SD. CD53 can pick up some parts of western CD52 closer to the coast.

    I don’t think this would fly on grounds of vote dilution. You would be taking the Latino population of Imperial and placing it in a majority-white, GOP-leaning CA-52 district. Meanwhile, the CA-51 district you’re proposing is clearly the one on my Imperial-to-Riverside map which has a lower percentage of Latinos than the district joining Imperial with Chula Vista.

    Beyond that, since Imperial isn’t going to Riverside in this alternative, then clearly you also have thinner CA-50/CA-49 districts along the coast, which almost surely adds the issue that you’re now dividing the coastal communities where I have endeavored to avoid that.

    So, you’ve lost a majority-minority district, replaced a more Latino district with a less Latino district, shifted a significant Latino population out of a majority-minority district and into a majority-white district, and subdivided more communities of interest. The only rationale I see for any of this would be to try and boost Duncan Hunter, and that’s not a consideration the Commission is permitted to accommodate.

  33. I think this is what you’re suggesting. Let me know whether it is or whether I need to make further modifications, and then I’ll figure up the partisan stats when I get back.

  34. Beyond that, since Imperial isn’t going to Riverside in this alternative, then clearly you also have thinner CA-50/CA-49 districts along the coast, which almost surely adds the issue that you’re now dividing the coastal communities where I have endeavored to avoid that.

    Scratch that part – the opposite is actually the case. My mind was working backwards I guess. LOL!

    So, you don’t have that issue, just the VRA issues that I listed. What you’ve essentially done is attach Imperial to the eastern environs of San Diego rather than to the southern environs of San Diego.

  35. of diltution as you increase the white % in CD51 in your map from 19 to 24%. The % of minority voters in CD51, in your map, naturally drops 5%. It appears that the % of hispanic voters goes down too.  How can it be dilution for me to suggest that imperial be attached to CD52 when I suggest putting in CD51 available hispanic voters in CD53.

    I want to see CD51 lose Imperial county but it should gain hispanic voters in the City of San Diego and from the Lemon Grove area now in CD53.

  36. I wonder if it would be possible to create a single North Coast district (DRA 2.0 is crashing on me, no memory) by combining Marin, a thin coastal strip of Sonoma and all points north. Sonoma has to get split either way, this would simply put Santa Rosa etc in with Napa County.

  37. I plead the 5th on local knowledge of San Diego.  

    The normal rule of thumb in CA is the higher the elevation the more republican but that’s a guess on my part.

  38. We’ll just roll with it then. If you attach Imperial to San Diego then you only have two viable choices: the arrangement I originally mapped that attaches it to the Latino population in the southern parts of the San Diego metropolitan area, or this arrangement which attaches it to eastern parts. This is how the districts have changed:

    Option 1

    CA-51 (D-Filner): 57% Obama – 42% McCain (76% Minority – 61% Latino)

    CA-52 (R-Hunter): 54% Obama – 44% McCain (55% Minority – 32% Latino)

    CA-53 (D-Davis): 63% Obama – 36% McCain (36% Minority – 18% Latino)

    Option 2

    CA-51 (D-Filner): 66% Obama – 33% McCain (83% Minority – 57% Latino)

    CA-52 (R-Hunter): 48% Obama – 51% McCain (47% Minority – 36% Latino)

    CA-53 (D-Davis): 63% Obama – 36% McCain (35% Minority – 16% Latino)

    I don’t think it’s even a remotely close call which option would be preferable from a VRA standpoint. It’s the one with two majority-minority districts and with the stronger majority-Latino district.

    I must say that you certainly give the appearance of being all for the VRA right up until it doesn’t give you the partisan results that you’re hoping for. LOL

    The current CA-51 district is 79% minority, btw. I doubt any reasonable person would think going from 79% to 76% minority while raising the Latino percentage from 53% to 61% is problematic from a VRA standpoint.

    The current CA-53 district is 49% minority and the current CA-52 district is 28% minority. So, my map goes from 79%-49%-28% minority districts to 76%-55%-36% minority districts. That is by no stretch vote dilution under any standard that I’m aware of.

  39. CD51 was 79% minority in 2001 and is now 83% minority.  I don’t see how any reduction in that % is allowed.   So Davis’s district is not minority majority as of 2001 or 2011 so under VRA the 1st rule is to avoid dilution.  CD51 cannot be diluted.  I might add that by making the district 24% white on a electoral basis it might exceed the 61% hispanic vote.  

    So yes I think option is the choice.  That’s the only way to make the hispanics voters a near majority in actual voters.  

  40. but there are “trojan horse” seats on boths in CA. In theory a seat that are  around  60% minority, say 40% hispanic, 5% AA and 20% Asian should be more likely to elect a hispanic.  There are 4 or 5 seats like this CA.    

    Here’s the % of hispanic in seats

    31st 70%

    32nd 62% (oops Asian now)

    Allard 77%

    38th  70%

    39th 61% (Sanchez #2)

    43rd 58% (Baca)

    47th 65%

    These are the hispanic seats. One at 61% and one at 58%

    CD5  57% minority 20% hispanic  

    CD7  57% minority 21%  hispanic

    CD9  65% minority  18% hispanic

    CD13 62% minority 21% hispanic

    CD 16 69% minority  37% hispanic

    CD 18 61% minority  41% hispanic

    CD20  79% minority 63% hispanic

    CD 28 69% minority 55% hispanic

    CD 29 61% minority 26% hispanic

    CD51 79% minority 53% hispanic

    There is a clear pattern in CA minorities do not build coaltions.  They will vote for the white candidate against the other minority candidate in primaries.

    If want an hispanic to have chance you need the white % to be below 20% and hispanic % to be above 60%.

    If you not you just making the seat another seat for a white democrat.  

  41. Let’s pretend you’ve persuaded me. The map below features an 81% minority CA-51 district. Here are the new stats:

    CA-49 (R-Issa): 47% Obama – 51% McCain

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

    CA-51 (D-Filner): 61% Obama – 37% McCain (81% Minority – 64% Latino)

    CA-52 (R-Hunter): 53% Obama – 45% McCain (52% Minority – 30% Latino)

    CA-53 (D-Davis): 62% Obama – 37% McCain (37% Minority – 14% Asian; 16% Latino)

  42. I’ve addressed whatever it is you’re saying in the main body of my original diary under the “VRA Implications” heading. If you have particular seats in mind that you think aren’t properly covered, then point them out.

    I don’t see any problems with my map on the basis of your reasoning here (as well as I can follow it). At best I see a general statement that amounts to: Wouldn’t it be nice if there were even more Latinos so that the Latino seats could be even more Latino?

    If you want my answer to that: Yes, it would. I’m glad we agree.

  43. and thanks for your work and comments.

    I have different view of VRA and I a have different view of how the CA map will be drawn.  

    1. I think for instance the bipartisan commission will set out to draw more hispanic seats and yes the standard will be 65% or more for hispanic population.  Don’t bother drawing seats for hispanics if its less then 65% in most cases.  

    2. I think most  VRA issues in CA will  eveolve around  retrogression. There are a dozen or so seats where hispanics/minorities are the majority of voters and they will need to maintain that status.  In most seats the key % is how many white voters are there.

    San Diego I believe is an example of what we will see.  There is nearly one million hispanics in San Diego county and your original map divides out the hispanics located near downtown & inner suburbs three ways.  That’s a sure way to ensure no hispanic ever represents San Diego county.  Even in Filner’s seat at 61% hispanic they are less then 50% of voters.

    I think we will see an hispanic district in the San Diego area and it will be 70%hispanic.  

  44. The Commission will bother. They’ll do what the law requires, which is to draw compact majority-minority seats wherever they can be drawn. And if they can be drawn to 65% Latino that’s what they’ll do, and to 60%, and to 55%, and to 50%.

    My maps are intended to approximate what the California Redistricting Commission would do, not what you would do. What you would do is self-evident.

    2. I think most VRA issues in CA will revolve around retrogression. There are a dozen or so seats where hispanics/minorities are the majority of voters and they will need to maintain that status.  In most seats the key % is how many white voters are there.

    There is a difference between “retrogression” under Section 5 and “vote dilution” under Section 2. The first step in effectively discussing the VRA is understanding that difference. I would recommend the NCSL Guide to Redistricting Law 2010 since that’s what the Commission will no doubt adhere to.

    In any case, I invite you again to show me where Latinos have failed to “maintain that status” on my map. Hint: The answer is nowhere.

  45. …a little bit. There is the CA-47 issue that I addressed in the main body of my diary. I don’t have anything further to add to that.

  46. retrogression and dilution plus the NCSL standards.

    In 99% of the cases retrogression and dilution are the same thing.  I do not want to beat San Diego county over again but its a classic case.  CD51 reaches 100 miles to Imperial county for hispanic voters to maintain its status as VRA minority seat.  Guess what? There are 200K hispanics in next door CD53 that could have been attached to it.  So yes you have clear dilution in San Diego that was solved by attaching hispanics 100 miles away.  You change the nature of CD53 why adding white voters while not adding nearby hispanic voters that clearly dilution and retrogression.

    Let me also be clear.  The creation of weird looking districts under VRA is not required but if the bipartisan board wants to stretch a point on compactness to draw a minority majority seat they can do it.

    Back to the same arguement that courts will not mandate racially gerrymandered seats but legislative bodies can and do.  Again I say look at PA, GA, FL, TX, IL and NC.  Legislatively drawn racial gerrymanders are allowed but courts will not force them on states. IMO.  

  47. We’ve gone over most of that, to which there’s nothing more to say.

    The part you left out, however, is what’s important. So, explain to me again how you would create two majority-minority seats in San Diego without Imperial County. Then we can take it from there.

    Oh, and on this:

    You change the nature of CD53 by adding white voters while not adding nearby hispanic voters that clearly dilution and retrogression.

    What silly nonsense.. Here, let me fix that for you:

    Do you feel better now? The VRA doesn’t care about district numbering. In actuality, I’ve increased the minority percentage of the CA-53 district.

  48. I was referring to 2001 map.  I am sorry if I was not clear.  

    There was clearly enough hispanics in the San Diego city and nearby inner suburbs to do a 65% seat with some voters left over.  In 2001 the state legislators did not  do a compact downtown San Diego district with nearby hispanics from suburbs included it.  Instead in 2001 the state legislators divided the hispanic voters-for the most part-between CD51 & CD53.  That was clearly dilution. The solution was to reach around the desert to bring Imperial county into CD53.  That allowed CD53 to be 60% hispanic and to avoid retrogression.  There was most certainly a nifty division of hispanic voters in San Diego county was it allowed as CD51 stayed 60% hispanic and retrogression was avoided.  I might add that some hispanic legislators voted for the map and that helped avoid court fights.  

    You can read the NCSL guidelines  and they warn against dividing minority communities and how that’s a no no under VRA.  Yet its clear that hispanic community was divided in San Diego in 2001.  

    The NCSL guidelines are make good fodder for political science discussions but  in reality they are just suggestions.  

  49. I’m sorry for what I said then. It didn’t make sense in reference to my map, but it does in reference to 2001.

    The 2001 issues about CA-51 were litigated. MALDEF filed suit and lost. I’ve said quite clearly that my own view is that the VRA does not require modifications anywhere near the extent of my maps here (much less to the extent that you suggest).

    The only reasons I’ve seen fit to spend the time to make these maps are: (a) the Commission may very well choose to take this route; and (b) to show that the VRA does not harm Democrats in California, no matter how extreme its application.

    I don’t mean particular incumbents, of course. I could care less whether Berman, Sherman, and Filner get replaced by Latino Democrats. What matters to me is that they won’t get replaced by Republicans. In fact, so far as I’m concerned, the more Latino Democrats there are in Congress the better, because it will help consolidate Democratic gains in Latino voting patterns.

    We have a long ways to go before Latinos in Congress match their proportion of the general population.

  50. lawsuit as I could not recall it. I want to call a truce on this.  Let  me jsut say this about the  hispanic community and how  aware in CA that they are under represented based on %.  The Filner district was one of their huge peeves.  Similar to the TX map of 2003 you see divided hispanic communities and disparage % of hispanic voter turnout between communities.  Yes There were lawsuits filed.  Yes there will be  testimony before the commission on hispanic concerns.  Will the 4 hispanic members have an impact?  I think so.  What will it be?  I guess we will see.

  51. A truce it is then. I do want to comment on one last thing, however, that I missed before.

    I think we will see an hispanic district in the San Diego area and it will be 70% hispanic.

    That’s impossible. You simply can’t map such a district unless the Latino percentage is much higher than the ACS projections. The absolute maximum you can get – including Imperial County – is the 64% seat on my last map. Without Imperial County you can’t even get to 60%.

  52. that it is impossible to get even a 60% hisp. seat in SD Co. when you don’t include Imperial — and even that percentage

    would entail a very convoluted district that the Commission will not draw … realistically only something in the mid-upper 50%-range is possible where the district maintains a semblance of compactness (there are indeed 100s of thousands of “extra” hispanics in SD Co. but they are spread out throughout the county) … I am not sure where the 70% hisp. number is coming from ???? (I thought we discussed this in a previous diary and everyone agreed that the 70% is an impossibility) …

  53. Thanks. CA is so big and daunting to attempt to map. Hopefully you plan to revise it in a few weeks when the Census data is out

    Re that article, I’m sure the GOP will still come out at least a couple seats ahead than if Dems with our trifecta (with big margins in the leg.) could have drawn up a gerrymandered map.

    GOP backing of Prop 20 or not had no impact on the outcome. The voters just had an opportunity to take away power from those effing politicians in Sacramento…

  54. Of course we’ll never know for sure, but I genuinely think that the Democrats would do another stand-pat map that protected all the incumbents – Democrat and Republican. I don’t think any of the California Democrats have the slightest interest toward reducing their victory margin, and some of them would have to do so if they were going to target Republicans. I mean, then they’d actually have to fly back and campaign for re-election, and that’s such a bother..

    And the GOP still holds enough of the trump card they had in 2001: tax or fee increases require 27 votes in the state senate and 54 votes in the state house.

Comments are closed.