How Democrats Fooled California’s Redistricting Commission

[muon2]

Erosity is like pornography - you know it when you see it. That is from SCOTUS (the comment being about porn).

No doubt. The problem is differentiating between use of man-made lines to gerry a result and natures own ragged edges. We know the difference, but it is hard, bordering on impossible, to establish a legal case based around compactness.

[muon2]

That 17th district is an interesting exercise, but it's remarkably ugly, and it doesn't need to exist. The Chinese in Cupertino and the Vietnamese in East San Jose have very little in common.

Do they tend to vote the same way?

[muon2]

There is another parameter I will probably use in So Cal from VRA cases. The Census Bureau has a disclaimer about use of CVAP for 2010 data since it is from an average of 2005-2009 statistics. That disclaimer was recognized by the CA commission in their report. In other circuits there has been recognition that 65% VAP can provide a controlling majority for a minority, and SCOTUS has accepted that from those circuits. If SCOTUS rules that the statistical nature of CVAP can't be relied upon, they may return to the supermajority 65% threshold. To play it safe, in areas where 50% CVAP may not be reachable, I will look for 65% VAP as a backup. For instance Anaheim/Santa Ana is one area where it will come into play for me.

I am not sure your comment has any nexus with mine, but OK. My Hispanic OC CD is 62.4% HVAP, and to hit 65% will either 1) be impossible, or 2) make an erose muni-chopping mess, and this for a CD which elects an Hispanic with no problem, and has despite determined, and well financed, runs at her. It's only 20% white, with the balance black (1.9%) or Asian (19.6%), the latter, particularly in this area, being a lower turnout group. The Commission's Hispanic OC CD is 60.88% HVAP btw.  So many little rules, so little time. Is Maldef bitching about this too?  Just curious.

I don't suspect any MALDEF issue there. They suggested a 63% HVAP district so you are in better shape than the commission. It's an area I will look at, but since I tend to look at compactness from the Roeck view, I'm probably more tolerant of erosity than you.

I don't know what "the Roeck view" means, but I assume it means, break every piece of china in the kitchen if it gets to the right number, based on some rogue court's "scrivinings" (I assume that you will enlighten me), but yes, as to erosity, definitely. You can hit 63% no doubt with an extra or two muni chop. You might sever the north end of Santa Ana (that is where the OC gentry used to live once upon a time), and then do two more mini-muni chops - a net of three more muni chops, although maybe just two, if there is enough Hispanic action left in Fullerton, which I chopped, to suck up a couple of very heavily Hispanic, low hanging fruit precincts, albeit kind of large ones, so you don't have to chop into both Tustin and Costa Mesa, or Placentia or something. And no, it won't happen in my map.

So this is my offering for Anaheim/Santa Ana. The block groups are imprecise, but the intent is to use all of Santa Ana and Stanton and use none of Orange, Buena Park or Costa Mesa. The Tustin piece will be my only muni chop into the Irvine district from the west and is needed for population equality. Anaheim has to be split due to its long eastern leg, but all parts are either in this district or an Orange-based district which wraps around the north. Likewise, the Fullerton split will be shared between this district and the Orange one. I confess to a Placentia split, but at least the part included in this district decreases erosity.

The HVAP is 65.0%.

[muon2]

So this is my offering for Anaheim/Santa Ana. The block groups are imprecise, but the intent is to use all of Santa Ana and Stanton and use none of Orange, Buena Park or Costa Mesa. The Tustin piece will be my only muni chop into the Irvine district from the west and is needed for population equality. Anaheim has to be split due to its long eastern leg, but all parts are either in this district or an Orange-based district which wraps around the north. Likewise, the Fullerton split will be shared between this district and the Orange one. I confess to a Placentia split, but at least the part included in this district decreases erosity.

The HVAP is 65.0%.

I would lose the Tustin chop and take the rest of west Anaheim, even if it reduces your HVAP percentage by a bit (it should not be that much).  Otherwise not bad from a chop standpoint. My CD did that, and went into Orange rather than Stanton and Placentia. You might be causing the Asian percentage in the north OC CD to drop some (my CA-40), as well as make CA-40 more erose of course. 

Garden Grove and Anaheim for this CD are "auto-chops" and really don't count as chops, so I consider that I did but one chop (Orange), plus the heist of those two precincts in Fullerton that were right next by and just so irresistible.

This is how I fit that district into my plan version of OC. As we've all agreed Imperial stays with SD so that defines the population need at the south end of OC. The main feature of this plan is that it preserves the LA/OC line. My algorithm drove me this way, but there was a lot of testimony before the commission to preserve that separation, too.

Hopefully my justification of the chop of Tustin becomes clear. My CD 48 stops before it reaches Orange, Santa Ana, or Costa Mesa, with population to spare. To bring CD 48 back into line I would have to cut into Tustin, Tustin Foothills, Irvine, or Newport Beach. If I have to chop, why not go where it helps a neighboring district, to wit CD 46.

It may be hard to see, but Cypress and La Palma are supposed to all be in CD 47 while Buena Park and all those pieces of Anaheim and Fullerton are in CD 45. You may notice that CD 45 takes the road less traveled to reach Los Serranos by way of Carbon Canyon Rd. I'll explain that more when I get to the Inland Empire.

[muon2]

OK. I am not in love which your cuts in OC. Tustin and Tustin foothills should be in your CA-45, and CA-45 should take Costa Mesa, with CA-47 moving up into Buena Park. In other words, twist the clock, clockwise around the Hispanic CD. I could explain why if you want, but I assume that you trust me that I know the lay of the land in OC at least.

And of course, my OC cuts are the best. Really. And Irvine is a good city to split, both from a geographical, and demographic, standpoint. The way, you get the wealthy beach zone, the Asian tinged zone, and the white inland more socially conservative zone, more cleanly defined.

I absolutely trust that you know OC better than I. However, it does us little good to draw maps from the same set of assumptions if we want to test the politics of the districts at the end. So, I'm letting myself be driven by geography more than sociology. That's what led to my version. My 48-49 line requires very little spillover across municipal lines and I can form nice compact districts for 47 and 48 that have very little in the may of split munis. CD 45 gets the leftovers.

As always if there are geographic items I've missed or other things that stay away from the squishy subject of communities of interest, I'm always game to refine my methods. Note that sbane's comments in Sacramento that led to a refinement have played out here as well. Mathematically there could be four districts entirely within OC, but my revised rule lets me back down from that to three as I keep the non-whole split pieces to two.

The rule also applies to SD county since it could have four districts, too. However, we want to link up Imperial and that leaves messy choices without the revision. This plan keeps three CDs entirely within the county. But my geographical treatment makes some changes.

In order to minimize municipal splits, CD 53 keeps Chula Vista whole (or would if I had block level controls) at the expense of the neighborhoods east of Balboa Park. That reduces the HVAP to 63.4 which is about 52.4% HCVAP, so it should be fine. I also let CD 50 follow the natural corridor from Escondido east to the Imperial line. That cuts CD 52 off on the east, which works well as it moves west to take up the parts of SD dropped by CD 53.

I had a choice for Encinitas vs Bonsall and Fallbrook. Either one could go with Escondido in CD 50 and the other would go with the coastal CD 49 into OC. You see my choice below based on my eyeballing the roads, but a reasonable argument could sway me the other way.

[muon2]

OK. I am not in love which your cuts in OC. Tustin and Tustin foothills should be in your CA-45, and CA-45 should take Costa Mesa, with CA-47 moving up into Buena Park. In other words, twist the clock, clockwise around the Hispanic CD. I could explain why if you want, but I assume that you trust me that I know the lay of the land in OC at least.

And of course, my OC cuts are the best. Really. And Irvine is a good city to split, both from a geographical, and demographic, standpoint. The way, you get the wealthy beach zone, the Asian tinged zone, and the white inland more socially conservative zone, more cleanly defined.

Addendum: Oh, and I would also cut Dana Point from CA-49 in exchange for taking more of Mission Viejo, for the reason stated in the above paragraph, and also because Laguna Niguel and Dana Point are so intertwined.

I'm somewhat averse to decisions based on the types of zones as they are usually proxies for specific political outcomes. I watched the Dems do that in IL last spring as they justified a whole host of gerrymanders on that type of logic. I recognize that VRA-based decisions also have political consequences, but that's in a different category for me.

To your addendum, it would work geographically to put Emerald Bay, Laguna Beach, and Laguna Niguel into CD 49 in exchange for the Rancho Santa Margarita area. It's nearly an even population swap and it doesn't impact the compactness of CD 48. Would that work for you?

[muon2]

OK. I am not in love which your cuts in OC. Tustin and Tustin foothills should be in your CA-45, and CA-45 should take Costa Mesa, with CA-47 moving up into Buena Park. In other words, twist the clock, clockwise around the Hispanic CD. I could explain why if you want, but I assume that you trust me that I know the lay of the land in OC at least.

And of course, my OC cuts are the best. Really. And Irvine is a good city to split, both from a geographical, and demographic, standpoint. The way, you get the wealthy beach zone, the Asian tinged zone, and the white inland more socially conservative zone, more cleanly defined.

Addendum: Oh, and I would also cut Dana Point from CA-49 in exchange for taking more of Mission Viejo, for the reason stated in the above paragraph, and also because Laguna Niguel and Dana Point are so intertwined.

I'm somewhat averse to decisions based on the types of zones as they are usually proxies for specific political outcomes. I watched the Dems do that in IL last spring as they justified a whole host of gerrymanders on that type of logic. I recognize that VRA-based decisions also have political consequences, but that's in a different category for me.

To your addendum, it would work geographically to put Emerald Bay, Laguna Beach, and Laguna Niguel into CD 49 in exchange for the Rancho Santa Margarita area. It's nearly an even population swap and it doesn't impact the compactness of CD 48. Would that work for you?

That sucks from a sociological standpoint, but is good from a geographic one, since there is an empty zone between Newport Beach and Laguna Beach.  In so much of CA freeways are the thing. They tend to define communities of interest. There is little reason for folks living inland from the 405 to cross it towards the beach, and visa versa.

I am sure that this communities of interest thing is abused for political purposes (of course it is!), but I assume that you believe me that I did not do that, and it is the commission's job to assess that, and evaluate the merits, and where a decision can reasonably go either way, probably go for the approach that makes for more competitive CD's. That is what I would do. Granted, I know some parts of the state better than others from an on the ground standpoint.

Anyway, in OC it makes no difference. All the CD's are safely GOP no matter how you draw them, after quarantining all those Hispanics in their little ghetto CD.

As I said earlier, I have no problem with your cuts, per se, and they do well for your goals. But one thing you mention in the comment above is competitiveness. In other states I have looked at, groupings primarily by socioeconomic and cultural factors tend to produce more homogeneous districts and hence less competitive ones. By using municipalities as substitutes for communities of interest, some research has found that more competitive districts can emerge. Certainly my application of that to MN on that thread recently ended up with a healthy blend of competitive and noncompetitive districts. But is CA like MN? I'm curious to see.

So back to OC. My hypothesis is that by following municipal lines I should get a natural balance of districts. Your plan split Mission Viejo right down the middle. I'd like to avoid splits entirely, but if needed just nibble a smaller piece of a town. That led me to try different arrangements like my map and the alternate I described. Sometimes that leaves flexibility, sometimes it works against it.

[muon2]

Yes, I understand sbane. I take your word for it about the black attitude (probably the second black incumbent is looking for a lifeboat, and was accommodated), I think the blacks are being short sighted about what their future is, but that is their problem. The Commission said they drew but one black CD in their text, but maybe something got lost in translation. And maybe Maldef will sue over this one too, assuming you have the numbers right. They should. This is one instance where I think they may have a pretty good case.

In the meantime, I am not changing my map, because of my own point of view about the legal exposure (I don't really think the CD looks all that bad myself - I have seen far worse racial gerrymanders), and because it doesn't matter for the purpose of my exercise, as you acknowledge. I can defend what I did without any embarrassment, if someone calls me on it.

What might help here is a Hispanic percentage map with your district's boundary overlayed with it. Just how Hispanic is that northwest extension, exactly?

The software does not allow that, but what I will do when I get home, is draw a "CD" that is just the NW extension of CA-33, and it will reveal the Hispanic percentage of that part of the CD.

By the way, Mike's map is really f'ing the Dems so far it looks like.  

And I haven't even looked at the political data. I'll run my usual fairness test after the maps are done to see if it's so. I'll have your political matrix as a cross check.

[muon2]

You have chopped central city San Diego to bits. You have a few precincts in San Diego there next to the Harbor filled with high income whites in condo towers.  Those should be in the Coronado, Pt. Loma, Ocean Beach, Pacific Beach CD. Why did you dislike my version of San Diego County (other than perhaps the Chula Vista chop, where I followed the commission's lines)?  

I spent some time looking at testimony about neighborhoods in central SD. I kept the neighborhoods together, but just attached them differently. After your comments I looked at MALDEF's border district, and saw that they also kept Chula Vista together and split the waterfront from areas east of I-805, so I assume they wouldn't sue over my map.

I noted your condo precinct in the file you sent, but it was boxed in by Hispanic areas and there was no bridge to connect to Coronado. I dislike water connections without bridges (or ferries) just like mountain connections without roads. I had to either lose the Hispanic areas by the bridge or add the condos to the border district.

On the east side of the county I look quite different, but I think that's because I didn't assume the historical district divide. The main east-west corridors are I-8 and CA-78. So I let CD-52 follow and I-8 until it ran into is the border district, and I let CD-50 follow CA-78. I won't object to a connection from El Cajon to Borrego Springs on someone's map, but getting there from Escondido looks much more natural.

I am sincere about advice for Encinitas. It's either part of a coastal district or it attaches to Escondido. I can make a case either way. It mirrors my choices at the other end of CD-49 in OC where I can hug the coast or shoot inland.

[muon2]

You may have kept neighborhoods together in central San Diego, but you tri-chopped central San Diego, appending part of it to the suburbs. I don't get that bit at all. What was wrong with my map in that respect? Just what about my CA-53 offends you, putting aside the trapping Hispanic precinct issue? Notice that I used the Miramar Marine Air Station as a natural barrier on its northern end.

Your map is fine, but I need to start with my assumption that Chula Vista stays together. That means that the area north of MLK and all of Bonita need to leave the Chula Vista district. If just move them in the natural way the downtown SD district is way too big and the El Cajon district is too small.

I noticed that you had chopped a piece of far northern SD onto CD 52 in what seemed like an unusual split. I had already determined that I wanted to try the CA-78 run east, so assembling all the areas between Escondido and Poway into the same district made sense. That meant shifting the CD-52 cut into SD from the northeastern part of the city down to the southeastern part of the city. I found that CD-52 became more compact as well with that shift.

That left the the downtown CD light on population. I didn't want to carve up the Mira Mesa neighborhood north of the base so I took that CD up the coast to Solana Beach. I could let the downtown CD pick up the area just east of Balboa Park and then drop the northerly boundary back to where you had it with the Escondido CD picking up the coastal communities. I do think it's worth thinking about the Escondido CD as an inland district, however.

[muon2]

Torie, I have an LA question on your CD 31 and 34. In the early copy you sent 31 had an HVAP of about 57% and 34 had over 75%. It looks like 31 will have to get over 60 or 61% HVAP to break 50% HCVAP. Did you look at a swap between 31 and 34 at any point?

[muon2]

And Lewis here are the stats for the NW quadrant of my CA-33 - 54% HVAP - with 210,658 people. And notice I minimized muni chops to boot, which I always try to do, absent a good reason not to. With more chops, I could have got it higher.

So not that Hispanic really. Population very much on the upper end of the range into which I'd have cautiously guessed though, which may be a problem in trying to cautiously reduce the erosity.
If I were trying to defend your map, I'd say that the extra Hispanic CD packed high enough to elect an Hispanic, being possible, was necessary and the Black pack just happened naturally.

Is there some grouping of communities now in your CD 37 that have at least 54% HVAP. If so, the district could violate Shaw, since a significantly more compact version would also exist that could elect a candidate of choice.

[muon2]

Torie, I have an LA question on your CD 31 and 34. In the early copy you sent 31 had an HVAP of about 57% and 34 had over 75%. It looks like 31 will have to get over 60 or 61% HVAP to break 50% HCVAP. Did you look at a swap between 31 and 34 at any point?

I got it up to 58.2% HVAP.  You could do a swap, but it would make the map a lot uglier, and violate natural lines and communities of interest. Is this HCVP 50% thing an absolute law, even if it is clear the CD will elect an Hispanic of their choice?  And the latter is only a Section 5 standard, not a Section 2 standard, correct, and Section 5 does not obtain here, or am I still confused on that one?

I doubt that CA-33 can get 210,000 54% HVAP out of the far north end of CA-37, but it deserves study. I took a look at it briefly the other day. Frankly it won't make it much less erose, since then it goes rather north from Long Beach, rather than doing an L thing.  It is the Carson thing that makes it look erose. I am not sure at the end of the day, it will fit any better into a circle with the same diameter, and all of that fun stuff.

An HCVAP of 50% is what the 9th circuit has established to satisfy the first Gingles condition for a majority Latino population for section 2. That usually works out to from 60-65% HVAP depending on the part of LAC. I would guess that as the income goes up, the needed percent comes down.

I looked at the CA-33 neighbors, and I see a solution. CD-38 is overpacked, and it can give up the South Gate side in exchange for Norwalk and La Mirada and still exceed 73% HVAP. If you add South Gate, Bellflower and the part of Carson south of I-405 you can get a 68% HVAP district. It does force CD-37 west of the Harbor, but the result is two VRA-compliant districts.

Similarly, a swap between CDs 31 and 34 will be needed as well. CD 31 only needs to get to 60% if it picks up East LA.

Edit:

Here's a version that has a 73.8% HVAP district on the west and 68.3% HVAP on the east. That keeps Bellflower and Norwalk together if it makes more sense that way.

[muon2]

Here's an updated map with all four south LAC Latino districts safely in compliance with section 2. I've drawn them for compactness, but Torie or sbane can say what pieces should be swapped.

East LA/Koreatown (yellow; goes up to the edge of Glendale and Pasadena): 64.5% HVAP
Downtown/South LA (green): 68.5% HVAP
South Gate/Paramount (peach): 75.3% HVAP
Downey/Norwalk (magenta): 67.5% HVAP.

[muon2]

Torie, I have an LA question on your CD 31 and 34. In the early copy you sent 31 had an HVAP of about 57% and 34 had over 75%. It looks like 31 will have to get over 60 or 61% HVAP to break 50% HCVAP. Did you look at a swap between 31 and 34 at any point?

I got it up to 58.2% HVAP.  You could do a swap, but it would make the map a lot uglier, and violate natural lines and communities of interest. Is this HCVP 50% thing an absolute law, even if it is clear the CD will elect an Hispanic of their choice?  And the latter is only a Section 5 standard, not a Section 2 standard, correct, and Section 5 does not obtain here, or am I still confused on that one?

I doubt that CA-33 can get 210,000 54% HVAP out of the far north end of CA-37, but it deserves study. I took a look at it briefly the other day. Frankly it won't make it much less erose, since then it goes rather north from Long Beach, rather than doing an L thing.  It is the Carson thing that makes it look erose. I am not sure at the end of the day, it will fit any better into a circle with the same diameter, and all of that fun stuff.

An HCVAP of 50% is what the 9th circuit has established to satisfy the first Gingles condition for a majority Latino population for section 2. That usually works out to from 60-65% HVAP depending on the part of LAC. I would guess that as the income goes up, the needed percent comes down.

I looked at the CA-33 neighbors, and I see a solution. CD-38 is overpacked, and it can give up the South Gate side in exchange for Norwalk and La Mirada and still exceed 73% HVAP. If you add South Gate, Bellflower and the part of Carson south of I-405 you can get a 68% HVAP district. It does force CD-37 west of the Harbor, but the result is two VRA-compliant districts.

Similarly, a swap between CDs 31 and 34 will be needed as well. CD 31 only needs to get to 60% if it picks up East LA.

Edit:

Here's a version that has a 73.8% HVAP district on the west and 68.3% HVAP on the east. That keeps Bellflower and Norwalk together if it makes more sense that way.

Is that 50% 9th circuit requirement an absolute one, or a safe harbor under Section 2?  What is the 9th Circuit case?   I guess I had better read it. I planned to review my map in this area anyway.

Here's what the commission cited in their report:

You must be logged in to read this quote.

[muon2]

So under the 9th Circuit, if you can't get to a compact 50% HCVAP CD, you don't have to draw one at all? 

That's what I was trying to say when Lewis was pushing for the Imperial-Coachella district. The commission did not cite section 2 compliance for CD 51 in their narrative. I assume it was because it is not compact enough at 50% HCVAP under the LULAC standards for common interests. They drew it based on common interests due to lying along the border.

I think that's also why they didn't do anything in Riverside even though they found racially polarized voting there. To link Perris or Riverside to SanB or Indio was not compact in the commission's analysis, so no action was required.

I'll put the San Jose connection to Salinas in that category, too. In other jurisdictions that might be required, but the commission is interpreting the ninth circuit to say it is not.

If a section 2 suit comes out of this, I think it will be in part because the commission has taken such a narrow view of both the 50% CVAP rule along with the compactness standards. SCOTUS has not had to rule on HCVAP, so it's ripe for litigation.

But in LAC there are plenty of ways to get to 50% HCVAP. I have 6 districts that meet that threshold, and MALDEF drew six as well in their official submission. However, the commission only drew 5, leaving a sixth at 49% while an adjacent district has HCVAP over 73%. In one part of their report they claim the 49% district is to satisfy section 2, but later they do not mention it as a reason for that district. I would not have left that 1% on the table in their position.

They also drew no Asian majority district in LAC, though it's clearly possible. They identified an AD for the San Gabriel Valley, but weren't inclined to link it over the hills to the east to make a CD. That seems to be another example of what the commission thinks compact means for section 2. There are courts that would not accept that view when the minority group shares the same media market and are in the same county (cf IL-04).

The more I look at their work the more I think this will find it's way to SCOTUS.

[muon2]

What is the HVAP percentage again for the San Diego-Imperial and OC districts that translates into 50% HCVAP again?  And I take it, under the 9th Circuit, there is no need to draw a 50% HVAP CD in Riverside. Is that correct? 

The commission CD 51 has 63.9% HVAP and they get 51.0% HCVAP. The MALDEF district is very similar except for the Chula Vista addition and they have 63.5% HVAP and get 52.5% HCVAP. How can that be, you may ask, since they are virtually the same district? I suspect this goes to the heart of the estimation aspect of CVAP, which are not part of the Census but estimated from prior years' data. The Circuits don't agree on its applicability, and I suspect that SCOTUS may not like using a number that is so imprecise on a hard and fast benchmark like the 50% rule. Any way, I would say that 63% HVAP would get you over 50% HCVAP in either case.

I don't think I've seen a specific ruling by the 9th Circuit that would indicate no district need be drawn in Riverside. The commission has said it would use CVAP instead of VAP for the 50% rule because the 9th Circuit says to. The commission is also making a judgement about what type of area is compact enough so as to apply the 50% rule, and I don't see that coming from the 9th. I can't find where the commission defines what constitutes a compact area, but based on their actions it's more compact than I would choose.

In my map I'll split the difference in Riverside and make a CD at 50% HVAP to satisfy the narrow reading of Bartlett. But I'll consider it an opportunity district, recognizing that it lacks the higher standard of 50% CVAP. I won't make the link to Indio because I think that would be two communities too far apart like in the TX LULAC case. I won't link Riverside to SanB because I've used SanB in a different section 2 district that I created because of other consistently applied redistricting principles, though I would consider them to be close enough to be in a compact area.

[muon2]

Returning to my plan, I want to talk numbers in the south region. It was divided into three subregions: Bakersfield with Kern. Kings, and Tulare counties; LA county; and Socal with Mono, Inyo, San Bernardino, Riverside, Orange, San Diego, and Imperial. The Bakersfield region has about 28K too many people, LA county is short 22K, and Socal is short 6K. If I use eastern Kern to give up population to both subregions that are short I end up with a four-way split of Kern and no CD entirely within the county due to the VRA district.

That means I either put 28K into LA or into Socal from Kern. I wanted to add population to LA from the Chino Hills so it makes more sense to shift the 28K into Socal from Kern. That chop perfectly covers Ridgecrest in the NE corner, and results in the following map for the Bakersfield subregion. The VRA district is the one I generated earlier and is 65.4% HVAP.

[muon2]

The next part of the numbers saga of my algorithm brings me to the population shift between LA and San B counties. I need to have a net 22 K go to LA, which can be done with a fragment that size, or with a larger fragment and a piece of LA county attaching to SanB in return.

The natural piece to go to LA is the Chino Hills. It fits well with the Asian majority district stretching from the San Gabriel valley, and it's cut off from the rest of the county by the necessary VRA district based in Ontario. The total cut off fragment is about 75 K population so if I use all of it in LAC, then I need to swing 53 K from LAC back to San B. I found three choices for that 53 K: in the north, at Claremont, or at Pomona. The north option required an ugly chop into Lancaster/Palmdale and could be ruled out. Using Pomona left Claremont and its neighbors isolated by the Covina VRA district and could only be connected by the mountains or by chopping a bunch of towns to follow a road along the foothills. It turns out that without the Chino Hills fragment there is the exact population for 2 CDs south of the San Bernardino mountains, and if 53 K is added by jutting into Claremont, then 53 K from south of the mountains has to added to the Cd on the north side, and that isn't particularly attrractive if there s another alternative.

However OC needed 46 K to complete its CDs and that's not very different from 55 K. So the Chino Hills are split into two with one part going to LAC and the other to OC. But I don't have to link Victorville to a piece of San Bernardino.

An Ontario to San B district has 65.5% HVAP, leaving the foothills in the other district. The only thing left in San B is to complete the high desert district. Here I can use Blythe linked to Needles, plus Desert Hot Springs and nearby areas north of I-10 with a good link to Yucca Valley.

In Riverside it would be nice to make a district that exceeds 50% HCVAP, but I've used the likely areas in San B to link, and the Coachella Valley is too far with low HVAP areas in between. Instead I've made CD 42 an opportunity district with 54.7% HVAP. The connection to Corona is not pretty, but was needed to get over 50%. If there is another way to get 50% HVAP there, I'm open to suggestions. The other two districts also fit entirely within Riverside county.

[muon2]

Here's an updated map with all four south LAC Latino districts safely in compliance with section 2. I've drawn them for compactness, but Torie or sbane can say what pieces should be swapped.

East LA/Koreatown (yellow; goes up to the edge of Glendale and Pasadena): 64.5% HVAP
Downtown/South LA (green): 68.5% HVAP
South Gate/Paramount (peach): 75.3% HVAP
Downey/Norwalk (magenta): 67.5% HVAP.

Wait, and what happens to the corridor here?

It has to be divided up between the other districts. In this map the Long Beach district is way short of population and could move west to pick up the corridor, or it could be given to the Torrance district and have the Long beach district encroach on the Palos Verdes area. Since it's not a VRA area, the mapper has choices.

In general the VRA will force a densely populated minority area (say 80-90%) to spread in multiple directions when multiple districts are required. There are other possibilities for those directions than in my map, such as going into OC and running east to the Anaheim district. The only direction off the table is northeast since that goes into the Asian majority district.

[muon2]

Wait a minute.  The 50% HVAP rule that is an interpretation of Bartlett, rejected by the 9th circuit, which rejection the Commission sometimes followed, but not in Riverside.  Isn't that inconsistent?  Is there some case law, that you have to draw a 50% HVAP CD if it is compact enough, or a 50% HCVAP CD, if it is compact enough, whichever is greater? 

The reason I ask, is that getting Riverside up to 50% HVAP violates other appropriate rules of redistricting in my view (at least in my map), and therefore, if no governing case law really supports the whichever is greater test, in my opinion, it is the Pubs job on the Commission to vote against a 50% HVAP CD in Riverside, if it creates more erosity, and hurts their case, which is does here.

I don't know that the 50% HVAP rule has been considered by the 9th Circuit since Bartlett came down in 2009. Bartlett says VAP, but declined to say if CVAP is needed, because they didn't need to to decide the NC case. If I was a minority group that could have a 50% VAP district, but didn't get one, I'd be in court claiming Bartlett gives me standing to sue under section 2. I'd say that CVAP is fine to see how likely it is that the minority group could control the election (as in LULAC), but I should get an opportunity district in any case, even if below 50% CVAP. Bartlett says that the state may create of crossover district where some of the white majority join with the minority, and such districts may enhance minority voting rights.

[muon2]

The Long Beach district is the only undersized district?

The Black-plurality district was also slightly undersized in that map. Here's my embedding of those districts in a complete LA county map. This follows from my algorithm that placed 13 whole districts in the county, then tried to minimize municipal splits after satisfying the VRA. A 14th CD (32) adds only 22K from outside the county.

Here are the stats for the districts that would have a majority minority CVAP:
The VAPs for minority districts are as follows:
CD 26: 63.3% H
CD 29: 61.3% H
CD 31: 64.5% H
CD 32: 52.0% A
CD 33: 68.5% H
CD 34: 46.9% B
CD 35: 67.5% H
CD 36: 75.3% H
I note that one legal vulnerability is that the Black groups that testified did not want a majority-minority district and preferred to be split to maintain their opportunity for two districts, and the commission agreed with them. Despite their threats that a packed district could be a section 2 violation, I could find no basis for that claim, and preferred to keep the neighboring Latino districts comfortably above 50% CVAP where they could elect candidates of choice.

Here is a big picture map (CD 32 crosses into SanB but it doesn't show), and an enlargement with municipal lines (where CD 32 shows correctly).

[muon2]

Wait a minute.  The 50% HVAP rule that is an interpretation of Bartlett, rejected by the 9th circuit, which rejection the Commission sometimes followed, but not in Riverside.  Isn't that inconsistent?  Is there some case law, that you have to draw a 50% HVAP CD if it is compact enough, or a 50% HCVAP CD, if it is compact enough, whichever is greater? 

The reason I ask, is that getting Riverside up to 50% HVAP violates other appropriate rules of redistricting in my view (at least in my map), and therefore, if no governing case law really supports the whichever is greater test, in my opinion, it is the Pubs job on the Commission to vote against a 50% HVAP CD in Riverside, if it creates more erosity, and hurts their case, which is does here.

I don't know that the 50% HVAP rule has been considered by the 9th Circuit since Bartlett came down in 2009. Bartlett says VAP, but declined to say if CVAP is needed, because they didn't need to to decide the NC case. If I was a minority group that could have a 50% VAP district, but didn't get one, I'd be in court claiming Bartlett gives me standing to sue under section 2. I'd say that CVAP is fine to see how likely it is that the minority group could control the election (as in LULAC), but I should get an opportunity district in any case, even if below 50% CVAP. Bartlett says that the state may create of crossover district where some of the white majority join with the minority, and such districts may enhance minority voting rights.

"May" is not the same as "shall."  Sorry to keep bothering you Mike. I really need to read both cases myself. Does enhancing "minority voting rights" mean electing a white Democrat, because in Riverside, it is not going to be that white Dems help to elect a Dem Hispanic.

Here is the operative quote from Bartlett, it also appears in the commission report.

You must be logged in to read this quote.

In my plan, I would note the creation of a >50% HVAP Riverside district as a mitigating factor that I did not make a majority CVAP district in Riverside, due to other rational redistricting principles. That may be the commission's thinking as well. In fact If you like the shape of their CD-41 better than my CD-42 I can swap to that shape.

I think that same language can be used to defend not splitting the Black population between two districts. It's clearly discretionary, and there is no other section 2 challenge related to that minority group.

Lots of chops there with your Asian CD Mike.

I wasn't particularly pleased with that part of the map, but I can shave the AVAP down in 32 somewhat and perhaps remove a chop, if you like.

[muon2]

What is the HVAP number for that 40th district? I think putting Yucaipa into that district really lowers the Hispanic %, but I could be wrong. I think splitting the SBD-Riverside line in the exurban areas is probably a better idea than in the desert. That way you can avoid chopping into LA county and deliver two 50% HVAP districts in SBD and Riverside in addition to the 65% HVAP. Might be worth playing with. I of course hold that the choice I made in including Pomona in the really Hispanic district is the correct one, but not feasible with your algorithm.

The HVAP for CD-40 is only 36.6% so the HCVAP is probably around 30%. Moving Yucaipa isn't going to get it anywhere near 50%. I already have a fully compliant section 2 district in the county as CD-41, so there is no obligation to try to form CD-40 into an opportunity district.

The problem for me wasn't where to split the SBD-Riverside line, but how to cross the SBD mountains from the north. I don't like crossing them just to grab some population and the piece I grab seems out of place compared to the rest of the high desert district. Keeping more of the eastern desert together seems like a much better fit. My move makes CD-43 more compact as well without changing CD-39 compactness much.

[muon2]

After I read the cases, if I am satisfied that a 50% HVAP Riverside CD need not be drawn as a legal imperative, then if to draw it involves as it appears more erosity, it is my opinion that the GOP members should have vetoed it, and that is how I would vote. In other words, in my mind, Riverside is back in play. I might withdraw my gift to the Dems here.   You do the minimum to satisfy the VRA if it otherwise violates appropriate redistricting principles, if it hurts your party.  To me that is rather basic.


My Asian tiger CD has no muni chops at all.

... except for Montebello.

In the meantime I'll see how smooth I can make a Riverside district. I'm still concerned that Riverside meets the Gingles test and something will have to be done there to answer to the VRA.