How Democrats Fooled California’s Redistricting Commission

[muon2]

Well, the dirty deed is done, and like Lady Macbeth, I have nightmares even while awake that I many never be able to cleanse the noisome stain from my sanguinary deed off of my hands. I really hated to do this. The map of Kern and Tulare is truly disgusting. CA-21 is 66.2% HVAP, and I got CA-20 up to 60.0% HVAP by nipping it into Merced County (to grab Dos Palos), while CA-21 in turn nips into Fresno County to pick up a few rather Anglo precincts (in lieu of Coalinga).  

Is everyone "happy" with the map now? Any more comments? If not, we shall commence to prepare the matrix grid charts.

I'm afraid I'm still worried about the CV. Merced is one of the section 5 counties and it seems like you've put all of it in an Anglo district. I think you'll have to link Merced to the Hispanic portion of Fresno to solve both Merced's section 5 issues as well as Fresno's section 2 situation. 61.5% HVAP gets it to a safe legal status. The commission thinks you'd be fine with anything over 50% HVAP, but MALDEF disagrees.

The best plan with the least splits is to take Merced and link it to Madera city and only split Fresno city. My version of that has 62.0% HVAP.

[Torie]

Why in this context is 61.5% HVAP "safe?" You are deleting from the Fresno Hispanic CD a bunch of Hispanic precincts, and then going to Merced County to make them up. Under Section 5 (which may not be long for this world, having read the transcript of the Perez oral arguments, but I digress), Merced Hispanics need to be represented by Hispanics, even at the cost of Fresno Hispanics being represented by an Anglo, and no additional Hispanic CD is created?  For what it is worth, the old Merced based CD per the Almanac of American Politics was 46.7% Hispanic population, so not remotely in an Hispanic CD.

In looking at the Commission numbers, where in their map Merced is appended to Fresno Hispanics like you did, I see that their Fresno Hispanic based CD, it is 58.01% Hispanic population, while my Fresno based Hispanic CD, which goes south rather than north, is 64.6% Hispanic population. (The numbers in the southern valley Hispanic CD are almost the same as mine (my CA-21), 70.6% Hispanic population, versus 71.3% Hispanic population for mine.) So are we supposed to substantially dilute an Hispanic CD so that it takes in Merced per Section 5, or chop Merced along with everything else (which the Commission didn't really do), shoving some Fresno Hispanics into an Anglo CD?

Addendum:  I "see" that there are but a handful of "trapped" Hispanic precincts in Merced in any event. It is just not all that Hispanic really.

[muon2]

Why in this context is 61.5% HVAP "safe?" You are deleting from the Fresno Hispanic CD a bunch of Hispanic precincts, and then going to Merced County to make them up. Under Section 5 (which may not be long for this world, having read the transcript of the Perez oral arguments, but I digress), Merced Hispanics need to be represented by Hispanics, even at the cost of Fresno Hispanics being represented by an Anglo, and no additional Hispanic CD is created?  For what it is worth, the old Merced based CD per the Almanac of American Politics was 46.7% Hispanic population, so not remotely in an Hispanic CD.

In looking at the Commission numbers, where in their map Merced is appended to Fresno Hispanics like you did, I see that there CD is 58.01% Hispanic population, while my Fresno based Hispanic CD, which goes south rather than north, is 64.6% Hispanic population. The numbers in the southern valley Hispanic CD are almost the same as mine (my CA-21), 70.6% Hispanic population, while mine was 71.3% Hispanic population. So are we supposed to substantially dilute an Hispanic CD so that it takes in Merced per Section 5, or chop Merced along with everything else (which the Commission didn't really do), shoving some Fresno Hispanics into an Anglo CD?

The idea that it might be legally safe was by using the HVAP to HCVAP numbers from the MALDEF plan. I concluded that 61.5% HVAP results in 50.0% HCVAP. That's going to be fine for the polarized voting in Fresno under section 2.

Merced is now in a district that is 47.2% HVAP according to the 2010 Census. Reducing that number would be a red flag for section 5. The commission chose to up it to 52.9%, but also placed the Hispanics in southern Fresno county in the Kings district with a 65.9% HVAP. It follows the similar pattern as the current districts and I presume is intended to avoid any vote dilution claims.

MALDEF advocated for a more aggressive plan which put Merced with central Modesto and Stockton at 48.5% HVAP. That covers section 5 there, and left room for two additional CV districts that exceeded 50% HCVAP - one with Madera and Fresno and one with Kings and Bakersfield.

You see that both plans decided to increase the HVAP for Merced to meet section 5 and both got an opportunity district for at least part of the Fresno Hispanic population to meet section 2. Given the shape of you map, I thought your easiest out was to meet both standards in a single district that represents an improvement on the 2000 map for the populations concerned.

I see a similar type of thinking has gone into section 5 concerns in Monterey. The current district is 44.2% HVAP and the commission made sure to exceed that in their district. Their commentary makes clear that they even took a slice of Gilroy, splitting the community, just to make the new district 44.4% HVAP.

[Torie]

Thank you for your comments, Mike. They were quite helpful.

Does this map make you happier Mike?  CA-20 is now 63.2% HVAP, with the Hispanic Merced folks now "liberated."  I strongly doubt that in exchange for eschewing  an extra 3 Hispanic points, that keeping Merced whole and not shoving a bunch of Fresno Hispanics into an Anglo CD in Fresno, and making a further hash of the map (including slashing municipalities here, there and everywhere), is required under Section 5 (or that the DOJ would demand it), but I understand, it is not the Commission's job to take a legal risk that is more than remote. I am also reasonably confident that SCOTUS will rule that at the end of the day, Section 5 does not trump Section 2, and whether the map is ultimately found legal, will turn on Section 2, with the role of the DOJ considerably more truncated procedurally. The conservative 5 seem to be going in that direction. But that remains to be seen.

[muon2]

Thank you for your comments, Mike. They were quite helpful.

Does this map make you happier Mike?  CA-20 is now 63.2% HVAP, with the Hispanic Merced folks now "liberated."  I strongly doubt that in exchange for eschewing  an extra 3 Hispanic points, that keeping Merced whole and not shoving a bunch of Fresno Hispanics into an Anglo CD in Fresno, and making a further hash of the map (including slashing municipalities here, there and everywhere), is required under Section 5 (or that the DOJ would demand it), but I understand, it is not the Commission's job to take a legal risk that is more than remote. I am reasonably confident that SCOTUS will rule that at the end of the day, Section 5 does not trump Section 2, and whether the map is ultimately found legal, will turn on Section 2, with the role of the DOJ considerably more truncated procedurally. The conservative 5 seem to be going in that direction. But that remains to be seen.

I'm with you on section 5. I'm watching the TX case as that may put it to bed. But, until then, ...

Now that I have the software running, it seems only fair that I put an offering out for criticism as well. As you may have noted I think that commissions need constraints, and that a generic community of interest is not sufficient. I also think that the rules should be out before the map. So here's what I would use.

Districts must have equal population which means that they must be within 500 on the DRA.
The VRA must be followed as currently construed.
County splits should be minimized.
    -Counties larger than a district should have as many whole districts as possible.
    -Split counties should have no more than two districts not counting districts entirely within the county.
    -Smaller split counties should preserve the majority of the county in one district.
    -County splits should minimize municipal splits.
    -County splits should not reduce compactness of districts, nor create bizarrely shaped districts.

I have an algorithm that furthers the above goals.
    -Divide the state into regions of whole counties that are each within 0.5% of a whole number of districts.
    -Divide regions into subregions of whole counties that are each within 5% of a whole number of counties.
    -Divide each subregion into districts.

[Torie]

You are going to draw an entirely new map of CA Mike?  Man, that will be a project, but I salute you!

Please use screen shots, with municipal lines, and voting districts not hidden, so I can read and evaluate your maps more easily. I sometimes have trouble with that.

One other thing. In CA, geography/topography/water matters as much as county lines, and I suspect that if you follow your algorithm religiously, that your map will be flawed, and not really hew to communities of interest. Subjectivity cannot be entirely exorcised from this exercise. So a computer will not be able to draw your map after your input your constraints as it were. I would also keep Imperial appended to Hispanic San Diego, so we can get a better comparison with the Commission's map. Where it made a major decision, affecting the whole map, that is reasonable, I think it should be followed.

And yes, the south Central Valley part of my map sucks (it is not as if I spent hours refining it - I was just trying to "fulfill" your "demands" for purposes of discussion. ). The Commission's version is probably better, although you don't like the lower Hispanic percentage in the Fresno Hispanic CD.

As to Fresno Hispanic percentage thing, has the DOJ signed off on the Commission's map, and is there any litigation pending? If with hindsight, it did not generate litigation, I think the lower percentage all other things being equal or actually in favor of doing so, should be accepted. The Commission's assessment of the legal risks proved correct in this instance.

[minionofmidas]

You must be logged in to read this quote.

Incredible what a difference in appearance one precinct can make sometimes. This is entirely superficial, but using the more southerly connector precinct looks so much worse.

[muon2]

You must be logged in to read this quote.

Incredible what a difference in appearance one precinct can make sometimes. This is entirely superficial, but using the more southerly connector precinct looks so much worse.

That's part of why I drew it the way I did. Once I had achieved my goal for the VRA, I didn't push it. For example I didn't cut into Visalia, because it wasn't needed to create the ability district.

[muon2]

You are going to draw an entirely new map of CA Mike?  Man, that will be a project, but I salute you!

Please use screen shots, with municipal lines, and voting districts not hidden, so I can read and evaluate your maps more easily. I sometimes have trouble with that.

One other thing. In CA, geography/topography/water matters as much as county lines, and I suspect that if you follow your algorithm religiously, that your map will be flawed, and not really hew to communities of interest. Subjectivity cannot be entirely exorcised from this exercise. So a computer will not be able to draw your map after your input your constraints as it were. I would also keep Imperial appended to Hispanic San Diego, so we can get a better comparison with the Commission's map. Where it made a major decision, affecting the whole map, that is reasonable, I think it should be followed.

I overlooked my rules for contiguity which addresses some of the mountain and water issues.

Whole counties at the regional and subregional stages are considered contiguous if their county seats are connected by numbered state and federal roads without crossing into another county. Split county pieces are contiguous if the census block groups are contiguous by local roads and not point contiguous. A road that makes a border of a block district can provide contiguity for areas on either side of the road.

You must be logged in to read this quote.

According to the commission the maps are still under DOJ review. After the DOJ issues an opinion groups will decide if they want to file VRA litigation. Until then I as a mapper remain cautious.

[Torie]

Are you using the block group DRA data base or the voting district data base? If the former, our maps will not "match."  What per your criteria prevents the north coastal CD from zipping into the Central Valley, which if you do, will just not fly no matter what other constraints you are using. That is an example of my skepticism that such an approach will prove workable. And sometimes doing an extra chop or two serves the greater good (beyond feeding the VRA monster).

Presumably the requirement that the map get votes from both parties constrains play partisan games, assuming the members are not flying under false partisan flags, and are reasonably competent (the charge here being that the Pub members were not).

Oh, and don't forget the class warfare angle, ala what evolved in my map in Silicon Valley, and how I drew the lines of the OC Gold Coast and LA westside CD's. Money is indeed a community of interest in California.

Are you also going to trash all my LA County CD's?  

[muon2]

Are you using the block group DRA data base or the voting district data base? If the former, our maps will not "match."  What per your criteria prevents the north coastal CD from zipping into the Central Valley, which if you do, will just not fly no matter what other constraints you are using. That is an example of my skepticism that such an approach will prove workable. And sometimes doing an extra chop or two serves the greater good (beyond feeding the VRA monster). Presumably the requirement that the map get votes from both parties constrains play partisan games, assuming the members are not flying under false partisan flags, and are reasonably competent (the charge here being that the Pub members were not). Are you also going to trash all my LA County CD's? 

I'm using the block groups, in part because that was what you originally had sent me for socal. They also tend to be smaller units than voting districts.

The northern end is an interesting problem. Reading the debate here and looking at past maps and public submissions leaves me with multiple opinions. If the Coast Range must be crossed the Humboldt to Shasta link seems as valid as one from Mendocino to Colusa. I can find advocates for both among Californians, and the commission didn't make the crossing. I'm trying to find proxies for communities of interest and taking sides on that one is not where I want to go in advance.

That same thinking applies to the cut of SF. I'm not going to preclude a link across the Golden Gate if that's where the path leads. In LA I have the advantage of the aforementioned districts from your file, and since they are generally within one county, made of whole municipalities and not bizarrely shaped, I expect that I will find myself drawing them in a similar way.

[Torie]

How did you manage to get the CA DRA software to start working for you?  I was never able to load the block group data myself.  Hey, there is one Hispanic voting district in San Jose with 70,000 people.

[muon2]

How did you manage to get the CA DRA software to start working for you?  I was never able to load the block group data myself.  Hey, there is one Hispanic voting district in San Jose with 70,000 people.

You must have had it working at one time, since the file you sent just after Christmas for SoCal loaded as a block group file. In any case, I found that part of the problem is finding the right wifi location. Once I found one where it would load, it's been fine at other locations that previously wouldn't load. I assume that it has to do with the files that go onto the specific computer. Of course I still save frequently as the map will freeze at inopportune times forcing a reload.

[muon2]

So here's the first step in my algorithm.

This first map creates four regions where each is within 1000 persons of a whole number of districts.
Northern: 10 districts (+940)
North Coast: 4 districts (-274)
Southern: 18 districts (-408)
South Coast: 21 districts (-257)



Very small shifts between regions in a single county are all that are needed to achieve exact equality. But, looking at this map, I fear a section 5 problem in Monterey. It would be largely combined with SLO which would almost exactly be a whole district, but the Hispanic percentage would drop compared with the current district. Courts might rule that its OK, but the commission map suggests that legal advice is to not diminish the Hispanic percentage. As pretty as the numbers are for this split, I'm going to head for a second plan.

As with the first plan I start with LA because it is the largest county. Picking up the southern counties leads to a 3-way split. The deviations are a bit larger, but still within the 0.5% rule. Monterey is in the same region as the other counties currently used to satisfy section 5, so there's good reason to believe that it will hold up through the rest of algorithm. I'm sure that Torie won't like the crossing into the northern CV, but given my rule about use of the VRA leads me to this plan. Of course if someone else has another region split to suggest I'd be happy to consider it.

Northern: 8 districts (-3178)
Coastal: 14 districts (+2158)
South Coast: 31 districts (-1019)

[Torie]

I don't think you are going to get any vote but yours for a plan that sets up that cut into the north central valley from the coast. You might rethink that. As I said, your constraints are just too tight.

The other regional map seems to allow Inyo and Mono to be appended to some central valley CD. That isn't going to happen either. They could be appended to the Anglo portion of a Kern based CD, but only in a pinch, and there had better be a damn good reason. That resort area is tied to the hip to Socal, not the Bakersfield dump, via that twisty road over the Tehachapis. Granted, we are only talking about 17,000 folks. Alpine's population is mostly centered on the west slope of of the mountains, but then it has almost no people. You could chop that county however, since no roads unite it, assuming it has more than one precinct.

As to your idea to remain flexible about crossing the Golden Gate, you will have an uphill battle convincing anyone to chop SF in half or something like that. It isn't going to happen. Now if SF were smaller than one CD, then going either north or south to pick up population would be OK.

Are all scientists as stubborn as you?  

In any event, the idea is to evaluate whether the Commission's choices were reasonable, and in particular whether the Pubs on it were really silly buggered by the Dems' gamesmanship, not if it met some Mike test about minimizing chops. The Commission is not a court subject to implementing some kind of Michigan law.

Good luck!  I have this vision, that if we were both Pubs on the Commission, we would be doing most of the talking, and fighting all the time, while the rest just sat back and enjoyed it all.

[krazen1211]

The South Carolina GOP has hired Paul Clement to take down Section 5 for good.

[Torie]

You must be logged in to read this quote.

As to this bit, I have no quarrel at all. It is entirely appropriate. As I said, I drew that part of the map on the fly, without refining it.

[muon2]

I don't think you are going to get any vote but yours for a plan that sets up that cut into the north central valley from the coast. You might rethink that. As I said, your constraints are just too tight.

The commission didn't go for a Humboldt to Shasta crossing and links from Lake to Colusa do exist in their version. As I noted other public submissions made the link all the way across Lake. I'm not convinced your plan gets more votes, but you can keep trying.

You must be logged in to read this quote.

I could shift Alpine and stay within my 0.5% rule. I just couldn't resist the near equality it gave in that location. And shouldn't I try for more equality? I haven't gone the next step to see where the Owen's valley would attach in that plan. Remember, I'm a lot less concerned with arbitrarily defined communities of interest, since I think they can be played by either side. Any given decade we'll one side or the other try to take advantage, why not reduce the chance?

You must be logged in to read this quote.

Clearly one of the problems with the 4-region split is that it would prevent whole districts from being in either SF or San Mateo, and that would also lead me away from that plan. However my VRA constraint takes me to the 3-region plan, so it's not an issue. If I keep one CD in SF then we chop the same amount, so it's just a question of where.

You must be logged in to read this quote.

Hypothesis, experiment, analysis, refine and repeat.

[Torie]

One other thing occurs to me. LA gets the bulk of its water from the Owens Valley. To append Owens to water competing Kern, is just totally unacceptable for that reason alone.

I don't think the Commission map impinged on the north central valley from the coast. In any event, the Commission eschewing that seems entirely reasonable to me, and to criticize them for not crossing the mountains to an area with which the north coast has nothing in common, seems entirely reasonable to me. The Central Valley has its own unique little issues and problems. So giving them demerits for taking that approach is just strange.

And chopping SF in half is just a cardinal sin. A chop in one place, just isn't the same as another. Frisco is a state of mind. They have a right to elect their own zany congressperson, without dilution from more prosaic types. Ditto Oakland and Berkeley. Just don't mess with them. Nobody in Sonoma
 will really mind if they are chopped (at least the way I did it), just because that is the way the map flowed. They don't give a damn. Really.

As to the matter, that we use these rigid tests to protect ourselves from a just go wild AZ type Commission, which while OK maybe for Iowa, certainly are not for CA, the answer to that, is that the law requires that any map be approved by at least one vote from each party. That is the check. Please respect it. I might add, from what I have seen, that with two or three potential exceptions, the Commission did in fact do a reasonable job, and that any partisan games were rather minor (and that due to the Pub establishment, I think (rather than a Pub Commission member flying under false colors), because they are dysfunctional and useless in general, and in this case just not being interested in what the Commission was doing until it was too late, maybe). That however is a tentative impression.

[muon2]

One other thing occurs to me. LA gets the bulk of its water from the Owens Valley. To append Owens to water competing Kern, is just totally unacceptable for that reason alone. I don't think the Commission map impinged on the north central valley from the coast. In any event, the Commission eschewing that seems entirely reasonable to me, and to criticize them for not crossing the mountains to an area with which the north coast has nothing in common, seems entirely reasonable to me. The Central Valley has its own unique little issues and problems. So giving them demerits for taking that approach is just strange.

I'm not criticizing them, I'm just trying to establish a consistent reason to prefer one route over the Coast Range over the other. Failing that I leave both options equally open.

The 2011 commission map does not provide much guidance to answer my query, though they do extend west from the Valley to Lake. The 2001 Congressional plan links Humboldt to Yolo, while the 2001 Senate plan attaches Del Norte and Shasta to the Valley as far south as Sutter. That leads me to the conclusion that some Californians are not opposed to a link from the coast to the Valley. Therefore both options remain on the table for me.

[Torie]

No, the Central Valley that is otherwise whole slipping into the mountains (heck I did that in a minor way), is entirely different than a coastal based CD poaching into the Central Valley. The mountains themselves can go either way really. But the way I drew the map, the north coast CD was desperate for population, so it took all of the mountains (plus the Napa Valley), except those in the NE corner of the state.

[muon2]

No, the Central Valley that is otherwise whole slipping into the mountains (heck I did that in a minor way), is entirely different than a coastal based CD poaching into the Central Valley. The mountains themselves can go either way really. But the way I drew the map, the north coast CD was desperate for population, so it took all of the mountains (plus the Napa Valley), except those in the NE corner of the state.

I'm not objecting to the choice you made. Your justification for it is completely rational and it makes for a reasonable map. I'm saying there are other bases to make a choice and I don't see a neutral way of distinguishing them.

Let me again cite the current map. Someone drew CA-1 for some reason, and it is a coastal district that poaches the Central Valley around Davis. I'll concede that partisan considerations probably drove that choice, but it's a reason. The point is that the mountains are not the kind of psychological barrier that say the York-Lancaster line is in PA, preserved despite overt partisanship.

Another example is the submitted MALDEF plan which I've studied to understand their take on the VRA. The VRA doesn't generally figure into Norcal, but it's included as part of the statewide plan. In that plan county lines are generally followed throughout Norcal, but they expand the current reach from the coast into the CV for CA-1. The adherence to county lines for them was rational, but led to a different choice linking the coast and CV.

To me this a lot like the debate on this site a year ago about WA. The problem was the new WA seat and the subsequent need for a crossing to pick up the excess population east of the Cascades. Most argued against the type of crossing eventually adopted for various community of interest reasons, yet that's what their commission drew.

[minionofmidas]

I'm not sure that Yolo fully counts as Central Valley for these considerations. And Torie poached it for a North Bay district, which really was functionally the same thing as what happened in 2000 but made for a nicer-looking map. (He also shifted all the Sacramento area districts westwards compared to the current map, as a consequence of using Redding - of using all of Redding, that is - which also enabled him to use the Solano-Napa county line. I don't know if this was the process by which he arrived at the map, but that is one purely rational reasonable way in which I might have come to draw the same map that he did.)
The counties to the north, especially without the areas in between on paths that people actually use, are a different matter.

[Torie]

If there is more than one reasonable choice as to how the draw the map, where it is a reasonably close decision, and particularly if it does not affect whether a CD is in real partisan play or not, then if the Commission made one choice, and I made another, the Commission (and the Pubs on it in particular), cannot really be criticized. That is the purpose of this exercise for me. If Mike wants to draw a map following his own metrics, and see what happens, fine, but that is a different exercise, and I am not quite sure what its relevance is, other than to see what a law incorporating his metrics would cause a map to look like.

If you don't chop SF, and you follow the class warfare theme (which I actually think is kind of appropriate), then you draw the Contra Costa CD the way I drew it, appending the NW corner to Solano. Solano is a transitional county between SF and the Central Valley. If the Contra Costa CD goes all the way to the Bay to take the downmarket precincts there, then the San Joaquin CD will cut more deeply into Contra Costa (not desirable), or you do the silly chop of Martinez (white middle class with no bridge, that the Commission did I see). And then the Solano CD needed more population,  and the only way to get it was to go into Yolo, with which Solano has a lot of common, and no mountains are crossed (with only Woodland really being agricultural, and Davis of course an academic node). So that is where the "poach" should be I think, to the extent it is a poach.

Finally, Napa has a lot in common with Sonoma, although it could be appended to Marin or Solano as well, but then you are back to a poach crossing the mountains into the Central Valley, and CA-01 gets ever larger, or you get a nasty chop of Santa Rosa, and that dog won't hunt. I tried hard, very hard, to avoid nasty chops of significant municipalities.

So as a Pub I would insist on this approach, if alternative approaches that I consider less desirable, hurt my party's chances. That is my job on the Commission, and should have been the job of the Pubs on it. Did the Pubs do their job on the Commission or not?  And where they didn't, we should find out why. Was it due to shill testimony, or incompetence, or did they have a rationale that seems reasonable, that we shall know better, when we compare the two maps and the data, or even better yet, their explanation.

And whatever happened to the concept of competitive districts?  Isn't that supposed to be desirable too, all other things being reasonable equal, particularly if it will make some members of both parties have to work a bit harder?

I will clean up the south central valley, now that I "know" the percentage HVAP benchmark percentages, as to which I take Mike's word are accurate. Not taking on Section 5, and the glacially slow DOJ (what the F is it doing?), is reasonable, and I accept that. Then I will put up my matrix chart, and Mike if he wants can use it as well for his map. The only sensitive CD's I suspect in Norcal, are CA-11, and maybe CA-03, using my CD numbers. We shall see. In Socal, we have in  potential play, CA-24, CA-23, CA-43, maybe CA-42, and CA-50.  And oh yes, the red Asian tiger CD, CA-32, which is barely in reach of the Pubs someday, maybe, if Chu retires, but certainly isn't in the Commission's map.

Make sense?

[muon2]

It sounds reasonable to me. I'll proceed apace to explore what my constraints would do to the map.

The competitiveness question is highly dependent on getting the right mix of elections. There's also the parallel question of fairness. I'd like to apply my partisan bias factor to test the maps, but DRA only has 2008 data, not 2004. In a state like CA it's hard to estimate within the big counties. I'm hoping Torie's matrix will provide a useful conversion to estimate PVIs.

In any case, I can move on to my subregional plan. Here the three regions are further subdivided into whole counties where each subregion is a whole number of districts to within 5% of a district (about 35,000). The idea is that there should only be one county line crossing between regions and subregions, except where required by the VRA.



NORTH REGION
Upper Sacramento (2) -15,869
Lower Sacramento (4) +7,777
Central Valley (2) +4,914

COAST REGION
North Coast (3) +11,821
Central Coast (11) -9,662

SOUTH REGION
Bakersfield (2) +28,982
Los Angeles (14) -22,065
Socal (15) -5,900

The central coast is technically two subregions, since San Mateo is 2.2% over population for one district and the remainder is then 3.6% under, but it was easier to see the map without that single district county.