US House Redistricting: Arizona

[jimrtex]

Trying to draw the most Native district possible - blue seat is 40% White, 28% Hispanic, 27% Native. Getting it to ~30% would be possible, but not while balancing the other districts. Urban Tucson district just happened as a result. (This has only one Hispanic majority district, so is quite fantastical.)

Isn't that going to be hard to maintain given the growth in Pinal County?

[jimrtex]

Arizona sues to overturn Section 5.

http://www.eastvalleytribune.com/arizona/article_3be614ce-cf53-11e0-bbb0-001cc4c03286.html

It's pretty interesting.

Arizona is underneath pre-clearance because of how Hispanics were tabulated.  At the time (1972) the Census Bureau did not have any separate tabulation for Hispanic's so depending on the state they included persons who had certain surnames, and in others they included persons born in Puerto Rico.  Nevada would have been covered if the same standard had been used as was used for Arizona, and Arizona would have been exempt if the Nevada standard had been used.

In addition, Arizona had provided for bilingual ballots in 1974, but the VRA of 1975 retroactively set a deadline of 1972.  It is similar to the California counties that are covered because military personal didn't vote (or didn't vote where they were based) and California had a literacy test that they no longer enforced.   In some cases, the military bases have been closed for over a decade.

[jimrtex]

California's Commission has 5 Dems, 5 Reps and 4 Indies, and requires at least 3 votes from each bloc for passage. (It passed 12-2 with both nays being Republicans, by the way, so it got the bare minimum of minority consent.) Similarly, you could make the final passage in Arizona require four votes.

That would help. Of course the Pubbies on the CA commission were pathetic too, but I really can't get too angry over that map, and a lot of CD's are close enough, that I kind of like the incentive it offers to Pubbies to stop being way out there in never-never land.  In a word, I fantasize it might strengthen a bit my little microscopic wing of the party. But hey, here on this forum, my wing is close to half the party! No doubt, that is in part due to my inspired "leadership." 

It would be a lot better to let ordinary decent citizens draw the maps, rather than trying to form panels of impartial experts chosen based on their partiality.

[jimrtex]

You must be logged in to read this quote.

I think there's a combination of ideas here that could work.

I suggest that the process allow the commissioners to have a partisan intent, but tie them to maps submitted by outsiders. Then force them to select a map based on criteria developed by neither the commission nor the mapmakers. Finally require a supermajority of commissioners to override the criteria and select a map other than the one with the best fit to the criteria.

The ordinary decent citizens would be a representative panel drawn from the voter rolls (perhaps 1 per 2000 persons) on a geographically stratified basis.

The following is based on Ohio:

A county apportionment area (CAA) is one or more contiguous counties with a population approximately equivalent to an integer number of representatives.

The maximum relative deviation would be 0.07 * sqrt(N), where N is the N number of representatives.  This would rule out Allen, Wood, and Columbiana counties, but could permit some multi-county single member districts to be created.   The intent is to prevent systematic overrepresentation or underrepresentation of larger counties, simply because any error can be evenly distributed.   If a county was entitled to 9.5 representatives, it could theoretically have 10 districts within the 5% safe harbor, but this would require total disregard of city and town boundaries.  The above limit is intended to avoid this situation.

Anyone could submit a plan that would provide CAA for the entire State with a total of 99 representatives.

A plan would be scored in the following manner: (1) 1 point per CAA; (2) An additional point for a single-county CAA; and (3) An additional point for a single district CAA.  (eg Wayne and Richland counties would be worth 3 points).

The highest ranking plan and those with 90% of the points would be presented to the panel of voters, who would rank the plans based on their immediate effect on them.  For example, the Summit County voters might rank plans based on whether they would pair Summit and Medina; Summit and Stark; or Summit and Portage, but would rank plans as equal that had no immediate effect (eg pairing Montgomery with Greene or Preble).

The top statewide plan would be determined using Condorcet methods.  There might be recursion where the rankings of the voters in each CAA of the top statewide plan would be evaluated to see if they concurred with the statewide plan.

CAA's with a single representative would become districts.

All other CAA's would be handled separately and independently.

A township apportionment area (TAA) is one or more contiguous townships (in a CAA) with a population approximately equivalent to an integer number of representatives.  Note, most large municipalities and many smaller ones in Ohio are also townships.

A TAA that includes parts of two counties is said to span the two counties.  No more than one TAA may span any pair of counties (this applies even if the counties are not contiguous).  This spanning rule replaces the current rule on county splits.   A county with 1.4 representatives could be placed in two (or more) districts going into adjacent counties (but not the same ones), rather than being restricted to one district wholly within the county, and the remnant in one other multi-county district. 

An excessively split county is split between more TAAs than the ideal representation of the county, rounded to the nearest integer and one (1) added.  A county with a population equivalent to less than 0.5 representatives is excessively split if it it split between two districts.  A county entitled to 1.6 representatives could be split among 3 districts, while one entitled to 1.4 could be split among 2 districts, without being excessively split.

Enclave townships (such as Bratenahl, Upper Arlington, and Norwood) would be treated as part of the containing township (Cleveland, Columbus, or Cincinnati) for purposes of creating TAA, unless they had sufficient population to form a separate TAA,  So if Cuyahoga County were a CAA, one TAA would include all of Cleveland plus some adjacent cities or townships.

Individuals could submit plans for a CAA comprised of TAAs.  All plans that proposed the most number of TAAs, and the fewest number of excessive county splits, would be submitted to the panel of voters from the affected counties, who would rank them based on their effect on themselves.

TAAs with more than one representative would be divided into Neighborhood Apportionment Areas (Neighborhoods would replace wards, and would be defined in advance of the census). Only one NAA could span a pair of counties or township/cities.  No neighborhoods could be split.

Qualifying plans would create the most NAAs and create the fewest city/township splits (in general, larger cities would be placed in as few districts as possible), while smaller townships and cities would be treated as if they were neighborhoods and not split.

NAAs with more than one representative (these could be extremely rare) would be divided based on proposals for splitting individual neighborhoods.  Only voters in those neighborhoods would vote on the actual split.

[jimrtex]

And what if the next chair doesn't obey the Republicans demands? Are they going to repeat the process over and over again until they find someone that will? This isn't going to stop, because the commission isn't going to draw the map they want, no matter who the chair is.

The problem for the GOP is that the commission is directed to create competitive districts. But the mix of 2008 and 2010 elections isn't particularly representative, and will tend to create slight Dem leans instead of even districts. Only if the commission is willing to adopt a better mix of elections will the GOP find them making better maps.

"To the extent practicable, competitive districts should be favored where to do so
would create no significant detriment to the other goals."

And the problem with creating competitive districts is that it also requires exacerbation of non-competitive districts.

[jimrtex]

You must be logged in to read this quote.

(emphasis added)

And that is the thing. Competitiveness is subordinated to the other parameters.

That interpretation of the law was explicitly ruled to be false by the AZ SC in 2002 (or 1 or 3 or whatever), fwiw.

Anyways, I presume Mathis is removed for good now, the SC has agreed? IIRC "impeachment" (which implies an actual trial) is a misleading word, and it's actually a power of the Governor and 2/3 of the Senate to remove Commissioners (presumably written with the expectation that 2/3 of the Senate should be unachievable without really serious violations).

The plain meaning of the text is subordination; if the AZ Supremes ruled otherwise, I cannot imagine the rationale. Impeachment is not a trial - it's a charge.

http://www.azcourts.gov/Portals/23/pdf2009/AZMinorityvFairRedistrictingOpinion.pdf

Actually the decision was made in 2009 (the commission website says this was it was a (2005) case, but it is definitely with the 2009 opinions.  The justice who partially dissented said it wasn't worth drawing up a new plan for 2010.  The lawsuit was over the legislative plan, rather than congressional map.

The first plan the redistricting commission drew wasn't precleared (this is probably a weakness of non-legislative commissions, since they are more likely to roll over for minority groups, and minority groups are likely to lobby for minority members to be on the commission), so a federal court drew the plan for the 2002 election. 

And then after the commission drew a new map, they were sued for not drawing enough competitive districts.  A state district court ruled that they should have drawn more competitive districts (7 vs. 4 or 30 legislative districts (Arizona elects one senator and two representatives from each legislative district) because the plaintiffs had drawn a plan with 3 more competitive districts, which at least in the eye of the court adhered to the other goals.

The state appeals court slapped down the district court, and said that that competitiveness was subordinate to the other goals.  The district court persisted, the appeals court overturned them again, and that was what ended up in the Supreme Court.

The Supreme Court most importantly said that the redistricting commission is equivalent to a legislative body, and that courts should judge the plans on the commission having a reasonable basis for their decisions.  Instead of the plaintiffs proving they could produce a "better" plan, they had to prove that the commission had not acted reasonably in producing their plan.

And the Supreme Court agreed that if any goals had to give, it was competitiveness.  But what they also said was that commission after they drew the grid plan, that they should have considered all goals in coming up with their next plan.  Instead the commission had drawn the grid plan, and then drawn up a plan for all the goals except competitiveness, and then had tried to fudge that.  The Supreme Court said that the 2000s commission had already recommended that the 2010s commission do that.

Of course that might lead the commission to attempt to draw a competitive plan that is not "too" non-compact, and doesn't violate city borders "too" much.

BTW, is there actually any relationship between the grid plan and the final map?  Or was that something to sell the commission in the first place?

[jimrtex]

It appears we have a 4th. The newly proposed Texas house court map violates the county line rule in Nueces County.

Nueces County is entitled to two districts.  The county has 340,000 people with 305,000 in the city of Corpus Christi (90%), so obviously two compact districts with a community of interest may be drawn.

The problem if you are trying to create two districts within the 5% limit, is that two districts can be drawn within 1%.   Since 1% is so much less than 5%, the obvious solution is to remove 11% of the county's population and attach it to a district based in Victoria.  Though this splits the city of Corpus Christi and happens to include about 20% of the Anglo population of Nueces County, it clearly is not based on race since the boundary is a continuous line and ignores city and school district boundaries.

Because you removed so much population it is no longer possible to create two districts in Nueces County (1.80).   Fortunately, you can add Kleberg County.  It is important not to split counties you understand.  Plus this places Bishop in the same district as Kingsville, albeit in a district dominated by Corpus Christi - though not the part of Corpus Christi nearest Kingsville.

And since you eliminated Kleberg County, you can draw that district you always wanted to from Willacy to Refugio,  So what if there is not internal road connectivity?  And since this uses whole counties it is obviously not driven by race.

[jimrtex]

http://maps.google.com/maps?q=http:%2F%2Fwww.azredistricting.org%2Fmaps%2Fpubmaps%2F100911%2FLeg-Merge_as_of_100911.kmz&hl=en&sll=34.168218,-111.930907&sspn=12.693328,26.784668&vpsrc=0&t=h&z=7

They reworked some parts of that. http://www.azredistricting.org/Maps/Tentative-Final-Maps/Legislative/maps/Tentative%20Final%20Legislative%20Map%20-%20Statewide.jpg

(add " - poster size" at the end of the address for a 140 million pixel version that may crash your computer)

I count 13 utterly safe GOP, 3 usually safe GOP (at least for now - the one in NE Phoenix might get competitive towards the decade's end), 3 tossups two of which tilt R, one D (and the tilt R in Pinal might be safe by decade's end), 2 usually safe Dem, 9 utterly safe Dem seats. After looking at their demographic and competitiveness charts.

As it stands, the Democrats won all the 'competitive' districts (basically 8, 9, 10, 26)

A lawsuit is being filed alleging Georgia style malapportionment.



Here are the population totals.

7   203,026
4   204,143
27   204,195
3   204,613
2   204,615
24   206,659
19   207,088
30   207,763
8   208,422
29   211,067
10   211,073
13   211,701
9   213,224
11   213,377
23   213,451
26   213,659
6   214,244
15   214,941
22   215,912
21   216,242
1   216,451
14   217,693
20   218,167
18   218,677
28   218,713
5   219,040
16   220,157
25   220,795
17   221,174
12   221,735

That said, the top 14 (and 14 of the top 15) overpopulated districts are Republican held. The bottom 10 (and 11 of the bottom 12) districts are Democratic held. There is only 1 GOP district with negative deviation.

There is a 0.804 correlation between district population, and percentage of non-Hispanic whites.

[jimrtex]

This lawsuit has been either refiled or something has been added to it. Given that the alleged defense for Mathis's malapportionment  was Section V, and Section V no longer applies, it is no longer a valid defense.

There are three lawsuits.

'Leach' is in state court alleging that the redistricting commission didn't follow the procedure set out in the constitution.  The agenda for a July 22, 2013 commission meeting had an item to discuss a discovery request from the plaintiffs, with executive session scheduled.

'Arizona State Legislature' is in federal court challenging the use of the initiative process to define the redistricting process for congressional districts, in violation of the US Constitution.  On May 29, 2013, the plaintiffs kicked the judges reminding them that they hadn't responded in over 180 days.

'Harris' is in federal court challenging a partisan malapportionment of the state legislature.  Trial was held in March.  On July 19, the plaintiffs filed a supplemental brief pointing out that Section V of the the VRA is no longer applicable.  The redistricting commission was using the pretext that Section V required malapportionment.