notes.txt

Problems with using a Genetic Algorithm for redistricting

I initially formulated the genetic string to be an array, one entry per block, noting which district the block belonged to. Crossover mating on this type was simple. Mutation was easy and I allowed either swapping the district value of two positions or randomly assigning a district to some position.

I used several different variations on measuring the fitness. They all had as components the "moment" of the districts and the inter district population variance.

	The moment is the sum of ((block to district center distance)^2 * block population) across all blocks for the district each is in. A district center is the population weighted average of the positions of the blocks in a district.

	The inter population variance is the sum of the squared differences between the population of each district and the average population of all the districts.

At first I tried to get an overall fitness score by multiplying moment and population variance. Lower scores are better in both components and in the resulting product. It turned out that variance was easier to solve for and the system would get stuck in a local minima where any improvement in moment would result in a relatively nasty degradation in variance. I then tried increasing the importance of moment by exponentiating that term by 2, 3 or 4. With fitness equal to variance * moment^4, I got one trial run to settle down to a nice solution. On another run, I found that I wanted to be able to tune the balance dynamically as it ran. That would not be practical even if I implemented it.

I settled on the solution of mapping each component onto the range (0,1] as appropriate for each population generation. So, find the most fit for moment, the least fit for moment, and scale all moments. Similarly for population variance. The final fitness for an individual is the addition of these scaled values. This method has been successfully running for days, neglecting neither component and moving both towards a better global solution.

But, it's been running for days an the results are mediocre.

Using just the zip code blocks, 1757 for California, or just 366 for New Mexico, the system rapidly settles on a "pretty good" solution. But this solution still has outlier blocks allocated to far away districts. Districts can even be more nastily intermingled, in the case of the Texas data which has been running for a day now.

Ultimately, I think the problem with this method is its lack of concept of a "region". I'm really looking for a region based answer, but I have no concept of a region in the solver. I assumed that the "moment" component would implicitly result in regions, but it seems to not quite be a strong enough influence. I am still running the California full census block data set. At that level of detail this implicit region building may yet happen, but that data set has been running for over 240 hours on a 1.25 GHz G4 processor. I'll leave it running until I write a better, region based, solver.

A couple little optimizations:
Neglecting the curvature of the earth and using longitudes and latitudes as cartesian coordinates provided a great speed boost. Not taking the square root of distances also provided a great speed boost, and simply biased the system to get those distances down.

2005-01-20 14:44:20 -0800

A region based solver will be able to use the same population variance and moment measures of district-plan fitness, but the solutions will be automatically constrained to be


Census Data URLs:
http://ftp2.census.gov/census_2000/datasets/Summary_File_1/
For some state (2 letter abbrev, lower case) "${st}", you just need the ${st}geo_uf1.zip files from Summary File 1. That has the per-block population and lat-lon in it. Download these into subdir 'data/' and preprocess.
data/
	${st}geo.uf1	-- what comes inside ${st}geo_uf1.zip . I recommend discarding the .zip and using bzip2 on this after preprocessing.
	${st}101.uf1	-- `grep 'uSF1  ..101' < ${st}geo.uf1 > ${st}101.uf1` Only the "summary level 101" lines contain block info we want.

http://www2.census.gov/geo/tiger/tiger2005se/
use the tiger/get.pl script to slurp a state worth of TIGER map data.
tiger/
	??/	-- a state's 2 letter abbrev, in upper case
		zips/	-- where get.pl slurps to
		zips/url	-- a file that just contains 'http://www2.census.gov/geo/tiger/tiger2004fe/??/' for some state '??', used by get.pl
		raw/	-- unzip the zips into this dir. from inside zips, `unzip -d ../raw tgr\*.zip`

Can't seem to find 109th congressional districts down to block level.
http://www.census.gov/geo/www/cd109th/tables109.html
http://ftp2.census.gov/census_2000/datasets/109_Congressional_Districts/109_CD_HundredPercent/California/cageo_h09.zip

NEW TIGER FORMAT
http://www.census.gov/geo/www/tiger/
http://www.census.gov/geo/www/tiger/tgrshp2009/tgrshp2009.html
http://www.census.gov/geo/www/tiger/tgrshp2009/TGRSHP09.pdf
started in 2008 but the old format was still working with the census 2000 data, but I'll need to upgrade!

http://www.esri.com/library/whitepapers/pdfs/shapefile.pdf

new format, for new data:
http://www2.census.gov/geo/tiger/TIGER2009/09_CONNECTICUT/tl_2009_09_tabblock.zip

new format, for census2000 data:
http://www2.census.gov/geo/tiger/TIGER2009/09_CONNECTICUT/tl_2009_09_tabblock00.zip

http://www2.census.gov/geo/tiger/TIGER2009/25_MASSACHUSETTS/
http://www2.census.gov/geo/tiger/TIGER2009/25_MASSACHUSETTS/tl_2009_25_tabblock.zip
http://www2.census.gov/geo/tiger/TIGER2009/25_MASSACHUSETTS/tl_2009_25_tabblock00.zip
http://www2.census.gov/geo/tiger/TIGER2009/25_MASSACHUSETTS/25017_Middlesex_County/tl_2009_25017_faces.zip

I may be able to get better block-block links information from combining the TFID and reconstructed BlockID from the *_faces.zip topological faces file and the TFIDL,TFIDR from the *_edges.zip all lines file.

*_tabblock00.dbf
	state		two digit fips code
	county		three digit code
	tract		six digit
	block		four digit
	blockid		which appears to be the previous four concatenated, 15 digits
	name		string(10)
	mtfcc		maf/tiger feature class code
	ur		urban/rural
	uace		urban area code
	funcstat	funcional status
	aland		land area
	awater		water area
	intplat		interior point, lat,lon
	intplon

*_tabblock.dbf
	statefp		two digit fips code
	statens		8 char ansi code
	countyfp	three digit code
	statefp00	two digit fips code for 2000 census
	countyfp00	three digit code for 2000 census
	tractce00	six digit code for 2000 census
	blockce00	four digit code for 2000 census
	suffix1ce	one char suffix to all that
	blkidfp		char[16] = statefp00+countyfp00+tractce00+blockce00+suffix1ce
	name		char[11]
	mtfcc		maf/tiger feature class code
	ur		urban/rural
	uace		urban area code
	funcstat	funcional status
	aland		land area
	awater		water area
	intplat		interior point, lat,lon
	intplon

And I get the sense from this that the specifics may change for the 2011 release.



TO DO

use unzip source in C++ utility for unpack/process? ftp://ftp.info-zip.org/pub/infozip/src/
use Java std library unzip for unpack/process?

Calculate voting-rights-act compliant good-gerrymandering majority-minority districts.

make automatic download/run client (might just be python with a web/cgi thing on
the server side) to do redistricting@home

http://www.redistrictinggame.com/

Micah Altman's Dissertation
"Districting Principles and Democratic Representation"
http://www.hmdc.harvard.edu/micah_altman/disab.shtml
http://www.hmdc.harvard.edu/micah_altman/dispdf/dis_full.pdf


http://scholar.google.com/scholar?q=redistricting+compactness&hl=en&lr=lang_en



"Nonpartisan Political Redistricting by Computer", 1965. Hess, Weaver,
Siegfeldt, Whelan
http://links.jstor.org/sici?sici=0030-364X(196511%2F12)13%3A6%3C998%3ANPRBC%3E2.0.CO%3B2-2

Gordon S. Harrison's review of the 2001 Alaska redistricting by 5 member commission
http://www.law.duke.edu/shell/cite.pl?23+Alaska+L.+Rev.+51

Alaska's brief history of redistricting is frought with lawsuits over every initially proposed map after the 1970, 1980 and 1990 Censuses.

2010-03-08
http://www.ombwatch.org/node/10815
http://rosereport.org/20100308/new-movement-for-redistricting-reform-in-illinois/

who are these people, what do they think?
http://www.americansforredistrictingreform.org/index.html
http://www.campaignlegalcenter.org/redistricting.html

http://www.fairdistrictsflorida.org/home.php

maybe the LWV has the contacts I want
http://www.lwv.org//AM/Template.cfm?Section=Home

http://www.brennancenter.org/content/resource/a_citizens_guide_to_redistricting/
http://www.brennancenter.org/content/section/category/redistricting/

Indiana Secretary of State Todd Rokita
http://www.in.gov/sos/
http://www.rethinkingredistricting.com/



clean, run inside data/??/:
rm -f *.jpg *.png *.mppb *.pb *.links *.dsz drendcmd basicargs handargs makedefaults .make *.uf1 mrun?? measure geometry.pickle ??_start_stats .uf1

TODO: fix Alaska, the map data has longitudes crossing -180 to +180. and if I get really crazy, do polar projections or something.

TODO: overlay the "urban area" data onto my maps.
http://www2.census.gov/geo/tiger/TIGER2009/tl_2009_us_uac00.zip

TODO: link problem may be due to disjointness in ubids between geo-101 data
and faces+edges

TODO: project states into something like azimuthal equal distance projection centered over the state  http://trac.osgeo.org/proj/


2011-02-10 Thursday 10:24 AM Boston
I have a first results on 2011 data! It's just 1000 generations of the nearest neighbor solver, so I'm sure I'll do better soon. So far I've run it for NJ and VA. Had to toss the super-fine face+edges based linking and fall back to common-line-segment linking based on the tabblock tiger maps. The faces seem to have lots of 'blocks' that don't exist in the redistricting data. Probably extra detail on water and unpopulated areas.

http://www.redistricting.wa.gov/


TODO: try to keep precincts/cities/counties together by imposing a distance penalty on blocks that are part of a region that isn't entirely within the district.

<b>Keep-Together regions</b> (city, county, other). If a region is not wholly within within one district, population in region counts as N miles more away from district centers. (KT region may not encompass more than one district worth of population, probably shouldn't encompass more than (WAG) 70% of a district population.)

<b>Do-Not-Cross lines</b>. If the line from district center to population crosses this line, a N-mile penalty is added to that distance. Apply this to rivers, mountains, other natural or man made well recognized dividing lines. (If the penalty is not high enough there might wind up being just a few people on the far side of it, and that could be weird. Placement and high penalties could result in weird constrained solutions in other ways too.)

http://www.legis.iowa.gov/Resources/Redist/redistricting.aspx
http://www.house.mo.gov/largemap.aspx

2012-05-28 14:38:05 EDT (Monday, May 28 02:38:05 PM)

TODO: set G+ on the website
https://plusone.google.com/_/+1/confirm?hl=en&url=${url}

2013-05-04 16:11:54 EDT (Saturday, May 04 04:11:54 PM)

analyze and render 2013 actual districts

./analyze -P ../../CT/ct.pb -d 5 --csv-solution --loadSolution 09_CT_CD113.txt

drend  --csv-solution 09_CT_CD113.txt -P ct.pb --mppb=CT.mppb --pngout=ct113.png

drend --csv-solution=25_MA_CD113.txt -P=ma.pb --mppb=MA.mppb --pngout=/tmp/MA_us2013.png -d=-1

Make drend easier to use, detecting number of districts from loaded solution. Make district allocation lazy to accommodate this.

~/src/redistricter/render_2013_actual.py --datadir=/Volumes/bulktogo/redata/2010 --districtdir=/Volumes/bulktogo/redata/2010/2013_actual/cd
~/src/redistricter/render_2013_actual.py --datadir=/Volumes/bulktogo/redata/2010 --districtdir=/Volumes/bulktogo/redata/2010/2013_actual/sldu
~/src/redistricter/render_2013_actual.py --datadir=/Volumes/bulktogo/redata/2010 --districtdir=/Volumes/bulktogo/redata/2010/2013_actual/sldl

t

~/src/redistricter/analyze -P FL/fl.pb -d=-1 --csv-solution=2013_actual/cd/12_FL_CD113.txt --compare :5,7,8,9,10,11,12,13 --labels total,white,black,native,asian,pacific,other,mixed --dsort 1 --html /tmp/fl.html --notext

~/src/redistricter/analyze_submissions.py --actuals=/Volumes/bulktogo/redata/2010/2013_actual --datadir=/Volumes/bulktogo/redata/2010 --soldir=/Volumes/bulktogo/redata/2010/solutions --report=/Volumes/bulktogo/redata/2010/solutions/report/report.html --outdir=/Volumes/bulktogo/redata/2010/solutions/report --rooturl=/2010/ --rehtml --config=FL_Congress

election data at precinct level:
http://www.stanford.edu/~jrodden/jrhome_files/electiondata.htm
http://projects.iq.harvard.edu/eda/data

http://www.washingtonpost.com/blogs/wonkblog/wp/2014/06/03/this-computer-programmer-solved-gerrymandering-in-his-spare-time/

http://www.washingtonpost.com/blogs/wonkblog/wp/2014/10/08/why-drawing-congressional-districts-by-race-is-a-terrible-idea/

2014-12-11 14:45:17 EST (Thursday, December 11 02:45:17 PM)

ftp://alt.ncsbe.gov/enrs/resultsVTD20141104.zip
ftp://alt.ncsbe.gov/enrs/resultsPCT20141104.zip
ftp://alt.ncsbe.gov/enrs/pritabulation11xx06xx2012.zip

TODO: draw map tiles projected with http://trac.osgeo.org/proj/
https://github.com/OSGeo/proj.4
git clone https://github.com/OSGeo/proj.4.git proj4

render tiles using:
projection EPSG:900913 'google' aka and more officially epsg:3857


20150630
The Richmond Times-Dispatch opined today [1], citing a recent article [2], that gerrymandering is making voting a pointless act and depressing voter turnout and depressing our democracy. In 600 state elections, 19 incumbents were defeated. In 280 party primaries there were 8 challenges to the incumbent. Voter turnout hovers at 8%.[3]


[1] http://www.richmond.com/opinion/our-opinion/article_42f07e65-aeb5-5d2e-bffb-f8cd34261e9a.html#.VYs95v4eYu0.wordpress
[2] http://www.richmond.com/news/virginia/government-politics/jeff-schapiro/article_784cffac-a447-53c0-afba-e97ecb83aa5d.html
[3] I haven’t checked these numbers; I assume the original sources are public data; I find these numbers plausible and believable.

http://bdistricting.com/2010/VA/


@RTDSchapiro Would some example non-gerrymandered compact district maps help? http://bdistricting.com/2010/VA/

TODO: http://vis4.net/blog/posts/no-more-mercator-tiles/

http://leafletjs.com/
http://polymaps.org/
http://openlayers.org/
http://modestmaps.com/
http://proj4js.org/
https://www.mapbox.com/tilemill/
http://mapnik.org/
http://en.wikipedia.org/wiki/Albers_projection
"Albers Equal-Area Conic"
https://github.com/mapbox/mercantile

2016-01-14 10:23:26 EST (Thursday, January 14 10:23:26 AM)

~/src/redistricter/analyze -P ~/src/redata/2010/OH/oh.pb -d 16 --places ~/src/redata/2010/OH/geoblocks.places --place-names ~/src/redata/2010/OH/st39_oh_places.txt --loadSolution ~/src/redata/2010/solutions/report/OH_Congress/solution.dsz --text -

~/src/redistricter/analyze -P ~/src/redata/2010/OH/oh.pb -d 16 --places ~/src/redata/2010/OH/geoblocks.places --place-names ~/src/redata/2010/OH/st39_oh_places.txt --csv-solution ~/src/redata/2010/2013_actual/cd/OH.csv --text -


Pennsylvania constitution on how to draw state legislature districts
https://en.wikisource.org/wiki/Constitution_of_the_Commonwealth_of_Pennsylvania_1874/Article_2#Section_16:_Senatorial_districts.

Section 16: Senatorial districts.
The State shall be divided into fifty senatorial districts of compact and contiguous territory, as nearly equal in population as may be, and each district shall be entitled to elect one Senator. Each county containing one or more ratios of population, shall be entitled to one Senator for each ratio, and to an additional Senator for a surplus of population exceeding three-fifths of a ratio, but no county shall form a separate district, unless it shall contain four-fifths of a ratio, except where the adjoining counties are each entitled to one or more Senators, when such county may be assigned a Senator on less than four-fifths and ex
ceeding one-half of the ratio; and no county shall be divided unless entitled to
 two or more senators. No city or county shall be entitled to separate represent
ation exceeding one-sixth of the whole number of Senators. No ward, borough or t
ownship shall be divided in the formation of a district. The senatorial ratio sh
all be ascertained by dividing the whole population of the state by the number f
ifty.

Section 17: Representative districts.
The members of the House of Representatives shall be apportioned among the several counties, on a ratio obtained by dividing the population of the State as ascertained by the most recent United States census by two hundred. Every county containing less than five ratios shall have one representative for every full ratio, and an additional representative when the surplus exceeds half a ratio; but each county shall have at least one representative. Every county containing five ratios or more shall have one representative for every full ratio. Every city containing a population equal to a ratio, shall elect separately its proportion of the representatives allotted to the county in which it is located. Every city entitled to more than four representatives, and every county having more than one hundred thousand inhabitants shall be divided into districts of compact and contiguous territory, each district to elect its proportion of representatives according to its population, but no district shall elect more than four representatives.


citation in Duke Journal of Constitutional Law & Public Policy, article by Micah Altman and Michael McDonald, http://scholarship.law.duke.edu/djclpp/vol5/iss1/5/ , "The Promise and Perils of Computers in Redistricting", 2010, Vol 5 issue 1


2016-02-08 21:03:09 EST (Monday, February 08 09:03:09 PM)

${HOME}/src/redistricter/shapefile.py \
--csvDist ${HOME}/src/redata/2010/solutions/report/OH_Congress/solution.csv.gz \
--outlineOut /tmp/outline \
--verbose \
${HOME}/src/redata/2010/OH/zips/tl_2010_39_tabblock10.zip


2016-03-19 13:39:25 EDT (Saturday, March 19 01:39:25 PM)

git clone https://github.com/OSGeo/proj.4.git proj4
cd proj4
./autogen.sh
./configure --prefix=${PWD}
make

https://trac.osgeo.org/proj4j/
https://github.com/Proj4J/proj4j
git clone https://github.com/Proj4J/proj4j.git proj4j
cd proj4j/trunk
mvn install
#cd proj4j/trunk/src
#ant




USGS professional paper 1395 (1987) supercedes USGS Bulliten 1532 (1982)
http://pubs.usgs.gov/pp/1395/report.pdf
but this appears to all be handilly encoded in PROJ.4


ohio (north)
--proj NAD83:3401

2016-05-09 13:35:04 EDT (Monday, May 09 01:35:04 PM)

xxxxx
oooxo
oxoxx
oxxoo
xoxxx



ooxox
oxoxo
xxxxx
xooxo
xxxxo

2016-08-03 18:29:43 EDT (Wednesday, August 03 06:29:43 PM)

https://www.legislature.ohio.gov/laws/ohio-constitution

2016-11-01 15:37:51 EDT (Tuesday, November 01 03:37:51 PM)

AK=${HOME}/src/redata/2010/AK
make linkfixup&&./linkfixup --plgeo ${AK}/geoblocks -p ${AK}/ak_base.pb
./drend -P ${AK}/ak_base.pb -pngout ak_base.png -loadSolution data/2010/solutions/report/AK_Senate/688/solution.dsz -d 20


make linkfixup&&./linkfixup --plgeo ${AK}/geoblocks -p ${AK}/ak_laea.pb --proj&&./drend -P ${AK}/ak_laea.pb -pngout ak_laea.png -loadSolution data/2010/solutions/report/AK_Senate/688/solution.dsz -d 20

make linkfixup&&./linkfixup --plgeo ${AK}/geoblocks -p ${AK}/ak_aeqd.pb --proj&&./drend -P ${AK}/ak_aeqd.pb -pngout ak_aeqd.png -loadSolution ${HOME}/src/redata/2010/solutions/report/AK_Senate/688/solution.dsz -d 20


2017-03-02 12:28:46 EST (Thursday, March 02 12:28:46 PM)

TODO: try computing a "Fryer & Holden" sum of squared distances - per district measure. sum across districts ( sum of squared distances between all pairs of persons in a district )

algorithmic redistricting papers? Fifield et al. 2015, Holden 2015
Fifield B, Higgins M, Imai K, Tarr A. 2015. "A new automated redistricting simulator using Markov chain Monte Carlo." Work. Pap., Princeton Univ., Princeton, NJ
http://imai.princeton.edu/research/files/redist.pdf

lots of citations of prior work in algorithmic redistricting.
'graph cut' problem. MCMC.
Somewhat more interested in getting a representative sample of possible district maps than generating one good one.





Fryer, R. and Holden, R. (2011). "Measuring the compactness of political districting plans." Journal of Law and Economics 54, 3, 493–535
http://scholar.harvard.edu/fryer/publications/measuring-compactness-political-districting-plans
http://scholar.harvard.edu/files/fryer/files/measuring_the_compactness_of_political_districting_plans.pdf

according to Fitfield,Giggins,Imai,Tarr(2015) "develops an approximate algorithm to efficiently compute the minimum of the sum of squared distances"

acknowledges distribution within districts, not just shape geometry!
tract-level data, not block

sum across districts ( sum of squared distances between all pairs of persons in a district )
they use that as a numerator with a power diagram (Voronoi) solution they assert is optimal as the denominator.




Holden R. 2015. "A process for non-partisan redistricting." Work Pap., Univ. New South Wales, Sydney, Aust.


Harvard Election Data Archive
http://projects.iq.harvard.edu/eda/home
https://dataverse.harvard.edu/dataverse/eda


https://sites.stanford.edu/sssl/stanford-election-atlas

2017-06-11 13:32:00 EDT (Sunday, June 11 01:32:00 PM)

A test of 'spherical cows on frictionless planes' of gerrymandering science...

Given randomly distributed people of two groups and random but perfect (contiguous, zero inequality) district spatial groupings, what is the distribution of outcomes of districts which prefer one group or the other?

Vary the group population split from 50-50 to 70-30.
Vary districts from 2-15.
Try 'clumpy' non-uniform distributions. (e.g. the urban self-packing group)

What's the chance of a MA 9-0 outcome given the MA statewide party registration ratio?


Google hilbert curve quad tree sphere spatial indexing
https://code.google.com/archive/p/s2-geometry-library/source/default/source
https://github.com/golang/geo
https://docs.google.com/presentation/d/1Hl4KapfAENAOf4gv-pSngKwvS_jwNVHRPZTTDzXXn6Q/view#slide=id.i148
http://blog.christianperone.com/2015/08/googles-s2-geometry-on-the-sphere-cells-and-hilbert-curve/


2017-07-12 12:25:48 EDT (Wednesday, July 12 12:25:48 PM)

https://www.vox.com/the-big-idea/2017/7/11/15949750/research-gerrymandering-wisconsin-supreme-court-partisanship

wasted vote: either above what was necessary to win, or what was cast for a loser
efficiency gap: (party a wasted) - (party b wasted) / (all votes)

https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2457468

2017-08-07 20:35:42 EDT (Monday, August 07 08:35:42 PM)

TODO: 'data load with filter' mode; e.g. filter to just include one city or county and make city council or county council districts

17000 core hours in 2011 on 2010 intel core-i7 970
17000 core hours in 2017 on google compute preemptible high-cpu instances: $130

2017-08-11 14:56:51 EDT (Friday, August 11 02:56:51 PM)

why I should quit population distance to land-area-center: land area is less precise than population counts!
(see arguments about why 'polsby popper' perimeter-vs-area measure is bogus due to varying level of detail on geographic data)

2017-09-23 11:20:58 EDT (Saturday, September 23 11:20:58 AM)

TODO: trim presentation down to 0.001 km

2017-10-06 08:29:11 EDT (Friday, October 06 08:29:11 AM)

Gill v Whitford (Wisconsin case) oral arguments
https://www.supremecourt.gov/oral_arguments/argument_transcripts/2017/16-1161_kjfm.pdf


Roberts, p5 l 18-17, '... right of association with other people elsewhere in the state'
Any citizen of a state has an interest in the district map of that state; and my bias says to look to at large proportional representation to best associate with people elsewhere in the state.

p15-18; "Legislatures can very much rest on conjecture whereas courts cannot" - Misha Tseytlin, arguing for WI before SCOTUS
The justices seem to go along with this. A gerrymander can merely 'probably' work whereas apparently a court needs to be 100% sure that a thing is a bad gerrymander before it can be thrown out.




Sotomayor argues the absurdity of the pro-gerrymander side.
"Could you tell me what the value is to democracy from political gerrymandering?
How -- how does that help our system of government?"
...
". it's okay to stack the decks so that for 10 years or an indefinite period of time one party, even though it gets a minority of votes, can't get a minor -- gets a minority of votes, can get the majority of seats?"
 -- Justice Sonia Sotomayor

2017-10-17 10:40:35 EDT (Tuesday, October 17 10:40:35 AM)

Radiolab "More Perfect" podcast 2017-10-03
http://www.wnyc.org/story/whos-gerry-and-why-he-so-bad-drawing-maps/

2017-12-06 16:17:27 EST (Wednesday, December 06 04:17:27 PM)

SCOTUS Gill v Whitford (Wisconsin case) notes

p.16 lines 10-19, SOTOMAYOR: "... they created three or four different maps, they
weren't partisan enough.  They created three or
four more maps, they weren't partisan enough.
And they finally got to the final map,
after maybe 10 different tries of making it
more partisan, and they achieved a map that was
the most partisan on the S-curve.
         And it worked.  It worked better than
they even expected, so the estimate wasn't
wrong.  The estimate was pretty right."

p.17 l22-23 Teytlin for WI: "... there was no constitutional requirement " to use non-gerrymandered maps

p.18 l12-13 "Legislatures can very much rest on conjecture
whereas courts cannot."

p 23 Gorsuch asks if there are measure thresholds for when gerrymandering is too bad. A softball question for WI to say: no, it's too messy, there will never be any reasonable standard.

p 24-25. RBG "I would like to ask
you what's really behind all of this.  The
precious right to vote, if you can stack a
legislature in this way, what incentive is
there for a voter to exercise his vote?
Whether it's a Democratic district or a
Republican district, the result -- using this
map, the result is preordained in most of the
districts.
Isn't that -- what becomes of the
precious right to vote?  Would we have that
result when the individual citizen says:  I
have no choice, I'm in this district, and we
know how this district is going to come out?  I
mean that's something that this society should
be concerned about."

p.27 Murphy for WI admits that a state law propping up one party through Gerrymandering would be unconstitutional.

p.28 SOTOMAYOR trolls it out "Could you tell me
what the value is to democracy from political
gerrymandering?  How -- how does that help our
system of government?"

p.30 ROBERTS want to ask the appealees (vs WI) about standing. :-/
Is any one person harmed by systemic gerrymandering?
Do parties matter? Why do people vote? What is democracy?

p.34 Smith vs WI "Their party has been punished by the law of the State of Wisconsin."

p.37 ROBERTS "if the claim
 2    is allowed to proceed, there will naturally be
 3    a lot of these claims raised around the
 4    country."
... and every one of those cases will have to go to SCOTUS.

ROBERTS is allergic to math and assumes Americans are:
"And if you're the intelligent man on
14    the street and the Court issues a decision, and
15    let's say, okay, the Democrats win, and that
16    person will say:  "Well, why did the Democrats
17    win?"  And the answer is going to be because EG
18    was greater than 7 percent, where EG is the
19    sigma of party X wasted votes minus the sigma
20    of party Y wasted votes over the sigma of party
21    X votes plus party Y votes.
22             And the intelligent man on the street
23    is going to say that's a bunch of baloney.  It
24    must be because the Supreme Court preferred the
25    Democrats over the Republicans."

...
p.38 "And that is going to cause very
 4    serious harm to the status and integrity of the
 5    decisions of this Court in the eyes of the
6 country."


p.40 ROBERTS math innumeracy:
"the whole point is you're
 5    taking these issues away from democracy and
 6    you're throwing them into the courts pursuant
 7    to, and it may be simply my educational
 8    background, but I can only describe as
 9    sociological gobbledygook."

p.40-41 BREYER brings it back in:
"if party A wins a majority of votes,
20    party A controls the legislature.  That seems
21 fair.
22             And if party A loses a majority of
23    votes, it still controls the legislature.  That
24    doesn't seem fair.  And can we say that without
25    going into what I agree is pretty good
gobbledygook?"

p.41
"3             CHIEF JUSTICE ROBERTS:  And if you
 4    need a convenient label for that approach, you
 5    can call it proportional representation, which
 6    has never been accepted as a political
 7    principle in the history of this country."

"partisan symmetry" vs "proportional representation"

Smith "... if
23    party A at 54 percent gets 58 percent of the
24    seats, party B when it gets 54 percent ought to
25    get 58 percent of the seats.  That's symmetry."

p.42 ALITO "... gerrymandering is distasteful."
but he wants a rigorous rule

and he's suspicious that 'efficiency gap' has only been around since 2014, too new.

p.46 KAGAN wants a threshold too

Smith proposes 'almost certain one party rule for a decade' as the threshold.

p.54 ALITO "... protection
 5    of incumbents, which has been said to be a
 6    legitimate factor"
 sad

p.63... closing arguments FUD for WI

end 11:03 am; started at 10:04 am; and so the fate of our country is decided in a one hour block.


2017-12-22 10:51:09 EST (Friday, December 22 10:51:09 AM)

We want to represent _who_ people are at least as much as _where_ people are.
Are we representing people or places?
Are we representing people or a line on the map?
Identifying a person only with the place they live is serfdom.

2019-01-08 22:47:30 EST (Tuesday, January 08 10:47:30 PM)

http://www.cleanmissouri.org/wp-content/uploads/2018/04/Packet-048-Petition-Text-copy.pdf

2021 reapportionment estimates
https://docs.google.com/spreadsheets/d/1MkSW6Y9TAMBq9mGPiuTNInDfJ3PQiDkNyRro25WxG5s/edit?usp=sharing

2021-01-19 17:35:05 EST (Tuesday, January 19 05:35:05 PM)

https://www.census.gov/programs-surveys/decennial-census/about/rdo/summary-files.html
https://www2.census.gov/geo/pdfs/maps-data/data/tiger/tgrshp2020pl/TGRSHP2020PL_TechDoc.pdf

States Posted on 01/19/2021: AR, CO, MD, NJ & VA
States Posting on 01/26/2021: CA, IA, IN, LA, MO, NC, NY, OH, & OR
States Posting on 02/02/2021: DE, HI, ME, MS, NE, NV, PA, SD, WA, WI

https://www.census.gov/geographies/mapping-files/time-series/geo/tiger-line-file.html
https://www2.census.gov/geo/tiger/TIGER2020PL/STATE/
https://www2.census.gov/geo/tiger/TIGER2020PL/STATE/08_COLORADO/
https://www2.census.gov/geo/tiger/TIGER2020PL/STATE/08_COLORADO/08/tl_2020_08_county20.zip
https://www2.census.gov/geo/tiger/TIGER2020PL/STATE/08_COLORADO/08073/tl_2020_08073_edges.zip
https://www2.census.gov/geo/tiger/TIGER2020PL/STATE/08_COLORADO/08073/tl_2020_08073_faces.zip

https://www2.census.gov/geo/tiger/TIGER2020PL/LAYER/FACES/
https://www2.census.gov/geo/tiger/TIGER2020PL/LAYER/EDGES/
https://www2.census.gov/geo/tiger/TIGER2020PL/LAYER/COUNTY/2020/
https://www2.census.gov/geo/tiger/TIGER2020PL/LAYER/TABBLOCK/2020/

http://www2.census.gov/census_2010/
-> ?
https://www2.census.gov/census_2020/


# xeof end of file search target