Monday, January 09, 2012

Redistricting Court Challenge: Leonard Lawson - Were There Other Ways to Draw the Districts and Meet VRA Requirements?

OK, there's a break now.  They are questioning Leonard Lawson who was the computer person for the Rights Coalition (which was one of the groups that submitted alternative maps to the Board).  He's also an employee of the Democratic Party.  I got to know him because he came to just about every redistricting board meeting and was a person I could get answers to technical and mapping issues.

Some key points that I'm trying to pull out without much time to think and before they come back into court:

I didn't make it.  They're back.  So here are the rough notes and I'll try to do an overview later.  These are very rough notes.

2:15 - Leonard Lawson - Mathematics and Physics major from ?? College.  I work for the Alaska Democratic Party.  Their primary data base manager.  I keep track of voter data bases, that's the biggest part of the job.  When elected officials want to know about a district.
Wallerie:  When you were deposed it was pointed out that you were listed as a political director for the Alaska Party.  Was that a surprise to you?

Lawson:  No, I give information.  My boss is Kay Brown, ED of Democratic Party Alaska.  Before that it was Deborah Williams.
Wallerie:  Poltical background before with Dems.
Lawson:  New Hampshire Democrats.
Wallerie:  Gilford County?
Lawson:  Center of North Carolina, volunteer for Al Gore?
Wallerie:  Coming from NC, you've had some experience with redistricting?
Lawson:  Laughter.  I went to National Conference on STate Legislators conerence on redistricting.  I think 4 day conference.  I saw John Ferguson and I I think Michael White.
Cover a wide variety of topics, but redistricting was big topic this year.
Other training:  Software, trained by company that makes map  software.  To help understand better how to use the software.  Already had a lot of experience, but wanted more.
Other:  Also took GIS class at UAA, but as it was starting I had to leave because redistricting was starting.  GIS is Geographic Information System.  Anything with geographic component - merges numbers and maps.
How does GIS relate to these maps:  It's what's use for most of these maps.  I think they were all made with GIS software.  Allows you to use Census data and realte the to maps.  Includes ethnicity, over and under 18, etc. and can be displayed visually.
What a census block is:  Small geographic area, created by the Bureau of, Is Census part of Commerce?  divides the us into georgraphic subunits so they can count the population.
How does that relate to Tiger files and shape files?   Tiger files are essentially US Census bureau's shape file.  Shape files are privately owned files - company owns about 65% - just a brand name like Microsoft word.  Other software can read shape files.  I use maptitude which can read shape files.
Can you imagine doing this with paper and pen?  No.  laughing.  But I can't imagine doing a lot of things wihtout computers.
Two different programs - White saying something in bakground.  Other type Arcview software.  Board is using Arcview and Div. of elections is using maptitude.
Wallerie:  Let's fire up the computer to see what you're talking about.
Lawson:  looking at different plans - this is hard to follow just the oral
This is the proclamation plan the only thing I moved out is the southern range connected to Ft. Wainright.
W:  Can you show us what a census block looks like?
Lawson:  They're really small, hold on.  OK each one of these is a census block.  I'll highlight one.
BASICALLY - Lawson is showing how to use the software.  Some of these blocks are only 19 people, or there could be only one or none.  the other thing that maptitude does well is to show me what happens when I make changes - it shows me that if I take this block out I will take out 24,  it shows me the ideal size 17,755, and what the deviation is from the ideal size.  We can see here if the census block doesn't distinction between Native and Indian so IND represents both.  Here no one identified self as black or Indian, but in this district there are some.  You can also see whether over 18.  If we move this block out we'd take away 18 voting age people from the district.

I've only added Indian and Native, but there are lots of different potential breakdowns.  They are really numerous.  Other states would have % as non-hispanic black and hispanic black.
Wallerie:  For Natives what does VAP stand for.
Lawson:  Voting Age Population.  Any percent Native - White Native Hawaian, and Other - that person could be counted in Native.
Native +one =  full blooded native or Native + White in the past.  We were reasonably sure from the past that Native+one  (either full blooded or Native+1 other).  But later the board said any combination with Native would count.
Wallerie:  What other attributes can you key in on the map?
Lawson:  You can see, this is airport - I don't have street names turned on because there would be so many of them - Outline of city limits of Fairbanks.  Generally follow city limits.  You can cut smaller than census blocks because you wouldn't know how many people there are.  Most are inside the city boundaries.  There are a couple places because there was an annexation after the Census bureau drew their maps.  So it doesn't really match because annexation after the shape files drawn.

This pink are is downtown Fairbanks.
Wallerie:  Border of 5 and 4.
Lawson:  I can move the annexed area back into the city, I can figure out how far I want to go - census blocks, voting districts, or the whole county - depending on purpose or need.
Wallerie:  Demonstration for the court.
Lawson:  I'm going to select these three census blocks.  Even before I move them in, I can check what the changes would do.  I would be moving 253 people from D5 to D4 - that would affect the deviation to 5.065%  - you can see in real time without having to make guesses.
Wallerie:  kill two birds with one stone.  With this district, if you just wan t to move this annexed area.
Lawson:  Not matching with the local boundary because annexed after the map made.  The total people in the block is change of 111 people - 94 white 2 American Indian, one mix of white and Indian.  I cannot break the census block.  With technology you could, but not allowed to unless a zero census block.  You'd have to guess what the results were which is why you can't do it.
Wallerie:  Sometimes you have to move Census blocks even if you don't want to.
Lawson:  Yes.  I'll move a census block and you leave areas not connected with the rest of the districts.  It will pull up area I as two different areas - something is 3/100 of a square mile.  Some are islands.  In district 5 you know there's no islands, so you know something is wrong.  this is how to correct the non-contiguous, so go back and correct them
Wallerie:  You actually have three separate blocsk.
Lawson:  I'm moving them now by clicking on individual census blocks.  It didn't highlight these because they are already in district 4.  Anything touching the circle gets listed.  Circle touched various census blocks.
Wallerie:  How, if we go to Tanana Flats area.
White questions.  Judge says:  You're just trying to demonstrate the software so I understand how it works.  Not working on the plan?
Lawson:  No, and not sure we'll get to the plan today.  Magnifying glass means the elements and features will zoom in.  These are large blocks because there are no people here.  How many people?  No one.
Wallerie:  did you have any questions about how program works?
You can actually go pretty small.  Here there are lots of empty precincts south of the river.
Wallerie:  These islands here?
2:49pm Lawson:  Each island is its own census block, each lake too. Not sure why, but they are.
Wallerie:  OK, I'll close that down.  Wait, how does this show you if a district is compact.
Lawson:  We can go to reports, and one is measure of compactness.  Which report to run or all of them.  The top one is accepted.
Wallerie:  Can you do a compactness test on one of he districts?
Lawson:  You can look at a particular district but it's designed to look at the plan as a whole. 
Wallerie:  You can shut down the computer and then retake the stand.  How are you familiar with the redistricting process?
Lawson:  I went to most of the board meetings.  Missed the first couple and the ones out of town.  I also worked on the plans for the Rights Coalition.  Getting more community input it would be better.  Leadership included:  Jake Metcalfe, attorney Myra Munson, ???  had traveled the state.  I'm not sure I can name them all.
Wallerie:  Most were Democrats?
Lawson:  Not sure, probably a lot were.  I don't know all of them.  A couple were non-partisan, I know because I had to ask them questions about SWest region.  Someone who worked in fish politics said people had ompeting interests adn I should look out for them.
Wallerie:  So not just computer based plans, I guess you were interviewing people?
Lawson:  I guess that's a good way to talk about it.  I called people around the state.  Asked them about things.  Sometimes someone would say, "You can't do that. . ."
Wallerie:  Other groups.  Who led AFFR?
Lawson:  Kay Brown.
Wallerie:  She's your current boss?  Were there differences among Democrats?
Lawson:  We're a big tent group, lots of different issues.
Wallerie:  What about AFFER?
Lawson:  I understand - White Objects, sustained.
Wallerie:  You loaded board plan on your computer too?
Lawson:  Requested them from the Board and some were on their websites.  The shape files.
Wallerie:  This is a report you did on compactness.  Move for admission.
White:  I think all joint exhibits are already ...Judge yes.
Wallerie:  This is two different reports, correct?
Lawson:  One I did and one generated by the computer software.
Lawson:  Two reports computer shoots out.  One on the proclaamation plan and one on the modified rights plan.  Proclamition was final plan the board released.
Wallerie:  This report would tell us what the compactness of Dsitrict 5.
Lawson:  In this one I ran all 8 tests in the Maptitude software including the ??? test.  I honestly don't know if areas of Fairbanks have enough population to actually test.
White:  Protest about his areas of expertise - not on political analysis.
Judge and White talking hard to hear White.  He can talk about data on Reod??
Wallerie:  In terms of District 5 you mention the popualation test and the accuracy.
White object and Judge overrules.
Lawson:  not accuracy of test but how it's applied.  Software says test draws a convex polygon around the district and looks at population in the polygon that is not in the district.  Each test is good for somethings and not others.  What would this test miss.  I cannot say whether or not you should apply that test when applying to Anchorage.  I'd be more comfortable applying it in Anchorage area.
Wallerie:  can you tell whree parts of the population concentrated and other's not?  For example here in these areas ????
Lawson:  pause.  Not that specific thing.  I could look at it, and I know from having drawn the maps so I know where the population is, but I don't think the software could.  I was looking at how the REod test affected by having that area of non-population in it.
White:  now he is doing analysis of politics of this. Judge:  I'll over rule.  I'm the one who will be judging ompactness.  He can draw maps and talk about what he's doing and how he did it.
Lawson:  To explain the Reod test - the perfect score is one.  Zeor is worst.  Draws the smallest circle it can around the district and looks at how much of the area in the circle was out of the district.
Wallerie:  What wold the .39 mean?
Lawson:  The mean of the scores for the districts was .37, so this is better than than mean.  Higher mean and lower standard deviation.
Wallerie:  What happens when you take the unpopulated area out?
Lawson:  Score drops from .39 to .25 (?)  beyond standard deviation.  Less compact by significant level, right on the border of standard deviation.  Judge asking questions - two measures.  I was supposed to know this was modified rights plan?  The MRP is actually less compact than the boards?
Lawson:  No, because the districts aren't numbered the same.
Judge:  OK, fair enough.
3:12 Wallerie:
Lawson:  Indicates it's part of military ground.  .... Also a military base.  Two different bases - testing range for Wainright and test firing range for Eilson.
Wallerie:  playing aorund with this area.  Does it make any difference in SEnate pairings where you ???
Lawson:  you could go with 38, 1,2, 5, 6.???  I'm familiare to know which districts they are in, but not what their street addresses.
Wallerie:  What happens if you move this around?
Lawson:  Affects District 5 because only connected by the military range.  District 5 could only be paired with District 3 or 4.
Judge:  You have to put the empty area somewhere.
Lawson:  Where you put the empty area affects the pairing.  If that empty area is not in 5 it changes everything and two senators wouldn't be paired.  If empty area were in 38, you'd 3:16  have to pair 5 with 4 or 6 (I think that's what he said.)
Judge:  how does a 3rd party not connected to the board understand.  How does anyone know what the benchmark is without .... no one knew what the benchmark was until very late and in fact when the benchmark came down finally was after when the board finished their plan was Thanksgiving.  How did people know what the benchmark was  - the rest of the population had to come in with plans that met the benchmark, but people didn't know what the benchmark was.  Dr. Arrington agreed  .....?? in the records of the Baord meeting on May 17 Dr. Handley  ..
Judge goes back to to Lawson.
White, modified benchmarks  ... can't quite hear.
Wallerie, it's a very clear issue.  Judge:  you can use your time wisely, but I'm looking for the connection.
Wallerie:  When did you first submit plan to the board.
Lawson:  Late March.  then guessing on benchmark based on previous redistricting.  Looking at Majority and minority disticts.  In May Lsia Handly made presentation to board and said drop Majority and Minority districts and start using effectiveness.
May 17 teleconference?  I think we needed 4 3ffective house districts and 3  and Senate districts and given a week to change that.
Judge 4 effective house and 3 ffective Senate districts.  How did you know what was effective district?
Lawson:  4.17 VAP  except fo district 6 which neeeed just under 50% 49,something and 39 that could be much lower about 39%.
Wallerie:  When did you learn that Dr. Handley was not using that standard of 4 and now was using 5.
White:  I'm confused.
When did you learn benchmark was now 5?
Lawson:  Maybe in late November.
Wallerie:  whole idea of influence districts, you undertand benchmark included influence districts and mapping on that assumption from the board?  Yes.
Five wasn't until final plans submitted?
Not until December of 2011 when I read Arrington's report and I saw the board was saying five was the number needed.

Wallerie:  What were you trying to do with the Rights plan.
Lawson:  Respect for local boundaries.  Should break local boundaries only if you had to.  We wante to show you could meet the VRA without breaking the city and borough boundaries.

Wallerie:  For senate pair - what's the population needed?
Lawson:  33,000 and some was the number for Fairbanks.  We tried to build one that was just the city of Fairbanks.  We tried to draw what was left of the city in a single district and pair that with the city house district into a Senate district.
Wallerie:  How many house seats can you fit in FB borough?
Lawson:  5.5
Wallerie:  How many in Matsu ?
White - object, Judge, where are you going?
Wallerie:  How things different in different parts ????
Judge - ten minute break, back at 3:40 and done at 4:30 on the dot. 

No comments:

Post a Comment

Comments will be reviewed, not for content (except ads), but for style. Comments with personal insults, rambling tirades, and significant repetition will be deleted. Ads disguised as comments, unless closely related to the post and of value to readers (my call) will be deleted. Click here to learn to put links in your comment.