Alaska Redistricting Board Meeting - Compactness v Deviation And Other Issues

I had trouble with the online broadcasts yesterday and today again they kept dropping off.  So I got to the Board Meeting around 10:30.  Here's an overview of the post:

OVERVIEW

• My observations of the day
• What they were doing - mapping Matsu, Kenai, and Anchorage for initial proposal due Sept 10
• Issues that Came Up 
• Census Block Issues - 45,000 blocks in 2010, only in 2020
• Balancing the State required criteria
• Socio-Economic coherence
• Compactness versus Deviation
• Aesthetic compactness versus meaningful compactness
• Deviation matters
• A few other things to consider
• My very, very rough notes of the meeting

My Observations

What they were doing - mapping Matsu, Kenai, and Anchorage for initial proposal due Sept 10

Background:  The State Constitution gives the Boar 30 days from the day they receive the official Census data from the Bureau of the Census to come up with the first draft proposal to present to the residents of Alaska.  This time the Bureau sent out data on August 12, but said it wasn't official yet.  Then later they said it was official.  Thus the Board lost a week or more while they were waiting for the official data.  

They didn't totally lose that time because they were training on the map software and the Board's staff was already working with the preliminary data (that turned out to be the official data).  But they were behind the 2010 Board which had their first big public meeting on the day they got the data on March 15, 2011. 

So the Board has tried to find a way to divide the parts of the state and have different members work on mapping them.  Apparently, yesterday, they agreed on much of the state, at least preliminarily, because today they were focused on Matsu, Kenai, and Anchorage.  

When I got to the meeting they were comparing different versions of Matsu prepared by Board Member Bethany Marcum and Member Nicole  Borromeo.  Later they moved on to Kenai which Board Member Budd Simpson had worked on with Executive Director Peter Torkelson.  And then they started looking at Anchorage - again maps from Bethany and Nicole.

Finally there was a Public Testimony - two people attending the Anchorage hearing live spoke - Robin  O’Donahue and Dave Dunsmuir  both from a group called Alaskans for Fair Redistricting that has  union and other support.  

Issues - Census Block problems

In the 2010 Census, Alaska was divided into 45,000 or so Census Blocks - the smallest units of geography as I understand it.  The 2020 Census divided Alaska into 28,000 Census Blocks.  So there are now fewer, but larger blocks for the Board to work with.  

Let me give you an example from Matsu - though for this example it doesn't matter where.  I was using the free public map making software available here on the Board's website.  

On the map below, the grey lines outline Census Blocks.  Some are much bigger than others.  Until you pick a block and add it to your map, you can't tell what the population will be.  So, I have picked  one fairly large block - colored in with blue.  On the right side you can get data.  Circled in red is where the population data goes.  But as you can see there is nothing there.  That means this census block has no population.  I'm guessing it's something like the Palmer Hay Flats.  It's a huge block with no population. 

Next, I've added a much smaller block - circled in red.  In the population column we've added 49 people.  

Now I've added to more small blocks (in the red circle) and now we have 261 people total of the four blocks.  

And now I've added this big squiggly block which added another 179 people for a total of 540.  

Now I've added one more block smaller than the last one (again circled in red) and that added 72 more people  and got the total to 712.

The point is that these census blocks are all sorts of shapes.  The Board staff says they're not sure how the blocks were drawn.  Small ones can have more people than much larger ones.  So when members are trying to make a smooth compact district that is as closely to the others in population, they are stuck with these odd shaped blocks that can make a district look weird to the average person.  Why is that jagged edge sticking out?  Well the answer could be - it was the only block we could pick that would get the right number of people into the district.  Remember, every district needs to be as close to 18,335 people as possible. (Total population divided by 40 districts.)

This final map looks at a much larger area than the previous maps.  You can barely see the districts I marked.   I did this to make the point that in the big picture, the blocks I picked out amounted to only 712 people out of a district that will need  18,335 people.  (This area which is mostly Matsu will have about six districts.)  And the Board will have to create 40 districts.  The real problems are where one district meets another.  That's when this sort of detailing has to happen.  

This is a very tedious process.  

Issues:  Balancing the State required criteri

• Socio-Economic comparability,  compactness, contiguity, and deviation (how much a district deviates from the 18,335 number that is the ideal size per district (total population divided by 40 House seats.)

The State Constitution requires the districts to have socio-economic compatibility.  That means the people in the district should have interests in common - they live in the same town or city or borough.  They face the same geographical and economic issues. This might be indicated by where they shop (they're was testimony in an earlier meeting about "people in our community got to the Fairbanks Fred Meyer, not the one in Palmer" and "cold for us is -40˚, not 10˚. Basically a representative from this area would be representing people who face the same issues.  So the Board is trying to get rural communities together and urban communities together.  Matt Singer, the Board's attorney told them today that Court cases have agreed that people in the same city and borough have socio-economic compatibility. Everyone in Matsu meets this.  Everyone in Anchorage meets this.  And when Member Bethany Marcum wanted to add a little bit of the northern Anchorage area into Matsu, Singer said the courts had said Matsu and Anchorage have socio-economic compatibility.  

I don't think that's a problem when you include places like Eklutna with parts of Southern Matsu.  But it would be more troubling if you were adding Sutton to Fairview (which would be hard to do because they're so far apart.  

But I also think we have a new measure for whether boroughs have socio-economic compatibility - vaccination rates.  Anchorage, as of today, has 65% of people over 12 with at least one vaccination.  Matsu is 44%.  That means Matsu has 66% unvaccinated!  That's a big gap that probably says more about compatibility than a lot of traditional measures.  

Issue:  Aesthetic Compactness versus Meaningful Compactness

Members Bethany Marcum and Nicole Borromeo both made maps for Matsu and for Anchorage and when they presented them, the two gave different weights to these two criteria.  Marcum was constantly trying to get rid of gaps and protrusions.  

These protrusions I've circle in red from this map of Matsu is what I mean.  The ideal of compactness means that you wouldn't have any of these things sticking out.  The ideal maps would be a bunch of square districts.  But topography is shaped by rivers and hills and marshes.  And populations aren't evenly divided into neat squares.  

Pennsylvania 
districts

The point of compactness in part is to make sure people in a district are compatible.  It's easier for a representative to get around the smallest possible district.  But in Alaska with its small population and huge physical size makes this impossible.  We have districts off the road where representatives can spend thousands of dollars to get to a remote village and districts where the representative can walk from one end of the district to the other in less than a day.  

But another key reason for compact districts without weird protrusions is to prevent gerrymandering.

On the right are silhouettes of some Pennsylvania districts.  These were designed in 2010 to tilt a state with more Democratic than Republican voters so that had majorities in most of the districts.  You can do this by cramming most Democrats into a few districts so that Republicans are competitive or have a majority in all the others.  But to do that you have to make very weird districts.

The protrusions on the Matsu map are there because of geography, population, and odd shaped Census blocks.  Getting rid of all the bumps makes the maps more aesthetically pleasing, but doesn't have a meaningful outcome.  Alaska's warts - at least on the maps we saw today of Matsu are not an issue.  

This link takes you to 538's page on Pennsylvania's extremely uncompact districts.  

Issue: Is Compactness More Important than Deviation?

The deviation standard derives from the US Constitution's requirement of one person - one vote.

Certainly it could be.  But in the maps we saw today, the answer is clearly no.  Deviation refers to deviation from the ideal district size of 18,335.  If all 40 districts have exactly that many people, then every representative and every senator would represent the exact same number of people.  Of course the Census' state population number was measured in 2020 and the population changes all the time (which is why the Census counts every ten years, so we can recalibrate.)

The deviation of any one district should not be more than 10%, and that's would be acceptable only in extreme cases.  Say a sparsely populated rural area where it was really hard to get enough people without grossly violating the other criteria of compactness, socio-economic, and contiguity.  (Contiguity didn't come up today.  It just means that all parts of the district have to be geographically connected to each other.)

The issue was that Matsu was about 2% under populated.  That means that six districts (Matsu has enough population with the Denali Borough to have six districts almost) will have the same number of representatives as six other districts but with a smaller population.  Two percent for any one district is well within the standards, but if you take a whole region that elects candidates of one party, it means other districts that are overpopulated aren't getting one vote per person.  It wasn't clear if the six districts in Matsu were cumulatively underpopulated by 2% or each district was 2% underpopulated. 

Issues - A Few Other Things to Remember

The maps the Board is now drawing will result in a proposed map for Alaska.  There will be several other maps submitted by other groups.  Then these maps will be shared with people around the state.  The Board will split up and travel around to meet with people and find out the issues that these maps present that are problems for local areas.    They have sixty days from September 10 to then develop their final proposed map.  

So there will be lots of feedback and lots of work before we get to the final proposed map.  Then they have to do things like pair up the 40 2House districts into 20  Senate districts.

NOTES

I did my best to record what people said.  I'm a reasonably good typist but I'm not court transcriber.  So there are lots of gaps and ??? where I wasn't sure.  And it's not verbatim.  But it will give you an idea of what was said.  I tried to put times in now and then.  When the tapes and transcripts come out, you  may have a better idea of what you're looking for.  

Ak Redistricting Board  Wed Sept 8, 2021

Arrived at 10:50

Taking about Denali Borough - only 1700?? People, has to be part of another district.

Which of the three Matsu’s should we use.  

Peter Tolkerson (Executive Dir)Matsu B would be under about 2%.  Reflects deviation  of Borough’s ??

Budd Simpson - all the community boundaries and local boundaries, ask Nicole that some of hers were not as tight. Otherwise easier to stick with these more compacts

Nicole - accepted city boundaries and didn’t let the rural creep into the city boundaries

Bethany - all had city boundaries in tact, but let some of rural in

Nicole - some rural areas  were put into cities

Bethany - you’ll see Wasilla and Palmer - green in between is a question of whether you add these to one side - core Matsu, suburban, not considered rural areas

Still feel Bethany’s version maintains socio-economic goals, cause hers seems more compact

Nicole - I think  our maps have more in common than not, she’s just gone out in a different direction to pull people in.  This is my first time looking at Bethany’s map.  Preliminary view I don’t see pitfalls, I get it, it looks prettier.

Peter:  This is a starting point and make adjustments from there

Nicole - wanted to be respectful of public testimony.  Heard from Delta Sutton area feel they are pulled in and they become Matsu Palmer seats and lose their distinction.  

Nicole - Way to overlay my map over Bethany’s?  

Peter:  Map overlay  [I had trouble figuring out what the map overlays showed]

John:  Within borders of borough, socio-economic ok, but next looking at compactness and one vote.

Nicole - I think my deviations are smaller than Bethany’s

John:  But compactness becomes important.  The more irregular looking, you potentially get into odd shapes that raise questions about compactness.  Within a percent or two of deviation, I think compactness becomes more important.  

Wraps around Wasilla

Nicole:  I see hers is more compact, Maybe I’m not seeing something.  Move on.

Nicole:  I’m fine with Bethany’s version.  And Anchorage creeps up into Matsu.

John:  When we put Denali Borough in, that won’t happen. 

Peter:  20 minute exercise to adjust it in, then wave of changes will probably touch every district.  

Palmer is one, Wasilla is one and Big Lake is one.  

John:  Let’s stand at ease.  20-30 minutes to integrate this into draft we’ve been working on.  Stand at ease.  

Break:  Bethany:  I like both of those better than the Anchorage stand alone.  Move ER, Eklutna up toward Matsu

11:16

Peter break  

45,292 census blocks to 28,000 blocks - makes fine tuning difficult.  Members want to add a little section and get some weirdly shaped block.  

11:42  Back in session

TJ:  Overview of Bethany’s map - added Denali B to Matsu, six districts 15, 16, 17 18, 19, 20

Added Anchorage north, but didn’t fill that out yet.  

SE, North, SW, Bristol Bay,  some discussion of inside Fairbanks, Finished Matsu area.  Haven’t discussed Kenai P. Or Gulf Coast, and Whittier is without a Borough.  

Peter:  Remaining:  reconcile new deviations by taking out Nenana and ????  - deviations have increased, B can refine those now or we can come back later for that.

5.92 districts - all will be under.  18 suffered loss when let Nenana go.  Equalize that loss when dealing with other district boundaries.  Does 18 abut to the north?  

That’s the Denali B line.  Stops around Houston.  

Bethany:  Where does Denali B. stop.  

Right here.  

John:  Greatest deviation?  A:  18 down 581  - they are all negative - [meaning they have more representation]

[They’re adding blocks to make things look cleaner - that is there aren’t strange spaces breaking up an area.  

19 and 20 deviation spread]

Bethany - we have that nice straight line though

Nicole, but if 

Made little change, doesn’t mess with compactness much.  Now they are just grabbing population and there is no consideration of neighborhoods being discussed.

John:  Let’s move on

Nichole:  Would like to see those numbers closer.  15 is too big.  Compared to 

Debate between Bethany and Nicole - Deviation is not part of the big three, 

John:  First look at socio-economical compatibility, then compact, then 

Melanie - we can still do this later.  

Attorney:  the change didn’t change compactness, but got more even deviation.  If you have odd shaped districts not good.  [Compact doesn’t mean smooth]

Peter:  

Bethany:  My preference.  Recommended not break B boundary in more than one direction.  South, including Whittier, still breaking to south

John:  I don’t recall that being critical importance

Bethany:  No problem with South Knik

Attorney:  Courts have said that Matsu and Anchorage are socio-economically the same.

Bethany:  Include Anchorage

Nicole:  John made strong argument for strong B boundary and here we have great deviation

John:  Difference between preservingAnchorage and Matsu with protecting Fairbanks

Nicole:  I don’t think so.  

Budd:  800 people leave in Anchorage or better used in Matsu.  

Peter:  word Better is something you can decide.  Anchorage is less underpopulated than Matsu.  

Budd:  Better area - it’s not in Anchorage Borough, nice line 

Bethany:  Equalization as opposed to deviations.

Attorney:  First is to have compact and then obligation one person one vote.  Where you feel that closer to equalize, create districts that are no longer compact.  

Bethany:  That last ??? completely changed compactness, but didn’t effect deviation that much.  Won’t be possible in Anchorage.  

Nicole:  Explain to me how that compactness.

Nicole:  The way that city is drawn is part of the problem.  

Melanie:  Time for a break, need to walk around even if lunch not here

12:11 

John - back, about 1pm  get finished by 3.   Lunch supposed to be here by now.

12:25 - everyone back in seats and discussing maps - CLEARLY this has become an illegal meeting and Peter is trying to scatter the Board members.  Several leave.  Now Budd is talking about coconut oil.  

Lunch arrived about 12:50

Back at 1:18

John:  Lunch late being delivered.  This room will have 3pm public meeting so we will get out by 3:45.  Should we move on to Anchorage?  Sounds like it.  Let’s move on

Peter:  Good to move to Anchorage, but we haven’t done Kenai and Anchorage will probably take a long time.

John:  Sounds ok.  

Peter:  Kenai Pen - goes to Yakatat Borough, Cordova.  Few members and I worked on this all day after last meeting.  Added additional members throughout following week.  Discovered the Cordova and Kodiak and NE Kachemak Bay have long time ties to Kodiak.  

Looking at 11, 12, 13, and 14.  Kenai-Cordova  3.3, 2.4   with Kodiak 4.09 districts.  Each is overpopulated by 1.2 to 2%.  Trying to spread that across the district.  Without Kachemak silo area, you can take city of Seward and give Kachemak back to Homer.  But then with Seldovia etc.  no one was happy.  But adding Cordova and Kodiak allows deviations that are reasonably close.   

This is just a starting point.  Could put Cordova with interior   or connect Seward and Homer.  Population comes out.  Kasilof Nikiski other options.  

Budd:  Comment, Peter and I spent a lot of time on this, seems really difficult coming up with rational explanation.  I think this is an elegant solution to difficult district.

Bethany:  Shape of appendage would be better if fixed.  Doesn’t change deviation.

Peter:  Want to improve that.  But hard to make a cleaner line.  Feature of underlying census geography government made.  

Melanie:  Socio-Economy pitfalls of this?  

Peter:  Critique Katchemak silo area, connections to Homer.  Not violating city boundaries, but just connections.  

Eric:  Fox River ???  Takes in several areas.

Bethany:  Part of map - 

Eric:  Katchemak Silo, ??? And ???  Not census blocks but are communities.

Bethany:  

Matt - North Kenai paired with Anchorage and that has passed Court rulings.  This Board seems to be going toward community boundaries.  Court has approved of Gulf Coast districts in the past.  Don’t see any show stoppers here.

John:  Let’s move on to Anchorage.

Peter:  How should we proceed.  We can show member versions.  

John:  Nicole?

Nicole:  Mine doesn’t work any more because I included Valdez.

Bethany:  I have one version.

Peter:  We can show the differences on the screen.  But merging, I think we should do tomorrow - long technical process.  Nicole’s Anchorage for her Northern portion 

TJ:  Bethany’s 

??  Bethany Markam on the left.  

Bethany:  Started with existing districts.  Then realized having existing districts in place misleading and then just turned them off.  I brought a map I’d worked on then kept getting odd shapes and still had a few appendages.  Let’s take this pop north and things would add up and get weird shape.  Then take it east and west instead.  Just by virtue of census block shapes that led to vertical rather than horizontal.  

Nicole - 

Bethany:  split east side and military.  Services and previous elmendorf and Richardson, split so they are socio-economic, not a big difference.  But also ER, trying to keep ER complete in one district instead of splitting it up.  Encompass ER, Chugiak, Eklutna - get more rural areas versus core of ER.  

TJ - two different maps - also Wasilla showed overlay one plan over the other.  Bethany versus Nicole Blue lines see differences.

District down 2%

Chukiak, Peters Creek, and ???  792 in there add to -256, added back to Matsu map.  

District 13  S. ER into one district

15 takes north and south forks with portion of east Anchorage.

Chugiak/Peters and majority of Fort Rich

Rest is self explanatory.  Nicole’s map on top

Nicole  - when I drafted didn’t have district boundaries and considering public testimony.  Didn’t get enough pop to make whole.  Wanted Chugiak, Peters whole without ER came down to border of ER proper and cover ER to be in own district.  Got much of base.  But service members all over community.

Inlet and Govt Hill (not Mt View) and tried to keep east districts in tact and not go into south Anchorage.  Census blocks made big differences.  West 17 and 15 are primary mid town districts and try to keep them east of Minnesota.  Then all the way over to Turnagain, Lake Hood, Spenard, then to Sandlake.  Follow creek boundaries.  14 needed to go into 15.  13 didn’t love this shape.  Once you get to end of mapping, have trouble.  Also, the way Census blocked this.  Some very wild census blocks.

10  big chunk of south Anchorage and hillside.  District 9 really deviates from map Board is producing.  Don’t look at 9.  Kitchen sink district.

John:  Hard for me.  I just don’t have an opinion on this.  Maybe Nicole and Marham work together and see what you come up with for those 16 seats.  OK    Do you want to take 20 minutes now.  We still have public testimony to hear.  In 20 minutes then get wrapped up to adjuourn.

Nicole:  I don’t think that’s realistic.  I think there’s a benefit for everyone participating.  

Fix Boundaries for Matsu and then 

Bethany:  Are we going to….

Matt Singer:  Going back to whether the 800 area of Knik.  AK Supreme Court said the two Boroughs are Socio-economically integrated.  If so, is moving those folks into Anchorage area to allow for less deviation is that practical.  We don’t combine communities where creates map that is not Socio-economically integrated.  So can’t use that as a strong justification for drawing the line.  Allow some blurring of line to get more population evenly distributed.  I know board focused on honoring political boundaries, but really one and the same. 

Peter:  Matsu districts are more underpopulated than Anchorage so moving that area to Matsu  improves the deviation.  Looking at the 22 seats rather than the 16 seats of Anchorage.

Matt:  I think that makes sense.  Try to balance pop among those 22 districts.  

Peter:  Will cause bringing Matsu district down into Anchorage because underpopulated.  

Bethany:  I’ve seen the court ruling that says Matsu and Anchorage are one socio-economic districts.  

Eric:  Anchorage, Denali Borough and Matsu together about 22 districts.  Every district down about 165.  

Peter:  But it will come up against compactness thing.  Still place for judgment.  We’re going to go south and get greater and greater underpopulation

John:  Let’s see what happens.  Let’s start at 18?

Matt:  Fair to say that board is being oversensitive to compactness when dealing with urban areas.  What’s a problem is weird corridors and appendages.  Intended to combat improper gerrymandering.  I’d say within these areas.  Is this generally a compact shape.  Does it make sense.  Not get overly caught up on jagged edge.  That’s a census block shape.  Wouldn’t concern yourself with that granular detail.  What’s troublesome, if you just focus on the numbers you can say, that looks like starfish or octopus.  Then courts say why.

John:  For the purpose of deviation we’re making them more odd shape.  Is it practicable to make odd shapes to get better numbers.  

Matt:  One side of the street or other side of street - does this make sense for the neighborhood.  If you’re down by 55 or 100 people that is a small deviation.  Good to make numbers match.  No court says you need exactly.  Why having computers do this versus experienced Alaskans.  A little more deviation ok.  

John:  Maybe this is a good time to take a break.  Need to get into public testimony.  Stop mapping for now.  Testimony until we need to vacate the room at 2:45.  

Testifying

Robin O’Donahue - Coordinate for Alaskans for Fair Redistricting and as life long Faribanksian.  

But Nicole’s question.  The split in Borough breaks community interest.  Does’t look at North south.  Ester, etc. connected to UA.  North Pole and ?  Are together.  Another way to think through school systems.  All feed to West Valley High School.  Don’t believe Chena Ridge and Salcha, you’d have to drive through all the others to get to that district.  Thank you.  

Nicole:  Is there anyway we can see your Fairbanks before next week?  If we have it in our binder before we finish.  

Robin:  Not bound to idea of keeping the Boroughs intact.  

John:  If asking AFFR giving us map early, then we should discuss with other groups.  

Peter:  If email map of just Fairbanks, in line with other groups

David Dunsmore:  Also AFFR, respectfully push back about historical ties of Kodiak to Kachemak .  The Old Believers split from New Believers.  Faced severe suppression and forced to Siberia.  Kodiak Island settled by Russians in 1784.  That split almost like Catholics and Protestants in Ireland.  The alternative would be Ninilchik.  Fox River has 634 Ninilchk has 845??  Ninilchik not a city but a census area.  Founded by Russians in 1700s.  

Peter:  Suggesting that old believers came over separately from Russians.  OB came in 1968 and founded Nikolias they shouldn’t be split into two different districts.  Stayed in Siberia until Soviet Union was oppressive, they moved to China, then Oregon, then Alaska. 

John:  There’s a good book on that 

Melanie:  Tried to find AFFR online  AKfariredistricting.org  Coalition of different orgs and individuals across the state.  AFL-CIO.  We’ve been internally meeting with Alaskans.  

2:39 - John:  No more testimony here or online.  Stand at Recess.