Idea for Redistricting Reform

[kwabbit]

The idea is for there to be larger congressional districts, each with about 3 times the population, and each with 3 representatives. For example, PA would get 6 districts, OH would get 5, NJ 4, TX 13, CA 17, and so forth. For small states they would have a single district that would elect less representatives. So HI, ME, NH would have a single district that would elect 2. WY, AK would have a single district that would elect one.

There would be a jungle primary with the top 4 advancing to the general election. This would generally be 2 Dems and 2 Reps, but would depend greatly on the partisan lean of the district and the ability of the party to clear their field. In the general election, each voter selects 3 candidates. So in essence, they would select the two candidates of their party, and their preferred candidate of the opposing party. In really lopsided districts, this might end up with 3 Dems or Pubs but in a PVI <15 district it would end up with a 2-1 split. This process repeated over a whole large state would end up with pretty proportional results. For NY it was 15-9 Dem, For NJ it was 7-5 Dem, For OH it was 9-6 Pub, For PA it was 10-8 Pub, for TX it was 20-19 Pub. This gets much closer to proportionality and encourages more consensus candidates.

[President Punxsutawney Phil]

Interesting idea.

[Sol]

You could just do full on PR, though in that case having even larger districts might be preferable when possible.

[kwabbit]

You could just do full on PR, though in that case having even larger districts might be preferable when possible.

The end result is pretty similar to PR, but this finds some common ground between proportionality and having representatives from a specific area. Like an 'Upstate New York' giga district might result in a favorable 5-5 proportional outcome, but then the representatives are not really representing anyone in specific.

[President Punxsutawney Phil]

You could just do full on PR, though in that case having even larger districts might be preferable when possible.

The end result is pretty similar to PR, but this finds some common ground between proportionality and having representatives from a specific area. Like an 'Upstate New York' giga district might result in a favorable 5-5 proportional outcome, but then the representatives are not really representing anyone in specific.

That is true. And it's a pretty clever way of getting there.

[President Punxsutawney Phil]

Would you mind if I did my interpretation of this idea?

[kwabbit]

Would you mind if I did my interpretation of this idea?

Be my guest

[President Punxsutawney Phil]

Would you mind if I did my interpretation of this idea?

Be my guest

So here are my specifications for my take on this.
Since it's no longer SMDs, the VRA is much, much more loose, if it applies at all.
Deviation for districts can also grow to a maximum of 5% from quota.
Likely 2020 district apportionment will be used, along with 2018 population apportionment.
No attempts will be made to try to create proportionality. All maps ought to be neutral in intent.
When not enough members are assigned to a state for it to be all 3-seaters, then a 1-seater or a 2-seater will be created to fill the gap. The number of 3-seaters created will be the maximum possible in all circumstances. If a 1-seater or 2-seater is formed, then it will be drawn in an area and in a manner that allows for whole counties and/or a reduction in overall county splits.
As per my usual preference, counties will be the most preserved sub-unit.

Do these seem agreeable?

[kwabbit]

Would you mind if I did my interpretation of this idea?

Be my guest

So here are my specifications for my take on this.
Since it's no longer SMDs, the VRA is much, much more loose, if it applies at all.
Deviation for districts can also grow to a maximum of 5% from quota.
Likely 2020 district apportionment will be used, along with 2018 population apportionment.
No attempts will be made to try to create proportionality. All maps ought to be neutral in intent.
When not enough members are assigned to a state for it to be all 3-seaters, then a 1-seater or a 2-seater will be created to fill the gap. The number of 3-seaters created will be the maximum possible in all circumstances. If a 1-seater or 2-seater is formed, then it will be drawn in an area and in a manner that allows for whole counties and/or a reduction in overall county splits.
As per my usual preference, counties will be the most preserved sub-unit.

Do these seem agreeable?

Yeah those seem pretty reasonable. When I made my quick maps to see what would happen, some states, like PA ought to have 5 3 member and 1 2 member rather than 6 3 members.

The idea was to allow for a lot of population deviation to keep counties whole and districts compact. But when dealing with very populous counties like in NYC, Westchester, Long Island, keeping counties whole with equal pop is hard. Like Nassau + Suffolk is way over but it makes sense as a district. Surprisingly, Brooklyn and Queens work as their own respective districts without pop too much deviation, as does a Bronx + Westchester district.

Maybe it can be a patchwork where some 1 member, 2 member, and 3 member districts are created to work with the political geography, but 3 members have the highest priority.

[President Punxsutawney Phil]

Would you mind if I did my interpretation of this idea?

Be my guest

So here are my specifications for my take on this.
Since it's no longer SMDs, the VRA is much, much more loose, if it applies at all.
Deviation for districts can also grow to a maximum of 5% from quota.
Likely 2020 district apportionment will be used, along with 2018 population apportionment.
No attempts will be made to try to create proportionality. All maps ought to be neutral in intent.
When not enough members are assigned to a state for it to be all 3-seaters, then a 1-seater or a 2-seater will be created to fill the gap. The number of 3-seaters created will be the maximum possible in all circumstances. If a 1-seater or 2-seater is formed, then it will be drawn in an area and in a manner that allows for whole counties and/or a reduction in overall county splits.
As per my usual preference, counties will be the most preserved sub-unit.

Do these seem agreeable?

Yeah those seem pretty reasonable. When I made my quick maps to see what would happen, some states, like PA ought to have 5 3 member and 1 2 member rather than 6 3 members.

The idea was to allow for a lot of population deviation to keep counties whole and districts compact. But when dealing with very populous counties like in NYC, Westchester, Long Island, keeping counties whole with equal pop is hard. Like Nassau + Suffolk is way over but it makes sense as a district. Surprisingly, Brooklyn and Queens work as their own respective districts without pop too much deviation, as does a Bronx + Westchester district.

Maybe it can be a patchwork where some 1 member, 2 member, and 3 member districts are created to work with the political geography, but 3 members have the highest priority.

So I need just one more piece of the puzzle before I can start - what is the best guess at the eventual 2020 apportionment that we have?

[President Punxsutawney Phil]

using this projection from Muon, for eventual 2020 changes:

AL -1
AZ +1
CA -1
CO +1
FL +2
IL -1
MI -1
MN -1
MT +1
NY -1
NC +1
OH -1
OR +1
PA -1
RI -1
TX +3
WV -1

total apportionment:
Alabama 6
Alaska 1
Arizona 10
Arkansas 4
California 52
Colorado 8
Connecticut 5
Delaware 1
Florida 29
Georgia 14
Hawaii 2
Idaho 2
Illinois 17
Indiana 9
Iowa 4
Kansas 4
Kentucky 6
Louisiana 6
Maine 2
Maryland 8
Massachusetts 9
Michigan 13
Minnesota 7
Mississippi 4
Missouri 8
Montana 2
Nebraska 3
Nevada 4
New Hampshire 2
New Jersey 12
New Mexico 3
New York 26
North Carolina 14
North Dakota 1
Ohio 15
Oklahoma 5
Oregon 6
Pennsylvania 17
Rhode Island 1
South Carolina 7
South Dakota 1
Tennessee 9
Texas 39
Utah 4
Vermont 1
Virginia 11
Washington 10
West Virginia 2
Wisconsin 8
Wyoming 1

[President Punxsutawney Phil]

Moving east to west and north to south.
New England's one-district states go first.
To start off: Maine is an obvious 1D, 1R, unless an independent with enough clout runs.
New Hampshire is an obvious 1D, 1R.
Vermont is an obvious 1D, unless an independent with enough clout runs.
Rhode Island is another obvious 1D.
Total: 4D, 2R

I assume that the each voter has 3 votes thing is dependent on there being 4 members elected. People would have one less vote than the number of members elected, unless it's a single-member district, with people having one vote.
In two-seaters, you still have 3 people in a runoff, and in a single-member district, only 2 people in a runoff.

[kwabbit]

Moving east to west and north to south.
New England's one-district states go first.
To start off: Maine is an obvious 1D, 1R, unless an independent with enough clout runs.
New Hampshire is an obvious 1D, 1R.
Vermont is an obvious 1D, unless an independent with enough clout runs.
Rhode Island is another obvious 1D.
I assume that the each voter has 3 votes thing is dependent on there being 4 members elected. People would have one less vote than the number of members elected, unless it's a single-member district, with people having one vote.
In two-seaters, you still have 3 people in a runoff, and in a single-member district, only 2 people in a runoff.

Yeah that was the idea. Reapportionment is not set in stone. I've seen a lot of different projections, so I think how thoroughly the census was conducted in each state will make a difference.

[Stuart98]

Moving east to west and north to south.
New England's one-district states go first.
To start off: Maine is an obvious 1D, 1R, unless an independent with enough clout runs.
New Hampshire is an obvious 1D, 1R.
Vermont is an obvious 1D, unless an independent with enough clout runs.
Rhode Island is another obvious 1D.
I assume that the each voter has 3 votes thing is dependent on there being 4 members elected. People would have one less vote than the number of members elected, unless it's a single-member district, with people having one vote.
In two-seaters, you still have 3 people in a runoff, and in a single-member district, only 2 people in a runoff.

Yeah that was the idea. Reapportionment is not set in stone. I've seen a lot of different projections, so I think how thoroughly the census was conducted in each state will make a difference.

Also if the annual census estimates were wrong then the projections could be fairly off.

[President Punxsutawney Phil]

Massachusetts

https://davesredistricting.org/join/0e209659-c174-4544-874e-69f3ad132fdf

MA-01: D+8.68, 55.6-36.3 Clinton 2D, 1R
MA-02: D+19.3, 68.6-25.1 Clinton 2D, 1S
MA-03: D+7.68, 56.5-36.4 Clinton 2D, 1R
Total so far: 6D, 2R, 1S
Total so far: 10D, 4R, 1S

[President Punxsutawney Phil]

Connecticut

https://davesredistricting.org/join/96770991-241e-4f26-aeb2-89cf6d40230b

CT-01: D+5.31 2D, 1R
CT-02: D+8.07 1D, 1R
Total: 3D, 2R
Total so far: 13D, 6R, 1S

[President Punxsutawney Phil]

New York is posing some difficulties. Suffolk is perfect for 2, but Nassau is more than 10% off from being 2. I think I'm allowing the creation of 4-seaters to deal with this, and to allow for more county integrity. However, 4-seaters are only to be used when county integrity will be well-served by them, and if 3-seaters are practical, then they by all means ought to made. I'm also expanding allowable deviation to 7.5% in either direction.

[President Punxsutawney Phil]

New York

https://davesredistricting.org/join/cf5f4c46-2d6b-4901-9156-1e8870c88528

NY-01: R+0.25 2D, 2R
NY-02: D+27.18 2D, 1S
NY-03: D+24.49 3D, 1S
NY-04: D+37.28 4D
NY-05: D+5.93 2D, 1R
NY-06: R+1.03 1D, 1R, 1S
NY-07: D+2.35 1D, 1R, 1S
NY-08: D+1.55 1D, 1R
Total: 16D, 6R, 4S
Total so far: 29D, 12R, 5S

[kwabbit]

New York

https://davesredistricting.org/join/cf5f4c46-2d6b-4901-9156-1e8870c88528

NY-01: R+0.25 2D, 2R
NY-02: D+27.18 2D, 1S
NY-03: D+24.49 3D, 1S
NY-04: D+37.28 4D
NY-05: D+5.93 2D, 1R
NY-06: R+1.03 1D, 1R, 1S
NY-07: D+2.35 1D, 1R, 1S
NY-08: D+1.55 1D, 1R
Total: 16D, 6R, 4S
Total so far: 29D, 12R, 5S

Thank you so much for putting the work into this!! Yeah, NY can be a problem. Nassau is too between 2 districts and 3 districts population wise to be not be an annoyance.

[President Punxsutawney Phil]

New Jersey

https://davesredistricting.org/join/651c2dff-b494-499c-a1c7-4852a6216edb

NJ-01: D+8.48 2D, 1R
NJ-02: R+1.05 1D, 1R, 1S
NJ-03: D+11.88 2D, 1R
NJ-04: D+8.85 2D, 1R
Total: 7D, 4R, 1S
Total so far: 36D, 16R, 6S

New Jersey's county populations, the fact that I didn't want to increase deviation range too much, as well as a desire to avoid having to rely on water contiguity, created a somewhat funky map. Oh well.

[President Punxsutawney Phil]

Pennsylvania

https://davesredistricting.org/join/8b369f3b-80f0-41d8-ab39-9287f3d21fb7

PA-01: D+33.57, 82.5-15.4 Clinton 2D
PA-02: D+4.66, 54.7-41.8 Clinton 2D, 1R
PA-03: R+6.69, 57.1-39.2 Trump 2R, 1D
PA-04: R+7.57, 53.7-41.9 Trump 2R, 1D
PA-05: R+15.95, 65.4-30.8 Trump 2R, 1D
PA-06: R+1.21, 48.9-47.7 Trump 1R, 1D, 1S
Total: 8D, 8R, 1S
Total so far: 44D, 24R, 7S
and that rounds out the Northeast.

[kwabbit]

New Jersey

https://davesredistricting.org/join/651c2dff-b494-499c-a1c7-4852a6216edb

NJ-01: D+8.48 2D, 1R
NJ-02: R+1.05 1D, 1R, 1S
NJ-03: D+11.88 2D, 1R
NJ-04: D+8.85 2D, 1R
Total: 7D, 4R, 1S
Total so far: 36D, 16R, 6S

New Jersey's county populations, the fact that I didn't want to increase deviation range too much, as well as a desire to avoid having to rely on water contiguity, created a somewhat funky map. Oh well.

Yeah, NJ is another state where it's hard to draw without splits. Being only two counties across most of the time as well as large county pops limits easy population equality.



This is what I had for NJ, resulting in 7-3-2.

[President Punxsutawney Phil]

New Jersey

https://davesredistricting.org/join/651c2dff-b494-499c-a1c7-4852a6216edb

NJ-01: D+8.48 2D, 1R
NJ-02: R+1.05 1D, 1R, 1S
NJ-03: D+11.88 2D, 1R
NJ-04: D+8.85 2D, 1R
Total: 7D, 4R, 1S
Total so far: 36D, 16R, 6S

New Jersey's county populations, the fact that I didn't want to increase deviation range too much, as well as a desire to avoid having to rely on water contiguity, created a somewhat funky map. Oh well.

Yeah, NJ is another state where it's hard to draw without splits. Being only two counties across most of the time as well as large county pops limits easy population equality.



This is what I had for NJ, resulting in 7-3-2.

You have a very good point about how only two counties wide is quite limiting. Simply ruling out water contiguity is enough to force my map, because it not only prevents a linking of Hudson and Union counties, it also prevents a linking of Atlantic and Ocean counties by default.

[President Punxsutawney Phil]

Maryland

https://davesredistricting.org/join/3deb6647-ee20-4caa-9031-f68b2248325d

MD-01: D+12.75, 60.5-34.3 Clinton 2D, 1R
MD-02: D+10.84, 58.7-35.2 Clinton 2D, 1R
MD-03: D+13.06, 62.3-31.5 Clinton 1D, 1S
Total: 5D, 2R, 1S


Delaware is an obvious 1D.

West Virginia with 3 members would be a solid 2R, 1D. With it losing 1 post-2020, both parties lose a safe seat and a new swing seat is created. Thus it is 1R, 1S.

Total so far: 50D, 27D, 9S

[President Punxsutawney Phil]

Virginia

https://davesredistricting.org/join/142fde4f-6f04-4a43-8b61-941476a3990e

VA-01: D+1.74, 49.8-44.7 Clinton 1D, 1R, 1S
VA-02: D+2.21, 51.5-42.4 Clinton 2D, 1R
VA-03: R+12.77, 60.1-34.8 Trump 2R, 1D
VA-04: D+16.03, 67.4-25.5 Clinton 2D
Total: 6D, 4R, 1S
Total so far: 56D, 31R, 10S