Fair redistricting: California

[muon2]

We should automatically disqualify AustralianSwingVoter from every state with 3 or more Districts, since he is constantly willing, and trying to violate fair redistricting criteria by trying to put as many County splits as possible all into the same County, and then trying to spin it off as a good thing like this:

Oregon Non-Partisan plan 2 (minimum county splits).

My second non-partisan redistricting plan for Oregon. Only one county is split.

Really dishonest.

It is just a personal choice. I respect the fact that many, like yourselves, prefer to not split counties between more than two districts. However I have a different approach I have previously outlined. I would also point out that the map you have quoted is very similar to muon's second plan.
However I most certainly agree with you that my Second Plan isn't good from a communities of interest view, and I far prefer my first plan. The second plan exists solely to prove that it is possible, but with grave shortcomings.
Also, what do you mean by "Really dishonest." in reference to my statements regarding the second plan of "minimum county splits" and "Only one county is split.". Those statements are true, and I cannot understand how I am dishonest in my making of such statements.

Here are my two plans for OR.

...


Plan B seeks to preserve the Portland UCC and both the pack and cover are maintained. Only Clackamas is chopped, and within it CD 3 picks up all of the city of Milwaukie and the CD 1-5 border follows school district boundaries in the county. The result is a CD 4 that includes almost the entire coast of the state. Though the districts are less competitive they still have the ideal skew of 1 for the state.



CD 1: (-1275) D+8.4
CD 2: (+2218) R+12
CD 3: (+307) D+28
CD 4: (-537) D+3.3
CD 5: (-714) R+6.1

The reaction of other posters to the plans by ASV and myself actually point to something important in redistricting. Many groups have suggested that computers would do a better job than people in drawing plans. Yet a plan like the ones above might be exactly the output of a fairly sophisticated algorithm.

My challenge to those who don't like such plans is to ask how you would define criteria that could be used by an algorithm to avoid a preference for plans like those above. The definition should be objective and not subjective. Also think about the fact that in 2020 OR will likely get a 6th CD and at least one district that spans the Cascades will become a necessity.

[Dr. MB]

We should automatically disqualify AustralianSwingVoter from every state with 3 or more Districts, since he is constantly willing, and trying to violate fair redistricting criteria by trying to put as many County splits as possible all into the same County, and then trying to spin it off as a good thing like this:

Oregon Non-Partisan plan 2 (minimum county splits).

My second non-partisan redistricting plan for Oregon. Only one county is split.

Really dishonest.

It is just a personal choice. I respect the fact that many, like yourselves, prefer to not split counties between more than two districts. However I have a different approach I have previously outlined. I would also point out that the map you have quoted is very similar to muon's second plan.
However I most certainly agree with you that my Second Plan isn't good from a communities of interest view, and I far prefer my first plan. The second plan exists solely to prove that it is possible, but with grave shortcomings.
Also, what do you mean by "Really dishonest." in reference to my statements regarding the second plan of "minimum county splits" and "Only one county is split.". Those statements are true, and I cannot understand how I am dishonest in my making of such statements.

Here are my two plans for OR.

...


Plan B seeks to preserve the Portland UCC and both the pack and cover are maintained. Only Clackamas is chopped, and within it CD 3 picks up all of the city of Milwaukie and the CD 1-5 border follows school district boundaries in the county. The result is a CD 4 that includes almost the entire coast of the state. Though the districts are less competitive they still have the ideal skew of 1 for the state.



CD 1: (-1275) D+8.4
CD 2: (+2218) R+12
CD 3: (+307) D+28
CD 4: (-537) D+3.3
CD 5: (-714) R+6.1

The reaction of other posters to the plans by ASV and myself actually point to something important in redistricting. Many groups have suggested that computers would do a better job than people in drawing plans. Yet a plan like the ones above might be exactly the output of a fairly sophisticated algorithm.

My challenge to those who don't like such plans is to ask how you would define criteria that could be used by an algorithm to avoid a preference for plans like those above. The definition should be objective and not subjective. Also think about the fact that in 2020 OR will likely get a 6th CD and at least one district that spans the Cascades will become a necessity.

In 2020 I’d assume all of Josephine and parts of Jackson would be transferred to a new district from the 2nd, so no significant change in Cascade-crossing would be needed.

[muon2]

The reaction of other posters to the plans by ASV and myself actually point to something important in redistricting. Many groups have suggested that computers would do a better job than people in drawing plans. Yet a plan like the ones above might be exactly the output of a fairly sophisticated algorithm.

My challenge to those who don't like such plans is to ask how you would define criteria that could be used by an algorithm to avoid a preference for plans like those above. The definition should be objective and not subjective. Also think about the fact that in 2020 OR will likely get a 6th CD and at least one district that spans the Cascades will become a necessity.

In 2020 I’d assume all of Josephine and parts of Jackson would be transferred to a new district from the 2nd, so no significant change in Cascade-crossing would be needed.

I understand what you are saying, but it's a subjective statement. But how does one objectively differentiate between connecting population along corridors to the north and corridors to the south?

I looked at the projections for 2020, and Jackson would have to be split such that Medford and Central Point go with the west, while the other suburbs like Ashland, Talent, and White City go with the east. Medford is the fourth largest metro in OR. I don't think metro Salem or Eugene would be treated that way as far as chops. Why is it better to chop the Medford metro between two CDs than to link population along the Columbia gorge to the Portland metro? It's a serious question to folks designing software.

[cvparty]

OREGON ENTRIES

plan A


CD 1: (+444) D+6.6
CD 2: (-1547) R+12
CD 3: (-156) D+28
CD 4: (+743) D+0.9
CD 5: (+515) R+1.0


Plan B


CD 1: (-1275) D+8.4
CD 2: (+2218) R+12
CD 3: (+307) D+28
CD 4: (-537) D+3.3
CD 5: (-714) R+6.1

1. D+7 (Northwest) -936
2. R+10 (East) -472
3. D+28 (Portland) -417
4. D+1 (Southwest) +1339
5. R+4 (Salem/Clackamas) +485

Oregon Non-Partisan plan 1
District 1 D+06.60 - 58.0 - 39.9
District 2 R+12.06 - 42.9 - 54.8
District 3 D+27.64 - 76.6 - 21.3
District 4 D+00.90 - 55.1 - 42.6
District 5 R+01.02 - 52.2 - 45.8

Oregon Non-Partisan plan 2
District 1 D+08.34 - 58.9 - 39.3
District 2 R+12.15 - 42.7 - 55.0
District 3 D+27.79 - 76.8 - 21.2
District 4 D+03.31 - 57.4 - 40.3
District 5 R+06.17 - 48.4 - 49.4

District 1: D+6
District 2: R+11
District 3: D+28
District 4: D+1
District 5: R+2

[Sol]

ID:
scarlet>MB>singletx>cvparty>ASV-A>muon2-B>ASV-B>muon2-A
Districts must be contiguous!!! Also, I prefered maps which put Canyon County with the panhandle as it is a less erose shape. Owhyee is part of the Boise area, so I liked Scarlet's approach.

Also those drawing WA may want to keep in mind that I will preference those maps which have any connection over Puget Sound, whether by bridge or by ferry, extremely low (except to islands obvs). Such maps are intolerably gerrymandered.

[muon2]

ID:
scarlet>MB>singletx>cvparty>ASV-A>muon2-B>ASV-B>muon2-A
Districts must be contiguous!!! Also, I prefered maps which put Canyon County with the panhandle as it is a less erose shape. Owhyee is part of the Boise area, so I liked Scarlet's approach.

Also those drawing WA may want to keep in mind that I will preference those maps which have any connection over Puget Sound, whether by bridge or by ferry, extremely low (except to islands obvs). Such maps are intolerably gerrymandered.

I'm confused by the statement about contiguous, so help me out. I assume you mean connected because all are contiguous.

My muon2 B is admittedly not connected, since there is no road that goes from Lemhi county to the rest of ID 1 in that plan. However, if you put Lewiston ID (northern panhandle) to Sun Valley (the famous ski resort) to Idaho Falls (in eastern ID) into Bing maps it gives you a preferred route that is entirely within the CD 1 of muon2 A. But you rated muon2 A last. I even pointed out the difference between my plan A and cvparty's plan in that he leaves out Boise county is needed in mapping software, so his is not connected. Yet you score cvparty's plan higher than muon 2 A (and muon2 B).

I've included the relevant posts below to illustrate the shift of Boise county with comments about the connection. Please help.

Plan A preserves to Boise UCC, has no county chops, a population deviation within 0.5%, and counties are connected by road, though in one case it is a former, not current state highway.

1: R+15
2: R+23
my map keeps the whole Boise metro area together

Without Boise county there's no way to get from northern to eastern ID. The mountains are as uncrossable as the Chesapeake Bay in MD without a bridge. It has the same problem as my map B.

One of the interesting things to observe here is the types of maps that Atlas produces and how they are judged. The actual map was done by a neutral group. It splits the Boise metro because the northern panhandle doesn't care much for eastern ID and they'd rather be grouped with Nampa and the Boise metro.

Also, more insight on your feelings (and others, too) about ferries would help. I've used the Island county ferry to the Olympic peninsula. It runs every hour, it's big, and it's full (with a long waiting line of cars). That indicates a significant level of interest in people going between the two points. Should the muon rules not count a connection by regular all-year ferry the same as any other state road?

[Strudelcutie4427]

1. D+3
2. D+17
3. R+3
4. R+13
5. R+8
6. EVEN
7. D+36
8. D+11
9. D+12
10. D+3

[KingSweden]

1. D+3
2. D+17
3. R+3
4. R+13
5. R+8
6. EVEN
7. D+36
8. D+11
9. D+12
10. D+3

I actually like this better than the real map. I can see something like this being drawn in the 2020s

[muon2]

Sol hasn't responded yet, but I give him credit for consistency.

Frankly, color me skeptical that ferry routes are the best ways to connect unconnected areas--it seems to me that they should only be used for islands not connected to the mainland by highway, and the like.

In reply to that post was this comment from d32123, a native to the area.

Yeah speaking as someone who lives in Whatcom County, crossing the Cascades in the North is a no-no.  The big issue with mapping Washington north of Seattle is that Whatcom+Skagit+San Juan+Island (which together form a reasonable community with lots of commuting between those regions and such) are nowhere near close enough to a full congressional district on their own and require chopping up Snohomish County, which is bad because that region is essentially just an extension of the Seattle suburbs.  Ferry crossings are weak too imo but should be acceptable if it means crossing the Cascades as little as possible and I think are understandable if it means not screwing up the Puget Sound area.

Those posts are from a thread that went on for about 10 pages on how to deal with the geography of WA. Included are a number of maps of the larger counties showing the boundaries for municipalities and school districts as rendered in DRA. We ended up using that combination as county subdivisions. We also wrestled with the question of whether it was better to keep the more small city and rural northern Puget Sound with suburban Seattle or with the similarly small city rural area across the sound. We did not impose a preference other than chops and erosity.

There are three UCCs in WA. Seattle (King, Pierce, Snohomish) has just over half the population of the state and is equal to 5.1 CDs. The other two are in eastern WA and much smaller: Kennewick (Benton, Franklin) and Wenatchee (Chelan, Douglas).

Here's my connection map. The numbers are percentages of a CD for each county.

[Sol]

Sorry for not responding Muon2, been very busy with final papers and the like.

The cvparty map which I ranked above yours was the only one included in the PM which I received from cvparty--this one:

one smol piece is taken from the northeast corner of canyon
1: R+14
2: R+25

I assume he wanted his other map excluded from consideration. Unfortunately it seems that ASV's maps have disappeared--not sure what happened to the images--but as far as I can recall I ranked ASV-A higher than muon2-A because it had a Boise-Panhandle linkage. Thus my lower ranking of muon2-A is because of shape/CoI concerns rather than connectedness.

I'd argue wrt: ferries that although they may be quite well used for determining connections for isolated islands (i.e. the San Juan Islands, or perhaps some of the islands in Alaska) when there's a road connection available the preference should be to defer to that. A lot of my feeling in this way is based on the example of Staten Island; it'd be very easy for an ambitious gerrymanderer to argue that Richmond County should be linked to lower Manhattan with the Staten Island Ferry, even though there's a much easier link to the city with I-278.

[muon2]

Sorry for not responding Muon2, been very busy with final papers and the like.

I'd argue wrt: ferries that although they may be quite well used for determining connections for isolated islands (i.e. the San Juan Islands, or perhaps some of the islands in Alaska) when there's a road connection available the preference should be to defer to that. A lot of my feeling in this way is based on the example of Staten Island; it'd be very easy for an ambitious gerrymanderer to argue that Richmond County should be linked to lower Manhattan with the Staten Island Ferry, even though there's a much easier link to the city with I-278.

I would counter that when the road connection is many hours longer than the ferry, the ferry is heavily used with an hourly schedule, and the ferry is officially part of the state highway connection, then the ferry should be given consideration. As d32123 says ferries aren't ideal, but they are understandable if used to avoid messing up the Puget Sound. In this case the Whatcom-Skagit areas seem far more like the Olympic peninsula than they do suburban Snohomish. The current map tries to deal with this in WA-01 by linking a bunch of poorly connected inland areas, and if you read the rest of d32123's linked post, he's not happy with that at all. So I think if CoI matters, then a cross sound CD can make sense.

With that in mind here's my plan from that same thread. It started by partitioning the state into groups of counties such that each was very close to a whole number of CDs. Because the Seattle UCC was just over 5 CDs the choice was to split it, or to chop a number of counties - I chose the former. There was also a choice of either the orange or purple line as a partition - I chose the purple line with lower erosity.



The southern region has 2 CDs (3 and 4) and the line was drawn through Yakima county to keep the reservation intact and follow school district lines in the county. It keeps Yakima city connected to the rest of CD 4 by way of WA-24. CD-4 isn't pretty, but it works with the geographical restrictions, cities, school districts and the reservation.

The western region has 3 CDs (2, 6 and 9) and the line there was drawn through Kitsap to follow school district boundaries and minimize the size of the chop. The chop in Pierce also follows municipal and school district boundaries.

The Seattle region has 4 CDs (1, 7, 8, and 10) and the lines were drawn to minimize erosity while keeping cities and school districts intact within each CD. Two CDs are nested within King and 1 CD is nested completely within Snohomish. The maps of the counties (King, Snohomish) showing the cities and school districts are on the older thread.





CD 1: (-2175) D+7.9
CD 2: (-687) D+4.0
CD 3: (-2677) R+1.8
CD 4: (+1258) R+14
CD 5: (+1933) R+7.7
CD 6: (+1971) R+0.9
CD 7: (-163) D+36
CD 8: (-41) D+12
CD 9: (+2519) D+7.7
CD 10: (-1938) D+12

There are 5 uncompetitive D and 1 competitive D (CD 2) districts.
There are 2 uncompetitive R and 1 competitive R (CD 3) districts.
There is one highly competitive tossup (CD 6) district.
With a three seat advantage for the Dems, the plan has a skew of 0.

[AustralianSwingVoter]

Washington Non-Partisan plan.

My non-partisan redistricting plan for Washington. Douglas County is the only unnecessary county split, in addition to the required splits of King, Pierce and Snohomish. Seattle is wholy within the 7th, and the districts follow city lines in the split counties.

District 1 D+02.59 - 55.1 - 42.9
District 2 D+14.84 - 63.6 - 34.7
District 3 R+04.53 - 50.3 - 47.7
District 4 R+13.13 - 39.0 - 59.1
District 5 R+07.66 - 46.5 - 51.0
District 6 D+04.82 - 56.5 - 41.5
District 7 D+36.23 - 84.0 - 14.5
District 8 D+09.14 - 58.3 - 40.2
District 9 D+06.44 - 57.2 - 41.0
District 10 D+05.73 - 57.5 - 40.6

[muon2]

Sorry for not responding Muon2, been very busy with final papers and the like.

The cvparty map which I ranked above yours was the only one included in the PM which I received from cvparty--this one:

one smol piece is taken from the northeast corner of canyon
1: R+14
2: R+25

I assume he wanted his other map excluded from consideration. Unfortunately it seems that ASV's maps have disappeared--not sure what happened to the images--but as far as I can recall I ranked ASV-A higher than muon2-A because it had a Boise-Panhandle linkage. Thus my lower ranking of muon2-A is because of shape/CoI concerns rather than connectedness.

I'd like to follow up with your thought on ID as well here.

Plan A preserves to Boise UCC, has no county chops, a population deviation within 0.5%, and counties are connected by road, though in one case it is a former, not current state highway.

There are three large CoI's in ID - the northern panhandle, the Boise metro, and the eastern Snake river. Boise is almost a whole CD while the other two are each only about a half CD. If I understand that right, you find it better to chop the metro than to connect the two half CD areas. Since in other states more weight seems to be given to metro areas, I'm curious to know what drives that thinking.

[muon2]

My non-partisan redistricting plan for Washington. Douglas County is the only unnecessary county split, in addition to the required splits of King, Pierce and Snohomish. Seattle is wholy within the 7th, and the districts follow city lines in the split counties.

For comparison I have two counties split outside the big three, but my plan has 2 fewer chops in counties (6 to 8 ), and UCCs (2 to 5). It goes to the debate earlier about the preference to chop more counties vs more chops in fewer counties.

[Torie]

Sorry for not responding Muon2, been very busy with final papers and the like.

I'd argue wrt: ferries that although they may be quite well used for determining connections for isolated islands (i.e. the San Juan Islands, or perhaps some of the islands in Alaska) when there's a road connection available the preference should be to defer to that. A lot of my feeling in this way is based on the example of Staten Island; it'd be very easy for an ambitious gerrymanderer to argue that Richmond County should be linked to lower Manhattan with the Staten Island Ferry, even though there's a much easier link to the city with I-278.

I would counter that when the road connection is many hours longer than the ferry, the ferry is heavily used with an hourly schedule, and the ferry is officially part of the state highway connection, then the ferry should be given consideration. As d32123 says ferries aren't ideal, but they are understandable if used to avoid messing up the Puget Sound. In this case the Whatcom-Skagit areas seem far more like the Olympic peninsula than they do suburban Snohomish. The current map tries to deal with this in WA-01 by linking a bunch of poorly connected inland areas, and if you read the rest of d32123's linked post, he's not happy with that at all. So I think if CoI matters, then a cross sound CD can make sense.

With that in mind here's my plan from that same thread. It started by partitioning the state into groups of counties such that each was very close to a whole number of CDs. Because the Seattle UCC was just over 5 CDs the choice was to split it, or to chop a number of counties - I chose the former. There was also a choice of either the orange or purple line as a partition - I chose the purple line with lower erosity.



The southern region has 2 CDs (3 and 4) and the line was drawn through Yakima county to keep the reservation intact and follow school district lines in the county. It keeps Yakima city connected to the rest of CD 4 by way of WA-24. CD-4 isn't pretty, but it works with the geographical restrictions, cities, school districts and the reservation.

The western region has 3 CDs (2, 6 and 9) and the line there was drawn through Kitsap to follow school district boundaries and minimize the size of the chop. The chop in Pierce also follows municipal and school district boundaries.

The Seattle region has 4 CDs (1, 7, 8, and 10) and the lines were drawn to minimize erosity while keeping cities and school districts intact within each CD. Two CDs are nested within King and 1 CD is nested completely within Snohomish. The maps of the counties (King, Snohomish) showing the cities and school districts are on the older thread.





CD 1: (-2175) D+7.9
CD 2: (-687) D+4.0
CD 3: (-2677) R+1.8
CD 4: (+1258) R+14
CD 5: (+1933) R+7.7
CD 6: (+1971) R+0.9
CD 7: (-163) D+36
CD 8: (-41) D+12
CD 9: (+2519) D+7.7
CD 10: (-1938) D+12

There are 5 uncompetitive D and 1 competitive D (CD 2) districts.
There are 2 uncompetitive R and 1 competitive R (CD 3) districts.
There is one highly competitive tossup (CD 6) district.
With a three seat advantage for the Dems, the plan has a skew of 0.

I remember this map drawing exercise of Washington. With one exception, Muon2's map is in my mind clearly the best map that can be drawn. The exception is that the map has a bridge chop. The rules on that have evolved, ending up with no sanction at all, after starting out with a ban, but with an odd and flawed in my opinion definition of what a bridge chop was. The rule that I still prefer is that all other things being equal a bridge chop should be disfavored. In the Seattle UCC area, other than cities, there are no agreed subunits, and the map below avoids a bridge chop with no subunit chops. So I prefer the map below. Perhaps its erosity score is higher (I don't know) but I m not sure that should enough to allow for a bridge chop, if the chop count is the same. Another issue is the connection for the NE corner of King County which is not internally connected to the rest of King. To establish a road connection, the chop into Snohomish needs to create that connection, and both Muon2's and my map do so.

So I submit the same map as Muon2, but with this one variation. Some disagreements are never resolved, and this is one example where Muon2 and I were unable to resolve the disagreement, so it is still extant.

[cvparty]

pretty similar to the current map


1: D+14
2: D+3
3: R+4
4: R+11
5: R+9
6: D+5
7: D+36
8: EVEN
9: D+16
10: D+6

[cvparty]

WASHINGTON ENTRIES

1. D+3
2. D+17
3. R+3
4. R+13
5. R+8
6. EVEN
7. D+36
8. D+11
9. D+12
10. D+3

CD 1: (-2175) D+7.9
CD 2: (-687) D+4.0
CD 3: (-2677) R+1.8
CD 4: (+1258) R+14
CD 5: (+1933) R+7.7
CD 6: (+1971) R+0.9
CD 7: (-163) D+36
CD 8: (-41) D+12
CD 9: (+2519) D+7.7
CD 10: (-1938) D+12

There are 5 uncompetitive D and 1 competitive D (CD 2) districts.
There are 2 uncompetitive R and 1 competitive R (CD 3) districts.
There is one highly competitive tossup (CD 6) district.
With a three seat advantage for the Dems, the plan has a skew of 0.

I submit the same map as Muon2, but with this one variation. Some disagreements are never resolved, and this is one example where Muon2 and I were unable to resolve the disagreement, so it is still extant.

District 1 D+02.59 - 55.1 - 42.9
District 2 D+14.84 - 63.6 - 34.7
District 3 R+04.53 - 50.3 - 47.7
District 4 R+13.13 - 39.0 - 59.1
District 5 R+07.66 - 46.5 - 51.0
District 6 D+04.82 - 56.5 - 41.5
District 7 D+36.23 - 84.0 - 14.5
District 8 D+09.14 - 58.3 - 40.2
District 9 D+06.44 - 57.2 - 41.0
District 10 D+05.73 - 57.5 - 40.6

pretty similar to the current map


1: D+14
2: D+3
3: R+4
4: R+11
5: R+9
6: D+5
7: D+36
8: EVEN
9: D+16
10: D+6

[AustralianSwingVoter]

Hawaii Non-Partisan plan.

District 1 D+18.23 - 70.8 - 27.8
District 2 D+17.83 - 73.0 - 25.3

[Strudelcutie4427]

1. D+14 (-678) 48% A, 20% W, 20% O
2. D+21 (+679) 46% A, 25% W, 20% O

[muon2]

Since over half the population of HI is on Oahu, that is where a split must fall. The island is covered by the combined city-county government of Honolulu, and any places on DRA are census places used for their internal purposes. Honolulu is divided into official neighborhoods with elected boards, and an interactive map where the neighborhood boundaries can be turned on. These are probably the best representative expressions of communities of interest on the island.

[jimrtex]

In my discussion with Muon2, I noticed that several counties such as Douglas, Lane, and Wasco, were large and contained disparate communities. Douglas and Lane contained coastal communities, and Wasco included vast tracts of inland territory. though most of the population was concentrated in The Dalles.

Instead, I used school districts. Oregon has 196 school districts. I consolidated these in a series of rounds, where in each round approximately 1/4 of the districts would be forced to consolidate. Pairs of adjacent small districts could merge, or a smaller district could join a larger neighbor. Occasionally there was a three-way merger as two small districts merged, and then a small neighbor would join them, rather than a larger neighbor.

I tended to base consolidation on factors that school districts might consider when forced to consolidate. Smaller districts would tend to resist absorption into a larger district, because it would likely mean closing of their schools, with the students bused into a larger town. If smaller districts consolidate, they may agree to operate the elementary school in one town, the middle school in another, and the high school in the third. Or they may agree to build a shiny new facility in a central location.

The first five rounds consolidated 196 => 168 => 141 => 114 => 95 => 79 districts.



In the eastern part of the state there were repeated consolidations. For example, the first three rounds saw consolidation of 8 districts in the southern part of Harney County, which were then merged with the Burns-based district, and finally the Harney County districts were merged with the Lake County districts.

There was some consolidation along the coast, and more rural districts in the Willamette Valley were merged, while generally resisting being brought into city districts.

Tne next 5 rounds consolidated 79 => 63 => 52 => 44 => 36 => 29 districts.



Oregon east of the Cascades has been consolidated into five districts; there are six districts in southern Oregon (Eugene, Roseberg, Grant Pass, Medford); five in central Oregon (Salem, Corvallis, Albany); eleven in the Portland area; and two along the coast.

In the central and southern parts of the state, county-sized districts are beginning to emerge. School districts were generally created within counties, as a county school board would organize small single-school districts. In the mid-1920's, Oregon had around 2300 school districts. It was only from about 1920 onward that there was suitable transport and roads for transporting pupils even a few miles to school, and prior to that many children did not attend high school. It would be normal for road networks to be built outward from major towns in a district, and these road networks would support creation of settlements, that could support a school, if only a one-room school. Consolidation would tend to reverse this process.

I came across a thesis from the the 1950's that was advocating for consolidation of schools districts in central Polk County, that would merge around 17 districts into the Dallas district. The thesis was pushing for inclusion of Perrydale and Falls City in this district. But apparently these districts remained independent because they had high schools. In my consolidation, these small districts of about 1000 (persons, not students) were merged into the Dallas school district of around 20K. This group in turn joined with the district in the eastern part of the county around Monmouth, and by this fifth round with the Salem-Keizer schools.

Note that some point, my districts accepted consolidation with a larger city. Maintaining identity by merging with a much more distant neighbor became unfeasible. Two small districts 15 miles apart might join forces rather than being absorbed into a large district 5 or 10 miles distant. This would not happen if the districts were 30 or 40 miles apart. As the consolidated districts became larger a broader community of interest would be recognized.

In the Portland area, Hillsboro and Oregon City have consolidated the hinterlands of Washington and Clackamas counties, but other suburban districts have just begun the merger process.

Along the coast, I tried to deliberately create a district, including three school districts in western Lane and Douglas. I also included Columbia, which though not on the coast might at least share more interests with the coast than with Portland. While the coastal strip does not have the population for a congressional district (about 3/10 of one), I wanted to explore keeping the area together in one or two districts.

The next four rounds consolidated 29 => 25 => 21 => 16 => 13



Consolidation of the east is nearly complete with only Bend holding out, and a single district has formed along the coast. There are five districts in the Portland area, with only the Portland and Beaverton school districts not having any changes.

In Round 15, Bend consolidated into the east, Douglas-Josephine combined with Jackson-Klamath, and Hillsboro consolidated into Beaverton. 13=> 10.

In Round 16, Albany-Corvallis merged with Eugene (Generally Benton, Linn, and Lane counties. In addition, the coastal strip was forced to consolidate. There didn't really seem like a reasonable way to keep the area whole. The northern portion including the areas downstream from Portland were added to the Washington-based district, and the southern areas with the southern district. But the central coast would have formed a panhandle for these areas, so was placed with the central district.



The 17th round reduced districts from 8 => 6, as the eastern district consolidated with the southern district, and the eastern Multnomah districts consolidated with the Portland district.

The 18th round reduced districts from 6 => 5, as the Salem district consolidated with the southern Portland suburbs.



We now have five districts, with one problem. They don't have equal populations. While asymmetric accretion makes it relatively easy to consolidate communities into ever larger communities, because it does not require agreement between the communities, it does not produce equal-population districts.

Symmetric accretion where we might start with 80 districts of equal size, and then do pairwise mergers 80 => 40 => 20 => 10 => 5 will produced districts of equal size, but the pairwise mergers may be difficult to arrange.

The 5 district populations are:

1. East + South 1.290
2. Eugene/Corvallis/Albany 0.776
3. Salem + Dallas + Clackamas 1.255
4. Washington + North Coast 0.709
5. Portland 0.970

OR-1 and OR-2 have a population equivalent to about two districts (2.066) as do OR-3 and OR-4 (1.964), while OR-5 is just below one district (0.970), so we can make most of the balancing of population as simple adjustments between OR-1 and OR-2, and OR-3 and OR-4.

Douglas (Roseburg) and Coos + Curry were moved from OR-1 to OR-2. Three school districts were moved for a final population balance. They were selected so as to not cut the road connection from Grant's Pass to Medford to Klamath Falls. If one were willing to split a school district, a division of the Three Rivers district in Josephine County might be preferred, with Grant's Pass and immediately surrounding areas placed on OR-4.

Six districts in Yamhill, including Newberg and McMinville, as well as Tigard-Tualatin in southern Washington county was transferred from OR-3 to OR-4. In addition Lincoln County was shifted from OR-2 to OR-4 to account for the difference in population between OR-1+2 and OR-3+4.

Sherwood remained in OR-3 for reasons of population balance. It is a reasoanble match for the Clackamas districts west of the Willamette (Lake Oswego and West Linn) and does itself extend into Clackamas county.

To get the Portland district to the quota, I added Oregon Trail which is a better fit with the eastern suburban districts than the southern suburban districts. In addition both Centennial and Gresham-Barlow school districts extend into Clackamas county. A final tweak was to place Corbett school district into OR-1. The district has about 40% of the area of Multnomah, but less than 4000 persons. Rather dragging Hood River into a Portland district, this places a rural district with the areas to the east.

This is my final map.



The districts largely conform to county boundaries:

OR-1: 18 counties east of the Cascades, all of Josephine, most of Jackson except Eagle Point, Butte Falls, and Prospect (0.039 quotas) and Multnomah: Corbett (0.005)

OR-2 Douglas; Coos; Curry; Lane, except a tiny bit of Lincoln County school district; Benton, except a bit of Central school district; a tiny bit of Philomath in Polk; and most of Linn, except all of Scio, a bit of Jefferson, and the Linn County portions of North Santiam and Santiam Canyon schools (the North Santiam River forms the boundary between Linn and Marion, so the two districts are really bi-county districts; and three districts in Jackson that are not in OR-1.

OR-3 Marion; Polk, except bits of Philomath, Willamina, Sheridan, and Amity; Clackamas, except for Oregon Trail, portions of Gresham-Barlow and Centennial, and bits of Tigard-Tualatin and Newberg; Multnomah: Riverdale; Washington: most of Sherwood, and a bit of West Linn; and Lane: Scio, and parts of Jefferson, North Santiam, and Santiam Canyon.

OR-4 Columbia, Clatsop, Tillamook, Lincoln, Yamhill, Washington, except Sherwood, and bits of West Linn and Portland; Multnomah: bits of Hillsboro, Beaverton, and Scappoose; Clackamas: tiny bits of Newberg and Tigard-Tualatin; Polk: Parts of Amity, Sheridan, and Willamina; Lane: southern end of Lincoln County school district.

OR-5 Multnomah, except Riverdale and Corbett, and tiny bits of Hillsboro, Beaverton and Scappoose; Clackamas: Oregon Trail, and portions of Gresham-Barlow and Centennial; and Washington: a tiny bit of Portland.

Population is based on school district population; demographic composition and political data from DRA 2.0. I matched the Census Block groups to school district boundaries as close as possible. I have no idea how DRA produces political data for census block groups.

OR-1 East and Far South Bend, Medford, Grant's Pass, Klamath Falls, Hermiston, The Dalles, La Grande, Ontario. +0.41%; A 86, H 9, O 2, AI 2, As 1; R+10.83.

OR-2 Southwest Eugene, Springfield, Corvallis, Albany, Roseberg, Coos Bay. +0.10%, A 88, H 5, O 3, As 2, AI 1, B 1; D+0.17.

OR-3 Central West Salem, Lake Oswego, Keizer, Dallas. -0.56%; A 82, H 12, As 3, O 2, AI 1, B 1; R+0.59.

OR-4 Northwest Hillsboro, Beaverton, Tigard, McMinnville, St. Helens, Newport, Astoria. -0.19%; A 78, H 11, As 7, O 2, B 1, AI 1. D+6.64.

OR-5 Portland Portland, Gresham. +0.24%; A 77, H 8, As 7, B 5, O 3, AI 1. D+26.52.

[cvparty]

imma just submit the current map cuz it's quite good

1: D+17
2: D+19

[cvparty]

HAWAII ENTRIES

1. D+14 (-678) 48% A, 20% W, 20% O
2. D+21 (+679) 46% A, 25% W, 20% O

imma just submit the current map cuz it's quite good

1: D+17
2: D+19

[muon2]

Unfortunately this last cycle coincided with a very hectic weekend, so I could make no submissions, and for some reason I thought the submission date was 4/18. I have maps for all the states applying the muon rules, but not with the new DRA updates. 4 days is very short when I'm in the hectic part of my RL spring schedule. I doubt I can do anything for this cycle either, especially for CA which requires a lot of download time.

[Strudelcutie4427]

Northern California



1. R+9 (Northeast)
2. R+8 (Northwest)
3. R+3 (Eastern Sacramento Suburbs)
4. D+21 (Sacramento)
5. D+22 (Marin to Humboldt)
6. D+17 (Napa Valley)
7. D+24 (Contra Costa)
8. D+4 (N. Central Valley)
9. D+14 (Eastern Alameda)
10. D+39 (Oakland)
11. D+38 (San Francisco)
12. D+26 (San Mateo)
13. D+22 (Southern Alameda)
14. D+27 (San Jose)
15. D+26 (Southern Santa Clara)
16. D+19 (Santa Cruz)
17. R+9 (Eastern CA Mountains)
18. EVEN (Stanislaus/Merced)

Central CA



19. D+22 (Monterrey)
20. D+6 (Fresno)
21. R+9 (Tulare)
22. R+9 (Bakersfield)
23. D+7 (Santa Barbara)
24. D+4 (Ventura)

Los Angeles Area

25. D+17 (N Coastal LA)
26. D+26 (East of 25)
27. D+33 (Coastal LA)
28. D+12 (Lompoc Area)
29. D+39 (South LA)
30. D+40 (Central LA)
31. D+13 (SE LA)
32. D+37 (East of 30)
33. D+26 (East of 32)
34. D+18 (NE Suburban LA)
35. D+24 (N Suburban LA)
36. D+21 (NW Suburban LA)
37. EVEN (N LA County)
38. D+12 (S Coastal LA)
39. R+2 (N Orange County)
40. D+16 (Central OC)
41. R+2 (Center Coast OC)
42. R+7 (Southern OC)
43. D+9 (NW Riverside/S SB )
44. D+8 (SW SB/E LA)
45. D+17 (Outer SB )
46. R+9 (Rural SB )

Southern California
https://snag.gy/GQvDus.jpg

47. R+1 (Western Riverside)
48. R+9 (Northern San Diego/SW Riverside)
49. D+3 (Central SD Coast)
50. D+17 (San Diego Downtown)
51. D+16 (S Coastal SD)
52. R+8 (Eastern SD County)
53. D+7 (Imperial/E Riverside)