Ohio - a celebration of the Muon2 rules

[Torie]

I highly doubt there will be a 50% BVAP CD in play sans Akron (absent something absurd, and even that I doubt), and the VRA won't demand Akron. But if Gingles is triggered so be it. We are just speculating.

Other than that, we are talking past each other. The commission draws say the map immediately below (oh, I see I missed a black line between Garfield Heights and Cuyahoga Heights, and another between Seven Hills and Brooklyn Heights), and then the legislature when picking a map, should be allowed by the law to pick one with a lower score that generates a higher BVAP CD subject to the rules I outlined above (such as the the second map below), and I said, I think it is likely that they will so choose to do so. And based on that assumption, I drew my state map. That is all. Nothing more.

[muon2]

Thanks for the detail. I understand exactly what you want. I am still missing how a higher BVAP map you describe gets into the set the commission sends to the legislature. The whole point of scoring is that the legislature is not free to consider all maps submitted, just those that make the commission's cut. The commission is bound by statutory rules so they can't play games with squishy CoI considerations.

[Torie]

Thanks for the detail. I understand exactly what you want. I am still missing how a higher BVAP map you describe gets into the set the commission sends to the legislature. The whole point of scoring is that the legislature is not free to consider all maps submitted, just those that make the commission's cut. The commission is bound by statutory rules so they can't play games with squishy CoI considerations.

The commission absent Gingles ignores creating a performing minority CD. However, if the legislature with both parties on board wants a CD with a higher minority percentage (I think I said that it needs to get up to at least 30% VAP, and no more than 50%), then the legislature asks the commission to revise the map to generate the highest scoring map that gets to the requested VAP percentage. Or the legislature does it, and the commission checks their work. If after the commission does its thing, the legislature does not like the map, they can then abandon the project, or change the percentage within the 30% to 50% parameter, and try again.

So probably the starting point is to wait until the commission generates the initial map, or maps, and then the legislature will negotiate whether it wants to increase the minority VAP of a CD, and if they do, then one goes through the ensuing steps. In the real world, the legislature will be doing what the commission does, and have a good idea of the end game.

[jimrtex]

FWIW, I don't think there is any precedent that a theoretical 50% BVAP CD, that is impossible to draw in practice, or would look ridiculous, triggers Gingles, and I don't think SCOTUS will ever go there. It is going in the opposite direction. I also think 41.5% BVAP is clearly performing. So I don't think there is any case to chop Cleveland, and in particular a macrochop.

Aren't we supposed to balance erosity against chops, so the right course would seem to be to look at both alternatives. If you are saying that there is insufficient black population to trigger Gingles then there is no need to try to draw a performing district - chops and erosity should be the only factors.

Yes, per your rules, you are absolutely correct. But in the real world, such a CD would never be drawn. That is why if I wrote the legislation, I would allow a deviation of the rules to allow drawing performing minority CD's, if both parties agree, and they would in this case, particularly if the NE corner CD is only marginal politically, which it is. If one does not allow this flexibility, I don't think your rules would ever become law, nor adopted by a court.

Doesn't your law violate the 15th Amendment?

Well that never occurred to me. A law that allows the legislature to agree to draw districts so that minorities can have adequate representation where the metrics of the state law in practice otherwise preclude that, violates the 15th amendment? Has anyone made that argument in an analogous context? I would be amazed if that argument ever got any traction.

My understanding is that you propose bypassing persons in order to include persons of one race who would tend to be a political minority in that district, such that they are in effect wasting their time time by even voting, and that you would be seeking to maximize the number of such persons.

Is that not deliberate state action to disenfranchise persons on the basis of their race?

[Torie]

In order to increase minority representation, one reduces non minority representation. It's a zero sum game. The idea is to encourage where reasonable that minorities get representation commensurate with their numbers, and that is the law.

[jimrtex]

In order to increase minority representation, one reduces non minority representation. It's a zero sum game. The idea is to encourage where reasonable that minorities get representation commensurate with their numbers, and that is the law.

The US Constitution is that you can not classify based on race (14th Amendment), particularly with respect to the right to vote (15th Amendment).

You can't take the right to vote from one person and give it to another person.

[muon2]

Thanks for the detail. I understand exactly what you want. I am still missing how a higher BVAP map you describe gets into the set the commission sends to the legislature. The whole point of scoring is that the legislature is not free to consider all maps submitted, just those that make the commission's cut. The commission is bound by statutory rules so they can't play games with squishy CoI considerations.

The commission absent Gingles ignores creating a performing minority CD. However, if the legislature with both parties on board wants a CD with a higher minority percentage (I think I said that it needs to get up to at least 30% VAP, and no more than 50%), then the legislature asks the commission to revise the map to generate the highest scoring map that gets to the requested VAP percentage. Or the legislature does it, and the commission checks their work. If after the commission does its thing, the legislature does not like the map, they can then abandon the project, or change the percentage within the 30% to 50% parameter, and try again.

So probably the starting point is to wait until the commission generates the initial map, or maps, and then the legislature will negotiate whether it wants to increase the minority VAP of a CD, and if they do, then one goes through the ensuing steps. In the real world, the legislature will be doing what the commission does, and have a good idea of the end game.

Unfortunately that defeats the essence of the IA system that forms the basis of the procedure to execute my rules. The mapper(s) follow strict rules and all the legislative body can do is give it (or one of a set) an up or down vote with an external arbitrator acting if the legislature can't. A clever legislative leader would use your method to force a favorable gerrymander under the guise of unsatisfactory minority representation.

I thought part of this exercise was to show non-commission states how a neutral process could work. There are states that have moved to commissions, and many others toying with the idea. I want to give those states a different way to use a commission. You seem to have given up on converting more states to an independent process.

[jimrtex]

This is the base map, showing projected populations for 2020, assuming 15 congressional districts.



Cleveland 2.504 districts. Stripping Lake and Medina, gives 1.979 for Cuyahoga and Lorain. This could result in Cleveland and the eastern suburbs in one district; and Lorain and the western and southern suburbs.

Columbus 2.411 districts. Franklin and Delaware together is 1.987, resulting in one district with the bulk of Columbus and southern suburbs in one district, and the remainder of Franklin and Delaware in the other. The smaller counties of Licking and Fairfield will form almost 1/2 of another district.

Cincinnati 2.101 districts. A small part of one of the suburban counties will be trimmed off. Hamilton is just over one district, but might have a tiny bit trimmed off.

Dayton 1.029 districts. The UCC can form a district.

Akron 0.905 districts. Adding Geauga has a population of 1.026.

Toledo 0.719 districts. Because of its isolation, the UCC will form the nucleus of a district extending into adjoining counties.

Youngstown 0.544 districts. Youngstown could conceivably be paired with Canton, but by taking up Ashtabula and Lake reacheds 0.960.

Canton 0.481 districts. Canton will be paired with Medina, to contain the five northeastern districts in the most urban areas.

Springfield 0.172 districts. Because Dayton will form a single district. Springfield (Clark County) will be treated as an isolated county.

Mansfield 0.153, Lima 0.132, Wheeling, WV 0.087, Steubenville 0.084, and Huntington, WV 0.077 will be treated as isolated counties.

[Torie]

Thanks for the detail. I understand exactly what you want. I am still missing how a higher BVAP map you describe gets into the set the commission sends to the legislature. The whole point of scoring is that the legislature is not free to consider all maps submitted, just those that make the commission's cut. The commission is bound by statutory rules so they can't play games with squishy CoI considerations.

The commission absent Gingles ignores creating a performing minority CD. However, if the legislature with both parties on board wants a CD with a higher minority percentage (I think I said that it needs to get up to at least 30% VAP, and no more than 50%), then the legislature asks the commission to revise the map to generate the highest scoring map that gets to the requested VAP percentage. Or the legislature does it, and the commission checks their work. If after the commission does its thing, the legislature does not like the map, they can then abandon the project, or change the percentage within the 30% to 50% parameter, and try again.

So probably the starting point is to wait until the commission generates the initial map, or maps, and then the legislature will negotiate whether it wants to increase the minority VAP of a CD, and if they do, then one goes through the ensuing steps. In the real world, the legislature will be doing what the commission does, and have a good idea of the end game.

Unfortunately that defeats the essence of the IA system that forms the basis of the procedure to execute my rules. The mapper(s) follow strict rules and all the legislative body can do is give it (or one of a set) an up or down vote with an external arbitrator acting if the legislature can't. A clever legislative leader would use your method to force a favorable gerrymander under the guise of unsatisfactory minority representation.

I thought part of this exercise was to show non-commission states how a neutral process could work. There are states that have moved to commissions, and many others toying with the idea. I want to give those states a different way to use a commission. You seem to have given up on converting more states to an independent process.

One can make up any set of rules as to how the legislature operates. The only gerrymander that could be effected is a bipartisan one (since both parties have to agree), and that is limited by the requirement that for a given increase (one cannot do this to reduce the otherwise obtaining VAP) minority VAP percentage (within the 30% to 50% range), only the highest scoring map that generates such VAP can be used. I don't think that will cause much mischief. As I have said before, and will say again, if you don't provide for some flexibility in the rules, to accommodate certain pressure points (and this is one of them), you rules rather than bending, will break. So on this one, we shall just have to disagree. In the end, it is not for us to decide what the rules are anyway. All we can do is try to persuade. And I have made my case!  

[muon2]

I find bipartisan gerrymanders particularly insidious, since they tend to result in few if any districts that can swing and allow the electorate to change the composition of a delegation. I also have observed that bipartisan gerrymanders can be quite popular with legislatures since they tend to protect the maximum number of members from challenges. In my scoring they may do well on SKEW but usually perform poorly on POLARIZATION.

The model I've been promoting (and not just here on Atlas) is similar to the Base Realignment and Closure Act that is used by Congress to determine what military installations get closed. The commission report goes to Congress, but it must be accepted or rejected in its entirety. It has the beauty of stifling vote-trading to protect projects in member districts, and was used five times from 1988 to 2005. As I noted IA has used this same model for adopting or rejecting redistricting plans since 1980.

[Torie]

I still don't think there is much room for your angst to be realized given the constraints I outlined, for bipartisan gerrymanders (and such constraints are no accident).  In Cleveland, if either your map (for 43% BVAP), or mine (41.7% BVAP) are used in Ohio, and they are the highest scoring maps for those BVAP's, then we create a performing minority CD in exchange for a Dem CD becoming a swing one. I don't consider that a policy problem.

If you can come up with a hypo where would your posited concerns are realized, that seems like an "attractive" bipartisan gerrymander, that would be interesting.

I don't like bipartisan gerrys either. Who knew?  

[jimrtex]

Here is my district map.



1,2. Cleveland -1.0% Cuyahoga divided
3,4. Columbus -0.6% Franklin divided.
5,6, and 7. Cincinnati and Southern (Hamilton and Clermont (or Warren) divided +0.7%.
8. Dayton +2.9%
9. Akron +2.6%
10. Toledo -0.5%
11. Youngstown -4.0%
12. Canton -0.6%
13. East -1.2%
14. North +1.4%
15. West 0.7%

[Nyvin]

In other news on this topic, it looks pretty likely that Ohio will have a bipartisan commission for congressional redistricting in 2020:

http://www.cleveland.com/metro/index.ssf/2017/05/congressional_redistricting_am.html

[jimrtex]

This is my detail for the two Cuyahoga districts.



1. Cleveland and eastern suburbs. -1.1%
2. Western and southern suburbs, and Lorain -0.9%. About 60% is in Cuyahoga, 40% in Lorain, which would make it about 75-80% suburbs, with the rest Lorain and Elyria and more rural parts of the county.

The Cleveland district is about as black as you can get in Cuyahoga County, and is pretty compact.

[jimrtex]

This is my detail for the Columbus area:



The OH-3 Green Monster includes Columbus, Bexley, and Perry and Clinton townships. 138,000 in Columbus needs to be transferred to OH-4 to equalize population. This would initially close the crevasse south of Dublin and link Prairie, Franklin, and Sharon townships, as well as provide a connection to Whitehall and Reynoldsburg. Some smaller remnants of townships may be switched to OH-3. The main criteria would be whether more of Columbus would need to be added, than the area that could be rescued.

Perry and Clinton townships are highly fragmented and only have about 4,000 each. Bexley is largely white, and to reach it would require a link through largely black areas. This would mean that the link would be as thin as possible, which gives a strong scent of race-sorting. Whitehall might also be included in the district. It is 19% Black (in 2010) and increasing.

This is a conceptual drawing.



The yellow boundary connects all of OH-4. The red X indicate areas that might be filled in to make OH-3 more compact, based on available population.

[jimrtex]

Cincinnati UCC detail.



OH-5 is Hamilton County minus a few slivers along the eastern edge.

OH-6 is Butler, Warren, the slivers from Hamilton, and northern Clermont, generally closer to the city.

OH-7 is the remainder of Clermont attached to 17 other counties in southern Ohio.

[jimrtex]

Final map including four divided counties: Franklin, Cuyahoga, Hamilton, and Clermont. The first three have population for more than one district, and one district is wholly within each.

[Torie]

Nice map, but you need to show the chops to score it.

[jimrtex]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

[Torie]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

Well the Dayton area CD needs a chop for sure, along with a chop into Geauga. You must not be using the 0.5% variance in population constraint. If that is the case, then we are doing an apples to oranges exercise, with you having your own set of rules.

[jimrtex]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

Well the Dayton area CD needs a chop for sure, along with a chop into Geauga. You must not be using the 0.5% variance in population constraint. If that is the case, then we are doing an apples to oranges exercise, with you having your own set of rules.

Is this a requirement in the US Constitution, or federal statute?

[Torie]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

Well the Dayton area CD needs a chop for sure, along with a chop into Geauga. You must not be using the 0.5% variance in population constraint. If that is the case, then we are doing an apples to oranges exercise, with you having your own set of rules.

Is this a requirement in the US Constitution, or federal statute?

I don't want to revive that discussion, which you had with Muon2. Different metrics make different maps.

[jimrtex]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

Well the Dayton area CD needs a chop for sure, along with a chop into Geauga. You must not be using the 0.5% variance in population constraint. If that is the case, then we are doing an apples to oranges exercise, with you having your own set of rules.

Is this a requirement in the US Constitution, or federal statute?

I don't want to revive that discussion, which you had with Muon2. Different metrics make different maps.

What if my map is qualitatively better?

[Torie]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

Well the Dayton area CD needs a chop for sure, along with a chop into Geauga. You must not be using the 0.5% variance in population constraint. If that is the case, then we are doing an apples to oranges exercise, with you having your own set of rules.

Is this a requirement in the US Constitution, or federal statute?

I don't want to revive that discussion, which you had with Muon2. Different metrics make different maps.

What if my map is qualitatively better?

Higher deviations in population tend to make for maps that fit the other constraints better. I have no idea what population constraint you are using. But it doesn't matter. It is going to be enough of a tough sell to push the 0.5% constraint, that seems pretty safe legally, assuming the states agree to it. Anyway, do your thing.  You will anyway.

[jimrtex]

Nice map, but you need to show the chops to score it.

It shows the chops, other than Columbus.

Well the Dayton area CD needs a chop for sure, along with a chop into Geauga. You must not be using the 0.5% variance in population constraint. If that is the case, then we are doing an apples to oranges exercise, with you having your own set of rules.

Is this a requirement in the US Constitution, or federal statute?

I don't want to revive that discussion, which you had with Muon2. Different metrics make different maps.

What if my map is qualitatively better?

Higher deviations in population tend to make for maps that fit the other constraints better. I have no idea what population constraint you are using. But it doesn't matter. It is going to be enough of a tough sell to push the 0.5% constraint, that seems pretty safe legally, assuming the states agree to it. Anyway, do your thing.  You will anyway.

A primary reason that the SCOTUS has resisted setting de minimis deviation standards, is that the limits become targets. A scoring system that favors certain thresholds makes them targets.

My map is based on seeking equality within the constraint of respecting county boundaries and UCC communities of interest. The constraint certainly was not to get within some predefined limit. So it is possible that the Muon2 method is not safe legally.

The equality of my plan can be measured by determining the number of persons who would have to be shifted across county lines in order to achieve equality. That is, a measure of whether you were attempting to hit the target or not. It is also a measure of the number of persons victimized by an obsessive compulsion to equalize population.

169,348 persons can be shifted to equalize population, but 77,000 of those are in Clermont. Otherwise it is only 91,519 statewide or about 1/126 Ohioans.

My districts are as equal as practicable, which is the standard set by the SCOTUS for congressional districts.