Applying Muon2 scheme for redistricting to Columbia County

[Torie]

My overactive brain has come up with a new project. Now that the petition process has been perfected, my thought is to have a proposition placed on the ballot come the 2020 election (it needs to be done in a POTUS year to maximize the chances of success, since this is something the Dems want, and the Pubs do not), to dump Columbia County's crazed system, and replace it with a county legislature.

The body would have 23 seats, just like the current board of supervisors does (that high number is needed to give a good chance that at least one black will be elected from Hudson from a combo of wards 2 and 4 plus Crosswinds and a few bits from the 5th ward). My thought is to give Muon2's scheme a test drive, and provide in the law for his scheme for restricting, where a computer spits out the map doing the chop (this time for towns and cities, with maybe special rules for villages (Columbia county has 3 villages), giving them preference in keeping whole where a town chop is involved (the village of Chatham straddles two towns), erosity minimization thing, using federal, state and county highways as the sensitive roads to test for cuts. Does this make any sense at all?

[Torie]

I am aware of the byzantine procedural rules.

I did draw a map. I don't have a census block map for the county (just a voting district map), so the lines are approximate as to where the chops lie in the town of Kinderhook and Taghkanic.

[Torie]

I don't know if the voting weights have been changed. There is a lacunae in the law about that potentially down the road, but I will leave you to figure out what that might be.

How do you avoid chopping Taghkanic, without chopping somewhere else? That would interest me.

[Torie]

I am aware of the byzantine procedural rules.

I did draw a map. I don't have a census block map for the county (just a voting district map), so the lines are approximate as to where the chops lie in the town of Kinderhook and Taghkanic.

If I assume a 5% max deviation to get to a 10% range, Ghent is too small to be a district on its own (-6.8%) but Claverack is just barely the right size for a district (+4.97%). I presume that you've compensated for Ghent elsewhere, but why not make Claverack whole instead?

With Claverack whole, the Greenport district would pull about 1000 from Hudson and 600 from Livingston. Ancram, Gallatin and Taghkanic would pull about 1200 from Livingston. That leaves the rest of Livingston with Germantown and Clermont. It's the same chop count in the south but removes the ugly connection (IMO) between Clermont and Gallatin.

In the north you would have to add a chop to bring Ghent up at least 46 people to get the range within 10%. OTOH you could probably make a case of a compelling state interest to accept that extra deviation for a range of 10.79% to keep Ghent whole as a district.

I interpret the law that the deviation cannot be more than 10%. The deviation is less than 10%, because the most populated district is about 2% over the quota or something like that. You interpret the federal law as requiring no more than a 5% deviation for each district from the quota amount? The state law does not have such a 5% requirement as I recall.

I give a pass for the ugly connection in the south, given the ugly shape of Clermont. But I will try out your map.

[Torie]

I don't know if the voting weights have been changed. There is a lacunae in the law about that potentially down the road, but I will leave you to figure out what that might be.

How do you avoid chopping Taghkanic, without chopping somewhere else? That would interest me.

It violates MHR to divide a town with less than 110% of a quota. In an 11-district plan, this prevents splitting of all but Hudson and Kinderhook.

Taghkanic is placed with the adjacent district with the smallest population.

Even if that causes a violation of the 10% deviation rule?  The 110% quota rule trumps the 10% deviation rule, when the two conflict?

[Torie]

Number 13 in section 10 of the MHRL sets forth the rule to which Jimrtex refers. I have seen in the weighted vote cases courts allow deviations of up to about 20%.

Morris cites earlier SCOTUS cases on the subject, including Connor. Connor distinguishes a fact pattern there that had in excess of a 10% deviation (16.5%) as not justified, because it was unlike the facts in Mahan, where there was uncontradicted evidence that the legislature's plan (which had a 16.4% deviation), "produces the minimum deviation above and below the norm, keeping intact political boundaries.'" Id., at 326. By contrast, the plaintiffs in this case submitted to the District Court an alternative Senate plan that served the state policy against fragmenting county boundaries better than did the plan the court ultimately adopted, and also came closer to achieving districts that are "as nearly of equal population as is practicable."

Here we have the same fact pattern as in Mahan. Mahan is our "spotted horse" case. I conclude from all of this, that particularly given the requirements of state law, which define the rules of the road for keeping municipalities together, a policy that SCOTUS endorses if the minimum deviation is achieved to hew to such policy (at least up to 16.4%), that the map must have a deep enough chop into Hudson by Greenport, to equalize the population between the two districts, without Greenport chopping into Claverack (each district having about 92% of the quota)), and that Taghkanic must be joined with the southern towns, if the population deviation does not exceed 20%. So we are good to go if the deviation is 19.4% as Muon2 suggests.  I think a federal court would be very sympathetic, given that the proposed deviation is in fact the only way to comport with state law, at least where the deviation does not exceed 20%. Anything over 20% would substantially increase the legal risk however.

After the next census, there might be more risk, since Hudson is losing population faster than the county (and perhaps Hudson and Greenport collectively losing population faster than the county even after taking into account the 95 person increase in Greenport through 2016 as Hudson expels residents to accommodate the rapid spread of Airbnb's here, there and everywhere in the city, particularly in the last two years (I tend to think Robinson Street from the combination of Airbnb's popping and gentrification might have lost close to half its population since 2010), which might push the deviation over 20%.

Addendum:  On a prisoner adjusted basis, the deviation is 21.57% for such a map, so now the answer is less easy. On page 1063 of this law review article is a statement that courts have allowed deviations over 20%.

I got the case cites read to me over the phone by the reference desk librarian at the state library in Albany, and one of them, Gorin v Karpan, makes for instructive reading. It revisits Wyoming redistricting a decade later, and it makes clear that Brown v. Thomson, is a very weird case that cannot be relied upon for much of anything, in particular that a deviation of 89%, rather than being possibly legal, is more akin to  legal suicide (the case has become, as they say, a mere vagrant on the waters of the law). From Gorin, here is the money paragraph:

"As mentioned, the Mahan Court warned that a 16.4% deviation "may well approach tolerable limits...." 410 U.S. at 329, 93 S.Ct. at 987. The Morris Court acknowledged that "no case of ours has indicated that a deviation of some 78% could ever be justified." 489 U.S. at 702, 109 S.Ct. at 1442. Furthermore, in Gaffney, 412 U.S. at 744, 93 S.Ct. at 2326, the Court found that a review of Supreme Court reapportionment cases made it "apparent that the larger variations from substantial equality are too great to be justified by any state interest so far suggested" (referring to Swann, (25.65%); Kilgarlin, (26.48%);[13] and Whitcomb v. Chavis, 403 U.S. 124, 91 S. Ct. 1858, 29 L. Ed. 2d 363 (1971), (24.78%)[14])."

[Torie]

Number 13 in section 10 of the MHRL sets forth the rule to which Jimrtex refers. I have seen in the weighted vote cases courts allow deviations of up to about 20%.

Morris cites earlier SCOTUS cases on the subject, including Connor. Connor distinguishes a fact pattern there that had in excess of a 10% deviation (16.5%) as not justified, because it was unlike the facts in Mahan, where there was uncontradicted evidence that the legislature's plan (which had a 16.4% deviation), "produces the minimum deviation above and below the norm, keeping intact political boundaries.'" Id., at 326. By contrast, the plaintiffs in this case submitted to the District Court an alternative Senate plan that served the state policy against fragmenting county boundaries better than did the plan the court ultimately adopted, and also came closer to achieving districts that are "as nearly of equal population as is practicable."

Here we have the same fact pattern as in Mahan. Mahan is our "spotted horse" case. I conclude from all of this, that particularly given the requirements of state law, which define the rules of the road for keeping municipalities together, a policy that SCOTUS endorses if the minimum deviation is achieved to hew to such policy (at least up to 16.4%), that the map must have a deep enough chop into Hudson by Greenport, to equalize the population between the two districts, without Greenport chopping into Claverack (each district having about 92% of the quota)), and that Taghkanic must be joined with the southern towns, if the population deviation does not exceed 20%. So we are good to go if the deviation is 19.4% as Muon2 suggests.  I think a federal court would be very sympathetic, given that the proposed deviation is in fact the only way to comport with state law, at least where the deviation does not exceed 20%. Anything over 20% would substantially increase the legal risk however.

After the next census, there might be more risk, since Hudson is losing population faster than the county (and perhaps Hudson and Greenport collectively losing population faster than the county even after taking into account the 95 person increase in Greenport through 2016 as Hudson expels residents to accommodate the rapid spread of Airbnb's here, there and everywhere in the city, particularly in the last two years (I tend to think Robinson Street from the combination of Airbnb's popping and gentrification might have lost close to half its population since 2010), which might push the deviation over 20%.

Addendum:  On a prisoner adjusted basis, the deviation is 21.57% for such a map, so now the answer is less easy. On page 1063 of this law review article is a statement that courts have allowed deviations over 20%.

The deviation range for the Vermont House is 19%. It would be worse if they went universally to single-member districts.

You can use weighted voting with legislative districts. In your map, eliminate the divisions of Hudson and Kinderhook and you're done. You could even put Taghkanic with Claverack, or simply let Taghkanic residents choose their district. Or simply start out with the 19 towns and eliminate one at a time, and let the residents attach themselves to another district.

But you have not addressed the fundamental question of why?

The town supervisors are not paid by Columbia County. They serve on the Board of Supervisors by virtue of their election as town supervisors. If you switch to a county legislature, you have to also provide for election. Town supervisors are elected to different terms, and at different times. While a town supervisor may be permitted to serve as a legislator, they have to run in two elections.

The only real problem with the current system is the five supervisors from Hudson, all of which have fewer persons than Taghkanic, the smallest town, and I have explained how to eliminate that problem.

Because the gold standard is one person districts with each representative having one vote. Anything else is worse than watered down beer. That is my opinion, and I am not changing it. I know you disagree. That is fine. The Columbia County system has a host of flaws, including inter alia a merging of the executive and legislative branches, and its governance works poorly and opaquely.  It is mostly a visionless patronage machine (almost all high ranking county employees are Pubs, while the employee peons are squeezed for contributions and their time to work on campaigns) - a celebration of Balkanized parochialism, largely devoid of long term planning and focus. It must go, and eventually will go when the time is right. My work in this part of the world has just begun. There is much left to do.

I assume the Vermont plan was not legally challenged as to the 19% figure.

[Torie]

I am aware of the byzantine procedural rules.

I did draw a map. I don't have a census block map for the county (just a voting district map), so the lines are approximate as to where the chops lie in the town of Kinderhook and Taghkanic.

If I assume a 5% max deviation to get to a 10% range, Ghent is too small to be a district on its own (-6.8%) but Claverack is just barely the right size for a district (+4.97%). I presume that you've compensated for Ghent elsewhere, but why not make Claverack whole instead?

With Claverack whole, the Greenport district would pull about 1000 from Hudson and 600 from Livingston. Ancram, Gallatin and Taghkanic would pull about 1200 from Livingston. That leaves the rest of Livingston with Germantown and Clermont. It's the same chop count in the south but removes the ugly connection (IMO) between Clermont and Gallatin.

In the north you would have to add a chop to bring Ghent up at least 46 people to get the range within 10%. OTOH you could probably make a case of a compelling state interest to accept that extra deviation for a range of 10.79% to keep Ghent whole as a district.

You really think try-chopping Livingston is a good idea? That seems like a pretty big dump on Livingston.

[Torie]

Well in other news, when a county adopts a charter, the charter dictates the rules of the road for districting, and not the MHRL.  In Ulster County, counsel opined that the charter had not referenced the 110% rule, so it did not apply. So assuming Columbia County were not so foolish as to get anywhere near adopting the silly and undoable 110% rule, one can chop any town one wants. Pity I spent a good part of yesterday reading all the cases on population disparities, but I least I think I know the law on that topic now. The good news is contained in  this document. I screen shot the page setting forth this conclusion. The county also need not exclude the prisoner population, and Ulster County in fact counted them, to avoid delays I think. That lacunae pops up elsewhere in the document.

[Torie]

I am aware of the byzantine procedural rules.

I did draw a map. I don't have a census block map for the county (just a voting district map), so the lines are approximate as to where the chops lie in the town of Kinderhook and Taghkanic.

If I assume a 5% max deviation to get to a 10% range, Ghent is too small to be a district on its own (-6.8%) but Claverack is just barely the right size for a district (+4.97%). I presume that you've compensated for Ghent elsewhere, but why not make Claverack whole instead?

With Claverack whole, the Greenport district would pull about 1000 from Hudson and 600 from Livingston. Ancram, Gallatin and Taghkanic would pull about 1200 from Livingston. That leaves the rest of Livingston with Germantown and Clermont. It's the same chop count in the south but removes the ugly connection (IMO) between Clermont and Gallatin.

In the north you would have to add a chop to bring Ghent up at least 46 people to get the range within 10%. OTOH you could probably make a case of a compelling state interest to accept that extra deviation for a range of 10.79% to keep Ghent whole as a district.

The problem with this analysis I now discern, is that you have exceeded the 10% deviation, where there is another map (mine), that does not, that has the same number of chops. You would have to argue that exceeding 10% is justified, because of the ugly shape of Clermont, and it is worth tri-chopping one town, Livingston, plus go over 10%, in order to get rid of that ugly shape, by smoothing out the lines via the tri-chop. So I see a fair amount of legal risk here.

[Torie]

Here is the chop for Taghkanic that gets the population split exactly right:



And the Kinderhook chops:



And the Hudson chop: