US House Redistricting: Ohio

[krazen1211]

Just checking; don't want to read through the entire thread. You're going for 13-3 here?

I think Torie is.

I'm going with the eliminate Kucinich/Sutton, screw Stivers map, which is 12-4. Stivers gets thrown into a Dem vote sink.

[Torie]

Just checking; don't want to read through the entire thread. You're going for 13-3 here?

Yes, Muon2 has inside info that that is the Pubbie plan. The Gerry is going to have to be brutal to do it. Terminating Sutton and Kucinich appears to be very doable. Whether a triple or quad chop of Columbus will contain the Columbus Dems, the ones now, and the additional ones in the future, without endangering one or two Pubbie seats that are part of the chop down the road, remains to be seen.

[I spent the winter writing songs about getting better]

Even if the Massachusetts Dems were to draw as Republican a district as possible, it still wouldn't guarantee electing a Republican. The best you could hope for is about an Obama +6 or so district in southeastern Mass. There's less of an excuse when a compact Columbus district would easily elect a Democrat.

I remember reading on SSP that it's apparently possible to draw a McCain seat in SE Mass, and an Obama by only a few hundred votes one in the NE. Obviously will never happen. I'm kind of skeptical though, I tried a McCain seat in Connecticut and don't think it's possible, I could only get to about 52% Obama.

[minionofmidas]

Republican PVI seat, IIRC, not actually a McCain district. It didn't look too bad, either (whole towns), and had the right population for another, Democratic, seat southeast of it.

[krazen1211]

Approximate McCain percentages on this map.

Chabot   51% McCain
Schmidt   54% McCain
Turner   53% McCain
Jordan   58% McCain
Latta   52% McCain
Johnson   53% McCain
Austria   56% McCain
Boehner   62% McCain
Kaptur   27% McCain
Gibbs   56% McCain
Fudge   21% McCain
Tiberi   55% McCain
Ryan   26% McCain
Latuor   50% McCain
Stivers   30% McCain
Renacci   53% McCain

[minionofmidas]

How hard would it be to add some extra points to Chabot without weakening Schmidt or Turner (well, Turner's hypothetical successor)?
Jordan and Latta's districts look horrid - since it doesn't serve to make both safe seats, what is the purpose behind that? Something to do with residence? Or would a more reasonable-looking geographic split be even more uneven?
Also, you're taking a bit of a gamble with Johnson's seat - it's not a given that the disconnect between presidential and congressional vote shares in these kinds of areas is gone for good just because it was obliterated in 2010.

[Torie]

Let's see. If the Dem core of Columbus was carried by Obama by 70-30, with 720,000 people, that means the Obama margin was 288,000 (720,000 x .4).  So even if we do a quin chop, with that margin evenly divided, that means each of the 5 chop CD's starts out with an Obama margin of 57,600 votes!  A quad chop is -72,000 each. So all the chop CD's are going to have a tough time even getting to 50% McCain. And if it is much less, than suddenly the CD's become marginal. Krazen, are you sure that the margin for your little Columbus CD was 70-30? That does seem high.

I have a feeling that at best, one of the chop CD's is going to have to be marginal, to have much potential at all to get all of the other chop CD's out of the danger zone. I wonder how much of Boehner's Pubbie PVI  that he currently enjoys, he is willing to give up. It is going to have to be substantially cut. And propping up Schmidt at all is out.  She is going to have to be cut too, and that might put her in political trouble, which is fine with me, but maybe not for the Pubbies, or some of them at least, who just love her red dresses and big stupid mouth.  Is there any way to have one of the chop CD's include Provo, Utah?  Just asking.

[krazen1211]

Let's see. If the Dem core of Columbus was carried by Obama by 70-30, with 720,000 people, that means the Obama margin was 288,000 (720,000 x .4).  So even if we do a quin chop, with that margin evenly divided, that means each of the 5 chop CD's starts out with an Obama margin of 57,600 votes!  A quad chop is -72,000 each. So all the chop CD's are going to have a tough time even getting to 50% McCain. And if it is much less, than suddenly the CD's become marginal. Krazen, are you sure that the margin for your little Columbus CD was 70-30? That does seem high.

Columbus itself voted:

107574 McCain
232494 Obama

That's 32% McCain, and there are some conservatives areas in the northwest corner (I think?),  that can be cut out and towns like Bexley added.

[krazen1211]

How hard would it be to add some extra points to Chabot without weakening Schmidt or Turner (well, Turner's hypothetical successor)?
Jordan and Latta's districts look horrid - since it doesn't serve to make both safe seats, what is the purpose behind that? Something to do with residence? Or would a more reasonable-looking geographic split be even more uneven?
Also, you're taking a bit of a gamble with Johnson's seat - it's not a given that the disconnect between presidential and congressional vote shares in these kinds of areas is gone for good just because it was obliterated in 2010.

Latta lives in Bowling Green (Wood County) and is probably better known in the immediate Toledo area (western Lucas County). Beyond that, I wanted to give Latta the most conservative rurals (Putnam, Allen counties), and pull out the less conservative rurals (Huron county)

I don't think Chabot can be helped much; that said, 51% McCain should be enough given turnout there in midterm years.

[Torie]

Let's see. If the Dem core of Columbus was carried by Obama by 70-30, with 720,000 people, that means the Obama margin was 288,000 (720,000 x .4).  So even if we do a quin chop, with that margin evenly divided, that means each of the 5 chop CD's starts out with an Obama margin of 57,600 votes!  A quad chop is -72,000 each. So all the chop CD's are going to have a tough time even getting to 50% McCain. And if it is much less, than suddenly the CD's become marginal. Krazen, are you sure that the margin for your little Columbus CD was 70-30? That does seem high.

Columbus itself voted:

107574 McCain
232494 Obama

That's 32% McCain, and there are some conservatives areas in the northwest corner (I think?),  that can be cut out and towns like Bexley added.

Oh yes, the entire population doesn't actually vote does it? My bad! I need a nap.

[Chancellor Tanterterg]

Approximate McCain percentages on this map.

Chabot   51% McCain
Schmidt   54% McCain
Turner   53% McCain
Jordan   58% McCain
Latta   52% McCain
Johnson   53% McCain
Austria   56% McCain
Boehner   62% McCain
Kaptur   27% McCain
Gibbs   56% McCain
Fudge   21% McCain
Tiberi   55% McCain
Ryan   26% McCain
Latuor   50% McCain
Stivers   30% McCain
Renacci   53% McCain

This map puts Gibbs and Tiberi in the same district.

[krazen1211]

Let's see. If the Dem core of Columbus was carried by Obama by 70-30, with 720,000 people, that means the Obama margin was 288,000 (720,000 x .4).  So even if we do a quin chop, with that margin evenly divided, that means each of the 5 chop CD's starts out with an Obama margin of 57,600 votes!  A quad chop is -72,000 each. So all the chop CD's are going to have a tough time even getting to 50% McCain. And if it is much less, than suddenly the CD's become marginal. Krazen, are you sure that the margin for your little Columbus CD was 70-30? That does seem high.

Columbus itself voted:

107574 McCain
232494 Obama

That's 32% McCain, and there are some conservatives areas in the northwest corner (I think?),  that can be cut out and towns like Bexley added.

Oh yes, the entire population doesn't actually vote does it? My bad! I need a nap.

339k turnout in a city that had ~750k people at the time (now 787k) is rather low. National turnout in 2008 was 64%.

Imagine if Columbus Dems started voting....

[Torie]

Well, I assume Obama moved heaven and earth to max Dem turnout there.

Moving right along with my next adventure in math, given your numbers Krazen, if we do a quad chop, to get each quad CD up to a 3 GOP PVI (the minimum acceptable, with 3.5 being better), the balance of the CD's needs to be 55.5% McCain. And we need a pad for Dem growth, so we are really talking about a 4 GOP PVI minimum, and that is about a 56.30% McCain CD for each of the chop CD's that needs to be amassed before hitting the city of Columbus GOP death zone for its final quarter of population. It does not really look good. One of the CD's probably will have to be marginal. Better that, than putting more than one CD in potential competitive play down the road.

Granted, the turnout in Columbus may be a bit lower than outside it, which might give us another 50 basis points of breathing room. We shall see.

[Torie]

In other news, I have the new population numbers up now, the precincts match beautifully, and I am finding a lot of GOP or acceptable precincts in addition to what I found in my previous map to cull from the existing OH-09. I suspect that OH-09 will be able to do a thorough cleansing of the Dem precincts assigned to it to cleanse in Cuyahoga County, just about all of them over 55% Obama that I have as targets. I also found the 4 precincts I wanted to excise from OH-05 in Maumee, and have another 4  or so on deck to dump into it which are between 55%-60% Obama, population permitting. The population drop in the Dem zone outside of Columbus is really, really shocking - considerably more disastrous than in Michigan. Which is what may make this jihad feasible. I found another 3 or 4 precincts in Toledo less than 55% Obama to cut from OH-09. This is going swimmingly.  If only Columbus were within reach of the Youngstown CD. If only. Pity.

This is very important, because we need to pump up the Pubbie CD's for their attack on Columbus. We need to push, push and then push some more. If we are doing a Columbus chop, no precinct can be afforded to be "improperly" assigned, thereby losing efficiency.

Update:

The map below reflects the Dem disaster. Just about everything bad in Cuyahoga has been cleansed now, except for a few lightly populated link precincts to the marginally Pubbie white zone. And the population OH-11 still needs (I have not massaged the Eastern edge of it, but the changes will be marginal), plus what the Youngstown CD will need, means that everything anti Pubbie in NE Ohio pretty much will be cleansed, too. In fact, I may run out of Dem precincts to pick up, and one might need to get creative to find more. One Pubbie CD will have to pick up the 234,000 people in the white zone pocket. Mirror, mirror, on the wall, which CD shall it be? The pink zone is of uncertain number (because there is "another" CD 10 as well, down in SE Ohio), but that zone is subject to being shared between Pubbie CD's.



And here is the Toledo side of CD-09, with more modest changes, but now that I know the precinct data exactly, just about everything over 55% Obama is in OH-09, and not OH-05.

[cinyc]

The population drop in the Dem zone outside of Columbus is really, really shocking - considerably more disastrous than in Michigan.

Michigan's redistricting data hasn't been released.  We'll get it next week, at the earliest.  Unlike Ohio, Michigan lost population.  Detroit's population loss might just outpace Cleveland's.  I wouldn't rule it out - many of the 2009 estimates have had large errors, especially in urban areas.

[Dgov]

Mirror, mirror, on the wall, which CD shall it be?

Have you considered giving it to Steve Latourette via the lake Erie Precinct?  He's probably the best Republican to take in more Cleveland suburbs, and it'll probably strengthen his district and free up the Summit-county portion of his district.

[Torie]

This is probably the most complex map I have ever drawn, or will draw. Man it was a bear!  I was going along swimmingly, cheerfully picking  up the precincts I wanted to pick up in Akron to put into OH-11 (which is everything with a significant black percentage (over 10%), which is west of that lime green Magellan Strait cutting through downtown Akron (and it has to be there, because downtown is less black, and except for one precinct rather unpopulated (office buildings). I had to do this, because I just knew that if I cut them back, I would be putting something like 70% Obama precincts in a Pubbie CD, which is obviously not acceptable. So that part of the plan was fixed in stone.

But after finishing, I had a little problem. OH-11 was only 49.5% black VAP! Now what? There were no other black precincts in Akron to pick up that would help. In fact, picking up more would just dilute the black percentage further. There is a block of black precincts next to the Akron Airport, but that is on the east side of the Magellan Strait, and thus OH-11 could not get at them. I tried to move the strait to the east, but that would just shove too many Dem precincts into OH-14, since everything west of the strait obviously cannot go into OH-13, which itself is blocked by the strait. Consequently by moving the strait east,  those would have to be dumped into a Pubbie CD, and so that was out. Uh-oh, there is trouble in River City! Now what?

Since cutting some minority black precincts in Akron was out,  back to Cuyahoga we have to go, to start culling out white precincts from OH-11. That was a long and laborious job, as I hunted and hunted for precincts to cut from OH-11 that were less than 60% Obama. I finally made it, barely, but had to such up each and every precinct at the edges of OH-11 in Cuyahoga County to get there. In the process, I found additionally some pretty good precincts to put into OH-14, but not enough. So I had to suck up a Euclid precinct, and three Pepper Pike precincts that were 55%-58% Obama, and so forth. The end result, is that the eastern edge of OH-11 in Cuyahoga has become quite ragged. But it had to be done.

Why am I bothering with the Magellan Strait? The reason is that I need to have OH-16 (or some other Pubbie CD) go through it to suck up population from OH-14, so that OH-14 can pick up marginal precincts in Warren County or Mahoning or wherever, so that OH-13 is sufficiently stripped of population, that it will be able to suck up the Dem precincts in Canton and so forth. In the end, maybe I don't need it, which would be grand, but I wanted it in place in case I do. We shall see what OH-13 can achieve with its marginal precincts stripped, enabling OH-13, with its shortage of population, to make it up by moving  Summit and so forth to neutralize all the Dem territory around. One doesn't like having marginal precincts in a Dem pack CD. That is not the road to maximum efficiency.

In any event, this 50 basis point shortage in black VAP probably cost me close to a point in efficiency, but it could not be helped. If it turns out that OH-13 runs out of things to do, and it will have marginal precincts irrespective as to whether they are in a zone OH-14 can reach, or elsewhere, as I say, I may be able to jettison the strait, and get some of my lost point back by picking up the black precincts around the Akron airport which are east of the Magellan Strait. But I can't get all of it back. To do that would push OH-11 below 50% VAP black, and that cannot be allowed to happen. Time will tell.

I am so glad I am finished with this particular headache. Oh my! I think I will call this the Twin Straits map, since I have two of these puppies, one cutting through downtown Cleveland, and the other cutting through a lightly populated industrial zone in east Cleveland up to Lake Erie and along Lake Erie in Lakewood to the Cuyahoga Pubbie friendly zone. Notice I put one precinct in Lakewood in OH-11, and had the Cuyahoga strait cut south of it. Why? Because that precinct has about 3,200 people (huge precinct), and it is 73% Obama. That simply cannot be permitted of course, so thus that little workaround.

[Torie]

Mirror, mirror, on the wall, which CD shall it be?

Have you considered giving it to Steve Latourette via the lake Erie Precinct?  He's probably the best Republican to take in more Cleveland suburbs, and it'll probably strengthen his district and free up the Summit-county portion of his district.

LaTourette will have far too many people to get anywhere near the white Pubbie pocket. In fact, some CD will probably go into his NE corner of Ohio through the green Magellan Strait, so that LaTourette can move heavily into suburban and rural Warren County, and maybe a bit of Mahoning, stripping OH-13 of population, thereby enabling OH-13 can suck up Dem precincts in Canton, and other places as needed. OH-13 needs to include the Youngstown Dem zone of course, the Dem precincts in the Akron area east of the Magellan Strait, the Canton Dem precincts, and maybe go elsewhere, so that the adjacent Pubbie CD's can be pumped it, generating a big enough Pubbie pad to with some measure of Pubbie safety "quadralize" Columbus. That is the plan at least!  

A lot of things need to be done at once, and the objective function as it were to make it happen given all of the constraints is tight, very tight. It will be a very close call, as to whether we achieve all of the Pubbie objectives, which objectives Muon2 tells us that they in fact have. The Pubbies are greedy, and my job is to help their greed keep them in mere pig rather than hog status (pigs get fat, and hogs get slaughtered). And this map is so complicated, that I am not entirely sure the Pubbies will come up with my little strategy. So, this is one map that I am sure I will be forwarding to Team Boehner. I just don't trust them to do it right. And thus Torie is entering on stage right on this one.

[TJ in Oregon]

One thing to keep in mind when drawing anything in the city of Cleveland is that the wards and precincts have been renumbered since 2008. The area Torie refered to as "Mordor" just north of Parma is really not a 80%+ Democratic area like the precinct results and a precinct map might suggest because this area changed wards. In 2008 results it is wards 15 and 16 (NOT 13!) and about 60% Obama at most. The same thing happens in the far west end of the city (wards 20 and 21 instead of 19).

The most Republican parts of the City of Cleveland are definitely not along the lake in the northeast corner where Voinovich lives. Those areas (the Collinwood neighborhood) are 70%+ black and about 90% Obama. Conveniently for the redistricting, the most Republican parts  of Cleveland are in the area just north of Parma and on the far west end of the city along the Rocky River.

In addition you could shave more of the eastern suburbs off the 11th as well and give them to the 14th. Lyndhurst, Pepper Pike, the northern half of Beechwood, the eastern half of University Heights and the southeastern part of South Euclid are only 50-60% Obama areas. I'd worry less about making LaTourette safe than any of the other Republican incumbents because he has the label of being moderate and tends to attract some Democratic support. Obviously he has to have a reasonable district to win reelection but it doesn't need to be any better than R+0 or D+1.

[Torie]

One thing to keep in mind when drawing anything in the city of Cleveland is that the wards and precincts have been renumbered since 2008. The area Torie refered to as "Mordor" just north of Parma is really not a 80%+ Democratic area like the precinct results and a precinct map might suggest because this area changed wards. In 2008 results it is wards 15 and 16 (NOT 13!) and about 60% Obama at most. The same thing happens in the far west end of the city (wards 20 and 21 instead of 19).

The most Republican parts of the City of Cleveland are definitely not along the lake in the northeast corner where Voinovich lives. Those areas (the Collinwood neighborhood) are 70%+ black and about 90% Obama. Conveniently for the redistricting, the most Republican parts  of Cleveland are in the area just north of Parma and on the far west end of the city along the Rocky River.

In addition you could shave more of the eastern suburbs off the 11th as well and give them to the 14th. Lyndhurst, Pepper Pike, the northern half of Beechwood, the eastern half of University Heights and the southeastern part of South Euclid are only 50-60% Obama areas. I'd worry less about making LaTourette safe than any of the other Republican incumbents because he has the label of being moderate and tends to attract some Democratic support. Obviously he has to have a reasonable district to win reelection but it doesn't need to be any better than R+0 or D+1.

Hi TJ. Welcome to the forum. Each and every precinct in OH-11, with maybe 2 or 3 exceptions (other than the few precincts which prong down to Akron), is more than 58% Obama, and 97% of them, or something, are over 60%. I have a spreadsheet of the returns, and the precinct labels match the precinct returns perfectly, with the exception of Euclid for some reason.

[Torie]

OH-11 and OH-14 are I think in final form now. OH-13 is a partial draw, but what has been drawn so far, probably won't change. It has reached the edge of the Dem zone in Summit County, and now will prong down over a few Pubbie precincts presumably to Canton, its next stop. The Magellan Strait is gone. It proved unnecessary to use, and also highly inconvenient to use as it turned out. In that pink zone in Cuyahoga County was Garfield Heights, and that was heavily Dem but quite white, and it had to go into OH-11, which really necessitated a maximum black efficiency run for OH-11 into Akron, and the gap would have interfered with that. And OH-13 was now free to pick up any Dem precincts in Akron that OH-13 shed. It made life a lot easier.

By the way, the most heavily black area of Akron that was put into OH-11 (almost all black), looks like it kind of hugs a river doesn't it?

The main trick was getting OH-14 to the right population, without getting any plus 55% Obama precincts. It was boxed in by that natural Dem bridge of OH-13 from Youngstown to Akron, and that marked its Southern frontier. So it had to fight to get enough population in Summit, and that ultimately necessitated moving the OH-11 southern prong down to Akron a bit west, to give OH-14 some more room. And then I noticed some Dem precincts there, and so OH-11 had to pick them up, and lose some more in Akron, and on and on it went!  

[Torie]

The map draw continues.  The Stark County map shows in brown the over 56% Obama precincts, and those in pink are the under 56% Obama precincts. Because OH-13 has a bit of excess population, I have started to put in "unassigned," precincts that are between 56% and 58% Obama (that would be the white precincts  precincts in Stark County, and there may be some more of them, that I gave to OH-13, which will have to go back out, along its edges).  In that regard, I am now putting anything under 56% Obama into a Pubbie CD vis a vis OH-13, rather than using the 55% cut off that I was using.

I am doing all of this in case I need to cut out of OH-13, and put into a Pubbie CD, 55% to 58% Obama precincts in order  to get OH-13's population down to where it needs to be.  This assumes that I can't find enough, or much if any, under 56% Obama precincts in the OH-13 portion of Mahoning County to cut them out of OH-13, and put in OH-06 (and I suspect that there will not be many, since the CD line between OH-13 (old OH-17) and OH-06 in Mahoning if I recall correctly was pretty much the demarcation line between where the marginal territory in Mahoning ended, and the heavily Dem area of Mahoning began). So I may have to cull OH-13 around the edges from where it is now outside of Mahoning, to get its population down. But I can't do that on its northern edges very well next to OH-14, because OH-14 is trapped by its two Dem pack CD neighbors, OH-11 and OH-13. So,  moving its southern line down farther south a bit, will screw up its population, and I would really like to avoid doing that.

Boy is Stark County polarized, which is good for gerrymandering purposes, since it makes the Dem pack more effective. However, on the other hand, you can see now can't you, where the land bridge will be to get OH-13 from where it is now, to where it needs to go in Canton, right? Which means that there will be a horseshoed shape Pubbie pocket between the two prongs of OH-13. The may create a problem, because given how polarized Stark is, and to some extent the southern Pubbie suburbs of Akron, that pocket has the potential of being very heavily GOP, in fact so heavy, that whichever CD fills it, might get overly Pubbied up. So either that CD needs to find some Dem area to neutralize to gets its GOP percentage down, or the map will get another scar on it as two Pubbie CD's go into the pocket. Or maybe the CD which fills the pocket will be a long snake that is one of the CD's to quadralize Columbus, taking more than its share of those heavily Dem precincts. But that would make the map look well - really butt ugly. Hopefully that can be avoided. We shall see.

And so it goes: one problem gets solved, and another arises. Also note that Summit County is going to be a quinchop!  Which creates a certain order to things. We have Summit as a quinchop, Franklin County as a quad chop (probably), Stark is going to step up the plate to be a trichop (well actually it may well be a quadchop too, if OH-06 sticks its nose into its SE corner, in which event we will have to hope we can effect a trichop elsewhere, just so that we have a nice integer staircase up of chop "numerology," without missing one) , and of course the will be multiple bichops, and then we actually have some counties that will be entirely in just one CD. What a novel concept!   5-4-3-2-1 - I like it - chop, chop, chop!

Hopefully OH-16 is adequately Pubbie; it should be, but it needs to be checked. Heck it may even be over-Pubbied, in which event it will need to be moved a bit to find more marginal territory. Hopefully not however, if it is in the right partisan range. One must admit that it is a kind of pretty CD, right?   Moreover, it is fairly important that the southeast corner of OH-16 be left alone if possible, because that is where the incumbent lives (and he was mayor of Wadsworth to boot (in the SE corner of Medina County), so it is probably an area in which he "over-performs." I would like to keep the western and northwestern Republican suburbs of Akron that are near Wadsworth in OH-16 if possible.

The moral of the story on that is that you adjust lines between Pubbies CD's to to make the map look prettier to the extent you can (without screwing up the partisan CD numbers), so as to mitigate the unaesthetic erosity of the Dem pack CD's. That way, the map looks at least facially less of a "go for the max gerrymander, taking no prisoners," than it really is. With luck, it may even fool a few folks when looking at the map, who don't involve themselves in these sorts of partisan games. It is just good for Pubbie PR.  Make sense?  

[Torie]

The 3 Dem pack seats are completed (along with OH-14), and if the partisan numbers are OK, also OH-06 and OH-16.

Once I calculate the numbers for the 3 Dem pack seats, we will know the PVI for the balance of the state, and whether it is feasible for Columbus to be prudently chopped, assuming the PVI can be reasonably evenly distributed between the 13 CD's slated for Pubbies. My back of the envelope calculations are that the GOP 2008 PVI for the 13 Pubbie seats if the the 3 Dem pack seats are 70-30 Obama overall, would be a GOP PVI of 6.6, and if the 3 Dem pack seats are 65-35 on average, it would be about 5.8 (assuming turnouts are relatively even across the state). My guess is that the numbers are somewhere in-between 70-30 and 65-35, so call it a GOP PVI of say 6. In short the chop of Columbus should work absent severe distribution problems between the Pubbie seats (unlikely).  In that event, the Pubbies lose no points in Ohio, as opposed to the anticipated one - assuming they go for something reasonably close to the max gerrymander.

[Southern Senator North Carolina Yankee]

Just plain beautiful, Torie. Definately outdoing all your previous mastery at screwing over good government.

[Brittain33]

Here is a clean-looking 12-4 map which keeps the Columbus crack and puts urban Canton in Tim Ryan's district, but doesn't go the full 13-3 route.

http://twitpic.com/4af7sr