2020 Redistricting in Pennsylvania

[Torie]

This map might actually be something like what we end up with.

[Torie]

I turned the western CD's counterclockwise to get rid of a chop of Beaver County to the west of Pittsburgh. So my map above has been changed a tiny tad. Below is the existing map.

 

I think Wolff would be hard pressed to veto a map that eliminates a Pub CD, and hews to good redistricting principles, without savaging a Dem incumbent.

A map that does savage a Dem incumbent is this one:



As I told the boys at RRH who disliked what I did by not putting Schuylkill with the northern part of Chester, it is a thought crime to combine the richest bit of PA with a rural fossil fuel rust belt county, and the howls would drown out everything else. Less is more, even for Pubs sometimes.

[Torie]

Other than Wolf's reputation, if the Pubs draw a reasonable map, that can defended on its merits aside from the partisan implications at the margins arising from hewing to such principles, even though the Court is Dem controlled, it might be that a court that takes an oath to just follow the law, and not be partisan hacks, just maybe will not go for something that is clearly a Dem gerrymander. And it is not clear to me that it will end up in state rather than federal court in the context of a deadlock. In NY, with a deadlock, it ended up in federal court, which drew the lines. So there are risks associated with going for the max.

[Torie]

Here is another iteration that may have some appeal.

[Torie]

FYI you can easily nest two districts in Dauphin, York, Cumberland, and Lancaster.

I saw that, but if Berks goes with the Philly metro, then there's the unfortunate connection of Lebanon to Schuylkill to points north. It looked even worse in terms of compactness than the CD 9 on my map, so I dropped the idea.

Why is putting Lebanon and Schuykill in the same CD a thought crime?

[Torie]

Geographically (in the lowlands south of Blue Mountain) and culturally (very German) it is a natural cousin to Lancaster County and is quite an awkward fit with the upland coal regions to the northeast.  Better to have Lebanon + Lancaster + Dauphin + York + east Cumberland as the two south-central districts, ceding western Cumberland to a central PA district. 

These maps have a very similar design, which is not surprising since the authors tend to hew to very similar metrics. Which do you like better? For me, lack of compactness though trumps perceived COI's. I guess assuming the chop situation is not exacerbated unduly that I give more weight to compactness than most (Muon2 characterizes it as the "artist syndrome" in me, which although not meant as a compliment, I take as one  ) The district that runs from Pike to Schuylkill (man that county is hard to spell for me), drives me nuts. The space in which the S county is wedged between just begs to be absorbed by the Harrisburg based CD. I do think the tri-chop of Allegheny might be superior however. I tend to dislike tri-chops, but accept them if the payoff is sufficient.

[Torie]

I think you persuaded me, outside of SEPA, where I like my choices.

As a caveat, sure the final population will almost certainly not be within one precinct of "perfect," but the issue that drives this preference is not that, but rather whether it is still a very small discrepancy - small enough that it would be legal under the US  Constitution as now interpreted to tolerate such degree of  discrepancy, at least in order to effect  a good end, e.g., in order to avoid a county or municipal chop. That degree of discrepancy figure at the moment is within 1% for CD's. So basically, for this round, no CD can be more than 0.5% from the "perfect" number, or within a bit less than 4,000 persons from the absolute parity number.

You probably know all of this already.

[Torie]

Due to the most excellent crowd sourcing, this is my synthesis.

[Torie]

The PA high court may be Dem, but in the case of a deadlock, below is to my mind the most reasonable least change map from the one the court drew in 2018, and the Dems have a problem. It looks like the Dems might lose 2 seats and the Pubs pick up 1.

https://davesredistricting.org/join/64b8193a-0849-45ac-805a-b004e03fe853

[Torie]

Least change is a protocol, used most particularly by federal courts. PA is a state court issue most probably, but the same court drew the last map, and it would be a starting point, based on the choices it made last time. That is the "why," but yes, the court can do whatever it wants, and it's not a law.

[Torie]

I am once again asking people to recognize that with both Lamb and Doyle out, there is no longer any need to preserve the ugly current PA-17 Pittsburgh reach-around. Just grab all of the Penn hills region to the west of the city and use the rivers as guidelines for district borders. This near-guarantees a Biden seat, one that could even be more Democratic than the nation if you know what you are aiming for.

Problem is Rs controlling the state legislature would have a collective stroke if Allegheny is drawn too unfavorably to them.

We know the map is either going to the D courts or end up a D-favoring compromise (that'll leave their seats intact and favorable) because the D's used said court as leverage and got a map they liked, so why do they matter? In the latter scenario the R's would care more about securing their own then spending little capital on offense, and the courts in the former clearly wanted to do it in 2018 except Lamb lived in Mt. Lebanon and Doyle had a base in the Swissvale area.

Didn't they draw the 6D/5 swing/7 R map before the special election Lamb won or am I misremembering? Either way as hard as it is to believe 6D/5 swing/6R is D-leaning considering how f---ed the geography of the state is (inb4 "land doesn't vote!" Sorry, we've already decided property >>>> life in this country.) Any more and Rs prolly strip the 5-2 Court of the authority to appoint the tie-breaker to the commission at the first opportunity.

Granted, knowing the um... authoritarian the PAGOP has taken they'll prolly do that anyway whenever they take back the governor's mansion so maybe you're right after all.

The commission is only for legislative lines. Congressional districts are a bill that (likely will not) pass the chamber and the governor's pen, which is why an outside authority will be needed - the same court as last time. And said court last time took steps to correct for the state's geographic lean, which means a mild D-favoring map.

TBH I don't like 'corrective mapping' like this in fair Congressional lines, cause various statewide geographic advantages and the VRA work to cancel out any real political benefit nationwide, so one does not need to correct in a single state. But that is not the rules we play by, and that is not the rules the PA court recognized in 2018.  

https://davesredistricting.org/join/e9498594-cb08-4a0a-afe8-96e209594042
This is a quick mock-up of a map trying to follow the rules the PA SC has used.
Did I do a good job?

I think you did a good job, and made reasonable choices assuming a Dem friendly but not hackish court that will follow a least change approach in general to the map they drew last time. My main issue is the exchange of real estate between the two CD's in Allegheny County. I understand that makes the Lamb seat more Dem, but aside from that, what is the justification for the exchange that would be based on neutral redistricting metrics?

[Torie]

I am once again asking people to recognize that with both Lamb and Doyle out, there is no longer any need to preserve the ugly current PA-17 Pittsburgh reach-around. Just grab all of the Penn hills region to the west of the city and use the rivers as guidelines for district borders. This near-guarantees a Biden seat, one that could even be more Democratic than the nation if you know what you are aiming for.

Problem is Rs controlling the state legislature would have a collective stroke if Allegheny is drawn too unfavorably to them.

We know the map is either going to the D courts or end up a D-favoring compromise (that'll leave their seats intact and favorable) because the D's used said court as leverage and got a map they liked, so why do they matter? In the latter scenario the R's would care more about securing their own then spending little capital on offense, and the courts in the former clearly wanted to do it in 2018 except Lamb lived in Mt. Lebanon and Doyle had a base in the Swissvale area.

Didn't they draw the 6D/5 swing/7 R map before the special election Lamb won or am I misremembering? Either way as hard as it is to believe 6D/5 swing/6R is D-leaning considering how f---ed the geography of the state is (inb4 "land doesn't vote!" Sorry, we've already decided property >>>> life in this country.) Any more and Rs prolly strip the 5-2 Court of the authority to appoint the tie-breaker to the commission at the first opportunity.

Granted, knowing the um... authoritarian the PAGOP has taken they'll prolly do that anyway whenever they take back the governor's mansion so maybe you're right after all.

The commission is only for legislative lines. Congressional districts are a bill that (likely will not) pass the chamber and the governor's pen, which is why an outside authority will be needed - the same court as last time. And said court last time took steps to correct for the state's geographic lean, which means a mild D-favoring map.

TBH I don't like 'corrective mapping' like this in fair Congressional lines, cause various statewide geographic advantages and the VRA work to cancel out any real political benefit nationwide, so one does not need to correct in a single state. But that is not the rules we play by, and that is not the rules the PA court recognized in 2018.  

https://davesredistricting.org/join/e9498594-cb08-4a0a-afe8-96e209594042
This is a quick mock-up of a map trying to follow the rules the PA SC has used.
Did I do a good job?

I think you did a good job, and made reasonable choices assuming a Dem friendly but not hackish court that will follow a least change approach in general to the map they drew last time. My main issue is the exchange of real estate between the two CD's in Allegheny County. I understand that makes the Lamb seat more Dem, but aside from that, what is the justification for the exchange that would be based on neutral redistricting metrics?

The goal was to keep it very competitive. The seat is expanding north into very R territory anyway, so giving it those parts of Allegheny County is a way to maintain its partisanship.
This of course is important for proportionality, which is part of what the court's criteria is. But it's defendable more generally on neutral grounds as well, as helping maintain a competitive district.

Are you sure that making seats competitive or proportionality was explicitly embraced by the court? In addition, the existing map has an ugly chop of the black community, creating an erose line to boot. What I did in the map below, is have the Pittsburgh seat take the balance of the two chopped cities, and then smooth out the line south of the river, and then take one small city in the other indented area to the south, and then chop a small city by about 300 people. The Lamb CD is still quite competitive. I would be surprised if the court goes for erosity and chopping the black community to make the seat a couple of points more Dem.

https://davesredistricting.org/join/0aba5052-7f1f-4d91-a9d7-399abe34d796

[Torie]

I am once again asking people to recognize that with both Lamb and Doyle out, there is no longer any need to preserve the ugly current PA-17 Pittsburgh reach-around. Just grab all of the Penn hills region to the west of the city and use the rivers as guidelines for district borders. This near-guarantees a Biden seat, one that could even be more Democratic than the nation if you know what you are aiming for.

Problem is Rs controlling the state legislature would have a collective stroke if Allegheny is drawn too unfavorably to them.

We know the map is either going to the D courts or end up a D-favoring compromise (that'll leave their seats intact and favorable) because the D's used said court as leverage and got a map they liked, so why do they matter? In the latter scenario the R's would care more about securing their own then spending little capital on offense, and the courts in the former clearly wanted to do it in 2018 except Lamb lived in Mt. Lebanon and Doyle had a base in the Swissvale area.

Didn't they draw the 6D/5 swing/7 R map before the special election Lamb won or am I misremembering? Either way as hard as it is to believe 6D/5 swing/6R is D-leaning considering how f---ed the geography of the state is (inb4 "land doesn't vote!" Sorry, we've already decided property >>>> life in this country.) Any more and Rs prolly strip the 5-2 Court of the authority to appoint the tie-breaker to the commission at the first opportunity.

Granted, knowing the um... authoritarian the PAGOP has taken they'll prolly do that anyway whenever they take back the governor's mansion so maybe you're right after all.

The commission is only for legislative lines. Congressional districts are a bill that (likely will not) pass the chamber and the governor's pen, which is why an outside authority will be needed - the same court as last time. And said court last time took steps to correct for the state's geographic lean, which means a mild D-favoring map.

TBH I don't like 'corrective mapping' like this in fair Congressional lines, cause various statewide geographic advantages and the VRA work to cancel out any real political benefit nationwide, so one does not need to correct in a single state. But that is not the rules we play by, and that is not the rules the PA court recognized in 2018.  

https://davesredistricting.org/join/e9498594-cb08-4a0a-afe8-96e209594042
This is a quick mock-up of a map trying to follow the rules the PA SC has used.
Did I do a good job?

I think you did a good job, and made reasonable choices assuming a Dem friendly but not hackish court that will follow a least change approach in general to the map they drew last time. My main issue is the exchange of real estate between the two CD's in Allegheny County. I understand that makes the Lamb seat more Dem, but aside from that, what is the justification for the exchange that would be based on neutral redistricting metrics?

The goal was to keep it very competitive. The seat is expanding north into very R territory anyway, so giving it those parts of Allegheny County is a way to maintain its partisanship.
This of course is important for proportionality, which is part of what the court's criteria is. But it's defendable more generally on neutral grounds as well, as helping maintain a competitive district.

Are you sure that making seats competitive or proportionality was explicitly embraced by the court? In addition, the existing map has an ugly chop of the black community, creating an erose line to boot. What I did in the map below, is have the Pittsburgh seat take the balance of the two chopped cities, and then smooth out the line south of the river, and then take one small city in the other indented area to the south, and then chop a small city by about 300 people. The Lamb CD is still quite competitive. I would be surprised if the court goes for erosity and chopping the black community to make the seat a couple of points more Dem.

https://davesredistricting.org/join/0aba5052-7f1f-4d91-a9d7-399abe34d796

Proportionality was definitely a factor, albeit only one factor, in how the PA SC drew the lines last time (just look at how the Bucks County CD was drawn).
It'd be surprising if a court were to place having as much of the blacks of Allegheny County in one seat above all considerations either. The existing court map has Penn Hills (37% AA) in the northern Allegheny seat. Let's not forget this is a county that is only 13% black. If it were, say, double that, then there might be a case. In any case, to the extent that it would be even relevant, it's probably likely that if given a choice, AAs in the county would prefer to be divided for sake of improving D chances in the swing district, as opposed to packing Ds into one safe seat.

Below is a link to the court decision last time. The court focused on minimizing splits and compactness. They failed to explain why two municipalities were chopped in Allegheny County rather than just one, and just said in general all the competing plans had more splits. Nor did they comment about the seeming unnecessary erosity in the county. The decision was quite conclusory. The litigants would be wise to channel its metric and point out where it seemed not to be followed. There was no mention of proportionality or competitiveness.

https://www.brennancenter.org/our-work/court-cases/league-women-voters-pennsylvania-v-commonwealth-pennsylvania

[Torie]

All things considered Cartwright's district is pretty compact for how D-leaning it is. Interesting to go after Smucker instead of Perry, though it may be prudent as York's pastorals/ex-sundown towns are more conservative than Lancaster's.

Lancaster is quite fertile ground in general for Democrats. The city itself is growing, the suburban areas of the county are trending Blue, and the blood red rurals are losing population quite quickly. Lancaster County could flip in a decade or two.

Aren't the rurals gaining population due to Amish growth ?

Technically you're right, the Amish f*** like rabbits. However, for political purposes, this fact is largely irrelevant, since barely any Amish people vote. They just aren't a significant factor in electoral politics. Plus, Lancaster City and its suburbs are growing at a much faster rate anyway.

The actual city of Lancaster shrunk still for the decade FYI.

That is unexpected.
Why did that happen? What factors are at work here?

The city is old and compact without much real estate so it would be expected to not have population growth.

[Torie]

All things considered Cartwright's district is pretty compact for how D-leaning it is. Interesting to go after Smucker instead of Perry, though it may be prudent as York's pastorals/ex-sundown towns are more conservative than Lancaster's.

Lancaster is quite fertile ground in general for Democrats. The city itself is growing, the suburban areas of the county are trending Blue, and the blood red rurals are losing population quite quickly. Lancaster County could flip in a decade or two.

Aren't the rurals gaining population due to Amish growth ?

Technically you're right, the Amish f*** like rabbits. However, for political purposes, this fact is largely irrelevant, since barely any Amish people vote. They just aren't a significant factor in electoral politics. Plus, Lancaster City and its suburbs are growing at a much faster rate anyway.

https://lancasteronline.com/news/local/was-2020-a-breakout-year-for-amish-voters-heres-what-the-numbers-show/article_f77af684-32a7-11eb-b3ec-13a56697652f.html

[Torie]

LOL. It apparently is impossible for the human species to avoid descending into the pit of partisan hackery and gerrymandering while claiming it is all about good government and the public interest. Most of the maps I glanced at made we want to laugh and cry at the same time. They are all going to hell, metaphorically speaking, where the Pubs have to listen to Kamala Harris species 24/7, and the Dems Trump speeches. 

[Torie]

Can someone explain why there’s such a stigma around splitting Bucks, when other counties around Philly that are simillar in size can be split like 3 ways no problem? One could argue that the Northern and South end of Bucks are very different.

To me a lot of it comes from 1) simple inertia, 2) it simply not being necessary in the math, and 3) the fact that areas around it either vary a lot internally (MontCo) or are opposite to that (NW Philly, Lehigh Valley).
There's also the fact that Bucks is just large enough to render itself almost a CD to itself, and the history of districts running from MontCo to Philly. Splitting Bucks just isn't necessary, never has been, and there are good alternatives.
Compare that to MontCo - in the center of it all and relatively easier to split.

In addition to the points you listed, Bucks is also on a state line on two of its four sides. It is also relatively homogeneous. It also is highly competitive politically, and perhaps more elastic than average (not sure about the latter point). Chopping it looks instantly suspect as part of a nefarious gerrymandering plot.

[Torie]

By the does anyone know more about the York-Gettysburg Carlisle area historically? Why was this area relatively Democratic relative to PA in the late 1800's/early 1900's? I assume the same type of settlers?

https://en.wikipedia.org/wiki/1900_United_States_presidential_election_in_Pennsylvania

[Torie]

The court has 4 pretty hard core Dem partisans so it may not matter, but the court in 2018 in the League of Women's Voters case (the same 4 hard cores provided the votes for that decision 4-3, with one concurrence that thought only truly hideous gerrymanders should be banned), was very hostile to gratuitous county and municipal chops, so I would be surprised if the court goes for such chops, particularly one as noticeable as Pittsburgh, when it can get its proportional map without doing such chops.

[Torie]

The court has 4 pretty hard core Dem partisans so it may not matter, but the court in 2018 in the League of Women's Voters case (the same 4 hard cores provided the votes for that decision 4-3, with one concurrence that thought only truly hideous gerrymanders should be banned), was very hostile to gratuitous county and municipal chops, so I would be surprised if the court goes for such chops, particularly one as noticeable as Pittsburgh, when it can get its proportional map with doing such chops.

Who was the concurring dem?

edit: Justice Baer.

It was fascinating listening to the Dem lawyers that muni lines are a community of interest except when they are not, and COI's may be smaller or larger. Aside from race, and I guess metro areas as a whole (town v country), I have trouble perceiving what that is other than 'values," aka partisan preference.   And the Dem lawyers did not enlighten me. It is a test drive for the same argument to be made in NYS, arguing the map is "fair" due to hewing to COI's (Dems?), even though it deems Pubs to be about as desirable as covid.

[Torie]

After listening to a fair amount of oral argument, I am persuaded that the court should go for a least change map. The maps submitted are all legal, and that is what a court should do when there is deadlock. As part of the change, the following order of priority should be maintained:  reducing splits and their size, keeping existing splits, hewing to metro areas, maintaining partisan balance, and finally reducing erosity, in that order of priority more or less (thus, e.g., losing the split of Penn Hills, but keeping it in the suburban Pittsburgh area CD rather than moving it to the city CD, which is what reducing erosity and keeping the black population together would otherwise suggest). I think the map below accomplishes the above very well. Which of the map options before the court comes closest to this map? I find neither the legal case nor how to draw that map that problematical it turns out. The map kind of drew itself almost. I should have tried to get someone to submit it.  

It will be interesting to see how hack the court is. While a couple of the justices sounded hack (one in particular, that woman who said it is better to split Pittsburg than Washington County, ignoring that if not Washington County, another county would have to be split, get a grip honey)), I don’t think I heard as many as four different voices on the court as hack. PA does not equal NC it seems.

https://davesredistricting.org/join/95acf932-1aae-478d-8806-5d8f11125c9a

[Torie]

Torie, are there 17 districts in that map or 18?

Oh F**K! No matter the map was so easy to draw! I used the existing map and just redrew. Sad!  


Time to take Roby for his walk. I am pleased to report that he seems to have recovered from a severe case of diarrhea  that has lasted 4 days. He finally started eating his white rice (the bland diet regime) at 2 am last night, and finally slept comfortably. If he had not, he was going to the emergency clinic this morning. It is quite a relief actually. And it all happened while Dan is being feted in Hawaii at an art opening of his work at the university there.

[Torie]

Torie, are there 17 districts in that map or 18?

Oh F**K! No matter the map was so easy to draw! I used the existing map and just redrew. Sad! 

Yes see the issue is a pub seat has to be cut but the dems have to take all those icky pub voters.

Yeah, I see the BS about splitting Pittsburgh, as if a split of 100,000 is the same as a split of 1,000 (if 4 justices buy that, I am sending them to hack hell, which is already way overpopulated with state high court judges), and Harrisburg will need to get more Pub.

[Torie]

OK, here is a map with the number of CD’s the census bureau thinks is correct. What map is closest to this effort from the oracle up on Mt. Olympus? I still think this is the only way to do a good faith least change map, following my metrics (see above), and allegedly the court regarding chops, assuming there are not 4 Dem hacks who ignore the size of the chops to boost the Dems. I am pleased and amused that the size of the requisite muni chop in Alleghany County would be 62 people. Take that queen justice hack!  

 

https://davesredistricting.org/join/f98065c8-c133-4483-86bb-2ffd314b13c6





And 88.165% of the population is in the same district, which is not the theoretical maximum but close it it. I wonder what the percentage was for the least change Carter map that was submitted.

[Torie]

Changes from 2020

PA-01: Biden +5.8 to Biden +4.6
PA-02: Biden +41.0 to Biden +42.7
PA-03: Biden +83.2 to Biden +80.9
PA-04: Biden +24.1 to Biden +18.9
PA-05: Biden +31.1 to Biden +32.3
PA-06: Biden +15.0 to Biden +14.8
PA-07: Biden +4.8 to Biden +0.6
PA-08: Trump +4.4 to Trump +2.9
PA-10: Trump +2.9 to Trump +4.1

Good for PA-08 but crappy for PA-07, while Dean is losing a chunk of Dems - where did they go? Are those the ones going to PA-05? Because that district is only 1% more Dem.

PA-01 and PA-10 kinda suck. Little to no chance that Dems can get those back even in a good year at this point with Fitz and now that PA-10 is more GOP and Perry won more than Trump in 2020.

You forgot the most important change, R's lose an entire seat yet 3/4 Dem swing seats get shored up.

Right, but the GOP should lose a seat. They have the most population loss and they lost the state in 2020.

That was always agreed upon since the begining of this cycle. The Democratic seats still needed to pick up 300k in population of blood red territory. Considering 3/4 swing seats for Democrats got shored up despite that, that's a swell courtmander they got.

https://davesredistricting.org/maps#viewmap::48bc041f-ff25-42bc-a74d-d90f325dde0e

Look at Palandios map. Democrats gain like 14 points from PA17 moving it from somewhere around Trump +8 to Biden +6. They gain like 10 points for PA06 with the tri chop of Berks to drown it out with Montgomery and Chester. . That's a whole 25 points worth gained in swing districts .

It's a neutral map in wbrock67's alternate reality where Biden actually won PA by 9.

Huh? What the f**k are you actually talking about? Biden won Pennsylvania. It's a 9-8 Biden map. Biden won PA slightly. The map has a slight Biden won edge. How is that objectionable? This is literally what a fair map looks like. If Trump won PA slightly, then you'd expect it to be a 9-8 Trump map. How are people actually having an issue with this?!

Still trying to understand why Dragging Montgomery County all the way out to Tri Chop Berks needs to happen for a 9-8 Biden map.

That is the least change approach. My map did exactly the same thing.