2020 Redistricting in Pennsylvania

[President Punxsutawney Phil]

https://davesredistricting.org/join/d2df5ff8-8b9b-4591-86ea-f668ac0dcbe1

I made this map.
The cascading impact from larger district sizes forces the 4th to take all of the 5th's share of MontCo and even part of DelCo which in turn disturbs PA-06, forcing it north and west. This is also visible with more of Monroe being placed in the 7th. The 8th is forced to take in yet still more of Luzerne, which combined with the 6th's movements, in effect merging the old 9th and 12th. Farther west, the 14th gains the remainder of Westmoreland. Additionally the Erie CD is largely out of the reach of Dems, and has to move east. Conor Lamb's CD has to take territory from the SW PA CD and the Erie CD, moving it to the right, but this is countered by it having more of the deep blue areas just east of Pittsburgh. In the end it shifts 2 points to the right, still winnable for Conor Lamb and still competitive especially in the wake of Dem trends in traditionally R areas in Allegheny County.

Overall, you have 7 Clinton seats (1, 2, 3, 4, 5, 7, 17), and 10 Trump ones (6, 8, 9, 10, 11, 12, 13, 14, 15, 16).
I'm not entirely sure if 6 voted for Trump or not, but the fact it is R+3 likely makes it too unlikely to be a Clinton district, notwithstanding the trends in Chester County - the most strongly D trending areas are probably placed in the 5th here anyway. But the margin would be too close for Rs to feel any comfort here.
Also Punxsutawney Phil is now in the 12th district but that doesn't really matter or anything.

[President Punxsutawney Phil]

Not crossing the Allegheny River would have made the map too R-friendly, and would have made the 16th hard for Dems. Not to mention its predecessor already crosses the river in that same area.
I'm also unwilling to alter the arrangements for Lebanon because 1) it being placed in with Lancaster makes a rational arrangement with the rest of York impossible and 2) it can't go with Dauphin because doing so destroys any competitive element in PA-10. Doing either also forces an additional county split.
Lebanon also, as far I recall, already is in the 9th. The only thing that changed now was it being more of a southern appendage of a more rural district due to PA-06 drifting north, but such things ought to be expected when seats get merged.

[President Punxsutawney Phil]

The Lehigh seat should clearly have Carbon in a fair map as its part of the Lehigh valley, honestly the fact the current map doesn't have Carbon is an absolute disgrace for a "fair map" as Lehigh+ Northampton+ Carbon is perfectly 1 seat with 2010 pop +6k people.

I don't disagree on the particulars of the shape of Lehigh CD, even though I think that the current map, for sure, absolutely qualifies as a fair map. Lehigh+Northampton+Carbon is just too good a pairing to avoid, though for optimal results, it ought to be paired with Luzerne+Monroe+Lackawanna (with Pike being sent into a CD running along the New York state border).

[President Punxsutawney Phil]

And Bucks has been a district community like that for decades lol, its not gonna be split so you can have your D hack gerrymander.

Even in a D gerrymander, Bucks shouldn't be split. A D gerry takes Bucks and gives it areas in southeastern Montgomery or snakes into more Dem parts of Philly; a fair map probably takes areas along the MontCo border; and an R gerry gives it far northeastern parts of MontCo.

[President Punxsutawney Phil]

Tbh, splitting Bucks can only really be called a Dem hack gerrymander, not that splitting Bucks is a precondition for qualifying. The only grounds I can think of that could justify it internally is "Dems are badly served by Bucks being whole, let's hive off the southern part and create a Likely D seat out of it". I've never seen a GOPer, online or otherwise seriously propose splitting it either.
It is my view that any map that splits Bucks is not truly good as a rule, and is quite possibly trash. It just makes no sense, has no real utility, and creates two additional county splits by default, unless the seat avoids MontCo entirely (which is possible but rare in my experience).

[President Punxsutawney Phil]

@blarite, did you split the city of Pittsburgh?

I can't tell.

From the looks of it, it doesn't.

[President Punxsutawney Phil]

28 seat PA.
Map was drawn with election data disabled. It was enabled after I finished drawing.
https://davesredistricting.org/join/6a7f7896-dcf2-42d4-b7fc-8d7584f31e29
This is the DRA link.

[President Punxsutawney Phil]

Fun fact: on 17 seats and 2018 population estimates you can make a district from Monroe+Pike+Lackawanna+Luzerne, with a deviation of exactly 0 from quota.

[President Punxsutawney Phil]

My effort at a fair PA. Shading is of 2012/2016 PVI, but for some reason seems to not be purely based on said metric? County splits are kept to the minimum when possible. Both Lehigh and Scranton CDs have very light R+ PVIs, at R+1.5 or less. PA-06 is not prevented from becoming R leaning; it becomes R+2.5. The Pittsburgh CD now takes in all of Pittsburgh+all of Allegheny south of and east of the three-way rivers except Dravosburg+about 57k in Washington County. PA-16, successor to old PA-17, moves about 1 point to the left due to the Pittsburgh CD following river boundaries.
https://davesredistricting.org/join/500e7b37-b19d-4fc8-afd0-cb8209ef30c5
DRA link

[President Punxsutawney Phil]

I don't regard a map as being fair if it's not proportional and accurately translates votes into election results. At some level FPP can't be fair, which is why I oppose FPP, but map-makers should try to make FPP as fair as possible.

Assuming you are responding to Tim, 9-8 with a bunch of swing seats is very reasonable for an R+1/2 state like PA.

Thoughts on my 17 seat fair map?

[President Punxsutawney Phil]

I'm currently doing a 51-seat map for Pennsylvania. With 51 seats things are not that bad for Democrats:

17 seats in SEPA, of which 16 are tilt D to safe D (by 2012/2016 PVI), one is tilt R

5 seats in Allegheny County, of which 3 are lean to safe D, while the two R-leaning ones are trending D

Scranton (D+6)
Allentown (D+4)
Erie (D+4)
Reading (D+3)
Harrisburg (D+3)
Bethlehem (tossup)
Easton, Stroudsburg (tossup)
Lancaster (R+1)
Wilkes-Barre (R+3)
State College (R+5)


19 likely to safe R

Isn't this basically a state senate map?

[President Punxsutawney Phil]

I'm currently doing a 51-seat map for Pennsylvania. With 51 seats things are not that bad for Democrats:

17 seats in SEPA, of which 16 are tilt D to safe D (by 2012/2016 PVI), one is tilt R

5 seats in Allegheny County, of which 3 are lean to safe D, while the two R-leaning ones are trending D

Scranton (D+6)
Allentown (D+4)
Erie (D+4)
Reading (D+3)
Harrisburg (D+3)
Bethlehem (tossup)
Easton, Stroudsburg (tossup)
Lancaster (R+1)
Wilkes-Barre (R+3)
State College (R+5)


19 likely to safe R

Isn't this basically a state senate map?

You're right. So I should better draw a State Senate map with 50 seats instead of 51 seats.

What this probably means is that with a ~50-seat map the Democrats' geographic disadvantage is smaller than with 17 or 18 seats. The current Pennsylvania State Senate has of course 28 Republicans and 21 Democrats (22 elected as Democrats) which would point to a R advantage.

How do the single areas compare?

SEPA has 14 Democrats which is close to the ceiling except for the 6th where Tomlinson seems to be a strong R incumbent in a D-leaning seat.

Allegheny has 4 Democrats, the 38th and the 43rd seem to be tossups, as expected (one D-held, one R-held).

Scranton is D.
Reading is D.
The Allentown/Bethlehem/Easton area has only one Democrat, due to both the Bethlehem/Easton seat being a D pack and Browne being a strong incumbent in the 16th.

Erie hasd Republican Laughlin elected in a D-leaning seat in 2016.
Harrisburg is drawn less favorably for the Democrats than is could be, Republican DiSanto gained this seat in 2016.
Lancaster is in traditionally R territory.
Wilkes-Barre elected Yudichak, but he left the Democrats.

Hence most of the Democrats' disadvantage is due to three Republicans in D-leaning seats and some is due to the map being drawn unfavorably for them (Harrisburg, Allentown).

How much priority do you place on reducing the # of county splits?

[President Punxsutawney Phil]

Hate to beat this CRR drum here, but imagine what could be done with ~28 seats?

PA has a problem since all of its D votes are packed into tight little circles, so larger districts done without spaghetti strips drowns them all out. Smaller districts on the other hand make things more interesting.

Here's a 16-12 D gerrymander I've drawn (using 2010 population figures). There are just 26 county splits and Pittsburgh is kept whole.

https://davesredistricting.org/join/856864b1-aeb8-4996-ab0c-124dc75cb423

Interesting. I would have expected you feeding Pittsburgh in to areas farther north, since northern Allegheny is the most R area in the county. But the way you have it also makes sense.

[President Punxsutawney Phil]

Hate to beat this CRR drum here, but imagine what could be done with ~28 seats?

PA has a problem since all of its D votes are packed into tight little circles, so larger districts done without spaghetti strips drowns them all out. Smaller districts on the other hand make things more interesting.

Here's a 16-12 D gerrymander I've drawn (using 2010 population figures). There are just 26 county splits and Pittsburgh is kept whole.

https://davesredistricting.org/join/856864b1-aeb8-4996-ab0c-124dc75cb423

Interesting. I would have expected you feeding Pittsburgh in to areas farther north, since northern Allegheny is the most R area in the county. But the way you have it also makes sense.

You're right, I suppose a more effective gerrymander would have a district snaking around from Beaver County to the Mon River valley. I did want to maintain some amount of compactness in my map.

Is it really the case that this would hurt compactness that much though?

[President Punxsutawney Phil]

Hate to beat this CRR drum here, but imagine what could be done with ~28 seats?

PA has a problem since all of its D votes are packed into tight little circles, so larger districts done without spaghetti strips drowns them all out. Smaller districts on the other hand make things more interesting.

Here's a 16-12 D gerrymander I've drawn (using 2010 population figures). There are just 26 county splits and Pittsburgh is kept whole.

https://davesredistricting.org/join/856864b1-aeb8-4996-ab0c-124dc75cb423

Interesting. I would have expected you feeding Pittsburgh in to areas farther north, since northern Allegheny is the most R area in the county. But the way you have it also makes sense.

You're right, I suppose a more effective gerrymander would have a district snaking around from Beaver County to the Mon River valley. I did want to maintain some amount of compactness in my map.

Is it really the case that this would hurt compactness that much though?

Okay, I made your suggested changes in Allegheny County (and fiddled with Chester County while I was at it). I'm happy with this version.
https://davesredistricting.org/join/856864b1-aeb8-4996-ab0c-124dc75cb423

iirc, didn't the old map have a Pittsburgh CD that is D+17? Meaning this lowers the Dem PVI a few points (not that it's a bad thing. It's a good thing)?

[President Punxsutawney Phil]

Splitting Bucks is a bad idea, was a bad idea, and will always be a bad idea in my book. It's not something I ever will do in a serious map and any post about a map that has such a split will not get a recommend from me, per a matter of principle. (unless it's about how it's a terrible idea)

[President Punxsutawney Phil]

Splitting Bucks is a bad idea, was a bad idea, and will always be a bad idea in my book. It's not something I ever will do in a serious map and any post about a map that has such a split will not get a recommend from me, per a matter of principle. (unless it's about how it's a terrible idea)

One of Montgomery, Bucks, or Chester will have to be split. Why is it inherently fairer to split Chester or Montgomery than Bucks?

1. it is...erm...extremely weird, never seriously floated IRL in any redistricting plan that came close to going into effect.
2. for all the talk of "oh, Bucks isn't a single CoI", Montgomery by that same yardstick is even less of a cohesive CoI, and Chester is no more of a single CoI by these metrics than Bucks is.
3. Bucks is just short of quota for a CD. I consider it highly unelegant to split it. Better an extra chop of Philly than a macro-split of an otherwise whole county.
4. Geography basically dooms at least one of either Chester or MontCo to not be whole in a CD or have a whole CD within them. Geography likely dooms Delaware to being split in some fashion.
5. Splitting Bucks means you have to do weird things to northern Bucks, which either has to go with the Lehigh Valley (yuck), or northern Montgomery and Berks (also yuck).

[President Punxsutawney Phil]

Splitting Bucks is a bad idea, was a bad idea, and will always be a bad idea in my book. It's not something I ever will do in a serious map and any post about a map that has such a split will not get a recommend from me, per a matter of principle. (unless it's about how it's a terrible idea)

One of Montgomery, Bucks, or Chester will have to be split. Why is it inherently fairer to split Chester or Montgomery than Bucks?

1. it is...erm...extremely weird, never seriously floated IRL in any redistricting plan that came close to going into effect.
2. for all the talk of "oh, Bucks isn't a single CoI", Montgomery by that same yardstick is even less of a cohesive CoI, and Chester is no more of a single CoI by these metrics than Bucks is.
3. Bucks is just short of quota for a CD. I consider it highly unelegant to split it. Better an extra chop of Philly than a macro-split of an otherwise whole county.
4. Geography basically dooms at least one of either Chester or MontCo to not be whole in a CD or have a whole CD within them. Geography likely dooms Delaware to being split in some fashion.
5. Splitting Bucks means you have to do weird things to northern Bucks, which either has to go with the Lehigh Valley (yuck), or northern Montgomery and Berks (also yuck).

1. Not really an argument in itself
2. Bucks is more cohesive than Chesco??
3. 130,000 people isn't really "just short," and this becomes more salient with every passing census
4. Do agree that Bucks is more geographically cornered, but it doesn't mean it can't be split, plus demography exists too
5. Reading is definitely more connected to the Pottstown area than to West Chester. It makes way more sense demographically to put Berks with upper Montco, heck even Upper Bucks than to put Chesco with Lancaster or Chesco with Berks (extreme yuck)

1. Agree to disagree
2. I didn't say that. I placed them on the same level. But come to think about it, you could very rationally argue that the West Nottingham doesn't belong in the same CD as Treddyffrin to a greater extent than, say, Springfield and Bensalem...
3. Bucks by itself can easily form a CD with the addition of parts of Montgomery or Philly or both (but, in a vacuum, preferably just one), either of which might be advanced by some kind of CoI-oriented argument. In a majority of cases, I prefer Montgomery.
4. The supposed "demography" reasons you cite are mere ether, or may as well be so in practice, compared to other, more weighty things.
5. Berks is generally best paired with whatever leftovers exist from MontCo or Chester, plus adjacent territory east of the Schylkill (such as Lebanon County, Northumberland County, etc).

[President Punxsutawney Phil]

It's not necessarily an either/or between keeping Bucks whole and keeping Delaware whole. This map does both, whilst containing two performing VRA districts: https://davesredistricting.org/join/6a9fa4a4-21e8-4144-b812-0d137b7c2ddd

I think the split of NE Philly is a little ugly here, and I personally think you'd have more cohesive districts if you gave it all to PA-1, hived off northern Bucks to the Berks district and adjusted the Montgomery district to fit, but it's perfectly serviceable.

Interesting middle ground proposal. I do agree it's a bit ugly how NW Philly is split. Imo, the linkage of Northern MontCo with Berks might have to be sacrificed.

[President Punxsutawney Phil]

If you reread the 2010 redistricting thread, you'll notice a lot of people talking about the traditional taboo on putting York with Lancaster, and that was smashed to bits in the 2018 reredistricting.

Map drawers often tend to draw districts because "it's always been done this way." See KY-01 and KY-02 for a good example. These taboos aren't usually rooted in a clear eyed understanding of communities of interest--rather blind adherence to previous map-drawing.

Your example is actually quite wrong here. Look at 1980, for instance, KY-01 and KY-02 have a logical border. KY-01's weird tail dates back to 1992 - it is not a "tradition", it is too young to even qualify as one. To compare Bucks being whole to KY-01 and KY-02 having their weird mutual border is like apples and oranges.

[President Punxsutawney Phil]

I guess I'll post my map, too, it's a hypothetical Democratic leaning court map

https://davesredistricting.org/join/4b0ab8e8-c4ac-4306-9eea-a120807c3a47

So this was another somewhat unique map.
The Philly CDs are bit messy, I do wonder exactly what the main driver of those lines were, it might be something very valid my eyes are missing. Imo I like the overall arrangement of SEPA; it's not exactly like what I would do but it does some novel things I haven't seen before, such as split Chester into a northern and southern half. The choice of precincts taken from Montgomery by the Delaware CD isn't terrible, though it would probably be better if they were directly bordering each other.
Tbh, I would remove the part of Dauphin within the Cumberland CD, take instead from York, and then compensate the Lancaster CD with parts of Lebanon County.
I would personally clean things up in SW PA by removing the tri-cut of Westmoreland and give it Blair County and, if necessary, at least part of Bedford and Fulton counties. If even that isn't enough, go into Somerset.
Also possible: give the Washington County CD Cambria in return for some of Westmoreland.
The lines linking Westmoreland with Pittsburgh can stay, given this is a Dem leaning map.
I would say it is decent, if somewhat unconventional in some areas.

[President Punxsutawney Phil]

Decided to go out of the box with this one.
Philly is divided right down Broad Street, and all leftovers are given to Bucks. MontCo cedes 9k or so to the Berks CD. Chester is the only county split in two in a huge way. I quite like this arrangement.
Particular emphasis was placed on compactness and reducing county splits. The Harrisburg CD moves right by about 2 points. I sacrificed having an ideal SW PA seat for large, compact CDs in the entirety of the rural areas of the state. One unique result of my decisions was the creation of a Trump+4 (R+1.8) seat combining all of Greene County and some of Fayette Counties with Swissvale and Plum, following river boundaries. The map ends up being a competitivemander in practice.
https://davesredistricting.org/join/987a1720-4a32-4313-82b8-c0a0efa72b35

[President Punxsutawney Phil]

Did a variation of East Anglia Lefty's map that Tim Turner will definitely hate. If you can get over the MontCo and Bucks chops, I actually think it reflects SEPA COIs extremely well. Plus it has two AA VRA districts.

Thanks. I hate it.
This is definitely not a good map in my book. You seem to know me well.
For starters, the border of 5th and the 12th, splits two counties.
The 11th and the 5th has a weird border that could be improved.
The 12th splits counties with almost all of its neighbors.
The 4th is a terrible district even leaving aside the MontCo/Bucks split (which, by itself, is not enough to make a map terrible, but is enough to disqualify it from being good). Its shape is instrumental in having all but one of Chester, Montgomery, Bucks, and Delaware split, places such divergent areas as far NW Philly and far NE Bucks in the same district, and somehow despite all of that, 4 districts take from Philadelphia.
You can't even call this a good effort at a CoI map. It's a poor imitation.

[President Punxsutawney Phil]

Did a variation of East Anglia Lefty's map that Tim Turner will definitely hate. If you can get over the MontCo and Bucks chops, I actually think it reflects SEPA COIs extremely well. Plus it has two AA VRA districts.

Thanks. I hate it.
This is definitely not a good map in my book. You seem to know me well.
For starters, the border of 5th and the 12th, splits two counties.
The 11th and the 5th has a weird border that could be improved.
The 12th splits counties with almost all of its neighbors.
The 4th is a terrible district even leaving aside the MontCo/Bucks split (which, by itself, is not enough to make a map terrible, but is enough to disqualify it from being good). Its shape is instrumental in having all but one of Chester, Montgomery, Bucks, and Delaware split, places such divergent areas as far NW Philly and far NE Bucks in the same district, and somehow despite all of that, 4 districts take from Philadelphia.
You can't even call this a good effort at a CoI map. It's a poor imitation.

What do you think of my Fair Map?

The number of overall county splits is a bit too high for me to be totally comfortable, but it's not too shabby. The Philly CDs aren't too bad, I've never seen that arrangement before. It's a decent alternative to splitting Philly by Broad Street. If I ever did something along those lines (probably will) I'd take care to not split any special wards, and if I have to, will split the minimum number necessary.
Personally I'd give the 5th territory less to the west, but you might well have had worthy objectives in mind with the choice of territory, so I won't be too negative in that department.
The 4th baconstripping is somewhat problematic in a vacuum but if you are more aimed towards proportionality, then it's not something I'd grumble about.
Delaware is kept whole, which is good. 8 and 9 are logical. Can't say I quibble that much with the districts in Central PA. 14 is pitch-perfect. Districts in Allegheny are weird on paper but they follow river boundaries so I'm fine with them at the end of the day.
Overall I say it's a decent map that, like the map I just posted upthread, makes some unique choices.

[President Punxsutawney Phil]

Again, I want to emphasize how absurd not splitting Montgomery is. Actually look at a map of the Delaware-Montgomery county line. It randomly slashed through the center of Bryn Mawr and Ardmore, cutting through a continuous built up area stretching from the blue route to the Schuylkill. There is nothing about this area that is more connected to Lansdale than Delco/Chesco but the county line was never adjusted to reflect that. This area should 100% use the Schuylkill River as the dividing line, not the county line. It's just a bad COI otherwise.

What is the population of Montgomery minus the part south of the Schuylkill?