Pennsylvania 2018 Court Map (Updated)

So since the PA Supremes look like they might overturn the congressional map, I decided to draw what I think a truly nonpartisan Pennsylvania map might look like. This map tries to keep in mind county splits, compactness, and COI to the point where I think this is a reasonable map the Court could implement if they’re not trying to be totally Dem hacks. There are no municipality splits other than Philly and just 11 counties are split (Allegheny, Butler, Chester, Cumberland, Delaware, Luzerne, Mifflin, Monroe, Montgomery, Philadelphia, Synder). This map is overall 9R-4D-5S, though I would bet the final result if implemented for 2018 is closer to 9-9 than 14-4. Thanks to RRR and Ryan_in_SEPA for their help, particularly on the SEPA portion.

Statewide map:

1. Brady (D-Overbrook, Philadelphia) and Meehan (R-Upper Darby) D+30 in 2008, D+34 in 2012/16, 44/41/5 W/B/H VAP, Safe D

Dividing Philly North-South is actually a bit more logical from a compactness point of view than the current East/West split. Brady’s district gives up the Hispanic areas of eastern North Philly for Black West Philly, making this seat likely to have a Black-majority Dem electorate. It also includes the heavily black-middle-class eastern towns of Delaware County to equalize population, meaning Meehan technically now lives in this seat. This seat is actually one of the few that is getting much more white as many neighborhoods south and west of Center City are gentrifying rapidly, so there’s a real possibility a white liberal and Black coalition could dislodge Brady if his ethics issues become more salient.

2. Evans (D-Oak Lane, Philadelphia) D+37 in 2008, D+40 in 2012/16, 27/50/17 W/B/H VAP, Safe D

This seat is a pretty good illustration of just how massive North Philly’s ghettoes are, as this becomes an entirely North Philly district, getting some Hispanic-heavy neighborhoods and some lower-middle-class areas of the lower Northeast, plus a few Roxborough wealthy white liberals. This seat is is probably one of the 10 poorest districts in the country and still Black-majority by a hair (though the D primary electorate will be easily majority-Black). It is totally safe for Evans in both the primary and general.

3. Kelly (R-Butler) R+2 in 2008, R+7 in 2012/16, Safe R

I tried to make this seat a purely small-town NWPA seat, without any rural mountain areas or Pittsburgh suburbs. The result is reuniting Erie and dropping some Pittsburgh suburbs to make a very clean, very blue collar district. This probably would have been very competitive or even flipped as late as 2012, but it is very Trump-friendly and should be totally Safe for Kelly nowadays.

4. Perry (R-Carroll Twp.) R+13 in 2008, R+17 in 2012/16, Safe R

Not much changes with this district, except it loses Harrisburg and gets more Republicans from Franklin County. Totally Safe for Perry in the primary and general.

5. Thompson (R-Howard Twp.) R+9 in 2008, R+17 in 2012/16, Safe R

This seat is State College plus a big swath of some of the Northeast’s most rural territory. Loses its arm into Erie and is still totally Safe for Thompson in the primary and general.

6. OPEN D+1 in 2008, R+3 in 2012/16, Tossup

Berks County has been epically sliced and diced since 2002, but this map reunites it in one district. Berks is actually a right-trending area, but the cleanest thing to do was attach it to some upscale outer Montgomery suburbs that are trending hard-left. The net result is a beautifully clean, very diverse district socioeconomically that’s a pure Tossup politically and likely to remain so. Costello represents a plurality of this seat, but Meehan also represents a sizeable chunk of it, and neither one has a real base here. Given the tradeoff of moving to a new base vs. running in the tougher 7th (see below) this seat could be open, have one incumbent, or even see a Costello-Meehan primary mashup in which neither would have an obvious edge.

7. Costello (R-West Chester) D+3 in 2008, D+4 in 2012/16, Lean D pickup

Now to the part that Democrats will really like about this map. The current 7th is a ridiculous gerrymander spreading across a huge area of suburban SEPA.  This seat becomes a clean, very COI-friendly, pairing of the middle-ring suburban parts of Chester and Delaware counties, which is almost entirely upscale (except for the city of Chester) and trending left hard. Both Meehan and Costello’s bases are here, meaning that they could wind up facing off in a primary that wouldn’t do the winner any favors in the general. Alternatively, either or both could move to the 6th or take on Smucker in the 16th, meaning this seat could even wind up open. This is a seat with strong GOP heritage that has a decent chance to be held by a strong incumbent like Meehan or Costello IL-10 style, even in a tough environment. But all in all this seat is more likely than not to wind up in Dem hands.

8. Fitzpatrick (R-Middletown Twp.) D+1 in 2008, R+1 in 2012/16, Lean R

Bucks County is almost the perfect size for a district and has to stay whole, which means this district doesn’t change much at all. Getting a little bit of the Northeast tip of Philadelphia makes it cleaner than taking some of Montgomery, so that’s the only change. Still a very typical swing district with a broad cross-section of upscale and downscale suburbs that will give Fitzpatrick a tough fight in 2018, though with his incumbency I’d mark him a very slight favorite.

9. Shuster (R-Holidaysburg) R+11 in 2008, ~R+20 in 2012/16, Safe R

This seat covers almost all of rural SWPA outside of the Pittsburgh metro area, including Altoona and Johnstown. This is a historically-R seat, and the historically-D parts of the district are trending right hard. Shuster gets a significant amount of new territory, but should still be a primary favorite.

10. Marino (R-Williamsport) R+9 in 2008, R+16 in 2012/16, Safe R

This seat covers a bunch of small towns in NEPA that have been sliced and diced on the current map. This seat is trending right strongly, and basically a clean and compact COI; while Marino gets a big chunk of new territory he should still be Safe in the primary and general.

11. OPEN R+8 in 2008, R+10 in 2012/16, Safe R

The Harrisburg area is a mess on the current map and this seat makes a compact COI of basically the entire metro and little else. Most of the candidates who are running for the current 11th are from that side of the district, so I would expect them all to run here. This seat is red enough for the primary winner to likely have little trouble in the general barring something unexpected.

12. Rothfus (R-Edgeworth) R+9 in 2008, R+8 in 2012/16, Safe R

This seat basically unites almost all of Pittsburgh’s white-collar suburbs. Beaver County doesn’t belong from a socioeconomic point of view, but it makes the seat much more compact and isn’t a horrible COI. This seat has some tension between SWPA’s downscale heritage and Pittsburgh’s development as a white-collar center, but overall the seat is Republican up and down ballot. It’s a fair amount of new territory for Rothfus, but he should be Safe in the primary and general barring something unexpected or a truly mammoth wave.

13. Boyle (D-Bustleton, Philadelphia) D+9 in 2008, D+12 in 2012/16, Safe D

This seat combines the white-liberal heavy southern half of Montgomery County with the lower-middle-class central part of Northeast Philly. Boyle may actually live just over the line in the 8th, though he could easily move back here. With less of the Northeast and more white liberals in the district, Boyle could potentially be vulnerable to a primary challenge, though I would guess that his incumbency is enough to carry him through.

14. Doyle (D-Forest Hills) D+13 in 2008, D+16 in 2012/16, Safe D

This seat becomes significantly cleaner and gets unpacked a bit by following municipal lines. Still includes the entire city of Pittsburgh and its inner, generally blue collar, southern and eastern suburbs. This seat is actually trending left with the influx of white liberals. Should be totally safe for Doyle both primary and general.

15. OPEN D+3 in 2008, EVEN in 2012/16, Tossup

This seat reunites the Lehigh Valley into one district, which is almost the perfect size for a seat. It has a tiny bit of the Poconos to equalize population. This is a purely swing seat, with a diverse mix of urban and suburban, white and blue collar areas, that should be very competitive in 2018 almost regardless of who the parties nominate.

16. Smucker (R-West Lampeter Twp.) R+7 in 2008, R+9 in 2012/16, Safe R

This is a beautifully clean pairing of all of Lancaster and the exurban western parts of Chester County; basically the Amish Paradise seat. This seat is very historically-Republican, and though the Chester part of the district is trending left, Lancaster is pretty static to even trending right. Smucker should probably be safe here barring a massive wave. Though there is a chance Costello would challenge him in a primary, Smucker would still be a fairly strong favorite.

17. Cartwright (D-Moosic) D+4 in 2008, R+1 in 2012/16, Tossup

This seat takes in the rural areas of NEPA and part of the Poconos to pair with essentially the entire Scranton-Wilkes Barre metro. The net result is that Cartwright’s purple to light-blue seat becomes light red by adding some very historically-R areas. This is a decent pickup opportunity for the GOP even in 2018 with a decent candidate, though Cartwright’s incumbency and the national mood probably keep it in the Tossup category.

18. VACANT (probably Saccone, R-Elizabeth Twp.) R+9 in 2008, R+13 in 2012/16, Safe R

This seat doesn’t change a whole lot, but now includes all of Washington, Greene, and Westmoreland Counties, along with some small chunks of southern Allegheny to equalize population. This seat is even more blue-collar than the previous version and very Trump-friendly. Assuming the GOP doesn’t blow it in the special Saccone should be safe in this seat for the 2018 primary and general.

UPDATE: Alternate SEPA Version

This is an alternate version that makes some different (but essentially equally valid) choices in SEPA. I like this slightly less than the above version (mostly because I feel the Berks split is a bit awkward and PA-1 is something of a COI mashup) but some people have liked it better. Only the 7 SEPA seats (1, 2, 6, 7, 8, 13, 16) are affected.

1. Brady (D-Overbrook, Philadelphia) and Meehan (R-Upper Darby) D+23 in 2008, D+26 in 2012/16, 53/36/3 W/B/H VAP, Safe D

This seat is something of an awkward COI mashup of urban south and west Philly and most of DelCo, though for parochial interests it’s probably not a terrible fit. The population split is almost a perfect 50-50, but Philly casts a majority of the D primary vote. Brady could have some real trouble from a coalition of blacks and white liberals here. The seat is white-majority by VAP but white-plurality by total population.

2. Evans (D-Oak Lane, Philadelphia) D+39 in 2008, D+42 in 2012/16, 26/52/14 W/B/H VAP, Safe D

North Philly’s ghettoes, without the white liberals in Roxborough/Manayunk but some White Liberals in Center City as well. Majority-black but has some Hispanic-heavy neighborhoods and some lower-middle-class areas of the lower Northeast. This seat is is probably in contention for the title of the nation’s poorest. It is totally safe for Evans in both the primary and general.

6. OPEN D+4 in 2008, D+2 in 2012/16, Tossup

This seat has most (but not all) of the Reading metro area, along with some rural areas of Eastern Berks, plus the suburban northern 2/3 or so of Montgomery. In spite of the huge portion of upscale left-trending MontCo suburbs in this seat, it’s actually moving right as the blue-collar Reading area has been stampeding rightward enough to offset it. Neither Costello nor Meehan lives here, but both represent sizeable chunks of the seat and either could easily carpetbag. If one of them runs here, I would say their incumbency probably leaves them a very slight favorite as long as the year isn’t too horrible, but Democrats are probably more likely than not to pick this one up if it were an open seat. However, the Dem primary electorate here is very liable to nominate a bold progressive (hello, Manan) that could underperform in the Berks part of the seat.

7. Costello (R-West Chester) D+1 in 2008, D+1 in 2012/16, Tossup/Tilt R

All of Chester plus the northwestern third or so of Delaware. Unlike the equivalent Delaware/Chester mashup, this is not a seat that Democrats would get without one heck of a fight. Costello lives here and Meehan’s base is here, and this is a seat that either could easily hold, even in a bad year. Trump underperformed here, but not by as much as you might guess. This seat would probably be a pure Tossup, even in a less than ideal environment, with either Meehan or Costello so long as they don’t both run here and bang each other up in a primary.

8. Fitzpatrick (R-Middletown Twp.) D+1 in 2008, R+1 in 2012/16, Lean R

Bucks County is almost the perfect size for a district and has to stay whole, which means this district doesn’t change much at all. With this configuration it’s a bit cleaner to have the seat take some border townships of MontCo rather than NE Philly, but that doesn’t really make any difference politically. Due to incumbency I’d put Fitzpatrick as a slight favorite.

13. Boyle (D-Bustleton, Philadelphia) D+13 in 2008, D+16 in 2012/16, Safe D

This seat combines the white-liberal heavy southern half of Montgomery County and Roxborough/Manayunk with all of lower-middle-class, blue collar Northeast Philly. Unlike the map above, Boyle will have no trouble here primary or general as Northeast Philly makes up nearly 2/3 of the seat.

16. Smucker (R-West Lampeter Twp.) R+8 in 2008, R+11 in 2012/16, Safe R

This seat now becomes something close to a GOP vote sink by pairing Lancaster with most of the deeply Republican rural parts of Berks and some Reading suburbs. Totally safe for Smucker both primary and general.

Previous Post Next Post

73 Comments

  • Red Oaks November 18, 2017 at 2:41 pm

    Yeah, that’s a pretty clean map – only 11 counties broken and no double crossings. Thanks for doing this.


    MI-03: Tired of Presidency; Focused more on downballot races; Chris Afendoulis for State Senate

    • Jon November 18, 2017 at 6:49 pm

      Main way it could be cleaner are in both in the counties containing Philadelphia and Pittsburgh:
      Both have two districts crossing the county line instead of just one.
      It might be within OMOV tolerance to just split the county containing Philadelphia into three equally sized districts none of them crossing the county the line. As is; it looks like an attempt to give Philadelphia one more congressman than they are entitled to.
      There might be enough population in the section of Allegheny within #18 that it was given to #12 instead would eliminate the need for it to cross into one of the rural counties.


      45, M, MO-02

      • FreedomJim November 18, 2017 at 10:57 pm

        I do not think Philly has enough people for three full districts. The 13th could get the 8th’s portion of Philly, but that would split Montgomery three ways or another county two ways.
        Putting Fayette county in the 18th would be more compact, but would probably cause another county split in western PA.

        • krazen1211 November 20, 2017 at 2:33 pm

          Philly is barely 2 districts of population.

          In any case, we should work on a Philly hacks + Rs coalition in the PA House to override any veto.

          • Jon November 20, 2017 at 6:41 pm

            That seems to make this map of 1 full + 3 partial even worse than I thought if Philly is much closer to 2 ideal CDs than 3; or worse yet slightly less than 2.


            45, M, MO-02

  • Ryan_in_SEPA November 19, 2017 at 10:17 am

    We discussed a cleaner map, but it would have been more favorable to the Republicans. Keeping Chester County whole while splitting Delco with the western half Montgomery being attached to eastern Berks and Lancaster being attached to western Berks makes more sense from a COI perspective, but it would definitely be more favorable to the Republicans.


    31, PA-6, fiscally conservative communitarian, Giant Meteor 2020 - Just End It Already!

    • CTIronman November 19, 2017 at 1:24 pm

      I think an objective argument could be made that keeping Chester in PA01 maintains the ability of an AA candidate to win a 2nd seat in PA & removing it creates VRA dilution problems. This would create a domino where 13 picks up more of Philly & 7 needs more of Chester or a sliver of Montco

      • shamlet November 19, 2017 at 1:39 pm

        Unfortunately this PA-1 is only 1% less black than the current version. And trying to put Chester in at all cleanly actually makes the seat less black.


        R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • CTIronman November 22, 2017 at 3:09 pm

        Did some history. The city of Chester has been in PA01 since the 92 election; when a map favored by Democrats was adopted. So bisecting Delco has a long bipartisan history.

        Also wonder if the multi cuts to Montco are primarily to ensure high propensity yuppie primary voters can’t hassle Brendan Boyle; surely he’d rather leave Lower Merion harmlessly in a compact AA dominated PA02.

  • CTIronman November 19, 2017 at 2:44 pm

    Hard to believe there aren’t replacement precincts that can be pulled out of PA01 less heavily AA than the City of Chester

  • CTIronman November 19, 2017 at 3:38 pm

    Thought: if you split Upper Darby Twp & swap out 80% white Drexel Hill for 74% black Chester you probably move PA1 to AA/White parity w/o another county split

    • Ryan_in_SEPA November 19, 2017 at 5:30 pm

      I would argue that 2/3 of Delco should be attached to Philly as it has more COI with Philly than it does even with the rest of the county.


      31, PA-6, fiscally conservative communitarian, Giant Meteor 2020 - Just End It Already!

      • CTIronman November 20, 2017 at 11:45 am

        Ran the numbers: the part of Delco north of Rt 1 has about the same demographics & PVI as the areas south of Rt 1 not in the pre-11 PA01

  • CTIronman November 19, 2017 at 5:55 pm

    Do you have an alternative map that does that? Suspect that it probably wouldn’t sit well with Philly D’s

    • shamlet November 19, 2017 at 6:15 pm

      This was the alternate version: https://i.imgur.com/Fgh1XXZ.png

      My main issues with it is that this PA-1 feels to me like a weird urban/suburban mashup, and I also didn’t like the split of Berks, which draws the line right through the middle of the Reading metro area.


      R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • CTIronman November 19, 2017 at 6:37 pm

        Could you split Berks so it’s more north/south? Also, is Springfield in 1 in this map?

        • shamlet November 19, 2017 at 6:54 pm

          Springfield is in 1. The northern edge of PA-1 here is UD, Springfield, NP, Rose Valley, Brookhaven, Chester Twp, Upper Chichester. And the seat is only 32% BVAP.

          There’s no clean way to split Berks if you pair it with all of Lancaster, because you need 200K people from a county of 400K, and the immediate Reading metro is more than half the county. No matter what you do you have to split the city from half of its first-ring suburbs which I considered less than ideal.


          R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • CTIronman November 20, 2017 at 11:49 am

        Wonder if a East Berks/North Chesco/West Monto PA06 passes muster? Tack Reading & west Berks to PA16 in Lancaster & south Chesco/north-middle Delco to PA07. Maybe PAGOP should try & preemptive the suit by trying to pass a new cleaner map before the court hearing?

        • shamlet November 20, 2017 at 11:54 am

          The problem is that you once again wind up with an ugly Berks split. The portion of Berks that needs to go with a non-split Lancaster invariably necessitates an ugly split of the Reading metro area. Think about it this way: Berks is kind of three concentric COI circles. A 100K rural area enclosing a 200K suburban area enclosing the 100K city of Reading. There’s no way to split that cleanly 200/200.


          R, MD-7. Put not your trust in princes. Process is more important than outcome.

          • CTIronman November 20, 2017 at 12:35 pm

            Ahh it’s not THAT ugly; it’s actually easy to do fairly geometrically; you just have to put some towns abutting Reading in a non Reading CD

            • shamlet November 20, 2017 at 12:37 pm

              Yeah, but it’s a worse COI fit than splitting Chester.


              R, MD-7. Put not your trust in princes. Process is more important than outcome.

          • CTIronman November 20, 2017 at 12:39 pm

            The GOP can try a mildly unattractive incumbent protection scheme that unwinds the worst of the indefensible 2011 map or take their chances the PA Supremes hose them with a D wet dream. They might want to be proactive. Just my opinion

            • shamlet November 20, 2017 at 2:24 pm

              No argument there. This map was more from the standpoint of trying to be as truly non-partisan as I could. I might try a legislative fairish mildly R-biased map and see how it might differ.


              R, MD-7. Put not your trust in princes. Process is more important than outcome.

              • CTIronman November 20, 2017 at 3:18 pm

                Oh this map wasn’t awful but D’s are already flogging ones that would be

              • CTIronman November 23, 2017 at 12:15 pm

                Thoughts in that regard:
                a) Leave PA08 as is from current map
                b) pretty much draw this draft map for the non SEPA seats
                c) leave city of Chester in PA01 & Lower Merion in PA02
                d) PA13 is NE Philly/rest of S Montco w/clean lines
                e) PA01 as plurality VAP AA; PA02 as 50% VAP AA
                f) clean line a PA06/PA07/PA16 divving Chester/Berks/Lancaster/outer Delco & Montco

    • w920us November 19, 2017 at 6:23 pm

      Yeah, that really is one of those things working in our favor. I cant imagine Philly Dems willingly be second fiddle to suburban Democrats, in one, if any Congressional districts.


      R, South Philly, 47, Gay, WFU Alum
      #TrumpVoter #NeverHillary

  • CTIronman November 19, 2017 at 10:14 pm

    I’m not sure how a clean map gets 3 Philly based House seats; you need all of em to reach well into the burbs & R’s will claim VRA violation if PA02 drops under 50% AA VAP

  • Torie November 21, 2017 at 2:45 pm

    Below is a link to a couple of more PA maps. I don’t like the quad chop of Philly myself in particular. Both maps have less chop in the Philly metro area.

    https://uselectionatlas.org/FORUM/index.php?topic=276987.msg5906153#msg5906153


    65 - NY-19 (D)

    • Jon November 21, 2017 at 6:38 pm

      Overall, I think that version better than what’s posted.
      The big difference if I were doing this is every place where there’s two districts crossing the county line, I’d eliminate one of the partial CDs within that county; even at the cost of splitting a non-split county so long as a CD that didn’t split any counties at all wasn’t impacted.

      Pittsburgh area: I’m pretty sure I’d shift 14 SE but stay entirely within the county. Then I’d give 12 all territory in that county not given to 14. 18 then gets western county(ies) in 12.

      Philly area: While indeed drawing 1 & 2 that way; I’d next look at that county split three ways between 6, 7, and 16 to see if it’s big enough for a CD. In any case, there’d either be one or two CDs in that county instead of three and consequently a major difference in the SE PA portion of the map as I’d try to avoid any other county being split into having a 3rd CD; which goes hand in hand with me also trying to avoid things like #6 having multiple partial counties.


      45, M, MO-02

    • shamlet November 21, 2017 at 8:09 pm

      The eastern half of that map is nice… I’d consider that equally fair to mine, though I think the NEPA seats are a bit non-compact. But the western half of that map is a bit of a mess IMO. 3 and 5 in particular are terrible from a COI perspective.


      R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • Jon November 22, 2017 at 6:32 pm

        #3 in his is actually my favorite district in what he drew: It appears to be composed entirely of whole counties. Looks about as clean as possible given that some of the counties obviously have a river as part of their boundary and also some other strange edges for some unknown reason. ( Person surveying the lines messed up?)


        45, M, MO-02

  • fzw November 21, 2017 at 3:07 pm

    @shamlet:

    The updated DRA is missing 2012/2016 PVI’s for some very Republican precincts in Berks County, which I think is skewing your overall PVI for PA-6. That seat should definitely be a lot more R-leaning than R+1. Northern Montgomery County is pretty Republican as well.


    Currently MO-5. From MO-3.

    • shamlet November 21, 2017 at 3:12 pm

      I’m well aware the DRA PVIs aren’t working for PA. The way I calculated 2016 PVIs is very rough – I took the 08 PVI and then did a weighted swing by county. I considered that seat to have a 1-point rightward swing since Berks swung right by around 6 and Montgomery swung left by around 3, 08->16


      R, MD-7. Put not your trust in princes. Process is more important than outcome.

    • Torie November 22, 2017 at 8:59 am

      There are two different maps in the link. Are you guys just referring to the first one? That was drawn by Muon2. I drew the second map.


      65 - NY-19 (D)

  • CTIronman November 21, 2017 at 5:02 pm

    Query what the Toomey/McGinty & Shapiro/Rafferty splits were on these seats

  • Ryan_in_SEPA November 22, 2017 at 9:25 am

    One idea I have tried is treating the Philly region as quarter semi-circles radiating from Philly. It would split county lines a lot, but arguably would make more sense from a demographic perspective.

    I would do it as follows:

    PA-1 and PA-2 – Philadelphia County minus Northwest Philly

    PA-7 – DelCo plus Montco South of 276

    PA-8 – Lower Bucks and Middle Bucks plus eastern Montco.

    PA-6 – Chester County, far western and northwestern Delco, and Pottstown area of Montco

    PA-16 – Bulk of Lancaster plus Reading dominated part of Berks

    PA-15 – Eastern Berks, Lehigh, Northampton and Upper Bucks

    Safe D – 1,2,7

    Tilt R – 8

    Lean R – 6, 15

    Safe R – 16

    Part of the problem with the existing map is that SEPA is overrepresented as part of the state delegation (8 congresscritters) when it should have 7. If you pull the Lehigh Valley, Berks, and Lancaster out, the region should have 5.

    Once you give up that paradigm, it makes it a lot easier to draw a Republican clean map.


    31, PA-6, fiscally conservative communitarian, Giant Meteor 2020 - Just End It Already!

    • FreedomJim November 22, 2017 at 5:56 pm

      Did you leave out PA-13 with the rest of MontCo and Philly?

      • Ryan_in_SEPA November 24, 2017 at 10:25 am

        PA-13 would be moved elsewhere probably south central PA. The portion of NW Philly (primarily Chestnut Hill and Roxborough plus a little more), would be in PA-7.


        31, PA-6, fiscally conservative communitarian, Giant Meteor 2020 - Just End It Already!

        • FreedomJim November 25, 2017 at 6:32 pm

          Thanks for clarifying. Can PA-1 and 2 be VRA compliant with all of the whites in NE Philly and none of the blacks in DelCo?
          If you are willing to split Bucks, how about putting Upper Bucks in 6 with all of Berks and some of north MontCo?

  • shamlet November 23, 2017 at 6:05 pm

    I figured out how to get DRA to spit out the correct 2-election PVIs… you have to use the regular 2010 districts instead of the “updated by the state” ones.

    1) D+34
    2) D+40
    3) R+7
    4) R+17
    5) R+17
    6) R+3
    7) D+4
    8) R+1
    9) R+20
    10) R+16
    11) R+10
    12) R+8
    13) D+12
    14) D+16
    15) EVEN
    16) R+9
    17) R+1
    18) R+13

    So this is pretty much a 9R-4D-5S map.


    R, MD-7. Put not your trust in princes. Process is more important than outcome.

    • Ryan_in_SEPA November 24, 2017 at 10:26 am

      This really shows that to get a map where the Democrats break even in a good year you have to have a Democrat gerrymander.


      31, PA-6, fiscally conservative communitarian, Giant Meteor 2020 - Just End It Already!

    • shamlet November 24, 2017 at 10:36 am

      For the “alternate” SEPA version with no Chester split, it’s:
      1) D+26
      2) D+42
      6) D+3
      7) D+1
      13) D+16
      16) R+11
      And all others same.

      So I don’t think there’s much partisan impact between the two versions… It probably means 6 is Lean D and 7 is Tossup to Tilt R rather than 6 being Tossup to Lean R and 7 being Lean D.


      R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • CTIronman November 24, 2017 at 10:54 am

        Damn. Love to see a side by side of 2 maps

        Played with new concept last night: keep Berks whole & link with central Lancaster in 16.

        Put southern Lancaster & Chester together in 6

        Link mid/north Delco in 7 with north/west Montco

        • shamlet November 24, 2017 at 11:46 am

          Working on doing just that. I’ll try to get it sometime today.


          R, MD-7. Put not your trust in princes. Process is more important than outcome.

  • shamlet November 24, 2017 at 5:15 pm

    Alternate SEPA has been added.


    R, MD-7. Put not your trust in princes. Process is more important than outcome.

    • roguemapper November 24, 2017 at 5:30 pm

      Your alternate SEPA would perform better for Ds in my view, but your original SEPA is a much neater division from a COI standpoint. I’ve been following this thread but, to be sure, is the SEPA debate about how many districts go into Philly? If so, then my personal view is that I don’t see a problem with it if it makes more sense from a COI standpoint. Not to mention our Republicans here in NC have been telling us for years that cities are so important they need to be divided among multiple districts even when they don’t have to and regardless of whether it makes sense from a COI standpoint. 😉 In any case, I prefer your first map, though I think it would be somewhat worse for Ds.


      Dem NC-11

      • shamlet November 24, 2017 at 5:45 pm

        I don’t think either configuration is intrinsically better for Dems – I think if 2018 is a mediocre year for them they’d lose both 6 and 7 on the second but still win 7 on the first. On the other hand if 2018 is a moderate wave they’d probably win 6 and 7 on the second but only 7 on the first. But if 2018 is a huge wave they’d win 6 and 7 on both – but have a small chance at 16 on the first and no chance at 16 on the second.

        The Philly thing isn’t a meaningful debate – whether you split it 3 or 4 ways is basically immaterial because 90% of the 8th is in Bucks and that part of Philly is basically the same as the MontCo areas near Bucks (which is to say, EVENish). I just did it differently based on what made sense from a compactness standpoint. The real point of contention between the two maps was the split of Berks vs. the split of Chester (and as a secondary point, which MontCo/DelCo split makes more sense)


        R, MD-7. Put not your trust in princes. Process is more important than outcome.

        • roguemapper November 24, 2017 at 5:58 pm

          Ah, OK. Well if the issue is how you split Berks, Chester, DelCo, and MontCo then I think your original map is greatly superior to the alternative map. To me that’s self-evident. With regard to electoral performance, PA-06 is better for Ds on the second map while PA-07 is better for Ds on the first map. I’m not just thinking in terms of 2018. However much Reading is “stampeding” to the right I think that it would easily be outweighed in the long run by both population growth and electoral shifts in alternative PA-06 MontCo. Meanwhile, Ds would get a slightly less winnable alternative PA-07 but a district that seems poised for a blueshift given current demographic trends. But, again, I think the initial SEPA is far preferable in terms of compactness and COI.


          Dem NC-11

          • shamlet November 24, 2017 at 6:05 pm

            Well, you’re only looking at a 2-election timeframe so long-term trends really aren’t an issue…


            R, MD-7. Put not your trust in princes. Process is more important than outcome.

            • roguemapper November 24, 2017 at 6:10 pm

              Not necessarily. If the next redistricting cycle is split control then it’s (a) more likely for a compromise map to retain the framework of a court map; (b) far more likely for a court map to retain the framework of a court map if it ends up back in the courts regardless.


              Dem NC-11

        • CTIronman November 24, 2017 at 10:37 pm

          What about leaving Berks whole, Chesco whole, & splitting Lancaster? Seems it yield a less vote sinky PA16 & a slightly safer red PA06

          • shamlet November 25, 2017 at 10:36 am

            Ironically this would actually be close to a best-case scenario for Dems. I just gamed it out and the Chester + Southern Lancaster seat is EVEN (the city of Lancaster is deep-blue and most of the deep-red territory is in the northern part of the county… and it double-bunks Smucker and Costello to boot) and the Berks + Northern Lancaster seat is R+10.

            Those two seats are not bad from a COI standpoint (Lancaster’s core metro winds up cleanly in the Chester seat). But the problem with that is that you then wind up with awkwardness in Delaware and Montgomery. Berks+Chester+Lancaster is basically the perfect size for two seats (17K over). So to avoid an unnecessary county split in the Philly/Delaware/Montgomery triad you have to keep DelCo whole *and* make a ~D+7 MontCo-only seat. Which kinda wreaks havoc on the Philly seats – Boyle’s seat becomes an all-Philly Navy Yard to far Northeast seat and Brady gets the DelCo seat, which gets a big chunk of the West Philly ghettoes – enough to make it really awkward from a COI point of view but only enough to bring it up to ~30% black.

            Or you can make a Delaware+Montgomery seat without breaching Philly, but that winds up forcing a recreation of the hideous lower Northeast to northern tip of MontCo 2002 PA-13. And you have the same problem of blanching PA-1 because you can’t get the black populations from the Darbies and the other DelCo border towns.

            So that’s actually probably the best way to configure SEPA as a cleanish Dem gerrymander that guarantees a 1-seat pickup and makes the second no worse than a Tossup.


            R, MD-7. Put not your trust in princes. Process is more important than outcome.

            • CTIronman November 25, 2017 at 12:54 pm

              Actually I had Lancaster city in the Berks/Lancaster seat & ran the Chesco/Lancaster seat to the Susquehanna. The Berks seat is about +6 & the Chesco seat is about +4.

              The easternmost Chesco Twp (unspellable Welsh name w/T) is part of a north/mid Delco/west Montco D +2 PA7 which resembles the pre gerrymandered ’00 iteration.

              NE Montco fills out PA8. PA 13 stays NE Philly/South Montco

              You jockey around precincts to get the AA % in PA01 & PA02 you want/need

              Basically GOP strategy IMO should be to keep the city of Lancaster & Reading in the same CD & the AA areas of Delco in a PA01

              • shamlet November 25, 2017 at 2:39 pm

                That could work but there’s no valid reason for it other than GOP advantage, especially since you have an unnecessary county split in the Philly/Delaware/Montgomery triad.

                Putting Lancaster City in the Chester district is far cleaner and how a court would do it if they split Lancaster.


                R, MD-7. Put not your trust in princes. Process is more important than outcome.

                • CTIronman November 25, 2017 at 2:47 pm

                  Some of my thinking is based on thinking the PAGOP would be better off trying to undo the current map themselves & even if Wolf vetoed it use the legislatively passed map as a good faith proposal for the PA Supremes

                  • CTIronman November 25, 2017 at 2:52 pm

                    Also the “unnecessary split” in Phil/Del/Montco is sorta inevitable unless you concede that PA01 has its AA% diluted in a court map or PA02 becomes a “coalition” district. Also, unless PA01 or PA02 go into the burbs Boyle faces a likely suburban primary challenger in PA13

                    • shamlet November 25, 2017 at 5:53 pm

                      No doubt, but once you start getting cute by splitting more counties than necessary you just invite the Supremes to tweak it into a Dem gerrymander, which is very possible with that configuration.

                      Re: the PA-1 issue, I don’t think there’s any way to keep the benchmark BVAP without either pushing PA-2 below 50% BVAP or making it utterly ugly. But as I’ve said before the PA-1 on the first map is only a point less black than the current one. I don’t think any court would consider that a meaningful distinction, especially since blacks will still be a clear D primary majority.


                      R, MD-7. Put not your trust in princes. Process is more important than outcome.

                    • CTIronman November 25, 2017 at 6:27 pm

                      If you’re willing to split UD township (keep 80%+ NHW Drexel Hill in PA07) you can probably bring the AA% in PA01 back to par in map no 1

    • roguemapper November 24, 2017 at 5:45 pm

      PS. The only thing I’d want to change about your original SEPA is to swap Germantown and Mayfair between PA-02 and PA-13. I assume the reason for the division you chose is to push PA-02 above 50% AA VAP. The court may very well insist on that, but as you well know I don’t find the arbitrary 50% line compelling in and of itself.


      Dem NC-11

      • shamlet November 24, 2017 at 5:54 pm

        Germantown? You mean Lawncrest?


        R, MD-7. Put not your trust in princes. Process is more important than outcome.

        • roguemapper November 24, 2017 at 6:02 pm

          Oh, that’s right. Lawncrest. I was thinking one thing and picturing another when I wrote that. 🙂


          Dem NC-11

  • Torie November 28, 2017 at 8:41 am

    Under the law, there is no requirement that a minority CD by 50% BVAP. The 50% figure (f such percentage can be reached with a “compact” CD), triggers the requirement that a minority CD be drawn, but it can be less than 50% BVAP as long as it is “performing.” 45% BVAP should be sufficient. The issue is whether, with a racially polarized voting pattern, the electorate which votes in the Dem primary is 50% BVAP.


    65 - NY-19 (D)

    • Jon November 28, 2017 at 6:54 pm

      I don’t read the SCOTUS decisions quite that way.
      SCOTUS had out right said during the 2000s redistricting that a district can’t qualify as a section 2 district if it’s less than 50% VAP of a single cohesive minority group. (There is some wiggle room in that the door is open as to only needing 50.0% CVAP if reliable CVAP figures can be found. Conversely, findings might be made that in particular instances a higher number is needed such as difference in turnout rates or registration.)

      However, the recent ones the past couple years have brought back into play the third prong that there must also be sufficient polarization for a section 2 district to be required in the first place. Basically either the PA legislature is fine to draw the Philly congressional districts however they want within reason (if there isn’t polarization *) or else there must be be majority (C)VAP district(s). The normal standard used for determining polarization is what happens within the D primary; particularly downstream seats when open as that happens more frequently than the CDs themselves. The only place a specific number lower than that might come into play is a group saying they’ll sue if the district doesn’t meet their criteria, but only because if no one cares enough to sue the district will stand.

      * Straw man example of not within reason would be intentionally cracking the AA community of Philly roughly in half between two districts.


      45, M, MO-02

  • Torie November 29, 2017 at 10:21 am

    Here is another effort by me of eastern PA. The map keeps all townships whole within a +/- 0.5% population deviance from precise population equality. There is a chop of a ward in Philly, but hey, the Philly chop looks pretty with that beautiful right angle. 🙂

    http://ibb.co/j0tOCw


    65 - NY-19 (D)

    • shamlet November 29, 2017 at 10:27 am

      Looks pretty good overall except for the way you drew Philly. I’d keep the 7th from crossing the Schuylkill, which also makes it more black, and bring the 1st down to the Navy Yard instead.


      R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • Torie November 29, 2017 at 5:37 pm

        Thank you for the suggestion. And the ward chop in Philly goes away as a most sumptuous lagniappe to boot. 🙂 PA-02 is 48.4% BVAP per the 2010 census. That is plenty to pass legal muster.

        http://ibb.co/kT6Bhw


        65 - NY-19 (D)

        • shamlet November 29, 2017 at 5:42 pm

          Yeah, that’s much better and a very solid map. If you want to put PA-2 up to 50% you can do it fairly cleanly by giving the southeast part of the city to PA-1 and moving the dividing line in North Philly east to (from N->S) Tacony Creek -> Roosevelt Blvd -> 5th Street, but this is probably fine with the new jurisprudence.


          R, MD-7. Put not your trust in princes. Process is more important than outcome.

          • Torie November 30, 2017 at 12:08 pm

            Thank you. Below is the whole state. I was able to clean up the lines some in west PA without an additional chop, making PA-10 look less emaciated at its waist in particular. PA-06 as I drew it I think went for Toomey by around 5 points, or 4 points better than his statewide margin. If one considers the Toomey figures to be a better read of PVI in PA than the Trump figures, than PA-06 has a GOP PVI of about 2 points. It’s slowly trending Dem though bit by bit.

            http://ibb.co/kiqmjb


            65 - NY-19 (D)

            • FreedomJim November 30, 2017 at 7:16 pm

              I think two incumbents live in your fifth and none live in your ninth. Where would you expect them to run?

            • shamlet November 30, 2017 at 7:21 pm

              I don’t love the 5/9/18 configuration on that map from a COI perspective (they throw together some culturally dissimilar areas) but it’s not a terrible choice. Overall very nice.


              R, MD-7. Put not your trust in princes. Process is more important than outcome.

              • FreedomJim November 30, 2017 at 7:50 pm

                I agree that is a good effort by Torie. I wish some of the western districts like 3 and 18 were more compact, but I am not sure how to do that. Perhaps replace the easternmost parts of 3 and 18 with more Venango (3) and Westmoreland (18). Then divide the rest of central PA between 5 in the north and 9 in the south. The 9th would be very Republican, but I think the rest of the western districts would still be pretty safe.

  • Torie December 9, 2017 at 7:41 am

    In the unlikely event anyone might be interested, below is a link that explains in some detail part of the algorithm that I try to follow when drawing CD maps. The rules for calculating erosity are not discussed. That involves a count of how many state highways running between county seats cross CD lines within the state (or within macro-chopped counties, any paved road between municipalities both of which are located in the county). If a municipality is macro-chopped, then sub-jurisdictions therein for this purpose function as municipalities (in Philly that would be its wards).

    https://uselectionatlas.org/FORUM/index.php?topic=276987.msg5949937#msg5949937


    65 - NY-19 (D)

  • Leave a Reply

    Do NOT follow this link or you will be banned from the site!