Draw It Yourself: What Would A Fair Pennsylvania Congressional Map Look Like?

As I’m sure you know, gerrymandering is a noxious threat to democracy. It violates the basic premise of separation of powers — it empowers legislators to draw their own districts and control their own elections, effectively entrenching them despite public disapproval of them at large. The executive and judicial checks on this power are often neutered, like when the governor is of the same party as the legislature or when the US Supreme Court refuses to make partisan gerrymandering a justiciable issue. With the pending Gill v. Whitford case, however, the practice may finally be limited and a semblance of sanity will be returned to drawing districts.

But some people aren’t waiting on the Supreme Court. In Pennsylvania a group of voters is suing to overturn the congressional map under state law, using Gill as an informative precedent. The state law case makes sense because Pennsylvania’s Supreme Court has a 5–2 Democratic majority, and they don’t need to wait for the Supreme Court to weigh in on whether their proposed efficiency gap measure is applicable under federal law. Further, the Wisconsin Republican Party’s expert witness Sean Trende even admitted in Gill that Pennsylvania had the “gerrymander of the decade.” To wit, in 2012 Republican and Democratic congressional candidates got nearly identical numbers of votes, suggesting a 9–9 delegation, or maybe 10–8 Democratic as Team Blue did slightly better. But rather the delegation was 13–5 Republican. How could that happen? Well, this is what Pennsylvania’s congressional districts look like:

Existing PA congressional map

Ooh boy, that map fell out of the ugly tree and hit every branch on the way down. The 7th is particularly egregious. Quite possibly the most hideous congressional district in the nation, it has been described as “Goofy kicking Donald Duck” and at points is no wider than a seafood restaurant. It snakes through the Philadelphia suburbs trying to pick up Republican areas without making the adjoining PA06 and PA16 too Democratic. And the gerrymander has worked — Democrats have never held more than 5 seats under this map, even as Pennsylvania has remained a purple state for the decade.

But the Pennsylvania Supreme Court could change that. In fact, it looks like they will change that, with four justices ruling that lower courts need to hold expedited hearings so new lines can be in place for the 2018 elections if necessary. The PA Supreme Court could sand down those rough edges and give Pennsylvania something they haven’t had in decades — districts with a passing resemblance to polygons and which keep communities of interest together, minimizing the crossing of political boundaries.

Which begs the question, what would a fair map look like? I tried making a fair map based on these principles

1. Equal populations. The most obvious legal requirement, each district is within 50 people of the ideal. The districts all need to be within 1 of ideal in real life, but the software that I use to draw districts (the invaluable Dave’s Redistricting App, which allows citizens to perform tasks legislators normally hoard to themselves) isn’t powerful enough to split precincts. So, use your imagination as the edges of these districts will get sanded down a little bit, like the border of PA04 and PA11
2. Obey the VRA. The Voting Rights Act requires minorities have a voice in electing their representatives in the two Philadelphia districts. In practice, this means the small parts of the city going to adjoining districts should be from the whitest wards and that one district should try to keep the growing Latino community together.
3. Keep communities of interest together. Keep cities with cities, suburbs with suburbs, rural with rural and the like. Pennsylvania has a lot of geographically distinct regions. The current congressional map ignores them. This map respects them (or at least tries to).
4. Minimize breaking political boundaries. Pennsylvania county lines may have been drawn by a whiskey-soaked surveyor 300 years ago, but they do correspond to discrete political entities and often communities of interest. So I tried to have district lines break through counties (instead of adhering to county lines for their boundaries) as few times as possible. For example, Dauphin County had 268,100 people in 2010, which is less than the ideal congressional district size of 705,688. But Montgomery County had 799,874 people, which means at best it’d be in 2 congressional districts. Accordingly, Dauphin would ideally be in 1 district while Montgomery would be in 2.

County Breaks measure how many districts each county is “broken” into. Pennsylvania has 18 congressional districts and you could draw 4 congressional districts entirely within one county, requiring at least 14 counties to break across more districts than necessary. Accordingly, you’d expect a fair map to have around 14 County Breaks. The map I provided has 13 Breaks (combining “leftovers” of large counties into one district can get you below the ideal baseline) while the existing one has 35. For comparison’s sake, in my map Dauphin is in 1 district and Montgomery is in 2 (zero Breaks) while in the existing map Dauphin is in 3 districts and Montgomery is in 5 (5 Breaks). Now, I’m not saying we have to be prisoners of geography and draw lines in strict accordance with county borders (and some gerrymanders like Michigan’s congressional map actually minimize Breaks too). But if a legislative map has nearly 3 times as many County Breaks as ideal it’s a pretty obvious gerrymander.

Guided by those principles, here’s what a new, fair, Pennsylvania congressional map would look like:

Proposed Fair Map

As for the drawing of the districts, I started from the cities and worked my way into rural areas. The City of Philadelphia is large enough to have two congressional districts entirely inside of it, so this map reflects that (PA-01 and PA-02). Bucks County to its north needs about 80,000 additional people for its congressional district, so it takes them from that remainder in Philadelphia (PA-08). Adjacent Montgomery County is big enough to support one congressional district entirely within its borders (PA-13). The remainders of Philadelphia and Montgomery County join in with Delaware County and a sliver of easternmost Chester County to create a district centered on the Main Line (PA-07). The rest of Chester is appended to the parts of Berks closest to Philadelphia (PA-06), while Lancaster and the remainder of Berks create a Philly exurbs district (PA-16).

To the north of these districts is the Lehigh Valley (Northampton, Lehigh and Carbon Counties). This area is almost perfectly sized for its own congressional district (PA-15), though these days Monroe County probably has more in common with the rest of the relatively prosperous valley than the old coal towns in Carbon. But Monroe is instead placed with most of Luzerne and Lackawanna Counties in a Poconos district (PA-17).

Shifting south, fast-growing York, Adams and Franklin Counties are also nearly ideally sized for a district (PA-04), as is the greater Harrisburg area centered on Dauphin County to the north (PA-11). This map tries to minimize County Breaks, but you could argue that these districts should swap parts of York closer to Harrisburg for the western fringes of Perry and Cumberland Counties.

Hopping to the other end of the state, one district takes in the city of Pittsburgh and its most densely populated suburbs to the south (PA-14) while the remainder of Allegheny County is combined with the closest parts of the most populous adjacent county — Westmoreland (PA-12) and the Pittsburgh suburbs in Butler County that don’t fit into the Northwest Pennsylvania district (PA-03). The Pittsburgh suburbs to the west and south of the city then get put into another district (PA-18).

After that only rural areas remain. The southernmost ones are placed in one district (PA-09), while the most sparsely populated ones go into a massive district bigger than Maryland and all but a few other seats east of the Mississippi (PA-05). That leaves rural Coal Country in its own seat, plus a few Central Pennsylvania cities like State College and Williamsport (PA-10).

Anyway, how would 2018 shake out under these districts? While I want these maps to be nonpartisan, we can’t pretend that they won’t have some partisan impact. So here’s a preview:

1st District (Bob Brady): Safe Democratic

Brady goes from representing a 46% AA district to representing a 56% AA one under this map. In the past the longtime boss of the Philadelphia Democratic Party could win a primary anywhere with a Wawa but his power has been on the wane and could be an attractive target for an up and coming African-American politician like Omar Woodard. Whoever emerged from that primary would represent West Philadelphia in Congress (Will Smith actually grew up pretty close to the county line).

2nd District (Dwight Evans): Safe Democratic

The 2nd is essentially Philadelphia east of Broad and north of Center City except for the Far Northeast and is the most racially diverse district in the state (36% White, 34% AA, 22% Latino, 7% Asian). Evans is used to representing a majority African-American district but he’s well-known to the city at large and shouldn’t have much of an issue winning election here.

3rd District (No Incumbent): Safe Republican

If this district had been in place for the 2012 elections it’d be considered a swing district. It had voted for Obama 50–48, making it slightly more Republican than the nation as a whole but its (slightly more Republican) predecessor was represented by Democrat Kathy Dahlkemper for one term. Last year, however, it voted for Trump by 20 points. Its current representative, Mike Kelly, lives just over the border from this district in the 12th but that shouldn’t stop him from running for re-election here (you only have to live in the same state as your district to represent it, and at least 23 congressmen live outside their districts).

4th District (Scott Perry): Safe Republican

Perry would face no issue winning re-election in an area that hasn’t sent a Democrat to Congress in more than 50 years.

5th District (Tom Marino): Safe Republican

Close to ⅔ of the voters in the current PA-05 are located in its new iteration, but GT Thompson, its incumbent congressman, isn’t one of them. Most likely he’d flip districts with Marino, who faces a similar predicament. Marino announced he was running for re-election after his Drug Czar nomination was torpedoed because he actually made the opioid crisis worse.

6th District (Ryan Costello): Tossup

In its new iteration the 6th voted for Clinton by 8 points, up from a narrow half a percentage point. Air Force veteran and former And1 COO Chrissy Houlahan is already running against Costello. No word on if she’s got Hot Sauce and Skip My Lou canvassing for her. This district would favor a Democrat in an open seat but this area has a habit of voting Republican downballot even in Democratic waves, so I’d rate it a tossup for now.

7th District (Pat Meehan): Likely Democratic

Meehan has been the subject of retirement rumors but has said that he’s still running for re-election. If a court gives him a district that voted for Clinton by more than 20 points, that’d change his mind. You can see why Republicans opted for Goofy kicking Donald Duck instead of a fair, compact district. Democrats would be immediately favored in PA07, though it’d be Leans Democratic if Meehan (who has a sizeable war chest) digs in and fights.

8th District (Brian Fitzpatrick and Brendan Boyle): Tossup

Brendan Boyle would face a conundrum under this map: run in his old seat, which no longer contains any of his Philadelphia base (and where he would likely face a Montgomery County-based challenger like Leslie Richards) or run where he lives, a more marginal seat against another incumbent but where he could surely count on party backing. If he runs in the 8th against freshman Fitzpatrick it’s a Tossup, but if he opts for the 13th this district switches to Leans Republican. The non-Bucks portion of the new 8th is slightly less Republican than in the old 8th so it’d flip from backing Trump by about 1,000 votes to backing Clinton by a similar margin.

9th District (Bill Shuster): Safe Republican

Shuster hasn’t had to face a contested race in his life — his father rigged it so he’d get the Republican nomination after he retired and he’s been in a safe seat since. He, Dan Lipinski, and other failson recipients of such sinecures are proof that America abolished aristocracy in name only. Sadly, a fair Pennsylvania map does not change that.

10th District (GT Thompson): Safe Republican

As noted above, Thompson gets a new but safe seat. The real issue would come in 2022, as Pennsylvania is certain to lose a seat and a slow-growing, centrally located one bordering 7 other districts would be easy to cannibalize.

11th District (No incumbent): Likely Republican

If Ryan Costello really is worried about his chances in the new 6th he could move to this now open seat, which contains a small number of his current constituents (something that Frank Pecora couldn’t say). Former Democratic Rep. Tim Holden, however, represented a district similar to this one and his moderate voting record helped him survive the 2010 Republican wave. He’s only 60 and lives just a few blocks outside the district so he could be coaxed into running here again. However, this district still voted for Trump by more than 10 points and could attract all sorts of ambitious Republicans. An open seat is still a prime target in a wave year, and if Holden runs the 11th leans Republican, but it’s an uphill climb for any Democrat. State Rep. Steve Bloom, already running for the currently constituted 11th, would be the likely Republican nominee.

12th District (Keith Rothfus and Mike Kelly): Likely Republican

The 12th takes in upscale suburbs that have trended Democratic and blue collar towns that have trended Republican. So, somewhat remarkably, it’s politically stable as a whole. It went from supporting John McCain by 8 points to supporting Trump by a similar margin. The shuffling around of Pittsburgh-area district lines makes the 12th a far more inviting target than the currently open 18th for Democrats, so perhaps Conor Lamb, Mike Crossey or Pam Iovino would run here instead.

13th District (no incumbent): Safe Democratic

Clinton won this district by around 20% so even an open seat would be difficult for Republicans to win. As noted above, the biggest threat to Brendan Boyle, if he ran here, would be from within his own party.

14th District (Mike Doyle): Safe Democratic

President Trump’s “Pittsburgh, not Paris” refrain after pulling out of the climate accords rang hollow because Pittsburgh supported his opponent by giving her nearly 80% of the vote. While the rest of western Pennsylvania trended against Hillary Clinton she held her own in Allegheny County, improving on Obama’s 2012 performance. The 14th takes in its most urban areas and remains safe territory for any Democrat.

15th District (No Incumbent): Tossup

The Lehigh Valley, traditionally defined as Lehigh, Northampton and Carbon Counties, is almost the perfect size for a congressional district. Naturally, Republican line-drawers ignored this in 2011 because it would’ve created a seat that voted for Obama 55–43 in 2008. But political winds have shifted since then and Trump carried this district 49–47. Incumbent Charlie Dent announced his retirement so a swing seat like this would start out as a Tossup. But if Democrats can recruit a strong candidate here (like former state party chair TJ Rooney) it would Lean Democratic.

16th District (Lloyd Smucker): Safe Republican

Smucker faced a spirited challenged in 2016 from Democrat Christine Hartman, but she still lost 53–43. Under this map his district would lose its most Democratic sections for a portion of Berks County that swung hard against Democrats. So he’d be safe under these lines. Of note, the western edge of Berks, about 65 people, would be placed in the 11th district to equalize populations (as noted above, I can’t split precincts). Apologies to anyone living west of Pine Grove Road in Bethel Township if I got your hopes up.

17th District (Matt Cartwright): Likely Democratic

If you want to know why Hillary Clinton lost the 2016 election look to this district, which Obama would’ve won 56–42 but she lost 52–45. Cartwright made it to Congress by challenging a conservative Democrat from the left, so it’s unclear if he could win a contested general election. Yet this district actually becomes slightly more Democratic than its current iteration, which was intended as a Democratic vote sink. Considering the district is currently rated Likely Democrat I don’t see any reason to change that under these lines. Notably Republican Dan Meuser, a former PA Revenue Secretary running for the currently constituted 11th, lives here.

18th District (no incumbent): Likely Republican

Noted pro-life abortion proponent Tim Murphy represented this seat, and regardless of what happens with the Pennsylvania state supreme court the special election for the remainder of his term will take place under the old lines. These new lines remove the Allegheny County portions of his district, replacing them with more suburban/rural areas that resemble neighboring West Virginia — safely Democratic in the ’90s but heavily Republican in presidential races today. But like West Virginia pro-life and pro-gun Democrats can win here locally. Most of Democrats’ current PA-18 candidates live outside this new version of the district but conservative Westmoreland County Commissioner Gina Cerilli does not. If she ran here against GOP PA-18 nominee Rick Saccone (who lives over the border in the 14th but could still run here) Saccone would likely win but Cerilli could beat him.

Right now Pennsylvania has seven safely Republican seats, three that could be competitive under the right circumstances (PA-15, PA-16 and PA-18), three that will be hotly contested next year but tilt Republican (PA-06, PA-07 and PA-08), one Democratic seat where we have no idea what’s going on (PA-17) and four safe Democratic seats. Under this map Republicans would have six safe seats with three that are competitive under the right circumstances, three Tossups, one likely Democratic pickup, one likely Democratic hold and four safe Democratic seats. So a fair map shifts the state toward the Democrats, but not overwhelmingly so. In fact, if Trump can somehow recover his popularity then the state could still have a 13–5 congressional delegation, with the parties trading a seat in Scranton for a seat on the Main Line.

But that assumes that these lines are actually what the Pennsylvania Supreme Court would and should implement. So Pennsylvanians, help me out. Does this map look good to you? What would you change about it? The comments are open below. My only request is that if you tell me that I should move a district line, tell me where I should draw it instead. After all, it has to go somewhere.