Draw the Lines Brief to the State Supreme Court

On February 7, Commonwealth Court Judge Patricia McCullough recommended to the Pennsylvania Supreme Court that they adopt the map chosen by PA House Republicans (HB 2146). Draw the Lines, with legal standing as an amicus participant in this case, finds her selection misguided for several reasons, and hopes that the State Supreme Court consider the PA Citizens' Map for adoption.

You can read our full brief here. The following summarizes our arguments.

HB 2146 does not represent “the will of the People”

Judge McCullough argues the map is “functionally tantamount to the voice and will of the People” (214) because it passed the General Assembly and originated from a map drawn by a single citizen. However, HB 2146 never became law, as it was vetoed by Governor Wolf because it failed to meet his previously publicized redistricting principles. To say that map, vetoed by the governor, represents the “will of the People” discounts the voices of the 57.8% of Pennsylvania voters who cast their ballot for the governor in 2018.

HB 2146 passed the General Assembly entirely on party lines, as Democratic members of the House State Government Committee claimed leading up the vote that they were not consulted about the selection of the map before it was introduced on December 8th, 2021 prior to a December 9th hearing, nor the last-minute amendments to it before a committee voting meeting on December 15th.

Rather than taking as fact that the General Assembly represents the “will of the People,” the Court should consider that a non-partisan, public mapping initiative such as Draw the Lines, with 7,200 participants representing Republicans, Democrats, and independents, living in communities across the Commonwealth, is just as representative, if not more than, the political professionals whose self-interest is heavily dependent on the results of PA’s redistricting processes.

Flawed analysis of maps

In her selection of the map in HB 2146 from PA’s Republican legislators, Judge McCullough uses a flawed process of elimination to determine the optimal map for Pennsylvania.

Before even considering the relative constitutionality of all submitted plans (beginning on page 206), she strikes a number of maps for extra-judicial reasoning. This includes:

A map that optimizes partisan fairness (the Gressman Plan); Maps where the party affiliation of incumbents that are paired in the same district is unequal (both Senate Democratic Caucus plans, the Carter Plan); Maps that split Pittsburgh or Bucks County, places that appear to be arbitrarily chosen by the Court for consideration (the Ali plan; Governor Wolf’s Plan, Draw the Lines Plan, both Senate Democratic Caucus plans).

Draw the Lines argues that the maps should first be compared on the constitutional merits, including equal population, contiguity, compactness, and limiting jurisdictional splits. They must also be in compliance with the Voting Rights Act. And in responding to the Court’s decision in League of Women Voters, which held that “the 2011 Plan was a partisan gerrymander and… It was designed to dilute the votes of those who in prior elections voted for the party not in power in order to give the party in power a lasting electoral advantage,” and thus violated Article 1, Section 5 of the Pennsylvania Constitution, the (“free and equal elections clause”), any map should attempt minimize partisan bias as much as possible.

In comparison to the Draw the Lines map below, HB 2146 lags on every one of those metrics.

To elevate HB 2146 above the Citizens’ Map, or other maps that may score higher on the constitutionally required metrics, subverts League of Women Voters and the PA Constitution.

Inaccurate interpretations of partisan bias

Several of Judge McCullough’s cited facts around the partisan bias of HB 2146 are either cherry-picked or inaccurate.

She claims that “Given the credible evidence of record, HB 2146 is predicted to result in 9 Democratic-leaning seats and 8 Republican-leaning seats and, consequently, is more favorable to Democrats than the most likely outcome of 50,000 computer drawn simulated maps that used no partisan data, which resulted in 8 Democratic-leaning seats and 9 Republican-leaning seats. Unlike other maps that leaned Democrat, here, it is the Republican majority in the General Assembly that developed and proposed a plan, HB 2146, that favors Democrats, which ultimately underscores the partisan fairness of the plan,” (211).

First, that statement reflects an incomplete interpretation of the data. Throughout the report, expert testimony uses measurements of partisan fairness like the efficiency gap and mean-median scores to provide a more comprehensive understanding of the overall bias of a redistricting plan. In that, there is near uniform consensus that most plans have an inherent bias towards Republicans. If constitutional factors are equal, the Court should be compelled to select the map that limits that bias as much as possible.

Dr. Barber’s analysis was utilized in the proceedings for the Legislative Reapportionment Commission, and several experts refuted his methodology, in which he simply averaged all election results without providing proper weight across different cycles to factor for things like varying levels of turnout. Dr. Barber also agreed that his analysis did not consider a number of variables, including the voting results of all recent statewide elections, Voting Rights Act requirements, equal population requirements (his simulations allowed for a variance of 30), the splitting of wards, or communities of interest concerns (92).

Second, if Judge McCullough’s interpretation of the data were accurate, would it not nullify HB 2146 on the grounds that it provides a Democratic advantage, like the other maps that Judge McCullough dismissed? Her logic here is inconsistent and compels the Court to return to first principles, above. Which map excels on constitutional requirements and meets the standards set in League of Women Voters to ensure that Pennsylvanians of any political party do not have their votes diluted.

Draw the Lines agrees with Judge McCullough when she states “the Court considers the degree of partisan fairness reflected within the maps as a substantial factor that is entitled to appreciable weight in the final calculus,” (197). Therefore, we encourage the Court to select a map that, when it excels on the constitutional benchmarks, also excels at neutralizing any partisan bias in a map.

As noted by Judge McCullough and each of the experts who testified before the Commonwealth Court, the Pennsylvania Citizens’ Map from Draw the Lines meets the basic redistricting standards set out in the Pennsylvania State Constitution. This includes:

Achieving contiguity - As noted in Judge McCullough’s report, the Citizens’ Map satisfies the basic contiguity requirement (page 137).

Achieving equal population - As noted in Judge McCullough’s report, the Citizens’ Map satisfies the standard that districts be created “as nearly equal in population as practicable,” with a deviation of 1 person (page 138). All districts in the Citizens’ Map are composed of 764,864 or 764,865 persons.

Minimizing jurisdictional splits - The Pennsylvania Citizens’ Map splits 14 counties, 23 municipalities (7 of which are already split by county lines), and 16 wards, or 53 total splits (page 145). Judge McCullough failed to include the Citizens’ Map in her final comparison (page 147), but it finished behind only the Gressman Math & Science map and the Senate Democratic Caucus 2 Plan for the least total splits of any map submitted to the Court.

Compactness - As noted by the table on page 141 in Judge McCullough’s report, aggregated by Dr. Moon Duchin, the Citizens’ Map (or “CitizensPlan” in the table) scores at or near the top of several compactness metrics (Polsby, Reock, PopPoly) and at minimum, within an acceptable range across all compactness metrics of maps being considered. Dr. Duchin claimed that the Citizens’ Map was third or fourth among all of the maps submitted in overall compactness (page 147).

Compliance with the Voting Rights Act - The Citizens’ Map creates a majority-Black district and a majority-coalition district, which appears to be in compliance with the VRA, according to Judge McCullough (183).

Judge McCullough did not recommend the Citizens’ Map for two primary reasons: that it split Pittsburgh, and that it was biased in favor of the Democratic Party.

Splitting Pittsburgh

Draw the Lines disagrees with Judge McCullough’s opinion that splitting Pittsburgh is disqualifying. She is incorrect in stating that Pittsburgh has never been split in a congressional district before. In fact, it was split regularly up until the 1980s redistricting cycle. (See 1943, 1951, 1962, 1972).

Additionally, there is no such statute or constitutional requirement that states any particular political subdivision be kept whole during the redistricting process, or that a county’s largest municipality must be kept whole.

Judge McCullough claims the Citizens’ Map “splits the City of Pittsburgh across two congressional districts for the first time without any convincing or credible expert explanation as to why this was absolutely necessary to achieve population equality,” (201). Draw the Lines argues this is a misinterpretation of the State Constitution. As it states, districts “shall be composed of compact and contiguous territory as nearly equal in population as practicable… Unless absolutely necessary no county, city, incorporated town, borough, township or ward shall be divided.” The constitution does not require that any particular municipality be held together to achieve population equality.

Indeed, HB 2146 splits Horsham Township in Montgomery County, population 26,564. District 4 also makes an awkward incursion into the 58th Ward of northeastern Philadelphia from Montgomery County, presumably to achieve the goal of equal population. Both of these regions would certainly argue that their communities are being cracked in this map. Unfortunately, as long as the standard exists that every district must have no more than +/- 1 person deviation from the target district population, splits of political subdivisions and communities of interest will be necessary.

Rather, the totality of the statewide map needs to be compact, minimize splits, and achieve equal population. As noted above, the Citizens’ Map accomplishes that.

Judge McCullough relies on the testimony of Dr. Keith Naughton, who gave analysis on how the different maps under consideration addressed communities of interest.

As noted by Judge McCullough (page 93), Dr. Naughton “has ‘no particular experience in redistricting,’ and has never served as an expert in redistricting litigation before.” Further, “Dr. Naughton explained that ‘much of [his] professional career has been dedicated to helping Republican candidates in Pennsylvania win their seats,’” (page 94).

The Court should take Dr. Naughton’s opinion about communities of interest with as much weight as it would take an opinion from any other Pennsylvanian that may have an opinion about what neighborhoods, towns, or counties belong together in a districting map under the notion of communities of interest.

One person may feel strongly that Pittsburgh’s municipal boundaries are sacrosanct and must be held together. However, another person may hold that as soon as you cross the Monongahela River and go through the Fort Pitt Tunnel, you may technically still be in Pittsburgh but you have entered an entirely new community, with different needs and a different culture.

Ultimately, Draw the Lines leaned on the weight of our mappers, particularly those from Allegheny County, that were drawing their own districts. From the 1,500 maps submitted to the Draw the Lines competition, a plurality of them used the three rivers confluence as a natural dividing line around Pittsburgh.

Thus, what makes the Citizens’ Map so strong is that it was developed using input from 7,200 Pennsylvanians, whose opinions are just as credible as Dr. Naughton’s on something as basic as Pennsylvania culture and what their neighborhood should be like.

We’d argue that beyond expressing the desires of our citizen mappers, there are benefits to splitting the city of Pittsburgh. First, a municipal split in Allegheny County is very likely necessary to achieve population equality. If a municipality in Allegheny County must be split, several of our mappers argued that splitting the county’s largest city into two districts enhances the city’s political power rather than detracts from it. They argued that a split in a smaller municipality in Allegheny County would make it much easier for that municipality to have their representation diluted.

Secondly, as McCullough admits (155), Pittsburgh’s irregular shape means that splitting does allow you to achieve a goal that is constitutional, i.e. greater compactness.

Lastly, splitting the city of Pittsburgh does allow for an additional competitive district in the region. Maximizing competitive districts across the state was a goal of our mappers, and creating a district that both parties could win, the 17th district. Both parties would have won that seat using different statewide election data over the last 6 years.

Limiting partisan bias

Draw the Lines strongly disagrees with Judge McCullough’s reasoning that due to the geographic clustering of Democrats in Pennsylvania, it is a fait accompli that any map that attempts to minimize the inherent advantage awarded to the Republican Party is a partisan gerrymander.

She writes: “[T]he Court finds that when lines are purposely drawn to negate a natural and undisputed Republican tilt that results from the objective, traditional, and historical practice whereby Democratic voters are clustered in dense and urban areas, such activity is tantamount to intentionally configuring lines to benefit one political party over another. The Court considers this to be a subspecies of unfair partisan gerrymandering and is legally obligated, pursuant to LWV II, to look up[on] such a practice with suspicious eyes.”

Yet, there is no law that says a political party is guaranteed a certain share of representation based simply on the geographic distribution of its voters. It defies logic that if all other criteria were equal, a map could be permissible if it allows for a Republican bias because of a more efficient geographic distribution of its voters, but then a map is not permissible if it attempts to neutralize that advantage.

In fact, if the map can achieve outstanding benchmarks on the constitutional requirements like equal population, compactness, and jurisdictional splits, as the Citizens’ Map does, then the map has the burden to attempt to minimize partisan bias for either party to the greatest extent possible under Pennsylvania’s Free and Equal Elections Clause. That is what the Citizens’ Map accomplishes.

Judge McCullough claimed that “based on its credited efficiency gap score, it provides a partisan advantage to the Democratic party in contravention to the natural state of political voting behavior and bias towards Republicans in Pennsylvania,” (201).

Draw the Lines strongly disputes the notion that the Citizens’ Map’s “credited efficiency gap score” provides a partisan advantage to the Democratic Party. Dr. DeFord noted that the score was 1.6% in favor of Republicans (174). Using the publicly available website PlanScore, it gives the Citizens’ Map an efficiency gap of 3.5% in favor of Republicans when not factoring in the power of incumbency. This means Republicans would win an extra 3.5% of 17 seats, or an extra half-seat. When factoring incumbency, there is a 0.2% gap in favor of Republicans. Each of the experts found that the map favored Republicans on the Mean-Median metric (170).

To claim the Citizens’ Map provides a partisan advantage to Democrats, Judge McCullough relies on analysis from Dr. Michael Barber for partisan fairness. Draw the Lines disputes his analysis and credibility. First, Dr Barber agreed that his analysis did not consider a number of variables, including the voting results of all recent statewide elections, Voting Rights Act requirements, equal population requirements (his simulations allowed for a variance of 30), the splitting of wards, or communities of interest concerns (92).

Dr. Barber was called to testify in front of the Legislative Reapportionment Commission on the state house maps, giving testimony on partisan results of their plans in a similar fashion to the testimony he gave in support of HB 2146 for the congressional map. However, as LRC Chairman Mark Nordenberg noted in his statement on February 4th, Dr. Barber “has not published a single academic article in the areas for which his expert testimony was being presented,” (Nordenberg). Chairman Nordenberg largely dismissed Dr. Barber’s analysis on the legislative maps because other academics could not accurately replicate his work. We would encourage the Court to do the same here.

As Dr. Duchin testified, the Citizens’ Map is “far superior at leveling the partisan playing field,” whereas she characterized House Bill 2146’s performance as “consistently converting close elections to heavy Republican representational advantages,” (82).

Incumbent pairings

This was not a significant focus for Draw the Lines, and we would argue that it should not be for the court either. However, we disagree with Judge McCullough’s method for determining fairness on these pairings by saying that both parties should have the same number of incumbents paired in the same district. Pennsylvania’s population declined in more Republican-represented districts (six Republican-held districts lost population from 2010 to 2020, while only two Democratic-held districts did), meaning those districts have to expand geographically and are more likely to pair Republican incumbents together.

Conclusion

The Pennsylvania Citizens’ Map meets or beats the House Republican map selected by Judge McCullough on the legal requirements, and it is far superior on the additional metrics that are important to Pennsylvanians, like competitiveness, limiting partisan bias, and preserving communities of interest.

More importantly, it was created with unprecedented public engagement and input. It is a composite map that attempts to represent what 7,200+ Pennsylvanians collectively mapped through public Draw the Lines competitions over the last four years. The Citizens’ Map meets or exceeds the statistical averages set by the 1,500 completed individual maps, and it reflects the values that mappers declared as important to them. The Citizens' Map, in effect, represents the everyday Pennsylvania mapper.

Draw the Lines presented it to both the State House and State Senate as a possible starting point for the General Assembly's work. DTL published a narrative explaining choices of the map, which has been available on our website (drawthelinespa.org) since September. DTL also has collected 110+ public comments on the map from Pennsylvanians across the commonwealth on the map. DTL then used those public comments to make revisions to the final version, which is then submitted to the General Assembly and the Court.

The Citizens’ Map and the work of Draw the Lines has been a model for what a functional redistricting process should look like.

[1] In the graph comparing HB 2146 and the Citizens' Map, split totals were aggregated from Judge McCullough’s report, starting on page 143. Compactness scores reflect those calculated by Dr. Duchin in Judge McCullough’s report. The efficiency gap is calculated from PlanScore, the non-partisan, publicly available. These scores were determined without factoring incumbency of current congresspersons. Since there were inconsistent methodologies to determine the mean-median scores in Judge McCullough’s report, Draw the Lines is relying on the mean-median measurements provided by Dave’s Redistricting App, the publicly available redistricting platform.