Carter, C. et al s. v. Chapman, L., et al. (concurring)

Annotate this Case

Download PDF

[J-20-2022] [MO: Baer, C.J.] IN THE SUPREME COURT OF PENNSYLVANIA MIDDLE DISTRICT CAROL ANN CARTER, MONICA PARRILLA, REBECCA POYOUROW, WILLIAM TUNG, ROSEANNE MILAZZO, BURT SIEGEL, SUSAN CASSANELLI, LEE CASSANELLI, LYNN WACHMAN, MICHAEL GUTTMAN, MAYA FONKEU, BRADY HILL, MARY ELLEN BALCHUNIS, TOM DEWALL, STEPHANIE MCNULTY AND JANET TEMIN, : No. 7 MM 2022 : : : ARGUED: February 18, 2022 : : : : : : Petitioners : : : v. : : : LEIGH M. CHAPMAN, IN HER OFFICIAL : CAPACITY AS THE ACTING SECRETARY : OF THE COMMONWEALTH OF : PENNSYLVANIA; JESSICA MATHIS, IN : HER OFFICIAL CAPACITY AS DIRECTOR : FOR THE PENNSYLVANIA BUREAU OF : ELECTION SERVICES AND NOTARIES, : : Respondents : : ---------------------------------------------------------- : PHILIP T. GRESSMAN; RON Y. DONAGI; : KRISTOPHER R. TAPP; PAMELA GORKIN; : DAVID P. MARSH; JAMES L. : ROSENBERGER; AMY MYERS; EUGENE : BOMAN; GARY GORDON; LIZ MCMAHON; : TIMOTHY G. FEEMAN; AND GARTH : ISAAK, : : Petitioners : : : v. : : : LEIGH M. CHAPMAN, IN HER OFFICIAL CAPACITY AS THE ACTING SECRETARY OF THE COMMONWEALTH OF PENNSYLVANIA; JESSICA MATHIS, IN HER OFFICIAL CAPACITY AS DIRECTOR FOR THE PENNSYLVANIA BUREAU OF ELECTION SERVICES AND NOTARIES, : : : : : : : : Respondents JUSTICE DONOHUE CONCURRING OPINION OPINION FILED: March 9, 2022 DECIDED: February 23, 2022 I agree with the selection of the Carter Plan, and I join in the Majority’s analysis, including its invocation of partisan fairness as a factor in its selection. Because this case requires the Court to select one of thirteen maps, most of which satisfy the four “floor” criteria identified in League of Women Voters v. Commonwealth, 178 A.3d 737 (Pa. 2018) (“LOWV”), we must use a tiebreaker. In my view, in this circumstance, the logic of LOWV compels us to consider the degree of partisan fairness among the plans. Contrary to Justice Brobson’s suggestion, none of us wish “to serve as the mirror on the wall and choose the fairest map of them all.” Dissenting Op. at 8 (Brobson, J.). And while Justice Brobson seems to be less opposed to our selection of the Carter Map than “the analysis that the majority uses to break a partisan impasse,” the fact remains that the political branches have unfortunately thrust the selection of a map on us. Justice Brobson fears that we have “invited, not discouraged this Court’s future involvement in the congressional redistricting process,” id., but does not set forth an alternative selection that would avoid his pessimistic prediction. While which map should be chosen is subject to good faith disagreement, we must choose, and “I don’t know” is the one answer we cannot give. [J-20-2022] [MO: Baer, C.J.] - 2 In LOWV, we held that to meet constitutional muster under our Free and Equal Election Clause,1 a map must satisfy four neutral “floor” criteria: “compactness, contiguity, minimization of the division of political subdivisions, and maintenance of population equality among congressional districts.” LOWV, 178 A.3d at 817. The submitted maps admirably complied with that dictate.2 The proponents of each map submitted the performance metrics corresponding to the neutral criteria.3 Pertinently, virtually all submissions contained an analysis of how each of their plans performed in terms of predicted partisan fairness.4 Undoubtedly, this was driven by the following passage from LOWV: As we have repeatedly emphasized throughout our discussion of [Article I, Section 5] the overarching objective of this provision of our constitution is to prevent dilution of an individual’s vote by mandating that the power of his or her vote in the selection of representatives be equalized to the greatest degree possible with all other Pennsylvania citizens. We recognize, then, that there exists the possibility that advances in map drawing technology and analytical software can “Elections shall be free and equal; and no power, civil or military, shall at any time interfere to prevent the free exercise of the right of suffrage.” Pa. Const. art. 1, § 5. 1 2 I acknowledge that the Carter Plan does not score the best on the floor criteria. See Majority Opinion at 27-33. I also agree with the Majority that there are trade-offs involved when giving one criterion more importance than others. See id. at 28. Moreover, unlike Justices Mundy and Todd, I do not view picking the best plan on these four criteria to be an objective exercise. The fact that both Justices wish to pick the plan that best complies with the floor criteria but end up favoring different plans illustrates the point. Additionally, the parties have largely acknowledged that the 2018 map implemented by this Court produced fair outcomes, and, further, that the maps now presented are comparable or superior to the 2018 map. Thus, I do not find that the differences on the floor criteria are so great that any map can be ruled out on that basis alone. Hence, we must turn to a tiebreaker. 3 See Majority Opinion at 28 n.23 (describing metrics used to evaluate compactness). 4 The Khalif Plan was the only one that did not analyze partisan performance. [J-20-2022] [MO: Baer, C.J.] - 3 potentially allow mapmakers, in the future, to engineer congressional districting maps, which, although minimally comporting with these neutral “floor” criteria, nevertheless operate to unfairly dilute the power of a particular group’s vote for a congressional representative. See N.T. Trial, 12/13/17, at 839–42 (Dr. Warshaw discussing the concept of an efficiency gap based on the number of “wasted” votes for the minority political party under a particular redistricting plan). However, as the case at bar may be resolved solely on the basis of consideration of the degree to which neutral criteria were subordinated to the pursuit of partisan political advantage, as discussed below, we need not address at this juncture the possibility of such future claims. Id. (emphasis added). Although the task of the Court in this matter is distinctly different than the constitutional challenge to the enacted redistricting plan at issue in LOWV, the parties in this matter obviously recognized that it was not enough to satisfy the neutral factors, because even though compliant with the drawing requirements, it was important that the plan did not “unfairly dilute the power of a particular group’s vote for a congressional representative.” Id. The purpose of our Free and Equal Election Clause is not to ensure that congressional district maps contain clean lines with few divisions and a minimum of irregular borders encompassing an equal number of people. It is not a cartography lesson. The overreaching objective of this constitutional provision is to prevent dilution of a citizen’s vote. Consequently, just as the political branches have an obligation to consider partisan fairness when enacting a redistricting plan, so too must this Court when put in the position of having to select one from the many that were submitted to us. Partisan fairness is not merely a subordinate factor to be considered. When, as here, all of the plans are compliant with the floor criteria, consideration of the degree of partisan [J-20-2022] [MO: Baer, C.J.] - 4 fairness must, in my view, drive the ultimate selection of a plan in the circumstances in which this Court finds itself.5 The degree of partisan fairness is measurable. Measurement is imperfect because it cannot account for, among other variables, the quality of candidates. Also, where, as here, the submitted plans have no performance record, the partisan fairness metrics are predictive, not actual. But the tools are available and widely used. The record in this case is replete with expert analyses of the predicted partisan fairness of the plans. Admittedly, the data sets used to calculate the metric and, in some cases, the methodologies within the designated partisan fairness tests differed among the parties’ experts. Nevertheless, I do not find that the lack of one perfect test for measuring partisan fairness precludes us from considering that factor. It simply means that we should look for the most comprehensive review available. Based on the record before us we have one comprehensive, comparative analysis of each of the submitted plans’ predicted performance on partisan fairness. The Gressman plaintiff’s expert, Dr. Daryl DeFord, performed an “apples to apples” analysis comparing all plans to each other. In other words, he reconciled the data set and methodologies used by the various experts. From my perspective, it forms a reliable basis to rank the predicted partisan fairness of the submissions. Unlike some other experts, who used limited data sets, Dr. DeFord’s 5 I do not suggest that any of the plans submitted for consideration reflect a degree of partisan unfairness that is disqualifying in a constitutional sense, nor do I suggest the level of partisan fairness that a duly enacted congressional district plan must attain. I do, however, believe that when this Court is forced to choose among plans, the plans that perform the best on partisan fairness metrics must rank above the others. [J-20-2022] [MO: Baer, C.J.] - 5 analysis examined “vote totals for [eighteen] statewide general elections[.]” Expert Report of Dr. DeFord at 5. He elaborated on this point: For each of my partisan-fairness metrics, I have used election results from [eighteen] statewide general elections that took place in the Commonwealth between 2012 to 2020. This represents the general elections races for U.S. President, U.S. Senate, Governor, Attorney General, Auditor General, and State Treasurer. This dataset includes examples of elections where each of the major political parties’ candidates won the overall statewide vote. Many of these races were decided by small margins, particularly those in which a Republican candidate won the overall election. Thus, I also included the 2017 Supreme Court Justice election in my analysis, as that election had a larger margin of victory for the Republican candidate than the other elections had. Looking at this breadth of election results helps us better understand and model the political geography of a state and related realistic vote outcomes. Id. at 22. Dr. DeFord explained that using general elections was useful because “the percentages reported reflect the two-party vote share from the two most successful candidates, which in these elections were always the Democratic and Republican candidates.” Id. at 22-23. Each of the partisan fairness metrics he used “requires one first to determine, for each of the [eighteen] general elections, which candidate, the Democratic or Republican, carried each of the districts in each redistricting plan at issue.” Id. at 23. Then, that information was used “to plot a seats-votes curve, and they also become inputs for the partisan-symmetry computations described below.” Id. These [J-20-2022] [MO: Baer, C.J.] - 6 results were then used to generate a mean-median score6 and an efficiency gap score.7 Dr. DeFord then compared all plans to each other on these two metrics, plus four other measures generated by the PlanScore.org website. The following table, which is copied from the Gressman’s Brief in Support of Exceptions at page 59 with slight alterations to the headings, reflects the results of that comparison. (In his report, Dr. DeFord indicates that a negative score indicates a Republican lean.) Partisan Fairness Metric [Tier one (least bias)] [Tier two] [Tier three (most bias)] Dr. DeFord’s Average Sen. Dems 2 (-0.3%) Carter (-1.6%) Reschenthaler 2 (-2.6%) Mean-Median Gressman (-0.8%) Ali (-1.8%) Reschenthaler 1 (-2.7%) (using all 18 elections House Dems (-0.9%) Sen. Dems 1 (-1.9%) Voters of PA (-2.7%) from 2012 to 2020) Governor (-1.0%) Citizen-Voters (-2.0%) HB2146 (-2.9%) (closer to zero is better) Draw the Lines (-1.2%) “The mean-median score is a metric related to partisan symmetry. In simple terms, a plan that exhibits partisan symmetry is one that is likely to treat the parties similarly in terms of seat outcomes given equal votes received by all candidates statewide. That is, if Party A is expected to turn a 55%-to-45% statewide vote advantage into a 10-to-7 seats advantage, then a symmetric result would require Party B to turn a similar 55%-to-45% statewide vote advantage into a 10-to-7 seats advantage.” Report of Dr. DeFord at 26 (footnote omitted). 6 7 We explained the concept in LOWV. Dr. Warshaw suggested that the degree of partisan bias in a redistricting plan can be measured through the “efficiency gap,” which is a formula that measures the number of “wasted” votes for one party against the number of “wasted” votes for another party. Id. at 840–41. For a losing party, all of the party’s votes are deemed wasted votes. For a winning party, all votes over the 50% needed to win the election, plus one, are deemed wasted votes. The practices of cracking and packing can be used to create wasted votes. LOWV, 178 A.3d at 777. [J-20-2022] [MO: Baer, C.J.] - 7 Dr. DeFord’s Average Carter (-0.4%) Draw the Lines (-1.6%) Voters of PA (-4.8%) Efficiency Gap Governor (0.6%) Sen. Dems 1 (-2.5%) HB2146 (-6.3%) (using the same 18 Gressman (0.8%) Citizen-Voters (-2.6%) Reschenthaler 1 (-7.8%) elections) Sen. Dems 2 (1.0%) Ali (-2.7%) Reschenthaler 2 (-7.8%) House Dems (3.3%) PlanScore Efficiency House Dems (1.2% D) Ali (2.4% R) Reschenthaler 2 (6.3% R) Gap Gressman (1.4% R) Sen. Dems 2 (2.4% R) Reschenthaler 1 (6.4% R) Carter (1.8% R) Sen. Dems 1 (2.5% R) HB2146 (6.6% R) Governor (1.9% R) Draw the Lines (3.5% R) Voters of PA (6.8% R) Citizen-Voters (4.6% R) PlanScore Declination Gressman (0.03 R) Ali (0.07 R) Reschenthaler 2 (0.18 R) House Dems (0.04 D) Sen. Dems 1 (0.07 R) HB2146 (0.19 R) Carter (0.05 R) Sen. Dems 2 (0.07 R) Reschenthaler 1 (0.19 R) Governor (0.05 R) Draw the Lines (0.10 R) Voters of PA (0.20 R) Citizen-Voters (0.13 R) PlanScore Partisan Bias Gressman (0.9% R) Sen. Dems 1 (1.8% R) Citizen-Voters (4.3% R) Governor (1.1% R) Ali (1.9% R) Reschenthaler 2 (5.9% R) Carter (1.3% R) House Dems (1.9% D) Reschenthaler 1 (6.2% R) Sen. Dems 2 (1.5% R) Draw the Lines (2.9% R) Voters of PA (6.5% R) HB2146 (6.3% R) PlanScore Mean-Median Gressman (0.4% R) Sen. Dems 1 (0.6% R) Citizen-Voters (1.7% R) Difference Carter (0.4% R) House Dems (0.7% D) Voters of PA (2.2% R) Governor (0.4% R) Ali (0.7% R) HB2146 (2.3% R) Sen. Dems 2 (0.5% R) Draw the Lines (1.0% R) Reschenthaler 1 (2.4% R) Reschenthaler 2 (2.4% R) This comparison establishes that four maps submitted for our consideration separate them from the field: the Carter Plan, the Gressman Plan, the Governor’s Plan, and the second Senate Democratic Caucus plan. The Gressman Plan performs the best, with the remaining three all scoring slightly lower. Although the Carter Plan is not the best performer, the other plans contain concerning anomalies in their physical configuration. Namely, as further explained, those plans make changes that depart radically from the historical treatment of certain established communities of interest. Because the Carter Plan does not contain these anomalies and its partisan fairness score is nearly identical to those other three maps, I agree that it is the best option. [J-20-2022] [MO: Baer, C.J.] - 8 The three maps which score better on partisan fairness draw districts that depart from historically recognized communities of interest that, in my view, are too drastic for this Court to adopt. The most salient of these are: the decisions to split the City of Pittsburgh (the Governor and Senate Democratic Caucus) and the decision to place Pittsburgh in a district with Washington County along with splitting Bucks County (Gressman Plan). Communities of interest are in the eyes of the beholder. A determination of what qualifies as a community of interest, and what those interests are, involves a mixture of local knowledge and political considerations uniquely determinable by the political branches within the confines of the floor constitutional criteria. If an adopted districting plan resulted in a map that split the City of Pittsburgh and otherwise met the LOWV criteria, then the split could be a valid choice. The same could be said for the Bucks County split that resulted in a Latino minority opportunity district and the combination of the City of Pittsburgh with Washington County based on the rationale that they are part of the same standard metropolitan statistical area. From where I sit, I have no legitimate way to decide whether the tradeoffs for more substantial compliance with the floor criteria involved with these significant changes in the historical treatment of these areas are acceptable.8 Therefore, I cannot endorse the selection of these maps when 8 For example, a bipartisan group of current and former Washington County elected public officials submitted an amicus brief urging this Court to select any plan but the Gressman Plan due to the fact it would create a new congressional district containing all of Washington County and the City of Pittsburgh. These individuals argued that Washington County and parts of Allegheny County, while “hav[ing] much in common,” actually “have little in common[.]” Amicus Brief at 5. Moreover, they predicted that the City of Pittsburgh would dominate Washington County. Id. at 6. [J-20-2022] [MO: Baer, C.J.] - 9 the Carter Map manages not to make those significant changes and still scores very highly on partisan fairness. Because the outcome achieved in the Carter Plan9 satisfies the LOWV floor criteria and is among the best in preventing dilution of an individual’s vote, as demonstrated in its partisan fairness metrics, without disrupting long recognized communities of interest, I join in its selection as the 2022 Congressional District Plan. As discussed in other opinions, the Carter Plan was designed using the “least change” approach. I agree with the Majority that our focus should not be on the method used in creating the map – it should be on the outcome. Majority Opinion at 27. 9 Regarding whether this Court can apply a clear standard in selecting a map, Justice Dougherty favorably cites the “least change” approach used by the Carter Plan mapmaker. See Concurring Op. at 3 (Dougherty, J.). Justice Wecht likewise cites that approach as a favorable criterion, albeit not as a sole tiebreaker. See Concurring Op. at 19-20 (Wecht, J.). Justices Mundy and Todd both desire to select the map which best follows the neutral floor criteria. See Dissenting Op. at 5 (Todd, J.); Dissenting Op. at 9 (Mundy, J.). However, this shared belief in the correct standard did not yield the same answer. I note that courts in analogous circumstances have asked parties to brief the question of whether a clear standard should be adopted. See Johnson v. Wisconsin Elections Comm’n, 967 N.W.2d 469, 476 (Wi. 2021) (“[W]e ordered the parties to address four issues. ... (3) The petitioners ask us to modify existing map using a ‘least change’ approach. Should we do so, and if not, what approach should we use?”). While the adoption of a fixed standard is desirable, without the benefit of advocacy I believe this Court is ill-equipped to clearly answer that question. For instance, Justice Mundy uses the “Borda system,” which was not used by any of the parties, and the weights Justice Mundy gives to the floor criteria were not subject to examination. In the absence of advocacy on the viability of a fixed standard, I believe that it is incumbent upon us to rely on the record. [J-20-2022] [MO: Baer, C.J.] - 10

Some case metadata and case summaries were written with the help of AI, which can produce inaccuracies. You should read the full case before relying on it for legal research purposes.

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.