Comments on the House and Senate preliminary maps
Monday January 17, 2022 To: Chancellor Mark Nordenberg Chairman of the Pennsylvania Legislative Redistricting Council Mr. Nordenberg, I commend you for your even tempered and courteous conduct of the LRC. However, I strongly believe that both Senate and House preliminary maps violate the “partisan fairness and proportionality” principle of your Pennsylvania Redistricting Advisory Council, and the map must be adjusted to be in accordance with that. 1. The Redistricting Principles of the Pennsylvania Redistricting Advisory Council state “…any proposed map must comply with the requirements of federal law, including most specifically, the constitutional requirement to maintain population equality among congressional districts and the provisions of the Voting Rights Act as they apply in Pennsylvania.” 2. These Redistricting Principles also state, “Ensuring partisan fairness and proportionality requires that parties have the opportunity to translate their popular support into legislative representation with approximately equal efficiency such that the proportion of districts whose voters favor each political party should correlate to the statewide preferences of the voters. Partisan fairness requires preventing structural advantage from being baked into the map so as to allow one party to more efficiently translate votes into seats in the delegation [bold underline mine]. In evaluating a proposed map, the Governor should analyze how it would have performed in a full range of prior statewide elections when compared to other potential maps which could have been drawn. A map with expected performance proportional to statewide voter preference should be favored as comporting with broad principles of fairness.” 3. The map experts testifying Friday January 14 in the afternoon all agreed that there is a Republican bias in the preliminary LRC State House and Senate maps proposed in December 2021. Based on the above principles, this should not be the case. Why would this condition be proposed in a state where the state Democrat candidates have recently received an approximately equal number of votes as the Republican candidates, and where registered Democrats outnumber registered Republicans by almost 6%? The partisan fairness and proportionality principle indicates that seats in the House and Senate should be approximately equal, or slightly favor the Democrats. To be otherwise would violate this principle, subject the LRC plan to lawsuits, and be a miscarriage of justice. 4. In the state House, “Republicans would be expected to win 2.5% R extra seats in a hypothetical, perfectly tied election, favoring Republicans in 94% of predicted scenarios.” In other words, “Predicted 48% D / 52% R seat share across scenarios vs. 50% D / 50% R vote share.” 5. In the state House again: Section 5003(c)(3) of the FTVA specifies that partisan fairness should be assessed using a state's two most recent elections for U.S. President and two most recent elections for U.S. Senate. When the preliminary LRC map is tested according to these scenarios, the results favor Republicans in all required scenarios, according to the Campaign Legal Center’s PlanScore analysis: a. U.S. President 2020: 0.9% R advantage b. U.S. President 2016: 2.9% R advantage c. U.S. Senate 2018: 4.7% R advantage d. U.S. Senate 2016: 7.1% R advantage 6. According to PlanScore, the LRC map would give Republicans the advantage in the following 4 measures of partisan fairness: a. Metric Value % of Scenarios favoring Dems b. Efficiency Gap 2.3% Pro-Republican 15% c. Declination 0.14 Pro-Republican 8% d. Partisan Bias 2.5% Pro-Republican 7% e. Mean-Median Difference 1.0% Pro-Republican 7% 7. Dave’s Redistricting Analysis summary shows that the LRC preliminary map falls one seat short of being fair to Democrats: LRC plan would likely result in 105 seats being won by Democrats, but 106 seats are the fairer representation based on the vote share: a. Dave's Redistricting "prominent measures of partisan bias: positive number favors Republicans b. Metric Description c. Proportional 0.50% The simple deviation from proportionality using fractional seat shares d. Efficiency gap 2.95% The relative two-party difference in wasted votes e. Gamma 1.82% The fair difference in seats at the map-wide vote share f. g. Seats bias 2.20% Half the difference in seats at 50% vote share h. Votes bias 1.22% The excess votes required for half the seats i. Partisan bias 2.86% The difference in seats between the map-wide vote share and the symmetrical counterfactual share j. Global symmetry 2.74% The overall symmetry of the seats-votes curve k. Partisan bias rating 72 The combined rating of seats bias & votes bias l. m. Declination 5.45° A geometric measure of packing & cracking n. Mean–median 0.63% The average vote share across all districts minus the median vote share o. Turnout bias -0.29% The difference between the map-wide vote share and the average district share p. Lopsided outcomes 4.34% The relative two-party difference in excess vote shares q. r. Proportional seats 106.48 The fractional Democratic seats for the map-wide vote share s. Geographic seats 109.06 The fractional Democratic seats implied by county political geography t. Geographic bias -1.27% The bias due to county political geography u. Map seats 105.48 The fractional Democratic seats for the map v. Boundary bias 1.76% The bias due to district lines 8. Looking solely at Presidential voting, here is a summary of recent past Democrat leaning Presidential voting: “Pennsylvania …. voted Democratic in the six Presidential elections prior to 2016 [with margins from 2.5% to 10.3%]. That election saw Donald Trump win the state by 0.7%, ….. Again competitive in 2020, Joe Biden won by 1.2% ... [in other words] since 2000, Pennsylvania has voted Democratic 83.3% of the time and Republican 16.7% of the time.” 9. Pennsylvania is now a Democratic leaning state, based on these data above. Your final map must lean Democratic, not Republican, in order to comply with your Redistricting Advisory Board’s partisan fairness and proportionality principle and to NOT be in violation of the Voting Rights Act. 10. To quantify the Pennsylvania House and Senate situation, the table below shows the most recent election results (2016, 18 and 20) making crystal clear the extreme lack of proportional representation in both the PA House and Senate. The Republicans have had a 7% advantage, when comparing the difference between the percent of the popular vote vs. the percent of the seats won. This extremely unfair Republican advantage is a violation of the partisan fairness and proportionality principle which the LRC is obligated to enforce. It is the obligation of the LRC to correct this unfairness by its final map! PA state election analysis 2016, 18 and 20 Year Branch Party Popular vote % Seats won % Advantage or -Disadvantage 2020 Senate Republican 50.87% 60.00% 9.13% Democrat 49.13% 40.00% -9.13% 2020 House Republican 52.73% 55.67% 2.94% Democrat 46.57% 44.33% -2.24% 2018 Senate Republican 45.51% 52.00% 6.49% Democrat 53.83% 48.00% -5.83% 2018 House Republican 44.43% 55.00% 10.57% Democrat 54.19% 45.81% -8.38% 2016 Senate Republican 50.87% 60.00% 9.13% Democrat 49.13% 40.00% -9.13% 2016 House Republican 50.49% 59.60% 9.11% Democrat 48.76% 40.40% -8.36% unweightd avg Republican 49.15% 57.05% 7.90% Democrat 50.27% 43.09% -7.18% 11. To summarize, the final LRC House map must be changed to eliminate the Republican bias currently embedded in the preliminary plan, and to become slightly Democratic leaning, to comply with all the redistricting requirements. Specifically, to make the map fairly represent Pennsylvania’s new voting realities the LRC must adjust the House map to lean slightly Democratic – where Democrats have a fair chance of winning 106 seats that the analyses overwhelmingly indicate to be fair. 12. Thank you for this opportunity to give citizen input!