In Gill v. Whitford, twelve voter-plaintiffs challenged the Wisconsin legislature’s 2011 redistricting as a violation of Fourteenth Amendment equal protection rights and a burden on First Amendment associational rights. After a bench trial, the district court concluded that the 2011 map had the intent and unjustified effect of “plac[ing] a severe impediment on the effectiveness of the votes of individual citizens on the basis of their political affiliation,” and entered judgment in favor of plaintiffs on both constitutional claims. The trial court based its findings of discriminatory effects on a compelling and extensive record that would have withstood appellate review under the deferential standard applicable to such causal and statistical inferences.
In a recent unanimous decision, however, the Supreme Court did not reach the merits of the claims, and instead vacated the district court’s judgment on the grounds that the plaintiffs had failed to demonstrate standing to bring their case. Describing the necessary standing as limited to “voters who allege facts showing disadvantage to themselves as individuals,” the Court determined that only four of the twelve voter-plaintiffs had complained of injuries specifically stemming from the packing or cracking of their own districts, and that the case at trial had improperly focused on statewide harm. The four plaintiffs pleading individual harms will have another opportunity to advance their claims on remand.
Without directly addressing the justiciability or merits of the claims, the Court noted that the plaintiffs had based their showing of statewide harms on measurements of partisan asymmetry, in accordance with the “social science tenet that maps should treat parties symmetrically.” The basic idea that courts can evaluate the severity of partisan gerrymandering in terms of “the extent to which a majority party would fare better than the minority party, should their respective shares of the vote reverse,” had been previously considered in LULAC v. Perry but found by Justice Kennedy to “shed no light on ‘how much partisan dominance is too much.’ ” The Gill plaintiffs had responded to this concern by offering the efficiency gap as a quantitative measure of partisan asymmetry, defined by political scientist Eric McGhee and legal scholar Nicholas Stephanopoulos as “the difference between the parties’ respective wasted votes in an election, divided by the total number of votes cast.” As the Court pointed out, however, a single statewide number could not describe the effects of the 2011 redistricting on individual voters in different parts of the state.
The efficiency gap has other shortcomings and peculiarities, as other commentators have pointed out. It can treat highly competitive 50-50 districts as problematic and lopsided 75-25 districts as fairly drawn, and is prone to imprecision in states having congressional delegations too small to closely reflect vote shares. It also enshrines a previously unrecognized principle that a party’s legislative majority should be twice the proportionate size of its statewide vote majority (at least when turnout is roughly uniform across districts).
A more fundamental problem with the use of the efficiency gap and other partisan asymmetry measures to detect and quantify the effects of gerrymandering is that the absence of asymmetry — i.e., symmetry — then becomes the normative baseline against which election outcomes are measured. But strict partisan symmetry is, in fact, rare. For example, Gill‘s lead plaintiff William Whitford was “naturally” packed in his district with other Madison Democrats, a situation that would favor Republicans even in the event of a hypothetical Democratic statewide electoral swing. Such urban concentrations of Democrats in Madison and Milwaukee led the trial court to find that “Wisconsin has a modestly pro-Republican political geography,” although it ultimately concluded that this natural tilt was much smaller than the “large partisan effect” of the challenged gerrymander. In focusing on standing, the Supreme Court did not review these findings. But Chief Justice Roberts’s characterization of the efficiency gap as “sociological gobbledygook” at oral argument bespoke his reluctance to weigh maps on a scale whose values had no meaning to “the intelligent man on the street,” let alone one whose baseline of zero had no corresponding real-world benchmark.
The simple uniformity of strict partisan symmetry belies the fact that value judgments about a state’s redistricting are shaped by the possibilities and limitations of the state’s unique political geography. The uneven distribution of voters, political boundaries, and other geographic features of a state can produce biased partisan outcomes (e.g., partisan asymmetries and packed and cracked districts), even in maps drawn by nonpartisan commissions. The purpose and effect of partisan gerrymandering, however, is to produce maps that tend to result in partisan outcomes that are not only biased, but atypically so.
We share the view of amicus Eric Lander that the most straightforward and objective way to support a valid statistical inference of partisan bias is to show that actual and/or expected partisan outcomes under a challenged map are extreme outliers when compared with those that would typically obtain under a nonpartisan redistricting process. (Justice Breyer also tentatively endorsed the manageability of this approach at oral argument in Gill.)
Our approach is to use an ensemble of compliant maps to investigate the typicality of outcomes from redistricting in a state. We generate this ensemble as a large random sample from the set of all maps that could have resulted from a nonpartisan process guided by a set of redistricting criteria. The criteria may include traditional redistricting principles, other legitimate legislative objectives, historical considerations, and any other factors stipulated as typical influences on the redistricting process. The choice of criteria is contestable, but the aim is to formulate a sufficiently robust nonpartisan conception of the choices and constraints facing the redistricting task that meaningful inferences of harm can be drawn. Having specified the redistricting criteria and generated the ensemble of maps, we can incorporate actual election data to calculate the partisan outcomes that would have resulted from the same voting behavior under each of the maps in the ensemble. A biased partisan outcome from a challenged map can then be compared against the distribution of outcomes in the ensemble to determine whether it is an extreme outlier. Thus naturally packed or cracked regions in the challenged map will also typically occur in the ensemble and will not give rise to an inference of bias. The ensemble approach absorbs these natural asymmetries as features of the state’s political geography, not as bugs degrading the inferential power of a partisan symmetry measure.
In Common Cause v. Rucho, North Carolina’s 2016 congressional redistricting (itself a replacement for racial gerrymanders enjoined in Cooper v. Harris) proved to be an outlier in comparison with our ensemble of 24,518 plans in many respects. Under most (55 percent) of the plans in the ensemble, North Carolina’s 2016 congressional voting would have produced a delegation of eight Republicans and five Democrats. In contrast, the ten Republicans and three Democrats elected under the challenged map represented an outcome that occurred in less than 0.7 percent of the ensemble. This was no accident: Republican legislators in February 2016 had specifically instructed their redistricting coordinator Thomas Hofeller to maintain the 10-3 partisan composition of North Carolina’s congressional delegation that had been elected under the racially gerrymandered plan. As the ensemble evidence demonstrated, the Republicans’ injection of partisanship into the redistricting process had the purpose and effect of achieving a result that could not confidently be left to chance.
Our ensemble approach also served to identify the specific districts that were packed and cracked to achieve the outlying result. The percentage of votes cast in 2012 and 2016 for Democratic candidates in the three most Democratic districts in the challenged plan were significantly higher than the corresponding percentages for plans in the ensemble, reflecting the packing of districts 1, 4, and 12. Also, the percentage of votes cast in 2012 and 2016 for Democratic candidates in the fourth through sixth most Democratic districts in the challenged plans were significantly lower than the corresponding percentages for plans in the ensemble, reflecting the cracking of districts 7, 8 and 11.
The district court found the ensemble analysis to “provide strong evidence that the General Assembly intended to subordinate the interests of non-Republican voters and entrench the Republican Party in power.” While the court accepted the evidentiary strength of the statistical inferences that could be drawn from the ensemble approach, the court pointedly stopped short of fashioning a constitutional standard from any of the empirical analyses presented in the case: “Plaintiffs need not show that a particular empirical analysis or statistical measure appears in the Constitution to establish that a judicially manageable standard exists to resolve their constitutional claims. Rather, Plaintiffs must identify cognizable constitutional standards to govern their claims, and provide credible evidence that Defendants have violated those standards…. The Supreme Court long has relied on statistical and social science analyses as evidence that a defendant has violated a standard set forth in the Constitution or federal law.”
At its final conference of the Term, the Supreme Court vacated and remanded the district court’s decision in Rucho for reconsideration in light of Gill. At this point, the North Carolina case appears to be the stronger of the two. Thanks to the ensemble approach, the record in Rucho already supports the trial court’s findings as to which specific districts were packed and cracked, a point the plaintiffs tried to drive home in supplemental briefs on standing filed immediately after the Gill decision.
Given the limits Gill has placed on the justiciability of partisan gerrymandering challenges brought under an equal protection theory, the plaintiffs in both states may be well advised to develop their claims of associational injury more fully. In a concurring opinion in Gill joined by Justices Ginsburg, Breyer and Sotomayor, Justice Kagan noted that where the complaint is that “the gerrymander has burdened the ability of like-minded people across the State to affiliate in a political party and carry out that organization’s activities and objects[,] … the valued association and the injury to it are statewide, [and] so too is the relevant standing requirement.”
While partisan gerrymandering jurisprudence remains unsettled (even more so with the impending change in the composition of the Court), we hope further proceedings in Rucho will establish the ensemble approach to statistical inference from empirical election data as a stable touchstone for judicial fact-finding in years to come.
Andrew Chin (UNC Law)
Greg Herschlag (Duke Math)
Jonathan Mattingly (Duke Math)