February 19, 2019
Will the Supreme Court Rein in Partisan Gerrymandering?
Senior Attorney, Voting Rights at the Southern Coalition for Social Justice
The author is a counsel for the League of Women Voters of North Carolina (“LWVNC”) in a partisan gerrymandering case scheduled to be argued at the U.S. Supreme Court on March 26, 2019.
“What’s different about the partisan gerrymandering cases in 2019 compared to the ones argued in 2018?” “How can you win without Justice Kennedy?” These are the most common inquiries lobbed at litigators by Supreme Court observers who watched the Court punt on the partisan gerrymandering cases from Wisconsin and Maryland cases last term, and who now hold a healthy amount of skepticism that the Court, with Justice Kavanaugh having replaced Justice Kennedy, will do anything to rein in partisan gerrymandering when it hears arguments in the North Carolina and Maryland cases on March 26.
The skepticism is understandable, but there are several things different about how this term’s gerrymandering cases have made their way to the docket, the facts in these cases, and the options available to the Supreme Court should it finally be ready to settle upon a theory of liability in partisan gerrymandering cases.
First, procedurally, these cases now are more properly developed for the Court’s consideration of the merits of the proposed legal standards. The Maryland litigants are not asking the Supreme Court to order new maps on the eve of an election. Following the Court’s guidance in Gill v. Whitford, the North Carolina litigants, on remand, supplemented the record to leave no question that plaintiffs had standing, and the district court so found. The Supreme Court can now turn its full attention to the merits of the cases, and the various standards for liability the district courts approved in each case. And having examples of both Republican and Democratic gerrymandering before the Court should not be undervalued. Evidence of the excesses by both parties on the same day may reinforce cries from democracy advocates that voters of every ilk recognize and suffer legal harms from gerrymanders that allow politicians to choose their voters rather than the reverse—as our constitutional system expects. The juxtaposition of these cases, along with their procedural posture, represents a significant change from the last term.
Second, while the facts in Maryland obviously remain the same, the Court will for the first time confront the full story of the egregious partisan gerrymander that North Carolina’s Republican legislature enacted in 2016. Significantly, North Carolina’s 2011 congressional map had already been invalidated for containing two districts that had violated the 14th Amendment —and that racially-gerrymandered map produced a congressional delegation with 10 Republicans and 3 Democrats in a state considered to be one of the most “purple” and competitive states in the nation. Once the courts struck down the racially gerrymandered map in 2016, the House Republican leading the redrawing effort, Rep. David Lewis, declared his intent to keep that lopsided margin, and he unabashedly bragged in on the record that the only reason the legislature would not pass a plan that would produce 11 Republican and 2 Democratic districts was because it was not physically possible. With no hesitation, he explicitly referred to the 2016 map as a partisan gerrymander, and he proclaimed his view that partisan gerrymandering was not illegal. Far more than just “smoking gun” intent evidence, the whole world witnessed the actual firing of the gun intended to silence the political voices of millions of North Carolina voters. The facts in Maryland are nearly as egregious, with the Democratic governor, charged with leading the redistricting process, blatantly acknowledging his intent to change the 6 Democratic and 2 Republican congressional delegations in the state to 7-1. With intent facts like those in North Carolina in Maryland, these cases hew as closely as one could imagine to the hypothetical that Justice Kennedy posed in the Gill argument: what if the face of the statute requires maximum political subjugation of one political party, which counsel for the Wisconsin map-drawers conceded would likely violate the Constitution. That is, to the extent, there are “close calls” of unconstitutionality that exist, the cases this term are not among them. Facts setting the outer limits of a standard ought to aid litigants who are asking the Supreme Court to establish a new rule.
Third, this year is different because Justice Breyer, one of the Court’s academics, is getting his dream seminar to explore this topic. In argument in the Benisek v. Lamone case last year, he asked counsel if it would not be better to just gather the different theories presented by the various litigants across the country and hear at once all of the options available to the Court in adopting a standard. Tuesday, March 26, 2019, will be just that day.
The three cases being argued before the Court on March 26 present an extensive menu of options from which the Court can choose its preferred means to curb the worst of what Justice Alito called “distasteful” gerrymandering without injecting mandatory judicial review into every redistricting process. The cases offer the Supreme Court multiple theories of liability—based on vote dilution unconstitutional under the Fourteenth Amendment; based on Article I, Section 2, which reserves for the people, not state legislators, the right to choose; and based on two variations of liability under the First Amendment (viewpoint discrimination and harms to the right to freely associate). It is not that the litigants cannot agree on a standard—it is that the practice of intentional partisan gerrymandering as happened in North Carolina and Maryland actually offends numerous constitution guarantees designed to protect our democracy. Just as both the lower courts in these cases have concluded, a viable legal rule is possible that can bring an end to these blatant partisan excesses.
In particular, the LWVNC case presents the court with a very logical and step-by-step test for unconstitutional partisan gerrymandering based on a vote dilution theory—a test that has been refined since Gill and embraced by the district court on remand following the Gill decision. The three-part test proposed asks whether, using three different metrics, there is a consistent pattern of severe and durable partisan asymmetry in the challenged plan. If the answer to the question is yes, the appropriate move for potential litigants is to, as one of the LWVNC’s experts described it, “throw a flag” and examine the other two factors that comprise the test. That is, to see if, first, there is evidence of discriminatory intent and second, whether there are any neutral redistricting norms that would otherwise explain the observed partisan asymmetry. Importantly, the proposed measures of partisan asymmetry would not require review of any plan based simply on a departure from partisan symmetry or proportional representation. Rather, metrics referenced in this test only trigger greater scrutiny in the most extreme and unfair maps, where voters in packed or cracked districts have no meaningful ability to influence electoral outcomes.
Too much has been read into Justice Roberts’ reference to the metrics that assess partisan asymmetry in Gill as “sociological gobbledygook,” implying that those metrics are too complicated for judges to understand. But these very measures are not new. District judges nationwide, with the aid of expert testimony, have been assessing very similar social science methods in voting rights cases for more than 30 years. In vote dilution cases brought under Section 2 of the Voting Rights Act, liability partly depends upon proof of legally-significant racially polarized voting. There exist at least three different ways of measuring racially polarized voting: homogenous precinct analysis, ecological regression, and ecological inference. In fact, when the Court approved the use of homogenous precinct analysis and ecological regression as tools for measuring racially polarized voting in Thornburg v. Gingles in 1986, ecological inference did not even exist as a methodology. But the social science analysis in this arena continued to evolve, and the ecological inference was developed as a variation of ecological regression that improves upon some of the weaknesses of the latter methodology. Most courts now will give greater weight to analyses employing ecological inference rather than ecological regression. All of this suggests that the judicial branch can handle social science metrics, including the tools that evolve over time.
On a related note, one can embrace the vote dilution theory of liability in a partisan gerrymandering case even without fully embracing Section 2 of the Voting Rights Act (VRA). Some academics and Court observers are suspicious that the Roberts Court may be looking for an opportunity to invalidate Section 2. Regardless of one’s views about the continuing need for Section 2 of the VRA (though, in my view, present evidence in this nation’s political realm demonstrate that this protection remains sorely needed), there is nothing remarkable about federal courts considering social science metrics as one part of an analysis of whether new districts are required in a redistricting case. Indeed, this approach is no more remarkable than reviewing economic modeling of competitiveness in an antitrust case. Lower federal courts have been successfully reviewing and drawing conclusions from statistical analyses of voting patterns for decades. This theory of liability tracks the well-developed path of Gingles-like cases that, for decades, have been viewed as offering the judiciary a manageable standard for reviewing the legality of redistricting plans. And the other theories, under the First Amendment and Article I, Section 2, likewise present the Court with additional manageable standards that could be employed.
Of course, there is still the reality that these cases will be argued before a different Court than the cases from the last term—a Court without Justice Kennedy. The fact that the cases are better postured for consideration on the merits does not preclude the Court from finding grounds to punt yet again—leaving untouched the Davis v. Bandemer holding that partisan gerrymandering cases are justiciable while declining to adopt any of the standards proposed by the litigants. And it does not foreclose the worst-case scenario—that five Justices might reverse Bandemer and conclude that partisan gerrymandering is non-justiciable. But the Supreme Court, regardless of present composition, generally looks to craft narrow rules, and the good news is that amongst the several theories of gerrymandering liability now being offered to the Court, several propose ways to impose limits on the very worst of the democracy-eroding maps seen in recent years without inviting judicial involvement any time redistricting happens. And this trend, given the options in front of the Court, is the reason for real hope.
On March 26, 2019, the Supreme Court will be presented with a wide range of menu options from which to choose a route to restore faith in the integrity of our electoral systems, particularly in the construction of districts that elect members of Congress. A win under any theory is a win for all citizens seeking a more fair democracy, with electoral structures that allow all voters to participate in the political process on equal footing.