The 33-year search for judicial standards in partisan gerrymandering cases came to a crashing conclusion this Term. As Justice Kagan noted in dissent, there is more than a touch of irony that the declaration of judicial incapacity by the Supreme Court arrives just as lower federal courts and some state courts had converged on a methodology for distinguishing the merely opportunistic from the pathological in political self-dealing. Much will be said about the Court’s analytic moves, the apparent impact of Justice Kennedy’s retirement, and the seriousness of the threat to democratic structures. I wish to leave these aside in favor of three reflections on thirty years of an “unfairness” approach to the problem of partisan gerrymandering, to borrow the Court’s term of opprobrium.
First: Ever since the Court refused to revisit the republican guarantee clause in Baker v. Carr, the Court has forced claims of distortions of the political process into the “familiar” precincts of equal protection law. This meant that challengers had to point to a particularized right and an aggrieved rights holder. Although it took decades of doctrinal development, this approach worked passably well for apportionment and for racial access claims. In the partisanship context, it led to the search for a measure of distortion and for some baseline range of reasonableness in electoral results. The forceful opening of Justice Kagan’s dissent unintentionally highlights the gulf between the nature of the systemic harms from partisan control of redistricting and the increasingly social scientific claims that have shaped litigation.
Beginning with the first piece I wrote in this area 25 years ago, I argued that the harm was systemic or structural. But the doctrine was otherwise, and the Court in Davis v. Bandemer anticipated the emergence of a technocratic approach to the mandated quantum of partisan representation, akin to that which defined minority vote dilution claims. That expectation might have reflected both the swift transition from Baker v. Carr to Reynolds v. Sims, and the contemporaneous transition from the hydra-headed White/Zimmer test of Section 2 of the Voting Rights Act to the manageable three-part quantitative inquiry of Thornburg v. Gingles, decided the same day as Davis. Yet partisanship proved less amenable to number-crunching than race. The social science offerings of swing ratios, efficiency gaps, and ever more complicated mathematical models could not successfully capture the basic intuition that a party’s electoral results should bear some relation to its legislative representation. By the time North Carolina Republicans could leverage 53% of the vote into 10 out of 13 seats, all boundaries of decency had been broken. Surely, it seemed, some test should be able to say that this was a step too far.
But what if the Court was asking the wrong question? Since Davis, and even since Gaffney 13 years earlier, the Court had rejected any constitutional challenge to the notion that incumbent authorities could control the conditions of their own accountability to the voters. Yet it is an elemental principle of fair play that the umpires be independent of the contestants, whether in sporting contests or in the sound advice given to every new democracy with regard to election administration. But for reasons of both substance and standing, the Court insisted that there must be a simple rights denominator to the claim, not a structural one. So if 10 was too much, what about 9, what about 8, and so on. As every first-year student must learn, at some point the reductio ad absurdum becomes just plain absurdum.
In retrospect, the doctrinal search for proof of too much partisanship seems a terrible shell game. There was a glimmer of a saner though poorly defined alternative in Justice Kennedy’s offer in Vieth of a First Amendment approach. It seemed unlikely that relocating the claim to the Elections Clause or the First Amendment by itself would make much difference. But the First Amendment might have shifted focus to the state conduct rather than the entitlement of the claimant. Viewpoint discrimination claims assess state conduct and the expressive component of its actions, not the precise entitlement of the claimholder. In this spirit, I joined with the impressive lawyers of the ACLU in trying to work through systematically the boundaries of a political line-drawing as expressive conduct approach, an effort that corresponded to Michael Kang’s academic take.
It turned out, however, that our argument ran into the same problem as the structural approach generally. The logic of viewpoint discrimination, like the logic of institutional independence, does not readily suggest a demarcation between some political self-dealing and too much of it. The point of the structural approach is to replace the institutionally difficult ex post judgments of outputs with ex ante protections. Think Miranda and Dickerson as just this kind of prophylaxis. The First Amendment similarly prohibits the infusion of viewpoint preferences into state decisionmaking so as to avoid any inquiry into how much was the action corrupted. The structural challenge was always easy to dismiss as unrealistic. Sometimes accommodation is also the path to defeat.
Second, Chief Justice Roberts makes much of the capacity of Congress to remedy the extremes of gerrymandering, echoing Justice Scalia’s early efforts in Vieth to euthanize this area of law. This is a more challenging jurisprudential move than appears. The fact that Congress can reject the product of state redistricting by refusing to seat a delegation is an odd surviving manifestation of Madison’s commitment to the legislative veto as a guarantor of the constitutional project. Every constitutional democracy needs some form of check on the immediate popular will in order to maintain its commitment to future accountability. Noah Feldman’s recent biography of Madison focuses on the centrality of congressional review of the constitutionality of state law in his grand design, a central feature that he introduced again and again at the convention, and always lost – except oddly in this one area.
The key to Madison’s design here was the indifference of the national legislature to local battles and hence the desire to protect the national institution against anything sounding in local political capture. But Madison’s constitution did not anticipate faction, made no room for political parties, and assumed that the scale of the Republic was a shield against local passion. We can dispute whether this broke down in the proselytizing for the Constitution itself, in the election of 1796, on the glades of Weehawken, or with the Van Buren party formation. But today, the national party is the fulcrum of partisan abuse at the local level, not its antagonist.
What does modern originalism say when the original design recognized an evil that must be addressed, but got the remedy wrong? It is as if the Constitution specified a governmental duty to protect against infectious disease, but mandated the distribution of leeches as the cure. It is hard to see how identifying congressional incapacity does any violence to the overall constitutional scheme. Paradoxically, it was Chief Justice Roberts in Shelby County who invoked the history of the legislative negative as reason to distrust the remedial authority of Congress as opposed to the courts in satisfying constitutional commands. The rise of the national party disables Congress from affording a remedy, but leaves the original constitutional concern intact.
Third is where this all heads now. For some time, I have worried that the failure to redress the partisan ambitions of gerrymandering would drive claims into what I once termed “the suffocating category of race.” I am no longer as convinced of this. First, much of the Shaw-style racial gerrymandering case law emerged from a time where empirical tools were much cruder and race was the key variable that could be manipulated by line-drawers as in Bush v. Vera. For jurisdictions formerly covered by Section 5, the fact of administrative reporting forced open consideration of the racial implications of redistricting. It took the evidentiary somersaults of Justice Breyer in Cromartie v. Easley to squeeze racial considerations from the record in North Carolina in the last cycle. But Section 5 is gone, sophisticated redistricting tools no longer rely on crude racial data, and partisan agitators are now counseled to trumpet their partisan aims unabashedly.
The most pernicious effect might be the collateral consequences on other aspects of the right to vote. Justice O’Connor posited that gerrymandering was self-limiting because too fine an effort at partisan gain made the beneficiaries electorally vulnerable. Chief Justice Roberts makes the same point indirectly by invoking the post-Vieth flip of seats in Pennsylvania. But if the Court declares a free pass for gerrymanderers and the temptation is toward slim margins, how do the victors protect themselves from demographic and political change? The unfortunate answer may be efforts to drive down the vote of potential adversaries. So far this has taken the largely ineffectual form of demand for voter identification. But re-registrations, pre-election purges, payment of civil debt prior to voting, and other mechanisms are on the horizon. Untrammeled partisan gerrymandering might be an invitation to open season on the right to vote.
We have passed through earlier periods when voting and political accountability themselves were the battleground of democracy. For much of this history, constitutionalism and legal reform were important means of rescue from immediate political temptation. Removing all constraints on partisan gerrymandering in an era of heightened polarization invites the basest impulses of political self-dealing.