On January 27, 1998, the Winston-Salem Journal featured an article discussing the lack of competitive Congressional races in North Carolina. John Hoeffel, Six Incumbents Are a Week Away From Easy Election, Winston-Salem Journal, Jan. 27, 1998, at B1. The article noted that more than half of North Carolina’s incumbent representatives were currently running unopposed in the upcoming election. Id. “You have virtually no competition anywhere across the state,” observed State Senator Mark McDaniel. Id. McDaniel, who had voted against the General Assembly’s latest redistricting plan, complained that the General Assembly was “in the business of rigging elections.” Id.
Gerrymandering has a long and colorful history in the United States. Most observers agree that partisan gerrymandering is, as Justice Alito recently observed, “distasteful.” But although federal courts have long limited the states’ ability to draw district lines in ways that disadvantage racial groups, they have been hesitant to limit the states’ ability to fashion districts that advantage one political party over another.
A case currently pending before the Supreme Court could test the limits of that deference. Gil v. Whitford involves a challenge to Wisconsin’s state legislative maps, under which Republicans have won more seats than their overall share of the statewide vote would suggest. For example, Republican candidates won 60 of the state’s 99 Assembly seats in 2012 despite receiving 48.6% of the statewide vote. See Whitford v. Gil, 218 F. Supp. 3d 837, 853 (W.D. Wis. 2016). The Court’s decision in Gil could set a new standard for evaluating partisan gerrymander claims and have a significant impact on American politics.
Background on Gerrymandering Claims
In the United States, the practice of manipulating district lines to promote a desired electoral outcome is nearly as old as the Constitution itself. Patrick Henry reportedly drew Virginia’s district lines in an unsuccessful effort to keep James Madison out of the first United States Congress. The word “gerrymander” was coined in 1812 after Massachusetts Governor Elbridge Gerry approved a salamander-shaped district that was drawn to benefit his party. In light of this historical pedigree and federalism concerns, federal courts have generally been reluctant to impose limitations on a state’s ability to fashion its own legislative maps.
In the mid-twentieth century, however, the Supreme Court began to recognize some restrictions on a state’s discretion in drawing legislative maps. In 1964, the Court held that the Fourteenth Amendment’s Equal Protection Clause requires states to apportion districts on a population basis according to the one-person, one-vote principle. Reynolds v. Sims, 377 US 533, 568–69 (1964). Under this principle, states may not dilute the voting power of citizens by grossly overpopulating a district. Id. The Court has also recognized that both the Constitution and the Voting Rights Act of 1965 prohibit states from using single or multimember districts that have the effect of diluting the voting power of racial groups. See, e.g, Thornburg v. Gingles, 478 US 30, 80 (1986). Although there have been many cases defining the boundaries of these principles, there is little debate that overpopulated districts and racial gerrymanders are unconstitutional.
In contrast with malapportionment and race, the Court has had difficulty articulating an approach to partisan gerrymanders. In Davis v. Bandemer, a divided Court rejected a challenge to Indiana’s state legislative districts. 478 US 109 (1986). Six of the nine justices agreed that partisan gerrymandering claims are justiciable under the Equal Protection Clause. But these justices were unable to muster a majority as to the proper standard for evaluating such challenges, particularly as to whether the results of a single election could be sufficient to establish that a particular map had a discriminatory effect on voters of a particular party. As a result, the plaintiffs’ claims were dismissed without a clear standard from the Court.
Nearly two decades later, the justices were still unable to reach a consensus. In Vieth v. Jubelirer, the Court rejected a challenge to Pennsylvania’s legislative districts. 541 US 267 (2004). Four justices proposed various standards to apply to the Equal Protection approach endorsed in Bandemer, while four others wrote that Bandemer should be overruled. Believing that no workable standard had yet been proposed, Justice Kennedy cast his tiebreaking vote in favor of the state. But Justice Kennedy also argued that partisan gerrymandering claims were justiciable under the Equal Protection Clause, and he suggested that a law that “has the purpose and effect of subjecting a group of voters to disfavored treatment by reason of their views” could also violate the First Amendment. Id. at 314 (Kennedy, J., concurring in the judgment). Justice Kennedy added, “Of course, all this depends first on courts’ having available a manageable standard by which to measure the effect of the apportionment and so to conclude that the State did impose a burden or restriction on the rights of a party’s voters.” Id. at 315 (Kennedy, J., concurring in the judgment).
The Gil Case
The plaintiffs in Gil purport to answer Justice Kennedy’s call for a “manageable standard by which to measure the effect of apportionment.” Drawing on the work of various political scientists, the plaintiffs proffer at least three different statistical methods designed to quantify the degree to which legislative maps advantage certain political parties. One such method estimates the total number of seats a political party would expect to win if the statewide vote were split evenly between two parties. Another method looks to the difference between a party’s mean or average vote share and the share it received in the median district. By far, however, the metric that has received the most attention is the so-called “efficiency gap,” which compares the number of votes each party “wastes” in a given election. The efficiency gap starts from the premise that the purpose of every vote is to win a legislative seat. Under this logic, a vote is “wasted” if (1) it was cast in favor of a losing candidate; or (2) it was cast in favor of a winning candidate who already had enough votes to win that district. The efficiency gap compares each party’s wasted votes to determine the degree to which one party benefitted from the drawing of district lines.
Although these techniques differ in their specifics, all seek to evaluate a state’s political map as a whole, rather than assessing the partisan tilt of any particular district in isolation. As a result, they represent a different approach than that taken in the Court’s one-person, one-vote and racial gerrymandering cases. In those cases, the Court has required voters to challenge the boundaries of specific districts, rather than relying on the overall effect of the map as a whole. See, e.g., Alabama Legislative Black Caucus v. Alabama, 135 S. Ct. 1257, 1265 (2015).
Perhaps for this reason, much of the oral arguments in Gil focused on defining the right at issue. Echoing Justice Kennedy’s concurrence in Vieth, the plaintiffs argue that, in addition to the Equal Protection Clause, partisan gerrymanders also implicate the First Amendment right to freedom of association. Justice Ginsburg suggested that partisan gerrymanders implicate “the previous right to vote.” Justice Gorsuch argued that partisan gerrymandering claims arise under Article IV, Section 4 of the Constitution, which requires the federal government to “guarantee to every State in this Union a Republican Form of Government,” and which the Court has long dismissed as giving rise only to non-justiciable political questions. See, e.g., Highland Farms Dairy v. Agnew, 300 US 608, 612 (1937).
The Court’s choice of framework is particularly important in because of the conceptual difficulties that statewide partisan gerrymandering presents. The first of these issues relates to a legislature’s intent. As with race, the Court has long acknowledged that political considerations, such as the protection of incumbents and the grouping of “communities of interest,” are legitimate considerations that legislatures may take into account when drawing districts. See, e.g., League of United Latin American Citizens v. Perry, 548 US 399, 434, 440–41 (2006). In fact, the Court rejected a racial gerrymandering claim from North Carolina after concluding that the district’s boundaries were drawn on the basis of partisanship, rather than race. See Easley v. Cromartie, 532 US 234, 257 (2001). If the Court recognizes a partisan gerrymander in Gil, it may need to more clearly differentiate between permissible political intent, such as the “protection of incumbents” or the desire to group “communities of interest” together, and impermissible political intent, such as the desire to “subject a group of voters to disfavored treatment by reason of their views.”
Another conceptual difficulty relates to the degree to which a map must be biased in order to establish a partisan gerrymandering claim. The Court has rejected the idea that the Equal Protection Clause guarantees a right to proportional representation, either in terms of race or partisan affiliation. See, e.g., City of Mobile, Ala. v. Bolden, 446 US 55, 75–76 (1980); Vieth 541 US at 287–88. If the Court sustains the challenge in Gil, it may need to articulate some method or test—statistical or otherwise—for determining how much partisanship is too much for constitutional purposes.
This task is especially difficult because political geography can produce partisan bias results in the absence of intentional gerrymandering. As the populations of urban and rural areas grow more polarized, the odds increase that a particular map may naturally favor one party over another. In Gil, for example, the plaintiffs proffered an expert report with information about the seventeen legislative maps across the country that exhibit the most partisan bias; ten of those maps were drawn by courts, bipartisan committees, or purportedly neutral redistricting commissions.
Finally, Gil raises the issue of how enduring a partisan bias must be in order to support a partisan gerrymandering claim. In Bandemer, the plurality rejected plaintiff’s argument that a single election cycle is sufficient to show discriminatory impact on members of a political party. 478 US at 135. In Gil, the plaintiffs produced evidence showing a similar impact across two election cycles. See 218 F. Supp. 3d at 853. On one hand, there is no obvious reason why two elections over a two-year period should be considered qualitatively different than a single election. On the other hand, a rule that requires many election results could effectively negate the utility of a partisan gerrymandering claim because most legislative maps are redrawn every ten years in response to the decennial Census.
The Court’s decision in Gil comes at an interesting period in American politics. Although gerrymandering is nearly as old as the country itself, improvements in technology have made it easier to divide the population into demographic groups with reasonably predictable voting behavior. And because state legislative and Congressional maps stand to be redrawn after the 2020 census, a decision for or against the plaintiffs in Gil will soon be felt nationwide.
Ryan Niland is a business litigator in the Winston-Salem Office of Womble Bond Dickinson. Before joining the firm, he served as a law clerk to the Honorable Thomas D. Schroeder.