In League of United Latin American Citizens ("LULAC") v. Perry, the Supreme Court, for the second time in two years, agonized over partisan gerrymandering. LULAC's rejection of a Democratic challenge to the Texas legislature's mid-decade pro-Republican congressional redistricting resembles the Court's 2004 dismissal of a Democratic gerrymandering suit against Pennsylvania's pro-Republican congressional redistricting plan in Vieth v. Jubelirer. As in Vieth, the Justices wrangled over justiciability, the substantive standard for assessing the constitutionality of partisan gerrymandering claims, and the interplay of justiciability and constitutionality. As in Vieth, the Court was highly fragmented: Vieth produced five separate opinions, while LULAC took that internal division one step further and generated six separate opinions on the partisan gerrymandering issue. As in Vieth, Justice Kennedy's was the Court's decisive voice, but none of the other Justices fully agreed with his disposition of LULAC's partisan gerrymandering question.
To be sure, unlike Vieth, LULAC actually resolved some of the legal issues presented by partisan gerrymandering. The Court confirmed that partisan gerrymandering is a justiciable issue but found that a districting plan driven solely by partisan concerns is insufficient to make out a case of unconstitutional gerrymandering. LULAC, thus, seems to clarify somewhat the law of partisan gerrymandering. By the same token, however, it still fails to define what makes a districting plan an unconstitutional gerrymander. As a result, the outlook for the future of gerrymandering litigation remains murky.
With respect to justiciability, a majority of the Court-Justices Kennedy, Stevens, Souter, Ginsburg, and Breyer-held that an equal protection challenge to partisan gerrymandering presents a justiciable claim. This removes one important uncertainty created by Vieth, or more specifically, by Justice Kennedy's Vieth concurrence. Prior to Vieth, in Davis v. Bandemer in 1986, the Court had held that gerrymandering claims are justiciable, although the Justices disagreed over what substantive standard to apply. In Vieth, four Justices, in an opinion by Justice Scalia, repudiated Bandemer's justiciability holding, while the four dissenters found that gerrymandering claims are justiciable but failed to agree on a standard for proving gerrymandering. Justice Kennedy's somewhat delphic concurring opinion agreed that, given the lack of "any agreed upon model of fair and effective representation," the Court ought to "refrain from intervention" but held open the possibility of judicial relief in the future "[i]f workable standards" are found. Vieth, thus, appeared to leave the question of justiciability up in the air. The LULAC majority, however, expressly concluded that Vieth did not disturb Bandemer's prior holding of justiciability, thus removing whatever uncertainty Vieth created.
On the merits, Justice Kennedy rejected the argument that a purely partisan motivation for a redistricting plan is enough to render the plan unconstitutional. Partisan motivation has always been a factor in assessing gerrymandering. However, as Bandemer found, when a legislature undertakes redistricting, some partisan motivation is inevitable, so partisanship by itself cannot be enough to invalidate a plan. In LULAC appellants argued that the Texas mid-decade redistricting presented an unusually clear case of sole, unvarnished partisan motivation. Texas had been redistricted in 2002 in order to satisfy the "one person, one vote" requirement in light of the 2000 census. The only reason for the unusual, constitutionally unnecessary, mid-decade redistricting was partisan-to increase the number of districts likely to elect Republicans. Whereas in other cases the role of partisanship or the balance between partisan and other factors might be debatable, here it was clear. By analogy to Shaw v. Reno's treatment of the role of race in redistricting, appellants argued that although some partisanship in districting is unavoidable, when partisanship is the sole motivating factor driving the enactment of a districting plan, that plan is presumptively unconstitutional.
Although Justice Kennedy quibbled somewhat with appellants' assessment of the extent of partisan motivation, noting that "partisan aims did not guide every line [the Texas legislature] drew," he ultimately concluded that even when partisan advantage is the sole motivation for the districting plan, that is not enough to make the plan unconstitutional. In his view, a challenge to the constitutionality of a partisan gerrymandering plan "must . . . show a burden, as measured by a reliable standard, on the complainants' representational rights." Justice Kennedy failed to define "representational rights," but the context indicates that he means the ability of a party to win a share of congressional seats that corresponds to that party's share of the vote in congressional races. Finding that the mid-decade plan produced a partisan balance in the Texas congressional delegation that more closely matched the Republican-Democratic split in the statewide vote than the plan it replaced, Justice Kennedy concluded it was a "fairer" plan notwithstanding its partisan motivation.
Justice Kennedy's assumption that the unconstitutionality of gerrymandering lies in the denial to a political party of its "fair share" of legislative seats is intuitively appealing. It grows naturally out of the equal protection case law that generated the "one person, one vote" and racial vote dilution doctrines, and it resonates with the idea of fair representation more broadly. Yet, it is problematic in several respects.
First, treating the denial of proportional representation as the gravamen of the constitutional harm implicitly reads a norm of proportionality into the equal protection clause, yet, as Justice Kennedy acknowledged in LULAC "there is no constitutional requirement of proportional representation."
Second, even assuming, in theory, that party proportionality ought to be part of the standard of fair representation, the single-member district system required by law for congressional elections and used in virtually every state for state legislative elections makes proportionality difficult to achieve in practice. The partisan division of voters across districts may not match the statewide partisan split. For example, if the backers of the two major parties are distributed evenly throughout a state, a party could win 51% of the vote in every district and 100% of the seats. On the other hand, if partisans are separated from each other and concentrated in different districts, then even significant shifts in voter support from one party to another might have little impact on district-specific electoral outcomes.
Justice Kennedy implicitly acknowledged the practical difficulty with the disproportionality standard when he discussed the "symmetry standard" proposed in an amicus brief. "Symmetry" recognizes that single-member districting plans do not necessarily produce proportionate results and instead looks to see if a plan treats both major parties symmetrically. In other words, if one party receives x% of the seats when it gets y% of the vote, then the other party should also get x% of the seats (and not more or less) when it gets y% of the vote. "Symmetry" thus defines fairness not so much as proportionality but as evenhandedness. Justice Kennedy rejected the use of the symmetry standard, finding it too dependent "on conjecture about where possible vote-switchers will reside." Moreover, in addition to its reliance on "a hypothetical state of affairs" and "counterfactual[s]," symmetry fails to "provid[e] a standard for deciding how much partisan dominance is too much."
Yet, the same criticism can be raised about any approach to gerrymandering based on the relationship between votes cast and seats won. How much disproportionality is needed to create a constitutional violation? If modest disproportionality is enough to support a constitutional claim, then districting plans will be constantly subject to litigation. If substantial disproportionality is a prerequisite to litigation, then few claims will succeed. The Bandemer plurality took the latter approach, requiring significant and protracted distortion of the seats-votes relationship in order to state a constitutional claim. As a result, over a decade and a half under the Bandemer standard, virtually all gerrymandering challenges were rejected.
Third, beyond the theoretical and practical difficulties of the proportionality approach to "representational rights," Justice Kennedy's opinion fails to appreciate the constitutional harm that occurs when the sole motivation for a districting plan is partisanship. Justice Kennedy emphasizes the democratic purpose of redistricting: "drawing lines for congressional districts is one of the most significant acts a State can perform to ensure citizens participation in republican self-governance." But an unabashedly partisan redistricting plan, adopted with the sole aim of reshaping the outcomes in particular districts constitutes the antithesis of "ensur[ing] citizen participation in republican self-governance." The purpose of such a plan is to prevent republican self-governance by creating districts that promote predictable outcomes, not competitive elections. With a purely partisan plan, the representatives are choosing their people, rather than the people choosing their representatives. Political competition and government accountability to the people both suffer when legislators are free to manipulate districts with the admitted purpose of creating safe seats for particular candidates or parties. Put another way, while Justice Kennedy expressed concern that the appellants show some impact on the "representational rights" of the plaintiff party, he failed to appreciate the potential impact of a highly partisan plan on the "representational rights" of the electorate as a whole. To be sure, in most gerrymandering cases the impact of a redistricting plan on the ability of the weaker party to win seats will be relevant to proving the partisan intent of the enacting legislature. But in LULAC the pure partisan intent was unquestioned, so further evidence of the kind required by Justice Kennedy was unnecessary.
LULAC's rulings on justiciability and on the sole-partisan-motivation theory bring some clarity to the law. Indeed, they appear to move it back toward Bandemer, which also found gerrymandering justiciable, also emphasized the electoral effect of a gerrymander over motivation, and also failed to achieve majority support for a substantive gerrymandering standard. But the extreme fragmentation of the LULAC Court makes it difficult to say anything about the future of gerrymandering litigation.
Strikingly, no Justice joined Justice Kennedy's rejection of the sole-motivation argument. Justice Stevens, joined by Justice Breyer, dissented; they would have held that sole-partisan motivation states a gerrymandering claim. Justice Scalia, joined by Justice Thomas, adhered to his Vieth position that gerrymandering is nonjusticiable and declined to address the merits. Chief Justice Roberts, with Justice Alito, joined in Justice Kennedy's disposition of the partisan gerrymandering claim but expressly declined to specify whether that was based on justiciability concerns or on the merits. Most perplexing of all, Justice Souter (with Justice Ginsburg) joined in the Court's finding of justiciability and joined Justice Kennedy's rejection of a subsidiary equal protection argument-that the 2003 plan's reliance on the 2000 census, violated the "one person, one vote" rule-but did not take any position on the merits of the gerrymandering claim. Seeing "no majority for any single criterion of impermissible gerrymander[ing]," Justices Souter and Ginsburg chose to "treat the broad issue of gerrymander much as the subject of an improvident grant of certiorari," even though as the case came before the Court as an appeal the "cert. improvidently granted" option was not actually available. Justices Souter and Ginsburg then observed that they did not "share Justice Kennedy's seemingly flat rejection" of the significance of a highly partisan process nor did they "rule out the utility of a criterion of symmetry as a test." As a result, the arguments dismissed by Justice Kennedy could have some future life.
LULAC's treatment of the partisan gerrymandering question, thus, may be as significant for the continuing divisions and uncertainties it reveals as for the result it achieved. A majority of the Court is willing to grapple with the gerrymandering issue but that majority is internally torn over what makes partisan gerrymandering a constitutional problem and when judicial intervention is appropriate. The Court's difficulty is understandable. Gerrymandering is a challenge to democratic self-government, but judicial intervention requires a judicially manageable theory of democracy compatible with the Constitution and our political institutions. It remains to be seen whether the Court can agree upon such a theory. Vieth and LULAC suggest that the outlook is not promising.