Today, I’m Thankful for: Whitford v. Gill

In a recent post, this blog examined the most significant racial and political gerrymandering cases unfolding in the year ahead. And, just in time for Thanksgiving, the United States District Court for the Western District of Wisconsin has issued an historic decision in one of those cases—Whitford v. Gill (3:15-cv-421)—and has struck down a state legislative redistricting plan as an unconstitutional partisan gerrymander for the first time ever. This is something we can all be thankful for, and special thanks are due to the Campaign Legal Center, which helped bring the case, and Nick Stephanopoulos and Eric McGhee, who set out the “efficiency gap” test put forth in the case.

Although the parties are yet to brief the district court on the proper remedy to be applied, an appeal to the Supreme Court is undoubtedly forthcoming. There, the case will await review along with Harris v. McCrory (16-166), which is likely to be heard or decided in the aftermath of the Court’s decision in McCrory v. Harris (15-1262). What are the odds that the decision is upheld? Let’s carve into the district court’s meaty opinion (sorry, give me one holiday pun), and find out.


In setting out the basis and elements of the claim, the Court held that “the First Amendment and the Equal Protection clause prohibit a redistricting scheme which (1) is intended to place a severe impediment on the effectiveness of the votes of individual citizens on the basis of their political affiliation, (2) has that effect, and (3) cannot be justified on other, legitimate legislative grounds.” Slip Op. at 56.

The Constitutional Basis

Taking this step by step, it is important to first point out that the Court grounded its decision in both the Equal Protection clause and the First Amendment. The Court began by tying the claim back to Fortson v. Dorsey—the origin of the racial and political dilution claim—and noting that the foundational cases “instruct that vote-dilution cases are governed by the same standards as other equal-protection claims, namely the plaintiffs must establish both a discriminatory intent and a discriminatory effect.” Slip Op. at 34. And yet, in tracing the lineage of the political vote dilution claim, the Court also held that “the specific [equal-protection] test for political gerrymandering set forth in Bandemer no longer is good law.” Slip Op. at 47.

How does the Court thread this needle? By looking to the Anderson-Burdick line of cases, which are based “directly on the First and Fourteenth Amendments” and which draw from equal-protection jurisprudence, but which do not “not engag[e] in a separate Equal Protection Clause analysis.” Slip Op. at 52. This line of cases is typically applied to electoral regulations, such as voter identification requirements or candidate registration requirements, that burden voters and/or candidates. This is what allowed the Court to rely heavily upon the Supreme Court’s prior equal-protection-based political gerrymandering jurisprudence, including the standard and logic of Bandemer, without being held accountable to the result of Bandemer itself. (In an important hedge, the Court also distinguishes Bandemer based on its lack of sufficient factual findings. Slip Op. at 79.)

It should be noted, however, that the Court invokes the AndersonBurdick line of cases only to draw an analytical parallel and not to actually rest its holding upon that line in cases. Slip Op. at 54 (“We therefore believe that there is a solid basis for considering the associational aspect of voting in assessing the gravamen of the harm allegedly suffered by the plaintiffs.”). In other words, the Court relies upon those cases to justify splicing the First and Fourteenth Amendment rationales together but does not go on to actually apply the Anderson-Burdick test to the facts before it.


For the first element of the claim, the plaintiff must prove that the plan was passed with the intent “to place a severe impediment on the effectiveness of the votes of individual citizens on the basis of their political affiliation.” As I have written, this formulation is critical because it accepts the existence of “political considerations” while distinguishing between legitimate and illegitimate justifications for state action. Slip Op. at 57 (“That some political considerations may intrude into the redistricting process without running afoul of the Constitution, however, does not answer the question whether partisan favoritism is permissible.”).

This formulation is also important because it focuses on a very specific and narrow form of intent: the intent to burden the votes of citizens based on their political affiliation. This is a far more targeted inquiry than the caricature of “partisan intent” attacked by the dissent. The dissent argues that “partisan intent” cannot form an element of the claim because “partisan intent” is inherent in politics itself. (“[P]artisan intent is not illegal, but is simply the consequence of assigning the task of redistricting to the political branches of government.” Slip Op. Dissent at 127.)

As discussed in my prior post, the dissent here improperly conflates a rule against “partisanship”—which would be impossible and undesirable to implement—with a rule against accepting “partisan advantage” as a legitimate justification for state action burdening or classifying citizens—which is not only possible to implement but absolutely necessary to the analytical foundations of our constitutional democracy. If “partisan advantage” constitutes a legitimate state interest, then rational-basis review becomes impossible. Every law could be justified on this basis. For example, the fact that a state legislature may be able to pass a voter identification bill and justify it in court based on an interest in “preventing voter fraud” does not mean that the same state legislature could pass the same bill and justify it in court based on an interest in “voter suppression.” Partisan intent itself cannot be unconstitutional as a matter of functional reality, but it also cannot be accepted as a rational basis for state action as a matter of constitutional law. The concept that partisan advantage itself can constitutionally justify burdening or classifying citizens based on their political beliefs is a dangerous idea that should find no quarter in a constitutional democracy grounded in rule of law.

This is no different than in racial gerrymandering cases. The dissent argues that “[t]he intent to weaken a racial group’s political power in drawing district lines is always and everywhere wrongful. The same is not true for political motivations.” Slip Op. Dissent at 121. For support, the dissent cites to Justice Kennedy’s contention in Vieth that “Race is an impermissible classification. . . . Politics is quite a different matter.” Slip Op. Dissent at 122. But neither is true. Protecting minority ability-to-elect under the Voting Rights Act necessarily entails “weakening a racial group’s political power in drawing district lines.” It is not possible to strengthen minority ability-to-elect in a district without weakening majority ability-to-elect in that district. Thus, it is hardly the case that this is “always and everywhere wrongful.” It is permissible and lawful under the Voting Rights Act because the justification is legitimate; namely, the law advances an interest in prophylactic protection against the unlawful, intentional dilution of minority voting power by racial majorities for raw racial advantage. A racial-vote-dilution plaintiff proceeding under the Fourteenth Amendment could hardly satisfy the “intent” element of his or her claim by pointing to the state’s reasonable efforts to comply with the Voting Rights Act’s remedial requirements. A racial-vote-dilution plaintiff could, however, satisfy the intent element of the claim if the state attempted to justify its weakening/strengthening by reference to raw racial advantage (whether for majority or minority voters). Such an interest, and such an intent, is illegitimate.

By narrowly focusing on an intent to suppress voters’ effective franchise on the basis of their party affiliation, the majority avoids sweeping in any legitimate political objectives of redistricting and any legitimate uses of political classifications. And, by implicitly rejecting partisan advantage as a legitimate state interest, the majority prevents normal-course partisanship from becoming a freestanding constitutional justification for burdening or classifying citizens.


For the second element of the claim, the plaintiff must prove that the plan had the effect of “plac[ing] a severe impediment on the effectiveness of the votes of individual citizens on the basis of their political affiliation.” Although the majority used various, alternative formulations for describing this effect, I rely upon this broader definition. Why? Because in the absence of a governing standard, the Supreme Court has an obligation to set out the proper test. In my view, this is the most likely formulation to survive.

Before discussing the nature of the constitutional injury described by the majority, however, there are two points worth noting. First, the majority held the efficiency-gap test to provide evidence of partisan gerrymandering rather than provide a dispositive measure of partisan gerrymandering. I have noted before the general problems with arguing that a “model’s specific kind of injury—as defined by its particular mathematical threshold—has always marked the constitutional bounds of political gerrymandering,” and have advocated “changing the ‘efficiency gap’ from a dispositive measure into an informative measure” in order to allow the Whitford plaintiffs to prevail and the claim to proceed. The majority adopted this view, holding that the efficiency gap, and disproportionality, was “evidence of a constitutional violation” and not the “violation itself.” Slip Op. at 85. This is important to prevent the claim from resting on a presumed right to “proportional representation,” which has been repeatedly rejected by the Supreme Court.

Second, the dissent trotted out a classic objection that is likely to be revived on appeal given the fact that it was largely ignored by the majority opinion: identifiability/mutability. In protesting the majority holding, the dissent states, “The assumption underlying Plaintiffs’ entire case is that party affiliation is a readily discernable characteristic in voters and that it matters above all else in an election.” Slip Op. Dissent at 123. The dissent goes on to question this “assumption,” noting that, unlike race, “party affiliation is not set in stone or in a voter’s genes.” Slip Op. Dissent at 124.

It is high time this argument be laid to rest. As an initial matter, the question is not whether party affiliation is immutable but whether the voter’s likely party choice is identifiable. As a matter of constitutional law, it must be considered identifiable. If this were not so, the Supreme Court’s decision in Cromartie would not be possible and the “party v. race” defense commonly relied upon by states in defending racial gerrymandering cases would not be legally available. If a state can successfully use identifiable gerrymandering by party as a shield against allegations of classification by race, then a plaintiff must be able to successfully use identifiable gerrymandering by party as a sword in stating allegations of dilution by party. One cannot credibly contend that political identities are easily ascertainable in one context and impossible to discern in the other.

Moreover, the mutability of the characteristic is irrelevant. Equal protection has never been the exclusive domain of classification by immutable characteristics, even if heightened scrutiny in warranted in such circumstances. If a legislature burdens or classifies its citizens on any basis for irrational or illegitimate reasons, the law does not survive just because that basis is mutable. As I have argued, “[t]he Court would not let an OPOV violation go unchecked upon the basis that residence is a ‘mutable’ characteristic, despite the fact that voters are regularly moving between districts and between states, lending an air of unpredictability to the districting exercise.” As the Karcher Court stated, “the well-known restlessness of the American people means that population counts for particular localities are outdated long before they are completed.” Residence information, like partisan identity, is reasonably identifiable despite its mutability and flux. And quantitative dilution, like qualitative dilution, can violate the equal protection clause nonetheless.

The one place the majority opinion stumbles, however, is in its description of the actual injury being vindicated. At one point in its opinion, the injury is posited as the statewide map’s impact on “the ability of Democrats to translate their votes into seats.” Slip Op. at 114. For the reasons pointed out by the dissent, this is a questionable approach in an otherwise robust decision.

It makes sense to focus on the effect of dilution of an “opportunity to elect the representative of choice” in a single district (the only race for which the vote is actually cast), but if one focuses on “an inability to convert statewide vote totals into seats in the legislature,” Slip Op. Dissent at 136, the injury becomes more difficult to credibly position as a violation of an individual constitutional right. The intent may be impermissible, but if there is no individual right to a certain degree of proportionality in statewide vote inputs and statewide seat outputs, then there is no injury when these inputs and outputs diverge.

Even worse, the majority describes the injury earlier in the opinion as a dilution of voters’ equal opportunity to elect representatives of their choice on a statewide basis. Slip Op. at 30. It is difficult to understand what this even means. The court has blended the typical “opportunity to elect” dilution language applicable on a district-by-district basis with the statewide outcomes concept it later endorses, but the latter very purposefully has little relation to the former. The “opportunities” language in the early vote-dilution cases was specifically devised to ensure that the violation of the right was not made dependent upon the specific proportionality of the “outcomes.”

The majority attempts to circumvent these questions by noting the hybrid equal-protection and associational roots of the claim. Slip Op. at 54. At the end of the day, perhaps this will be enough. But the question of “injury” is where the dissent makes its strongest stand and unloads the greatest number of legitimate counterarguments. The dissent points out that this seems to “enshrin[e] . . . a phantom constitutional right” to proportional representation, Slip Op. Dissent at 120, and that, “in voting, a citizen is simply expressing a choice about who he believes is a better candidate to represent his own district,” not necessarily a preference for the overall partisan makeup of the legislative body, Slip Op. Dissent at 147-48. As the dissent points out:

“A given voter might like an incumbent Republican in his own district, even if that voter leans Democratic in other respects, and so such a voter will vote for the Republican assembly candidate even while preferring that his vote does not translate into additional Republican seats in the assembly. Such a voter would be surprised if his wasted Republican vote were used in some sort of ex post facto calculus to determine whether the Republican candidates won “enough” seats that year. And what of ticket-splitters and independents? Imagine a voter who votes for a Democratic assemblyman, a Republican state senator and a Republican governor. What are we to make of such a ballot, except to conclude that the voter is expressing individual preferences about individual races, rather than some kind of global desire to increase seats for a given party?”

The majority’s opinion does not seem to adequately respond to these concerns, especially given the Supreme Court’s longstanding objections to grounding dilution claims—political or otherwise—in any kind of entitlement to proportional outcomes.

There is, however, a simple workaround to this dilemma. Even if there were no injury based on disproportional seat outcomes or dilution of statewide partisan influence, there would still be a district-specific dilution of a voter’s opportunity to elect their state representative of choice based on political identity. This is why the majority’s broader formulation of the harm—namely, that the plan had the effect of “plac[ing] a severe impediment on the effectiveness of the votes of individual citizens on the basis of their political affiliation”—provides a standard that the Supreme Court can adopt even if it were to disagree with the more specific articulation of the “statewide injury” adopted by the district court.

The statewide map achieved its aggregate partisan purpose by cracking and packing voters based on their political affiliations on a district-by-district basis. Because the majority was careful to stipulate that the efficiency-gap measure is evidence of unconstitutional dilution and not determinative of unconstitutional dilution, there is no reason why voters should be prohibited from challenging the dilution of their district-by-district ability-to-elect by making reference to, among other things, statewide evidence such as a large efficiency gap. Where there is an intent to create proportional numbers of safe seats on a statewide basis, such as in Gaffney, or an intent to create competitive districts, changes in district-by-district ability-to-elect have a legitimate justification to defeat any showing of “invidious partisan intent.” However, where there is evidence of a statewide intent to entrench one party in power through district-by-district dilution and where there is evidence that the aggregate exercise is effective in obtaining this entrenchment, this should be able to help carry the burden of proving district-by-district dilution under a totality of circumstances evaluation.

None of this is premised on any kind of constitutional right to “proportionality”; instead, it is merely premised on the notion that one’s ability-to-elect in a district cannot be intentionally diluted on the basis of political affiliation for the purpose of a party obtaining a “partisan advantage” on a statewide level.

Just as it was the district court’s responsibility to take the analytical tools and proposals set out by the parties and to craft a judicial standard for reviewing partisan gerrymandering claims, so too will it be the Supreme Court’s responsibility to take the district court’s opinion and the parties’ arguments into account in fashioning its test to govern such claims. Statewide entrenchment is, and was, effectuated using district-by-district dilution. If the Supreme Court has any reservations about the district court’s definition of a statewide injury, the easy solution is to adopt the same approach to vindicate district-specific injuries.


Finally, the majority sets out a “third” element of the claim: that the plan “cannot be justified on other, legitimate legislative grounds.” Importantly, the majority declined to specify which party carries the burden of making this proof. Slip Op. at 91 (“The record before us does not require us to anticipate how the Supreme Court will resolve the allocations of proof on this issue. . . . Assuming the plaintiffs have the ultimate burden of proof on the issue, they have carried that burden.”).

Although this “step” is positioned by the district court as being distinct, there is no need for it to be so, especially if the majority is not necessarily going to apply a burden-shifting framework. If the burden remains with the plaintiffs, then it becomes difficult to see how this consideration of “alternative justifications” is distinguishable from the plaintiff’s affirmative responsibility to prove discriminatory partisan intent and dilutive partisan effect under the totality of the circumstances.

For example, the dissent points out that if the plan were drawn “with the intent of making it competitive,” a narrow loss in a tight race “would produce massive numbers of wasted votes (and thus inefficiency) for the loser.” Slip Op. Dissent at 152. This state interest in “competitiveness” would not be a separate “justification” apart from the burden of proof facing the plaintiffs on intent and effects. If the state adopted the plan to enhance competitiveness, then the requisite showing of intent would be lacking. If the efficiency gap is the result of close, competitive races, then the requisite showing of effect would be lacking (since the evidentiary value of the efficiency gap figure would be undermined and the dilutive effect would be largely nonexistent on both a statewide and district-by-district basis).

Similarly, the majority noted that one expert failed to incorporate the impact of incumbent-pairing prevention on the alternate maps. This is a legitimate state interest and, as the majority notes, the expert’s “lack of attentiveness to this concern well might diminish the Demonstration Plan’s worth as a viable, legislative alternative.” Slip Op. at 111. If the state interest were limited to “incumbent-pairing prevention”—and did not extent to “incumbency advantage”—then these explanations would undercut the plaintiffs’ case.

As discussed above, the intent element defined by the majority made a key distinction between legitimate justifications for districting decisions (i.e., the consideration of party in service of a legitimate state interest) and illegitimate justifications for districting decisions (i.e., the consideration of party in service of partisan advantage). These justifications are intertwined with and inextricable from the elements of the claim itself.

The Appeal

A direct appeal to the Supreme Court in this case is almost certain to occur. Here, the plaintiffs should reiterate, as they did before the district court, that the ultimate burden of establishing a coherent and viable test for adjudicating the violation of a constitutional right rests with the Court itself. Over the past thirty years, litigants have searched for various theories, standards, tests, and measures to help the Supreme Court find its way to vindicate voters’ rights. But only the Supreme Court itself can say what the proper constitutional test is.

Similarly, public interest organizations planning on filing in support of the district court’s decision should be clear in pointing out to the Supreme Court that this is not a normal situation where amici typically file briefs saying “yes, because.” Here, many amici are also likely to file briefs saying “yes, and” or “yes, or.” Without any set standard, the Supreme Court should be reminded that different approaches may not be mutually exclusive and may not undermine the district court’s ultimate holding.   For example, the court in Whitford v. Gill distinguished between legitimate state justifications, such as establishing a “fair” system of representation, and illegitimate state justifications, such as entrenching partisans. This is the same distinction found at the heart of the Harris v. McCrory appeal presently pending before the Supreme Court. A constitutional rule that prohibits states from diluting voters’ ability-to-elect with the intent of achieving a partisan advantage is entirely consistent with a constitutional rule that prohibits states from classifying voters for the purpose of achieving a partisan advantage.

Indeed, recognizing that these prohibitions are mutually reinforcing would actually address one of the dissent’s main arguments head on. The dissent rejected the idea “that gerrymandering can occur without distortions of district boundaries,” stating that such a “definitional novelty . . . flies in the face of Vieth.” Slip Op. Dissent at 128. Hardly. Here, the dissent simply fails to distinguish between the Court’s “dilution” gerrymandering cases and the Court’s “sorting” gerrymandering cases. While a sorting claim requires distortions to district boundaries, a dilution claim focuses on the intentional diminishment of the targeted voters’ effective franchise. This can be accomplished in the racial or political context without deviations from traditional, neutral districting principles. (Indeed, the dissent points to the Supreme Court’s line of racial sorting cases in an attempt to show that “there is every reason to believe that any political gerrymandering test Justice Kennedy might adopt would include the plan’s adherence to such principles as part of its analysis.” Slip Op. Dissent at 131. This may be true for a political sorting case, but the Supreme Court’s line of racial vote dilution cases do not require such deviations. As such, there is every reason to believe that a political vote dilution case need not adopt this approach.)

Admittedly, the majority seemed to dispose of both traditional equal-protection-clause bases for liability. Slip Op. at 47, 60. Yet, this is not necessary for the majority’s decision to hold true. It is not inconsistent to argue that Bandemer can be distinguished in any number of ways. The decision can be distinguished on a legal basis because of its failure to “consider the associational aspect of voting in assessing the gravamen of the harm.” Slip Op. at 54. The decision can be distinguished on a factual basis because of its failure to make any number of requisite findings, Slip Op. at 79, or its failure to “credit any [statistical] evidence” because “it did not ‘wish to choose which statistician is more credible or less credible,” Bandemer, 478 U.S. at 116 n.3. And, the decision—like the Vieth decision—can be distinguished for its failure to recognize that there is more than one type of gerrymandering harm in the political context just as there is more than one type of gerrymandering harm in the racial context. A dilution-based offense represents an impermissible burden separate and apart from a sorting-based offense, which represents an impermissible classification.

Far from being mutually exclusive, these arguments are mutually reinforcing and, in fact, necessary for analytical consistency. It would make little sense to hold that a state’s intent to burden citizens by political affiliation to obtain partisan advantage is unconstitutional while simultaneously holding that a state may expressly classify its citizens by political affiliation to obtain the same partisan advantage.

The Court’s failure to adjudicate such claims over the past thirty years has had a tragically predictable consequence: the nationwide deprivation of citizens’ equal protection and associative rights under the law due to discriminatory state treatment on the basis of political opinion, preference, and affiliation. Between Whitford v. Gill and Harris v. McCrory, the Supreme Court will soon have everything it needs before it to finally bring this shameful chapter to a close and to empower voters to vigorously defend their rights in federal courts across the nation. And that is something we can all be thankful for.