Can a State Pack Black Voters in a District to Comply with the Voting Rights Act?

Steven D. Schwinn, John Marshall Law School

The Court will hear oral arguments on Monday in Whittman v. Personhuballah, a case testing whether a state's move to pack black voters into a congressional district supposedly to comply with Section 5 of the Voting Rights Act, but with the net effect of diluting black voters' influence, violates equal protection. This is the second time in two Terms that the Court has dealt with the issue: last Term the Court ruled in Alabama Legislative Black Caucus v. Alabama that the lower court applied the wrong standard and remanded the case for further proceedings. Whittman deals with a slightly different question, as described below. There's also a significant question of standing.

Here's my oral argument preview, from the ABA Preview of United States Supreme Court Cases, with permission:

Does a redrawn congressional district, which is based on maintaining a minimum fixed percentage of black voters in a district and in fact increases the percentage, violate equal protection?

Case at a Glance

The Equal Protection Clause forbids a state legislature from unjustifiably using race as the predominate factor in redrawing state legislative and congressional districts. At the same time, some states were required under the Voting Rights Act to ensure against retrogression, the diminution of a minority group’s ability to elect a preferred candidate of their choice. This means that a covered state had to consider race in redistricting. This case tests how a covered state can consider race.

INTRODUCTION

The Virginia legislature adopted a redistricting plan that increased the percentage of black voters in a majority-minority congressional district. The legislature based the plan on maintaining a fixed percentage of black voters in a certain congressional district, supposedly to comply with the Voting Rights Act. A three-judge district court struck the plan as a racial gerrymander and ordered the implementation of its own map.

ISSUES

1. Does a member of congress from an adjoining district have standing to challenge the court-ordered map, on the theory that the map may make it harder for him to win re-election?
2. Did the Virginia legislature’s use of race predominate when it drew congressional district 3, and, if so, was its use of race justified in order to comply with the Voting Rights Act?

FACTS

In 1991, as part of its redistricting plan after the 1990 Census, Virginia created its Third Congressional District, or “CD3.” The state created CD3 as its only majority-minority district, so that racial minorities in the district could elect a candidate of their choice. At the time, CD3 had a black voting age population, or “BVAP,” of 61.17 percent. The U.S. Department of Justice, or “DOJ,” precleared the plan and CD3 under Section 5 of the Voting Rights Act, “VRA.”

In 1997, however, a three-judge court invalidated CD3 as a racial gerrymander. The court in Moon v. Meadows, 952 F. Supp. 1141 (E.D. Va. 1997), found the evidence “overwhelming that the creation of a safe black district predominated in the drawing of the boundaries.” As a result, the General Assembly redrew the district, and lowered its BVAP to 50.47 percent. DOJ precleared the plan, and CD3 was not challenged.

After the 2000 Census, the state redrew CD3 again, along with its other congressional districts. As part of this redistricting, the state shifted a number of black voters from CD4 into CD3 and CD5. As a result, the BVAP in CD3 increased to 53.1 percent. DOJ precleared the plan, and CD3 was not challenged. (This plan is sometimes called “the Benchmark Plan,” because it immediately preceded the challenged plan and thus sets the benchmark against which the challenged plan is measured.)

After the 2010 Census, the state once again undertook to redraw its congressional districts. This time, CD3 was underpopulated by 63,976 citizens, so it needed additional citizens in order to reach the state’s benchmark population for compliance with the one-person-one-vote principle. Delegate Bill Janis introduced a plan that added population to CD3 and increased its BVAP from 53.1 percent to 56.3 percent.

Janis said that he based his plan on several criteria. These included the one-person-one-vote principle, the VRA rule against retrogression of minority voter influence, respecting the will of the Virginia electorate as reflected in the November 2010 elections, and maintaining current boundaries as much as possible. Throughout the floor debates on the plan, Janis repeatedly said that Section 5 of the VRA prohibited retrogression of minority voter influence, that compliance with Section 5 was “nonnegotiable,” and that compliance with the non-retrogression mandate was a “paramount concern” in drafting the plan.

The Virginia legislature failed to enact a plan in its 2011 special session. But Janis’s plan was reintroduced in the 2012 session (although Janis was no longer a member). At least two members of the legislature (Senators Locke and McEachin) protested that the plan packed black voters into CD3 and some surrounding districts, leaving them “essentially disenfranchised.” The House and Senate nevertheless passed the Janis plan, and the governor signed it. The plan maintained an 8-3 partisan division in favor of Republicans in the state and protected all incumbent members of congress. DOJ precleared the plan in March 2012.

In June 2013, in Shelby County v. Holder, 133 S. Ct. 2612, the Court invalidated the preclearance coverage formula in Section 4 of the VRA. This meant that Virginia (along with other previously covered jurisdictions) were no longer subject to the non-retrogression requirement in Section 5.

In October 2013, Dawn Curry Page, Gloria Personhuballah, and James Farkas, three voters in CD3, filed this case, seeking to invalidate CD3 as a racial gerrymander. Republican members of congress from districts surrounding CD3, including Representative Randy Forbes, Republican from CD4, intervened in the case to defend the plan.

The plaintiffs alleged that CD3 was designed to pack black voters in the district, which would dilute black voters’ influence in CD3 and in other districts. During trial, the plaintiffs called an expert, Dr. Michael McDonald, who testified that CD3 was drawn as a majority-black district for predominantly racial reasons. (McDonald based his conclusion in part on an Alternative Plan, produced by the plaintiffs, that resulted in a 50.1 percent BVAP in CD3. The parties disagree over the meaning of the Alternative Plan and whether it supports McDonald’s conclusion.) The state called its own expert, John Morgan, who testified that CD3 was explainable by race-neutral factors of politics and incumbency protection.

Importantly, evidence suggests the General Assembly applied a 55-percent-BVAP floor in drawing CD3. In particular, some evidence shows that at least some in the legislature thought that CD3 needed a 55 percent BVAP in order to pass DOJ preclearance under Section 5 of the VRA. (Remember, the state created CD3 before the Court ruled in Shelby County.)

The district court, by a 2-1 vote, concluded that CD3 was an unconstitutional racial gerrymander and enjoined the state from conducting any further congressional elections under the 2012 plan. The intervenors appealed, and the Supreme Court vacated the district court’s judgment and remanded the case in light of Alabama Legislative Black Caucus v. Alabama, 135 S. Ct. 1257 (2015). (Alabama Legislative Black Caucus involved the same kind of challenge to a similar plan, which also packed black voters into a district supposedly to comply with Section 5 of the VRA. The Court held that the lower court used an incorrect standard for judging this kind of case and gave some guidance for applying the correct standard. As discussed below, a portion of this ruling is relevant here.)

On remand, the district court again ruled 2-1 that CD3 was an unconstitutional racial gerrymander. The court wrote that the “legislative record here is replete with statements indicating that race was the legislature’s paramount concern in enacting the 2012 Plan,” and that the legislature had impermissibly used “a 55% BVAP floor” in redrawing CD3. (Because race predominated in redrawing CD3, the court applied strict scrutiny. The court held that compliance with Section 5 was a compelling state interest, but that the state’s use of race to increase the BVAP from 53.1 percent to 56.3 percent was not narrowly tailored to avoid retrogression in CD3. That’s because Congressman Bobby Scott, “a Democrat supported by the majority of African-American voters,” had been repeatedly reelected under the prior BVAP by large margins.) The intervenors appealed to the Court.

Meanwhile, the remedial phase of the litigation proceeded in the district court. The court invited the parties and any interested non-parties to propose plans and appointed Dr. Bernard Grofman as special master. Grofman rejected the proposed plans and recommended his own. The court found that the Grofman plan cured the racial gerrymander by redrawing CD3 according to neutral districting criteria, and not race. The court also found that the plan complied with the VRA, despite the drop in CD3’s BVAP (to 45.3 percent), because the lower BVAP combined with significant white-crossover voting preserved “African-American voters’ ability to elect the representative of their choice.”

The court-ordered plan also changed CD4. The plan increased the BVAP in CD4 from 31.3 percent to 40.9 percent, creating a “realistic possibility,” according to Grofman, that black voters could elect a candidate of their choice. The plan also increased Democratic representation (as measured by the election results for the 2012 Presidential election) from 48.8 percent to 60.9 percent, turning a “safe seat for the Republican incumbent” into a “competitive” district, according to Grofman. (The current and previous maps, and alternative maps, are available at the web-site for the Virginia Division of Legislative Services, http://redistricting.dls.virginia.gov/2010/RedistrictingPlans.aspx#41,list.)

The intervenors again appealed to the Supreme Court. The plaintiffs in the original case (minus Dawn Curry Page, who was dismissed by stipulation) defend the court-ordered plan as appellees before the Court. The Virginia State Board of Elections also defends the plan as an appellee. The government defends the plan as amicus in support of the appellees. The Court divided oral argument to permit each of the parties and the government to participate.

CASE ANALYSIS

The case involves two principal issues. Let’s take them one at a time.

Standing

In order to bring a case in federal court, a plaintiff must demonstrate (1) that he or she suffered an “injury in fact,” (2) that the challenged action caused the injury, and (3) that the lawsuit will redress the injury. The injury-in-fact requirement means that a plaintiff must show “concrete” and “particularized” harm, “actual or imminent.” The causation requirement means that the plaintiff has to show that the challenged action (here, the district court ruling and the court-ordered map) caused the injury. And the redressibility requirement means that the plaintiff must show that a successful lawsuit would redress their harm.

Standing is a threshold requirement. This means that the Court has to be satisfied that the intervenors have standing before it will rule on the merits (the discrimination claim, discussed below). Here, the parties focus particularly on intervenor Forbes. If Forbes has standing, then the Court will consider the merits. If not, the Court will dismiss the case.

The intervenors argue that they have standing, because the court-ordered plan transforms at least one of their districts (CD4, Forbes’s district) from a majority-Republican district to a majority-Democratic district. (Indeed, they say that every proposed plan would have made at least one Republican district a majority-Democratic district, so that they would have standing under any of the proposed plans.) The intervenors contend that this injures at least one of them, because it harms his or her chances for reelection, replaces his or her “base electorate” with “unfavorable” Democratic voters, and undoes his or her recommendations for the district. (The Board sides with the intervenors on standing and makes similar arguments.)

The plaintiffs argue that the intervenors lack standing. They claim that the intervenors have no responsibility for drawing or enforcing the 2012 redistricting plan (and therefore cannot complain that they were harmed by losing their redistricting power), and that they do not live in or represent CD3, the only challenged district. The plaintiffs assert that the intervenors’ only claim to standing is that the court-ordered plan might make it harder for some of them to win, if they choose to run, and if they defeat their primary challengers. The plaintiffs say that this alleged harm is too speculative and not sufficiently connected to the court-ordered plan. And in any event, they say, there are many other factors that might contribute to this harm. (The government sides with the plaintiffs on standing. The government adds that the intervenors have no right to “fence out those voters to enhance their odds of electoral success,” and therefore no harm when that happens.)

Discrimination

State legislatures can use a variety of factors in redrawing state legislative and congressional districts. Most of these factors are neutral—for example, preserving the compactness of a district, preserving the contiguity of a district, maintaining communities with like interests within a single district, and even advancing political interests—and do not alone raise constitutional problems. But the Equal Protection Clause prohibits a state legislature from using race as a factor, when its use of race predominates over other race-neutral factors without a sufficient justification (that is, without satisfying strict scrutiny).

States that were subject to the preclearance requirement in Section 5 of the VRA, including Virginia, had to consider race in their redistricting decisions. That’s because in order to obtain preclearance under Section 5, a covered state had to show that its new map, as compared to the immediately preceding map, would not result in retrogression, that is, diminishment of a minority group’s ability to elect its preferred candidate. (No state is subject to the preclearance requirement today. The Supreme Court in Shelby County struck the coverage formula for preclearance. This means that preclearance remains on the books, but currently there are no covered jurisdictions.)

This raises an important question: If a state legislature uses race in redistricting in order to comply with Section 5, does that use of race violate equal protection? The Supreme Court gave us some guidance to work that out last Term in Alabama Legislative Black Caucus v. Alabama. As relevant here, the Court held that Section 5 does not require a state to maintain a particular minority percentage; instead, it requires the state to maintain a minority’s ability to elect a preferred candidate of choice. This means that when a state legislature uses race as a predominate factor in redistricting in order to comply with Section 5, it cannot use a mechanical percentage—because that’s not what Section 5 requires.

That principle would seem to answer the question in this case (in favor of the plaintiffs and the Board), except that this case involves an additional wrinkle. Here, the intervenors claim even if the legislature used race as a predominate factor, CD3 would have looked the same if the legislature hadn’t used race. The intervenors rely on language from Easley v. Cromartie, 552 U.S. 234 (2001), to argue that because CD3 would have come out the same under neutral principles (without considering race), then race couldn’t have predominated, and the 2012 map satisfies equal protection.

The parties and amicus frame their equal protection arguments around these principles.

The intervenors argue first that the district court erred in finding that race predominated in drawing CD3 in the 2012 plan. The intervenors concede that race was a factor in the 2012 plan—that the legislature recognized that compliance with Section 5 of the VRA was “non-negotiable” and “paramount.” But they say that this use of race was necessary (because the state had to comply with Section 5), and that if it is considered predominant, then every use of race to comply with the VRA will automatically be deemed predominant. Moreover, they contend that the legislature’s racial goals were coextensive with neutral redistricting principles that governed all districts (like protecting incumbents by preserving the cores of existing districts) and with the legislature’s political objectives. They say that because the use of race resulted in the same district lines that would have resulted without the use of race, race cannot have predominated over neutral redistricting principles. Finally, the intervenors contend that the district court’s approach requires the legislature to treat majority-minority districts differently than majority-white districts, because under the district court’s approach the legislature could not use neutral principles to draw CD3, so long as the VRA also required the legislature to draw CD3 the same way.

Next, the intervenors argue that the district court failed to properly determine whether the legislature’s racial considerations subordinated other neutral redistricting criteria. They claim that the legislature would have drawn CD3 the same based only on neutral criteria, and so race could not have predominated. (They even say that achieving a 55 percent BVAP floor was the best way to achieve the legislature’s neutral redistricting objectives, irrespective of any racial purpose in using that floor.) The intervenors contend that the district should have determined whether there was an inconsistency between the neutral motives and the racial motives in order to determine whether racial motives predominated. But they say that the court never looked at this question.

Third, the intervenors argue that CD3 in the 2012 plan served permissible political purposes. They say that the 2012 plan treated CD3 the same as all other (majority-white) districts in the state, making only minor changes to district cores for the permissible purpose of benefitting incumbents. They claim moreover that changing CD3’s shape or reducing its BVAP would have sent a significant number of Democratic voters into the adjacent districts, all of which had Republican incumbents. Again, according to the intervenors, this means that CD3’s shape and BVAP serve the permissible purpose of benefitting incumbents. Finally, they contend that even the plaintiffs’ expert conceded that CD3 benefitted Republican incumbents and could be explained by a political purpose.

Fourth, the intervenors argue that the plaintiffs failed to show that the state could have achieved its political goals by drawing CD3 any other way. They contend that drawing CD3 with a BVAP of 56.3 percent was the only way for the state to retain all Republicans incumbents. They say that the plaintiffs’ alternative plan proves their point: this plan, which itself subordinated neutral redistricting principles to race, would have converted CD2 from a toss-up district with a Republican incumbent into a Democratic district, in order to achieve a lower BVAP.

Finally, the intervenors argue that CD3 in the 2012 plan meets the Alabama test. They claim that of all the alternatives, CD3 best advances the legislature’s political purposes and thus least subordinates those principles to race. In particular, they say that the court-ordered plan, with its reduction to a 30 percent BVAP, was not only based on race but likely would have failed DOJ preclearance. They contend that the legislature therefore had a “good reason” under Alabama to adopt the 2012 version of CD3.

The plaintiffs argue that race impermissibly predominated when the legislature redrew CD3 in 2012. They say that Janis, who originally introduced the plan, said as much, when he adopted the 55-percent-BVAP threshold based on a mistaken belief that any decrease in the BVAP would violate Section 5 of the VRA. They claim that circumstantial evidence shows this, too: the 2012 version of CD3 was the least compact district in the state, using water continuity to connect disparate black communities along the James River; it moved over 180,000 people to address underpopulation in CD3 of only 63,976; and the legislature disproportionately moved black voters into and white voters out of CD3.

The plaintiffs argue next that the intervenors are wrong to assert that the district court failed to apply Alabama—a legal error. Instead, the plaintiffs say that in truth the intervenors challenge the district court’s factual findings. But the plaintiffs contend that the intervenors cannot show that the district court’s findings were “clearly erroneous,” the standing for reversal on appeal.

The plaintiffs argue that the intervenors are also wrong to assert that the legislature’s use of race could not have predominated, because CD3 would have looked the same based on race-neutral redistricting criteria. The plaintiffs claim that if the legislature used race as a proxy for race-neutral criteria (as the intervenors seem to argue), then the legislature impermissibly used race. Moreover, the plaintiffs say that Comartie II is distinguishable: in that case, the direct evidence showed a partisan purpose, and the plaintiffs advanced a largely circumstantial case to prove otherwise; but in this case, the direct evidence (Janis’s statements) reveals a clear racial purpose behind the 2012 version of CD3.

Finally, the plaintiffs argue that Alabama supports their position. They say that the legislature made the same mistake as the legislature in Alabama, by focusing on how it could meet the arbitrary threshold of a 55 percent BVAP. But the plaintiffs argue that under Alabama the legislature should have focused on this question: “To what extent must we preserve existing minority percentages in order to maintain the minority’s present ability to elect the candidate of its choice?” Alabama, 135 S. Ct. at 1274. The plaintiffs argue that the intervenors try to escape the plain factual record, but they cannot: the record clearly reflects that race was the predominate purpose in drawing the 2012 version of CD3.

The Virginia State Board of Elections (as appellees) and the government (as amicus) make substantially similar arguments. The government clarifies a couple of points, however. First, the government says that mere statements that the legislature has to comply with the VRA does not mean that race predominated; instead, the legislature’s use of the 55-percent-BVAP threshold, along with other circumstantial evidence, means that race predominated. And the government, like the plaintiffs and the Board, argues that the court made no clear error in finding these facts. Next, the government contends, contrary to the intervenors, that a racial gerrymandering claim does not depend on a showing that race and politics conflicted. Instead, the government says that the constitutional harm comes from the predominate use of race, and that a plaintiff who can show that race predominated need not also show that the district’s actual configuration was different than an alternative configuration under race-neutral criteria. Finally, the government emphasizes that “Section 5 does not require jurisdictions to adhere to mechanical and factually unsupported racial targets, uninformed by a functional analysis of a minority group’s ability to elect.” And because that is exactly what the legislature did here, race impermissibly predominated.

SIGNIFICANCE

This is the second time in two years that the Court will consider a case of a previously covered jurisdiction packing black voters into a district supposedly to comply with Section 5 of the VRA. The Court limited this practice last Term in Alabama by ruling that (as relevant here) a state legislature cannot use mechanical percentages in order to comply with the VRA; instead, it must use the retrogression standard. That principle seems to answer the question in this case (because the Virginia legislature used a mechanical percentage), except that the intervenors claim that CD3 would have looked the same even without race—and therefore that race did not predominate in drawing CD3. Look for the Court to test this claim at oral argument.

The history of this case and the result in Alabama both suggest that the Court may lean toward the plaintiffs and the Board (and against the intervenors). First, the Court denied the intervenors’ application for a stay of the lower court’s decision pending appeal. Ordinarily, that would not (necessarily) suggest anything about the likely outcome. But in this case, Virginia has to know the shape of its districts, because it has to conduct 2016 elections. Because Virginia is set to run its elections based on the court-ordered map, the Court would throw a real wrench into Virginia politics by reversing course. (For one, it could affect Forbes himself. Forbes has announced that he will run for re-election in 2016 in CD2, not CD4, his current district. That’s because under the court-ordered map, CD4 leans much more Democratic. If the Court reversed the lower court, this could affect Forbes and the new representative in CD4, among others. It’ll be interesting to see if Forbes’s decision to run in CD2 becomes a factor for standing purposes.) Second, Alabama was a 5-4 decision, with Justice Kennedy siding with the progressives. This is a different case, to be sure, but it could turn on a similar line-up.

If the Court gets to the merits, look for it to rule narrowly. Given the Court’s approach in Alabama and given certain features of this case, this case seems an unlikely vehicle for the Court to make a grand statement about the constitutionality of the practice of packing districts for the supposed purpose of complying with the VRA, but with the effect of diluting the impact of all black voters in the state.

But that’s only if the Court rules on the merits. Indeed, we have good reason to think it might not, or at least that the Court will take the standing issue very seriously. The Court itself introduced the issue and ordered the parties to file supplemental briefs on it—twice. The first time, the Court simply asked the parties to brief whether the intervenors had standing. This apparently didn’t satisfy the Court, however, because it then asked the parties to brief whether they had standing because none of them lived in or represented CD3. These orders raise the real possibility that the Court could simply dismiss the case based on lack of standing, and not even address the merits.