SCOTUS Voting Rights Case Foreshadows "Evil Day"
by Michael C. Dorf
In his concurring opinion last week in Merrill v. Milligan, Justice Kavanaugh, joined by Justice Alito, said that the 5-4 order staying the lower court's decision that Alabama's new congressional map violated the Voting Rights Act (VRA) was primarily based on timing, not the underlying merits. The concurrence (and thus presumably the majority) relied on the so-called Purcell principle, which generally proscribes judicial intervention too close in time to an election.
As Justice Kagan's dissent explained, however, the invocation of the Purcell principle was badly flawed in the Alabama VRA litigation. The plaintiffs and the district court acted with alacrity and a new VRA-compliant map could be drawn lickety-split, thereby allowing adequate time for primary campaigns and elections.
Indeed, I would go further to say that the Purcell principle itself should be renamed the avocado principle. The delicious but temperamental fruit (technically a berry!) from which guacamole and an expensive spread for toast can be made is unripe for days, then edible for about fifteen minutes before it turns rotten. So too, election litigation as constrained by SCOTUS doctrine is not ripe until it runs afoul of the Purcell principle.
Yet while Merrill v. Milligan is thus chiefly an exemplar of the Purcell principle's shortcomings, it also holds dark hints about what the Court could do to the VRA and to antidiscrimination law more broadly.
But wait! The Kavanaugh concurrence expressly disclaims a view on the merits: "The stay order is not a ruling on the merits, but instead simply stays the District Court’s injunction pending a ruling on the merits." So why do I think it portends evil? Because even though the Court does not rule on the merits, it does take account of the merits.
Ordinarily, Justice Kavanaugh says, SCOTUS will grant a stay of a lower court decision when, among other things, the party seeking the stay shows "a fair prospect that the Court would reverse." However, he goes on to say, election litigation is somewhat different because of the Purcell principle. That principle, he says, does not utterly preclude judicial relief close in time to an election, but it does change the calculus of the merits. Kavanaugh writes that instead of needing to make the ordinary showing of "a fair prospect" of relief, a petitioner seeking a stay of a lower court order altering voting procedures on the eve of an election need only show that "the underlying merits are [not] entirely clearcut in favor of the plaintiff."
As the dissenting opinions by Chief Justice Roberts and Justice Kagan show, the petitioner (the Secretary of State of Alabama) failed to meet even that incredibly slight burden. Roberts writes: "the District Court properly applied existing law in an extensive opinion with no apparent errors for our correction." The Kavanaugh concurrence does not point to a single error in the (three-judge) district court's long and thorough opinion. Instead, he cites a statement in the Roberts dissent acknowledging that there is some confusion in the Court's VRA precedents, without any indication that that confusion led to an error in this case.
The real substantive basis for the majority's action appears to be encapsulated in the suggestion that if the VRA is construed to require the drawing of two rather than one majority-minority congressional districts when two districts would lead Alabama's congressional delegation to closely mirror the proportional racial population of the state as a whole, that could violate the Equal Protection Clause. To be clear, given the very light substantive burden that Justice Kavanaugh says that Alabama bears--showing merely that the plaintiffs' victory is not "entirely clearcut"--Merrill v. Milligan cannot be fairly understood as holding that the Equal Protection Clause precludes interpreting the VRA to mandate drawing enough districts to provide roughly proportional representation for racial minorities. But given the failure to identify anything else that might be wrong with the district court opinion, last week's stay appears to foreshadow such a ruling when the Court does decide the merits.
We can now expect the Supreme Court's supermajority to say that the VRA should not be construed to require two majority-minority congressional districts in Alabama simply due to the fact that two out of seven districts would bring the representation of Black voters in rough proportion to their share of the state's population. Yet that would work a substantial change in the law.
As the district court opinion explained effectively, redistricting operates under two constraints: (1) a line of SCOTUS equal protection precedents that forbids consideration of race as the predominant factor in drawing district lines; and (2) the VRA's prohibition of vote dilution or, as Section 2 puts it, the drawing of district lines so that members of a class of citizens defined by race or color "have less opportunity than other members of the electorate to participate in the political process and to elect representatives of their choice." The plaintiffs credibly alleged intentional racial discrimination--Black voters were "cracked" among six white majority districts and "packed" into one majority-Black district--but the three-judge district court did not rely on and, under the leading SCOTUS precedent of Thornburg v. Gingles, was not required to rely on, any finding of intentional racial discrimination. It was sufficient to make out the VRA claim that the Alabama legislature, which was obviously aware of what it was doing, drew district lines that had the effect of diluting minority voting strength in Alabama.
Why then, would Justices Kavanaugh, Gorsuch, and likely other members of the SCOTUS conservative bloc think that the Alabama case presents a possible clash between the VRA--if construed to require two rather than one majority-minority districts--and the Equal Protection Clause? The most plausible explanation is that they wish to overrule Gingles and hold that the effect of diluting minority voting strength cannot justify express consideration of race in districting, even if such consideration is only one factor among others, such as compactness, contiguity, and preservation of political units.
Does that possibility sound familiar? It should. In the Harvard and University of North Carolina affirmative action cases in which the Court recently granted certiorari, the plaintiffs seek a decision overruling Bakke and Grutter, which allow the use of race as a plus-factor. In the higher education context, conservatives have long argued for a principle of color-blindness in which race may be considered by a government actor only as a remedy for adjudicated race discrimination. So too here, Justices Kavanaugh and Alito appear to be signaling that likewise in the VRA context, use of race--even as a mere plus factor alongside traditional districting criteria--is impermissible absent an adjudication of intentional discrimination. Such a ruling would have far-reaching consequences in at least three areas.
First, as I've just noted, receptivity to Alabama's argument in the VRA case signals receptivity to the plaintiffs' argument in the Harvard and UNC cases for the elimination of all race-based affirmative action except perhaps in the very rare case in which a college or university has recently been found to be engaged in intentional discrimination against Black and other disadvantaged minority applicants.
Second, in principle, the impact should be modest in VRA cases themselves. After all, as I noted above, there was powerful evidence in the Alabama litigation that the legislature drew district lines with the intention of diluting Black voting strength. However, the practical impact could be much more serious. Proving intentional discrimination means proving that the government actor adopted the challenged plan because of, not merely in spite of its disparate impact on the plaintiffs. It is open to Alabama and other state legislatures to say that while they were aware of the impact on Black voters, they were really only aiming to increase the number of Republican seats in the state's delegation and that Black voters were thus "cracked" because most of them are Democrats, not because of their race. That will often be false. Often the demographic data on which legislative map drawers rely will include racial information. Using race as a proxy for political affiliation is forbidden race discrimination, but proving that this was what the legislature did in any case could be challenging, given that the Court has held that challenges to political (as opposed to racial) gerrymandering are not justiciable.
Third, if I'm right about the bread crumbs dropped by Justices Kavanaugh and Alito, we could be approaching what Justice Scalia, concurring in Ricci v. DeStefano, called "the evil day on which the Court will have to confront the question: Whether, or to what extent, are the disparate-impact provisions of Title VII of the Civil Rights Act of 1964 consistent with the Constitution’s guarantee of equal protection?" If "mere" disparate impact on Black voting strength doesn't suffice to require remedial action, then the same logic could imply that mere disparate impact in the workplace doesn't suffice to require remedial action.
To be clear, none of the foregoing is foreordained. It's possible that the Court will ultimately reaffirm Gingles and rule for the plaintiffs in Merrill v. Milligan. Perhaps last week's ruling really was only an exemplar of the Purcell/avocado principle. It's also possible that the Court might rule for Alabama on some ground that does not have broader implications. But there are certainly signs pointing to a gathering storm--a hyper-conservative Court that sees the Fourteenth Amendment's Equal Protection Clause chiefly as a vehicle for protecting white-dominated institutions against justice and equality claims by disadvantaged racial minorities.