Main points: (1) Ohio’s elimination of Golden Week has some differences than NC’s elimination of same-day registration throughout all of early voting; and (2) NC’s reversion to a rule disqualifying out-of-precinct ballots has an outright disenfranchising effect lacking in Ohio’s case, but present in Wisconsin’s.
Of the three election administration cases that have made it to the U.S. Supreme Court in the past couple of weeks—Ohio’s, Wisconsin’s, and North Carolina’s—I find North Carolina’s the most difficult. I can see the Court granting a stay, as it did in the Ohio case, because the Fourth Circuit made a late change in North Carolina’s voting procedures. Indeed, the Fourth Circuit’s October 1 ruling, which reversed the district court’s denial of a preliminary injunction, came almost a month after the Ohio preliminary injunction, entered on September 4. Given the Supreme Court’s previous admonition against such last-minute judicial disruption to a state’s voting procedures, the Fourth Circuit’s October 1 order would appear particularly vulnerable, and indeed Judge Motz (a Clinton appointee) dissented from the Fourth Circuit’s ruling essentially because of this timing concern.
Like Judge Motz, however, I think the underlying merits of the case may be much trickier than the timing issue alone would suggest. The underlying merits, at least has the case has developed to this point in its preliminary injunction posture, primarily concern the application of section 2 of the Voting Rights Act to recent changes in a state’s voting procedures. The two rule changes that the Fourth Circuit preliminarily enjoined were (1) the elimination of “same day registration” during North Carolina’s early voting period and (2) the reinstatement of a requirement that a provisional ballot be cast in the voter’s correct precinct in order for any vote on the ballot to count.
The first rule change makes the North Carolina case appear, at least superficially, similar to the Ohio case. Ohio eliminated “Golden Week,” a five-day period at the beginning of early voting in the state during which it had been possible to register and vote at the same time. Despite this superficial similarity, however, there are factual differences that ultimately may prove important on the merits. Ohio’s “Golden Week” was an anomaly in the context of that state’s overall early voting process; during the bulk of early voting pre-registration had been required. Moreover, this anomaly had been the result of an accidental legislative oversight: the five-day overlap between the end of registration and the availability of absentee ballots had existed when absentee voting in Ohio had been a very limited, excuse-based system; then, when Ohio moved to no-excuse absentee voting and decided to permit in-person absentee voting (which the world calls “early voting”), Ohio suddenly realized that it had inadvertently created a “Golden Week” during which was possible to register and in-person absentee vote, without an excuse, at the same time. By contrast, North Carolina had made a deliberate policy choice to permit same-day registration throughout its early voting period. Thus, Ohio’s subsequent legislative decision to jettison an inadvertent anomaly, so that its entire early voting period is consistent with the state’s overall policy goals concerning registration and early voting, is arguably a very different kind of statutory correction than North Carolina’s decision to reverse course on the entirety of its same-day registration policy.
There is also the issue of partisanship. As I discussed previously, Ohio’s new early voting rules are derived largely from a recommendation of the bipartisan Ohio Association of Election Officials. Although my Moritz colleague Dan Tokaji sees the structure of OAEO as flawed insofar as it over-represents low-population rural counties and under-represents high-population urban counties, it nonetheless is a bipartisan body that negates the charge that Ohio’s new rules are a thoroughly partisan manipulation of the electoral process designed to rig the rules in one party’s favor. By contrast, I did not see any evidence in the North Carolina case to inoculate that state from the charge of blatant partisanship in manipulation of the voting rules.
The second rule change in North Carolina, concerning out-of-precinct ballots, presents the risk of outright disenfranchisement not associated with an elimination of same-day registration. Under the new North Carolina rules, any ballot cast in the wrong precinct will not count. Period. There is no way for the voter to validate any portion of that ballot, even with respect to statewide or countywide offices for which all precincts are eligible to participate in the election. Thus, a voter who mistakenly is told to vote a provisional ballot in the wrong precinct—it happens—will be disenfranchised with no remedy under state law. By contrast, a voter who registers in advance of the registration deadline will be able to vote a ballot that counts, either during early voting or on Election Day. Thus, the consequence of North Carolina’s rule change concerning out-of-precinct ballots is much more severe to any of the adjustments to early voting, either in North Carolina or Ohio. In this respect, the disenfranchising consequence of the new disqualification of out-of-precinct ballots in North Carolina makes that case more like the Wisconsin one concerning the addition of a new stringent voter identification requirement.
This fact alone does not make the new disqualification of out-of-precinct ballots in North Carolina a violation of section 2 of the Voting Rights Act—just like a new voter ID rule does not automatically violate section 2. But insofar as the new rule causes a disproportionate disenfranchisement of minority voters, the stakes are obviously much higher and directly implicate the primary concerns of the Voting Rights Act. The North Carolina case thus raises a key question about the proper interpretation of section 2. What more (if anything), beyond the discriminatory effect of disenfranchising more minority than nonminority voters, must a plaintiff show in order to prevail on a section two “vote denial” claim?
North Carolina is hardly the only state that disqualifies out-of-precinct provisional ballots. Florida also does, as does Ohio (except to the extent constrained by Sixth Circuit precedent concerning “right church, wrong pew” ballots cast in the correct multi-precinct polling location). But, unlike North Carolina, neither Florida nor Ohio enacted a law that permitted the counting of out-of-precinct ballots before reverting back to a practice of disqualifying them. Assuming that Florida’s and Ohio’s rules disproportionally disenfranchise minority voters as much as North Carolina’s, are they equally unlawful under section 2 as North Carolina’s? Or is the reversionary rule change a relevant factor in calculating liability under section two? If so, then does not section 2—at least to that extent incorporate a kind of anti-retrogression principle—despite protests to the contrary?
These are vexing questions for which there are no obvious answers, which is why I find the North Carolina case particularly difficult on the merits. It raises, to my mind, profoundly important questions about the proper interpretation of section 2, a hugely significant statute to our national identity as a democracy committed to equal voting rights. It may be surprising that such an iconic statute about to celebrate its fiftieth birthday lacks a clear standard for determining liability under its most important provision. Yet that is the situation today. It will be necessary for the Supreme Court to clarify the standard, either in the North Carolina case or one like it.
But the time for that clarification is still down the road a bit. Not now, during the flurry of these emergency stay petitions. Meanwhile, like many others, I will watch the development of the records in the pending cases, including North Carolina’s, as well as the development of the briefing on the merits of what that clarifying standard should be. Maybe any state that disqualifies out-of-precincts ballots violates section two if that disqualification causes disproportional disenfranchisement of minority voters, and it does not matter whether this disqualification rule is a new, old, or reversionary law? Maybe this kind of discriminatory disenfranchisement violates section 2 only when it is the consequence of new or reversionary laws, but longstanding unchanged rules are somehow protected from section 2 liability, at least absent some showing of racially discriminatory intent? Or maybe liability hinges in part on the extent to which a state has been a bad actor in the past, so that North Carolina’s history of racial discrimination will condemn current practices for which there is no showing of intentional discrimination? If so, then the same disproportional disenfranchisement of minority voters may be permissible under section two in a northern state, like Massachusetts or Michigan, which lacks the same segregationist history as a southern state, like Florida or North Carolina. Although this interpretative approach would seem to run counter to the Court’s recent decision in Shelby County, perhaps it is an option available under section 2 even though it was not under sections 4 and 5.
North Carolina might end up being decided as a special case insofar as it involves a package of rule changes that should be evaluated as a package rather than as a series of separate changes. The “totality of circumstances” inquiry that Congress explicitly built into section 2 would seem to permit, and perhaps even require, that sort of analysis. But at this juncture, I wish to leave my own thinking on this case as undecided and tentative, while watching how it and the others unfold.