Posted: September 6, 2012
Non-Retrogression, Equal Protection, and Ohio’s Early Voting Case
The United States Court of Appeals for the Sixth Circuit has set an expedited briefing schedule in the Obama campaign’s case over early voting in Ohio. The state’s brief is due this coming Monday (9/10), with Obama’s response a week later (9/17), and the state’s reply (if any) the Friday of that same week (9/21). As this appellate process gets underway, I wish to make one observation about an innovative and intriguing aspect of the federal district court’s unexpected order, issued last Friday. (In separate development, the district court has ordered Ohio’s Secretary of State Jon Husted to appear at a hearing next Thursday (9/20) to explain his response to the court’s Friday order.)
The district court ruled that the state must restore for Ohio’s entire electorate the three days of early voting immediately preceding the traditional Election Day. These three days existed in 2008 and more recently, until taken away in 2011 by a convoluted series of legislative enactments (combined with some implementing directives from the Secretary of State). The district court did not base its ruling on the ground that these three days of early voting are constitutionally compelled. Rather, the court relied on the ground that the state had left open the possibility that these three days of early voting would be available only to military voters this year, and that the state did not have an adequate justification for differentiating among military and non-military voters in this way. (For further details on the court’s ruling, see my colleague Steve Huefner’s insightful analysis from the day of the district court decision.)
But critical to the district court’s conclusion was the fact that these three days had been taken away for non-military voters rather than added for military voters. In other words, the district court built into its constitutional analysis a kind of “non-retrogression” analysis that is usually reserved for statutory claims under section 5 of the Voting Rights Act, as Rick Hasen noted in his initial report of the decision. This case, however, was not a statutory case, since Ohio is not a jurisdiction subject to the non-retrogression requirement of the VRA’s section 5.
To see that non-retrogression was essential to the district court’s reasoning, one needs to consider only the hypothetical comparison in which the state decides to extend to military voters a benefit with respect to three extra days of early voting that previously had not been available to any voter. To set the baseline, therefore, suppose that at Time One for all voters the state offered five weeks of weekday early voting (but no weekends), coupled with no-excuse absentee voting, as well as the traditional Election Day, but no early voting on the Monday immediately before Election Day. There would be no doubt whatsoever that that situation would be constitutional. All voters would have ample opportunities to vote, much more than the longstanding tradition of just Election Day with absentee voting limited to those with a permissible excuse.
Suppose, then, at Time Two the only move the state makes is to open up the possibility of early voting just for military voters on the last three days immediately preceding the traditional Election Day. I don’t think that limited move would be considered unconstitutional, as it would be seen as justifiable by the special circumstances confronted by military voters (for instance, the possibility of an unforeseen last-minute deployment). In comparison to Time One, non-military voters would be no worse off at Time Two by this limited extra benefit given to military voters, and the differential treatment between military and non-military voters would seem no more constitutionally problematic than the different rules for absentee voting that currently exist between military and overseas absentee voting, on the one hand, and conventional “domestic” absentee voting, on the other.
Nor do I think it makes a difference, when thinking about this hypothetical move from Time One to Time Two, if the extra benefit that the state extends to military voters depends on the discretion of local election officials. In other words, we can imagine two different versions of the Time Two scenario. The first, which we can call Time Two-A, is a state rule that requires all local county boards of elections to make the three extra days of early voting available to military voters. The second, which we can call Time Two-B, is a state rule that permits, but does not require, local election boards to extend the three extra days of early voting to military voters. Each local board can decide whether within its community there might be enough military voters who might wish to take advantage of the extra opportunity to warrant the board making the necessary arrangements to accommodate the interest. Obviously, the extra benefit to the military voters in the Time Two-B scenario is much more limited, and contingent, than the one in the Time Two-A scenario. But I don’t think that would make the move from Time One to Time Two-B a violation of the federal Constitution. Instead, the very modest additional benefit to military voters, with no harm to non-military voters relative to their previous position in Time One, would still seem justifiable, even though the move to Time Two-B didn’t do very much to help military voters.
Thus, although the district court in its Friday order relied on the fact that the state left it to the discretion of local boards to decide whether to make the three extra days of early voting available to the military, that fact alone would not be enough to render the situation unconstitutional. (The discretion of the local boards, to be precise, is limited to the Saturday and Sunday of this three-day period.) It was essential to the district court’s conclusion that this situation was the result of the state having taken away those three days of early voting from non-military voters, not that the state merely extended a modest benefit to military voters that previously had not existed for any voters. Hence, the conclusion that some sort of non-retrogression principle was the key animating linchpin of the district court’s reasoning.
But where does this non-retrogression principle come from in a constitutional case, rather than a statutory one under section 5 of the Voting Rights Act? It is not as if the idea of non-retrogression is entirely unknown to federal constitutional law. Indeed, in 1998 law professors John Jeffries and Daryl Levinson co-authored an extensive law review article entitled, The Non-Retrogression Principle in Constitutional Law (86 Calif. L. Rev. 1211). There, they observed that the Supreme Court had employed the idea in a line of race discrimination cases, including one in which a state had withdrawn busing as a remedy for school desegregation when that remedy had been previously available. Jeffries and Levinson also explained that the then-recent decision of Romers v. Evans, in which the Supreme Court invalidated Colorado’s withdrawal of anti-discrimination protection previously available under state law to gays and lesbians, arguably rested on a similar non-retrogression rationale. That same rationale, it should also be noted, was the basis of the Ninth Circuit’s ruling earlier this year in the case involving California’s Proposition 8, which overrode the California Supreme Court’s recognition of same-sex marriages under state law. (The Prop 8 case is now at the Supreme Court, waiting for its decision on whether to grant review.)
But Jeffries and Levinson saw the non-retrogression principle as very much the exception, and not the rule, in federal constitutional jurisprudence. In most circumstances, as the authors explained, constitutional analysis of state law is not path-dependent. For example, if a state decides to tax riverboat gambling at a lower rate than racetrack gambling, the validity of that differential rate of taxation under the federal Equal Protection Clause does not depend on whether the previous situation was that both had been taxed at the same higher rate rather than both having been taxed at the same lower rate. In other words, it does not matter whether a more favorable tax rate had been taken away from the racetracks, instead of a more favorable rate been extended to riverboats. (This example is taken from an actual Supreme Court case, Fitzgerald v. Racing Association, upholding this differential rate of taxation.)
Nor did Jeffries and Levinson look kindly on the use of the non-retrogression principle in even a rare constitutional case. They saw it as a kind of procedural cop-out that masked substantive values that the Court should acknowledge explicitly. Even worse, they saw non-retrogression unmoored from normatively attractive substantive values as potentially a mere preference for tradition, enabling an activist judiciary to invalidate social experimentation that a democratic legislature legitimately should be entitled to undertake.
Perhaps, however, the non-retrogression principle deserves consideration in the specific context of voting rights jurisprudence under the Equal Protection Clause of the federal Constitution. After all, it is in the specific area of voting rights that the non-retrogression principle plays its powerful statutory role, under section five of the Voting Rights Act. (And, without getting too technical, one can understand Justice Kennedy’s opinion interpreting the separate section two of the VRA in the LULAC Texas redistricting case as importing something of the non-retrogression idea into the vote dilution context—Latinos there were losing an opportunity that they previously had, and that fact seemed central to Justice Kennedy’s reasoning.) Moreover, voting has long been recognized as being a fundamental interest for purposes of Equal Protection analysis, and thus arguably whenever the government cuts back on the availability of voting opportunities for some group of citizens, the judiciary should look upon that cutback with heightened scrutiny. Simply put, given the centrality of voting to the democratic process, there are reasons for the courts to be more suspicious when the government curtails voting opportunities than when the government extends them.
Even so, just because the government cuts back on voting opportunities does not make the government’s conduct unconstitutional. If Ohio had eliminated the three days of early voting for all voters, military and non-military alike, there would be no Equal Protection problem (assuming this elimination was not motivated by an improper racial animus, something that has not been alleged in this lawsuit). Moreover, if Ohio had done a better job legislatively in deciding to eliminate the three days of early voting for non-military voters, while guaranteeing their continued availability for military voters (and not leaving it to the discretion of each local election board), that differential rollback would have been justifiable. That point is what I heard the Obama campaign’s counsel to acknowledge at the oral argument in the case. Indeed, the district court appears to have acknowledged the same point, since it is only because the rollback of early voting for non-military voters was combined with the contingency of early voting for military voters being dependent on the discretion of local officials that the current situation is unconstitutional according to the district court.
But was it appropriate for the district court to combine in this way, first, the non-retrogression principle with, second, the point about local discretion? Neither factor, by itself, would be enough for a conclusion of unconstitutionality. Together, do the two of them suffice? Put differently, while the existence of retrogression might be a reason to elevate the level of judicial scrutiny, is the mere fact that the state made its limited benefit for military voters discretionary, rather than mandatory, enough to cause the state to flunk that heightened scrutiny?
As this case moves to the federal court of appeals, I worry about Equal Protection jurisprudence being sufficiently indeterminate to permit different appellate judges to see the case differently depending on their political point of view (a concern that Rick Hasen has also expressed). There are liberal Democrats on the Sixth Circuit, and also conservative Republicans. In 2008, they were deeply divided over an important election case from Ohio, Ohio Republican Party v. Brunner, and needed a unanimous Supreme Court to rectify their disturbingly partisan treatment of that case. It would not be good if liberal Democrats on the Sixth Circuit find an Equal Protection violation in what Ohio has done with early voting, whereas the conservative Republicans on the court do not see one. Therefore, while the non-retrogression idea might be a worthy one to consider in the context of Equal Protection jurisprudence applicable to voting rights, I am not sure if it can carry the day where it depends on the additional distinction between discretionary and mandatory treatment of military voters on the part of local election boards.
Perhaps the non-retrogression idea can be coupled with another one: the suspicion that the rollback in voting opportunities was motivated by a legislative desire for partisan advantage, rather than any genuine policy justification (including any concerning the need of military voters). After all, the reason to elevate judicial scrutiny, as John Hart Ely famously put it in his classic book Democracy and Distrust, is to “smoke out” improper legislative motives. Therefore, if a rollback in voting opportunities can be shown to be motivated solely by the bare desire for partisan advantage, then perhaps it can be condemned as unconstitutional.
But even here a couple of cautionary notes are in order. First, the judiciary is reluctant to attribute partisan motivation to legislation that can be defended on policy grounds. A benefit to military voters, even a very minor and discretionary one, is susceptible to a policy rather than partisan explanation. Second, if the Sixth Circuit were to condemn Ohio’s current rules on early voting as the product of partisanship, not policy, it would seem doubly important that this condemnation come from a decision that did not itself appear partisan in motive. In other words, it would be unseemly to have only Democrats on the Sixth Circuit complaining that Ohio’s early voting rules were unconstitutional as a product of Republican bias.
Edward B. Foley is Director of the Election Law @ Moritz program. His primary area of current research concerns the resolution of disputed elections. Having published several law journal articles on this topic, he is currently writing a book on the history of disputed elections in the United States. He is also serving as Reporter for the American Law Institute's new Election Law project. Professor Foley's "Free & Fair" is a collection of his writings that he has penned for Election Law @ Moritz. View Complete Profile