Posted: April 26, 2012
Provisional Ballots, Consent Decrees, and the Balance Between the Federal and State Governments
The question of when and how to count provisional ballots is an important and ever-changing area of election law, with developments coming on a variety of fronts. The focus here is on Ohio, as there has been significant, continuing litigation regarding provisional ballots in that state. However, these issues are present throughout the United States. The Ohio litigation addressed below raises incredibly difficult issues, raising questions about the mechanics of state election law, the Fourteenth Amendment of the U.S. Constitution, the relationship between federal and state courts, the law of consent decrees, and more. I have attempted in this article to grapple with these issues in a fair and unbiased manner. Given the complexity of this problem, however, these matters are open to a variety of interpretations and this is just one approach. I hope, however, that this article can provide a useful starting point for a discussion about the future of provisional voting in Ohio and in the nation at large.
I. A Brief Background on the NEOCH Consent Decree
The NEOCH Consent Decree was issued by Judge Algernon Marbley of the United States District Court for the Southern District of Ohio on April 19, 2010 and is in effect until June 30, 2013. The Secretary of State of Ohio—at the time Jennifer Brunner—was a party to the litigation as a defendant, and the State of Ohio intervened as a defendant as well. The Plaintiffs had alleged that Ohio provisional ballot laws violated the Equal Protection and Due Process Clauses of the U.S. Constitution. However, the consent decree itself does not “constitute an adjudication or finding on the merits . . . nor [does it constitute] an admission by the Defendants [the Secretary of State and the State of Ohio] of any wrongdoing or violation of any applicable federal or state law or regulation.” Northeast Ohio Coalition for the Homeless v. Brunner, No. 2:06-cv-00896, at 2 (S.D. Oh. Apr. 19, 2010), available at http://moritzlaw.osu.edu/electionlaw/litigation/documents/NEOCH-Decree-4-19-10.pdf.
The consent decree requires provisional ballots to be counted if the provisional voter provided the last four digits of his or her Social Security Number and either of the following occurred:
•The provisional ballot was cast in the correct polling location but in the wrong precinct due to pollworker error
•The provisional ballot envelope or affirmation was filled out incorrectly based on instructions from the pollworker, or was not signed by the pollworker
Notably, the consent decree does not directly bind Ohio Boards of Elections. Instead, it requires the Secretary of State to issue a Directive to the Boards of Elections to have the Boards comply with the NEOCH consent decree. That Directive is binding on the Boards of Elections under Ohio law.
II. Potential Conflict with Ohio Law
Under Ohio law, there is no explicit exception to the ban on counting miscast provisional ballots for those provisional ballots that are miscast due to pollworker error. See Ohio Rev. Code §§ 3505.181(C), 3505.183(B). The Ohio Supreme Court reinforced this idea in State ex rel. Painter v. Brunner, 128 Ohio St.3d 17, 28 (2011). The Ohio Supreme Court further held in Painter that Ohio law presumes no pollworker error occurred when a provisional ballot is miscast and that the Ohio Secretary of State cannot order Boards of Elections to investigate provisional ballots for potential pollworker error.
Arguably, the Painter decision is not in conflict with the NEOCH consent decree. The decreedoes not require an investigation into pollworker error. Further, while Ohio law prohibits counting provisional ballots cast in the wrong precinct, the Ohio Supreme Court in Painter did not say that federal law could not require an exception due to pollworker error. The Ohio Supreme Court did note that it is only bound by the pronouncements of the U.S. Supreme Court with regards to federal law. Painter, 128 Ohio St.3d at 31. Even so, neither Judge Marbley nor Judge Susan Dlott in the Hunter v. Hamilton Cnty. Bd. of Elections litigation purported to hold that federal law required such an exception or that the Ohio statutes were unconstitutional at the time Painter was decided.
III. Effect of the Hunter Litigation
In her most recent decision in the Hunter v. Hamilton Cnty. Bd. of Elections litigation, Judge Dlott specifically declined to address a substantive Due Process argument that would have essentially held that the relevant Ohio statutes are unconstitutional due to their lack of a pollworker error exception for miscast provisional ballots. Judge Dlott’s rationale was that the Federal Rules of Civil Procedure require Plaintiffs challenging the constitutionality of a state statute to notify the State Attorney General and to have the State Attorney General certify that the statute has been challenged and intervene in the suit, something that had not occurred in Hunter. See Hunter v. Hamilton Cnty. Bd. of Elections, No. 1:10-cv-00820-SJD, at 82-85 (S.D. Oh. Feb. 8, 2012), available at http://moritzlaw.osu.edu/electionlaw/litigation/documents/ Hunterorder.pdf.
Following the issuance of this judgment, the Plaintiff did file the notice required by the Federal Rules of Civil Procedure. See Plaintiffs’ Notice of Giving Notice to Attorney General of the Unconstitutionality of Ohio Statute, Hunter v. Hamilton Cnty. Bd. of Elections, No. 1:10-cv-820 (S.D. Ohio Feb. 13, 2012), available at http://moritzlaw.osu.edu/electionlaw/litigation/documents/noticetoAG.pdf. Furthermore, prior to the Judge Dlott’s judgment discussed above, a Sixth Circuit panel cast serious doubt on the constitutionality of the relevant Ohio statutes: “Arguably, these . . . provisions operate . . . in a manner that is fundamentally unfair to the voters of Ohio, in abrogation of the Fourteenth Amendment’s guarantee of due process of law. . . To disenfranchise citizens whose only error was relying on poll-worker instructions appears to us to be fundamentally unfair.” Hunter v. Hamilton Cnty. Bd. of Elections, 635 F.3d 219, 243 (6th Cir. 2011). Even so, the Sixth Circuit did not rule on the constitutionality of the relevant statutes either. See id. at 244 (“It is premature, however, to decide a due-process challenge to Ohio’s election laws as they relate to poll-worker error . . . .”). As such, their statements about the constitutionality of the relevant statutes were pure dicta.
IV. The Arguments Against the NEOCH Consent Decree
In the Hunter litigation, the Hamilton County Board of Elections tried to attack the NEOCH consent decree on several grounds. The main thrust of their argument was that Article 1, Section 18 of the Ohio Constitution states that only the Ohio General Assembly can suspend or alter Ohio laws; the Secretary of State cannot. Therefore, the Board argued that because the Ohio General Assembly was not a party to the NEOCH litigation, the consent decree could not be used to change Ohio law. Judge Dlott dismissed these concerns, stating that the consent decree merely “restate[d]” Ohio law, it did not suspend it. She also pointed out that the State of Ohio, through the Ohio Attorney General, was a party to the NEOCH litigation and that the Ohio General Assembly was thus represented by extension.
In their request for a writ of mandamus from the Ohio Supreme Court in Niehaus v. Husted, the Republicans members of the General Assembly who are challenging the NEOCH consent decree are not pursuing the argument that the Ohio General Assembly was not a party to the NEOCH litigation and that therefore the consent decree cannot change Ohio law. The principal arguments made in Niehaus are that the Secretary of State exceeded her authority when she entered into this consent decree, since she does not have the power to alter Ohio law, and that the consent decree is invalid because it amends Ohio law in a way not provided for in the Ohio Constitution. Thus, the Niehaus Plaintiffs are asking that the current Secretary of State be restrained from complying with the consent decree and that the Secretary instead be forced to comply with the strict letter of Ohio law.
V. Arguments in Favor of the NEOCH Consent Decree
The Secretary of State has not filed a Reply in Niehaus as of this writing (the Secretary is required to file a Reply or a Motion to Dismiss by May 8), so it is unclear what arguments he will raise against those made by the Plaintiffs. The main argument against a writ of mandamus in this case is likely to be based on the Supremacy Clause. The Secretary entered into a consent decree issued by a federal court, and such an order—which is akin to federal law—trumps state law by virtue of the Supremacy Clause of the U.S. Constitution. A second argument is that the Ohio law may well be unconstitutional, as noted by the Sixth Circuit in Hunter, and that therefore the federal court had within its power the ability to amend or restate Ohio law in a way that complies with the Due Process Clause of the U.S. Constitution.
There are some flaws with these arguments. The first is that while federal law does trump state law by virtue of the Supremacy Clause, it is unclear what federal law requires here. Typically, a state supreme court is only bound by what the U.S. Supreme Court says federal law is, not what a U.S. District Court says federal law is. Furthermore, Judge Marbley never actually adjudicated the merits of the NEOCH case or issued any rulings on the merits; instead, he merely ratified what the litigants agreed on. Can that really be federal law? A final, related point, is that no federal court—whether in NEOCH or in Hunter—has found the Ohio statute to be unconstitutional under the Due Process Clause. If a federal court did make such a finding, it could then interpret the statute in a manner consistent with the Constitution (i.e. with a pollworker error exception, if indeed the Constitution requires such an exception) or strike it down, but that has yet to happen.
All of this being said, the NEOCH consent decree is still an order issued by a federal court and is therefore entitled to some deference from the Ohio Supreme Court. Even though the consent decree is not federal law in the sense that it does not tell us what the U.S. Constitution requires, it is federal law in the sense that the U.S. District Judge ordered the parties to comply with the contents of the decree and they are bound under federal law to comply with that order. Indeed, the U.S. Supreme Court has treated consent decrees as being akin to final judgments. See, e.g., Rufo v. Inmates of the Suffolk County Jail, 502 U.S. 367, 383-4 (1992). Therefore, a writ of mandamus by the state supreme court telling a party not to comply with what a federal court has required that party to do would in turn cause that party to violate federal law.
Given that the Ohio Supreme Court would be telling the Secretary to disobey a federal court, it would seem difficult for the Ohio court to reach such a result without explaining why such an order is not in conflict with the Supremacy Clause. It is also possible that the Ohio Supreme Court would not even need to address the question of whether the Fourteenth Amendment requires an exception in Ohio law for ballots miscast due to pollworker error. In the wake of a decision by U.S. Court of Appeals for the Third Circuit (discussed at greater length in Part V), it is unlikely that the Ohio Supreme Court can justify the writ of mandamus solely based on an argument that the Secretary of State lacked the authority to enter into the consent decree.
Therefore, it seems as though the Ohio Supreme Court would have to argue that Judge Marbley got federal law wrong and that therefore his order is not valid. The Ohio Supreme Court is well within its rights to take such an action—it is not bound by the judgments of the Sixth Circuit or U.S. District Courts on this issue, let alone Sixth Circuit dicta or implications made by Judge Marbley and Judge Dlott. It seems as though reaching that issue would be necessary, however, unless the Ohio Supreme Court can find another reason why the Supremacy Clause does not elevate a valid federal order over state law.
V. What Happens if the Ohio Supreme Court Issues the Writ of Mandamus and Orders the Secretary not to Comply With NEOCH?
The Ohio Supreme Court has demonstrated a willingness to challenge the role of federal courts in this area of election administration. In Painter, the court specifically stated that it would only be bound by pronouncements of federal law from the U.S. Supreme Court and paid no heed to the fact that its order in Painter was in some tension with what the U.S. District Court had ordered in Hunter, despite the fact that Painter and Hunter both arose out of the same election dispute. That being said, it is one thing for the Ohio Supreme Court to issue a decision that is in tension with a federal court order; it is quite another to order an individual not to abide by a federal court order. Therefore, Painter is not necessarily evidence that the Ohio Supreme Court will issue the writ of mandamus in this case.
But what would happen if the Ohio Supreme Court did issue the writ of mandamus and ordered the Secretary not to comply with federal law? Given the current Secretary’s vigorous opposition to the decision of the federal court in Hunter, he seems likely to obey the writ of mandamus and to stop complying with the NEOCH consent decree. Therefore, the NEOCH Plaintiffs would have to go to federal court in an attempt to have the consent decree enforced.
Their only option in this regard would probably be to go back to Judge Marbley. While the Ohio Supreme Court’s action might have some implications for the Hunter litigation, Judge Dlott’s decision was based on an Equal Protection violation, because some miscast provisional ballots were counted, but other miscast provisional ballots that were similarly situated—at least in the view of Judge Dlott and a majority of the Sixth Circuit panel—were not. The Hunter decision therefore did not truly depend on the need to have an exception for ballots miscast due to pollworker error and NEOCH was only relevant because the Board was attempting to comply with NEOCH when it counted some provisional ballots but not others. This, combined with the fact that during the Hunter litigation Judge Dlott declined to enjoin the Secretary from complying with the Painter decision, makes it seem unlikely that the NEOCH Plaintiffs could obtain relief based on Hunter.
Therefore, they would likely need to turn to Judge Marbley, which makes the most sense at any rate, given that he issued the consent decree. If Judge Marbley continues to believe that the parties need to abide by the consent decree, he has two options. First, he could enjoin the actual issuance the writ of mandamus so that it would never take effect. Generally, the Anti-Injunction Act (28 U.S.C. §2283) bars federal courts from enjoining state courts. There is an exception, however, where an injunction is necessary to “protect or effectuate [the federal court’s] judgments.” While the U.S. Supreme Court has not addressed the issue, every Circuit Court to address the issue has decided that “judgments” is broad enough to cover injunctions and other orders issued by a federal court. Therefore, Judge Marbley could take such an action. As a practical matter, however, this would seem to be a tough sell for the Plaintiffs, given that they would be seeking an injunction to enforce a settlement between the parties, rather than an actual ruling of some kind on what the law requires by Judge Marbley himself.
Judge Marbley’s second option, which seems a more likely response, is to issue an injunction preventing the Secretary of State from complying with the Ohio Supreme Court’s writ of mandamus and requiring him to abide by the terms of the NEOCH consent decree. This would put the Secretary in the unenviable position of having to decide which court he wants to hold him in contempt—the U.S. District Court or the Ohio Supreme Court. Given this reality, the Secretary would almost have no choice but to challenge the federal court injunction—and thus the validity of the consent decree—before the Sixth Circuit.
Per Sixth Circuit rules, the NEOCH appeal would first go to the same panel that has heard other appeals in the case: Judge Julia Gibbons, Judge David McKeague (both George W. Bush appointees) and U.S. District Judge Arthur Tarnow (a Bill Clinton appointee), who was sitting by designation in the case. Judge Tarnow has since taken senior status, so if he proved unavailable then the Clerk of Courts for the Sixth Circuit would draw a name from among the active judges to sit on the panel in his stead. This panel could decide whether it would hear the case, or whether the case should be assigned to a different panel chosen from among the active judges of the Sixth Circuit. 6 Cir. I.O.P. 34(b)(2). If the original panel declines to hear the case, therefore, it is impossible to know which judges would hear the appeal.
The issue then turns to whether the consent decree is valid. Consent decrees are a bit odd in that they have elements of both a contract and a court judgment. See Rufo v. Inmates of Suffolk Cnty. Jail, 502 U.S. 367 (1992). In Rufo, the Court created a two-prong test for modification of a consent decree: 1) “a significant change in circumstances warrants revision in the decree” and 2) the revisions “must be suitably tailored to the changed circumstances.” Id. at 383-84. This is somewhat reflected in Federal Rule of Civil Procedure 60(b), which the Supreme Court and the Circuit Courts have generally used in determining whether or not a consent decree should be abandoned.
FRCP 60(b) states that “the court may relieve a party . . . from a final judgment, order, or proceeding for the following reasons . . . (5) the judgment has been satisfied, released, or discharged; it based on an earlier judgment that has been reversed or vacated; or applying it prospectively is no longer equitable; or (6) any other reason that justifies relief.” (emphasis added). The Supreme Court has explicitly relied on these provisions, particularly 60(b)(5), in its consent decree jurisprudence. See, e.g., Horne v. Flores, 557 U.S. 433, 129 S.Ct. 2579 (2009).
Rule 60(b)(5) actually cannot be used to challenge the legal conclusions upon which the consent decree rested. Horne, 129 S.Ct. at 2593. However, the Court noted that consent decrees often raise “sensitive federalism concerns.” Id. Therefore, courts must apply 60(b)(5) in a flexible manner in order to avoid binding a state for years in a way that no longer makes sense due to changing circumstances or equities. See id. at 2594-95. The Court has also noted that consent decrees issued by federal courts “must be directed to federal interests.” Frew v. Hawkins, 540 U.S. 431, 437 (2004). Finally, “‘principles of federalism . . . require the [district] court to give significant weight’ to the views of [state] government officials.” Id. at 442 (quoting Rufo, 502 U.S. at 392 n. 14).
The Sixth Circuit has recently addressed some consent decree issues as well. The Sixth Circuit reviews district court decisions on whether to modify or vacate a consent decree for abuse of discretion. See U.S. v. Wayne Cnty., 369 F.3d 508 (6th Cir. 2004). In John B. v. Goetz, 2010 U.S. App. LEXIS 25589 (6th Cir. Dec. 16, 2010), the Sixth Circuit vacated part of a consent decree because the statute relied on by that portion of the decree was not privately enforceable. This was the result of a change in the law, however, as the non-existence of a private right of action had been established by subsequent court decisions. Similarly, the Sixth Circuit invalidated a consent decree that was based on a district court’s misunderstanding of federal law in Doe v. Briley, 562 F.3d 777 (6th Cir. 2009). As in Goetz, the original consent decree had relied on a U.S. Supreme Court case that the Court then rejected years after the consent decree was issued; that change in the law justified vacating the consent decree based on 60(b)(5).
In the Supreme Court’s decision in Horne, the consent decree was issued to remedy a violation of federal law and the argument was over whether that violation had been corrected, thus rendering the consent decree unnecessary. The facts of that case are not particularly applicable to the Niehaus litigation in Ohio. In Frew, the consent decree at issue required states to follow certain procedures in applying federal Medicaid laws. The Court conceded that federalism required that the views of state officials be considered and that “enforcement of consent decrees can undermine sovereign interests . . . .” Such concerns, the Court noted, were “legitimate” because if a consent decree went too far, it “may improperly deprive future officials of their designated legislative and executive powers.” Frew, 540 U.S. at 441.
That reasoning seems much more applicable to the issues and concerns raised in the Niehaus litigation. In Frew, however, the Court reversed a Circuit Court decision that had held the consent decree unenforceable on the basis of sovereign immunity. The Court determined that sovereign immunity alone could not block a consent decree because the Supremacy Clause meant a federal order like a consent decree trumped the state’s sovereignty. The decision in Frew was unanimous.
However, Frew was about state officials’ compliance with federal law. The NEOCH consent decree is about state officials’ compliance with state law, which only heightens the federalism concerns. That being said, all of the jurisprudence still points to needing some sort of change in circumstances or change in law in order to vacate a consent decree; as the Supreme Court has noted, a litigant cannot challenge the legal conclusions of the consent decree absent some change in facts, law, or equities.
Arguably, the equities in this case lie on the side of the NEOCH consent decree, as it seeks to protect voters from mistakes made by pollworkers, though there is certainly an equitable argument to made about a Secretary of State—who is elected as the nominee of a political party—binding future Secretaries, who may be of different political parties, and preventing them from following their view of Ohio election law. That means some sort of change in facts or law would presumably be needed to vacate the NEOCH consent decree.
This case is odd because there is no federal law upon which the consent decree rests. It argues for a proposition of Constitutional law that no court has yet espoused directly. Presumably a change in law justifying the vacating of the decree could come if, for example, a court held that the U.S. Constitution does not require a pollworker error exception. So far no court has taken this step. Furthermore, this would presumably have to happen outside the context of the litigation over the consent decree itself, since there would be no change in the law prior to the appeal of the consent decree.
The issue is unlikely to come up in the appeal of the Hunter litigation, which may leave state court as the only viable option for a change in the law in the foreseeable future, as no other federal court litigation on this issue appears to be pending at the moment. The Niehaus litigation is unlikely to result in such a change in the law justifying the vacating of the NEOCH consent decree. First, a holding by the Ohio Supreme Court that the Secretary of State lacked authority to enter into the consent decree is likely ineffective to vacate the decree.
In Delaware Valley Citizens’ Council for Clean Air v. Pennsylvania, 755 F.2d 38 (3d Cir. 1985), cert denied 474 U.S. 819 (1985), the Pennsylvania Supreme Court held that a state agency had no authority to enter into a consent decree and a lower court enjoined the agency from complying with the decree after that state supreme court issued that ruling. The State of Pennsylvania then went to federal court in an attempt to vacate the consent decree. The district court refused to vacate the consent decree and the Third Circuit affirmed, holding that federal courts have the power to determine whether litigants before them are able to enter in consent decree. See also AMTRAK v. Pa. PUC, 342 F.3d 242 (3d Cir. 2003) (reaffirming Delaware Valley Citizens’ Council’s holding that a state court cannot nullify a federal consent decree).
It does not appear as though a federal court has addressed the question of whether a state court decision on the meaning of federal law is the kind of change in federal law that would justify vacating the consent decree. Presumably it would not be, as a state supreme court’s decision regarding federal law would not be binding on a U.S. District Court. Therefore, it seems unlikely that the Sixth Circuit could justify vacating the consent decree on the basis of Niehaus.
What about the Ohio Supreme Court’s decision in Painter? The Sixth Circuit may be able to hold that the requirements of Ohio law were unclear when the consent decree was issued, but that the Ohio Supreme Court subsequently clarified Ohio law in Painter in a way that precluded the consent decree. This would be consistent with the aforementioned Goetz case, where subsequent court decisions held that a statute relied on in the consent decree was in fact inapplicable. The Sixth Circuit in that case ended up vacating the portion of the consent decree that was undermined by subsequent court decisions, and the Sixth Circuit panel could theoretically do that again in this case, holding that the Painter decision necessarily requires a reconsideration of the NEOCH consent decree by Judge Marbley. This is a bit different than those other Sixth Circuit decisions, however, where subsequent decisions by a federal court were enough to vacate the consent decree. It is unclear whether a state court’s judgment, even one relating to an issue of state law, is the kind of change in law needed to vacate a federal consent decree.
The Sixth Circuit could potentially duck that issue by noting that no federal court has yet reached this conclusion in connection with the Ohio statutes and that the NEOCH consent decree was not based on a trial and thus there is no record on which to determine whether the Constitution requires this pollworker error exception. Essentially, all the NEOCH Plaintiffs would have to rely on is the Sixth Circuit panel dicta in Hunter in terms of past precedent. However, there was plenty of briefing on the issue at the district court level and there would presumably be much more at the Circuit Court level, as well as oral arguments. Even so, the Sixth Circuit could perhaps vacate the consent decree on the grounds that it is not clear whether the decree reflects what federal law requires. Furthermore, as noted above, the Supreme Court has cast doubt on the idea that litigants can challenge the legal conclusions of a consent decree absent some subsequent change in the law.
If the Sixth Circuit vacated the consent decree, such an action would probably trigger the trial on the merits that was avoided in 2010 with the issuance of the consent decree. The result of this proceeding could be an actual judgment from Judge Marbley that the statute is unconstitutional, which in turn would likely trigger another appeal. This trial could theoretically be avoided if the Sixth Circuit decided the merits and declared that the U.S. Constitution does not require a pollworker error exception, however.
On the other hand, if the Sixth Circuit panel sustains the consent decree, Secretary Husted will once again find himself caught between either obeying a federal court or obeying the Ohio Supreme Court. Perhaps the Ohio Supreme Court would back down, but otherwise the Secretary would have to hope that either the Sixth Circuit would overturn the consent decree in an en banc proceeding or, failing that, that the U.S. Supreme Court would settle the matter once and for all. The U.S. Supreme Court seems wary of wading into these disputes on election administration in the post-Bush v. Gore era, but as this particular scenario involves a direct conflict between a Circuit Court of Appeals and a state supreme court, the U.S. Supreme Court might be more amenable to hearing the case.
This situation presents a delicate issue and ultimately comes down to the balance between the federal government and that of the states. These issues are of incredible importance for American democracy and that importance is only heightened here, where issues of elections and voting are at stake. I hope that this article can serve as a starting point for a discussion about these important issues, which are so critical for the effective operation of our democracy.
The bottom line based on my analysis is as follows: the Ohio Supreme Court should not attempt to restrain the Secretary of State from complying the NEOCH consent decree, as such an order would be problematic under the Supremacy Clause. If it does, however, the fate of the consent decree is very uncertain, as it was not based on any federal precedent, despite strong hints from the Sixth Circuit panel in Hunter that the Ohio statute is unconstitutional without the kind of exception found in the consent decree. Even so, the lack of a change in federal law may mean that the consent decree should remain in place for the time being. In the meantime, litigation should be brought to directly challenge the Ohio statutes at issue in order to determine once and for all whether the Ohio statutes are constitutional. Whatever the answer to that question, it would be better than the reigning uncertainty that has helped precipitate this potential conflict between federal and state courts.