Why counting presidential votes is not for federal district courts
For the past several months, Election Law at Ohio State and SCOTUSblog have teamed up to track significant election-related lawsuits with the potential to reach the Supreme Court and affect the presidential election. Now, two weeks after Election Day, litigation over the outcome of the election is rapidly diminishing, but it hasn’t yet completely disappeared. Still scheduled for Tuesday, Nov. 17, is a hearing in the Trump campaign’s federal-court lawsuit seeking to delay certification of the popular vote in Pennsylvania. The remaining litigation almost certainly will not have a practical effect on the election’s outcome for a variety of reasons (it would be necessary for President Donald Trump to overturn Democrat Joe Biden’s apparent victory in three states, not enough ballots are at stake even assuming the merits of the legal theories raised, factual evidence is lacking on many of the claims made, and so forth). But as others have observed, there are still some legal principles at stake before this ends up just a matter for the history books.
In a recent Washington Post column, I set forth the basic reasons why litigation over the counting of votes in a presidential election belongs in state, rather than federal, court. If this procedural point is correct, as is also argued in an important amicus brief filed by prominent former GOP officeholders, the pending federal-court case that is the subject of tomorrow’s hearing should be dismissed without consideration of its merits, so that any such claims can be pursued in an appropriate forum established by state law. Here I elaborate on additional details in support of the basic point. Many of these are drawn from research for the book Ballot Battles: The History of Disputed Elections in the United States as well as work done for the American Law Institute’s Principles of Law—Election Administration project.
One can conceive of these points as adding up to the conclusion that federal courts should stay out of cases involving the counting of votes in presidential elections as a matter of equitable discretion — a court of equity historically always can decline to intervene, particularly on a request for a preliminary injunction or temporary restraining order, when the public interest as part of the “balancing of the equities” so dictates — rather than as a strict jurisdictional barrier to federal-district court review. Another way one could think of this is that vote-counting litigation in a presidential election warrants its own special form of an “abstention” doctrine, or at least yields the conclusion that traditional abstention doctrines as applied to this context calls upon federal district courts to abstain rather than getting involved. But whatever doctrinal label one wishes to attach to this conclusion, these factors combine to provide a strong basis for federal district courts refusing to intervene in the litigation over the counting of presidential ballots.
First, at the beginning of the 20th century, in a case involving allegations of ballot-box stuffing in a Kentucky gubernatorial election, Taylor v. Beckham, the Supreme Court invoked the political question doctrine to create a strict jurisdictional barrier to any federal court consideration of 14th Amendment claims arising from the vote-counting dispute. The political question doctrine had been articulated prior to the adoption of the 14th Amendment in the Rhode Island case of Luther v. Borden. Chief Justice Melville Fuller’s opinion for the Court in Taylor v. Beckham extended the prohibition on Article III federal courts becoming involved in this kind of fight over the outcome of an election to claims resting on the 14th Amendment. The court’s opinion is explicit on this jurisdictional point: “We must decline to take jurisdiction on the ground of deprivation of rights embraced by the 14th Amendment.”
Second, the jurisdictional barrier imposed by Taylor v. Beckham was well settled by subsequent precedents, including Snowden v. Hughes, to the point that Justice Hugo Black considered it straightforward when he granted an emergency order to deprive the federal district court of jurisdiction over the vote-counting dispute in the infamous case involving Ballot Box 13 in Lyndon Johnson’s race for the U.S. Senate in 1948. The dramatic story of this litigation is well chronicled by Robert Caro in his prizewinning volume, Means of Ascent, as part of Caro’s multi-volume biography of Johnson. But drama aside, Black considered the legal proposition involved mundane. As one of his biographers put it, “The legal principle involved was not complicated: federal courts are supposed to stay out of state elections.”
Third, this straightforward lack of federal-court jurisdiction in a vote-counting case is why Richard Nixon in the 1960 presidential election had no recourse to federal courts as an option for the possibility of overturning allegations of fraud in favor of the Kennedy-Johnson ticket in both Illinois and Texas. There has been much comparison in public discourse between this year and 2000, but not nearly as much public discussion of the possible comparison between this year and 1960, which might be the more apt comparison because Nixon would have needed to overturn the results in two states, not just one (whereas in 2000 the fight for the presidency focused on just a single state, Florida). When comparing Nixon’s position in 1960 to Trump’s position this year, Nixon would have had the much better case for claiming irregularities that might actually have made a difference, and yet there was no avenue for pursuing those vote-counting claims in federal court.
Fourth, the Warren court’s “reapportionment revolution” starting with Baker v. Carr and Reynolds v. Sims changed the jurisprudence. But until Bush v. Gore, the Supreme Court never had applied that new jurisprudence to a vote-counting case. Bush v. Gore, in its haste, failed to address the precedent of Taylor v. Beckham, which did involve vote-counting, and thus these two precedents — at the opposite ends of the 20th century — remain in conflict, which the court never has acknowledged or explained. The treatment of Taylor v. Beckham in Baker v. Carr is pretty flimsy and does not squarely overrule that precedent along the way to holding the redistricting claim at issue there justiciable. Thus, there is still more work for the Supreme Court to do in putting all of its relevant precedents together into a coherent whole.
Fifth, even accepting that Bush v. Gore implicitly overruled Taylor v. Beckham on the power of the Supreme Court itself to address 14th Amendment issues that arise in the context of a vote-counting dispute when reviewing a judgment of a state supreme court, there is the separate question of whether a federal district court should entertain jurisdiction over this kind of case. If the research for Ballot Battles is correct, it was not until 1994, in an election for the chief justice of the Supreme Court of Alabama, that a lower federal court ever took jurisdiction over a vote-counting dispute in a statewide election, in direct opposition to Black’s decree in the “Ballot Box 13” dispute in 1948. Thus, for the first two centuries of the republic, nothing like this federal-court intervention into a state’s vote-counting proceedings had ever happened in a statewide election.
Sixth, although the Bush campaign in 2000 went to federal district court in an effort to get relief, it did not prevail there. It needed to go to the U.S. Supreme Court on review from the Florida Supreme Court in order to get relief, as it did twice. That is how it should be, as it would contravene the premises of the Electoral Count Act of 1887 and its “safe harbor provision” — invoked by the Supreme Court in both of its 2000 Bush decisions — for a federal district court to interfere with the state’s own adjudicatory procedures, including its use of state courts, to resolve any disputes over the counting of votes in a presidential election.
Seventh, the Electoral Count Act was adopted before Taylor v. Beckham was decided, but it rests on the same basic premise of no federal court involvement in a disputed presidential election. Moreover, for the entire decade between the resolution of the dispute between Rutherford B. Hayes and Samuel Tilden in 1877 and the enactment of the ECA in 1887, Congress considered various institutional options for what should be the tribunal to resolve a ballot-counting dispute in a presidential election. They were informed by their experience of the Electoral Commission they created for the Hayes-Tilden dispute. The leading authors of the ECA served on the commission itself (Sens. George Edmunds and George Hoar) or were lawyers for Hayes arguing before the commission (Sen. William Evarts). They rejected another federal electoral commission as a model. They rejected federal judicial involvement in the key legislative debates on the ECA. They made a conscious and explicit decision to leave this kind of dispute to the tribunals of the states. They were thoroughly “states’ rights” in this post-Reconstruction philosophy, even if our own generation would reject that kind of “states’ rights” philosophy were we to amend the ECA. Consequently, a proper construction of the ECA requires keeping vote-counting disputes over presidential ballots out of federal district court, so that state courts can handle these cases pursuant to state law.
Eighth, as a practical matter, it’s worth considering just how dangerous to safe harbor status even well-meaning federal district judges might be. (To achieve the safe harbor of 3 U.S.C. § 5, a state must do two things: One, it must resolve all disputes over its popular vote for president, including any applicable lawsuits, by the specified deadline, which this year is Dec. 8; and two, it must resolve those disputes using rules and procedures that state law had in place before Election Day itself, Nov. 3. If a state achieves these two prerequisites, then Congress promises to treat as “conclusive” the state’s “final determination” of its popular vote, upon which its appointment of electors is based, when Congress meets to count the electoral votes on Jan. 6.) Tuesday’s hearing in the Trump campaign’s federal-court case in Pennsylvania is exactly two-fifths of the way through the entire five-week safe harbor period. There is not a lot of time to wrap things up.
When the American Law Institute’s Principles of Law-Election Administration project considered this topic, over a multi-year period and on a nonpartisan basis involving a group of advisers who included leading recount lawyers associated with both major political parties, the strong consensus judgment was that it was extremely difficult to fit in all of the state-law proceedings that may be necessary in time to meet the congressional five-week safe harbor deadline. The project even developed a model five-week calendar to demonstrate how, with careful planning, a state might be able to complete judicial review of the canvassing of returns, any recount that might be necessary, and any judicial contest of the election’s results, including all state-court appeals of these proceedings. There simply was no room within this tight five-week schedule for collateral federal district-court involvement, however well-intentioned, to delay or interfere with state-law steps. Any issues of federal law, like due process or equal protection, can be handled as part of the state-court proceedings, but there is no time for both state-court and federal-court litigation over the same ballots in dispute.
Think of the factual issues to be litigated. How many ballots are allegedly in the category of having been wrongfully counted because of an impermissible opportunity to “cure” that should not have been granted? This key fact is unclear to me from the Trump campaign’s brief seeking emergency relief in its federal case.
Pennsylvania law explicitly provides for a judicial contest of the result of an election for the office of “presidential elector” — after completion of the certification of the counting of the popular vote. When is this contest supposed to happen, in order to be complete by the safe harbor deadline — especially if the federal district court gets involved and delays certification, even for a “brief pause” as requested by the Trump campaign?
How long is the federal court going to take to do the factfinding required by the Trump campaign’s motion, assuming that there are any credible factual claims to pursue? And what about the inevitable appeal to the U.S. Court of Appeals for the 3rd Circuit? Does the judicial contest under Pennsylvania law need to wait until all this federal court litigation is resolved?
This is precisely the problem — except even worse because it is in federal court — that occurred in Florida in 2000. Because the Florida Supreme Court inappropriately rewrote the deadlines for the pre-certification “protest” phase of the process, there ended up being inadequate time for the post-certification “contest” phase before arrival of the safe harbor deadline. If the federal district court in Pennsylvania (or elsewhere) gets involved with any factfinding necessary to adjudicate an equal protection claim prior to certification of the popular vote, it’s going to prevent the ability of the state to conduct its own proceedings, including the kind of judicial contest envisioned in Pennsylvania law, as the state is entitled to do under the safe harbor provision. It is one thing for a state supreme court to mess up the deadlines and procedures set by state statute — and that might raise an Article II issue — but it would be quite another affront to the Article II power of state legislatures to determine the “manner” of appointing the state’s electors for a federal district court to mess up the state-law procedures for adjudicating any claims involving the counting of the state’s popular vote upon which the appointment of electors is based.
There is more that could be said in support of federal-court abstention in order to permit the state’s own legal system to achieve safe harbor status under 3 U.S.C. § 5. But the preceding points should suffice. Whether the barrier to federal-district court involvement in this kind of case is considered strictly jurisdictional or just prudential, as a matter of sound discretion given the nature of the relief requested in the circumstances involved — including the tight timeframe for meeting the safe harbor deadline — the answer should be crystal-clear: These claims should be litigated in state court, with any review if necessary in the U.S. Supreme Court, as in Bush v. Gore. It would be wholly without precedent in all of U.S. history, and entirely contrary to the fundamental premises of the Electoral Count Act itself (as well as the Article II premises on which that act rests) for a federal district court to insert itself into this kind of dispute.