In my lengthy high-level overview of the King-Klobuchar-Durbin “Electoral Count Modernization Act,” I looked at the federal causes of action, including, “extending popular elections” if (1) it’s beyond a reasonable doubt, that (2) a catastrophic event has prevented a potentially outcome-determinative number of voters or ballots (3) for the challenging presidential candidate. (It’s worth noting that this provision is not, strictly speaking, in the Electoral Count Act, but in the Presidential Election Day Act of 1845.) The time for voting extends throughout the state, for the “shortest duration necessary in light of the catastrophic event,” no longer than 14 days.
My initial reaction was that it was something within congressional authority. But after thinking about it, I have a lot of other, separate concerns about its practical application. In short, any time you create a novel federal cause of action, there are bound to be repercussions, and there there are many.
First, what happens to congressional elections? It seems a little odd to create a mechanism for extending Election Day in presidential elections but not congressional elections.
Second, can we separate presidential from non-presidential ballots in an extended election? Setting aside the congressional question, states run a lot of other state and local elections on their presidential election ballots. Would it be able to account for a system where some ballots come in “on time” under existing state and federal law, while others come in on an “extended popular election” basis? For many ballots, it’s easy. For others (say, those without a postmark that we rely on being received by a certain deadline), it becomes trickier. Feasible, but more complicated.
Third, does this preempt state emergency law deadlines? In Florida, for instance, the governor can “suspend or delay any election” in an emergency. In Utah, there are broad delegations to handle the times, places, and manner. Many states postponed primaries during Covid in 2020. (The work of Professor Michael Morley is extremely useful to see how some states handle these matters.) Does the new “time” designation simply replace the old one, per Ex parte Siebold (1880)? Even if the old one might be “better” for voters and candidates, or more flexible?
Fourth, if preempted, how does it interact with existing state emergency law deadlines in non-presidential elections? This becomes much more complicated when trying to separate presidential from non-presidential ballots, of course. But it could also set up very difficult situations where you can vote for President on Wednesday at Site X (if that’s what the federal court holds), but you have to vote for Senate on Thursday at Site Y (if that’s what state law requires). I don’t know how you try to administer two separate sets of emergency rules simultaneously–and, let’s face it, in an emergency, at precisely the moment when logistics will be most difficult.
Or if your statute allows for same-day registration on “Election Day,” as a defined term under state law, would a federal court necessarily fashion relief in such a way that addresses whether same-day registration should be extended to later days
Fifth, how does the law interact with existing constitutional law on election emergencies? There is no existing federal statute, but that’s not to say that the United States Constitution hasn’t been construed to apply to election emergencies. In RNC v. DNC, for instance, a federal district court would have essentially extended election day in the primary to account for Covid concerns. When the United States Supreme Court reinstituted the original deadline, it did not do so on the basis that the district court lacked authority under the Anderson-Burdick framework (while it remains good law) for doing so. Instead, it was on the grounds of Purcell, and that it was a novel remedy that the plaintiffs’ hadn’t originally sought. As the Court wrote, “By changing the election rules so close to the election date and by affording relief that the plaintiffs themselves did not ask for in their preliminary injunction motions, the District Court contravened this Court’s precedents and erred by ordering such relief.”
In other words, an appropriate constitutional remedy in the right case might be extending an election deadline. It’s also possible to conceive of other existing constitutional causes of action (perhaps a robust Equal Protection violation if voters are disparately treated across the state because of a disaster).
But, these forms of action would not be constrained by the limitations in the draft ECMA. Unlike the bill, any litigant (voters, interest groups, etc.) could file an action, not simply presidential candidates. Anyone could intervene under existing law, not simply presidential candidates. Any appropriate forum could be used, not simply where the state capital is. Remedies might extend beyond those enumerated in the draft ECMA.
In other words, you might not simply have competing rules for presidential and non-presidential elections occurring simultaneously; you might have competing orders from federal courts on presidential elections occurring simultaneously. That’s particularly likely if most prospective litigants are blocked from litigating under this provision.
Sixth, how does the statute work simply as a matter of statutory interpretation? There are some uncertainties in the language, which I noted earlier (e.g., language like “potentially” but simultaneously “beyond a reasonable doubt”). But, for instance, what does it mean that a catastrophic event has “prevented” voters from casting ballots? Is it the event itself that has prevented the voters (e.g., the hurricane prevented voters from casting a ballot at the polling place because they could not leave their homes because of flooding in their street), or is it that the voters acted on their own in reaction to the event (e.g., the hurricane prevented voters from casting a ballot at the polling place because they did not venture outside because of their own sense of safety), or is it that election administration was not available to voters (e.g., the hurricane prevented voters from casting a ballot at the polling place because the polling place was closed due to the hurricane). How do we determine which counties’ voters to include in an “outcome-determinative” tally? It’s a post hoc assessment of an “emergency” by a federal court, not a determination of those jurisdictions that have been declared emergency. What if there’s a statewide emergency declared by the governor, but after the election only some counties meet the statutory definition? What if voters fled in anticipation of a hurricane, but the hurricane didn’t actually hit in those counties? These are actions that must be filed–and proof that must be introduced–within hours of the election, and they are not terribly obvious terms.
All this is to say, it’s within Congress’s power to dictate the time of holding presidential elections. There are likely some other peripheral concerns, too, like how it fits with campaign finance deadlines that require certainly disclosures within a period of time around Election Day (which may result in some over-disclosure and under-disclosure), or with state laws that set up timelines for certification of results or deadlines for recounts (that would likely need to be bumped back in some states). But it’s worth discussing how even a seemingly small change to create a federal cause of action for disasters has significant ripple effects in how we administer elections in our federalist system.