A couple of weeks ago, I was privileged to testify before the Senate Rules Committee in support of the Electoral Count Reform Act of 2022. You can view my testimony (along with those of other witnesses and a video of the hearing itself) here.

I want to focus on one line of misunderstanding that’s gotten a lot of repetition in some media accounts: there’s only a “six day window” to litigate issues related to a certificate of ascertainment of appointment of election of presidential electors, which is not enough time.

This significantly misrepresents the timing of litigation under the ECRA. In the vast majority of cases, there is several weeks’ time for resolution of election disputes; in a very narrow band of very late-breaking disputes with little need for new factual development, there is a shorter window. And it’s tough to move that window around.

Let’s focus on the big dates for timing for a moment, as they exist right now under the Electoral Count Act:

November 2-8: Election Day (first Tuesday after the first Monday in November)

December 7-13: “Safe harbor” deadline [35 days after Election Day, 6 days before electors convene]

December 13-19: Electors convene (first Monday after the second Wednesday in December) [41 days after Election Day]

January 3: New Congress sworn in (fixed by the Twentieth Amendment) [15 to 21 days after the electors convene]

January 6: Congress convenes to count votes [18 to 24 days after the electors convene]

January 20: Inauguration Day (fixed by the Twentieth Amendment) [14 days after Congress convenes to count]

That first tranche of time, the 35-day window between Election Day and the “safe harbor” deadline, is where the bulk of election litigation takes place. The canvass, recount, administrative audit, and election contest happen here, as does post-Election Day litigation in state or federal court. In 2020, for instance, Georgia canvassed its votes, then held a risk-limiting audit, then did recount. And myriad litigation in Florida in 2000, in Ohio in 2004, in a few states in 2016, and in several states in 2020 also mostly took place in this window.

The ECRA makes two changes to this first window, one smaller, one bigger.

First, a smaller change: it pushes back the date the electors convene one day, to the first Tuesday after the second Wednesday in December.

Second, a bigger change: the “safe harbor,” which was something of an optional deadline, becomes a hard deadline for state executives to complete the certification of the state’s results.

In the ECA, this “safe harbor” was a deadline to bind Congress and treat as “conclusive” any results that came from a state, if they were resolved six days before the electors met. But, it appeared that state disputes could stretch well beyond that date–with some suggestion, then, that any extension might thwart the presumption of treating the votes as “conclusive” (a fear in 2000 in Florida, a point raised by some in Congress in 2004 about Ohio, and a source of persistent litigation strategy of 2020).

Under the ECRA, however, it’s a firm deadline. Certification must be done by then.

It’s actually less to put pressure on state executives, and more to put pressure on the rest of the state apparatus. The Hawaii recount of 1960 stretched into January. Lawsuits in 2020 were filed up through the counting of electoral votes on January 6, 2021. The lack of clarity incentivized protracted litigation or dilatory practices. It also suggested that the results of an election could be unknown even after the electors met, precisely the opposite of what the ECA hoped to achieve after the contested election of 1876.

States, then, under the ECRA, must resolve their elections 36 days after Election Day. Most of the time, this is entirely unremarkable. And, in fact, most of the time, certification of an election happens well before that day. Leaf through the certificates of ascertainment filed in 2020. Alabama certified its results on November 24. Delaware, November 18. Idaho, November 20. States set their own internal deadlines of when certification must take place.

And, obviously, smaller states, states with Election Day ballot receipt deadlines, states with fewer provisional ballots, states that can pre-process absentee ballots, and states that have state-of-the-art ballot counting machines are all in an advantageous place for faster certification (to name a few factors). (Of course, states may well want to consider how they can expedite counting to meet such a deadline in the future!) Additionally, states sometimes have to delay, or amend, certificates as the result of recounts, audits, contests, or litigation.

This ECRA update offers tremendous advantages compared to the present system. It sets a firm deadline for conclusive resolution of state election contests. It does not allow some meandering litigation after the electors meet, to suggest there was some other winner (repudiating the Hawaii 1960 precedent).

Now, to the narrow point.

There may be issues relating to the executive’s certificate of election: the refusal to issue, or issuing an incorrect one. This is typically a ministerial task, and state mandamus proceedings and related processes (injunctions, state judicial issuance, Rule 70 substitutions, and so on) may well apply. It could run afoul of state deadlines, and state remedies would apply (all well before that six-day window begins).

Alternatively, an executive could issue an erroneous certificate that, say, treated votes unequally under the Equal Protection Clause, which would enable a federal lawsuit. That could happen on November 18, or on December 19, or sometime in between. It isn’t necessarily something that happens only six days before this deadline.

But, by six days before the electors are supposed to convene, there may be a situation where an executive has failed (or refused) to issue a certificate. Or, there may be a situation where on that date the executive finally issued a flawed certificate.

That is a six-day window: if there’s a problem that arises or becomes choate at the last possible date.

Again, however, look back the proceedings: for several weeks, any canvass, recount, audit, contest, or litigation issues will have worked their way through the process. At this drop-dead date, the opportunity for further factual development approaches zero. The executive’s role is reduced to the ministerial. We saw a recent mandamus case in New Mexico go from “county board refuses to certify” to “state supreme court grants mandamus” in about 48 hours.

The tradeoffs are these: if we extend the six-day window at the very end, in this narrow slice of cases where the issue arises, where do we add time?

One solution could be, of course, to move Election Day earlier. That seems essentially politically impossible (although many have spoken in favor of a weekend Election Day!), and certainly undermines the anti-“lame duck” spirit of the Twentieth Amendment.

Another, then, would be to move into the period of the canvass and recount. That seems particularly imprudent. By cutting into the period for state resolution of disputes, we undermine, in my view, what is an even more important avenue for factual development, recounts, and the like.

A third option would be to move the counting of electoral votes back, along with the convening of electors. January 6 is not set in stone, and it has been moved by Congress in the past when it has fallen on a Sunday. But a counting date 2 weeks before Inauguration Day avoids problems in the event of true election crises, witnessed in elections like 1800, 1824, and 1876. If no candidate receives a majority, Congress must elect a president after it counts votes and finds that no one has a majority. Counting of electoral votes (as we saw in 2020!) can stretch beyond a single day. Ensuring the final formal step occurs well before Inauguration Day, for reasons like security clearances and presidential transition resources, is simply a necessary mechanism.

The last, then, is to push back the convening of electors. That was suggested in the Electoral Count Modernization Act “discussion draft” earlier this year. It’s also been the focal point of some suggestions about how to fix this “six day” period. If the electors convene later (closer to when Congress counts electoral votes), it allows more time for litigation after the executive’s (federal) certification deadline.

But there’s a problem. The ECRA pushes back the electors’ convening by one day, so we’re at 17 to 23 days before Congress counts votes. And that’s also 14 to 20 days before a new session of Congress is sworn in. Couple that with the intermediary Christmas and New Year’s Day holidays (setting aside weekends), and the schedule looks tight.

Tight, because, recall, it is a new Congress that counts electoral votes. As parliamentarians, clerks, and staff gather preparing for counting electoral votes on January 6, they are preparing simultaneously for ~435 new House members and ~33 new Senators to join Congress at noon on January 3. (Some staff don’t formally begin their jobs until after January 3!) That’s a lot of preparation from states across the country, certificates of election, and preparation for a new Congress–herding attendance of new members, handling absences, and so on.

There might be places, a day here or a day there, to accommodate the schedule if one were really pressing for additional time. But each choice is a tradeoff. There is a fixed period of time to handle these matters.

When one reflects on the fact that the “six day” window is narrow in scope and application, then, I think it becomes much more justifiable. We prefer more time before that window for states to get their affairs in order. We cannot move much later than that window to handle all of the transition of government that happens. And the window, given its limited scope with expedited judicial review, is about what we’d want, where reasonable minds could differ (a day here or there), but not in ways that suggest the ECRA, even if enacted in its present form, is inadequate for the task at hand.