“…in Case of Disagreement between [the House and Senate], with Respect to the Time of Adjournment, [the President] may adjourn them to such Time as he shall think proper.” — Article II, Section 3 

The presidential power of adjourning Congress has never been used. There are no precedents and scant commentary about what it means or exactly what triggers it. But now, there is credible consideration of the idea being discussed as part of Trump’s demand for an adjournment to let him use recess appointments to completely bypass the Senate confirmation process. It wouldn’t be the first time the issue has come up, as Trump briefly floated it during his first term, and Speaker Mike Johnson has refused to rule out possible support for the scheme. 

It is important, then, for us to quickly get a handle on exactly what this obscure bit of constitutional text means and what it does, and doesn’t, allow the president to do. 

The possibility is uncomfortably suggestive of one of the most firmly repudiated ideas in Anglo-American law, the attempt by Charles I to rule without Parliament. That did not end well, to put it mildly, for either side of the dispute or the nation as a whole. It was an example the Framers of the Constitution were very aware of and consciously sought to avoid by explicitly constraining the chief executive’s power to interfere with the legislature. 

As Ed Whelan writes at NRO, the basic outline of the idea is as follows: the House (presumably more amenable, though that’s far from certain with a razor-thin Republican majority) would pass a concurrent resolution adjourning Congress, which is the normal procedure. The Senate would not concur. Trump could then cite this as the two chambers being in “disagreement,” and adjourn them to whenever he wants. To allow for recess appointments would require an adjournment of at least ten days. But in theory it could extend nearly an entire year, until the next constitutionally mandated annual convening of Congress on January 3, 2026, per the 20th Amendment. 

Would this scenario actually be a “disagreement … with respect to the time of adjournment” as contemplated by the Constitution? It would appear to be so as a practical matter, but I would contend not as a formal procedural posture. In short, the House passing an adjournment resolution while the Senate simply takes no action would not be a disagreement between them. (For clarity, I will assume the scenario of the House trying to adjourn while the Senate opposes it, though the same analysis would apply vice versa). 

Disagreement requires conflicting action on the part of both the House and the Senate. They must each pass adjournment resolutions, specifying different proposed dates. This is both a fairer reading of the clause’s plain language and a necessary construction to avoid absurd results. In practice, this means the Senate can prevent a presidential adjournment by not taking any action to officially disagree with the House. 

Even in the normal course of business, the two chambers do not act simultaneously. Between an adjournment resolution passing one chamber, some time elapses… minutes, hours, days… before the other chamber can concur. If mere non-action suffices to create a “disagreement,” then the president could intervene during this window any time Congress adjourns. This was plainly not the intent, nor would it be a reasonable definition of the word disagreement. 

 

A disagreement only occurs when one chamber actually expresses its disagreement with the other. This requires action, not simply inaction. It means adopting a conflicting adjournment resolution of its own. Or at the very least, to take a floor vote and affirmatively reject the proposed adjournment resolution, which the Senate has no obligation to do. Until then, the Senate might still agree with the House and has not said otherwise. While the political reality might be apparent, no formal mechanism of disagreement would have been triggered. 

Another reasonable argument could also be had: if the House wishes to adjourn while the Senate does not wish to adjourn, is this is a disagreement over the “time” of adjournment? This qualifier must be read to mean something, under the canon against interpreting constitutional language as mere surplusage. Otherwise, the provision would read disagreement over adjournment and not only over the time of adjournment. The former would not be a conflict over scheduling the action in question, but over whether to do it at all, at any time. Compare the Constitution’s other time-setting powers, for Congress to exercise over presidential and congressional elections, and it becomes obvious that setting a time is something different from deciding whether to ever do the thing in question. 

More to the point, what happens if the president plus the House purports to adjourn Congress but the Senate does not accept this as valid? Either on one of these interpretations or some other, such as the idea that presidential adjournment only applies to special sessions called by the president, as suggested by my colleague Thomas Berry. Other theories would involve the interaction between dates set by statute and adjournment by non-statutory mechanisms, distinctions between the power to adjourn both houses of Congress as opposed to adjourning only one of them, or arguments over who has the power to re-convene the Senate and on what authority. 

At this point, the correct course of action would be for the Senate to simply continue to meet at its previously appointed times and conduct business. A majority would still constitute a quorum. If the president’s proclamation of adjournment is null and void, there is no way for the Senate to assert that position other than treating it as a nullity, an act with no more force than if you or I tried to adjourn Congress. 

This scenario would be well into a full-blown constitutional crisis, to be sure, a Senate whose very existence is denied by the president and the House. Vice President Vance, as president of the Senate, could show up in person at the Capitol and repeat the position that the body is adjourned. However, this quickly becomes a rather absurd contradiction: if the Senate is not in session, there is nothing for the Vice President to preside over. Claiming the gavel would be conceding the point. And if he declines to claim his constitutional power to preside, the job would pass per the normal rules of the Senate to the president pro tempore or their designee. 

All of this so far offers little opportunity for direct judicial intervention, any clear path for anybody with standing to present a justiciable case. But since the whole goal of the exercise would be for Trump to make recess appointments to the executive branch, this would squarely tee up a challenge in line with NLRB v. Canning to the legitimacy of those appointments, by anybody affected by their purported actions. And though Canning was not directly about the presidential adjournment power, it does stand for the general proposition that the Senate is in session when the Senate says it is in session. This follows from the fact that each house has the power to determine the rules of its proceedings, including deciding disputes over those rules. If the court were to either reject or validate the claimed recess appointments, that would effectively decide the underlying issue. 

All this may well be moot, and we can certainly hope it remains theoretical. The White House could back away from the idea of trying to adjourn Congress. As few as two or three Republican defectors in the House could deny the needed majority there, so there would be no basis to claim disagreement between the houses. The Senate may proceed to confirm most if not all of Trump’s Cabinet nominees, with the demand for recess appointments falling by the wayside as a consequence. Or the Senate could hold firm and reject nominees, defending its constitutional prerogative.

But if the presidential adjournment power is invoked, it should be recognized as not just a serious, norm-defying abuse of a procedurally valid constitutional power. Instead, it would be unconstitutional. The Senate would have firm grounds to ignore it, and the Supreme Court could reasonably agree. 

In September I became the fourth director of Cato’s Robert A. Levy Center for Constitutional Studies. This is the biggest honor of my professional life, and it is also a tremendous responsibility. The Levy Center (or CCS, as we call it here) has been shaped by the three directors who came before me. And those three directors have also served as professional mentors to me over the last nine years. The best way to explain why CCS is so unique and so important is by sharing what I’ve learned from each of its previous three directors: Roger Pilon, Ilya Shapiro, and Anastasia Boden.

First, Roger Pilon. Roger founded CCS and led it for 30 years. Roger was the CCS director when I interned here in 2015 and when I served as a legal associate in 2016–17. When Roger founded CCS, the main philosophical dispute in constitutional interpretation was framed as “judicial activism” versus “judicial restraint.” Roger founded CCS to advocate for a then-radical third approach: judicial engagement. As Roger put it, CCS was founded “to encourage judges to be more engaged than many conservatives believed proper, and to locate the authority for that engagement in the Constitution itself, properly understood.” 

Roger made CCS the intellectual home of the idea that the Constitution strongly protects personal liberty in all its facets, including economic liberty. As Roger summed it up, the goal was “to help change the climate of ideas to one more conducive to liberty under limited constitutional government.”

On my first day as a legal associate, Roger sent me a reading list. If I had been starting at a law firm, such a list might have comprised style guides and legal hornbooks. But Roger instead sent works of legal theory and political philosophy. Thanks to Roger’s example, I’ll always remember that CCS is fundamentally about ideas—about legal and political philosophy put into application. The most important test for everything we do at CCS is whether it will help to advance the acceptance of those ideas by both the courts and the general public.

Next, Ilya Shapiro. Ilya made Cato’s amicus brief program what it is today. It was under Ilya that Cato began its practice of hiring four legal associates (recent law school graduates) per year to help draft amicus briefs that would be filed in state and federal courts across the country. Writing more than 30 amicus briefs per year with roughly a half-dozen lawyers is an enormous undertaking, but Ilya made it look easy. To run the amicus program, Ilya had to make many quick but informed decisions about which cases to file in, what arguments to make, and how much time to devote to each brief. By his example, Ilya taught me the judgment and organizational practices needed to make these decisions and run this program. As I now begin to manage Cato’s amicus program myself, I know that the task is feasible only because I watched Ilya do it firsthand.

Perhaps even more important, Ilya taught me how to write a Cato amicus brief. Again, he taught by example. As a legal associate, I was tasked with writing the first draft of several amicus briefs. And like most recent law school graduates, I thought I could show off by sprinkling in legalese, jargon, Latin phrases, and obscure words found deep in my thesaurus. As my editor, Ilya dutifully took out all of them. It was during my year as a legal associate that I learned the point of our briefs is to be understood, not to show off. 

I saw how Ilya transformed my drafts from middling to crackling. Short, snappy sentences. Section headings of just a couple of words. Direct language. And arguments that can be understood not just by lawyers, but by any educated layperson. As I now take on the role of final editor for Cato’s briefs, I still have Ilya’s editorial voice in my ear.

Finally, Anastasia Boden. I was fortunate to work with Anastasia when we were both attorneys at the Pacific Legal Foundation. When she came over to Cato and we were professionally reunited, she reinvigorated our amicus brief program with a litigator’s perspective. Yes, our briefs are about ideas. But they’re also filed in real cases involving real people. These human stories help explain why the ideas and principles we argue for matter. And these human stories deserve pride of place in amicus briefs just as much as they do in the parties’ briefs. Anastasia’s keen instincts as a litigator gave her a sense of what grabs a judge’s attention and what puts them to sleep. For the rest of my career, I will never be satisfied with a brief’s opening paragraph unless I think Anastasia would be satisfied with it.

My hope as CCS director is that I can put into practice what I have learned from each of my predecessors. My mission is for CCS to produce accessible legal writing that is both principled and personal. Our briefs and papers should explain the philosophical and historical case for our view of the Constitution. They should demonstrate the human cost of departing from that vision. And they should clearly communicate that message to both judges and the general public alike.

I’m keenly aware that Cato’s CCS fills a role that no other legal institution inhabits. We have filed amicus briefs defending a constitutional right to same-sex marriage and a constitutional right not to be forced to create speech endorsing a same-sex marriage. We’ve aligned with conservative legal groups in economic liberty and Second Amendment cases, and we’ve aligned with progressive legal groups in police accountability and Fourth Amendment cases. 

We defend every constitutional right on an equal footing, and we collaborate with anyone willing to defend that right with us. That has been CCS’s guiding principle since it was founded, and I won’t be the director who changes that approach.

Going forward, what do I see as the likely focus of CCS? As always, we will meet the moment with whatever legal arguments are most needed. In this moment, officeholders of both parties seem more interested in using political power to fight in the culture war than in getting the government out of the culture war. The CCS has opposed both “conservative” efforts to compel speech by private tech companies and “progressive” efforts to compel speech by private retailers. The CCS has opposed both a Texas prosecutor’s misguided crusade against a risqué movie and campus regulators’ overzealous suppression of edgy speech. 

Whenever the government tries to put a thumb on the scale in the free arena of public debate, Cato will step in and defend the principle of government neutrality.

In addition, this era continues to see the leaders of both parties try to circumvent Congress and erode the separation of powers. Cato has opposed both the Trump administration’s attempt to build a border wall without congressional authorization and the Biden administration’s attempt to waive student loan debt without congressional authorization. I’m immensely proud of this record, and it will continue no matter which party wins the next election.

Does the work that we’ll be doing at CCS sound exciting to you? If so, I’d love for you to join us. If you’re a lawyer with three or more years of legal experience, apply to be a legal fellow, which is a permanent scholar position at Cato and the position that I held for the last four years. If you’ll be graduating law school in 2025 or if you’re a lawyer with fewer than three years of experience, apply to be a legal associate. That’s a one-year position that will allow you to draft amicus briefs across all of Cato’s issue areas. And that was my first job out of law school. Finally, if you’re a current law student, apply to be a summer, spring, or fall legal intern. That was the first position I ever held at Cato. And the rest is history.

At his rally in Madison Square Garden on Sunday, Donald Trump offered an aside to Speaker Mike Johnson about a “little secret” concerning the election and the House of Representatives. “[W]e can take the Senate pretty easily. And I think with our little secret, we’re going to do really well with the House, right?” Trump said to Johnson. “Our little secret is having a big impact. He and I have a secret. We’ll tell you what it is when the race is over.”

Many took this as a nod to possible efforts to obstruct the joint session where electoral votes are counted on January 6. It’s not entirely clear what Trump might have meant, and Johnson has been evasive about whether he would seek to lead, as he did among House members four years ago, with ill-founded objections to a possible Democratic victory. 

But because the topic is attracting concern, it’s worth putting to rest the idea that the Speaker has any procedural tools to change the outcome of the presidential election. 

The Speaker’s role during the electoral count is, in fact, nearly non-existent. There are no lawful mechanisms for the Speaker to derail the joint session of Congress without the backing of concurrent majorities in both the House and the Senate. The job of the Speaker is much more constrained here than their usual authority over the House. 

To start, the statutory rules in the Electoral Count Reform Act of 2022 (ECRA) are in place as the default unless both chambers concur in any deviation by means of a concurrent resolution. Normally, both chambers also adopt a concurrent resolution simply reiterating the statutory rules. But the statute, which directs that both chambers convene in joint session in the House chamber at 1 p.m. on January 6, and lays out detailed rules for what then follows, is valid and binding of its own force.

It has, after all, already been adopted by both houses and in many respects it simply reiterates what is already commanded by the Constitution. While Congress could by its internal rule-making power adopt some theoretical changes, should unexpected circumstances arise, doing so for the joint session would require the House and Senate to agree. The Speaker alone has no such power, and neither does the House alone. 

Nor does the Speaker preside over the joint session, a task falling to the vice president as President of the Senate, who happens to be Kamala Harris. (She would not be the first VP to preside over certification of her own victory or defeat). The President of the Senate has no power to make decisions about the electoral count itself, about rejecting or accepting votes, but as presiding officer she will still have the “power to preserve order” as specified in 3 USC § 18. 

Though the electoral count might take place in the House chamber, the space is not, for that time and purpose, the House’s alone or the Speaker’s to control, as it would be in the course of normal business. It is under the control of Congress as a whole with the President of the Senate tasked with maintaining, if need be, literal physical control of the room, directing the sergeants at arms and however much backup is necessary to that end. 

The only scenario in which the Speaker retakes the gavel is if an objection to a state’s electoral votes is made by at least one-fifth of each chamber, or at least 87 representatives and 20 senators. In that case, the Senate will withdraw to its side of the Capitol and both chambers will separately consider and vote on the objection, with the agreement of both chambers required to sustain it. 

This period of debate and vote is limited by ECRA to no more than two hours. During that time, the Speaker (or as always, another member they designate) would preside over the House debate but not have any lawful power to postpone the reconvening of the joint session after those two hours expire. 

One of the more outlandish ideas some have suggested is that Speaker Johnson could, when the newly elected House convenes on January 3, refuse to seat some number of Democratic members-elect—even, potentially, if Democrats have regained the majority at the election. 

However, the speakership does not automatically carry over into the new Congress. Rather the position is initially vacant unless and until the members-elect choose a Speaker in the process of “organizing” the House on the first day of their terms. After the election of a Speaker, there is a procedure for objections to a member-elect’s qualifications to be seated, with the rest of the House excluding that member then voting on the question. 

But such objections made on a blatantly partisan basis could quickly escalate into a rather absurd tit-for-tat spiral. Baseless objections to seating Democratic members could be met in-kind with retaliatory objections to seating Republicans. The whole thing would spiral into a farcical deadlock, inviting almost philosophical debates about who “the House” really is on January 3. In any event, the Speaker alone does not decide, having no individual power to simply refuse to let members take their oaths. 

More substantively, such a lawless purge of Congress on spurious grounds would not be constitutional. At that point, the rump “House” would only be destroying its own constitutional legitimacy, rather than making a feasible play for control of the White House. While it would be a grave constitutional crisis, it is not one the putative Speaker, with a dubious claim to that office under the circumstances, would be in a position to win. 

Any attempt by the Speaker of the House to subvert the electoral count might observe a hollow pretense of following lawful procedures, but it would not be lawful. Hopefully, we will not face such a senseless confrontation. And it’s worth emphasizing these are extreme worst-case scenarios. 

Again, who knows what Trump was really talking about, or what Johnson might have in mind in the event Trump loses the election. It is very unlikely that the needed majorities of both the House and Senate would go along with this hypothetical coup attempt, even in the event Republicans have narrow majorities. 

But if such a thing was nevertheless attempted by the Speaker acting unilaterally, the legally and constitutionally required course would be for other actors and institutions—including the members of the House, the Senate, the vice president, the courts, and the executive branch—to treat such actions as null and void. 

There is no one neat trick or magic words the Speaker can use to reverse the outcome of a presidential election, forcing everyone else to accept an illegitimate power grab. The law and the Constitution remain the law and the Constitution even in the face of a rogue officer’s attempt to violate them. 

As we get closer to the election, a number of articles have been published offering legal explainers for how the process under the new Electoral Count Reform Act of 2022 would work. These have been broadly correct in most details, but one common claim reflects a confusion worth clarifying. Specifically, the conditions under which a “contingent election” could be triggered, sending the race to the House under the procedures of the Twelfth Amendment. 

A contingent election has been a particularly acute fear among Democrats, because in the House it is not each representative who would have one vote, but each state’s delegation. An absolute majority of states, twenty-six, is required to win. It is likely Republicans will control more state delegations than Democrats even if Democrats have an overall majority of the House. It is also possible, because of states being evenly split, that no party would control the needed twenty-six votes. This was the case last time, even if a contingent election had been triggered, due to Republican members opposed to the effort in Wyoming and Michigan. 

The most obvious way a contingent election could happen, though it’s unlikely, would be an exact 269–269 tie in the Electoral College. Even less likely would be a third candidate winning enough states, a strategy attempted several times in American history but never successful. There are no prominent third party candidates with such a chance in this election. Under these scenarios, no candidate would then have “a majority of the whole number of Electors appointed,” which the Twelfth Amendment requires to avoid a contingent election. There are 538 electoral votes, so normally a majority would be 270. 

The other possibility frequently discussed is that Congress, during the electoral count on January 6, 2025, would reject enough electors so that no candidate reaches 270 votes. However, the form of the objection matters here, and the kind of objection most often floated by Trump supporters would not result in a contingent election. 

This distinction arises from the implications of the Constitution itself, but is codified by ECRA with two different provisions: 3 USC §15(d)(2)(B)(ii)(I) and 3 USC §15(d)(2)(B)(ii)(II). Try reading those citations out loud five times fast. But more simply, the first kind of objection is that the electors were not “lawfully certified.” The second kind of objection is that the votes of valid electors were not “regularly given.” 

An objection that electors were not lawfully certified would be the implication of claiming a candidate did not really win a state’s popular vote, or did not do so lawfully, and so that party’s electors were not lawfully certified. This kind of objection is intended to be foreclosed by ECRA altogether, by requiring only a single certified slate of electors as determined under state law and confirmed by possible litigation, as provided under other sections of the law. The whole point is that sitting in judgement of a state’s popular vote result is not within the proper purview of Congress. The provision remains only for the slim possibility that those mechanisms have failed and Congress is presented with a slate of electors not in compliance with the law’s detailed procedural definition of “lawfully certified.” Electoral certificates to the contrary would never reach the joint session unless the law is already being broken. 

An objection that votes were not “regularly given” would mean the electors themselves were properly appointed but they somehow then cast invalid votes. This would cover, for example, if the electors voted for an ineligible candidate, not entirely hypothetical in this race, but lacking serious support in Congress. It would also apply if the electors voted for two candidates for president and vice president who are both residents of the same state as the electors, which is not allowed. It would in theory also include failure of the electors to meet and vote on the required day, which the Constitution requires to be the same throughout the United States, or if the electors failed to certify their votes in the form required. 

The first kind of objection, lawfully certified, would have the effect of reducing the “whole number of electors appointed,” because its contention is that the rejected electors were not really electors at all. This would mean the denominator for determining a winning majority is likewise reduced. It would be the same as saying the state simply failed to appoint any electors, as southern states did during the Civil War. In such a case, whoever has a majority of the votes cast by the remaining electors, unless there is a tie, would be directly elected. There would be no contingent election in the House. 

Only the second kind of objection could result in the total number of electors still being 538, thus leaving the winning majority threshold at 270. But this kind of objection has little relevance to the common theories advanced by Trump supporters in the last election and being readied for a possible repeat attempt. To make an objection of this nature would be conceding that Harris really did win the relevant state, the Democratic electors were properly certified, and yet those electors somehow cast invalid votes. 

The law makes this distinction clear, carefully tracking the Constitution’s text. 3 USC §15(e)(2) provides that the whole number of electors is reduced only if a state has failed to appoint the number of electors to which it is entitled, or else an objection is sustained under “subsection (d)(2)(B)(ii)(I)” (emphasis added). That is, a “lawfully certified” objection, but not a “regularly given” objection.

In other words, the kind of objection which could trigger a contingent election is also the kind of objection made effectively impossible by ECRA, because “lawfully certified” objections can only be raised if Congress is somehow presented with a slate of electors in defiance of a federal court ruling. ECRA requires such a question to be decided in the courts prior to January 6, with that outcome being binding on Congress. The concept of Congress choosing between dueling slates of claimant electors, as envisioned under the previous Electoral Count Act of 1887, is no longer applicable. 

As a practical matter, even with the possibility of narrow Republican majorities in both the House and Senate, it is plain not enough would vote in favor of an attempt to overturn the election. ECRA was co-sponsored by ten Republican senators, and passed out of committee with only one Republican vote against it. In 2021, about two-thirds of House Republicans voted for the objections, and in the Senate, only eight voted for one or both of the two objections considered. Considering that any plausible Republican majorities would have only a few votes to spare, possibly as few as just one, majorities in both chambers will be out of reach. Even if higher shares of the GOP members vote that way, it would require near-unanimity for an idea already firmly rejected by many Republicans.

Attempts to subvert the outcome of the 2024 election are serious and ongoing, and merit serious opposition. But a contingent election in the House is not how such an effort would play out, and popular explainers of the law should avoid erroneously suggesting otherwise.