Can the Midterms Be 'Stolen' On Jan. 3? It's Complicated.

The date Jan. 6 has outsize significance in the United States for the worst reasons. Jan. 3, however, most Americans have never heard of. Ideally, it will stay that way.

President Trump’s demands in 2020 that Vice President Pence do “the right thing” have prompted speculation that he or his allies will attempt a similar gambit in 2026. The 2020 scheme depended on the vice president unconstitutionally rejecting votes in his ceremonial role presiding over the Electoral College count from states that favored Joe Biden.

Of course, the midterm elections do not involve an Electoral College; instead, members of the House and Senate are elected directly by the voters of their states, whose officials count, canvass, and then certify the returns. But there are still some parallels to the presidential process.

Jan. 6 was an essential moment in 2020 because, as the appointed day for the electoral count, it marked a point of no return after which the presidential election was effectively lost—by the losers, that is. Jan. 3, the day the old Congress ends and the new one begins, is the congressional equivalent: Only the swearing-in process still stands between the winners and their seats. This vulnerability lately has filled some commentators and activists with fear.

They’d probably be wrong to panic. But they’re right to pay attention.

What Happens on Jan. 3

Some of the Jan. 3 doomsday speculation hinges on the fear that Speaker of the House Mike Johnson or the current Republican House majority will “refuse to seat” incoming Democrats. Despite the prominence of these warnings in mainstream media, they are largely nonsense.

Why? The 20th Amendment provides that the terms of senators and representatives expire at noon on Jan. 3. Their successors’ terms “shall then begin.” Except it’s not that simple. The complication is that the House of Representatives is not a continuing body. This means that its entire membership turns over every two years—and the transition isn’t seamless. The old members’ terms uniformly expire at noon, but the new ones don’t uniformly begin at that same moment. Instead, the chamber enters “organization,” a reprieve between sessions during which it must constitute itself. (The Senate, only one-third of whose membership is elected at a time, is a continuing body.)

Practically, the members-elect during the organization period can speak, make motions, and vote on them, operating under parliamentary procedure and the most recent rules package—but that’s according to custom, not law. Officially, the House of Representatives does not exist in this limbo. In other words, at noon on Jan. 3, Mike Johnson will no longer be the Speaker of the House, and there will be no Republican majority. Hakeem Jeffries will no longer be the minority leader, and there will be no Democratic minority.

But embedded in this structure is also a little-understood vulnerability. A single unelected official whose authority carries over from one session to the next plays a central role in moving the House of Representatives from organization into order: The clerk of the House is responsible for presiding over the chamber during the transitional period. Crucially, this includes assembling the roll of members-elect entitled to vote on who will become speaker of the House—at which point the speaker assumes the presiding role and swears in the rest of the chamber.

So what happens if a clerk decides to seat not the lawful Democratic winner of a district’s election but the defeated Republican challenger?

This is the possibility analysts—including, most prominently, Jeffrey Toobin in the New York Times—have mused about. Right now, the clerk of the House is Kevin McCumber, who, despite having been chosen for the job by a Republican, has accumulated the bipartisan bonafides to be regarded as a reliable civil servant. But the speaker of the House—this House—can fire the clerk unilaterally. And, while installing a permanent replacement would require the majority’s approval, the speaker may instead unilaterally appoint an acting clerk authorized temporarily to perform the clerk’s duties. McCumber himself, in fact, was acting clerk for about a year and a half: between July 1, 2023, and Jan. 3, 2025.

All of this forms the plot of a highly speculative disaster tale in which Trump could demand that Johnson oust McCumber and replace him with, in effect, a political hack willing to “steal” precisely the number of seats necessary to ensure that the Republican Party reclaims its majority. (Obviously, this would prove a much easier task in the event of a squeaker 218-217 victory by the Democrats than in the case of a 235-200 blowout.) This nefarious clerk would ensure Johnson’s reelection, at which point Johnson would swear in whomever he pleased—presumably, the same majority that appeared on the clerk’s roll.

That’s very scary, but not very likely.

The Rogue Clerk Meets Reality

The so-called rogue clerk theory relies in part on the idea that Article I, Section 5, of the Constitution provides that “each House shall be the Judge of the Elections, Returns and Qualifications of its own Members.” Certainly, an aggrieved candidate, and possibly the state or its certifying officer, could bring suit over a member-elect being left off the roll. But the courts, the thinking goes, would point to that clause and decline to intervene in any dispute over the clerk’s decision to draw up a dubious roll—regarding the matter as a political question.

Maybe—but maybe not. For the same reason the clerk purportedly has the ability to cause a crisis, the courts may well have more ability (and inclination) to defuse one. Remember, at such a moment the House of Representatives, which under this theory would supposedly judge the elections, returns, and qualifications of its members, doesn’t exist. It has been dissolved and has not yet been reconstituted.

Meanwhile, the clerk—who, unlike the House, still exists—has a statutory obligation, defined in 2 U.S.C. § 26, to “make a roll of the Representatives-elect, and place thereon the names of those persons, and of such persons only, whose credentials show that they were regularly elected in accordance” with the law. This constitutes what’s known as a ministerial duty for the clerk not merely to make the roll but to make it in a particular way: putting on it only those individuals who have the legally required credentials. Those credentials, in this case, would be a certificate of election from the entity responsible under state law—sometimes the governor or secretary of state, sometimes the legislature, sometimes a state board.

Jurisprudence around ministerial duties stipulates that courts may compel compliance with obligations where the law leaves no room for discretion. 2 U.S.C. § 26, at least on its face, doesn’t appear to leave any such room. The statute doesn’t invite the clerk to doubt or to determine for himself an election’s legitimacy. Instead, it requires him to accept a state’s assessment and assemble the roll in accordance with that assessment.

All of this is the basis for the theory, ably articulated by Stanford Law School’s Matthew Seligman, that the lawfulness of the clerk’s actions should not be considered a political question: The only reason the clerk remains in his position when the House has ceased to exist is because the clerk isn’t a member of the House. And because the clerk isn’t a member of the House, ordering him to perform his ministerial duty wouldn’t seem to infringe on the House of Representatives’s constitutional prerogative to judge the elections, returns, and qualifications of its own members. The courts, then, may well be willing to issue a writ of mandamus ordering the clerk to place the lawful winner on the roll—or ordering another official to do it if he won’t.

(It’s worth noting that this suggests the roll-making process is only as robust as the certification process. If a state presents a single certificate from a single statutorily designated source, the clerk is obligated to put that person on the roll regardless of whether state officials acted lawfully or disregarded the election results. Still, it’s unlikely the clerk would end up in this position: Certification of the lawfully elected candidate isn’t a choice election officials get to make but, rather, an enforceable ministerial duty itself. Pulling off a certification coup would therefore require complicity from every actor in the state and local ecosystem, from county clerks to elections boards to executives to prosecutors to courts. Federal courts also presumably wouldn’t permit such a scheme under the electoral due process doctrine of Roe v. Alabama and related cases.)

The clerk’s obligations are straightforward when only one certified winner presents herself, but they’re less so if two winners present themselves with two conflicting certificates—because it may not be clear, at least superficially, which candidate the clerk is obligated to put on the roll. Particularly in the event of a certification dispute that is at least plausibly in good faith, judges might be inclined to read the clerk’s statutory duties as leaving room for discretion after all, making it improper for the judiciary to issue a writ of mandamus. However, to the extent that state law is clear—which it largely is these days—only one set of credentials would truly show a candidate was regularly elected. So there’s still a strong argument that the clerk’s duty would therefore be to put that candidate on the roll, that duty would be a ministerial one, and courts would have the authority to compel the clerk to comply with it.

However ironclad a state’s certification processes are, there also remains the possibility—particularly in the event of a ballot seizure breaking the chain of custody on the physical results of the vote—that litigation could drag on past the Jan. 3 deadline (as, in some past instances, it has). There would be no credentialed candidate in this situation, and the “such persons only” provision of 2 U.S.C. § 26 would actually impose on the clerk a ministerial duty to leave the seat vacant. The courts would, theoretically, have to enforce that duty, too.

But leaving a seat vacant has only half the effect of installing the wrong candidate. Say an election yields a 220-215 split in favor of the Democrats. If a clerk replaced three Democrats on the roll with three Republicans for the purposes of the speaker vote, the breakdown would change to 218-217, with the Republicans now in the majority. But if the clerk can only remove Democrats but not replace them, leaving three seats vacant would result in a 217-215 split, with Democrats still in the majority. And, of course, any Republican majority would have to vote in lockstep to subvert the election—something that law professor Derek Muller points out in a piece on Jan. 3-related fearmongering this disorderly Republican majority in particular doesn’t seem well-suited to bring about.

Nonetheless, while a hypothetical rogue clerk may find it difficult to singlehandedly overturn the democratic order, he still has plenty of opportunity to create chaos. Whether or not judicial intervention ultimately arrived in any dispute, the members-elect of the party from whom the majority was being improperly wrested would surely object to an incorrect roll. At that point, considering multiple claimants for contested seats would presumably show up to work on Jan. 3, the proto-House of Representatives would have to vote on who should be allowed to vote for speaker. This would likely result in a vote on who should be allowed to vote on who should be allowed to vote for speaker, and so on—with the chamber descending into a parliamentary vortex.

The eventual outcome, as law professor Ned Foley put it to me, is ultimately more of a political analysis than a legal one: “The reorganization of the House is the barrier between law and non-law.”

The Rogue Clerk Meets History

One reason to take the rogue clerk concept seriously is that something like it has happened before. Foley chronicles multiple instances in his book “Ballot Battles.”

The closest analogue is the Broad Seal War of 1839, a contest between the Whigs and the Democrats involving five New Jersey seats. The governor, per the usual practice, confirmed the election of the Whig winners of these districts by granting them certificates affixed with the official “broad seal” of the state. The problem was that the secretary of state certified different winners, issuing self-styled credentials to Democrats on the grounds that a complete tally of the vote—including two towns improperly excluded from the county—revealed that they, not the Whigs, had actually won.

This dispute landed before the House clerk, Hugh Garland, who was selected 106-104 by an outgoing Democratic House hoping for his compliance. Whichever set of claimants was put on the roll would solidify a majority; that majority could then elect a speaker. The clerk responded by throwing up his hands, leaving the five seats vacant until the rest of the House was sworn in. No business could occur until there was a quorum, but there could be no quorum until the clerk called the roll, and the clerk could not call the roll until there were no objections.

This led to a deadlock that neither side could surmount without those five disputed seats, with multiple votes on who was allowed to vote—in at least one of which four Whig claimants and four of the Democratic claimants participated. Ultimately, the House voted to leave the seats vacant until the roll was called. The Whigs succeeded in electing a Whig speaker, and the Democrats succeeded in having their claimants seated when enough Whigs decided the exclusion of those two towns’ votes wasn’t fair.

The second relevant episode occurred nearly a quarter-century later, during the Civil War. The Republicans controlled the House heading into the 1862 elections. Sensing a close contest, they passed the very statute that defines the clerk’s duties today: now codified as 2 U.S.C. § 26. But back then, the Republicans took the expansive view that this law allowed the clerk, Emerson Etheridge, to exclude candidates (mostly from border states) whom they viewed as disloyal to the Union and thus improperly credentialed. Unfortunately for them, Etheridge decided he liked the Democrats better after all. He did end up excluding 16 claimants from the roll, except that they were Republicans.

This effort failed. When Republicans moved to add their claimants to the roll, only the members Etheridge had already recognized—and not the candidates whose victories were challenged—were allowed to vote. But the Democrats lost regardless: They couldn’t scare up a majority to disregard the election’s apparent results. What would have happened if the clerk simply attempted to rule the Republicans’ resolution to overrule him out of order is uncertain. The Republicans might have tried to install a temporary speaker; they even considered whether they’d have to forcibly eject Etheridge from the floor.

Game Over

These stories teach a few lessons. One lesson is that the organization period that begins at noon on Jan. 3 really is a Rubicon: Whichever party manages to cross it with a majority should feel secure in its continued control of the chamber.

Yes, after a speaker election, the minority could dispute a number of seats greater than or equal to those needed for a flip in power. Law professor Brian C. Kalt has written extensively about the “stand-aside principle,” by which the members whose election was challenged customarily wait until everyone else takes the oath. After that, those sworn in vote on whether to seat the disputed claimants—to which the majority, having turned itself into a majority through the contestations, would say nay, before vacating the chair, electing a new speaker, and seating their preferred candidates instead.

But there are ways out of that bind. The majority speaker originally presiding—let’s say it’s Hakeem Jeffries—could simply refuse to make the challenged members of his party stand aside. Even if Jeffries complied, the Democrats in this version of the story could turn around and object to a matching number of Republicans so that they, too, were forced to stand aside and the whole rigmarole would end in a wash. This is more or less what the resolution to the Bloody Eighth reveals: Majority rules the moment there’s a majority at all.

The only outstanding question is whether, in the case of a flagrantly lawless switch of a seat from one party to the other after a speaker has been elected, a court would intervene. The answer is likely no, thanks to the same part of Article I, Section 5, of the Constitution that doesn’t apply to the clerk but does apply to congresspersons: that “each House shall be the Judge of the Elections, Returns and Qualifications of its own Members.” This is precisely where courts would likely refuse to answer what they viewed as a political question. That reality is especially concerning given the aforementioned possibility that delayed certificates could result in the clerk leaving enough seats vacant to flip the speakership election. Once the speaker is elected, after all, he can fill the rest of the seats how he pleases regardless of whom those certificates eventually designate the winning candidate.

One caveat concerns the 1969 case Powell v. McCormack, in which the Supreme Court ruled that the House can’t exclude a properly credentialed member-elect who satisfies the constitutional qualifications of age, citizenship, and residency for the position. A majority of the House had voted to not to seat New York’s Adam Clayton Powell Jr., who was accused of misappropriating federal funds, after his reelection. The Court found that while the House could expel a member for misconduct by the required two-thirds vote, it couldn’t exclude a duly elected official from membership in the first place.

But this narrow ruling focused on constitutional qualifications rather than certification credentials; there wasn’t, in this case, any dispute over who had won the race. There’s still a chance that a court, confronted with an election conclusively certified with no colorable claim to the contrary, would similarly find that the House had unlawfully interfered “with the right of the people to choose their own representative.” The moment an election is plausibly disputed, however, is the moment the courts may well bow out.

So What’s the Good News?

These tales of doubt and disorder involving the clerk may seem discouraging. But before readers despair, they should take note that even when non-law seemed to win out over law, political compromise prevailed over partisan uniformity. Now that the legal system surrounding certification is stronger, not weaker—laws having replaced what once were only norms—it seems even less probable that a rogue clerk could enable a minority party to undo our democracy.

That’s no reason for complacency. There’s always room to strengthen the system further, for these midterms where it’s still possible and for future contests where it’s not. Delayed certifications open up the widest opportunity for mischief—so states should take every available step to ensure certification can’t be delayed, at least not unduly. That involves everything from tightening the process itself (a subject deserving of thousands of words of its own), to clarifying its mandatory nature, to ensuring that any litigation is resolved ahead of Jan. 3. Preserving records so that results are readily verifiable even when the chain of custody around ballots themselves is broken is also essential.

Valuable as it is to take the rogue clerk scenario seriously, it’s also imperative not to take it too seriously. Treating a relatively unrealistic hypothetical as a foolproof electoral hack implies the move is lawful: a way to slip through a legal loophole. In reality, the would-be victims of such an effort would be wise to challenge the gambit as a lawless contravention of norms, statutes, and the Constitution all at once. It also assumes that the move would work. Defenders of democracy should be on notice that this ploy not only is preventable but also has been prevented. Subverters of the same should be aware that the most they’re likely to achieve through the stratagem is mayhem.

The real worry is that, for some, mayhem is more than enough.