To protect America’s democracy, members of Congress should dust off their copies of Homer’s Odyssey and re-read the part about the song of the sirens. On Circe’s advice, Odysseus ordered his sailors to tie him to the mast of his ship and to fill their own ears with beeswax so that he and his sailors could not be lured by the sirens’ irresistible song. Once within earshot, however, Odysseus ordered his men to untie him—such was the song’s power—but the crew appropriately disobeyed that immediate order, following instead the antecedent instructions about the necessity to avoid succumbing to the sirens’ temptations. As a result of adhering to that anticipatory strategy for self-protection, Odysseus and his ship managed to sail safely past the danger and thus escape their demise.

Congress needs to adopt the same approach when it comes to revising the Electoral Count Act of 1887 in preparation for the 2024 presidential election. This law sets forth the rules and procedures for Congress itself to follow when it receives, opens, and counts the votes for President sent from each state. As the insurrection at the Capitol on January 6, 2021 tragically illustrated, the existing law too easily permits members of Congress—like Senators Josh Hawley and Ted Cruz, or Representatives Mo Brooks or Paul Gosar, among many others— to be tempted by the notion that they can repudiate the valid votes that the states send to Congress. Like Odysseus ordering himself tied to the mast before hearing the sirens’ song, Congress now needs to enact unambiguous provisions in advance of the 2024 election that, when the moment arrives for Congress to count the votes, will prevent its members from succumbing to the partisan lure of subverting a valid victory by the other side.

The task of modernizing the 1887 statute should not be too difficult—as long as there is the political will in Congress, on both sides of the aisle, to lock in this mandatory self-restraint. The 1887 law itself embraced this philosophy, although it did so with nineteenth-century verbiage that is so convoluted and archaic to contemporary ears that it’s become functionally incomprehensible and thus unworkable. Simply modernizing its approach in straightforward twenty-first century formulations would be a huge public service.

But more than that, the role of courts—and especially federal courts—in resolving disputes over counting ballots has evolved considerably since the Gilded Age. What therefore needs to happen in Congress now, in order for the law to be well-tailored to current threats of electoral manipulation by whichever side is disappointed by the outcome, is for enough Democrats and Republicans to come together over the principle that the courts are capable of assuring an honest and accurate result according to the rules established before the ballots are cast. If Congress can pin this point down in advance of the 2024 election, then America can rely on its judges to do their job if the outcome of the election ends up disputed. The law adopted by Congress will prevent partisan politicians—including members of Congress themselves—from attempting to repudiate the results that the courts have confirmed, or revised, as required by the relevant rules (including applicable constitutional provisions, like equal protection and due process).

Presidential versus Congressional Coups

As Americans look back at the January 6 insurrection in an effort to comprehend what went wrong and ponder how to prevent its recurrence, the main focus has been on what role President Trump played in the effort to derail the official congressional declaration of Joe Biden’s election. This focus has caused the inquiry by the House Select Committee to Investigate the January 6th attack on the United States Capitol to define the unprecedented attack on democracy that occurred on that date as an attempt by an incumbent President to stay in office for a second term even when the voters decided to turn him out of office and to replace him with his challenger. Many commentators (including a federal judge) have characterized this as an attempted “coup” of a certain kind—a “selfcoup,” according to the terminology of some analysts—and there’s been much attention on the extent to which Trump endeavored to employ aides and other assets within the executive branch as part of this antidemocratic effort to cling to power despite the will of the electorate. The role of senior Department of Justice officials after the resignation of Attorney General Barr (who famously told Trump that his claim that the election had been wrongfully stolen from him was “bullshit”), has come under particular scrutiny. And the explosive revelations from late March of the text messages between Ginni Thomas and Mark Meadows are especially significant because Meadows was White House chief of staff and thus President Trump’s most senior aide, and so it supports the contention that a conspiracy to subvert the congressional completion of the presidential election process was coordinated from the West Wing.

But in truth the January 6 plot was not an abuse of presidential power. Trump did not exercise an authority granted presidents under Article II of the Constitution as a part of his efforts to stop Congress from recognizing Biden’s victory against him. Under the Constitution, an incumbent President plays absolutely no role whatsoever in the entire Electoral College process, except potentially as a candidate for reelection (as Trump was). And it was as a candidate, not as an officeholder, that Trump acted in attempting to subvert the January 6 session of Congress that the Constitution specifies for the counting of electoral votes. Even the pressure that Trump was attempting to exert in Georgia, by having compliant DOJ officials threaten local election administrators with bogus federal charges, would have required Georgia’s officeholders to do what Trump wanted and then for Congress to have accepted these fabricated Georgia results instead of the real ones. The President of the United States has no power to control the vote count in Georgia (or any other state) or to insist on how Congress handles the electoral votes received from a state. (This is why, by the way, Trump’s claims of “executive privilege” in this context ring so hollow; there was no “executive” function that he was performing in connection with the January 6 congressional session—except the key one he did not perform once violence at the Capitol erupted: the constitutional obligation under Article II of the nation’s chief executive to deploy executive branch forces to protect the lawful functions of the federal government, including congressional proceedings, from forcible disruption.)

Even Trump’s efforts to persuade Vice President Mike Pence to abuse the latter’s role in the January 6 process was not an exercise of presidential power, but rather one running mate attempting to assert political leverage over another. Pence’s role was the constitutionally anomalous one caused by the fact that the Vice President, despite being part of the executive branch, is also an Article I legislative officer as the Senate’s president. It is only as Senate president that Pence was in a position to potentially disrupt the January 6 session of Congress: The Twelfth Amendment to the Constitution specifies that electoral votes from each state are sent to the Senate president, who opens them in a special congressional proceeding in which both chambers of Congress are in attendance. Thus, when Trump implored Pence to disqualify electoral votes cast for Biden, Trump was not acting as the nation’s chief executive directing his second-incommand to engage in any executive capacity. Instead, it was Candidate Trump exhorting Pence to abuse his special legislative authority, as Senate president, to act on behalf of their shared campaign as the Republican ticket for reelection.

Thankfully, Pence did not do what Trump wanted. But even if Pence had, it would not have been an abuse of executive power on either Pence’s part or Trump’s. Instead, it would have been an abuse of Pence’s legislative power in his special capacity as an officer of Congress, and it would have been a politically inappropriate request by Trump as a presidential candidate regardless of whether he was an incumbent seeking reelection or not.

This point is important because it is essential for understanding the nature of the threat next time—on January 6, 2025, when Congress meets again under the Twelfth Amendment to count the electoral votes from the 2024 presidential election. If next time, unlike last time, the attempt to subvert the outcome of the election is successful, this subversion of democracy will have come from an abuse of legislative power in Congress itself, with no danger from abuse of executive power by an incumbent President. We can be confident in this truth not because President Biden, who will be the incumbent President then, would never attempt the kind of subversion of democracy that Trump wanted to perpetrate (although this contrast between the two may also be true), but because Biden—like Trump in 2020—will be in no position to use presidential power to subvert the January 6 congressional process. Instead, if there is going to be an antidemocratic coup on January 6, 2025, it will be perpetrated in Congress itself, by its own members (or, although perhaps less likely, by the incumbent vice president acting as Senate president).

To put the matter bluntly, what Josh Hawley and Mo Brooks wanted to do last time does not require an incumbent President (or even an incumbent vice president). It just takes enough senators and representatives willing to reject the actual electoral votes sent by the states and to replace these actual votes with a bogus substitute set of electoral votes purported to be cast in favor of the candidate who did not actually win. Hawley and Brooks did not have enough fellow members willing to go along with their legislatively perpetrated coup attempt on January 6, 2021. But what is the risk that there might be enough members of Congress willing to engage in this antidemocratic abuse of legislative power on January 6, 2025? It is to ward off this possibility that Congress should tie itself to the mast in advance, so that it maximizes the likelihood of resisting the siren song of a partisan power grab that secures the White House for one side’s candidate regardless of the choice the voters actually made.

What if the Coup Occurs in the States?

Democrats in Congress are currently nervous about limiting the body’s power to reject electoral votes sent from the states. They fear that in 2024 Republicans will be able to rig the results in the states and thus send to Congress fabricated outcomes that do not truly reflect the choice of the voters in those states. If Congress has bound itself to accept what the states send, then Democrats worry that they will have denied themselves the power to undo in Congress the falsification of results that Republicans may perpetrate in the states. As Chuck Schumer, leader of the Democratic majority in the Senate, put it: “If you’re going to rig the game and then say, ‘Oh, we’ll count the rigged game accurately,’ what good is that?”

The Democrats’ fear is predicated on a valid premise—there is a real risk that Republicans may attempt to subvert the result of the 2024 presidential election in the states, before the confirmation of the result reaches Congress. But the conclusion that Democrats draw from this valid premise is less persuasive. One need only to follow David Perdue’s campaign this year for governor of Georgia, in which he condemns incumbent Republican Brian Kemp for upholding the triply-confirmed 2020 tallies (including by means of a statewide manual recount), to worry that if Perdue were to replace Kemp as Georgia’s governor, he would refuse in 2024 to certify a valid victory in the state by the Democratic presidential ticket. Perdue has staked his entire campaign on his promise to be loyal to Trump, not loyal to the actual choices that voters make as an exercise of their right to self-government in a democracy. Perdue is hardly alone in this respect among the Trump loyalists running in this year’s midterms as part of an orchestrated effort to take control of the state-based procedures by which ballots are counted and vote tallies are certified. Jody Hice, for example, is running for Georgia’s secretary of state to replace honest vote-counter Brad Raffensperger on the same Trump-over-truth platform as Perdue.

Perdue, Hice, and their ilk present a real threat to democracy. The remedy to that threat is not, however, to empower Congress to review and invalidate the electoral votes that Georgia, or any other state, sends to Vice President Kamala Harris to be opened and counted on January 6, 2025. Instead, the remedy is to rely on the courts, including the federal judiciary, to prevent Perdue, Hice, or anyone else from brazenly endeavoring to subvert the result of what an honest and accurate count of the ballots in the state actually shows. The courts, both state and federal, demonstrated their fidelity to honest vote-counting in 2020, and there is no reason that they would not do the same in 2024. Yes, there are increasing pressures in America for courts to be political, but they are still the most rule-of-law and evidencebased institutions of American government. If there really is a situation in 2024 where it’s clear that the voters of the state want one result, but the governor (or other partisan elected official, like secretary of state) refuses to accept the will of the voters, then we can reasonably expect the courts to require the governor (or other official) to abide by what the voters have chosen.

We can expect this not because courts are committed to doing whatever democracy requires. Rather, we can expect this because it is what the law itself demands in the circumstance. The federal Constitution—the explicitly selfproclaimed “supreme law of the land”—empowers state legislatures to choose the “manner” of appointing each state’s presidential electors. There is a popular vote to appoint a state’s electors, in Georgia or any other state, only because the state legislature itself has chosen a popular vote as the state’s preferred manner of appointing its electors. (The practice in all states now is for each political party to nominate a slate of electors pledged to vote for that party’s presidential and vice-presidential candidates. When citizens cast their ballots in the November election, although their ballots list the name of the candidates themselves, by law these ballots are cast for the party’s slate of electors pledged to the party’s ticket. Thus, the ticket that wins the most votes causes, by operation of state law, its party’s slate of electors to become the officially appointed electors for the state.) Once the state legislature has chosen to use a popular vote as its method of appointing the state’s electors, rather than for example the legislature appointing the state’s electors directly itself, the rule of law requires that the appointment of electors be completed according to that chosen method. Thus, once the popular vote has actually been held as the state’s method for appointing its electors, the ballots must be counted honestly and accurately consistent with that legislative choice. If a governor or secretary of state, or other official, attempts to deviate from what the law requires in this instance—simply because the official does not like the result of the popular vote and wishes to repudiate it—the courts will insist upon compliance with what the state’s own law itself mandated for the appointment of the state’s electors.

The best strategy for making sure that the result of the election declared by Congress on January 6, 2025 accords with the will of the voters is, therefore, to require Congress to accept whatever results the courts themselves have determined to be required by the rule of law. If the governor, secretary of state, and other state officials associated with the vote-counting process simply follow the directives of the courts, or if there is no dispute over the outcome of the election that entails judicial involvement, then on January 6, 2025, the congressional treatment of that state can be straightforward: Congress accepts what is unambiguously the submission of the state’s electoral votes pursuant to the rule of law. But if the governor, secretary of state, or other vote-counting official attempts to repudiate what the courts themselves have declared the law to require regarding the result of the popular vote in the state—and if as part of this attempt have sent to Congress a submission of electoral votes from the state that purports to rival an alternative submission of electoral votes from the same state, one which corresponds to the appointment of electors according to the rule of law, as declared by the courts—then the goal must be for Congress on January 6, 2025 to count the electoral votes from the state that correspond to what the courts have declared reflects the rule of law regarding the appointment of the state’s electors, and to discard whatever rival submission is backed by a partisan governor or other official attempting to undermine the rule of law.

The Specific Risk of Electoral Subversion by a State Legislature

What if it is a state legislature—and not the governor, secretary of state, or some other state official—that attempts to repudiate the result of the state’s popular vote simply because the legislature for partisan reasons does not like the result? Given the authority of the state legislature to choose the manner of appointing the state’s electors, as granted by the federal Constitution, isn’t it dangerous for Congress to bind itself to accept whatever submission of electoral votes reflects state law as chosen by the state’s legislature? Wouldn’t this leave Congress, and thus the nation, vulnerable to a state legislature’s nullification of a popular vote that does not go its way?

No. A state legislature cannot repudiate the result of a popular vote when the legislature previously chose a popular vote to be the state’s method of appointing its electors for that year’s presidential election. To be sure, a state legislature, in advance of the election, could have chosen a different method of appointing the state’s electors, including direct appointment by the legislature itself (as many states did in the early years of the republic), and if a state legislature did choose that different method of appointing electors, then Congress would be stuck with that choice. There’s nothing in the Constitution that gives Congress the right to reject a state’s electoral votes because Congress disapproves of a state legislature’s chosen method of appointing the state’s electors. But if a state legislature already has chosen a popular vote as the method for appointing electors in a given year, and, crucially, if the popular vote already has been held for the purpose of appointing electors because that’s what the state legislature chose, then the state legislature itself is stuck with the choice it made, and Congress can insist upon counting the electoral votes cast by the electors appointed by the popular vote and at the same time reject any attempt by the state legislature to substitute a different method of appointing electors after the popular vote has been held.

The reason underlying this key point is that the federal Constitution gives Congress, not the state legislature, the authority to determine the time of appointing the electors. What Americans call “Election Day” is the date that Congress constitutionally has specified for the appointment of electors in each state. Thus, the state’s legislature is entitled to choose the method of appointing the state’s electors, but the appointment must occur on that date, as mandated by Congress. Therefore, when the state’s legislature already has chosen a popular vote as the method of appointing electors for that year’s electors, and when the popular vote has been held on Election Day in November as specified by Congress, then the appointment of electors has occurred pursuant to state law, and the state’s legislature is powerless to choose a different method of appointment for that year’s election because the time for appointment has passed and the appointment irrevocably has already been made.

After Election Day, it still may take some more time to complete the counting— and if necessary, recounting—of the ballots in the popular vote upon which the appointment of electors was based. In the technical language of the Electoral College process, this counting and potential recounting is part of ascertaining the appointment that already occurred in the casting of ballots that formed the popular vote that occurred on Election Day, the specified time for the appointment. But the fact that it may take days or even weeks to complete the ascertainment of the appointment does not negate the legal proposition that the appointment itself occurred on Election Day. More significantly, nothing about the fact that the ascertainment of the appointment requires additional time beyond Election Day itself gives a state legislature power to repudiate the appointment of electors that occurred on Election Day. Election Day remains when the appointment happened, and the state legislature is stuck with the appointment made by the popular vote that it chose as the method of appointment for that year’s election.

Thus, if in 2024 a state legislature were to attempt to do what’s impermissible and to try to rescind the appointment of its state’s electors by means of a popular vote after the popular vote already has taken place on Election Day in November, it is clear that the procedures that Congress adopts for counting electoral votes on January 6, 2025 should require Congress to count the votes cast by the electors appointed by that popular vote. Imagine that the procedures for counting the popular vote have occurred, and the state legislature does not like the result. The state legislature might attempt to nullify the popular vote and to appoint the state’s electors directly. The electors purported to be appointed directly by the legislature might meet to cast their electoral votes and then send those votes to Congress. But the original electors, the ones appointed by the popular vote, the result of which has been revealed by the procedures set forth by state law for counting the popular ballots, can also meet to cast their electoral votes and then send their electoral votes to Congress. In this situation, the rules and procedures for governing the joint session of Congress on January 6, 2025 should unambiguously provide that it is the lawful electoral votes cast by the electors appointed pursuant to the popular vote on Election Day that must be counted, because that appointment occurred at the time constitutionally required by Congress according to the method chosen in advance by the state legislature, and the contrary unlawful electoral votes cast by electors claiming to have been appointed after Election Day in a procedure not authorized by the Constitution must be discarded by Congress on January 6, 2025 as invalid.

There’s one additional element to this point about the inability of a state legislature to rescind an appointment of electors already made on Election Day by means of a popular vote. Federal law currently permits a state legislature to choose after Election Day a new method of appointing the state’s electors if but only if there has been a “failure” to achieve the appointment of electors by means of a popular vote on Election Day itself. The idea here is to give the state a second chance at appointing electors if something like a natural disaster (think earthquake or hurricane or the like) or a terrorist attack (as occurred on 9/11, the day of a primary election in New York) prevents holding the popular vote on the congressional designated day for appointing electors. But there is no “failure” to appoint electors by means of a popular vote just because the state legislature might dislike the result that counting the popular vote reveals. Thus, a revision to the relevant provision of federal law should make abundantly clear at a minimum that a state legislature cannot subsequently assert a “failure” to appoint by means of a popular vote whenever the state’s own chosen procedures for counting the popular vote, including any recounts or litigation over the counting of ballots, are capable of being used in the particular election. Disagreement with the result reached by the state’s own vote-counting procedures does not suffice to qualify as a failure.

The Key Role of the Federal Courts

For the reasons we have seen, the overarching approach that Congress should adopt for counting electoral votes should be to bind itself in advance to counting the electoral votes cast by electors whose appointment was made on Election Day in November pursuant to state law enacted by the state legislature before Election Day. Figuring out exactly what that state law may require regarding a particular election may necessitate involvement of the state’s judiciary. As part of the process of ascertaining the appointment made on Election Day by means of a popular vote, there may need to be litigation concerning the specific state laws for counting the ballots or how those state laws apply to particular ballots at issue. But whatever the resolution of this litigation, its conclusion will be the rule-oflaw answer to the question of which candidate won the popular vote in the state.

There is also potentially a role for federal as well as state courts to play in this litigation. While state courts undoubtedly must play the main role in determining what state law requires with respect to counting the state’s popular vote, issues of federal law may emerge from the counting process, as they did in Bush v. Gore, and as they have in a myriad of other elections (including the non-meritorious federal claims that the Trump campaign and its allies raised in 2020). Indeed, egregious misapplication of existing state law by the state’s own courts may itself violate federal law for the reasons spelled out in the Bush v. Gore concurrence (and reiterated by several current Supreme Court justices during the litigation over the 2020 election).

Thus, any revision of the rules for counting electoral votes that binds Congress to the judicially determined rule-of-law conclusion of which candidate won a state’s popular vote should appropriately recognize the potential role that federal courts may play in reaching this conclusive judicial determination. There is no need to create new federal-court jurisdiction in this regard; this jurisdiction already exists, as evidenced by Bush v. Gore, all the federal court decisions in the 2020 election, and numerous other precedents. Still, it would be valuable for Congress to acknowledge the existence of this federal court jurisdiction as part of its revision of the procedures for counting electoral votes on January 6, 2025. In binding itself to the judicially determined result of the popular vote in state, Congress would be recognizing the essential role of federal courts to assure that the counting of the state’s popular vote complies with all applicable federal-law requirements, including the equal protection and due process clauses of the Fourteenth Amendment.

Republicans as well as Democrats in Congress should value the confirmation of this federal judicial role. It is easy to imagine the federal judiciary ordering a state supreme court, as well as the governor, to comply with the counting of the popular vote that favors the Republican presidential ticket. Consider Pennsylvania, for example. Given what happened in that state in 2020, where the state supreme court extended the deadline for when absentee ballots must be received to be counted in a ruling that Republicans found objectionable (although it did not change the ultimate outcome of the presidential election in the state), we can hypothesize something similar happening in 2024—only this time it would determine which candidate won. The state’s courts and governor claim the Democrats to have won the popular vote, but the federal courts order otherwise. In this situation, Congress should be bound by what the federal court decrees given the supremacy of federal law under the Constitution.

But the principle does not favor either party. It is just as easy to imagine another situation in which the federal courts order counting the popular vote in a way that favors the Democratic presidential ticket, contrary to the ruling of the state’s judiciary and over the objection of the state’s governor. Return to our previous consideration of Georgia. It is easy to imagine the state’s judiciary ruling ineligible ballots that the federal courts require to be counted, and the governor insisting that the state’s courts must prevail over the contrary federal court determination. But if the U.S. Supreme Court ultimately holds that the ballots must be counted, then that is the conclusive rule-of-law determination that should be binding on Congress when it comes time to count the state’s electoral votes.

Congress Can’t Trust Itself to Resist to Temptation of Partisanship

Fair enough, one might acknowledge—let the judicial resolution of any disputes over a state’s popular vote, including any resolution achieved by the federal judiciary, prevail ordinarily. Even so, shouldn’t Congress still retain for itself some residual authority to overturn the outcome in a state if something seriously goes wrong? Consider it a kind of safety-valve or insurance policy. Congress presumes to let the courts decide, but keeps for itself the power to have the last word just in case.

Here’s where we must return to the story of Odysseus and the sirens’ song. If Odysseus hadn’t been tied to the mast completely and had retained the power to untie himself if he had wished, all would have been lost. It was only by successfully preventing himself in advance to succumbing in the moment that he was able to avoid the calamitous enticement of the sirens.

The same point applies to Congress and the counting of electoral votes. If Congress gives itself some sort of safety-valve authority to reject electoral votes it considers to be intolerably objectionable, this safety-valve authority is immediately susceptible to partisan abuse. As long as one party’s members can round up enough support to claim the need to invoke the safety-value provision, that party is capable of repudiating the result of the election based on a pretext.

Far better for Congress to rely on the judiciary to reach the right answer. No human institution, including the courts, is guaranteed to be perfect. There is unavoidably a risk that the courts will get it wrong. But that risk is far lower than the risk that Congress will abuse whatever authority it retains for itself to second-guess the conclusion that the courts have reached. If the goal is to have Congress count the state’s electoral votes in the way that most likely accords with the outcome of the popular vote in that state—which is what the goal should be—then that goal is best achieved by requiring Congress to count the electoral votes in accordance with what the courts determine the rule of law requires with respect to counting the popular vote, and not what members of Congress in the heat of the moment assert when they assemble two weeks before Inauguration Day.