How to deal with a political leader who has broken the law is one of the oldest questions in the history of political thought. Among the Athenians, the first people to experiment with democratic government, political crimes as severe as treason and as trivial as proposing a bad law were treated with the same severity. Punishments, often severe, ranged from ostracism – whereby one was stripped of citizenship and exiled – to execution.

Today, we are less punitive in political temper. Modern views are closer to those of the French revolutionary statesman Emmanuel-Joseph Sieyès, who, himself having participated in the regicide of Louis XVI, lamented in 1793 that ‘it will always be a great misfortune to see the head of a great official fall to the sword of the law’.

So disquieted are we by the idea of political trials that we only rarely think about what it means to subject politicians to standards of law and justice. We think of ‘political tribunals’ as tools of arbitrary despotism and the punishment of legitimate dissent, which conjure up images of politicians persecuted by their opponents before sham courts. But politicians can, and do, commit crimes in office, often in furtherance of their political ends. When such cases arise, our understandable hesitancy about granting the state the capacity to punish political figures can lead us to prevaricate about what can or ought to be done.

Recent efforts to prosecute Donald Trump have spurred significant interest in these questions in the United States – and revealed how little has been done to establish firm answers across the country’s tumultuous history. Although two other presidents, Richard Nixon and Ronald Reagan, faced threats not only of impeachment but of being put on trial after leaving office, one was pardoned pre-emptively and the other was found innocent by two official inquiries. In a controversial recent ruling, the Supreme Court has found that the president is immune from prosecution in his ‘official capacity’, although this does not mean, as some commentators have suggested, that the president is wholly unaccountable and above the law, since, after all, the method of presidential accountability favoured by the Founding Fathers – impeachment – remains on the table. At any rate, the boundary between ‘official’ and ‘unofficial’ acts is so ambiguous, as to leave as many questions unresolved as solved when it comes to holding sitting or former presidents to account.

The question of holding the president to account troubled the Founders deeply. Even impeachment, now so familiar to the American public, roused controversy during the Constitutional Convention in 1787. Gouverneur Morris, a friend of Alexander Hamilton and fellow delegate to the Convention from New York, warned his colleagues that the impeachment mechanism ‘would render the Executive dependent on those who are to impeach’ and transform impeachment into a tool of partisan persecution. Bill Clinton might have taken some sorry comfort from Morris’ remarks when he suffered such a fate in 1998. Nor could the Founders settle originally on the mechanism for impeachment. Madison and Washington’s ‘Virginia Plan’ suggested impeachments be conducted by the Supreme Court, while the rival ‘New Jersey Plan’ proposed that the removal of the president be dependent on a majority of the state governors. It was only quite late in the day that the mode eventually adopted, in which impeachments are initiated by the House and tried by the Senate, where a two-thirds majority is required for conviction, was proposed by Hamilton.

What seemed to sway the Founders more than anything was Benjamin Franklin’s chilling prophecy that, without some mechanism within the constitution to hold the first magistrate of the republic to account, then the people might turn back to the political brutalities of the past. There, perusing the history of all prior republics, Franklin warned that they might readily take inspiration from those occasions on which ‘recourse was had to assassination in which [the head of the executive] was not only deprived of his life but of the opportunity of vindicating his character’.

In both Athens and Rome, unaccountable leaders had been put to death in the absence of institutional mechanisms of accountability. But while the ancients had lauded so-called tyrannicides, the Founders recognised the perils of assassination and of the breakdown in political order it would herald – a republic gripped by political violence and the prospect of civil war. The stark relevance of Franklin’s warning across the centuries needs no elaboration. Impeachment, as Morris came to recognise, was a perilous process, but one which provided an outlet for far more dangerous political passions.

This power has been used only four times by Congress against three men, Andrew Johnson, Bill Clinton, and Donald Trump, although the threat of impeachment pushed Richard Nixon from power. In none of these cases was the occupant of the White House removed by the Senate, although Andrew Johnson was found guilty by the majority of senators and acquitted by just one vote. In general, impeachment has been seen as a last resort, and impeachment votes conducted along strict party lines, just as Morris feared.

Attempts to find other methods of holding presidents to account have enjoyed far less success over the long arc of American constitutional history. Although Democratic President Andrew Jackson was formally ‘censured’ by the Whig majority in the Senate in 1834 over his attempts to undercut the Second Bank of the United States, when the Democrats gained the majority in 1837 they quickly voted to ‘expunge’ the censure after a gruelling 13-hour debate that ended with a vote along party lines. Jackson’s censure only further intensified partisan polarisation between the Democrats and Whigs, and attempts to censure presidents have not been seriously attempted since. The only legacy of the whole affair was a brief attempt by supporters of Trump to ‘expunge’ his multiple impeachments in the House of Representatives, backed by the then Speaker of the House, Kevin McCarthy.

In fact, there has only been one attempt to prosecute a former head of state in the history of the United States – and he was not a president of the United States at all, but the first and only president of the Confederate States of America, Jefferson Davis.

The effort to prosecute Trump has inspired renewed interest in the postwar debate around the prosecution of Jefferson Davis. As the historian Jill Lepore has argued in the New Yorker: ‘Sometimes it feels as if the century and a half separating the trial of Jefferson Davis from the trials of Donald Trump were as nothing.’ Similar comparisons have been advanced by Jeffrey Abramson and Dennis Aftergut in Salon, and by Joshua Zeitz in Politico, but these have missed the profound dissimilarity between the major legal and constitutional questions that pertained to the trial of Davis and those of the trials of Trump.

While efforts to prosecute Trump relate to a number of specific, concrete, charges, in 1865 it was not entirely clear what Davis could even be charged with. Initially, there had been some speculation that Davis might be put before a military tribunal for violations of the laws of war, particularly in connection to his suspected involvement in Abraham Lincoln’s assassination, or in the mistreatment of Union soldiers at Andersonville prison during the war, but neither case could effectively be made. Davis was soon found to have been uninvolved in the assassination, and the governor of Andersonville, Henry Wirz, refused the commutation of his sentence in exchange for implicating Davis in crimes against prisoners of war.

The only charge against Davis which could be prosecuted was treason against the United States. On the surface, this would seem an easy case for any prosecutor. Davis had, after all, led a government which had waged war on the United States for four years at the cost of more than 300,000 Union soldiers’ lives. Article III Section 3 of the Constitution’s definition of treason as ‘levying War against [the United States], or in adhering to their Enemies, giving them Aid and Comfort’, suggests Davis’ was almost the archetypal case.

Treason can only be committed by a citizen of the prosecuting state and Davis, his supporters, and later his lawyers, argued, had not been a citizen of the United States when he took office as president of the Confederacy, but a citizen of the state of Mississippi, which had already seceded. Drawing on the ‘compact theory’, which held that the states were sovereign entities who had joined and could leave the Union at any time, the Confederates had always maintained that their secession had been entirely legal and, therefore, that they had ceased to be subject to the US Federal Government from the moment that their legislatures willed it. On this basis, as the historian Cynthia Nicoletti puts it in her book Secession on Trial: The Treason Prosecution of Jefferson Davis, Davis’ defence could contend that secession was constitutionally permissible, and, since his state had left the Union in 1861, Davis was no longer an American citizen who owed a duty of loyalty to the United States. He was thus incapable of committing treason against the nation.

It was the complicated legal dilemma which this argument aroused that made prosecuting Davis so difficult, because if Davis was found innocent at trial, then it stood to reason that it could only be because secession had been legal and the Confederacy’s cause legitimate.

Davis himself was among the first people to recognise that this line of defence could secure both his freedom and vindicate the cause of the Confederacy in the eyes of posterity. As he would later argue in his 1881 book, The Rise and Fall of the Confederate Government, ‘an elaborate argument or plea in behalf [sic] of Jefferson Davis in particular, and of the Southern Confederacy in General’, as the Atlantic put it at the time, his actions had been neither treacherous nor illegal, because his home state had legitimately seceded and he had therefore ceased to be a US citizen long before becoming Confederate president. Davis was eager for a trial and willing to risk execution, should he be found guilty, in order to bring the issue before the courts, despite the strenuous objections of his wife Varina and his principal lawyer, the Irish-born New York attorney Charles O’Connor.

A defeat for the government in a treason trial for Davis could thus have catastrophic implications and ‘had the potential to shatter the fragile postwar settlement’, in Nicoletti’s words. On the one hand, as Columbia Law Professor Francis Lieber speculated at the time, if Davis was restored to his full citizenship, there would be nothing to prevent his return to the US Senate, or even to the presidency. On the other, it was possible that charging Davis with treason would plunge the country back into civil strife as every former Confederate, from the highest cabinet official to the lowest private in the defeated army, faced treason charges and the gallows.

If Davis was found guilty and executed for treason, particularly if he was prosecuted before a military rather than a civil tribunal, there was the possibility of renewed violence between North and South. Northern politicians, Robert Icenhauer-Ramirez argued in his book Treason on Trial: The United States v. Jefferson Davis, feared that Davis would become a Confederate martyr.

Such were the perils of the case that the Johnson administration ultimately demurred from pursuing a prosecution at all, running down the clock until, on Christmas Day 1868, Johnson issued ‘a full pardon and amnesty for the offence of treason against the United States’ to all former Confederates. Instead of dragging Davis, and other leading Confederates, before the courts, the Federal Government contented itself with a political solution. The amnesty would close off the legal issue of culpability for secession, while the strict enforcement of a ban on former Confederates holding political office enshrined in the 14^th Amendment would prevent Davis and his former compatriots from returning to the political arena. This, too, was overturned in 1872 by the ‘Amnesty Act’, with the exception on prohibitions on officeholding for those who had served in the United States Government at the time of the Civil War and had defected to the Confederacy, including Davis himself.

The various prosecutions facing Donald Trump, and the constitutional questions which they raise, are very different in nature. In Davis’ case, it was never at issue whether the president of the United States possesses immunity from prosecution or whether only the path of impeachment laid out by the Founders could be used to hold the president to account, because Davis was not a president of the United States. All that was at issue was whether or not the power exercised by Davis was lawful or treasonous.

In Trump’s case, by contrast, there is no doubt that he held office legally. Instead, he stands accused of various crimes while holding office and exercising the executive power legitimately, particularly in the case of his – indefinitely postponed – trial relating to the 6 January 2021 attack on the US Capitol, his trial relating to efforts to pressure Georgia’s secretary of state to ‘find 11,780 votes’ for him in the 2020 election, and his trial for mishandling classified documents under the Espionage Act. Further, the question of whether or not the former president can, or should, be tried revolves around a different set of constitutional questions about presidential immunity for actions committed while in office, and issues about the separation of powers which arise from them.

These are questions which have only partly been resolved by the decision given by the Supreme Court in Trump v. United States on 1 July 2024. This leaves open the thorny question of where one draws the boundary of ‘official actions’ for the courts to decide in the no doubt torturous bout of trials and appeals to come.

And yet, Davis’ case can still offer important political insights about what it takes to punish a former officeholder who still retains the loyalty of a significant section of the country. In both cases, constitutional and legal arguments were only one aspect of the contentious issue of whether or not it was politically prudent to put the leading politician of half the nation before an ordinary court, wherein it would be hard to find an impartial jury and from which any verdict would be essentially suspect. Although it was almost universally condemned at the time, the same might be said for Gerald Ford’s decision to pardon his predecessor Richard Nixon in September 1974, which, Ford claimed, was the only way to bring to an end a bitter and divisive crisis.

Whether the trial of a political leader is fair or not, both supporters and opponents of the individual on trial will always be quick to condemn the verdict if it does not go ‘their way’. The entire process will gain the colour of political persecution even if conducted entirely within the bounds of normal procedure. This problem, over which Johnson and his cabinet agonised in the 1860s, haunts our own debates on the feasibility of prosecuting and even imprisoning a former president today.

Nor can it be ignored that, in both Trump’s and Davis’ case, some of those who supported prosecution did so less in the interest of justice than in the hope of politically neutralising a dangerous opponent, as Lieber hoped Davis’ prosecution would achieve. In neither case does this delegitimise the case for prosecution, but in both it raises questions about whether the courts and the justice system are appropriate venues for the pursuit of political ends and about whether ordinary justice is corrupted even by the assumption of political motivations. In the abstract world of law, all criminals must be put on trial, but what happens if falsely assumed illegitimacy casts doubt upon the impartiality of the entire system?

It was these pragmatic concerns that, among other reasons, led the Federal Government to pursue a political rather than a legal course of action against Davis and his fellow ex-Confederates, lifting the judicial sword of Damocles from above their heads with one hand while imposing a ban on future political activity with the other. Davis never appeared before any federal or state court to answer allegations of treason. Instead, he was free to return to private life and the production of apologia for the Confederate cause, unpunished for helping to tear a nation apart.

By failing to prosecute Davis, the United States government prevented the constitutional issue of secession from being re-opened and closed off the possibility of a return to civil strife – but it did so, many felt, at great cost to the sacred notion that in America the law alone was king. Can this morally unjust yet politically necessary compromise offer guidance to Americans today? Perhaps. But the decision to pardon Davis was made possible in no small part because the Civil War had been so decisively settled on the battlefield. By the time he was pardoned in 1868, Davis could no more hope to return to office than he could change the course of affairs of the Confederate surrender at Appotomax in April 1865.

Therein lie the further limitations of the analogy between the trial that never quite was of Jefferson Davis and the trials of Donald Trump. Davis, unlike Trump, never attempted any kind of return to power, possessed no potential means of self-exoneration, and could not hope to use the powers of the presidency to avoid punishment. While much can be learnt from those who re-made America after the Civil War, it is not clear that their solution to the problem of prosecuting a president can save the United States today. Regardless of what one thinks of the legality or probity of prosecuting Trump, the bitter reality is that it is unlikely that justice will be furnished until the political struggle that colours it is brought to an end on the electoral battlefield. In a democracy, ultimate accountability can only flow from the voters, and our willingness to throw our lot in with majority rule leaves open forever the possibility of popular injustice – but at what cost the alternative?