You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Trying Political Leaders

Why George W. Bush is not like Louis XVI.


Trying political leaders: I do not mean trying them out, in advance, to see if we are likely to find their leadership disastrous, though that might be a good idea if we could find a way of doing it. In politics, judgment does not have to be, and often cannot be, after the fact. But it is post facto judgment that I wish to discuss: the morality and wisdom of putting political leaders on trial after we have endured their leadership and, perhaps, their crimes. In the aftermath of the Bush administration, this question has been furiously debated (for example, in Elaine Scarry’s recent book Rule of Law, Misrule of Men). But political trials have a long history, and the judgments we make of their procedures and verdicts have always been contested. So here is a comparative politics of political trials, aimed at illuminating the controversies.

I will consider four kinds of trials. All are from the modern period; I am not able to discuss the Greek and Roman precedents. The four are, first, revolutionary trials (such as those that took place in 1649 and 1793 in London and Paris); second, show trials that are part of a political purge (such as the Moscow Trials of the 1930s); third, post-war trials (such as those held in Nuremberg between 1945 and 1949); and, fourth, what I will call post-election trials (like those proposed for Bush administration officials). My perspective throughout is that of a political theorist, not a lawyer. The legality of political trials is often dubious, but that is not my central concern. It is possible for trials to be legally dubious but morally and politically right (and perhaps even necessary). It is also possible for them to be perfectly legal but radically imprudent.

Let us begin with the trials of Charles I and Louis XVI by the revolutionaries who overthrew them. These are critical moments in early modern political history—our political history, for we, citizens of a republic, are the heirs of these trials. Kings had been killed before; indeed, untimely death was a regular feature of kingship all through its history. But they had been killed by single assassins, like Henry IV of France, stabbed to death by François Ravaillac, or murdered in corners, out of sight, like Richard II in Pontefract Castle. When Oliver Cromwell said of Charles that “we will cut off his head with the crown on it,” he was aiming not simply at the death of a king but at the death of monarchy itself. And that required, he thought, a public accounting—an indictment and a trial that challenged the legal inviolability of kings and treated this particular king as a public official, answerable to the public, as capable of committing a crime as any other subject of the realm.

Not all the revolutionaries wanted to put the king on trial (and not all those who favored the trial wanted to kill him—but I am not going to get into the question of the death penalty here). The debate in France over what to do with Louis raised issues that re-appeared in the run-up to the Nuremberg trials, and they are worth considering in some detail. The Jacobins opposed a trial, arguing that Louis had committed no crimes to which the law reached. His only crime “was to have been king”—and this was obviously not a crime under the law of the ancien régime. “No one can reign innocently,” said Saint-Just, but in fact he was not interested in the guilt or the innocence of Louis. What he and Robespierre believed was that ruling as a king, reigning, was an act of war against the people the king ruled.

This was an unjust war, a war of aggression, but Louis was not first of all a criminal: he was an “enemy of the people” (I believe that this phrase has its origin in the French debates), and since he was dangerous even in captivity, military necessity required that he be killed. He could not be treated as a prisoner of war; the legal phrase “benevolent quarantine for the duration” comes later, but it describes very well what the Jacobins thought impossible in the case of kings (and what the Bolsheviks still thought impossible in 1917, when they acted on the Jacobin doctrine—though with their own special ruthlessness, killing the czar and all his family: no Jacobin suggested anything like that in 1793). Like Cromwell, Robespierre and Saint-Just wanted to kill the king with his crown on, but they were not interested in a public indictment. They thought that a trial was something owed only to citizens—and soon enough they would be putting a lot of citizens on trial. But they saw no need to make a case, to persuade men and women who had been ruled by kings for centuries that this king had betrayed them. It was self-evident that kings and patriots were in a state of war.

The underlying assumption of the Jacobin argument was that Louis was not a Frenchman—that is, not a member of the nation or a citizen of the state, and not a party to the social contract. Kings, and aristocrats too, belonged to an international class of parasites and predators, who took no responsibility for the well-being of the people they ruled because they were not “of” the people. And so they had no claim on the protection of the revolutionary state that the people had created. Once, a king of France had arrogantly proclaimed that he was the state (“L’état, c’est moi”); now the Jacobins proclaimed that the king was stateless. I am not sure that this description, despite its fierceness, was entirely inaccurate. Louis himself might have agreed that he never signed on to the social contract. Still, it was a dangerous description since it could easily be extended to other “enemies” of the people.

I much prefer the Girondins, who were the chief advocates of a trial for Louis, and who carried the day, though they did not get the trial that they wanted. They argued that Louis was indeed (and obviously) a Frenchman: he was Louis Capet, a public servant, who had committed the crime of treason by conspiring with and actively assisting enemy armies attacking the republic. And, of course, the law reached to this crime. Even under the ancien régime, treason was a criminal act, though it was thought then, mistakenly, that the king could not be a traitor (since treason was defined as a crime against himself). But the king had always been, in fact, an officer of the people, and treason was always a betrayal of the people and not merely of its chief officer. The real innovation of the revolutionaries was to deny the king’s inviolability: if the law’s reach extended to the crime, then, they insisted, it extended to the criminal. All Frenchmen were equal before the law. This was an ex post facto equality, but the Gironde was willing to live with that, since equality, its members thought, was natural and eternal—even if it was new to French law.

It followed from the Girondin argument that Louis should be accorded all the rights of a French citizen accused of a crime—that he should be represented by an attorney, and that he and his attorney should know the charges against him, see the documentary evidence collected by the prosecution, and be allowed to question hostile witnesses and to call friendly witnesses to the stand. “What justice truly demands,” argued Condorcet, the leading intellectual of the Gironde, “is that ... the general principles of jurisprudence in favor of the accused should be preserved or even extended.” This sort of thing, Robespierre responded, was a sign of “delicacy”—the term was disdainful; delicacy in the case of kings demonstrated nothing more than a lack of revolutionary energy. Robespierre and his friends called Louis “the last king”—their politics was resolutely forward-looking. Condorcet and his friends called Louis “the former king”—their politics aimed at a just reckoning with the past.

But the trial that the Girondins wanted was also intended to have consequences for the future. In a sense, it was a show trial and, indeed, a splendid spectacle: the king judged by his subjects, his peers! Louis Capet, ci-devant roi, standing before the court, exposed as a mere man, all the mythic masks of kingship torn off, but charged with crimes he had committed when he wore the crown—this could not help but be, Condorcet believed, a great moment in the education of the human race. Yet if it was to make this kind of “showing,” it had to be a real trial. Its educational value depended on its procedural integrity.

Many years ago, in a book defending the trial and the execution of Louis XVI, I adopted what was pretty much the Girondin position. It still seems to me to be the right view. “Revolutionary justice is defensible whenever it points the way to everyday justice. This is the maxim that marks off morally legitimate trials from proscription and terror. The legitimacy of the king’s trial lies first in its resemblance to an ordinary judicial proceeding. Condorcet was entirely right to argue that the nearer the resemblance, the better the trial: the more just for the king and, not inconsistently, the more effective for the revolutionaries. But justice in revolution cannot be measured in a purely formal way, for what is usually at stake is not the validity of this or that procedure—there was no disagreement in the Convention [the revolutionary assembly] as to what constituted a fair trial—but the status of the men and women involved. In the case of the king, what was at issue was whether he was justiciable at all. Here the whole purpose of the revolution was to change the law, not merely to apply it, and it is obvious that the proposed changes have to be defended in political and moral, not in legal, terms. Hence the second and more important defense of Louis’ trial is simply the claim that to turn kings into citizens is to advance the cause of justice.”


The trials that accompany political purges are also “show trials,” but their educational value depends on their lack of procedural integrity. The aim is to inspire fear among opponents and potential opponents of the regime, and so it is necessary that everyone watching the trial understand that what Condorcet called “the general principles of jurisprudence in favor of the accused” have no place here. The law offers no protection: if you are accused, you are dead.

But then why go to the trouble of organizing a trial? The “purge” metaphor does not suggest the need for any kind of judicial procedure: corrupt elements are simply removed from the body politic. Indeed, many enemies of the regime, in the Soviet Union in the 1930s, for example, were simply arrested and shot—and something like that has probably happened, though on a much smaller scale, in Iran today. “Disappearances” of this lethal sort certainly have an educational effect. But the organizers of a purge want something more than this, which they apparently think only a trial can provide. Considerable care went into the preparation of the Moscow Trials, and today we can see something of the same care in Tehran, where it is similarly difficult to understand. What point can there be in organizing a trial if the procedures that constitute a “trial” are ignored? Fear can obviously be produced in other ways.

It is the confession, I think, that is crucial in the purge trials. Procedural justice is rendered unnecessary by the abject confession of the accused. But these two have different purposes: procedural justice, in the case of Charles and Louis, was meant to convince the political public that kings were as liable to prosecution as any other citizens; they were public officials who had in fact committed crimes. Confession is meant to degrade and disgrace former revolutionary leaders suspected of “oppositional tendencies.” The charges and then the confessions are unbelievable, contradictory, absurd, and precisely for this reason the recitation by the defendant of his “crimes” is an act of public self-abasement and political self-destruction. He is not guilty so much as helpless, beaten, cowardly, irrelevant. The sabotage, espionage, and treason to which he confesses are inventions of the regime, which require a radical rewriting of history, and the trial’s message is that the regime can transform the past in any way it likes and is therefore in firm control of the future. It can do anything, whereas its opponents can do nothing but acquiesce. The trial provides an occasion for their ritual acquiescence to their own political oblivion.

But this cannot be the whole story. Stalin insisted that confessions be extorted, by torture if necessary, from tens of thousands of his victims who were never brought to trial—who were simply shot or transported to the camps without any public appearance. Their confessions were sent to the archives, marked “to be kept forever.” Robert Tucker offers a psychological explanation of Stalin’s insistence: insecure in his paranoia, Stalin needed this confirmation that the people he was purging were indeed enemies of the revolution—which is especially odd, since he certainly knew that the hostile acts described in the confessions had never taken place.

Everyone else should have known that, too. And yet the charges and the confessions that I just called “unbelievable” were believed by many people—not only faithful communists prepared to believe anything, and not only clever philosophers, such as Merleau-Ponty, who argued “dialectically” that even if the charges were factually false, they were politically true. Independent observers, journalists from many countries, liberal and conservative intellectuals, diplomats (such as Joseph E. Davies, the American ambassador) watched or read about the trials and concluded that the defendants must be guilty of something—as if political defeat was itself a sign of guilt. Hence it was really important to insist that the purge trials were not “trials” at all, for nothing was being tried: every outcome was known in advance, and all the evidence was made up. The failure of so many politicians and activists, especially (but not only) on the left, to join in this insistence is a lasting shame.

Of course, purge trials do bear some resemblance to real trials, else they would not “work” at all. Confession is a common way of giving up procedural rights in exchange for other benefits, sometimes life itself. But the resemblances are superficial, and it is still very hard to accept the fact that so many intelligent people were fooled by the Moscow Trials. The absurdity of the charges, the groveling confessions, the broken men in the dock—this was terrorism in the guise of justice.


Justice after a war is commonly called “victor’s justice.” It resembles the justice done (or attempted) by triumphant revolutionaries, and after World War II it prompted similar debates—though with one important difference. Many of those who opposed the establishment of the Nuremberg tribunals did so out of a strong commitment to procedural justice. Their arguments were different from those of the Jacobins, though their conclusion was similar: the leading Nazis should be shot as soon as they were captured. A fair trial, in accordance with British and American principles, was more than they deserved, but even more importantly, it was unlikely that any trial would be, or would be seen to be, fair.

The British War Cabinet worried that since “the result would be a foregone conclusion, these proceedings would tend to bring judicial procedure into contempt.” As Gary Bass writes in Stay the Hand of Vengeance, his fine study of war crimes tribunals (and my chief source here), the British “showed an unwillingness to compromise the integrity of their traditional judicial procedures—and a willingness to be blunt about [their unwillingness].” This bluntness is evident in Churchill’s proposal that the Allies produce a list of the fifty or one hundred leading Nazis and then when any of these persons fall into the hands of any of the ... armed forces of the United Nations, the nearest officer of the rank ... of Major-General will forthwith convene a Court of Inquiry, not for the purpose of determining the guilt or innocence of the accused but merely to establish the fact of identification. Once identified, the said officer will have the outlaw ... shot to death within six hours and without any reference to higher authority.

I should add that Churchill responded with high indignation to Stalin’s proposal at the Tehran Conference in 1943 that at least fifty thousand Germans should be proscribed and shot. “The British Parliament and public,” he told Stalin, “will never tolerate mass executions.” Churchill aimed to kill no more than one hundred; there would then be trials for all the “subordinate war criminals,” preferably in the country in which their crimes had been committed. Despite the designation of these people as “criminals,” the outcome of their trials was presumably not a foregone conclusion.

Churchill’s “outlaw” plays something of the same part as Robespierre’s “enemy of the people.” Both terms describe people who are somehow beyond the law. With regard to the Nazis, Anthony Eden wrote, “the guilt of such individuals is so black that they fall outside ... the scope of any judicial process.” Their fate, Sir John Simon, the lord chancellor, told the Cabinet, is “a political, not a judicial, question.” That was also the Jacobins’ view of the king’s fate, though they expressed it in quasi-military terms and without, obviously, the “delicacy” of the Cabinet’s legalism.

In the United States, Henry Morgenthau represented the Jacobin position, and Henry Stimson, with support from Felix Frankfurter, was the Girondin. He was working out the position that Robert Jackson defended at the London Conference, and then expressed with great eloquence in his opening statement for the prosecution in Nuremberg: “That four great nations, flush with victory and stung with injury, stay the hand of vengeance and voluntarily submit their captive enemies to the judgment of the law is one of the most significant tributes that Power has ever paid to Reason.”

The reason for the tribute was twofold. First, the Americans had a strong commitment—often called “legalistic,” though it seems to me deeper than that word suggests—to the value and the integrity of judicial procedures and, in contrast to British jurists, were optimistic about the possibility of sustaining procedural integrity in a post-war court. They believed, as Otto Kirchheimer wrote in 1961 in Political Justice, that victor’s justice “did not exclude judicial freedom of appreciation of the role of individual German leaders”—and, indeed, that freedom led to acquittals as well as verdicts of guilty. The Nuremberg results were not, then, a “foregone conclusion.” Second, they believed, much like Condorcet, in justice as moral education. Only a trial, Stimson argued, only the detailed testimonies and the massive documentation that a trial would require, would “demonstrate the abhorrence which the world has for [the Nazi system of terror] and bring home to the German people our determination to extirpate it and all its fruits forever.”

I am not going to argue that Nuremberg was a successful demonstration—just as I did not argue that the trial of Louis was an effective or a great moment in the education of humankind. Certainly, the impact of Nuremberg was resisted in Germany at the time (and in some quarters ever since), though I believe that the trials of Nazi war criminals have had significant, if hard to measure, effects—legally, politically, morally. Curiously, perhaps, they have had the greatest effect where they were legally most dubious. This is the argument that Judith Shklar made in her discussion of the Nuremberg trials in 1964 in her book Legalism. “Crimes against humanity” was the most innovative of the charges brought at Nuremberg; Kirchheimer called the legal formula “problematic,” and if you read the lawyers’ arguments in 1945 it was certainly that. But Shklar predicted, correctly, that this was the charge (with “genocide” later incorporated into it) that would resonate over the years. In the sixty-five years since 1945, by contrast, nobody has been brought to trial for the crime of aggression, and no such trials have been seriously contemplated, though the criminality of aggression was established in international law well before Nuremberg, by the Kellogg-Briand Pact in 1928.

It is astonishing to remember that there were lawyers and jurists in Britain, the United States, and France who thought that the Nazi leaders could be tried only for “crimes against humanity” committed in the occupied countries. They argued, as Bass reports, that “the killing and persecution of German Jews” was “protected by German sovereignty”—not permitted or justified by sovereignty, but “protected” in the sense that they did not fall within the jurisdiction of an international court. This claim was rejected by the prosecution, whose argument eventually won the day. International jurisdiction was explicitly affirmed in the Genocide Convention in 1948 and in the Rome Treaty in 1998: sovereignty does not extend to mass murder. Today it is widely (though not universally) accepted that a government engaged in the massacre of its own citizens can be overthrown by foreign military forces—and that its officials can then be brought to trial in either a national or an international court. This liability was first defended in the Nuremberg trials, which pointed the way, like Louis’s trial, to “everyday justice,” though the examples of Rwanda and Darfur suggest that we are not there yet.


Imagine the following case. A group of political leaders, who have been in power for almost a decade, lose an election. Many supporters of the winning party believe that the policies of the defeated leaders were criminal, because they involved the violation of specific laws, domestic and international. These victors want to put the defeated leaders on trial, and this, they claim, is a perfectly reasonable and entirely legal thing to do. It is not revolutionary justice, since the turnabouts in democratic elections are not revolutionary: the regime is not being changed, only the government. The constitution, the whole legal system, is intact; no new understandings of agency and responsibility are required to justify the trials that are now proposed.

Nor would this be “victor’s justice” as in the Nuremberg trials, because competitive elections are not like wars, even when we use the language of war to talk about them (we run campaigns; we try to outflank our opponents on the right or the left; we design strategies and tactics; we maneuver for advantage; and so on). But we are not, in fact, fighting. Moreover, post-election trials would not be a case of “victor’s justice,” because the trials would be conducted by independent judges who are not newly installed and who are not active members of the winning (or the losing) party. The decision to bring charges, or to launch an investigation that might result in charges being brought, would have to be taken in the United States by the attorney general, who is a political appointee (and whose appointment is politically very sensitive). Still, the trials themselves could be and would be, so their advocates claim, insulated from partisan politics. Nor, finally, would the proposed trials be anything like a purge. The charges would not be absurd, and the defeated leaders would not be tortured. Their families would not be threatened. They would not be forced to confess. Indeed, it is most unlikely that they would even think about confessing.

If revolutionary trials and post-war trials can be distinguished from purges, and if they can be justified, morally and politically, as I have tried to suggest they can be, then surely, all the more so, can post-election trials be justified. In fact, however, this a fortiori argument does not work, or rather, it works the other way. It is easier to justify the trials of political leaders after the major break of a triumphant revolution or a victorious war than it is after the minor break of a democratic election. Trying kings for treason, or prime ministers and war ministers for aggression or “crimes against humanity,” may involve understandings of treason, aggression, and humanity different from those that once prevailed, but these understandings have been vindicated by revolution and war. The trials enact a new consensus about domestic or international justice. But there is no new consensus, about the foundations of the political order or about the policies of the last administration, after a democratic election. Instead, the defeated party, in opposition now, defends what it did in government, insists on the rightness of its own ideology, and plans for the time when it can return to power and do the same things again.


Democracy depends on the willingness of political leaders to go into opposition—and on their expectation that they might one day return to power. We call this rotation in office, and though it does not occur on a regular schedule, it does regularly occur. By contrast, kings and republican politicians do not alternate in office—first Louis XVI, then Robespierre, and then Louis again (actually the restoration and the revolutions of 1830 and 1848 did produce a kind of rotation, but not as a settled constitutional arrangement). Nor did we want, after World War II, to see a Nazi regime ever again in power: first Hitler, then Adenauer, then Hitler (or Göring) again. Endings such as those of 1793 (the year of the king’s execution) and 1945 are meant to be definitive; but endings such as those of 2000 or 2008, when one political party and then the other was toppled from power, are not in any sense definitive.

The truth about democratic elections is that the stakes are fairly low, compared to the stakes in revolutions and wars. Critically important questions about public policy are (sometimes) decided democratically through party politics and election campaigns; but they are rarely decided in any conclusive way. The argument goes on after the decision has been made, and so do the arguing politicians. Perhaps if the electoral defeat is decisive, the defeated leaders will retire from politics (though usually with the hope of being called back), but their place will be taken by younger leaders with many of the same ideological commitments. This is the sense in which the stakes are low: no one expects to be killed or imprisoned after they lose an election.

Aristotle speaks of the relationship among democratic citizens as “friendship.” That may seem to us too strong a term, but there does have to be a sense among the citizens that they are engaged in a common project, and some readiness (however frayed it may now seem) to recognize the good faith of fellow citizens who disagree about how that project ought to be pursued. Of course, we also recognize individual greed, power lust, and ideological zeal that seem to leave no room for the common project, but with regard to critical policy debates, it is important to be able to assume that all the parties have the general welfare in mind. It is not a disaster if our opponents win this time, because they will not always win. This is how democracy works: all political victories and all political defeats are partial and temporary. They do not put our common project in radical jeopardy. As the saying goes, you win some, you lose some.

So we should think very carefully before we do anything that raises the stakes of democratic politics—and putting your opponents on trial after they lose an election would certainly raise the stakes. There is a narrowly prudential argument against doing that (in addition to the broader prudential argument that I have been making): if we raise the stakes for them, they can raise the stakes for us. Paraphrasing Saint-Just, we might say that nobody can rule innocently: there will always be reasons for a trial. There will always be political maneuvers or policy decisions that violate the law or that can be made to look as if they violate the law. And once the game gets serious in that way, anyone holding office would have a very strong incentive to do whatever was necessary to win the next election. Winning some and losing some would then look like a fool’s politics. It would be very dangerous, I think, to start down this path.

The opposing argument, of course, is the argument of the Gironde: political leaders are not, and should never be, inviolable. We are a political community of equal citizens; no one is exempt from the law. This is a very strong argument, and I do not have a principled response. I want only to suggest that the advocates of absolute justice do not understand how fragile democratic politics is—even in the world’s oldest democracy—and how dependent it is on the willingness of incumbents to leave office. In any case, trials are not our only recourse. There are also congressional investigations, truth commissions, and legislative action (to pay reparations, for example, to victims of the former administration)—all these are political, not judicial, responses, and they figure in the ongoing practice of democracy. Indeed, truth commissions aim at reconciliation, which is the best guarantee of “ongoingness.” Reconciliation certainly seems unlikely in the current political climate, but it is a far better democratic ambition than the imprisonment of your opponents.

I do not mean to deny the responsibility of political leaders for what we might call personal crimes. Not at all. A president who takes bribes (or who commits rape or murder) should be impeached or defeated and put on trial, without regard to his or her office. But it makes sense to treat political crimes differently. Impeachment should be possible here too, but trials are best avoided. When Lincoln suspended habeas corpus, and when Franklin Roosevelt ordered the internment of Japanese-American citizens, and when Bush and Cheney authorized the torture of suspected terrorists, they were acting for what they thought was the public good, not for any personal good. And the best way of dealing with actions of that sort, if we believe them to be wrong, is to “throw the rascals out” in the next election and find some decisive way to repudiate their policies.

But we might reconsider here the British position in 1945—which may be more relevant to post-election than to post-war justice. The fate of democratically elected leaders is indeed, in Sir John Simon’s words, “a political, not a judicial, question.” It has to be worked out, one way or another, through the political process. But the people that Churchill called “subordinate” criminals could conceivably be brought to trial—for specific illegal actions—without dangerously raising the stakes of democratic politics. I am very uneasy with this position, since it violates the principle of equality before the law. But the violation is common—consider the trials that followed the My Lai massacre and the abuse of prisoners at Abu Ghraib; and it may serve the rule of law, even if inadequately. It targets only people who are not subject to democratic (that is, electoral) repudiation.

So perhaps we are not a community of absolutely equal citizens. (Absolutism is always a problematic position.) If I capture and torture somebody, I should be tried and punished. If the president orders that done, not just to somebody, but to many bodies, acting, so he says, in the name of national security, his only punishment is political defeat: we should organize in opposition to his policies and vote against him as soon as we can. But there is a line here that must be carefully drawn. If the president’s violations of the law or the Constitution aim at the creation of a tyrannical regime, if it is his own political enemies, not actual enemies of the republic, that he is arresting and torturing, or if he is systematically confusing these two, then we have to find a definitive way to defeat him or to overthrow him. And as part of that process, or after it, a full-scale trial might well be the best available means not only of doing justice, but also of demonstrating, to paraphrase Henry Stimson, our determination to extirpate tyranny and all its fruits forever. But we have not yet had trials like that in the United States, and we should not imagine holding them until necessity forces our hand. May this never come to pass.

Michael Walzer is Professor Emeritus at the Institute for Advanced Study at Princeton University. A version of this essay was delivered as a Dewey Lecture at the University of Chicago Law School

For more TNR, become a fan on Facebook and follow us on Twitter.