Impeachment clauses and the different meanings of ‘constitution’ in the United States of America

From an article published in La Responsabilité du chef de l’État, coll. ‘Colloques’, vol. 12, Paris, Société de législation comparée, pp. 23-54


The purpose of the present article is not to review all the aspects of executive accountability in the United States. Rather, I will attempt to make a few steps towards a better understanding of the removal procedure applicable, among others, to the President of the United States and that is commonly termed “impeachment”. Besides, impeachment is so central to the constitution that it commands the overall interpretation of the accountability of the US President. From this standpoint impeachment exerts both an attractive and an inhibitive effect. It is attractive because the political will to bring the President to account must generally fit into the mould of impeachment. It is inhibitive because impeachment has such solemnity, such gravity about it that it is not an instrument for everyday use. Attempts to inflict lesser sanctions or disapprobation on the President are halted by that solemnity and that gravity. From a dogmatic standpoint, all of that is capture in the fact that impeachment is approached by lawyers as a procedure of an exceptional nature.


The purpose of the present article is not to review all the aspects of executive accountability in the United States. Rather, I will attempt to make a few steps towards a better understanding of the removal procedure applicable, among others, to the President of the United States and that is commonly termed ‘impeachment’. Besides, impeachment is so central to the constitution that it commands the overall interpretation of the accountability of the US President. From this standpoint impeachment exerts both an attractive and an inhibitive effect. It is attractive because the political will to bring the President to account must generally fit into the mould of impeachment. It is inhibitive because impeachment has such solemnity, such gravity about it that it is not an instrument for everyday use. Attempts to inflict lesser sanctions or disapprobation on the President are halted by that solemnity and that gravity. From a dogmatic standpoint, all of that is captured in the fact that impeachment is approached by lawyers as a procedure of an exceptional nature[1].


1.1 Interpreting the impeachment clauses

In the 1787 Constitution, lawyers identify a series of provisions known as ‘impeachment clauses’. These include Section 4 of Article II of the Constitution with which it is customary to begin.

This text provides that ‘The President, Vice President and all civil Officers of the United States, shall be removed from Office on Impeachment for, and Conviction of, Treason, Bribery or other high Crimes and Misdemeanors’.

The best starting point here consists in observing that the Constitution creates a removal procedure without specifying whether it is ‘criminal’, ‘political’ or whatever else. The constitution, il should be noted, is silent about the nature of the procedure. That silence is deeper still. Identifying a set of ‘impeachment clauses’ within the Constitution is a fairly elementary synthetic operation but already a step forward with respect to the Constitution’s initial neutrality. The Constitution does not construe itself. Yet it is not possible to stop there and say nothing about a legal rule. It cannot remain in this state of non-being to which lasting silence would condemn it. Law is not properly speaking like ‘a presumptuous and ignorant man, who does not let anyone do anything other than what he himself has decided, nor anyone put any question to him’ as in Plato’s metaphor (Laws, 294). It is more exactly a silence that is only changed into words through the interpreter’s voice. Having become law – the dead letter of the law’ – the word of the constituent power has returned to silence. Only when the word of the interpreter takes over will that silence will be broken. It will not be broken arbitrarily in a wholly indeterminate space. The language of the constitution acts as a constraint on the interpreter. But they are not the only constraints on him/her. Maybe we should use the word constitution’ to denote not only the entrenched constitution but also all the constraints that surround the interpretation process to which the entrenched constitution is subjected.

So, to return to the matter before us, there is no dividing line between what is in the rule, its intrinsic meaning as it were, and what is incorporated in it or ‘added’ to it by the interpreter. In a case like the one before us, it can readily be seen that anything that can be said, and that is usually said, is based on the adjunction of terms, qualifications and ways of thinking that are external to the actual text of the 1787 Constitutional Act. If we were to keep to material that was exclusively ‘positive law’, we could not even say that the US Constitution contains ‘impeachment clauses’ or ‘a removal procedure known as impeachment’. Once it is accepted that it is important to say what it is about, to find a meaningful name, even if it is a fairly abrupt sort of attribution of meaning that happens when an object is placed into an abstract category, the difficulty arises from the commonly used options being barely satisfactory. Can we speak of ‘accountability’? The answer is probably yes because the idea of accountability is so vast that it can cover any operation by which a competent authority holds a subject of law to account with legally identifiable consequences. Does the procedure in question fit into the category of criminal liability, or that of ‘political’ accountability? The question is an open one. It is self-evident, besides, that we are not bound by these options. For that matter, their respective implications are not sharply defined anyway.

When it comes to interpreting the impeachment clauses, two observations of a more general order relating to the problem of constitutional interpretation deserve to be made. The first is also the more general of the two. It is particularly interesting not to treat the problem of interpretation ex ante, that is, from the position of an abstract interpreter on the verge of making an interpretation. It seems more fruitful to reason ex post, envisaging interpretations that actually have been made. I shall try to lend some plausibility to this approach by analysing the impeachment articles like samples of such concretely made interpretations. These articles are texts that it is helpful to apprehend as a whole, if only to observe their cumulative nature, the propensity to piling up and combining assertions. The second observation is an application of the first. In the samples of interpretations of impeachment articles we have, we are led to detect different languages, that is, different forms of presentation of law[2] that are each expressed with their own vocabulary, and that correspond to fairly distinct registers of argument. Thus in the case of impeachment, we shall come upon the register of the separation of powers (impeachment as a mechanism of checks and balances), the register of judicial reading of the Constitution (impeachment as Congressional competence in the context of a ‘Constitution of enumerated powers’) or of ways of speaking that are separate from these first forms. These registers sometimes combine, which brings me back to my first observation. From this point, I shall confine myself here to outlining three series of problems of interpretation posed by the impeachment clauses: (1) Who can be subjected to impeachment? (2) To what extent and in what way does the Constitution determine the behaviour of those who intend to bring an accusation? (3) What does ‘removal’ mean? I shall then have come to the point where we shall have to look beyond the impeachment clauses and question the meanings to be given to the idea of Constitution with a view to understanding these clauses.


I shall make some remarks later about what Section 4 of Article II says about which people the removal procedure may be aimed at. Notice only here that, from the standpoint of the addressees, it does not concern the President alone but also the Vice President and the civil officers. The delimitation of its scope is itself fairly mysterious. It is a halo of mystery rather than a sharp line. In particular, it is accepted that federal judges, including Supreme Court judges, come within its scope. It is also accepted that only certain holders of high office of the federal state are concerned. Why? That cannot be deduced from the impeachment clauses themselves. It arises from a sort of interpretive convention. Such conventions are very common in impeachment law. It might even be said that this law is paved exclusively either with questions for which there are no answers or with answers consisting only in stating such constructive conventions. Controversy has arisen over just about every possible question, with each party being able to assert reasonable arguments.


The law of impeachment contains certain terms characterizing the allegations that may be brought, let’s say (in neutral language) the offences that may come within the scope of the removal procedure. These are ‘Treason, Bribery’ and ‘other high Crimes and Misdemeanors’. To these could be added, if we were trying to extend the palette of references, even more varied formulas drawn from state constitutions and relating to comparable removal procedures. And added to this is what is made of these formulations in the impeachment articles adopted by the House of Representatives, not to speak of the impact on their interpretation of the Senate’s votes of guilty or not guilty. It can be seen that, depending on the standpoint we take up, the bonds tying the interpretive process can just as easily be seen as loopholes through which the interpreter can escape. In other words, trying to define the act of interpretation as entirely free raises as many problems as defining it as entirely unfree. Trying to apply a coefficient of freedom (including infinity) to anyone in the position of interpreter is a task that can genuinely be thought to be a dead end.

What is to be done with these terms? There seems to be little discussion that impeachable acts are not limited to the characterizations of criminal law, in any event as far as ‘high Crimes and Misdemeanors’ are concerned. Except for ‘Treason’, which is defined elsewhere in the Constitution[3], and perhaps for ‘Bribery’ (for which, here too a sort of constructive convention points to the 1790 Bribery Act)[4], there is no principle of strict definition of offences in law[5]. It is to be observed, for that matter, that all of the impeachment articles against a US President pertained to the commission of high crimes and misdemeanours. Never have accusations of treason and bribery been levelled against a President in impeachment articles passed by the House of Representatives.

In some respects, the terms employed by the Constitution have the same determinative effect as the statutory elements that constitute a criminal offence. So, the drafter of an impeachment article does not have the power to create new labels, say by adding ‘Robbery’ to ‘Treason, Bribery, high Crimes and Misdemeanors’. This restriction on the grounds for impeachment comes from outside. The Constitution being a ‘Constitution of limited powers’, this bears on the clause in question: through it, the founding fathers did not grant an unlimited power to impeach and condemn, but a limited one[6]. This understanding was to command the common reading of the works of the Philadelphia Convention. We cite the terms in which Madison criticized Mason’s proposal to extend removal to take in ‘maladministration’: ‘so vague a term will be equivalent to a tenure during pleasure of the Senate’. That was well summarized by Michael Gerhardt who said that ‘Readers may not be confined by particular constitutional language, but they are confined to the language’[7]. The determining factor here is in the form of an obligation to use certain words, to the exclusion of all others. What is not within the power of the competent organs—in other words, what determines the persons involved in this action so long as they intend to act as ‘agents’ or ‘officers’ of the Constitution—is to use terms other than those of the Constitution. In the following stage of the process of drafting an impeachment article, however, it is another problem that arises. The Constitution is not its own dictionary. In determining the use of certain terms to trigger a removal, the Constitution does not determine the import that is to be ascribed to those words.

In 1970 Gerald Ford declared that an impeachable offence ‘is whatever a majority of the House (considers it) to be at a given moment in history’ and that ‘conviction results from whatever offense or offenses two-thirds of (the Senate) considers to be sufficiently serious to require removal’[8]. This is not a context of total arbitrariness, though. The decision process is created by the constitutional act and that act contains a number of determinations on the subject. The presence of constitutional characterizations (however vague) has as its result, for example, to prevent impeachment being like an act of attainder whereby Congress could drum up an offence for the purpose of charging the person with it forthwith[9]. For there to be an act of attainder, there must be a legislative element combined with a simultaneous judicial element. With the constitutional characterizations applying to impeachment, this legislative element is necessarily prior to the exercise of a sort of judicial power in the context of the removal ‘trial’. Impeachment mobilizes the said constitutional characterizations and does not create them. That does not eliminate all the difficulties, for while Section 4 of Article II is still the mould from which the characterization operation is shaped, it is far from determining its content in full. Here, the danger of retroactivity, the fear of which led the Constitution to proscribe acts of attainder, is difficult to avoid. The looser the constitutional characterizations, the more readily they can be applied retroactively to actions that are deemed objectionable.

In other respects, the same terms imply, not a limitation but a freedom. More exactly, they are permissive in nature. This is how one might interpret the fact, remarked upon by Hamilton, that there should be no strict delineation of the offence by the Senate[10]. The impeachment clauses open up a sort of empowerment of the authorities referred to by Sections 2 and 3 of Article I. They authorize the House of Representatives to ‘fill’ them freely with new meaning. ‘Freely’ may not be not the right word here because we are not free with language. But everyone knows that language has a property of being equivocal, in that when we say something, we do not say just one thing, and we do not control everything that is said. The impeachment clauses authorize the Senate not to instigate such a creation but by a ‘guilty’ or ‘not guilty’ on each impeachment article, to say whether the creative work done by the House of Representatives corresponds or not to the labels used in the Constitution. This attribution has a remarkable property of being exclusive: ‘The House of Representatives ... shall have the sole Power of Impeachment’ while the Senate ‘shall have the sole Power to try all Impeachments’.


What results from Section 4 of Article II is that when one of the persons named commits one of the punishable acts defined by the clause, a consequence must follow in the form of what the Constitution calls ‘removal’. The text does not refer to what are ordinarily considered as criminal sanctions such as imprisonment or fines. Removal severs the bond between the individual and the constitutional office held by this individual. Section 3 of Article I specifies that ‘Judgment in Cases of Impeachment shall not extend further than to removal from Office, and disqualification to hold and enjoy any Office of honor, Trust or Profit under the United States’. At one and the same time, a legal characterization must be made and it is thought that evidence can and should be adduced to prove the acts so characterized. The question arises for that matter, recurrently in scholarly dogmatic studies of impeachment as to the standard of proof applicable to the procedure. But what is not in doubt is that some sort of evidence is required.

1.2 Impeachment as a form of constitutional accountability

1.2.1 Impeachment and higher lawmaking

In the impeachment clauses, the absence of any definition of offences is not a contingent fact; it is not the result of the legislator’s silence that could be broken should the legislator so decide. It seems to be agreed that the offences in question do not lend themselves to any statutory definition. This is what Joseph Story meant by saying that ‘political offenses are of so various and complex a character, so utterly incapable of being defined, or classified that the task of positive legislation would be impracticable, if it were not almost absurd to attempt it’[11]. The ‘political’ nature of the offence seems to be such that it impedes the statutory identification of its constituent factors.

In a parliamentary government, the political accountability of ministers before parliament likewise escapes any a priori definition. Attempts to define cases of accountability in advance have ended in failure. But not for the same reason: political accountability in parliamentary regimes is a test of the identity of political views between executive and parliament. It is fundamentally a governmental accountability. The question is Who will govern and how? The logical matrix and the discursive forms that parliamentary manifestations of distrust must take on do not consist in identifying facts corresponding to statutory factors abstractly determined by a legislator. In a word, the parliamentary type of political accountability does not rest on an act of characterization. In that it eludes the jurist’s art a little further. What makes the terms of the Constitution escape for the most part from any prior definition cannot lie in it being impossible, as is characteristic of a parliamentary regime, to ascertain beforehand what is or is not the subject of consensus between a government and a parliamentary majority. When we say it is the political nature of the offence that prohibits its statutory characterization, we mean something else. It can seriously be doubted that Story meant to say that any crime against the state eludes by definition any statutory definition. Counter examples were already abundant in English law and in the law of the young American Republic.

Here I think rather that we must construe the term ‘political’ offence as follows: the question posed is the question of the conduct of a governor and the response may consist in his removal. The evil and its cure lie firmly within the province of the political. It is not a question of curing an individual evil (punishing the guilty party or helping him/her atone) nor even of a social evil in the broad sense (defending society), but more specifically of an evil affecting government. However, at the same time, it is a government made up of real people, men and women of flesh and blood. This is the level at which US constitutional law recognizes that there has to be ‘government of men’ alongside a government of laws. This government has close ties with society. Politics has various levels of sedimentation. This is especially true of politics in modern constitutional states. In such states, things are arranged so as to distinguish what is fundamental, what serves as the foundations – and that is called ‘constitutional’ – from what is a matter of immediate political action, relating to the immediate state of affairs and subject to that fundamental law. The constitutional domain is therefore an area of intense sedimentation of politics, a fairly stable tectonic plate. At least, this is what it is expected to be. And yet governmental policy and constitutional policy are not entirely dissociated. This is a central property of modern constitutionalism, especially in the United States and that is asserted in the opening lines of the letters of The Federalist: political arrangements are a matter of choice, they are part of human freedom. The Constitution is the very manifestation of that freedom. There is in that therefore a contradiction that goes to the very core of the constitutional question in the broad sense of the term: on one side, what is ‘constitutional’ is separated from what is political and overarches it; on the other side what is ‘constitutional’ is nothing but political, and is entirely the outcome of man’s political activity.

That said, the way in which government action and the constitutional sphere are related depends on how we interpret the idea of a ‘violation of the Constitution’. That it is so difficult to give meaning to this expression has much to do with the proposition that the Constitution is a supreme law. This proposition acts as a kind of ‘thought screen’. This idea of the Constitution as the supreme law has considerable importance in modern constitutionalism, but that importance is very largely attributable to its character as a dogmatic proposition. In the context of normative reasoning, it makes it possible to grasp a logical contradiction between a statement of the Constitution and a statement found elsewhere in the body of existing law. If we accord the Constitution the status of higher law, we can understand this contradiction as a conflict of laws. Jurists are mentally equipped for dealing with conflicts of law. To say that the Constitution is a law removes us from the realms of the unknown into that of the known. That in many circumstances the Constitution can be dealt with as a positive law, a certain kind of entrenched enactment, there is no doubt. But in other contexts, reasoning in this way clouds the meaning of propositions that are no less present in existing law. This is the case of the violation of the Constitution. The Constitution contains a number of statements for which it cannot readily be seen in what their violation would consist. But if we accept to go behind the analogy between the constitution and statutory law, something else comes to light. The Constitution is also understood, not as a succession of norms, but as legality itself in the sense of the general validity of law. It can also cover a certain understanding of the political action required of the people to whom a fragment of constitutional power is given. With this in mind, ‘violating the Constitution’ takes up other possible significations.

It goes without saying that this interpretation depends in turn on what is meant by ‘Constitution’ and the way we envisage relations between individuals and the Constitution. This question arises over the various rules and constitutional regimes in each state. It is also greatly implicated in the use of a certain constitutional vocabulary. The use of the term ‘office’ is correlated with the use of the term ‘duties’ to characterize the President’s constitutional powers. One speaks of his ‘duties of office’, which does not exclude the use of the term ‘powers’ for the same subject matter, but points in potentially different directions. The office holder is not merely ‘authorized’ to intervene but is also under a duty to do so.

Most of the interpretative difficulties encountered with impeachment, and especially, the way we must understand the constitutional characterizations of punishable acts, relate to the point that impeachment opens up a window, a point of communication between the sphere of everyday, governmental policy and the sphere of constitutional policy. Each impeachment affects the constitutional stratum: this is not ordinary compliance with the law that is at issue; this, it seems to me, is the central reason why it is not possible to turn to the ordinary law, the ‘law of the land’ to understand any of the factors constituting ‘high Crimes and Misdemeanors’. That is supposedly why impeachment is an exception to the principle of the strict definition of offences in law.

This is where presidential impeachments take on a special prominence. Because of the President’s central constitutional mission, because he takes an oath to abide by and uphold the Constitution, the accusations that may entail his removal must bring out the character of his relation with the Constitution. They must state what the holder of the presidential office must or must not do. This is a form of what Bruce Ackerman has called ‘higher lawmaking’. Each impeachment is a new opportunity to state the Constitution. Conversely, resort to impeachment is a danger for the state and its stability[12]. This calls for discernment, caution and moral maturity. James Wilson wagered before the Pennsylvania Convention that impeachments would not be frequent. They have something in common with the constituent power: like it, they go to the very foundation of constitutional order. Like it, they constitute, to take up Marshall’s expression, ‘a very great exertion (...)’ that should not ‘be frequently repeated’[13].

The context of impeachment is one of abnormality: the procedure is surrounded by an atmosphere of exception. Uncertainty about the motives of an accusation, the political context that surrounds the proceedings (since the organs of investigation and trial are political organs) mean that it is never very clear whether this atmosphere of exception exists before the impeachment, or whether it is the impeachment that creates it, and if it will bring it to an end. All that is known is that the constitution has entered a zone of turbulence. The impact of an impeachment is a separate question from the matter of whether the President is acquitted or not. Acquittal or abortion of the procedure is not a decisive factor. The consequences of the procedure immediately extend beyond the accused individual. What counts above all is whether the turbulence of political life attains the territory of the Constitution and destabilizes it. Is that not one of the implications of the idea of a trial: a disagreement arises over the law. The law, which is generally regarded as what is stable, what must not change, is then set in motion, or more exactly is shown to be fundamentally uncertain. The indeterminacy it contains is made apparent.


For all these reasons, it seems appropriate to define the kind of accountability which is at play in the impeachment procedure as a ‘constitutional’ one. By ‘constitutional accountability’ I mean that which is incurred for calling into question the constitutional equilibrium as a whole. In fact, such a reproach appears in all of the impeachment articles passed by the House of Representatives against Johnson, Nixon and Clinton.

In all these instances, the head of state was criticized for violating the Constitution: by ignoring the obligation it imposed (especially ‘that he should take care that the laws be faithfully executed’)[14]; by ignoring his oath of office by which he undertakes to uphold the Constitution; or by a more general formula observing that he has acted ‘in violation of the Constitution and laws of the United States’[15]. The common ground between constitutional responsibilities and governmental political accountability is that the exercise of political power is at issue. Executive power is conferred by the Constitution on the President alone. He is not a head of state on the model of the German Federal President, for example, on whom Article 61 of the Basic Law places a form of accountability fairly close to that of impeachment but who otherwise has far less powers than a U.S. President.

What does it mean, in this context, to ‘violate the Constitution’ and in what way is impeachment a procedure that safeguards the Constitution, a procedure that calls into question some ‘constitutional accountability’? To gain a better understanding of this, we must try to apprehend the meanings of the word Constitution that impeachment reveals. Impeachment is a moment at which it is necessary to ascribe a meaning, not only to this or that clause in the constitution, but to the concept of a constitution as such. In every impeachment, the constitution is at stake.


The background to these provisions about impeachment consists in a certain understanding of the idea of constitution. In France, Michel Troper popularized a distinction between two possible definitions of the Constitution: as a norm (or a set of norms) and as a mechanism[16]. The constitution can also be understood as a certain arrangement of political authorities, or as Montesquieu famously put it, an ‘arrangement of things’ whereby ‘power should be a check to power’. In the US case, there are factors that go in both directions. On one side the Constitution declares itself to be the ‘supreme law’ of the land. This position has been consolidated by the development of controls on the constitutionality of laws by the Supreme Court. The US Constitution is also a certain arrangement of the constitutional ‘branches’ or ‘departments’. This double aspect was quite well expressed in a famous speech by Gerald Ford, when, further to the attempted impeachment of Richard Nixon, he had been designated President of the United States. He then declared: ‘Our Constitution works; our great Republic is a government of laws and not of men. Here the people rule. But there is a higher Power, by whatever name we honor Him, who ordains not only righteousness but love, not only justice but mercy’[17].

And yet, the US Constitution is also something other than that. It is reducible neither to a supreme law governing men nor to a mechanism capable of a proper ‘functioning’. Oddly enough, what impeachment tells us about the Constitution points back to some old, maybe even archaic definitions of the Constitution. One envisions the Constitution as a set of laws (as the law in its entirety). The other views the constitution as a set of judicial offices.

One might wonder here about what I mean by ‘definition’ of the Constitution. The definitions in question are present in our working material, the impeachment articles and the related forms of political discourse. In these texts we encounter a tendency to speak of the Constitution as a set of laws or of law as a whole. That implies certain figures of speech that in turn are dependent upon certain logical schemes. The same is true of another central way of speaking of the US Constitution namely as a set of offices. This language itself, which triggers the use of a certain vocabulary (for example, whoever says ‘office’ also says ‘powers and duties’ of its holders) is often echoed by another register, which is that of the moral language of the Constitution. Where one speaks of office, rights and duties, one is also induced to reason in terms of trust, censure, blame or criticism, and so on[18].

2.1 The Constitution and ‘the laws’ (or the constitution as the law as a whole)

The impeachment procedure refers us to an extensive definition of the Constitution. It is not a question here of applying a normative type of reasoning in which what matters is to ascribe its proper coefficient of normative value to this or that ‘norm’. The question here is to take into account the law as a whole. It is a matter of contemplating the problem not as one of specific violations but of an overall challenge to law. This way of reasoning thwarts any analysis in terms of the hierarchy of norms. ‘To protect the Constitution’ does not mean the same thing as ‘to give a sanction to each constitutional norm’. It would not be pointless adopting this directive for interpreting Article 5 of France’s 1958 Constitution (which invites the President of the Republic to ‘ensure due respect for the Constitution’).

A procedure of ‘constitutional’ accountability such as the impeachment apprehends law as a whole and makes it an offence to infringe not some part or other of the law but the law taken as a whole. In constitutional language we encounter standards such as ‘the law of the land’ or ‘the Constitution and the laws’. The purpose of these standards is to bring out the unity of the legal system and the fact that the issue here is to protect that unity.

Such a standard is to be found in the formula of ‘supremacy’ clause of Section 2 of Article VI of the Constitution: ‘This Constitution, and the Laws of the United States which shall be made in Pursuance thereof; and all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land’.

Here I might cite by analogy the provisions of Article 61 of the German Basic Law: the impeachment of the President ‘for deliberate violation of the Basic Law or any other federal law’. At the core of this type of standard lies the possibility of ‘the Constitution’ on one side and ‘the laws’ on the other being either the same thing or two different things. It is certain that as a rule of federalism, the supremacy clause makes a clear distinction between the Constitution and the laws, even if the purpose is to unite them with the same superiority over state laws. But in other instances, it seems that this standard invites us to bundle together, with the least possible distinction, both the ‘Constitution’ and ‘the laws’. The use of the plural for ‘laws’, which it seems was more frequent at the time of the Founding Fathers than nowadays, points in this direction. James Wilson declared before the Pennsylvania Convention that although the President ‘is placed high […] yet not a single privilege is annexed to his character; far from being above the laws, he is amenable to them in his private character as a citizen, and in his public character by impeachment’[19]. Wilson’s language (amenable to the laws) prompts an analogy between impeachment and a sort of prosopopoeia comparable to that of the Criton, in which the President at fault is confronted with the laws of his body politic. By impeachment, and also by other ordinary legal channels, it is as though the President is summoned before the laws in person.

Strikingly, this standard is also to be found in several articles of impeachment against US Presidents. It may be present directly: in the first impeachment article against Andrew Johnson, the President is accused of acting ‘in violation of the Constitution and laws of the United States’. I liken these impeachment articles to the Senate’s pseudo-motion of censure passed against Jackson in 1834: ‘Resolved that the President, in the late executive proceedings in relation to the public revenue, has assumed upon himself authority and power not conferred by the Constitution and laws, but in derogation of both’[20].

The same idea of the unity of the Constitution and of the laws of the land can also be found more indirectly: a very striking thing in practically all of the impeachment articles drafted in the House of Representatives against US Presidents is the reference to the violation of one or other individual legal rule. Yet one point seems clear: there is no need for such a specific breach of a rule to justify an impeachment. But that does not mean there is no connection between impeachment and positive law: impeachment is a constitutional ‘guarantee procedure’ of the law as a whole. What is sanctioned is not the violation of the law but the fact that it entails consequences that go beyond violating a certain rule. The end point of the demonstration in an impeachment article is not that a violation of the law has been proved but that this violation of the law entails a violation of the Constitution. And more exactly, we must dwell on what is to be understood by violation of the Constitution. To determine whether Clinton set himself ‘above the laws’ presupposes that the constitutionality of his behaviour is called into question. The violation of the laws by a President who is tasked with enforcing them is the model of infringement of the Constitution, just as, for the English of the Glorious Revolution, the king flew in the face of the Constitution by suspending the legislative measures by an act of his prerogative. Impeachment is one way of facing up to the threat of tyranny, the absolute political evil that despotism represents in liberal doctrine. In the heroic version of this doctrine, it is a question of confronting a situation where the law has been dismissed lock, stock and barrel. Legality has been disregarded by the tyrant (who defines himself as such by that very fall into lawlessness) and therefore ‘government’ is destabilized. A certain atmosphere of political Lockeanism seems to hang over all that. Impeachment is a weapon given to society against those who govern it. In the constitutional role play game, society is played by the Congress and the governor par excellence – as well as the potential tyrant – is the holder of executive office[21].

The defining feature of political Lockeanism was to put centre stage the question ‘Who will be judge?’. This question is side-lined in part with the parliamentary form of political accountability: ministerial censures in a Wesminster-style form of government[22]. Such accountability knows of no impartial third party called upon to decide disputes between two parties. This is not the case in the impeachment procedure. There is an impeachment trial with an accusatory organ and an organ for judging, and with procedural guarantees akin to those of a criminal trial, and offences. But at the same time, the appropriate ‘judge’ is a body that is also tasked with a part of the legislative function. There are therefore an accuser and a judge, but at the same time the accuser and the judge are both representatives, in their different ways, of the American people. The expected impartiality of senators in their judgement is not of the same kind as the impartiality expected of a judge sitting in a court of law. It is quite obvious that the senators are ‘judges in their own cause’. The impartiality expected of them consists in their laying aside their partisan ties and prejudices at the moment of voting. They must, as it were, summon up the impartial spectator that is within themselves.

In any event, what impeachment aims to protect is the rule of law[23].

This is apparent in the charges against Clinton, for example: ‘In doing this, William Jefferson Clinton has undermined the integrity of his office, has brought disrepute on the Presidency, has betrayed his trust as President and has acted in a manner subversive of the rule of law and justice, to the manifest injury of the people of the United States’[24].

The accusations of obstructing the course of justice and such like would suggest, further, that impeachment can be used in order to guarantee the constitutional principle of due process. The violations of due process can count as impeachable offences[25]. At least, articles of charge such as the one just quoted would point in that direction. Conversely, one of the most acute legal questions of the impeachment regime is whether the accusation and removal procedure complies with the requirements of due process. ‘The President, no less than the lowliest citizen, is entitled to the protection of due process’.

In any case, the question here is not to provide an effective sanction at some level or other of the hierarchy of norms. It is a question of overall protection against what is itself a comprehensive challenge. There is a rather grandiloquent aspect in the impeachment articles: each seems to be saying that the accused President, through this or that behaviour, called into question the rule of law as a whole, in its entirety. Judicial rhetoric joins forces with constitutionalist rhetoric for the occasion.

We can conclude this point with a remark. One of the languages that can be used in the context of impeachment is the language that gravitates around the ‘state’. English law recognized crimes against the state. In a number of interpretations of the impeachment clauses, we encounter the state or terms approaching it. Any number of quotations can be listed showing that impeachment is to protect the state. It is also common to speak of ‘great injury to the community’; of ‘great misdemeanours against the public’[26].

The reference to the state has an encompassing virtue which is in many respects more effective, more ‘blanketing’ in a way, than the reference to the ‘law’. The term ‘state’ defines concretely the victim of the wrongdoing. It is closer to the republican idea of a collective danger, a threat to the political community. In doing so, it wards off the need to adduce evidence of specific behaviour, or to prove that specific offences were committed. After listing a whole series of misdeeds ascribed to Johnson, representative Boutwell concluded that ‘When you consider all these things, can there be any doubt as to his purpose, or doubt as to the criminality of his purpose and his responsibility under the constitution? It may not be possible, by specific charge, to arraign him for this great crime, but is he therefore to escape? These offenses which I have enumerated… are the acts, the individual acts, the subordinate crimes, the tributary offenses to the accomplishment of the great object which he had in view. But if, upon the body of the testimony, you are satisfied of his purpose, and if you are satisfied that these tributary offenses were committed as the means of enabling him to accomplish this great crime, will you hesitate to try and convict him upon those charges of which he is manifestly guilty, even if they appear to be of inferior importance, knowing… that in this way, and this way only, can you protect the State against the final consummation of his crime?’[27].

2.2 The Constitution as a set of offices

In the context of the impeachment procedure the constitution is not only defined as a repository of rules, or as another term for the law as a whole (or legality). Rather, ‘constitution’ also seems to point towards a set of institutions.


I begin with a prominent element for anyone examining impeachment as a procedure for engaging the accountability of the head of state of the United States: the peculiarity of impeachment in the 1787 Constitution is that this regime of accountability is not specific to the head of state but extends to a broad category of civil office holders.

It is pointed out, admittedly, that it is the President who was discussed above all at the Philadelphia Convention and that the Founding Fathers (...) were quite prepared to believe that Presidents might abuse their power and were therefore determined to provide the new republic with a way of removing any who did[28]. Treatment of the President is understood even so as an eminent case[29], indeed, but all the same one case among others of addressees of a more comprehensive legal regime. This character has grown over time. It can be conceded that in[30]. But, as time has passed by, the outline of the procedure was drawn in much more detail by the comparatively more numerous impeachments of federal judges than of Presidents.


From the point of view of impeachment, the interpreter is therefore induced to apprehend the President as an office holder, admittedly a high profile one, but even so an office holder of the Constitution among others. Indications that the President of the United States can be characterized as an ‘officer’ within the meaning of the Constitution are not wanting. The term is found in the case law of the Supreme Court (‘officer of the executive branch’[31]; ‘the office of the President of the United States’[32]). It is also found in scholarly literature[33]. I do not intend here to set out a discussion about the scope of the term ‘officer of the United States’ in the Constitution[34] but I would like to use the term as an explanatory concept with respect to the institutions arising from the Constitution. In France, the higher political institutions are organs of the nation. This concept of organ does not translate so well into the idiom of American constitutional law, mostly because of a different theory of political representation. Admittedly, letter 14 of The Federalist contains an assertion that might have suited someone like Sieyès: ‘In a democracy, the people meet and exercise the government in person; in a republic, they assemble and administer it by their representatives and agents’. However, the underlying idea of what it is to be a ‘representative’ or an ‘agent’ widely differs in the United States and in France. An evidence of this is that the word ‘office’ belongs to the idiom of constitutional law in the U.S. while it is next to absent in French law, especially as far as the executive is concern.

I would suggest that the model of office is one of the main models for making sense of US political institutions. The US Constitution is at the same time a ‘government’ in the broad sense of the term that obtains in the English language. The Supreme Court is not unaware of this. For example, in setting aside the argument that executive privilege is not explicitly granted by the Constitution, it observes that ‘The privilege is fundamental to the operation of Government and inextricably rooted in the separation of powers under the Constitution’[35].

Any interpretation of the concept of constitution that confined itself to the idea of the ‘supreme law’ would be incomplete; very incomplete even. This normative dimension is only one aspect of things. Seeing this dimension only, or giving it undisputed precedence in the analysis, is to give in to a normativist prejudice. What the Supreme Court’s case law shows is that this normative aspect is in reality inseparable from the governmental aspect. The normative dimension and the institutional dimensions (the Constitution as government) are intertwined. The Constitution creates a limited power, arising from the people, but it also creates a ‘strong’ power (as implied by its preamble and explained as regards the President in letter 70 of The Federalist) and a living power. It is as such too that it is a ‘living Constitution’. This fact is taken into account in the letters of The Federalist: a republic, says Madison in letter 39, is ‘a government which derives all its powers directly or indirectly from the great body of the people, and is administered by persons holding their offices during pleasure, for a limited period, or during good behavior’.


What are the implications of characterizing as offices the individual positions of power created by the Constitution? The Supreme Court has engaged in the work of relating the term of office to other terms in the system of constitutional denominations. ‘The term [officer] embraces the ideas of tenure, duration, emolument, and duties’[36]. The articles of impeachment also help us to understand what an office of the US Constitution is. Such offices are the creation of the Constitution. The articles of impeachment almost systematically recall that the office of President is an office ‘of the Constitution’. Because of this status as creatures of the Constitution, they invariably act upon delegation (and never initially) and the powers attached to them are always limited. The provisions relating to impeachment are manifestly means of expressing such limits. The English practice was to allow private individuals holding no public office to be impeached[37]. On the contrary, US impeachment is limited to office holders.


Speaking in terms of offices has other legal implications. When the question is one of the respective delimitation of powers, the Constitution is described as being divided into ‘branches’ or into ‘departments’. But impeachment is precisely not a procedure tending to guarantee the distribution of competences. The President, or any other holder of a civil office, is not criticized for exceeding the bounds of his competence. Indeed, it is necessary to reason in terms of office here because what they are reproached with is of another nature. The central implication of reasoning in terms of offices is that all individuals invested by the Constitution with even a parcel of constitutional power are viewed as moral agents, whose action is governed by their conscience.

In a first meaning of the term ‘conscience’, what is meant is each agent’s or institution’s individual capacity to define the scope of its powers. It is a matter of recognizing to each a power to personally apprehend the scope of their function and the requirements it implies. Nothing could replace this. In a previous study, I endeavoured to show how the judges of the founding period of the American republic had defined their own constitutional office in such terms and how it meshed together with reasoning about their oath of office[38]. This obligation – which is at the same time a freedom – to delimit the limits and implications of one’s own power, is not a matter for judges alone. It extends to the other offices of the Constitution. The term office is quite accurately suited to the position of the President and of the ‘civil officers’ covered by the impeachment clauses. It is also suitable for judges. The term office has admittedly been rejected with respect to members of Congress. And yet, when it comes to understanding how their status and functions are contemplated in US constitutional law, it remains a very useful approximation. The appeal to one’s own conscience to delimit the powers one holds and the way one should exercise them is no less frequent among members of Congress than in the executive or judicial branches[39]. Such reasonings are not merely more or less fuzzy moral justifications floating above positive law. They are a component part of legal reasoning performed on the basis of the Constitution. They are therefore an integral part of the dogmatic work of interpreting and implementing the Constitution.

This can readily be seen by re-reading Jackson’s famous words in his veto message upon the United States’ Bank bill.

If the opinion of the Supreme Court covered the whole ground of this act, it ought not to control the coordinate authorities of this Government. The Congress, the Executive, and the Court must each for itself be guided by its own opinion of the constitution. Each public officer who takes an oath to support the Constitution swears that he will support it as he understands it, and not as it is understood by others[40]. Although voiced in a situation of intense political controversy, these words are not just the outraged expression of a polemical point of view. They say of the concept of Constitution something about which there is a degree of consensus. Hearing an action brought by a judge who had been subjected to an impeachment, the Supreme Court stated in 1973 that ‘In the performance of assigned constitutional duties each branch of the Government must initially interpret the Constitution, and the interpretation of its powers by any branch is due great respect from the others’[41].

This goes some way towards lending some credence to the ‘departmental’ analysis of the U.S. constitution. The Democrats who took Johnson’s defence when he dismissed Stanton in disregard of the Tenure of Office Act insisted that ‘The President of the United States has a primary right to judge of the Constitution of the United States and the laws passed under the constitution, subject to all the penalties to which he may be liable if he violates any law when that law is adjudged to be constitutional by the Supreme Court of the United States’. Otherwise, it would suffice for Congress to pass an unconstitutional law so as then ‘impeach the President because he wishes to test the constitutionality of your rule’[42].

Significantly enough, the Republican radicals who took the initiative to impeach Andrew Johnson claimed a similar power for Congress. They took the view, as did George Boutwell, that the power of impeachment given to Congress by the Constitution ‘was not reviewable and should only be guided by ‘the conscience of this house’. But on the office of President, they replied significantly through James Wilson that when the President accepts office ‘he merges his individuality with into that official creature which binds itself by an oath as an executive officer to do that which, as a mere individual, he may not believe to be just, right or constitutional’[43]. The Republican argument in 1867-68 was that the office of US President was subordinate to Congress. This was the only ground for charging Johnson with a violation of the Tenure of Office Act. Otherwise, it was asserted, the consequence would be that the President held absolute power to decide which laws he executed or refused to execute. This reasoning was perfectly acceptable. Under other circumstances, it could have proved a useful rampart against certain executive claims. It is not absolutely invalid, if only because the absolute standpoint from which it could be declared invalid does not exist. But it seems legitimate to say that it was incompatible with another way of speaking of the Constitution. By ‘way of speaking’ I mean both the basic materials (especially a certain vocabulary) and reasonings into which that material fits. According to this other way of speaking of the Constitution, the President is armed with a power of veto allowing him to oppose any measure he deems unconstitutional. Even when he has given his assent to a statute, he still uses signing statements addressed to the administration to neutralize any interpretations of the said law that he judges unconstitutional. The result of the interpretation of Johnson’s Republican opponents was to turn the President of the United States into a pouvoir exécutif, French style. That does not just mean a certain limitation of his powers, but also the way the institution is conceived. For them, he was basically a higher-ranking agent, in whom the characteristics of a human being were as little apparent as possible, and who is placed in a relationship of subordination and of execution with respect to the legislative organ and the measures it passes. That was almost tantamount to claiming that in the US there was a sort of general will, expressed by Congress, and which it was the executive office to carry into execution. That interpretation of the U.S. Constitution did not prevail in 1868 nor subsequently. The U.S. president did not become a constitutional ‘organ’ à la française or in the sense of the German Allgemeine Staatslehre.

It is impossible to maintain this view and at the same time to consider, as orthodoxy requires, that the Constitution creates coordinate powers. The difficulty arises from the fact that some of the President’s powers cannot in any way be understood as subordinate powers[44]. In fact, the whole difficulty that was to lead in 1952 to the decision in Youngstown Sheet and Tube Co. v. Swayer (343 U.S. 579) and to the typology drawn up by Justice Jackson at this occasion is that the President is both a constitutional office holder and an institution to which Congress, through its laws, can attribute powers or keep under subjection. The US Constitution is organic in the sense Kant gives to the word: the executive branch is, depending on cases, the branch that determines the ends to pursue, or a means to ends determined by another branch. The President may be the arm of Congress or an officer of the Constitution that holds the reins of state. In the 1868 precedent, the Republican radicals’ interpretation of the constitutional position of the President would have been maybe more convincing if one of the central complaints levelled against Johnson, the breach of the Tenure of Office Act, had not relied on a statute by which Congress had tried to recover one of the President’s constitutional prerogatives (the appointment of executive officers) by a legislative regulation which very significantly altered the constitutional equilibrium, and especially the role of the president as the bearer of a unitary executive power. More than in other cases, perhaps, it could be observed that the President was indisputably qualified to interpret the scope of a power that the Constitution bestowed on him and that Congress was attempting to restrict. The position of Johnson’s supporters on this point was that he had good cause to persist in fully using his constitutional prerogative of revocation[45]. That view was based on the belief that the president had ‘a right independent of Congress or the Supreme Court to decide the constitutionality of the laws’[46].

The Supreme Court’s claim to be the ultimate interpreter of the Constitution is no impediment to this state of affairs. First of all, this claim relates to the Supreme Court’s power to state the law with a view to protecting the rights of individuals. It is when such rights are at issue that its power of interpretation is at play and is held to be sovereign. Furthermore, the Supreme Court itself is respectful of the power of interpretation of the other high constitutional institutions. That is the sense of its case law on political matters. The two main criteria set out in 1962 by Baker v. Carr (369 U.S. 186) can be reduced from this standpoint to just a single criterion. Where the Constitution makes a ‘textually demonstrable constitutional commitment of the issue to a coordinate political department’, the outcome is that the language used by the Constitution ceases to be one about which the judicial branch of the Constitution has an authentic power of interpretation. We therefore find ourselves in a case where the courts must recognize there is ‘a lack of judicially discoverable and manageable standards for resolving it’. An eminent illustration of this possibility of each coordinate branch of the Constitution to construe the Constitution lies in the President’s power to veto congressional legislation. The Supreme Court thus recalls that the presidential veto ‘reflects the Framers’ careful efforts to check whatever propensity a particular Congress might have to enact oppressive, improvident, or ill-considered measures’[47].

2.3 The Constitution as a system of positive moral values


In a second meaning of the term ‘conscience’, that is inseparable from the first, the holder of a constitutional office is also understood as the seat of an individual moral conscience. He is a being of flesh and blood, a being of passions and of reason. The Constitution apprehends him as such and does not seek to reduce him to the organ of a higher will, that goes beyond him and erases in him the characters of an ordinary moral agent. Several aspects of the way the Constitution relates the constitutional office with its holder point in this direction. The oath of office is not the least. It recalls that an office is an institutional form in which the man and the function cannot be fully separated. We can return to Gerald Ford’s investiture speech: ‘In all my public and private acts as your President, I expect to follow my instincts of openness and candor with full confidence that honesty is always the best policy in the end’.

The Impeachment procedure is also an indication of the same phenomenon. This can be seen by contrasting the presidential right of immunity and the impeachment regime. When the question of the President’ immunity is at issue before the courts, they ask themselves whether the acts are ‘official’ or not; were they exercised within or outside the scope of the said functions? In Nixon v. Fitzgerald (457 US 731, 1982), the Supreme Court ruled that the President must be recognized to have immunity from liability for civil damages in matters ‘within the outer perimeter of his official responsibility’. By contrast, in Clinton v. Jones (520 US 681, 1997) civil immunity was denied on the grounds that the litigious conduct did not come within the scope of the President’s official functions. The Court argued that the Nixon v. Fitzgerald precedent could not be invoked in support of immunity for ‘unofficial conduct’. ‘But we have never suggested that the President, or any other official, has an immunity that extends beyond the scope of any action taken in an official capacity’[48].

Now, it may well be that this distinction between official and extra-official becomes meaningless when it comes to impeachment. In the removal procedure, office-holders are envisaged as moral agent in connection with other moral agents; hence the use of the word ‘trust’. This would be the place to explore the connections between impeachment and what the founding fathers, or at least the authoritative interpreters of the Constitution in its beginnings, had to say about the concept of moral law. The fact is that in a context of written law and of republicanism, the American Constitution was treated as a depository of legitimacy and of reason. Despite its being an entrenched enactment, and thus positive law, the constitution was expected to perform a role somewhat similar to that of natural law in the classical tradition. The province of the constitution was a choice area for transposing an ethics that was eager to conceive the individual conditions assigned by the law of nature as ‘offices’ correlatively determining both the rights and duties of their holder[49].


We come to observe therefore that the analysis of the U.S. constitution in terms of either a norm or a mechanism do not tell the whole story. They express truths that should be interwoven with other truths so as not to be lies by omission. To keep to our subject, it is easy to see that the standards used in Section 4 of Article II refer to the violation of a system of ‘positive’ moral values. Each of these terms (moral, values, positive) deserves a few words of comment. Firstly, these are moral values because what is at issue is the judgement of an authority defined in moral terms. Each impeachment article states these values as in a mirror, in which things appear in reverse: it says negatively in what way this or that President violated the values in question.

The articles of charge drafted in English impeachments are not the only antecedent of this way of going about things. Since what is in question today is the responsibility of heads of state, it is not pointless noticing a proximity between the articles of impeachment against Johnson, Nixon and Clinton and the paragraphs in the great English declarations of freedom, especially the 1628 Petition of Right and the 1689 Bill of Rights, that recapitulate the illegal and liberticidal acts held against first Charles I and then James II. In both instances, the complaints conclude by observing that the monarch has imperilled the ‘fundamental law’. The restoration of this principled and presumed pre-existing lawfulness is judged to the be necessary and sufficient condition for the protection of freedom. From this standpoint, it is somewhat inadequate to observe, as is often done, that the British monarch ‘can do no wrong’ whereas the President of the United States is responsible for his acts of office. That is true from the strict standpoint of law and the maxim that ‘the king can do no wrong’ is eminently a maxim of strict law. On the other hand, British political culture also bears fairly recurrent forms of reactions to the tyranny of its monarchs whether real or supposed. The English declarations of rights are an illustration of this type of reaction.

Secondly, the moral values to which the impeachment clauses refer are moral values and not moral rules. What is at issue is ‘misconduct’, improper behaviour. This behaviour is not improper in that it violates a rule but in that it is a shortcoming relative to a moral standard, which is at the same time a political standard, that of the good governor. The closest and most appropriate formula is perhaps that of ‘good behaviour’. This standard is used explicitly in the Constitution about the office of judges. It is not our purpose here to embark on another discussion of technical interpretation of the Constitution, pertaining this time to the question of whether the term ‘good behaviour’ is a further offence, concerning federal judges alone, in the case of removal proceedings and adding to the group ‘Treason, Bribery, high Crimes and Misdemeanors’ applicable to the other civil officers subject to the same removal procedure.

The most convincing opinion on this matter is that this is not the case: for a federal judge to be impeached, in any case, his behaviour must come under one of the characterizations of ‘Treason, Bribery, high Crimes and Misdemeanors’. So the standard of ‘good behaviour’ is probably not an additional characterization[50]. But whether it is or not, it is not dead for all that. It seems more appropriate to me to understand it as a principle of interpretation of the sense of the constitutional characterizations (‘Treason, Bribery, high Crimes and Misdemeanors’). If that is so, it is equally applicable to the President. That seems to me entirely compatible with the way in which the authors of The Federalist used this formula of ‘good behaviour’ or expressions of a similar nature, such as ‘mal-conduct’ (the opposite of good behaviour). These formulas are used both in Letter 65 about judges (‘they are liable to be impeached for mal-conduct’) and in Letter 71 about the office of the President (‘inducements to good behaviour’). We do not, though, stumble on the difficulty that had led the Philadelphia Convention to dismiss George Mason’s proposal to include ‘maladministration’ among the offences for which one could be removed. That was well seen by Raoul Berger. Although he says, one should not ‘import the standards of good behaviour into ‘high Crimes and Misdemeanors’, breaches of good behavior ‘may amount to high crimes and misdemeanours’[51].

In what way is this idea of ‘good behaviour’ an idea that regulates the interpretation of impeachment clauses? In the sense that, for any office of the Constitution, what is at issue is not a rule-based, legalistic morality, but a morality based on virtue, one that defines the Good in terms of appropriate behaviour. It is part of the constitutional ‘regime’ of political offices. And it does cover most of the complaints made in the impeachment articles. Thus in article 10 of the impeachment of Johnson (in fine), Lincoln’s successor is accused of ‘utterances, declarations, threats and harangues, highly censurable in any, (but which are) are peculiarly indecent and unbecoming in the Chief Magistrate of the United States, by means whereof the said Andrew Johnson has brought the high office of the President of the United States into contempt, ridicule and disgrace, to the great scandal of all good citizens, whereby said Andrew Johnson, President of the United States, did commit, and was then and there guilty of a high misdemeanor in office’.

Here there is no great difference between the President and any other civil officer. When impeachment of Judge George W. English was under consideration in 1926, the House of Representatives noted that:

A civil officer may have behaved in public so as to bring disgrace upon himself and shame upon the country and he would continue to do this until his name became a public stench and yet might not be subject to indictment under any law of the United States, but he certainly could be impeached. Otherwise the public would in this and kindred cases be beyond the protection intended by the Constitution ...’[52].

In this text, as more broadly in the American Constitution, the term ‘censure’ takes on a particular moral shade. More exactly, it moves closer to its traditional meaning: a social institution, employing legal means, that protects the moral values of a community. It is in this context that one must try to understand the attempts to pass ‘censure’ votes against certain Presidents, in particular against Clinton. The fact is that in parliamentary regimes the term ‘censure’ has lost almost all of its moral weight. In this context, censure is a political act, not a moral judgement on a person or an act. On the contrary, this meaning has been maintained in the United States. Even after 1834, Congress has regularly had the chance to censure public agents of various standings (President Buchanan in 1860, an attorney general in 1886, the Secretary for the Navy in 1924, etc.). But such censure has not become a normal form of challenging the legitimacy of the head of state.

Thirdly, these standards are positive moral values in that the Constitution provides a procedure aimed at ensuring their effectiveness. With impeachment, the Constitution has created a special procedure for guaranteeing certain moral values. This is the meaning of the term ‘trial’ used for impeachment. In Nixon v. United States (506 US 224 (1993)) the Supreme Court insisted that the term ‘try’ used in the Constitution[53] was not a characterization that could be apprehended by the courts.

That is justification for granting the power of investigation and judgment to the two Houses of Congress. This was put very appositely by Hamilton: ‘Would it be proper that the persons who had disposed of his fame, and his most valuable rights as a citizen in one trial, should, in another trial, for the same offense, be also the disposers of his life and his fortune?’[54]


The specific feature of the US idea of constitutional office is therefore that there is no separation between personal morality and detention of a part of public authority. Impeachment only makes sense in this context. The ‘constitutional’ nature of the accountability so created, to which is added the inevitable imprecision of the words of Section 4 of Article II, has as its result that each impeachment of a President is a moment of constitutional innovation. Each impeachment is also a moment for redefining public morality. Every impeachment is a precedent in which that morality is newly expressed and made manifest. The Constitution is supreme law in the sense too that it is a supreme moral law. An incessant moral conflict is waged in any modern society. A distinctive feature of the American Constitution is that there is a ‘constitutional common good’ of which the Constitution is held to be the expression and the depository. This is a brilliant and historically significant solution to the problem of the moral disorientation that is specific to the modern human condition. We could go even further and say that the Constitution expresses a form of the ‘sovereign good’. The constitution is a legal enactment, which is entrenched and rigid (in the sense that it can only be modified through a specific amending procedure). Therefore, it could – at least in the hopes of its 18th century drafters – take up the characters of objectivity and fixity of the ‘sovereign good’ of the classics, while on the other side being the product of the will of the individuals making up political society. In a sense, the form of the entrenched constitution could reconcile the ancient desire to attain the common good and the modern view of the good as something subjective (a bonum per se). The moral achievement of American constitutionalism lies in the astonishing success of such a seemingly idealistic undertaking. But the moral disorientation, the moral uncertainty, that are also part of the condition of modern man are not absent here for all that. There is perhaps even an additional danger in expecting the Constitution to be the sanctuary of choices pertaining to good and evil, a sort of legal equivalent of Kant’s kingdom of ends.

The constitution is a bulwark against the republican fortuna and Presidential impeachments are moments when fortuna bears hard on the body politic. They always correspond to times when moral conflict, which is constantly ongoing in society, has found an outlet in the constitutional domain. By characterizing the supposedly immoral behaviour of the head of state as violations of the Constitution, one camp tries to take advantage of the authority of the supreme moral law that is the Constitution. It claims that a governor to which it was politically hostile was at the same time blameworthy in moral terms. This must be understood to mean blameworthy in an absolute way and not from a subjective standpoint. When one says ‘what you did was wrong’, one seldom thinks ‘wrong for me’. The moral reproach absolutizes what it denounces. In the case of impeachment, the blame is absolutized by its conversion into an accusation of breach of the Constitution, that is of the supreme moral law of the political community. In a society where life choices are fundamentally left up to subjective appreciations, having an adulterous relationship with a White House trainee can be variously appreciated. Even if few people think that it is praiseworthy, many will think that it is not necessarily blameworthy, especially as the blame in question is public and expressed in a way that singularly increases the solemnity and consequences. But the camp that takes the initiative of bringing the impeachment believes itself in a position to establish that the Constitution has been violated by that behaviour. We must therefore move out of the area where the moral neutrality of the modern liberal state, its alleged indifference to anyone’s life choices of individuals, as well as the principle of pluralism of the moral orientations within society are prevailing. A persuasive article of impeachment must be drafted in a way that overcomes this moral diversity. Everyone in the community must agree that something wrong has been done. It then becomes difficult to abide by the recommendation framed by Holmes in his dissenting opinion in Lochner v. New York[55], and significantly in the opening paragraphs of the Court’s opinion in Roe v. Wade. ‘The Constitution’, said Holmes in Lochner ‘is made for people of fundamentally differing views’. With impeachment, one cannot stop at this pluralistic observation. The Senate is bound by the constitution to decide according to a binary figure of moral judgment inspired by the forms of criminal law (guilty, not guilty) about the behaviour of a man with official functions of the highest importance. Johnson’s impeachment followed on the major moral crisis of the War of Secession. It was the illustration that a profound moral evil lurked within the ‘sovereign good’ expressed in 1776 (in the Declaration of Independence) then in 1787 and 1790 (in the Constitution and in its first ten amendments). The impeachment of Clinton illustrates how the questions of individual morality and public morality could not be separated in the United States. The true partisan divide of the last thirty years has been on this moral ground: it has opposed to different conceptions of the ‘good life’, of the just and blameless terrestrial life. Each will judge whether he/she supports the moral conceptions that impelled the Republicans who undertook to have Clinton removed. The view that the procedure has been abused can be supported with some serious arguments. But, in admitting that one of the important points of the entire business was disapproval of Bill Clinton’s conduct, it is difficulty to criticize his accusers of hijacking a constitutional procedure to attain their ends. The fact is that impeachment and through it the American idea of constitutional office establishes a connection between individual morality and political function. The sort of ‘judgment’ that is expected of members of Congress in the context of impeachment consists precisely in making or refusing to make such a connection in a concrete case. Out of the Pandora’s box thus opened may escape the worst or the best. And worse yet, in a modern and pluralistic society, there will be no agreement as to whether it is the worst or the best that has escaped from the box.

[1] This has often been observed by the Supreme Court. In Bowsher v. Synar (478 US 714 (1986)): ‘Congress cannot reserve for itself the power of removal of an officer charged with the execution of the laws except by impeachment’. Or in Morrison v. Olson (487 US 654 (1988)): ‘Indeed, with the exception of the power of impeachment – which applies to all officers of the United States – Congress retained for itself no powers of control or supervision over an independent counsel’.

[2] By these words, I am trying to capture what German-language philosophy has sometimes called ‘die Formen der Abbildung’.

[3] Section 3 of Article III: ‘Treason against the United States, shall consist only in levying War against them, or in adhering to their Enemies, giving them Aid and Comfort. No Person shall be convicted of Treason unless on the Testimony of two Witnesses to the same overt Act, or on Confession in open Court. The Congress shall have Power to declare the Punishment of Treason, but no Attainder of Treason shall work Corruption of Blood, or Forfeiture except during the Life of the Person Attained’.

[4] Cf. M.J. Gerhardt, The Federal Impeachment Process: A Constitutional and Historical Analysis (Chicago: University of Chicago Press, 2000), p. 103.

[5] Cf. E. Zoller, De Nixon à Clinton, malentendus juridiques transatlantiques, coll. Béhémoth (Paris: PUF, 1999), p. 28 ff.

[6] R. Berger, Impeachment: The Constitutional Problems, (Harvard University Press, 1974) p. 86-7.

[7] Gerhardt, The Federal Impeachment Process, (n 3) p. 67.

[8] Ibid. p. 103.

[9] L.H. Tribe, however, sees impeachment as an exception to the prohibition of acts of attainder. Cf. American Constitutional Law, 3rd edn, (City, Foundation Press, 2000) vol. 1, p. 158.

[10] Cf. The Federalist, Letter 65.

[11] Gerhardt, The Federal Impeachment Process (n 3), p. 106.

[12] Berger, Impeachment: The Constitutional Problems (n 5), p. 91.

[13] Marbury v. Madison, 5 U.S. (1 Cranch) 137 (1803).

[14] Cf. the first article of impeachment against Andrew Johnson.

[15] Ibid.

[16] M. Troper, ‘La machine et la norme. Deux modèles de constitution’, in Troper La théorie du droit, le droit, l’État (Paris: PUF, 2001) p. 147 ff.

[17] Remarks on taking the Oath of Office as President, August 9, 1974.

[18] See E.G. Jeremy Waldron’s comments about the U.S. constitution in terms of offices and their respective ‘ethos’: ‘judicial office is not (for the most part) an elective office, (...) the judiciary is not permeated with an ethos of elections, representation, and electoral accountability in the way that the legislature is. (...) the business of the courts is not normally conducted, as the business of the legislature is, in accordance with an ethos of representation and electoral accountability’. Jeremy Waldron, ‘The Core of the Case Against Judicial Review’, Yale L.J. 115:1346 2006, 1363.

[19] P. Kurland and R. Lerner, The Founders’ Constitution, Volume 3, Article 2, Section 1, Clause 1, Document 8; The University of Chicago Press.

[20] Resolution of 28 March 1834. See J. Maskell, Censure of the President by the Congress (8 Dec. 1998), CRS 98-843-A.

[21] This Lockean tone is found for example in the words of Hamilton in Letter 65 of The Federalist: [impeachable] offenses‘relate chiefly to injuries done immediately to the society itself’.

[22] It is not completely set aside because censorship remains a form of political judgement and also because the majority that passed it does not escape from the ambiguity of being both judge and party.

[23] Tribe, American Constitutional Law (n 8), p. 187.

[24] First article of impeachment against W.J. Clinton.

[25] Berger, op. cit., p. 277.

[26] Berger, Impeachment: The Constitutional Problems (n 5), p. 88.

[27] M. L. Benedict, The Impeachment and Trial of Andrew Johnson (New York: Norton, 1973) p. 76 (session of 5 Dec. 1867).

[28] A.M. Schlesinger, The Imperial Presidency (Boston: Houghton Mifflin, 2004) p. 10.

[29] See Berger, Impeachment: The Constitutional Problems (n 5), p. 92.

[30] Berger, Impeachment: The Constitutional Problems (n 5), p. 4.

[31] Nixon v. US, 418 US (1993) 696-7 in fine.

[32] Clinton v. Jones, 520 US 681 (1997).

[33] E.g. E.S. Corwin, The President, Office and Powers, (New York: New York University Press, 1957).

[34] On this point see Gerhardt, The Federal Impeachment Process, (n 3) p. 76 ff.

[35] United States v. Nixon 418 US 707 (1974).

[36] United States v. Germaine, 99 US (1879) 508-511.

[37] Gerhardt, The Federal Impeachment Process, (n 3) p. 80-81.

[38] ‘Le Serment des juges dans Marbury v. Madison – ou : qu’est-ce qu’obéir à une Constitution écrite ?’ in E. Zoller (ed.), Marbury v. Madison : 1803-2003 – Un dialogue franco-américain (Paris: Dalloz, 2003) p. 51-71.

[39] When he left the Republican ranks to join the Democrat majority in the Senate in 2009, Senator Arlen Specter declared ‘Whatever my party affiliation, I will continue to be guided by President Kennedy’s statement that sometimes party asks too much (...) When it does, I will continue my independent voting and follow my conscience on what I think is best for Pennsylvania and America’, New York Times, 28 April 2009.

[40] The Founders’ Constitution, Volume 3, Article 1, Section 8, Clause 18, Document 20 The University of Chicago Press.

[41] Nixon v. US 703, 506 US 224.

[42] Benedict, The Impeachment and Trial of Andrew Johnson (n 24), p. 106.

[43] Benedict, ibid. (n 24), p. 109.

[44] Such as the function of Commander in Chief.

[45] Cf. Benedict, The Impeachment and Trial of Andrew Johnson (n 24), p. 165. This main argument was weakened, however, by the fact that Johnson’s veto of the Tenure of Office Act had been overridden by Congress. Therefore, the issue of constitutionality was resolved in a positive sense. The defence was therefore compelled to turn to far less sound subsidiary defence arguments. However, it still had to be accepted that the overturning of a veto ended the President’s power to interpret the constitution on the act in question. The practice of presidential signing statements suggests on the contrary that the conception of successive Presidents has been that they conserved a possibility of neutralizing a provision they judged unconstitutional in a statute while it is in force.

[46] Benedict, The Impeachment and Trial of Andrew Johnson (n 24), p. 153. However, Johnson’s position was not consolidated by the fact that in addition to dismissing Stanton, he had appointed a successor to him without having asked for the Senate’s confirmation.

[47] INS v. Chadha, 462 US 919 (1983).

[48] 520 US 681, 694 n. 19, cited by Tribe, American Constitutional Law (n 8), p. 765.

[49] K. Haakonsen, ‘From natural law to the rights of man: a European perspective on American debates’ in M.J. Lacey and K. Haakonsen (eds.), A Culture of Rights – The Bill of Rights in Philosophy, Politics and Law – 1791 and 1991 (Cambridge: Cambridge University Press, 1991) p. 35-6.

[50] Gerhardt, The Federal Impeachment Process, (n 3) p. 83 ff.

[51] Berger, Impeachment: The Constitutional Problems (n 5), p. 91.

[52] J. Turley, ‘Senate trials and factional disputes: impeachment as a Madisonian device’, Duke Law Journal, vol. 49, Oct.

1999, p. 74.

[53] ‘The Senate shall have the sole Power to try all Impeachments’ (Section I of Art. III).

[54] The Federalist, Letter 65.

[55] 198 US 45 (1905) 76.