What Is “The Unitary Executive”?

A Brief Response to Larry Lessig

For reasons I will explain, I was surprised on several levels by my colleague Larry Lessig’s recent reflections on a post of mine (The Head and Body of Leviathan), in which I laid out a maximalist version of the unitary executive theory, deriving most famously from Myers v. United States. What follows is a brief reply.

It seems to me that Lessig’s main idea is mistaken; it rests on a conceptual error, one that I am hardly the first to identify. Lessig has run together two distinct questions: what power the executive branch holds under statutory and constitutional grants of authority, on the one hand, and on the other hand which officials within the executive branch may exercise that power, however broad or narrow it may be. The former question goes to the scope of executive power, the latter to the internal allocation of that power (whatever its scope) within the executive branch. The unitary executive theory addresses only the latter issue, not the former.

Lessig’s idea goes like this: In Marbury v. Madison, the Court distinguished between executive acts that embody political discretion vested in the President, and executive acts that are ministerial, simply the direct implementation of a specific statutory duty. A subordinate officer’s violation of such a duty is, at least presumptively, remediable in court, assuming jurisdictional limits and other limits on judicial power are satisfied. According to Lessig, the unitary executive theory is inconsistent with that distinction.

To distill this claim, Lessig relies on the following passages from Marbury:

[W]hen [a subordinate executive officer] is directed peremptorily to perform certain acts; when the rights of individuals are dependent on the performance of those acts; he is so far the officer of the law; is amenable to the laws for his conduct; and cannot at his discretion sport away the vested rights of others…. [W]here the heads of departments are the political or confidential agents of the executive, merely to execute the will of the president, or rather to act in cases in which the executive possesses a constitutional or legal discretion, nothing can be more perfectly clear than that their acts are only politically examinable. But where a specific duty is assigned by law, and individual rights depend upon the performance of that duty, it seems equally clear that the individual who considers himself injured, has a right to resort to the laws of his country for a remedy.

Lessig calls this the “two-hat” version of executive power: “Sometimes an officer is wearing the hat of the President; sometimes he’s wearing the hat of the law. When wearing the hat of the President, the maximalist [unitary executive] theory is correct. But when wearing the hat of the law, his acts are not the President’s acts. And depending upon the nature of the law, his acts could well be checked by the Courts.”

On its face Lessig’s thesis is surprising. Despite occasional efforts along Lessig’s lines, Marbury is not widely thought to speak in any direct way to the issue of the unitary executive. And that is probably because it doesn’t. The passages from Marbury quoted above speak, not to the unitary executive theory, but to a separate question, one that the Court would soon address and clarify in cases like Little v. Barreme: the question whether subordinate executive officials, and for that matter the President, are bound by specific, non-discretionary or ministerial statutory duties. The Court’s answer to that question, both in Marbury and later cases, was “yes, so long as the statute is otherwise constitutional.” When executive officers act as agents of the President in the exercise of a discretionary judgment authorized by the law, courts cannot substitute their own judgment for the executive judgment; but assuming jurisdictional and remedial limits are complied with, courts can provide a remedy when executive officers act ultra vires, beyond statutory authority. Or, as Marbury itself said: “This is not a proceeding which may be varied if the judgment of the Executive shall suggest one more eligible, but is a precise course accurately marked out by law, and is to be strictly pursued. It is the duty of the Secretary of State to conform to the law, and in this he is an officer of the United States, bound to obey the laws. He acts, in this respect, as has been very properly stated at the bar, under the authority of law, and not by the instructions of the President. It is a ministerial act which the law enjoins on a particular officer for a particular purpose.”¹

The unitary executive theory, by contrast, addresses the vertical allocation of executive power — whatever the scope of that power — within the executive branch, between the President and subordinate executive officers. The unitary executive theory says that the President alone holds the executive power, and that his subordinates hold executive power only derivatively from the President, not in their own right. Myers drew several corollaries from that axiom, such as the presidential power to remove (principal) executive officers at will, because otherwise the President’s power over his subordinates would be rendered effectively nugatory. Whatever the validity and proper reach of the Myers analysis, the whole point of the unitary executive theory is to address the vertical relationship between the President and his own subordinates. Its point is that in our constitutional scheme, as the Supreme Court told us only last year, “the President is a branch of government” unto himself.

In short, nothing in unitary executive theory, by itself, tells us how broad or narrow executive power is.² The Marbury distinction on which Lessig relies, however, operates not vertically within the executive branch, but horizontally across branches; it is about what discretionary power and what nondiscretionary duties the executive as a branch has in virtue of statutory and constitutional grants of authority. Marbury’s observation that sometimes, executive power is discretionary, whereas at other times the executive is bound by law to execute a specific, nondiscretionary legal duty, is a proposition that no version of unitary executive theory has ever denied.³ As I put it in The Head and Body of Leviathan, “[n]othing in the maximalist view of the unitary executive suggests that the President is entitled to act ultra vires, beyond the boundaries of his constitutional and statutory authority, a constraint established long ago and repeatedly reaffirmed.” Even the maximally unitary executive, which holds plenary power internally over its own subordinates, is constrained by law. (Indeed this is so whether or not that law is enforceable in court; the rule of law is not the same as the rule of courts).

To be sure, to say that even the unitary executive is constrained by law is not the same thing as saying that Congress can limit the president's control of the executive branch or limit the exercise of presidential powers. Congress may attempt to impose a statutory duty or constraint on the executive — and hence on the President — that is inconsistent with constitutional grants of preclusive or exclusive power to the executive (for example, the power of removal or the exclusive power to recognize foreign sovereigns). When this happens, we have what the jargon describes as a “Youngstown Category III” case, and the statute may be invalid for that reason. In such cases, the problem is not that the President or his officers have acted ultra vires, but that Congress itself has acted unlawfully, in attempting to enact a duty that it lacked constitutional power to impose. But that is entirely consistent with the conventional understanding of Marbury, as opposed to Lessig’s reinterpretation of it.

One might adduce secondary points. Lessig seems to think (although I am not confident I follow his discussion here) that any version of the unitary executive that is not based on originalism must be an example of “living constitutionalism.” I was most surprised to see this dichotomy as well, given that Lessig himself once articulated a “translation” theory of constitutional interpretation that is self-consciously different than both conventional originalism and Warren Court-style living constitutionalism. Indeed that translation approach was once used to advocate a non-originalist justification for the unitary executive — the advocates being Lessig himself and a co-author, Cass Sunstein. But evidently times and circumstances change. In any event, these secondary points amount to gilding the lily. Lessig’s main argument stands or falls on its inference from Marbury, and for the reasons given above, that inference seems to me unsuccessful.

1. Some confusion arose from an opaque discussion in Marbury of the five-year term of Justices of the Peace for the District of Columbia, an Article I officer. Marshall seemingly opined that Justices of the Peace were not removable at will by the President (even though Jefferson had already fired all of those officers in 1801). In any event, later cases gently dismissed Marbury’s discussion of the removal issue as either dictum, or as resting on unique features of the constitutional and statutory scheme governing the District of Columbia, and made clear that a term of years does not restrict presidential removal power. Lessig does cite Marbury’s discussion of removal of Justices of the Peace, but says that his main argument is “wildly more important” — possibly because he is aware of the later decisions, which he does not cite. Hence I focus on that argument.

2. John Harrison recounts an excellent metaphor Samuel Alito gave at his confirmation hearing for the Supreme Court: the unitary executive theory is like a table, which stands vertically, but whose lateral dimensions are not determined by the theory.

3. On Marbury’s own reasoning, there is of course a different question if statutes confer discretionary duties on subordinate executive officers. As I put it in The Head and Body of Leviathan, “[n]o matter how clearly Congress speaks, statutory duties committed to the discretion of subordinate officers or agencies are only ever grants of executive power, which is the President’s own power and which can be directed as the President sees fit” (emphasis added).