Caleb Nelson has recently weighed in with a long and thoughtful blog post on the so-called unitary executive debates and whether the president of the United States has a constitutional right to remove principal executive officers. Nelson is a famous and prominent originalist who is usually right about everything. His foray into this debate has caused something of a stir.
With all due respect to Nelson—and I mean it, since he’s absolutely right about the public rights doctrine, about sovereign immunity, and about much more besides—he has not quite considered all the arguments for the proposition that the president has a constitutional right to remove. His intervention, though, is critically important because he discounts, probably correctly, many of the older, standard theories of executive power and removal. He relies on works from Julian Mortenson and Jed Shugerman, among others. I have learned much from Professors Mortenson and Shugerman, too—and as I shall explain presently, I agree with Mortenson about the meaning of executive power and with Shugerman about the uncertainty regarding the so-called “Decision of 1789.” I do not, however, think that either of them has shown as a consequence that the president does not have a constitutional right to remove principal executive officers.
Much of the following will summarize the argument of my own longer article in the Journal of Legal Analysis, entitled “The Original Presidency: A Conception of Administrative Control.” As an initial matter, though, it is important to establish what that article doesn’t argue, and where I agree with Nelson.
Congress can no more restrict the president’s power to remove—once it is acknowledged he has that power—than it can restrict his power to pardon.
The conventional view of executive power among formalists is that the Vesting Clause grants a “residuum” of executive powers, including, for example, foreign affairs-related powers traditionally exercised by the British monarch. If removal is executive in nature, and the Constitution does not assign that power elsewhere or otherwise limit the president’s exercise of it, then it vests in the president by virtue of the residuum.
As I have written elsewhere, I agree with Nelson and Mortenson that this account of the Vesting Clause is likely incorrect. My view is that “the executive power” is a substantive grant of power, but of only one: the power to oversee the execution of the laws. But that power, I argue, includes removal—not because removal was a freestanding executive prerogative, but rather because it was part and parcel of the power to oversee the execution of the laws. (More on all this presently.)
I also agree with Nelson that nothing about Article II supports the strong version of the so-called unitary executive thesis that the president must “be in charge of all exercises of executive power by the federal government.” But, as I argue in my Journal of Legal Analysis paper, one can believe—and many did at the time—that the president has no constitutional right to interfere with the duties of subordinate officers in whom Congress has vested discretion (unless Congress has said otherwise), but also believe the president always has the right to remove, for any reason, if he thinks those officers have not exercised their discretion well. This view would also make sense of the Opinions Clause, which Nelson, along with many others, suggests is otherwise superfluous. That clause would be necessary to impose a constitutional obligation on the principal officers to obey the president in this one respect—to provide written opinions about their duties—so that the president may receive information so that he may intelligently exercise the power to remove.
Whatever the Vesting Clause might otherwise grant the president—that is, regardless of whether the president can direct officers or has some other residuum of power—the executive power is, at a minimum, the power to oversee the execution of the laws by others. Because the king, a governor, or the president could not possibly execute the laws alone, the executive power was understood to include the power to appoint officers to assist. One need not accept the royal-prerogative-residuum-vesting thesis to accept this proposition. Here is Mortenson himself: “The executive power was often viewed as either logically entailing or functionally implying the appointment of ‘assistances.’”
The evidence for this proposition is legion. For example, George Mason thought that the Senate should have no role in “the appointment” of “public officers” because it was an executive power. James Wilson thought similarly: “There can be no good Executive without a responsible appointment of officers to execute.” In the Constitutional Convention, Madison argued that the “extent of the Executive authority” was the “power to carry into effect[] the national laws” and “to appoint to offices in cases not otherwise provided for.”
The Antifederalist writer Hampden wrote that “the most important and most influential portion of the executive power” was “the appointment of all officers.” Brutus, Centinel, and Richard Henry Lee, among others, agreed. As did Publius: “the appointment to offices … is in its nature an executive function.”
The more controversial question is the connection between appointment and removal. Giles Jacob’s law dictionary—the most prominent in America at the Founding—stated plainly that the king “names, creates, makes and removes the great officers of the government.” As Shugerman has written, however, it is “rare to find much discussion of removal power on the Founders’ bookshelf.” But there is an explanation for such silence: the power to remove was incident to the power to appoint. Thus, the appointment power included removal by default because an officer could be removed by the very act of appointing someone new.
Shugerman himself has canvassed numerous authorities for this proposition. This tradition was “enshrined in Latin” maxims, “unumquoque dissolvitur, eodem modo, quo ligatur” and “cujus est instituere ejus abrogate,” translating to “every obligation is dissolved by the same method with which it is created” and “whose right it is to institute, his right it is to abrogate.”
In 1780, Thomas Jefferson wrote in a private note: “The power of appointing and removing executive officers inherent in Executive. Executive inadequate to every thing. Appoint deputies. … He who appoints may remove.” On this point, Alexander Hamilton agreed. When he appointed Tench Coxe as the assistant secretary of the Treasury pursuant to the act establishing the Treasury Department, he noted in the commission that he could remove Coxe even though the statute was silent.
The Supreme Court subsequently adopted this proposition. At least “in the absence of all constitutional provision, or statutory regulation,” the Court held early on, “it would seem to be a sound and necessary rule, to consider the power of removal as incident to the power of appointment.” As did other prominent antebellum Americans. Chancellor Kent wrote in a letter to Daniel Webster that “the power to appoint and reappoint, when all else is silent, is the power to remove.” And Daniel Webster wrote in another letter, the power to remove is “incident to the power of appointment.”
As the Kent and Webster letters suggest, however, the real question is whether the Constitution derogates from the default rule. It does with respect to appointments, at least, because of the Appointments Clause. But what of removal? The reason the Constitutional Convention did not talk about removal was because the initial drafts to come out of the Committee of Detail assigned the appointment power over ambassadors and Supreme Court judges entirely to the Senate, and the appointment of other officers to the president alone. There was no need to think about removal at all; the Senate, acting alone, would remove ambassadors, and the president, acting alone, would remove other officers. The appointment power was not shared between the president and the Senate until the Committee of Postponed Matters in the final days of the Convention on September 4. The Framers had no time to think about the implications that sharing the appointment power would have for removal.
The power to remove may follow from the power to appoint, but the power to appoint is—as noted previously—ultimately an incident of executive power.
Thus, in the famous debates over the removal power in 1789, when Congress established the first executive departments, representatives could draw different conclusions from this maxim. Several representatives believed that, because of this maxim, the president and Senate together had the power to remove, as both together had the power to appoint.
Even James Madison agreed that in general “the power to annul an appointment is, in the nature of things, incidental to the power which makes the appointment.” He argued that if all the Constitution said on this score was that the president and Senate shall appoint, then he would agree that the president and Senate together must remove. Because the Constitution included both the Executive Vesting Clause and the Take Care Clause, however, he thought otherwise. The power to remove may follow from the power to appoint, but the power to appoint is—as noted previously—ultimately an incident of executive power. “If any thing in its nature is executive,” Madison stated, “it must be that power which is employed in superintending and seeing that the laws are faithfully executed.”
Thus, Madison argued, both appointment and removal were ultimately incidents of the executive power. The Constitution then derogates from that principle with respect to appointments, but not for removals. The “association of the Senate with the President in exercising” the appointment function, Madison argued, “is an exception to this general rule” that the executive power is vested in the president, which exception does not apply to removal.
The Take Care Clause further supports this proposition, Madison argued, because it implied that the president has the power “necessary to accomplish” the duty of faithful execution. That is, giving the Senate a check on removal effectively transfers the take care duty from the president to the Senate. Importantly, Caleb Nelson agrees with all of this. He writes in his post that he agrees that a senatorial check would be equivalent to an impermissible legislative veto. But that gives away the whole game because that is a recognition that the removal power is in fact the president’s.
The question then becomes whether Congress can, pursuant to its Necessary and Proper Clause, limit the reasons why the president can remove a principal officer. Nelson doesn’t see why it couldn’t, if it can establish the offices. But just because Congress can establish, say, federal crimes, it does not give Congress a “lesser” power to limit the reasons why the president can pardon individuals convicted of those crimes. It is true that Congress establishes offices, and doing so furthers the president’s ability to oversee faithful execution. Even tenure rules, qualifications for office, bipartisanship requirements, and the like can help the president carry law into execution. But Congress can no more restrict the president’s power to remove—once it is acknowledged he has that power—than it can restrict his power to pardon.
To repeat, it doesn’t follow that the president must have a constitutional right to control the exercise of an officer’s discretion, although many executive power theorists do seem to think so. As I’ve written, however, Presidents Washington, Jefferson, and even Taft—who subsequently authored the Myers v. United States opinion—thought Congress could insulate the individual decisions of an officer. But, Taft wrote, the president “may consider the decision after its rendition as a reason for removing the officer, on the ground that the discretion regularly entrusted to that officer by statute has not been on the whole intelligently or wisely exercised.”
Nelson’s intervention in this field is welcome, but it seems to me he hasn’t sufficiently shown why Taft was wrong about the bottom line.