A debate is raging and litigation has already begun over whether President Trump’s selection of Matthew Whitaker to serve as acting attorney general is constitutional. Just recently, three Democratic senators filed suit arguing that Whitaker’s appointment has unconstitutionally deprived them of their right to vote on his nomination. As this debate unfolds, one crucial piece of the puzzle will be the precedent of the two earliest Vacancies Acts, passed in 1792 and 1795. In this piece, I will examine both their language and the actual practice of how they were used to fill cabinet vacancies during the first 76 years they were in effect, to see if they might shed light on interpreting the relevant constitutional provision.
First, an explanation of the current controversy. The Appointments Clause of the U.S. Constitution states that the President shall appoint all “officers of the United States” with the advice and consent of the Senate. U.S. Const., Art. II, §2, cl. 2. However, the clause also includes a potential exemption from the need for Senate confirmation: “Congress may by Law vest the Appointment of such inferior Officers, as they think proper, in the President alone, in the Courts of Law, or in the Heads of Departments.” Thus, some officers are “inferior officers” for whom Congress has this option, while others are not “inferior.” These “non-inferior” officers have come to be known as “principal officers,” and they must be confirmed by the Senate; Congress could not exempt them from this requirement even if it wanted to. A crucial question in the current debate is where the dividing line falls between “principal” and “inferior” officers, but one point on which there is no dispute is that all cabinet secretaries, including the attorney general, are principal officers.
When a vacancy occurs in any office requiring Senate consent, a federal statute called the Federal Vacancies Reform Act (FVRA) states that the first assistant to that office shall immediately become the “acting officer.” 5 U.S.C. § 3345. Alternatively, the President may instead select anyone currently serving in any other office requiring Senate consent, or any high-ranking employee in the same department as the vacancy. Id. In any of these three alternatives, the acting officer begins serving immediately, without the need for Senate consent. The acting officer serves until either a successor is confirmed or the FVRA’s generous time limit runs out.
But as Justice Clarence Thomas pointed out in a concurring opinion last year, the FVRA’s grant of appointment power when used to select an “acting” department head seems unconstitutional. See N.L.R.B. v. SW General, 137 S. Ct. 929, 945–54 (Thomas, J., concurring). No statute may waive the requirement of Senate consent to fill any “principal office.” Yet the FVRA does exactly that, allowing the president to select an “acting attorney general” who assumes the powers of that office immediately, without the need for Senate confirmation. (Indeed, the entire point of the FVRA is for positions to remain filled during the time it takes for a permanent successor to win Senate confirmation.)
Does Thomas’s argument mean every acting attorney general is unconstitutional? Not necessarily. When the person selected for acting service is already a Senate-confirmed officer in the same department, the reasoning of two Supreme Court cases, called Shoemaker and Weiss, likely saves their constitutionality.
Both of these cases were about the Senate, and specifically what the Senate reasonably understands itself to be doing when it confirms someone to an office. The holding of Shoemaker is more easily illustrated with a recent example. Kathleen Sebelius was confirmed by the Senate to be Secretary of Health and Human Services in 2009, before the passage of the Affordable Care Act. When that act was passed in 2010, Secretary Sebelius acquired significant new powers and duties as the cabinet secretary in charge of administering the act. But the Senate, when it confirmed her, did not know she would acquire these duties. Did this mean that a new Senate confirmation vote was necessary for her to remain in office?
The Shoemaker doctrine tells us no, a new Senate confirmation vote was not necessary. When the Senate confirms someone to an office, it is on notice that the office may naturally acquire new powers by new statutes that are of the same general nature as the ones the office currently holds. Shoemaker held that so long as an office is given new duties that are reasonably in line with what would be expected for that office—what the Supreme Court called new duties “germane” to the office—no new Senate vote is required. Shoemaker v. United States, 147 U.S. 282, 300–01 (1893).
Weiss took the Shoemaker doctrine a step further. What if a Senate-confirmed officer is selected by a superior to take on new duties temporarily, such as a military officer being selected to serve for a term as a military judge? Is a new Senate confirmation vote required when such new duties are granted? The Supreme Court, by analogy to the reasoning of Shoemaker, once again held no. Just as the Senate is on notice that an officer may acquire new statutory duties germane to the office, so is it also on notice that an officer may be temporarily detailed with new assignments that are germane to the office. Weiss v. United States, 510 U.S. 163, 176 (1994).
Under the Shoemaker/Weiss line of reasoning, the selection of any Senate-confirmed officer in the Department of Justice to be acting attorney general is likely constitutional. The Senate is on notice when it confirms the deputy attorney general or solicitor general or a U.S. attorney that future circumstances might require that officer to temporarily take on the duties of the attorney general. That is why there was no constitutional controversy when Sally Yates and Dana Boente each served as acting attorney general early in the Trump Administration; each had already been confirmed by the Senate to serve in other DOJ positions.
But Matthew Whitaker was not serving in a Senate-confirmed position at the time of his ascension, and so the Shoemaker/Weiss doctrine cannot apply to him. And without the Shoemaker/Weiss doctrine, we are back to Justice Thomas’s serious argument against the FVRA’s constitutionality. That argument has convinced many legal scholars and commentators, who have called Whitaker’s acting service unconstitutional.
What is the response to this? Modern defenders of the FVRA’s constitutionality reason that the appointment of acting secretaries is constitutional because the FVRA “allows the President, by virtue of the Inferior Officers Appointments Clause, to appoint people to temporary, inferior offices distinct from the original, permanent office.” E. Garrett West, Note: Congressional Power Over Office Creation, 128 Yale L.J. 166, 215 (2018). The Office of Legal Counsel (OLC) took this same position in a 2003 opinion: “Although the position of Director is a principal office, we believe that an Acting Director is only an inferior officer.” Designation of Acting Director of the Office of Management & Budget, 27 Op. O.L.C. 121, 123 (2003). Since the Appointments Clause allows the “president alone” to appoint inferior officers when granted that power by statute, under this reasoning the FVRA’s grant of unilateral acting appointment power to the president would be permissible.
This brings us back to the crucial distinction between a principal and an inferior officer. Why would an acting attorney general, who has virtually all the powers of a permanent attorney general, be an inferior officer? Many (including the OLC) have pointed to the reasoning of the 1898 Supreme Court case United States v. Eaton, 169 U.S. 331 (1898). In that case, a foreign diplomat named Boyd, who was serving in the Senate-confirmed position of consul general in Bangkok, had fallen seriously ill. Boyd tapped a missionary named Eaton to serve as the acting consul general while Boyd attempted to recuperate and sailed back to America. Eaton continued to serve in this acting role until a senate-confirmed vice-consul arrived in Bangkok to take over.
The Supreme Court held that Boyd had only temporarily delegated his powers to Eaton, not made Eaton a new consul general. The Court reasoned that to rule Eaton’s service as a promotion to consul general (and therefore as a promotion to principal officer) would be to “void any and every delegation of power to an inferior to perform under any circumstances or exigency the duties of a superior officer, and the discharge of administrative duties would be seriously hindered.” Id. at 343.
Does Eaton settle the question? Are all acting officers merely inferior officers temporarily delegated the powers of a superior? Once again, it’s not that simple. The OLC and others who rely on Eaton gloss over a crucial distinction between absences (when a principal officer remains in office but is temporarily sick or travelling) and vacancies (when the principal officer has permanently left office due to death or resignation). The State Department’s Nineteenth Century records for all uses of the Vacancies Act to fill cabinet positions from the Jackson to the Fillmore Administrations put temporary service into two distinct categories: “acting” service, when the full-time secretary was in office but temporarily sick or travelling, and “ad interim” service, when the temporary officer filled a gap between one permanent secretary leaving office and the next permanent secretary being confirmed. See Government Printing Office, Trial of Andrew Johnson 585–88 (1868). Congress in the 1868 Vacancies Act likewise treated these two cases as very different, placing a 10-day time limit on ad interim service but no time limit on acting service. Act of July 23, 1868, ch. 227, § 3, 15 Stat. 168.
Why does this distinction matter to the question whether an acting department head is an inferior officer? As the Supreme Court has explained, “the term ‘inferior officer’ connotes a relationship with some higher ranking officer or officers below the President: Whether one is an ‘inferior’ officer depends on whether he has a superior.” Edmond v. United States, 520 U.S. 651, 662 (1997). An acting secretary standing in for a particular permanent secretary who has temporarily left the capital or temporarily fallen ill can plausibly be described as having a superior. The stand-in knows that the permanent secretary aims to return, and that the permanent secretary will be able to reverse any decisions the stand-in might make. Thus, so long as an acting secretary does not have the power to make any irreversible decisions, it is plausible that an acting secretary who serves while the permanent secretary is only temporarily absent remains an inferior officer with only sub-delegated and supervised powers. That was the situation in Eaton. But no such inferior-superior relationship exists when an office is entirely vacant due to death or resignation. Thus, applying the holding of Eaton to vacancies as opposed to absences likely extends Eaton beyond its original rationale.
But even if Eaton alone is not enough to justify the inferior-officer theory, what about the text and history of the original Vacancies Act? The fact that the earliest Congresses believed the act to be constitutional is certainly relevant to interpreting the meaning of the Appointments Clause. “The practices of the First Congress are often considered to be of extra importance in constitutional interpretation because they reflect the understanding of the Framers and the public at the time of the Founding.” Hanah Metchis Volokh, The Two Appointments Clauses: Statutory Qualifications for Federal Officers, 10 U. Pa. J. Const. L. 745, 774–75 (2008).
The first Vacancies Act was passed in 1792 by the Second Congress, and it did not place any limit on who the president could choose to serve as an acting officer. It read as follows:
In case of the death, absence from the seat of government, or sickness of the Secretary of State, Secretary of the Treasury, or of the Secretary of the War department, or of any officer of either of the said departments whose appointment is not in the head thereof, whereby they cannot perform the duties of their said respective offices, it shall be lawful for the President of the United States, in case he shall think it necessary, to authorize any person or persons at his discretion to perform the duties of the said respective offices until a successor be appointed, or until such absence or inability by sickness shall cease.
Act of May 8, 1792, ch. 37, § 8, 1 Stat. 281.
Three years later, the Third Congress made two amendments: The phrase “death, absence from the seat of government, or sickness” was replaced with the more general term “vacancy,” and Congress added a provision that “no one vacancy shall be supplied, in manner aforesaid, for a longer term than six months.” Act of Feb. 13, 1795, ch. 21, 1 Stat. 415.
In passing these acts, the early Congresses obviously believed them to be constitutional. But one strong piece of textual evidence suggests that Congress did not believe so on the theory that acting officers were “inferior officers.” The early Congresses were explicit when they created subordinate positions that were inferior officers. When creating the Departments of both War and Foreign Affairs, the First Congress “expressly designated the Secretary of the Department as a ‘principal officer,’ and his subordinate, the Chief Clerk of the Department, as an ‘inferior officer.’” Edmond, 520 U.S. at 663. But in the 1792 and 1795 Vacancies Acts, Congress did not refer to the acting official as an “inferior officer,” never using that term at all.
Why then did those who passed these acts believe they were constitutional? Several pieces of evidence suggest that rather than viewing acting officials as inferior officers, Congress instead viewed these officials as not officers at all.
Research into the meaning of the word “officer” at the time of the Framing has shown that one defining feature of an office is that it is “ongoing” or “continuing.” “Both under the Articles of Confederation and during the First Congress, there was a category of contractors or other nonofficer persons whom officers hired for services outside the Article II appointment process. Therefore, one additional requirement for federal officer status appears to be responsibility for ongoing duties.” Jennifer Mascott, Who Are “Officers of the United States”? 70 Stan. L. Rev. 443, 534 (2018). The defining characteristic of these non-officer contractors was that they were “hired to perform discrete services.” Id. at 535. And these services were not necessarily menial. In some cases, “government officials conducted discrete high-level diplomatic missions without being commissioned as foreign affairs officers.” Id. at 535–36
There is also evidence in the Constitution’s text that only those with ongoing duties are officers. “The Constitution refers to an office as something that one ‘holds’ and ‘enjoys’ and in which one ‘continues,’ and these descriptions suggest that an office has some duration and ongoing duties.” Officers of the United States Within the Meaning of the Appointments Clause, 31 Op. O.L.C. 73, 101 (2007).
In drafting the text of the 1792 act, Congress seemed to take pains to avoid describing an acting officer as actually “holding” an office. Instead, these officials are “authorize[d] . . . to perform the duties of the said respective offices.” While such semantic distinctions should not necessarily make the difference between whether a statute is upheld or struck down, they do provide valuable insight into the reasoning of Congress when it passed the act. And they strongly suggest that Congress viewed an “authorization” under the act as an assignment to temporarily perform a set of duties for the express purpose of achieving a single project: that of caretaking. Congress most likely viewed such an assignment as distinct from holding an office.
What about post-enactment practice? What, if anything, do the actual choices made by presidents tell us about how the act was interpreted? The 1868 Vacancies Act required, for the first time, that all acting officers chosen by the president must have been confirmed by the Senate to some other office. Thus, the period of 1792–1868 was the only period in which presidents had the option to appoint non-Senate-confirmed persons to be acting department heads (until the Vacancies Act was amended in 1998, adding the statutory provision under which Whitaker was selected).
Using several sources, I attempted to make as complete a list as possible (though one in which I have likely missed at least a few) of acting Secretaries of State, War, and Treasury during this period. What does this history show? First, although the early Vacancies Acts made no distinction between who could be appointed during absences and vacancies, presidents showed a clear pattern in their selections. In the case of absences of the secretary due to travel or sickness, the chief clerk in the department, an inferior officer not confirmed by the Senate, was usually chosen to serve as acting secretary. In the case of vacancies, however, it was more common for presidents to select another Senate-confirmed secretary to serve as the ad interim secretary, especially if the vacancy was to last a significant time.
Specifically, for Secretary of State, fourteen total ad interim appointments were made. Eight appointees were Senate-confirmed cabinet officers, whose tenure of ad interim service ranged from 12 to 197 days and averaged 64 days. Five appointees were chief clerks of the State Department, whose tenures ranged from 2 to 6 days. In one unusual case, the ad interim appointee was not a federal official at all: James Alexander Hamilton (son of the founding father), an ally of Martin Van Buren in New York politics, served as ad interim secretary of state for 24 days at the beginning of Andrew Jackson’s administration, when Van Buren was forced to remain as governor of New York in Albany. See James A. Hamilton, Reminiscences of James A. Hamilton: or, Men and Events 92 (1869).
For Secretary of War, ten total ad interim appointments were made. Four were sitting cabinet members, with tenures ranging from 11 to 154 days and averaging 85 days. Two, Winfield Scott and Ulysses S. Grant, were Commanding Generals of the U.S. Army, a position requiring the advice and consent of the Senate. They served for 24 and 155 days, respectively. Four were chief clerks of the War Department, serving for 18, 30, 33, and 352 days. The longest of these, George Graham’s service from 1816 to 1817, is notable for apparently violating the Vacancies Act’s then-six-month time limit on acting service.
The Secretary of the Treasury had fewer and shorter vacancies during this period, and the practice for filling them also differed. Only once did a cabinet secretary serve as acting treasury secretary, for 2 days in 1860. The other seven ad interim secretaries were all chief clerks of the treasury or assistant secretaries of the treasury (also not senate-confirmed), whose tenures of ad interim service ranged from 2 to 62 days and averaged 19 days.
What does this history show? Presidents certainly did, on some occasions, take advantage of the wide latitude of the early Vacancies Act and appoint non-Senate-confirmed persons to be ad interim department heads—precisely the same choice that was made in appointing Whitaker. But in general, presidents seemed hesitant to do so if the vacancy was expected to last for a significant length of time. The year for which George Graham served as ad interim Secretary of War is the greatest outlier, but it was apparently caused by four people turning down the permanent position unexpectedly, until John Calhoun finally accepted. See Frances Packard Young, John C. Calhoun as Secretary of War, 1817–1825, The Quarterly of the Oregon Historical Society, Vol. 13, No. 3, 297, 300 (1912). And if Eaton is read with proper narrowness, the Supreme Court never endorsed the constitutionality of such an appointment.
Where does this leave us with regards to the current controversy? The majority of those who defend the constitutionality of Whitaker’s appointment do so on the “inferior officer” theory, but such a theory is incompatible with the original understanding of “inferior” as meaning “subordinate.” While an acting department head who serves during the principal’s sickness or distant travel might plausibly be inferior, an ad interim head like Whitaker, who serves after a resignation, is inferior to no one but the president.
Whether an “acting” official is an officer at all is a closer question; David Rivkin for one has argued that such an official is not. I myself lean toward the view that possessing all the powers of a department head is such an open-ended charge that it cannot be considered a discrete “contract,” even if it only lasts for a single day. And although early Congresses are given significant weight in constitutional interpretation, they were not infallible. (After all, the Sedition Act was passed only five years after the Bill of Rights). It is not heresy to argue that an early act of Congress was unconstitutional, and in fact others have done so with regard to the 1792 Vacancies Acts. See Michael B. Rappaport, The Original Meaning of the Recess Appointments Clause, 52 UCLA L. Rev. 1487, 1516 n.80 (2005). In my view, this act was passed because Congress quickly realized that the Constitution should have included a provision analogous to the Recess Appointments Clause to deal with unexpected vacancies that occurred while the Senate was in session. Faced with the need for such a clause, Congress passed the act with a good faith belief in its constitutionality, writing it in such a way as to minimize the officer-like qualities of acting officials.
But in the context of the current version of the Vacancies Act, resolving this debate is probably unnecessary. Under the current law, an acting officer may serve for up to seven months plus the pendency of up to two nominations, no matter how long those might take. 5 U.S.C. § 3346. Such lengthy periods of acting service can make “acting” officers indistinguishable from permanent ones. Cf. SW General, 137 S. Ct. at 946 n.1 (Thomas, J., concurring) (noting that an acting officer had “served for more than three years in an office limited by statute to a 4–year term”). I believe the lack of any hard statutory deadline eliminates any plausible argument that an acting officer today is a mere “contractor.”
The recent OLC opinion defending Whitaker’s appointment does not provide a convincing argument to the contrary. First, the opinion notably fails to engage with the text of the Appointments Clause itself. Rather than firmly arguing in favor of either the “inferior officer” or “non-officer” theory, the opinion instead hedges by pointing out that Whitaker’s appointment satisfies the requirements for inferior officers and therefore “it does not matter whether an acting official temporarily filling a vacant principal office is an inferior officer or not an ‘officer’ at all within the meaning of the Constitution . . . .” Steven A. Engel, Assistant Attorney General, Office of Legal Counsel, Designating an Acting Attorney General: Memorandum for Emmet T. Flood, Counsel to the President, at 6 (Nov. 14, 2018). In refusing to commit to one textual theory, the opinion does not make the case for either.
It is instead post-1792 practice upon which the opinion almost exclusively relies. But here, the opinion’s flaw is that, like previous OLC opinions, it once again glosses over the distinction between acting and ad interim officers. The opinion lumps these two categories together in order to reach the figure of “at least 160” non-Senate-confirmed temporary department heads serving between 1809 and 1860. Id. at 10. But examining these appointments in more detail shows that the significant majority were only acting heads, not ad interim. To give one example, the opinion states that at least 110 chief clerks temporarily led the Departments of State, War, and Treasury between 1809 and 1860 Id. at 9–10. Of the 93 that I could identify, 80 were acting heads serving during travel or sickness, and only 13 were ad interim. By not differentiating these two types of temporary service, the opinion gives the impression that non-Senate-confirmed officials were appointed to lead departments after a death or resignation far more frequently than actually occurred.
Finally, the opinion’s discussion of tenure focuses only on actual length of acting service, rather than the more relevant issue of potential length. The opinion accepts that “‘the tenure of an Acting Director should not continue beyond a reasonable time.” Id. at 15 (quoting Status of the Acting Director, Office of Management and Budget, 1 Op. O.L.C. 287, 289–90 (1977)). It then declares Whitaker’s tenure to be sufficiently temporary because it “was made one week ago” and “would lapse in the absence of a presidential nomination . . . .” Id. at 15. But more relevant than how long Whitaker has served is the fact that he could serve indefinitely, so long as a nomination is pending. This potential for indefinite acting service under the current Vacancies Act fundamentally differentiates ad interim officers today from those who served in the 1795–1868 period.
When acting officers served for only a limited, non-extendable amount of time, a more plausible argument could have been made that they were brought on to perform only a distinct and non-continuing project. Indeed, this is one potential explanation for why Congress amended the first Vacancies Act after only three years to add a six-month upper limit on acting service. (And although it would be difficult to enforce, an even stronger case could be made that acting service is non-continuing if acting department heads were limited to only overseeing the projects started by their predecessors, not starting any new projects.) Whether the early Congresses were correct in their view that the time-limited 1795 act complied with the Appointments Clause is a close question. But under the current Vacancies Act, the question is less close. Given that Whitaker has not been confirmed by the Senate, has no superior but the president, and has no certain end-date for his duties, I believe it is highly likely that his service violates the Appointments Clause.
Thomas Berry is an attorney at Pacific Legal Foundation. He co-authored the amicus brief of the Cato Institute supporting the respondent in N.L.R.B. v. SW General, and authored the first published law review article analyzing the opinions in that case.
 My main sources were three incredibly helpful lists compiled by the State Department in 1868 and entered into the Senate records during the impeachment trial of Andrew Johnson. See Trial of Andrew Johnson 575–81, 585–88, 590–94. (Hat tip to Michael Ramsey and Andrew Hyman of Originalism Blog for pointing out the existence of these records). Unfortunately, these lists only covered the administrations of Jackson through Johnson, and so I also used two other purportedly complete lists of ad interim secretaries of state and war. See Department of State, Office of the Historian, “Secretaries of State ad interim,” last accessed Nov. 13, 2018; William Gardner Bell, Secretaries of War and Secretaries of the Army, Appendix A (1992). I checked all three of these records against the list of permanent and ad interim secretaries appearing in The Biographical Directory of the American Congress, 1774–1971 13–20 (1971). Where there was a discrepancy between the sources, I favored the original source of the State Department records. In some instances, a successor to an ad interim officer is listed as taking office on one date but not “entering upon duties” until a later date. In those cases, I have chosen to calculate the length of ad interim service as ending on the earlier date. When a successor has been confirmed but is absent, the temporary service becomes acting rather than ad interim.