Aditya Bamzai and Saikrishna Bangalore Prakash

How to Think About the Removal Power

In an earlier article titled The Executive Power of Removal, we contended that Article II gives the President a constitutional power to remove executive officers, at least those who are presidentially appointed. In this Essay, we expand on, and reply to a critique of, that article. We discuss the meaning of the clause vesting “executive Power” in the President and the clause authorizing Congress to make laws “necessary and proper for carrying into Execution” the powers of the federal government. We contend that the former vests authority to remove in the President and the latter does not allow Congress to treat that allocation of authority as a default. We discuss how constitutional developments in the Commonwealth of Pennsylvania—specifically, a 1784 report authored by the Council of Censors—support our understanding of the federal Constitution’s text and structure. We also discuss early practice under the federal Constitution—specifically, high-profile instances where presidents removed executive subordinates without Senate participation. These sources and episodes, along with those we discussed in our previous article, support the conclusion that the Constitution confers on the President the authority to remove presidentially appointed executive officers.

Introduction

Does the President have a constitutional power to remove executive branch subordinates? In a recent article entitled The Executive Power of Removal,1 1.See Aditya Bamzai & Saikrishna Bangalore Prakash, The Executive Power of Removal, 136 Harv. L. Rev. 1756 (2023).Show More we joined the Supreme Court2 2.For recent cases, see Free Enter. Fund v. Pub. Co. Acct. Oversight Bd., 561 U.S. 477 (2010); Seila Law LLC v. CFPB, 140 S. Ct. 2183 (2020); Collins v. Yellen, 141 S. Ct. 1761 (2021).Show More in defending the proposition that Article II of the Constitution gives the President authority to remove executive officers, at least those who are presidentially appointed.3 3.See Myers v. United States, 272 U.S. 52 (1926). As explained, we did not address those officials performing functions for the territories and the District of Columbia—where Congress might have greater authority to structure administration. See Bamzai & Prakash, supra note 1, at 1802–18. Moreover, we did not dispute that a non-executive institution, like Congress, can initiate “removal” of an executive officer, say, by impeachment. See U.S. Const. art. II, § 4. And we bracketed the question of whether the President has similar removal authority over inferior officers appointed by others. See Bamzai & Prakash, supra note 1, at 1830–35.Show More Without such a power, it is hard to see how the President could exert control—on behalf of an electoral coalition—over the vast American bureaucracy. Without such a power, it is easy to see how a temporary coalition could entrench long-term control over the bureaucracy by creating an officer insulated from presidential control through, for example, the conferral of statutory life tenure with removal only by impeachment. The Constitution’s conferral of removal authority on the President thus has a deep and important connection to the concept that electoral majorities should be able to control the executive branch. In our previous article, we focused on historical sources that had embraced the perspective that the President has just such a power of removal.

In this Essay, we reply to a critique of that article—Removal Rehashed by Professors Andrea Katz and Noah Rosenblum.4 4.See Andrea Scoseria Katz & Noah A. Rosenblum, Removal Rehashed, 136 Harv. L. Rev. F. 404 (2023).Show More We part ways with their analysis in several significant respects. But despite our disagreements, we are grateful for the chance to sharpen our own thinking on these issues. Just as the hammer and the anvil forge the metal, so too in the realm of intellectual discovery the critic forces the author to refine arguments that would otherwise remain untested. In that spirit, we offer this reply. We continue to believe that, although our theory is not the only one possible, it best fits text, structure, history, and early practice, and is therefore preferable to the alternatives.

In contrast to our views, Katz and Rosenblum reject altogether the notion that the Constitution confers a removal power on the President.5 5.As best we can tell, Katz and Rosenblum do not concede that the President has a removal authority that Congress can regulate using the Necessary and Proper Clause. Thus, it appears to us that their view departs from the perspective of those, like Justice Kagan, who believe that the President possesses an indefeasible constitutional power to remove close military or diplomatic advisors, because the absence of such power would impede the President’s performance of his constitutional duties. See, e.g., Seila Law, 140 S. Ct. at 2233 (Kagan, J., concurring in the judgment with respect to severability and dissenting in part).Show More Starting from that perspective, they advise that readers “will find little new” in our article6 6.Katz & Rosenblum, supra note 4, at 425.Show More—a refrain they repeat so many times and so fixatedly that it takes on the air of a government official advising a passerby to “move on; nothing to see here!”7 7.See id. at 404 (“We are not convinced that the Article says much that is new.”); id. (“[I]t was unclear to us which materials were new or what the new materials added.”); id. at 406 (arguing that “Bamzai and Prakash’s argument is not new”); id. at 416 (claiming that our argument “largely rehashes old arguments with old sources”); id. at 417 (“We are unsure what Bamzai and Prakash believe is new about their argument.”).Show More To be sure, if one starts from the premise, as Katz and Rosenblum do, that it is “intellectually indefensible” to believe that Article II grants a presidential removal authority,8 8.Id. at 405.Show More then we agree: there is nothing to see here. But for those who are more open-minded about one of the most significant (and historically, most debated) questions of the separation of powers, read on.9 9.Of course, feel free to read The Executive Power of Removal too.Show More For as we explain below, Katz and Rosenblum misdescribe several of our arguments and several of the underlying sources. The case for a presidential removal power is stronger than they are willing to acknowledge.

Consider, for example, one of Katz and Rosenblum’s claims about the historical pedigree of the President’s removal authority. They contend that “[t]he historian might wonder why th[e] argument [for an executive power of removal,] if once so widespread, disappeared so quickly.”10 10.Katz & Rosenblum, supra note 4, at 416. Parenthetically, as anyone familiar with history knows, arguments can come and go, such that even if an argument disappeared at some point, it would not necessarily dispose of a claim that the argument existed in, for example, the eighteenth and early nineteenth centuries. At any rate, Katz and Rosenblum’s assertion that our claim “disappeared” is contradicted by the very source that they cite.Show More To support their claim that the argument for a removal power “disappeared so quickly,” Katz and Rosenblum rely on a quotation from a 1916 book by Frank Goodnow providing that courts “have held that [the Vesting Clause] has little if any legal effect, and that for the most part it is to be explained by the powers which are later specifically mentioned.”11 11.Id. at 416 n.94 (quoting Frank J. Goodnow, Principles of Constitutional Government 88–89 (1916)).Show More But Katz and Rosenblum’s use of this quotation does not properly characterize Goodnow. Two pages after the quoted language, Goodnow explained that the “practice” with respect to removal is “that the President has the power to remove arbitrarily almost all civil officers of the United States, not judges. This power has been recognized as belonging to the President as a part of the executive power granted to him.”12 12.Here is Goodnow’s paragraph, which we quote in its entirety to allow readers to assess the relevant context:The United States Constitution, as we have seen, vests the executive power in a President. The meaning of the power thus granted is, however, to be obtained from the powers subsequently specifically enumerated. These are:1. The power to appoint all officers of the government except inferior officers, who, if so provided by law, may be appointed by their superiors or by the courts. This power, where not otherwise provided by law, is to be exercised with the approval of the Senate. No mention is made in the Constitution of any power of removal from office. All that is said with regard to the termination of office is contained in the provision with regard to impeachment, which is applicable to all civil officers, and that giving the judges a term of office during good behavior. The practice is, however, that the President has the power to remove arbitrarily almost all civil officers of the United States, not judges. This power has been recognized as belonging to the President as a part of the executive power granted to him.Frank J. Goodnow, Principles of Constitutional Government 91 (1916). As this language makes clear, Katz and Rosenblum simply misinterpret Goodnow. Even while questioning whether other presidential powers derived from the Vesting Clause, Goodnow plainly acknowledged that the executive power was thought to encompass removal. (We leave to one side the question of whether a 1916 disappearance would count as “quickly” following the Constitution’s adoption.)Show More Rather than demonstrating the “disappearance” of an “executive power of removal,” Goodnow’s 1916 book demonstrates how into the twentieth century, it was widely recognized that Article II conferred removal power on the President.13 13.In a 1905 book, Goodnow foreshadowed the perspective that he explicitly articulated in 1916. See Frank J. Goodnow, The Principles of the Administrative Law of the United States (1905). There, Goodnow described “the interpretation of the constitution made by the first Congress relative to the President’s power of removing officers” as having been “that the power of removal was a part of the executive power, and therefore belonged to the President.” Id. at 76. He claimed that this was “the recognized construction of the constitution” until the Civil War. Id.; see also id. at 77 (remarking that, after the repeal of the limits on presidential removal imposed by the Tenure of Office Act, “the early interpretation of the constitution must be regarded as the correct one at the present time” and describing the conferral of removal authority on the President as having “been of incalculable advantage in producing an efficient and harmonious national administration”). We might part ways with some of the details in Goodnow’s account, and we take no position on whether Goodnow expressed different views in other writings that we have not mentioned. At a minimum, however, these discussions from Goodnow’s 1905 and 1916 books flatly contradict Katz and Rosenblum’s assertion that the concept of an executive power of removal “disappeared . . . quickly.” Katz & Rosenblum, supra note 4, at 416.Show More Professor Goodnow once remembered what some modern historians have forgotten.

As we discuss below, this is not the only occasion where we part ways with Katz and Rosenblum’s characterization of our article or the underlying sources. Specifically, they spend a significant portion of their response on the question of removal under the Pennsylvania Constitution—claiming, for example, that “Pennsylvania’s charter made no mention of executive removal”14 14.Katz & Rosenblum, supra note 4, at 407 (citing Pa. Const. of 1776, §§ 20, 22–23, 30, 34).Show More and that a 1784 report by Pennsylvania’s Council of Censors “probably means nearly the opposite” of our characterization of it.15 15.Id. at 417–18.Show More Respectfully, we disagree. Katz and Rosenblum fail to recognize that the Pennsylvania Supreme Court explained at an early date that, despite the Pennsylvania Constitution’s silence, “it has been generally supposed, that the power of removal rested with the Governor, except in those cases where the tenure, was during good behavior.”16 16.Commonwealth ex rel. Lehman v. Sutherland, 3 Serg. & Rawle 145, 149 (Pa. 1817).Show More And they bury in a footnote a concession that a passage in the Censors’ Report “might be read to suggest that the Censors believed removal was ‘an executive power’ and so support Bamzai and Prakash’s argument.”17 17.Katz & Rosenblum, supra note 4, at 420 n.111.Show More As we explain at length below, their concession is appropriate; the Censors’ Report supports our position. And the early history of gubernatorial removal in Pennsylvania is itself a fascinating case study with parallels to federal removal practice.

In addition, Katz and Rosenblum claim we were mistaken to rely on data from a study by the political scientist Carl Fish to show how often presidents removed executive subordinates in the early Republic. They claim that many such removals occurred on appointment of a successor, because for positions requiring Senate advice and consent, “removal was incident to appointment: the appointment and confirmation of someone new removed the previous officeholder.”18 18.Id. at 421. As an initial matter, Katz and Rosenblum suggest that removals that occurred upon the appointment of a successor did not happen “in the way Bamzai and Prakash use the term in their Article.” Id. at 422. But that misdescribes our article, which purposefully did not take a position on how removal had to be accomplished. Cf. Bamzai & Prakash, supra note 1, at 1787 (discussing issues that might arise due to a dispute over the timing of a removal).Show More But leaving to one side that nothing in our claim turns on Fish’s precise number of removals (which we did not even cite), we explain below that, in high-profile instances, presidents removed executive subordinates without Senate participation. Katz and Rosenblum’s theory of removal-by-appointment fails to explain such removals. And as demonstrated by the at-pleasure commissions conferred on executive branch officials, along with statements by executive branch officers, it was certainly the view of many that presidents could unilaterally remove.

Our Essay proceeds as follows. In Part I, we set forth the analytical framework for a presidential removal power. At the risk of “rehashing”—which we now understand to be strictly verboten—we rely upon some of the same material we previously surfaced in The Executive Power of Removal. In Part II, we turn to a significant state-law antecedent to the federal Constitution—removal in the context of the Pennsylvania Constitution. This portion of the Essay introduces “new” sources and arguments from one State that might have played a role in the drafting of the federal Constitution.19 19.We use the word “new” with some trepidation, given how hawkishly it appears some of our interlocutors police that language.Show More In Part III, we address the role of early federal practice. Finally, in Part IV, we address some overarching methodological points.

  1.  See Aditya Bamzai & Saikrishna Bangalore Prakash, The Executive Power of Removal, 136 Harv. L. Rev. 1756 (2023). ↑
  2.  For recent cases, see Free Enter. Fund v. Pub. Co. Acct. Oversight Bd., 561 U.S. 477 (2010); Seila Law LLC v. CFPB, 140 S. Ct. 2183 (2020); Collins v. Yellen, 141 S. Ct. 1761 (2021). ↑
  3.  See Myers v. United States, 272 U.S. 52 (1926). As explained, we did not address those officials performing functions for the territories and the District of Columbia—where Congress might have greater authority to structure administration. See Bamzai & Prakash, supra note 1, at 1802–18. Moreover, we did not dispute that a non-executive institution, like Congress, can initiate “removal” of an executive officer, say, by impeachment. See U.S. Const. art. II, § 4. And we bracketed the question of whether the President has similar removal authority over inferior officers appointed by others. See Bamzai & Prakash, supra note 1, at 1830–35. ↑
  4.  See Andrea Scoseria Katz & Noah A. Rosenblum, Removal Rehashed, 136 Harv. L. Rev. F. 404 (2023). ↑
  5.  As best we can tell, Katz and Rosenblum do not concede that the President has a removal authority that Congress can regulate using the Necessary and Proper Clause. Thus, it appears to us that their view departs from the perspective of those, like Justice Kagan, who believe that the President possesses an indefeasible constitutional power to remove close military or diplomatic advisors, because the absence of such power would impede the President’s performance of his constitutional duties. See, e.g., Seila Law, 140 S. Ct. at 2233 (Kagan, J., concurring in the judgment with respect to severability and dissenting in part). ↑
  6.  Katz & Rosenblum, supra note 4, at 425. ↑
  7.  See id. at 404 (“We are not convinced that the Article says much that is new.”); id. (“[I]t was unclear to us which materials were new or what the new materials added.”); id. at 406 (arguing that “Bamzai and Prakash’s argument is not new”); id. at 416 (claiming that our argument “largely rehashes old arguments with old sources”); id. at 417 (“We are unsure what Bamzai and Prakash believe is new about their argument.”). ↑
  8.  Id. at 405. ↑
  9.  Of course, feel free to read The Executive Power of Removal too. ↑
  10.  Katz & Rosenblum, supra note 4, at 416. Parenthetically, as anyone familiar with history knows, arguments can come and go, such that even if an argument disappeared at some point, it would not necessarily dispose of a claim that the argument existed in, for example, the eighteenth and early nineteenth centuries. At any rate, Katz and Rosenblum’s assertion that our claim “disappeared” is contradicted by the very source that they cite. ↑
  11.  Id. at 416 n.94 (quoting Frank J. Goodnow, Principles of Constitutional Government 88–89 (1916)). ↑
  12.  Here is Goodnow’s paragraph, which we quote in its entirety to allow readers to assess the relevant context:

    The United States Constitution, as we have seen, vests the executive power in a President. The meaning of the power thus granted is, however, to be obtained from the powers subsequently specifically enumerated. These are:

    1. The power to appoint all officers of the government except inferior officers, who, if so provided by law, may be appointed by their superiors or by the courts. This power, where not otherwise provided by law, is to be exercised with the approval of the Senate. No mention is made in the Constitution of any power of removal from office. All that is said with regard to the termination of office is contained in the provision with regard to impeachment, which is applicable to all civil officers, and that giving the judges a term of office during good behavior. The practice is, however, that the President has the power to remove arbitrarily almost all civil officers of the United States, not judges. This power has been recognized as belonging to the President as a part of the executive power granted to him.

    Frank J. Goodnow, Principles of Constitutional Government 91 (1916). As this language makes clear, Katz and Rosenblum simply misinterpret Goodnow. Even while questioning whether other presidential powers derived from the Vesting Clause, Goodnow plainly acknowledged that the executive power was thought to encompass removal. (We leave to one side the question of whether a 1916 disappearance would count as “quickly” following the Constitution’s adoption.) ↑

  13.  In a 1905 book, Goodnow foreshadowed the perspective that he explicitly articulated in 1916. See Frank J. Goodnow, The Principles of the Administrative Law of the United States (1905). There, Goodnow described “the interpretation of the constitution made by the first Congress relative to the President’s power of removing officers” as having been “that the power of removal was a part of the executive power, and therefore belonged to the President.” Id. at 76. He claimed that this was “the recognized construction of the constitution” until the Civil War. Id.; see also id. at 77 (remarking that, after the repeal of the limits on presidential removal imposed by the Tenure of Office Act, “the early interpretation of the constitution must be regarded as the correct one at the present time” and describing the conferral of removal authority on the President as having “been of incalculable advantage in producing an efficient and harmonious national administration”). We might part ways with some of the details in Goodnow’s account, and we take no position on whether Goodnow expressed different views in other writings that we have not mentioned. At a minimum, however, these discussions from Goodnow’s 1905 and 1916 books flatly contradict Katz and Rosenblum’s assertion that the concept of an executive power of removal “disappeared . . . quickly.” Katz & Rosenblum, supra note 4, at 416. ↑
  14.  Katz & Rosenblum, supra note 4, at 407 (citing Pa. Const. of 1776, §§ 20, 22–23, 30, 34). ↑
  15.  Id. at 417–18. ↑
  16.  Commonwealth ex rel. Lehman v. Sutherland, 3 Serg. & Rawle 145, 149 (Pa. 1817). ↑
  17.  Katz & Rosenblum, supra note 4, at 420 n.111. ↑
  18.  Id. at 421. As an initial matter, Katz and Rosenblum suggest that removals that occurred upon the appointment of a successor did not happen “in the way Bamzai and Prakash use the term in their Article.” Id. at 422. But that misdescribes our article, which purposefully did not take a position on how removal had to be accomplished. Cf. Bamzai & Prakash, supra note 1, at 1787 (discussing issues that might arise due to a dispute over the timing of a removal). ↑

  19.  We use the word “new” with some trepidation, given how hawkishly it appears some of our interlocutors police that language. ↑

©Virginia Law Review Association. All rights reserved.