Seven words stand between the President and the heads of over a dozen “independent agencies”: inefficiency, neglect of duty, and malfea­sance in office (INM). The President can remove the heads of these agencies for INM and only INM. But neither Congress nor the courts have defined INM and hence the extent of agency independence. Stepping into this void, some proponents of presidential power argue that INM allows the President to dismiss officials who do not follow presidential directives. Others contend that INM is unconstitutional because it prevents Presidents from fulfilling their duty to take care that the laws are faithfully executed. This Article recovers the lost history of INM, explaining its origins and meaning, inverting our current under­standing of its purpose, and rejecting both challenges to agency independence. It shows that INM provisions are not removal “protec­tions” that prevent at-pleasure removal; they are removal permissions that authorize removal where it is otherwise prohibited by an officer’s term of years, a tenure long understood to bar executive removal for any reason. INM provisions are narrow exceptions to term tenures: Neglect of duty and malfeasance in office cash out an official’s failure to faith­fully execute official duties, while inefficiency relates to government waste and ineptitude. INM provisions do not permit the President to remove agency heads for failing to follow presidential directives. But they do not clash with the Take Care Clause either, because even on an expansive reading of the clause, INM provisions authorize Presidents to remove unfaithful or incompetent officials.

The full text of this Article can be found by clicking the PDF link to the left.


Independent agencies are government bodies whose leaders do not serve at the pleasure of the President or other government officials. 1 Most definitions of “independent agency” encompass only government bodies headed by officials who are not removable by the President at will. See, e.g., Jacob E. Gersen, Designing Agencies, in Research Handbook on Public Choice and Public Law 333, 347 (Daniel A. Farber & Anne Joseph O’Connell eds., 2010); Marshall J. Breger & Gary J. Edles, Established by Practice: The Theory and Operation of Independent Federal Agencies, 52 Admin. L. Rev. 1111, 1138 (2000); Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2376 (2001); Geoffrey P. Miller, Introduction: The Debate over Independent Agencies in Light of Empirical Evidence, 1988 Duke L.J. 215, 216–17. This Article applies a slightly different definition that includes officials, like the Comptroller General, whose “dependence” or “independence” mostly involves another branch of government. See, e.g., Bowsher v. Synar, 478 U.S. 714, 737 (1986) (Stevens, J., concurring) (“The fact that Congress retained for itself the power to remove the Comptroller General is important evidence supporting the conclusion that he is a member of the Legislative Branch . . . .”). It also explicitly excludes government bodies whose leaders are not removable by the President at will but are removable by officials who serve at the pleasure of the President. See, e.g., 31 U.S.C. § 303 (2018) (the Bureau of Engraving and Printing). This definition is narrower than those that treat agency independence as a function of several different factors of which tenure in office is only one. See, e.g., Free Enter. Fund v. Pub. Co. Acct. Oversight Bd., 561 U.S. 477, 547 (2010) (Breyer, J., dissenting) (adopting a broader definition); Robert E. Cushman, The Independent Regulatory Commissions 3 (1941) (same); Kirti Datla & Richard L. Revesz, Deconstructing Independent Agencies (and Executive Agencies), 98 Cornell L. Rev. 769, 772 (2013) (same). Although independent agencies are common creatures in our political ecosystem, 2 See Free Enter., 561 U.S. app. A at 549 (Breyer, J., dissenting) (identifying forty-eight federal independent agencies); infra Appendix B; see also Datla & Revesz, supra note 1, at 786. their legality and independence are hotly contested. Promi­nent jurists argue that some or all conflict with Article II of the Constitution, which vests “executive [p]ower” in the President and requires the President to “take [c]are” that the laws are “faithfully exe­cuted.” 3 U.S. Const. art. II, §§ 1, 3. Other proponents of presidential power contend that there is no constitutional problem with independent agencies because inde­pendent agencies are actually subject to a good deal of presidential con­trol. On this view, the President already has the power under existing law to remove the heads of independent agencies for failing to follow directives or achieve White House policy goals. 4 See infra section I.B.1. In an important new article, Professors Cass R. Sunstein and Adrian Vermeule stake out a modified version of this position, focusing on an official’s neglect of their statutory duties under the Administrative Procedure Act. Cass R. Sunstein & Adrian Vermeule, Presidential Review: The President’s Statutory Authority over Independent Agencies, 109 Geo. L.J. (forthcoming 2021), (on file with the Columbia Law Review).

Much of this debate centers on the statutory provisions that define the President’s removal authority. These provisions typically permit the President to remove independent agency heads for cause. Acts creating the Federal Reserve System, 5 12 U.S.C. § 242 (2018). the Postal Service, 6 39 U.S.C. § 202 (2018). and the Federal Housing Finance Agency (FHFA) 7 12 U.S.C. § 4512(b)(2). use precisely these words (“for cause”). But most laws specify three causes: inefficiency, neglect of duty, and malfeasance in office (INM). 8 See infra section I.A, Appendix B. Federal agencies with INM provisions in their enabling acts include the Federal Trade Commission (FTC), 9 15 U.S.C. § 41 (2018). the National Transporta­tion Safety Board (NTSB), 10 49 U.S.C. § 1111(c) (2018). the Office of Special Counsel (OSC), 11 5 U.S.C. § 1211(b) (2018). the Federal Energy Regulatory Commission (FERC), 12 42 U.S.C. § 7171(b)(1) (2018). and the Consumer Financial Protection Bureau (CFPB). 13 12 U.S.C. § 5491(c)(3) (2018). In recent decades, courts have even read INM provisions  into  statutes,  like  the  Securities  Exchange  Act, 14 See 15 U.S.C. § 78d(a).   that  do  not  include them. 15 See, e.g., Free Enter. Fund v. Pub. Co. Acct. Oversight Bd., 561 U.S. 477, 496 (2010) (noting that the President might “intervene” in the SEC only if its determinations are “so unreasonable as to constitute ‘inefficiency, neglect of duty, or malfeasance in office’” (quoting Humphrey’s Ex’r v. United States, 295 U.S. 602, 620 (1935))); PHH Corp. v. Consumer Fin. Prot. Bureau, 881 F.3d 75, 94 (D.C. Cir. 2018) (“[T]he independence of financial regulators . . . is so well established by tradition and precedent that courts have assumed these agencies’ heads have removal protection even in the absence of clear statutory text so directing.”), abrogated by Seila L. LLC v. Consumer Fin. Prot. Bureau, 140 S. Ct. 2183 (2020).

Yet despite the critical role these terms play in shaping the relation­ship between independent agencies and the President, there is no consensus about what they actually mean. Neither Congress nor the Supreme Court has ever defined INM provisions, and in recent years, appeals court judges have been unable to agree on their scope and, hence, on the extent of agency independence. 16 See PHH Corp., 881 F.3d at 127–28 (Griffith, J., concurring) (“[T]he meaning of the standard’s three grounds for removal remains largely unexamined. Congress has nowhere defined these grounds and the Supreme Court has provided little guidance about the conditions under which they permit removal.”); Breger & Edles, supra note 1, at 1144–45 (noting that there is no accepted definition of inefficiency, neglect of duty, or malfeasance in office); Datla & Revesz, supra note 1, at 787 (same); Lawrence Lessig & Cass R. Sunstein, The President and the Administration, 94 Colum. L. Rev. 1, 110 (1994) (same); Richard H. Pildes & Cass R. Sunstein, Reinventing the Regulatory State, 62 U. Chi. L. Rev. 1, 29 (1995) (same). The meaning of the related term, “cause,” is also undefined, see John F. Manning, The Independent Counsel Statute: Reading “Good Cause” in Light of Article II, 83 Minn. L. Rev. 1285, 1306 (1999) [hereinafter Manning, Independent Counsel] (noting that the Court has not defined “good cause” in a removal decision), and actively disputed, with the Supreme Court facing the question this term, see infra note 367. Can the President remove members of the Federal Reserve’s Board of Governors for keep­ing interest rates too high? 17 See, e.g., Peter Conti-Brown, Can Trump Fire Jerome Powell? It’s a Political Question, Wall St. J. (Dec. 10, 2018), (on file with the Columbia Law Review) (suggest­ing that Trump might be permitted to remove Powell from his position as Chairman of the Federal Reserve’s Board); Jeanna Smialek, Trump Redoubles Attacks on Fed Chair, Saying ‘I Made Him’, N.Y. Times (June 26, 2019),‌/business/jerome-powell-donald-trump.html (on file with the Columbia Law Review) (contrasting President Trump’s assertions of his ability to fire the Federal Reserve chair with the requirement that Fed governors only be removed “for cause”). Do statutory limits on the President’s power to remove agency officials conflict with the President’s constitutional duty to take care that the laws are faithfully executed? 18 U.S. Const. art. II, § 3; see also, e.g., Seila L., 140 S. Ct. at 2191. The Supreme Court indirectly addressed these questions eighty-five years ago in Humphrey’s Executor v. United States, 19 Humphrey’s Ex’r, 295 U.S. at 632. but judges and scholars alike are unsure why the Court decided that case the way that it did. Among other things, the origins of the INM standard are forgotten, as are the goals of the legislators who incorporated it into the federal code. 20 For an excellent summary of the open issues, see Peter M. Shane, Harold H. Bruff & Neil J. Kinkopf, Separation of Powers Law: Cases and Materials 514–15 (4th ed. 2018).

This Article seeks to recover this lost understanding. It reconstructs the history of INM and examines its role in federal law. In so doing, it refutes the conventional interpretation of removal provisions as “protections”—text that prevents the President from removing inde­pend­ent agency heads at pleasure. 21 See, e.g., PHH Corp., 881 F.3d at 173 n.1 (Kavanaugh, J., dissenting) (“In general, an agency without a for-cause removal statute is an executive agency, not an independent agency, because the President may supervise, direct, and remove at will the heads of those agencies.”); supra note 15. Rather, it shows that the default runs in the other direction—against removal, not for it. When officers are appointed for a “term of years” with the stipulation that the President may remove them for inefficiency, neglect of duty, or malfeasance in office, the language that protects them from removal at pleasure is not INM—it is the term of years. 22 For more on the lost history of this tenure and its implications for modern doctrine and practice, see Jane Manners & Lev Menand, Recovering the Forgotten Tenure of a Term of Years (June 2020) (unpublished manuscript) (on file with the Columbia Law Review). Since before the Founding, offices held for a term of years, in the absence of constitutional or statutory language to the contrary, were designed to be inviolable: Short of impeachment, their holders could not be removed before the end of their terms. Statutory words like “ineffi­ciency” and “malfeasance” that qualified this protection were permissions—they authorized the removal of officers who were otherwise not removable. 23 See infra section II.A. Marbury v. Madison—a case typically read today for its holding on judicial review—reflects this understanding. When Chief Justice Marshall describes the dispute as “a plain case for a mandamus,” he is relying on the fact that the statute authorizing Marbury’s appointment sets a five-year term and makes no mention of removal. 5 U.S. (1 Cranch) 137, 173 (1803); see also An Act Concerning the District of Columbia, ch. 15, § 11, 2 Stat. 107 (1801). Once the office had vested, there was no legal mechanism for the President to remove Marbury before his five years were up. See infra section II.A.

Term-of-years offices, like good-behavior offices, have been a feature of English and American law since at least the eighteenth century. 24 See infra section II.A. As late as 1978, terms of years were still widely understood as tenure protections. See infra note 377 and accompanying text. See generally Manners & Menand, supra note 22 (describing the use of such provisions throughout history). They protect officials from the uncertainty and vulnerability of an “at pleasure” appointment while still ensuring regular review of their work. Removal permissions, when added to such offices, serve as a safeguard. They limit, rather than protect, officeholder independence by authoriz­ing removal under certain discrete circumstances.

When Congress first used the now-talismanic INM phrase in 1887, it defined these circumstances using terms that were already well-known. “Neglect of duty” and “malfeasance in office” were old common law concepts employed by courts and legislators to connote an officer’s failure to faithfully execute statutory duties. Neglect of duty indicated instances of “nonfeasance”—a failure to perform one’s duties in a way that caused injury to others. At common law, neglect had been grounds for removing the officers of English towns and boroughs for hundreds of years. It also constituted a type of “misdemeanor”—or “bad behavior”—that could trigger the removal of clerks, judges, and other officers appointed for life to “good behavior” positions. “Malfeasance in office,” meanwhile, referred to a wrongful act committed in the execution of one’s duties that caused injury to others. Malfeasance was another type of misdemeanor that warranted removal from a good behavior office, and it could also lead to removal in the municipal context. 25 See infra sections II.B.1–.2. Inefficiency, by contrast, was of newer vintage: a term increasingly used over the course of the nineteenth century to describe wasteful government administra­tion caused by inept officers who gained their positions through political connections rather than merit. Inefficient officials lacked  the  skills  to  perform  their  duties,  rendering  them  incapable  of  doing  their jobs. 26 See infra section II.B.3.

Congress was not the first legislature to codify INM. All three terms appeared in state law, with neglect and malfeasance appearing in the laws of the colonies before that. As Professors Andrew Kent, Ethan Leib, and Jed Shugerman have shown, early American legislatures often required officials to take oaths to faithfully execute their duties. 27 See Andrew Kent, Ethan J. Leib & Jed Handelsman Shugerman, Faithful Execution and Article II, 132 Harv. L. Rev. 2111, 2159–62 (2019). These laws author­ized suit against officials who violated their oaths. Often, these oath violations were liquidated as “neglect of duty” or “malfeasance in office.” Over the course of the nineteenth century, state legislatures also used neglect of duty and malfeasance in office in removal provisions to define the behavior that might forfeit an office. Sometimes, they made these words removal grounds for officers otherwise granted tenure for a term of years, using the security of term-tenure to insulate proficient administra­tors from partisan political pressure while employing neglect of duty and malfeasance as a safety valve. This approach became increasingly common as legislators created offices to oversee ambitious infrastructural projects such as prisons, canals, banks, and railroads—offices for which term-of-years administrators who neglected their duty or engaged in malfeasance could cause immediate and significant harm. 28 See infra section II.C.

In 1843, Indiana became the first state to combine neglect and malfeasance with “inefficiency.” Confronting a massive public finance crisis caused by defaulting railroad and canal projects, the state included inefficiency as a ground for removing government officers who were incapable of performing their duties promptly and effectively. 29 See infra section II.B.3.a. Over the next thirty years, New York, Ohio, and several other states incorporated inefficiency into removal statutes as well. 30 See infra section II.B.3.

When Congress imported INM into federal law in 1887, it used the terms to establish the Interstate Commerce Commission (ICC), a federal railroad regulator. It empowered the new commissioners to serve for terms of six years, but it also authorized the President to remove them for INM. 31 See infra section II.C. As the country’s economy grew increasingly technical and complex, Congress drew repeatedly on this structure, creating “inde­pendent commissions” to regulate the activities of private companies, especially those providing public infrastructure. Legislators thought of these entities as “arm[s] of the Congress” operating in a quasi-legislative, quasi-judicial manner. 32 91 Cong. Rec. 11,965 (1945) (statement of Rep. Bland). They gave the President removal power, not so the President might direct the commissions, but so there would be a ready alternative to impeach­ment, especially when Congress was out of session. This was how judges and scholars understood removal statutes when the Court decided Humphrey’s Executor. And this was how legislators continued to understand these provisions when they drafted the Federal Reserve Act, created the Maritime Commission, designed the Civil Service Merit Systems Protection Board, and set up the FEC. 33 See infra notes 363–364 and accompanying text. See generally Manners & Menand, supra note 22 (analyzing the history behind the enactment of these provisions).

Two conclusions follow from the history. First, the law was not designed to permit the President to remove the heads of independent agencies for inefficiency or neglect of duty if they do not follow presiden­tial policy directives or if they depart from the President’s agenda. INM permits removal only in cases where officials act wrongfully in office, fail to perform their statutory duties, or perform them in such an inexpert or wasteful manner that they impair the public welfare. In reaching this conclusion, this Article looks beyond evidence regarding early under­standings of INM. It interrogates legislative intent, statutory design, and the relevant case law. Its results are largely consistent with 150 years of practice by Presidents, legislators, and agency officials. To accept this Article’s definitions of INM, one need not accept meanings from centu­ries ago, frozen in time. On the contrary, the evidence suggests that the under­standings recovered here were widely shared until relatively recently.

Second, there is no need to expand the concept of neglect of duty, or to rely on the concept of inefficiency at all, to square independent agencies with the Take Care Clause, as some scholars have argued. Neglect of duty and malfeasance in office, as traditionally interpreted, encompass what we call a failure of “faithful execution”: the official mis­behavior that the Take Care Clause purportedly obliges the President to prevent. In other words, even assuming that the Take Care Clause creates a role for the President in overseeing independent agency officials, most existing independent agency statutes already allow Presidents to perform this role by permitting them to remove those who engage in malfeasance or neglect.

The Court’s recent decision in Seila Law v. Consumer Financial Protection Bureau raises the salience of the analysis presented herein. In that case, Chief Justice Roberts, writing for a divided Court, held that the design of the CFPB—with a single director appointed to a five-year term, removable by the President only for INM—violates the Constitution’s “separation of powers.” 34 140 S. Ct. 2183, 2192 (2020). In reaching this conclusion, five Justices cast doubt on the idea that INM allows the President to remove officials on the basis of policy disagreements and stated that the Court had not been presented with “any workable standard derived from the statutory language.” 35 Id. at 2206. This Article supplies such a standard.

This Article also provides support for Justice Kagan’s suggestion in her dissent that there is an equivalence between neglect and malfeasance, on the one hand, and a failure of faithful execution, on the other. Explaining that INM provisions permit the President to remove for “incompetence” and a “failure to ‘faithfully execute[]’ the law,” 36 Id. at 2238 (Kagan, J., concurring in the judgment and dissenting in part) (quoting U.S. Const. art. II, § 3). Justice Kagan concluded that statutes limiting the President’s authority to remove domestic officers who execute the laws do not conflict with the Take Care Clause so long as the President can remove such officers for cause. 37 Id. at 2235 & n.9. The potential implications of the dissent’s interpretation of Article II—including the extent to which it would permit statutory restrictions on the President’s power to remove principal officers outside of the independent agency context—are beyond the scope of this Article. But the majority’s unwillingness to adopt a broad reading of INM, coupled with the dissent’s conclusion that INM permits removal for incompetence and a failure of faithful execution, underscores the significance of this Article’s analysis—analysis that offers legally grounded definitions of these terms and provides historical ballast to the hypothesis that neglect and malfeasance correspond to a failure to faith­fully execute the law. 38 An earlier version of this Article was publicly available under a different title (“Faithful Administration and the Limits of Agency Independence”) during the pendency of this case and was cited by that title in one of the amicus briefs submitted to the Court. See Brief of Harold H. Bruff et al. as Amici Curiae in Support of Court-Appointed Amicus Curiae, in Support of the Judgment Below at 7, 19, 20, Seila L., 140 S. Ct. 2183 (No. 19-7).

This Article proceeds in three Parts. Part I reviews recent scholarly and judicial treatments of for-cause removal statutes and identifies unsettled questions. Part II excavates the lost history of removal law and examines the origins and function of INM. Part III returns to the questions Part I raises and examines them in light of the evidence Part II uncovers.