Presidential Review: The President’s Statutory
Authority over Independent Agencies
CASS R. SUNSTEIN* & ADRIAN VERMEULE**
Many presidents have been interested in asserting authority over in-
dependent regulatory agencies such as the Federal Trade Commission,
the Federal Communications Commission, the Nuclear Regulatory
Commission, the Securities and Exchange Commission, and the Federal
Reserve Board. The underlying debates raise large constitutional questions,
above all about the meaning and justification of the idea of a “unitary exec-
utive.” In the first instance, however, the President’s authority over inde-
pendent agencies depends not on the Constitution but on a common
statutory phrase, which allows the President to discharge the heads of such
agencies for “inefficiency, neglect of duty, or malfeasance in office.” This
phrase—the INM standard—is best understood to create a relationship of
presidential review—and a particular remedy for legal delinquency flowing
from that review. It allows the President to discharge members of independ-
ent agencies not only for laziness and torpor (inefficiency) or for corruption
(malfeasance) but also for neglect of their legal duties, which includes egre-
giously erroneous decisions of policy, law, or fact, either repeatedly or on
unusually important matters. Connecting this understanding to the Take
Care Clause, we reject both a minimalist approach, which deprives the
President of any kind of decisionmaking authority over policy made by inde-
pendent agencies, and also a maximalist approach, which would treat the
independent agencies as essentially identical to executive agencies in terms
of presidential oversight authority. This approach has strong implications
for how to understand the President’s supervisory authority over independ-
ent agencies. It suggests that he has such authority insofar as he is attempt-
ing to ensure against “neglect of duty,” but not if he is displacing their
policymaking discretion.
T
ABLE OF CONTENTS
I. DOES THE PRESIDENT HAVE AUTHORITY OVER INDEPENDENT
REGULATORY AGENCIES? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638
II. INM,
WITH AN EMPHASIS ON THE N. . . . . . . . . . . . . . . . . . . . . . . . . . 644
* Robert Walmsley University Professor, Harvard Law School. © 2021, Cass R. Sunstein & Adrian
Vermeule.
** Ralph S. Tyler, Jr. Professor of Constitutional Law, Harvard Law School. We are grateful to
Daphna Renan and Peter Strauss for valuable comments on a previous draft. Thanks too to participants
in a valuable legal theory workshop at Harvard Law School. We are also grateful to Dustin Fire for
superb research assistance.
637
A. THE UNSETTLED MEANING OF THE INM STANDARD . . . . . . . . . . . . . . 644
B. WHAT THE INM STANDARD MEANS . . . . . . . . . . . . . . . . . . . . . . . . . . . 647
C. NEGLECT OF DUTY: THE PRESIDENT AS REVIEWER. . . . . . . . . . . . . . . 649
D. THE STANDARD APPLIED: REMOVAL . . . . . . . . . . . . . . . . . . . . . . . . . . 652
1. Easy Cases: Group One . . . . . . . . . . . . . . . . . . . . . . . . . 652
2. Easy Cases: Group Two . . . . . . . . . . . . . . . . . . . . . . . . . 653
3. Easy Cases: Group Three . . . . . . . . . . . . . . . . . . . . . . . . 654
4. Harder Cases: Group Four . . . . . . . . . . . . . . . . . . . . . . . 654
III. P
RESIDENTIAL SUPERVISION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
A. EXECUTIVE AGENCIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
B. INDEPENDENT AGENCIES. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 660
C. ON POLICING LINES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 662
C
ONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 664
I. D
OES THE PRESIDENT HAVE AUTHORITY OVER INDEPENDENT REGULATORY
AGENCIES?
Does the President have authority over the independent regulatory agencies?
1
May he fire their members?
2
May he control the decisions of the Federal Trade
Commission (FTC), the Nuclear Regulatory Commission (NRC), the Federal
Communications Commission (FCC), the National Labor Relations Board
(NLRB), the Federal Reserve Board, and the Consumer Financial Product
Bureau (CFPB)?
3
These are among the most fundamental questions in American
public law.
Such questions are usually approached as a matter of constitutional law.
4
The
key text is Article II, Section 1, which vests “executive [p]ower” in one person:
1. For present purposes, we understand “independent regulatory agencies” to refer to agencies whose
heads are not subject to at-will discharge by the President. For additional discussion, see generally
Adrian Vermeule, Conventions of Agency Independence, 113 C
OLUM. L. REV. 1163 (2013).
2. The defining decision, offering a qualified answer of “no,” is Humphrey’s Executor v. United
States, 295 U.S. 602, 629 (1935). In the same vein are Wiener v. United States, 357 U.S. 349, 356
(1958), and Morrison v. Olson, 487 U.S. 654, 691–92 (1988).
3. See Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2192 (2020) (answering the question of control with
a firm “yes” in the context of the CFPB, as a matter of constitutional law).
4. See id. See generally S
TEVEN G. CALABRESI & CHRISTOPHER S. YOO, THE UNITARY EXECUTIVE:
P
RESIDENTIAL POWER FROM WASHINGTON TO BUSH (2008) (providing constitutional and historical
analysis to the unitary executive theory); Saikrishna Prakash, New Light on the Decision of 1789, 91
C
ORNELL L. REV. 1021 (2006) (discussing an early legislative construction of the Constitution regarding
executive power).
638 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
“[A] President of the United States.”
5
It is broadly agreed that the presidency is
“unitary,” at least in the sense that the Constitution rejects a plural executive-by-
committee, and that the “Decision of 1789,” resulting in broad presidential
authority over executive agencies, is relevant to the discussion.
6
Notwithstanding
that agreement, the precise meaning of “unitariness” is disputed,
7
and reasonable
people disagree intensely about the appropriate understanding of the constitu-
tional text, the relevant history, and the relevance of institutional and other
changes over time, including the rise of the modern administrative state.
8
For many decades, the Supreme Court has agreed that Congress has the consti-
tutional authority to make some agencies “independent” of the President.
9
But
that conclusion has been under continuing and newly intense pressure,
10
and
under the Constitution, the meaning of “independence” remains highly uncer-
tain.
11
The Supreme Court’s decision in Seila Law LLC v. CFPB, which struck
down the provision granting independence to the CFPB, emphasized those
Sections of Article II that grant the executive power to the President (and no one
else), and that give to the President (and no one else) the power to “take Care that
the Laws be faithfully executed.”
12
Seila Law emphasized that under the
Constitution, the Executive Branch is unitary and the President is in charge of
it.
13
At the same time, Seila Law emphasized that the CFPB was headed by a sin-
gle
director (rather than a multimember commission) and suggested that, so long
as an independent agency is headed by a multimember commission and “do[es]
not wield substantial executive power,”
14
it is constitutionally acceptable. The
Court’s complex and ambiguous decision has magnified the long-standing uncer-
tainty because the Court showed a great deal of skepticism about the whole idea
of independent administration. Indeed, Seila Law might well be taken to cast cur-
rent independent agencies, which wield substantial executive power, into serious
constitutional doubt.
15
See Cass R. Sunstein & Adrian Vermeule, The
Unitary Executive: Past, Present, Future, 2021
S
UP. CT. REV. (forthcoming 2021) (manuscript at 2–3) (available at https://papers.ssrn.com/sol3/papers.
cfm?abstract_id=3666130 [https://perma.cc/6DLJ-USKZ]).
5. U.S. CONST. art. II, § 1; see Seila Law, 140 S. Ct. at 2197.
6. See Prakash, supra note 4, at 1022–23.
7. Compare Seila Law, 140 S. Ct. at 2197, with id. at 2227–28 (Kagan, J., dissenting).
8. For various views, see generally Rachel E. Barkow, Insulating Agencies: Avoiding Capture
Through Institutional Design, 89 T
EX. L. REV. 15 (2010); Lisa Schultz Bressman & Robert B.
Thompson, The Future of Agency Independence, 63 V
AND. L. REV. 599 (2010); Kirti Datla & Richard L.
Revesz, Deconstructing Independent Agencies (and Executive Agencies), 98 C
ORNELL L. REV. 769
(2013); Aziz Z. Huq, Removal as a Political Question, 65 S
TAN. L. REV. 1 (2013); Geoffrey P. Miller,
Independent Agencies, 1986 S
UP. CT. REV. 41; Paul R. Verkuil, The Status of Independent Agencies
After Bowsher v. Synar, 1986 D
UKE L.J. 779; and Vermeule, supra note 1.
9. For example, Free Enterprise Fund v. Public Co. Accounting Oversight Board, 561 U.S. 477, 483
(2010), takes the independent agency form as acceptable, though it would not be difficult to read the
opinion as uncomfortable with it.
10. See Seila Law, 140 S. Ct. at 2192.
11. See id.
12. Id. at 2191 (quoting U.S. C
ONST. art. II, § 1, cl. 1; id. art. II, § 3).
13. See id.
14. Id. at 2199–200.
15.
2021] PRESIDENTIAL REVIEW 639
In the midst of that uncertainty, it is important to recognize that the President’s
authority over the independent regulatory agencies turns on relevant statutes, not
directly on the Constitution at all—although the constitutional background will
influence judicial interpretation of those statutes.
16
In the usual formulation,
Congress allows the President to discharge heads of independent agencies for
“inefficiency, neglect of duty, or malfeasance in office.”
17
The INM standard,
18
as we call it, can be taken both to limit and to grant presidential authority. More
than a century after the INM standard was first used, its meaning has yet to be
settled.
19
It is easy to identify two polar views of presidential authority over independent
regulatory agencies. The first insists on policy independence.
20
For a
valuable discussion with a historical focus, see generally Jane Manners & Lev Menand,
Presidential Removal: Defining Inefficiency, Neglect of Duty, and Malfeasance in Office, 121 C
OLUM. L.
R
EV. (forthcoming 2021) (available at https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3520377
[https://perma.cc/4CSJ-8EYT]).
Some language
from Humphrey’s Executor inclines in this direction; for example, the Court sug-
gested that Congress wanted the FTC to “be independent of executive authority,
except in its selection, and free to exercise its judgment without the leave or hin-
drance of [the President].”
21
According to that view, the President may discharge
heads of independent agencies only for serious improprieties that do not involve
policy in any way. With respect to presidential authority, we can characterize this
view as minimalist. If, for example, a commissioner takes a bribe, he has engaged
in “malfeasance,” and he may be discharged for that reason. So too, a decision to
take a six-month vacation in Hawaii would constitute “neglect of duty.”
Crucially, even on the minimalist view, this amounts to de facto abdication of the
16. See John F. Manning, The Independent Counsel Statute: Reading “Good Cause” in Light of
Article II, 83 M
INN. L. REV. 1285, 1288 (1999) (arguing that, to avoid a “serious constitutional
question,” the “good cause” removal provision in the Ethics in Government Act should be
interpreted to allow removal for insubordination). Examples of relevant statutes with language
restricting the President’s removal power include 5 U.S.C. § 1211(b) (2018) (Office of Special
Counsel); 12 U.S.C. § 5491(c)(3) (2018) (CFPB); 15 U.S.C. § 41 (2018) (FTC); and 49 U.S.C.
§ 1111(c) (2018) (NTSB).
17. A helpful overview is Marshall J. Breger & Gary J. Edles, Established by Practice: The
Theory and Operation of Independent Federal Agencies, 52 A
DMIN. L. REV. 1111, 1135 (2000)
(describing the “inefficiency, neglect of duty, or malfeasance in office” (INM) standard as the basic
removal provision). See, e.g., Act of Oct. 6, 2006, Pub. L. No. 109-304, § 301(b)(3), 120 Stat. 1485,
1488 (applying the INM standard to the removal of commissioners of the Federal Maritime
Commission); ICC Termination Act of 1995, Pub. L. No. 104-88, § 701(a)(3), 109 Stat. 803, 933
(Surface Transportation Board); Federal Mine Safety and Health Act of 1977, Pub. L. No. 95-164, §
113, 91 Stat. 1290, 1313 (Federal Mine Safety and Health Commission); Federal Aviation Act of
1958, Pub. L. No. 85-726, § 201(a)(2), 72 Stat. 731, 741 (Civil Aeronautics Board); Bituminous
Coal Act of 1937, ch. 127, § 2(a), 50 Stat. 72, 73 (National Bituminous Coal Commission); Act of
Sept. 26, 1914, ch. 311, § 1, 38 Stat. 717, 718 (FTC).
18. Judge Griffith used the same abbreviation in PHH Corp. v. CFPB, 881 F.3d 75, 127 (D.C. Cir.
2018) (en banc) (Griffith, J., concurring).
19. The Court expressed such a view in Seila Law, 140 S. Ct. at 2206; we take up that view below.
See infra Section II.A.
20.
21. Humphrey’s Ex’r v. United States, 295 U.S. 602, 625–26 (1935) (emphasis omitted).
640 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
office and thus a neglect of duty. But the President may not discharge commis-
sioners because of any kind of policy disagreement.
Suppose, for example, that a member of the FTC makes a series of merger
decisions that
the President believes to be arbitrary, or adopts a lenient approach
to mergers that would, in the President’s view, lead to the creation of monopolies
throughout the economy. Or suppose that the Consumer Product Safety
Commission adopts a pro-market approach that, in the President’s view, is
incon-
sistent
with the governing statute and will leave consumers (and their children)
vulnerable to serious risks. If commissioners have policy independence, no
President has, in such circumstances, a statutory basis for discharge. Of course, it
remains possible that the President could contend that, so construed, the INM
standard is unconstitutional under Article II.
22
But the statutory question would
be at an end.
The idea of policy independence is important at all stages of a presidential
term, because
any President is likely to think, on occasion, that members of
inde-
pendent
agencies, or independent agencies as a whole, are making at least some
decisions that he abhors. But Presidents are especially likely to be hampered by
independence in their initial years, when independent agencies contain large
numbers of commissioners who have been appointed by their predecessors.
The second
polar view is that the President has broad policymaking control.
23
According to that view, the President is allowed to discharge commissioners if
their conclusions on significant issues seem to him to be inconsistent with what
sound understandings of policy and law demand. With respect to presidential
authority, we can characterize this view as maximalist. A President might
conclude that FTC commissioners have neglected their duty if they interpret
their organic statute incorrectly (by his lights) or make bad choices about public
policy—for example, by giving consumers too little protection against fraudulent
advertising. A President might condemn an insufficiently aggressive regulation
from the Nuclear Regulatory Commission, subjecting people to excessive risks as
a demonstration of “inefficiency.”
24
This view would ensure a thin and perhaps even nonexistent distinction
between the President’s authority over independent agencies and his
correspond-
ing
authority over executive agencies. But to those who embrace this view, it is
the best reading of the INM standard. Those who support this view believe that if
Congress genuinely wants to ensure independence then it must choose different
and clearer language—not by INM, but instead by allowing removal only for
22. The Executive Branch made this argument in Humphrey’s Executor, but the Court rejected it,
finding “it plain under the Constitution that illimitable power of removal is not possessed by the
President.” Id. at 626–29.
23. See, e.g., PHH Corp., 881 F.3d at 124 (Griffith, J., concurring).
24. See id. at 131–32. Judge Griffith focused in particular on dictionary definitions of inefficiency,
urging that “it is the broadest of the three INM removal grounds and best illustrates the minimal extent
to which the INM standard restricts the President’s ability to supervise the Executive Branch.” Id. In our
view, this is a contentious understanding of inefficiency, and the broadest of the three grounds, most
tightly connected with the President’s constitutional authority, is neglect of duty.
2021] PRESIDENTIAL REVIEW 641
“gross improprieties.” The maximalist position would seem especially attractive
if it is deemed necessary to avoid serious constitutional questions about the power
of Congress to infringe on presidential authority under the Take Care Clause and
the Vesting Clause of Article II; it might be an appealing use of the avoidance
canon.
25
As we shall see, the Supreme Court has firmly rejected maximalism.
26
We aim
to defend a third view here—a middle way which rests on the idea that INM
clauses create a relationship of presidential review vis-a
`
-vis the independent
agencies. More specifically, and with a particular emphasis on the N (and hence
on the crucial idea of neglect of duty), the President is entitled to discharge
com-
missioners
for acting in a way that he reasonably believes to be not merely wrong
but a clear violation of their statutory responsibility. Arbitrary action counts as
neglect of duty, though for reasons to be explained, it is doubtful that a single ar-
bitrary
action, of the sort that courts strike down under the Administrative
Procedure Act (APA), counts as a legitimate basis for discharge.
27
Rather, a pat-
tern
of arbitrariness would almost always be necessary. Although we connect the
INM standard to the language of the Take Care Clause and the Vesting Clause—
and hope that originalists might be willing to accept our conclusions—ours is not
an originalist argument focusing on the original public meaning of constitutional
clauses. Rather, it is a textual and structural one,
28
premised on the best reading
of statutory text and taken in light of institutional roles in the administrative state
and background constitutional principles.
Our approach firmly rejects minimalism; it does permit discharge for reasons
that involve policy,
at least in an aggregate sense over an array of decisions.
Under our approach, emphasizing neglect of duty, the President may discharge a
commissioner for making a series of decisions that he reasonably believes to be
arbitrary (and so in violation of legal duty). But like the Supreme Court in Seila
Law,
29
we also reject maximalism because our approach forbids the President
from discharging commissioners merely because he believes that their decisions
are wrong. The difference is significant. Under our view, the President must show
that a commissioner has acted in ways that clearly represent a neglect of
duty. As compared with maximalism, that is an important limitation on his
25. The avoidance canon argues in favor of construing statutes so as to avoid serious constitutional
questions. See generally John F. Manning, The Nondelegation Doctrine as a Canon of Avoidance, 2000
S
UP. CT. REV. 223. The maximalist position would do exactly that.
26. See Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2192 (2020).
27. We deal with the question of single decisions below. See infra Section II.B.
28. Cf. C
HARLES L. BLACK, JR., STRUCTURE AND RELATIONSHIP IN CONSTITUTIONAL LAW (1969)
(exploring an approach to constitutional interpretation based not on the original understanding but on
structural considerations). Jane Manners and Lev Menand reach a similar conclusion, with an emphasis
on the historical origins of the INM standard and in particular with reference to neglect of duty and
malfeasance, which they identify with “unfaithful execution.” See Manners & Menand, supra note 20
(manuscript at 8–9).
29. See 140 S. Ct. at 2206 (declining to revisit the Humphrey’s Executor Court’s implicit rejection of
an interpretation of the INM standard that would leave the President free to remove an officer based on
disagreements about agency policy).
642 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
authority. We are aware that, briefly stated, the position we aim to defend raises
many questions—on one side because of the risk of presidential overreach, and on
the other, because of the risk of unduly constraining the constitutionally grounded
authority of the President. We will explore these concerns in due course.
The INM language governs removal, but it also has strong implications for the
question of presidential supervision.
30
As a matter of practice, presidents have
avoided interfering with the substantive policy choices of independent agencies
partly because of concerns about their legal authority.
31
Rejection of minimalism
suggests that those concerns are easily overstated. Under our approach, empha-
sizing neglect of duty, the President has considerable authority to cabin the judg-
ments of independent agencies, certainly through establishing general policies
and processes, and on appropriate occasions by influencing their decisions with
respect to particular issues. If the President can control independent agencies for
neglecting their legal duties, it follows that he can supervise independent agen-
cies, not to overrule their decisions simply because he disagrees with them, but to
ensure that they come within what he reasonably believes to be the permissible
bounds of law, fact, or discretion.
It is true that aggressive presidents might try to collapse the line that separates
the
two—a problem that could put considerable pressure on both government
lawyers and federal courts. But that is no unique objection to our reading of INM
clauses; it is a general problem that can arise wherever the line between
presiden-
tial
authority and that of independent agencies is set. The only rule that would
eliminate it is one that entirely subordinates all agencies wielding executive
power to presidential control in all their functions—a rule that the Supreme Court
has never once entertained, even in Myers v. United States,
32
the apogee of the
unitary executive. As we will see, Myers recognized several exceptions to both
presidential removal and direction, and the same is true of the Court’s ruling in
Seila Law, which can be seen as Myers 2.0.
The remainder of this Article is organized as follows. Part II explores the
meaning of the INM
standard, taken in its context. Part III applies our proposed
standard to the problem of removal with reference to both easy and hard cases.
Part IV investigates the question of supervisory authority, asking whether and
when the President can oversee policy judgments by independent agencies. Part
V briefly concludes.
30. See Peter L. Strauss, Overseer, or “The Decider”? The President in Administrative Law, 75 GEO.
W
ASH. L. REV. 696, 716–17 (2007).
31. The best evidence can be found in past and present executive orders governing White House
review of regulations, which do not apply to independent agencies. See, e.g., Exec. Order No. 12,291, 3
C.F.R. § 127 (1982), reprinted in 5 U.S.C. § 601, at 431–34 (1982) (revoked 1994); Exec. Order No.
13,563, 3 C.F.R. § 215 (2012), reprinted in 5 U.S.C. § 601, at 816–17 (2012). In Executive Order
13,579, President Barack Obama suggested that independent agencies “should” follow Executive Order
13,563, but he pointedly avoided the word “shall.” See 3 C.F.R. § 256 (2012), reprinted in 5 U.S.C. §
601, at 817–18 (2012).
32. 272 U.S. 52 (1926).
2021] PRESIDENTIAL REVIEW 643
II. INM, WITH AN EMPHASIS ON THE N
In federal law, the INM standard seems to have first appeared in 1887, with
the creation
of the Interstate Commerce Commission.
33
As noted, the stand-
ard has become common. It is often adopted by Congress with little or no dis-
cussion of its meaning; it is a kind of “off the rack” provision—a legal
convention or phrase of art—meant to signal independence of some kind.
34
Although it has a revealing and long history,
35
to which we will make refer-
ence,
we will focus mostly on the natural meaning of the text, read in light of
the constitutional background
36
and structural goals—emphasizing that the
legislative history is sparse and inconclusive even if deemed relevant.
37
It is
remarkable but true that after well over a century of practice, the meaning of
INM remains unsettled.
38
And in spite of its central importance, the literature
on the topic remains puzzlingly undeveloped.
39
A. THE UNSETTLED MEANING OF THE INM STANDARD
In the Supreme Court, the meaning of the INM standard remains unsettled, and
the Justices have given some sharply conflicting signals. Bowsher v. Synar—an
apparently defining decision—raised an assortment of questions about the INM
standard.
40
There, the Court dealt with a statute that authorized the Comptroller
General to be removed by joint resolution of Congress on any one of five
grounds: (1) permanent disability, (2) inefficiency, (3) neglect of duty, (4)
mal-
feasance,
or (5) a felony or conduct involving moral turpitude.
41
For the Court, a
key question was whether the Comptroller General was thereby made an agent of
Congress, subject to its will, or essentially independent. The Court chose the
for-
mer
view. In doing so, it pointed to a statement by a member of Congress, declar-
ing
that: “Congress at any moment when it found [the Comptroller General] was
inefficient and was not carrying on the duties of his office as he should and as the
Congress expected, could remove him
. . . .”
42
That statement seems to suggest
33. See Act of Feb. 4, 1887, ch. 104, § 11, 24 Stat. 379, 383. The best discussion of the historical
context is Manners & Menand, supra note 20 (manuscript at 68–81).
34. See Breger & Edles, supra note 17, at 1144–45.
35. See Manners & Menand, supra note 20 (manuscript at 41–81).
36. We do not outline that background here. For discussion, see Seila Law LLC v. CFPB, 140 S. Ct.
2183, 2232–36 (2020) (Kagan, J., dissenting); Steven G. Calabresi & Saikrishna B. Prakash, The
President’s Power to Execute the Laws, 104 Y
ALE L.J. 541, 642–45 (1994); Elena Kagan, Presidential
Administration, 114 H
ARV. L. REV. 2245, 2319–31 (2001); and Lawrence Lessig & Cass R. Sunstein,
The President and the Administration, 94 C
OLUM. L. REV. 1, 106–12 (1994).
37. See Manners & Menand, supra note 20 (manuscript at 74–81). On some of the snippets of
legislative history, see PHH Corp. v. CFPB, 881 F.3d 75, 132–33, 136–37 (D.C. Cir. 2018) (en banc)
(Griffith, J., concurring).
38. See generally PHH Corp., 881 F.3d 75 (including multiple concurring and dissenting opinions
that reach different conclusions on the meaning of INM).
39. A valuable discussion appears in Manners & Menand, supra note 20.
40. 478 U.S. 714 (1986).
41. Id. at 727–28; see 31 U.S.C. § 703(e)(1)(B) (2018).
42. Bowsher, 478 U.S. at 728 (quoting 61 C
ONG. REC. 1081 (1982)).
644 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
that under the INM standard, the principal (whether Congress or the President)
has a great deal of authority over the agent.
43
In a crucial passage, the Court added: “The statute permits removal for ‘ineffi-
ciency,’
‘neglect of duty,’ or ‘malfeasance.’ These terms are very broad and, as
interpreted by Congress, could sustain removal of a Comptroller General for any
number of actual or perceived transgressions of the legislative will.”
44
That con-
clusion, apparently essential to the Court’s holding, would seem to apply to pre-
cisely the same terms (INM) used in statutes specifying the President’s power to
remove independent agency heads. And if so, the statutory and constitutional
questions are settled simultaneously. If the INM standard is “very broad” and
could allow removal of an agency head “for any number of actual or perceived
transgressions of the” presidential will,
45
we are squarely in the domain of max-
imalism. If Congress has not even tried to exercise its authority to create a genu-
inely independent agency—if the INM standard leaves the President with broad
removal power—there is no constitutional problem.
In the decades since Bowsher was decided, no
consensus has emerged on
whether the Court’s ruling applies to the INM standard. In fact, the Court’s more
recent cases plainly suggest that it does not—the standard creates at least some
degree of policy independence. Free Enterprise Fund v. Public Co. Accounting
Oversight Board, which struck down a statute offering two layers of insulation
through the INM standard,
46
is unintelligible without that assumption. If the INM
standard does not create policy independence—notwithstanding Bowsher—then
two layers (or three or seven) should not matter because the layers would be thin
indeed. Free Enterprise Fund seems to reject the maximalism of Bowsher or
any-
thing
close to it. And in speaking of the INM standard, the Court did not even
mention Bowsher.
That is a genuine puzzle. Why wasn’t Bowsher relevant? The best explanation
points to context. Even if the statutory text is identical, Congress’s use of that
standard to specify and limit its own authority over the Comptroller General
might be understood differently from Congress’s effort to specify and limit the
President’s authority over a member of an independent regulatory commission.
In the latter context, after all, it is agreed that Congress is seeking to create some
real space between the President’s policymaking authority and that of the
com-
mission member.
47
Recall these four words from Bowsher: “as interpreted by
Congress.”
48
It is not clear to what those words are meant to refer, but nothing of
43. See Verkuil, supra note 8, at 797 n.100 (noting that the Humphrey’s Executor standard “could be
construed so as to encompass a general charge of maladministration, in which event even if the terms of
removal are deemed to be exclusive they could still be satisfied by a removal by the President on the
ground of policy incompatibility”).
44. Bowsher, 478 U.S. at 729.
45. Id.
46. 561 U.S. 477, 514 (2010).
47. See id. at 495.
48. Bowsher, 478 U.S. at 729.
2021] PRESIDENTIAL REVIEW 645
the kind can be said about the INM standard as applied to presidential removal
authority.
The rejection of maximalism became unambiguous in Seila Law,
where the
Court struck down the INM standard insofar as it limited the President’s authority
to remove the CFPB Director,
49
principally on the ground that an independent
agency is constitutionally unacceptable unless it consists of a multimember
com-
mission.
50
Importantly, the Court might have avoided the constitutional question
by ruling that INM allows the President the kind of policymaking control that
Article II requires. That conclusion would, of course, have had large implications
for the President’s relationship to a host of independent agencies, and the Court
pointedly refused to offer it. The Court’s analysis was surprisingly brisk:
Humphrey’s Executor implicitly rejected an interpretation that would leave
the President free to remove an officer based on disagreements about agency
policy. In addition, while both amicus and the House of Representatives invite
us to adopt whatever construction would cure the constitutional problem, they
have not advanced any workable standard derived from the statutory language.
Amicus suggests that the proper standard might permit removals based on gen-
eral policy disagreements, but not specific ones; the House suggests that the
permissible bases for removal might vary depending on the context and the
Presidential power involved. They do not attempt to root either of those stand-
ards in the statutory text. Further, although nearly identical language governs
the removal of some two-dozen multimember independent agencies, amicus
suggests that the standard should vary from agency to agency, morphing as
necessary to avoid constitutional doubt. We decline to embrace such an uncer-
tain and elastic approach to the text.
51
The Court added, “Without a proffered interpretation that is rooted in the statu-
tory text and structure, and would avoid the constitutional violation we have iden-
tified, we take Congress at its word that it meant to impose a meaningful
restriction on the President’s removal authority.”
52
In that way, the Court rejected
an approach of constitutional avoidance and (importantly) firmly rejected
maximal-
ism.
At the same time, these words need not mean that the President entirely lacks
policymaking control over independent agencies. But they certainly mean that the
statutory provision that grants independence is a “meaningful restriction”—which
indicates, in the Court’s view, that the constitutional issue could not be avoided.
To get slightly ahead of the story, we will suggest an approach that is firmly
rooted in the statutory text and structure, that is indeed a workable standard, that
lies between minimalism and maximalism, and that should serve to avoid
49. Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2191–92 (2020).
50. See id. at 2211. The constitutional issue is beyond the scope of the present discussion. For
analysis, see generally Sunstein & Vermeule, supra note 15.
51. Sunstein & Vermeule, supra note 15 (manuscript at 19) (quoting Seila Law LLC v. CFPB, No.
19-7, slip op. at 28 (U.S. June 29, 2020)).
52. Id. (manuscript at 20) (quoting Seila Law, slip op. at 29).
646 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
potential constitutional doubts about the independence of multimember commis-
sions. In fact, that is our principal goal.
B. WHAT THE INM STANDARD MEANS
Interpretation of the INM standard would be easier if it were accompanied by
an assortment of legal materials, giving a rich sense of context. And indeed, the
standard does have a long history, dating back to English common law sources
involving both municipal corporations and officeholding.
53
In the United States,
the phrase “neglect . . . of duty” can be found as early as 1796, and it seemed to
create authority, on the part of higher level officials, to remove people who would
otherwise be entitled to serve specified terms.
54
In 1842, Indiana specifically
allowed a court to remove a court’s sheriff “for inefficiency, neglect of duty, or
maleconduct in office.”
55
Nonetheless, we think that it would be a considerable
overstatement to say that the INM standard was a term of art prior to its initial use
in the Interstate Commerce Act enacted in 1887.
56
The history of that Act does offer useful hints.
57
Corruption was a central con-
cern,
but so were incompetence and failure to execute the duties of the office.
58
Maximalism seems to have been ruled out of bounds.
59
It is plain that the text
must be understood in the context of a congressional effort simultaneously to
authorize and to limit presidential authority over certain officials and also in the
context of the constitutional structure and background principles, properly under-
stood. As we shall explain, the ordinary meaning, taken in context, strongly
favors the approach for which we shall be arguing, and the avoidance canon sup-
ports it as well.
With respect to the meaning of INM, it seems self-evident that the minimalist
position captures a significant part of the picture—as a set of sufficient conditions
for removal, regardless of whether they are also necessary conditions. If an FTC
commissioner accepts a bribe, the commissioner has engaged in malfeasance; so
too if the commissioner steals from the government or misuses public funds. If
FCC commissioners fail to do their job, in some literal sense (for example, by
refusing to show up, spending time on crossword puzzles, or perhaps repeatedly
failing to meet legal deadlines), they have neglected their duty.
60
Finally, if a
53. See Manners & Menand, supra note 20 (manuscript at 41–42).
54. Id. (manuscript at 59) (internal quotation marks omitted).
55. Id. (manuscript at 62–63) (internal quotation marks omitted).
56. See id. (manuscript at 69). We have learned a great deal from Manners and Menand, supra note
20, and would add only that the long and instructive history of the Interstate Commerce Act, on which
they elaborate, does not offer much specificity about the meaning of the INM standard.
57. See id. (manuscript at 69–81) (discussing the historical context and enactment of the Interstate
Commerce Act).
58. Id. (manuscript at 65–66).
59. See id. (manuscript at 71–72).
60. Here the difference between the minimalist reading of neglect of duty and our reading is that the
former allows the President to discharge when (but only when) an officer has de facto abdicated an
office. Our reading, by contrast, will allow presidential review for a more broadly understood neglect of
duty, as discussed below.
2021] PRESIDENTIAL REVIEW 647
commissioner on the NRC works slowly and takes months to do what would take
most people days, that commissioner has been inefficient. None of this should be
controversial.
It also seems apparent that the ordinary meaning of the text rules the
maximal-
ist
position out of bounds. Suppose that a President disagrees with a decision of
the FCC—say, for example, about how to understand the statutory ban on
“obscene, indecent, or profane language.”
61
The President would understand it
more broadly than the FCC would, though (let us stipulate) both views are
reason-
able as
a matter of policy. Or suppose that the President disagrees with some fac-
tual
conclusions of the NRC with respect to safety issues; the President assesses
the risks differently from the agency. The mere disagreement cannot be regarded
as sufficient for removal. On the basis of its plain text, the INM standard does not
allow removal merely because the President has a different view on a question of
law, policy, or fact.
On either originalist or non-originalist grounds, it would be possible to
con-
clude
that the INM standard is unconstitutional if it is understood as a restriction
on the President’s authority to control policymaking by those who implement the
law.
62
We have said that we do not mean to resolve the constitutional questions
here. And it would not be impossible to use the avoidance canon to move in the
direction of maximalism. The only point is that it would be an unusually
aggres-
sive
use of that canon, and note that in Seila Law, the Supreme Court refused to
engage in that enterprise.
63
Strictly as a matter of statutory interpretation, if the
INM standard means anything, it means that the President cannot discharge a
member of an independent agency simply because he disagrees with the agency’s
conclusions about policy or fact. The avoidance canon applies only in the
pres-
ence of
ambiguity, suitably defined; it is not a license to obliterate the core appli-
cations of a statutory phrase.
64
We could imagine understandings of the term “inefficient” that would tend to-
ward
maximalism. One dictionary defines “inefficient” as “not producing the
effect intended or desired.”
65
Inefficient, M
ERRIAM-WEBSTER, https://www.merriam-webster.com/dictionary/inefficient [https://
perma.cc/WY9K-ZUGR] (last visited Dec. 25, 2020).
Pressing on that definition and motivated by the
avoidance canon, a textualist might argue, in good faith, that the President could
remove commissioners on the ground that their decisions will not have the
intended or desired effect. Intended or desired by whom? A fair question.
However it is answered, Seila Law plainly rejected such a broad understanding of
the “I” in the INM standard. The Court did not carefully parse the text, but it did
61. 18 U.S.C. § 1464 (2018); see FCC v. Fox Television Stations, 556 U.S. 502, 505–10 (2009)
(addressing the FCC’s authority to interpret the statutory ban).
62. See Manners & Menand, supra note 20 (manuscript at 3–4, 82–83).
63. See Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2206–07 (2020); see also Rust v. Sullivan, 500
U.S. 173, 190–91 (1991) (counseling against use of the avoidance canon to distort statutes rather than to
choose one of two fair readings).
64. See Rust, 500 U.S. at 190 (noting the threshold requirement of “two possible interpretations of a
statute” (quoting Blodgett v. Holden, 275 U.S. 142, 148 (1927)).
65.
648 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
say this: “Humphrey’s Executor implicitly rejected an interpretation that would
leave the President free to remove an officer based on disagreements about
agency policy.”
66
Whatever inefficiency means, it does not allow the President to
remove an officer simply because he thinks that the officer has made the wrong
policy choice.
C. NEGLECT OF DUTY: THE PRESIDENT AS REVIEWER
Does the President have any room to remove commissioners when he disap-
proves
of their decisions? In our view, the INM standard is a unitary phrase that
is best understood as a kind of legal drafting convention, usually adopted as a
whole.
67
For concreteness and simplicity, however, and because of their clear
relationship to Article II, let us pause over three words: neglect of duty.
68
In our
66. Seila Law, 140 S. Ct. at 2206.
67. It should be noted, however, that Congress has enacted at least one other statute that includes
only two of the three INM removal grounds: soon after the Court decided Humphrey’s Executor,
Congress added a removal provision to the National Labor Relations Act but turned the INM standard
into the NM standard by eliminating inefficiency. See Act of July 15, 1935, Pub. L. No. 74-198, ch. 372,
§ 3(a), 49 Stat. 449, 451 (codified at 29 U.S.C. § 153(a) (2018)) (“Any member of the Board may be
removed by the President . . . for neglect of duty or malfeasance in office, but for no other cause.”).
68. In an impressive and lengthy opinion, plainly rejected in Seila Law, 140 S. Ct. at 2197, Judge
Griffith took a different approach. See PHH Corp. v. CFPB, 881 F.3d 75, 131–32 (D.C. Cir. 2018) (en
banc) (Griffith, J., concurring). He “concentrate[d] on ‘inefficiency’ because it is the broadest of the
three INM removal grounds and best illustrates the minimal extent to which the INM standard restricts
the President’s ability to supervise the Executive Branch.” Id. Relying on dictionaries from the relevant
period, Judge Griffith wrote:
Dictionaries consistently defined the word “inefficiency” to mean ineffective or failing to
produce some desired result. For example, one prominent turn-of-the-century dictionary
defined “efficient” as “[a]cting or able to act with due effect; adequate in performance; bring-
ing to bear the requisite knowledge, skill, and industry; capable; competent.” . . . These dic-
tionaries indicate that an individual acts inefficiently when he fails to produce some desired
effect or is otherwise ineffective in performing or accomplishing some task.
Id. at 132. Judge Griffith urged that the same view is reflected in relevant legislative history. For
example, he pointed to a remark of a senator who, during the debate over whether to include the INM
standard for officials of the Civil Aeronautics Authority, feared that inefficiency did not provide
sufficient independence for agency officials and even lamented:
If we provide that the President may remove a man for inefficiency, to my mind we give him
unlimited power of removal. Under such authority he could have removed Mr. Humphrey,
had he assigned that as a reason. . . . I do not see anything to be gained by discussing the legal
question if we are to leave the word ‘inefficiency’ in the provision.
Id. at 133 (alteration in original) (quoting 83 C
ONG. REC. 6865 (1938) (statement of Sen. William
Borah)). But the dictionary definition could easily be understood more narrowly, see Kent H. Barnett,
Avoiding Independent Agency Armageddon, 87 N
OTRE DAME L. REV. 1349, 1386 & n.191 (2012) (citing
N
EW OXFORD AMERICAN DICTIONARY 867 (2001)), and the isolated comment by one senator in one
debate does not tell us much.
Judge Griffith did muster additional grounds for his broad understanding of inefficiency. See, e.g.,
Budget Hearing—Consumer Financial Protection Bureau: Hearing Before the Subcomm. on Oversight
& Investigations of the H. Comm. on Fin. Servs., 112th Cong. 8 (2012) (statement of Rep. Barney Frank,
Ranking Member, H. Comm. on Fin. Servs.) (“[T]his notion that the Director cannot be removed is
fanciful. . . . No one doubts that if a change in Administration comes, and the new President disagrees
with the existing Director, he or she can be removed. And proving that you were not inefficient, the
burden of proof being on you, would be overwhelming.” (emphases added)). But in our view, these and
2021] PRESIDENTIAL REVIEW 649
view, those three words deserve especially careful attention; they provide impor-
tant clues about how best to read the INM standard.
If commissioners act in patent defiance of law, they are neglecting their
(statutory) duty,
at least in the literal sense. The same is true if a
commis-
sioner
acts arbitrarily, at least if some governing law, such as the APA,
makes it unlawful to do that.
69
And if commissioners find a fact without sub-
stantial evidence, or arbitrarily, they are once more neglecting a (literal)
duty. If the President is allowed to discharge a commissioner for neglect of
duty, it would seem clear that the President has the authority to ensure obe-
dience to law.
From the constitutional point of view, this neglect-of-duty approach has an im-
portant
advantage: it links the President’s statutory authority to the Take Care
Clause.
70
It recognizes the President’s authority to oversee the independent agen-
cies insofar as he is attempting to ensure faithful execution of the laws.
71
By rec-
ognizing
that authority, it softens the constitutional concern and perhaps even
eliminates it, at least in one view of constitutional requirements. In Free
Enterprise Fund, the Court placed a great deal of emphasis on the Take Care
Clause, suggesting that the problem with the two layers of insulation was that the
President was deprived of his authority under that Clause.
72
If the President can
discharge commissioners for neglecting their (legal) duties, it would seem that he
has, under the Take Care Clause, exactly what he needs. He does not have
policy-
making
control; nothing we have said is inconsistent with Seila Law. But the
INM standard squarely refers to neglect of duty, and hence drawing attention to
the natural understanding of those words is hardly in tension with the statutory
text.
But with
its emphasis on neglect of duty, what precisely does this approach
entail? We do
not mean to support the distinction squarely rejected in Seila Law,
between general and particular policy disagreements.
73
We do not mean to point
to policy disagreements at all. The question is legal duty. In our view, the INM
standard should be read so as to put the President in a position somewhat
analo-
gous
to that of a reviewing court. The President can find a neglect of legal duty
other materials are quite thin. As a matter of ordinary understanding of the English language, at every
relevant time, the term neglect of duty is much meatier, and it becomes more relevant still in light of its
evident connection with the Take Care Clause, emphasized in Free Enterprise Fund v. Public Co.
Accounting Oversight Board, 561 U.S. 477, 484 (2010).
69. See 5 U.S.C. § 706(2)(A) (2018).
70. For a similar effort making that link, see Manners & Menand, supra note 20 (manuscript at 17–
18).
71. A valuable, relevant treatment appears in Barnett, supra note 68, at 1364–67. Barnett might be
read, among other things, to support the view that, although the President is obliged to take care that the
laws be faithfully executed, one of the laws that he must faithfully execute might limit his authority over
administrators. See id. The problem is that such laws might be independently unconstitutional. If the
President is constitutionally authorized to ensure faithful execution of the laws, a statute that limits his
ability to do so offends the Constitution, regardless of what Congress says.
72. See Free Enter. Fund, 561 U.S. at 496.
73. See supra note 29 and accompanying text.
650 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
under standards analogous to those outlined in Auer,
74
Chevron,
75
State Farm,
76
or Universal Camera
77
and respond accordingly. Under this approach, the
President might ask not whether the agency has exceeded its authority, but
whether it has clearly done so (Chevron); not whether the agency has chosen the
right policy, but whether it has considered legally relevant factors and addressed
reasonable alternatives (State Farm); or not whether the agency has found the
facts correctly, but whether it has adequate evidence for its conclusions
(Universal Camera).
78
Under this approach, we also need to know what form a presidential response
would take. Are we speaking of removal or supervision? The two must be dis-
cussed separately. If we are speaking of removal, it would be a grant of consider-
able power to the President to say that in the face of what he considers to be a
single Chevron violation, the President would be entitled to discharge a member
of an independent regulatory commission. That would mean that a commissioner
of, say, the FCC could be removed for voting in favor of even a single unlawful
decision. That would be an implausibly aggressive understanding of what it
means to neglect one’s duty. In the course of any substantial period—for exam-
ple, two years—many agencies will make at least one decision that does not sur-
vive judicial review. By itself, a mistake should not be counted as neglect of duty.
There is significant space between the two.
For removal, what is necessary would something much more than an isolated
mistake—such as a
series of serious mistakes, suggestive of what the President
reasonably believes to be consistent or glaring indifference to legal requirements.
We acknowledge that this formulation leaves ambiguity and, accordingly, hard
cases.
79
And we would add that a truly egregious isolated case, in which an offi-
cial has not merely made a mistake but has made a patent and highly consequen-
tial departure from a legal duty, could plausibly justify discharge. We offer an
example below.
80
But our focus is on a pattern of neglect and not on imaginable
hard cases.
The locus classicus for this
distinction between isolated error and a pattern of
neglect of legal duty is, again, the opinion in Myers
81
—a more nuanced opinion
than casual discussion acknowledges. In the Court’s discussion of formal adjudi-
cation in the Executive Branch, it wrote:
[T]here may be duties of a quasi judicial character imposed on executive offi-
cers and members of executive tribunals whose decisions after hearing affect
74. Auer v. Robbins, 519 U.S. 452, 461 (1997) (deferring to an agency’s interpretation of its own
regulation because it was neither plainly erroneous nor inconsistent with the regulation).
75. Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 844–45 (1984).
76. Motor Vehicles Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983).
77. Universal Camera Corp. v. NLRB, 340 U.S. 474, 487–88 (1951).
78. See supra notes 75–77.
79. For discussion of certain harder cases, see infra Section II.D.4.
80. See infra Section II.D.4.
81. Myers v. United States, 272 U.S. 52 (1926).
2021] PRESIDENTIAL REVIEW 651
interests of individuals, the discharge of which the President cannot in a partic-
ular case properly influence or control. But even in such a case he may con-
sider 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. Otherwise he does not dis-
charge his own constitutional duty of seeing that the laws be faithfully
executed.
82
The passage is not without ambiguity, but one reading is that although a presi-
dential
intervention into one particular formal adjudication—a “ticket good for
one day [and train] only”
83
—risks favoritism or political bias, it is entirely legiti-
mate for the President to review the adjudicator’s performance over a series of
decisions (on the whole), by definition involving a number of different parties,
and to decide that the adjudicator’s performance is beyond the acceptable boun-
daries of expertise or prudence. Our submission then is that if members of an in-
dependent agency have neglected their duty on multiple occasions by voting for
and supporting outcomes that the President believes to be unlawful, there is suffi-
cient ground for removal. This formulation is not meant to create an arithmetic
straitjacket. It is natural to wonder about how many occasions are “multiple.”
Two are not; ten certainly are. There is no escaping a degree of judgment here
from both the President and Executive Branch lawyers, focusing on both the
number of cases of neglect and their egregiousness.
With respect to supervision, taken up in more detail below, the analysis should
be
similar. If the President reasonably believes that a commission has acted in
violation of existing law in a single case, he cannot overrule the decision for that
reason alone. In such cases, we might have nothing more than a reasonable
differ-
ence
of opinion, and in those circumstances, the President cannot tell an inde-
pendent agency that it must do as he prefers. But the President can take steps to
ensure that independent agencies act in conformity with their legal obligations.
84
This authority gives him considerable supervisory authority. Let us spell out these
points by reference to particular cases.
D. THE STANDARD APPLIED: REMOVAL
1. Easy Cases: Group One
Under any approach, including ours, many imaginable cases are
straightfor-
ward:
the President has the power of discharge. In such cases, the INM standard
is unmistakably clear. For example:
82. Id. at 135; cf. Nash v. Bowen, 869 F.2d 675, 680–81 (2d Cir. 1989) (upholding Social Security
Act reforms based on aggregate performance of administrative law judges).
83. See Richard M. Re, On “A Ticket Good for One Day Only,” 16 G
REEN BAG 2D 155, 155 (2013).
84. In some cases, the statutory obligation is not sharply defined and grants the agency broad
discretion. Even in such cases, there are prohibitions on arbitrariness or on decisions lacking substantial
evidence, and agencies must do their duties by avoiding inconsistency with such prohibitions.
652 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
(a) A member of the Federal Reserve Board has repeatedly missed meetings.
This member has not done the required work and has neglected to complete
numerous assignments. The member’s colleagues on the Board are disap-
pointed and embarrassed; they have to pick up the slack. The President can
discharge the member on grounds of palpable inefficiency.
(b) A member of the Consumer Product Safety Commission has taken a bribe
from a company under the Commission’s investigation. That action is plainly
malfeasance, and the President can discharge the member for that reason.
(c) The Social Security Commissioner has repeatedly missed statutory deadlines.
The evidence suggests that the Commissioner is not performing the job.
Although honest, the Commissioner has failed to produce multiple reports
required by Congress or to issue mandatory regulations. These failures count
as neglect of duty, and the Commissioner can be discharged for that reason.
In cases (a) through (c), our approach encompasses minimalist approaches.
2. Easy Cases: Group Two
Conversely, if the President has only the authority to correct what he
reason-
ably
believes to be a clear neglect of legal duty, numerous actual and imaginable
cases are just as straightforward: the President lacks the power of removal. For
example:
(d) The Nuclear Regulatory Commission issues a new regulation, designed to
reduce risks of a nuclear accident. The President believes that it is unneces-
sary. In his view, the regulation will do little good and will impose exces-
sive burdens on the nuclear industry. Though many and perhaps most
people agree with the President’s assessment, the Commission’s decision
falls within the bounds of reasonableness. The President may not remove
members of the Commission who voted in favor of the regulation.
(e) The Securities and Exchange Commission (SEC) issues a rule governing
“conflict minerals.”
85
In the President’s view, the rule goes well beyond
what is required by the authorizing statute, and its costs greatly exceed its
benefits.
86
This case is different from case (d) because there is a legal ques-
tion, which in the President’s view, the agency has resolved incorrectly.
But so long as the agency’s view is not self-evidently unreasonable, the
same (easy) result as case (d) obtains, and for the same reason.
(f) The Federal Trade Commission (FTC) allows a merger between social
media companies. In the President’s view, the merger is inconsistent with
the Commission’s regulations and also with the underlying statute. The
President’s view is perfectly reasonable, but the Commission has also acted
85. For the details of the actual rule, see Conflict Minerals, 77 Fed. Reg. 56,274, 56,275–77 (Sept. 12,
2012) (codified at 17 C.F.R. pts. 240, 249b).
86. For relevant discussion, see National Ass’n of Manufacturers v. SEC, 748 F.3d 359, 369–70 (D.C.
Cir. 2014).
2021] PRESIDENTIAL REVIEW 653
in a way that is easily defensible by reference to governing law. Same result
as (d) and (e).
3. Easy Cases: Group Three
Unlike minimalist approaches, our approach would allow the President to
dis-
charge
a member of an independent regulatory agency in the following
circumstances:
(g) A member of the Nuclear Regulatory Commission believes that because of
private incentives, nuclear power plants are now safe, and that they will be
for the imaginable future. The member opposes any and all federal regula-
tion even when statutes require it. The President may discharge the member
on two grounds: the member is neglecting their (legal) duty under the gov-
erning statute and the member has adopted a theory of regulation that—let
us stipulate—essentially all reasonable people roundly reject, which means
that the member’s approach is arbitrary within the meaning of the APA. In
either case, we are dealing with neglect of duty within the meaning of the
relevant statute.
(h) A member of the National Labor Relations Board (NLRB) votes regularly
with labor unions. More than that, the member does so even in cases in
which the law and the facts do not favor the union. For example, the Board
member has said that there is evidence of “anti-union animus” on the part
of employers even in cases in which no reasonable person could find such
evidence. The President may discharge on the ground that the member has
neglected a legal duty. The Board member is frequently acting in ways that
are consistent with legal obligations.
(i) A member of the Federal Reserve Board believes that, even at modest lev-
els, inflation is a serious problem for the economy and also that unemploy-
ment is tolerable, even if it becomes quite high. The member votes to keep
interest rates high during a period in which the national unemployment rate
is skyrocketing and increasing numbers of people find themselves without
work. As a matter of law, the member is nearly isolated in their opinion
about how to discharge their statutory duty; the member is also acting in a
way that is inconsistent with mainstream economic thinking, broadly
defined, such that their actions are arbitrary under the APA. The President
can discharge the member.
4. Harder Cases: Group Four
We acknowledge that, under our approach, some cases cannot be counted as
straightforward. For example:
(j) A member of the Federal Trade Commission concludes that certain adver-
tisements are neither unfair nor deceptive; within the Commission, that
member’s is the only vote to this effect, and the underlying reasoning is
654 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
difficult to understand. The member appears to have ignored both the law
and the facts. At the same time, this particular vote is a departure from the
member’s normal behavior, which is well within the bounds of reason. The
best conclusion is that the President cannot discharge the member for an
isolated error, even if it is unquestionably an error. The qualification is that
if the isolated error, reasonably deemed an egregious neglect of duty, occurs
in a case of extraordinary importance, the issue becomes more difficult.
(k) The Social Security Commissioner calls for and oversees a large-scale revi-
sion of the Social Security Disability Guidelines,
87
with the purpose and
effect of making disability benefits more broadly available. The revision
causes a public outcry. Most people, including the President, believe that it
goes too far insofar as it makes disability benefits available to people who
are not, in their view, truly disabled. The new regulation will be challenged
in court, and most people think that it will be struck down, but given princi-
ples of judicial deference to agency action, many people believe that it will
be upheld. This is not an easy case. On the one hand, the President can rea-
sonably believe that the Commissioner has acted arbitrarily and inconsis-
tently with the governing statute in an extraordinarily important
rulemaking, helping to define the agency’s entire mission. On the other
hand, the Commissioner can reasonably believe that the action is both law-
ful and reasonable. Because of the reasonableness of the Commissioner’s
belief, the better view is that there has been no neglect of duty; it is easier
than case (j).
(l) The Federal Communications Commission (FCC) issues a new regulation
restoring the fairness doctrine.
88
The President believes that the regulation
violates the First Amendment and transgresses statutory limits on the
FCC’s power, the President also believes that the new regulation is arbitrary
and capricious under the APA. Most observers agree with him on at least
two of those points, and in fact, the regulation is invalidated in court on con-
stitutional grounds. Though this is not an easy case, the better view is that
the President cannot discharge a commissioner for voting in favor of the
regulation. We have noted that many agency decisions are invalidated, and
invalidation cannot be enough to justify a presidential discharge. The ques-
tion is whether there has been a neglect of duty in our sense, and so long as
the FCC reasonably believed that it acted lawfully, its members should not
be vulnerable to presidential discharge.
(m) A member of the NLRB has a strong view: labor unions have been system-
atically mistreated by both federal courts and the NLRB. The member
votes for unions in numerous cases, not in a way that flouts the law, but in
a way that suggests a kind of “result orientation.” The President believes
that this amounts to neglect of duty. This is not an obvious case; we need
to know more. If the member is repeatedly resolving difficult issues in
favor of unions, there is no neglect of duty, and the President cannot
87. For relevant discussion, see generally Heckler v. Campbell, 461 U.S. 458 (1983).
88. See Red Lion Broad. Co. v. FCC, 395 U.S. 367, 369 (1969).
2021] PRESIDENTIAL REVIEW 655
remove the member. All we have is a dispute over policy. But if the mem-
ber is not taking hard questions seriously, and a thumb is always on the
scale in a way that suggests indifference to factual disputes and hence a
kind of repeated arbitrariness, the President probably has the authority to
remove the member.
Note that there is an underlying concern here involving the precise relationship
between judicial invalidation and presidential review. Our focus has been on
presidential judgments, but what if some court has actually spoken? Roughly
speaking, let us say that under standard doctrines of judicial review in
administra-
tive
law, agencies lose only when they are clearly wrong. In routine cases, agen-
cies’ statutory interpretations will be upheld unless the interpretations violate the
text under Chevron step one or are found unreasonable under Chevron step two;
89
their policy choices will be upheld unless arbitrary or capricious; their findings of
fact will be upheld unless a court concludes they are arbitrary or unsupported by
substantial evidence; and so on. Does this imply that whenever a court overturns
an agency, the agency neglected its duty for purposes of the INM standard, thus
justifying presidential removal?
Certainly not. There is an ample domain of cases in which an agency and a
court can reasonably disagree
over whether the agency has acted lawfully.
90
Such
situations are the routine stuff of judicial review and should not be mysterious.
The President is not entitled to discharge an agency head merely because that
head has voted in favor of a decision that was invalidated. If, by contrast, a court
has upheld a decision or a series of decisions, the President would be forbidden
from removing an agency head who has voted in favor of such decisions.
Validation is a demonstration that the agency has not violated its legal duty.
III. P
RESIDENTIAL SUPERVISION
The supervisory power refers to the President’s capacity to influence and over-
see
an agency’s decisions, not merely to remove its leaders. Supervisory authority
is usually taken to follow from removal authority, which means that the existence
and extent of the power to supervise are inferences from the power of removal. It
would be possible to say that under the INM standard, a President can remove
those leaders only ex post, and cannot influence or oversee their decisions in an
ongoing way. But according to long-standing understandings, the power of re-
moval
implicitly carries with it a power of supervision.
91
89. This is a simplification, of course. See Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467
U.S. 837, 842–44 (1984) (describing the two-step inquiry in greater detail).
90. See, e.g., Michigan v. EPA, 576 U.S. 743, 750–51, 759 (2015); Babbitt v. Sweet Home Chapter
of Cmtys. for a Great Or., 515 U.S. 687, 708 (1995).
91. See Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2191–92 (2020) (“The President’s power to
remove—and thus supervise—those who wield executive power on his behalf follows from the text of
Article II, was settled by the First Congress, and was confirmed in the landmark decision Myers v.
United States.” (citation omitted)).
656 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
A. EXECUTIVE AGENCIES
What does that supervisory power include, and what are its limits? For orienta-
tion,
let us begin with executive agencies, over which the President is understood
to have a great deal of supervisory power. For present purposes, we will understand
executive agencies as those whose heads are not protected from presidential removal
by the INM standard or anything close to it. For example, the Departments of
Agriculture, Commerce, Defense, Energy, Health and Human Services, State, and
Transportation are executive agencies; so is the Environmental Protection
Agency. And although an understanding of the President’s supervisory power
over executive agencies is not, strictly speaking, necessary to an
understand-
ing
of his supervisory authority over independent agencies, it is more than
helpful by way of background.
It is agreed that the President may impose purely procedural
require-
ments
on executive agencies—for example, by requiring them to produce
cost–benefit analyses before issuing certain important regulations.
92
It is
also agreed that the President may exercise substantive supervisory power
by requiring agencies to adhere to certain principles or to follow certain
policy commitments.
93
Any such requirements may not transgress statutory
requirements. But if the President says that an executive agency may not
proceed unless the benefits justify the costs, or may issue a regulation only
if it has removed two other regulations,
94
he is probably on firm ground, at
least insofar as presidential control is exercised within the confines of the
law.
The reason for the word “probably” is the continuing debate about
whether the supervisory power
includes the directive power.
95
That power
is understood to exist if the President can order an agency head, in particular
cases, to act or to refrain from acting—for example, to issue a regulation to
reduce greenhouse gas emissions, to issue a regulation to close or open air
travel during a pandemic, to refrain from issuing a regulation to protect
food safety, or to adopt a specific policy with respect to asylum applicants.
There are three questions relevant to assessing the directive power:
(1) whether the President has the directive power as a matter of
constitu-
tional
right; (2) if, contrary to (1), Congress has the constitutional authority
to eliminate the directive power, whether it does so by conferring relevant
authority (such as rulemaking) on agency heads, including the heads of
92. See Exec. Order No. 12,866, 3 C.F.R. § 638 (1994), reprinted in 5 U.S.C. § 601, at 802–06 (2012)
(requiring agencies to create a regulatory plan for “the most important significant regulatory actions that
the agency reasonably expects to issue,” which must include a summary of the “anticipated costs and
benefits” and available alternatives).
93. See id.
94. See Exec. Order No. 13,771, 82 Fed. Reg. 9339, 9339 (Feb. 3, 2017) (stating that “it is important
that for every one new regulation issued, at least two prior regulations be identified for elimination”
because the Executive Branch must be “prudent and financially responsible in the expenditure of
funds”).
95. See Kagan, supra note 36, at 2326–28.
2021] PRESIDENTIAL REVIEW 657
purely executive agencies; and (3) whether Congress has the authority sug-
gested by (2).
With respect to (1), it would be possible and perhaps tempting to read the vest-
ing of executive
power in “a president of the United States” to mean that the exec-
utive power is the President’s alone; the vesting clause might seem to recognize
the directive power.
96
Realistically, any cabinet head is likely to do as the
President wishes regardless of the legal status of the directive power. But as a
for-
mal
matter, any President would like to know that the directive power exists, and
in some (admittedly rare) cases, it might matter if a cabinet official wants to take
a firm stand.
With
respect to (2), in Myers v. United
States, the Court raised a serious
question about the existence of the directive power.
97
The Court endorsed the
idea of a strongly unitary executive.
98
At the same time, it pointedly and
somewhat mysteriously said, “Of course there may be duties so peculiarly
and specifically committed to the discretion of a particular officer as to raise a
question whether the President may overrule or revise the officer’s
interpreta-
tion
of his statutory duty in a particular instance.”
99
The basic idea in Myers,
then, seems to be that Congress may deny the President the directive power,
at least when the issue is the proper interpretation of an officer’s statutory
duty.
100
In view of the Myers Court’s emphasis on the Take Care Clause, that
fact is worth underlining.
101
Though the Court invoked the constitutional text
to support the idea of a strongly unitary executive, the Justices were nonethe-
less willing to acknowledge the possibility that when an officer has one inter-
pretation of his statutory duty, and the President another, the officer cannot
be overruled.
102
Importantly, however, these words are tentative, with not one but two indi-
cations
that they were not meant to settle much (“there may be” and “as to
raise a question whether”).
103
But the sentence signals the possibility that
Congress has the authority to forbid the President from ordering executive
officers to do as he wishes—and thus to insist that such officers have the
authority to make the ultimate call. If that is correct, then the real question is
one of statutory interpretation: Has Congress exercised that authority in
par-
ticular
cases? Has Congress said, as a matter of law, that duties have been
“peculiarly and specifically committed to the discretion of a particular
96. See CALABRESI & YOO, supra note 4, at 3–4. Attorneys general, in the early days of the republic,
expressed varying views on the question. For a clear summary, see Robert V. Percival, Presidential
Management of the Administrative State: The Not-So-Unitary Executive, 51 D
UKE L.J. 963, 977 (2001).
97. 272 U.S. 52, 106 (1926).
98. See id. at 134–35.
99. Id. at 135.
100. See id.
101. See id. at 117.
102. See id. at 135.
103. See id.
658 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
officer”?
104
Has it said that with respect to executive officers who are given,
for example, rulemaking authority?
With respect to (3), there are competing views with respect to Congress’s
authority in this
area.
105
Some people believe that whenever Congress has given
specific authority to an agency head—for example, a grant of rulemaking
author-
ity
to the Secretary of Transportation—the best conclusion is that the President
lacks the directive power.
106
In their view, the relevant statutes are plain, because
the agency head, and not the President, has been given the authority to make
rules. The only question is whether the grant of that authority to the agency head
transgresses constitutional boundaries. But embracing (3), Justice Kagan has
argued that because Congress has the authority to create independent agencies,
the very grant of power to an executive official ought to be taken as a matter of
convenience only and as reflective of a congressional expectation that, if the
President likes, he may tell his subordinates what to do.
107
Without resolving
whether Congress has the constitutional authority to deprive the President of
directive power over executive officials, Kagan urges that the mere grant of
authority to an executive official ought not to be taken as reflective of a congres-
sional desire to do that.
108
Her conclusion might well be supported by the avoid-
ance canon, at least if we believe that depriving the President of the directive
power would raise problems under Article II.
104. See id.
105. The Supreme Court seems to have rejected an unlimited directive power in Kendall v. United
States, 37 U.S. (12 Pet.) 524, 610 (1838):
The executive power is vested in a President; and as far as his powers are derived from the
[C]onstitution, he is beyond the reach of any other department, except in the mode prescribed
by the [C]onstitution through the impeaching power. But it by no means follows, that every
officer in every branch of that department is under the exclusive direction of the President.
Such a principle, we apprehend, is not, and certainly cannot be claimed by the President.
There are certain political duties imposed upon many officers in the executive depart-
ment, the discharge of which is under the direction of the President. But it would be an
alarming doctrine, that [C]ongress cannot impose upon any executive officer any duty they
may think proper, which is not repugnant to any rights secured and protected by the [C]onsti-
tution; and in such cases, the duty and responsibility grow out of and are subject to the con-
trol of the law, and not to the direction of the President.
But the passage is hardly unambiguous. It can be taken to say that “political duties” of certain kinds are
“under the direction of the President,” without clarifying whether that is a matter of statutory
requirement or constitutional compulsion. To say that some duties may be freed from “the direction of
the President” is not to say which duties may be so freed.
106. See Kevin M. Stack, The President’s Statutory Powers to Administer the Laws, 106 C
OLUM. L.
R
EV. 263, 295–96 (2006); Percival, supra note 96, at 1005–07; Robert V. Percival, Rediscovering the
Limits of the Regulatory Review Authority of the Office of Management and Budget, 17 E
NVTL. L. REP.
10,017, 10,019–20 (1987).
107. See Kagan, supra note 36, at 2326–27.
108. See id.
2021] PRESIDENTIAL REVIEW 659
B. INDEPENDENT AGENCIES
Now let us turn to independent agencies, recalling that for present purposes,
we are understanding them to be headed by people who are protected by the INM
standard. The question is how to understand the meaning of that standard insofar
as it bears not on removal, but on supervision. If Congress has said that the
President may discharge agency heads only for INM, what kind of supervisory
power does he possess over those agency heads?
Our approach to presidential authority offers a distinctive answer to that ques-
tion. We have emphasized
that with respect to removal, the key question is
whether agency heads have neglected their duty. The same is true with respect to
supervision. If the President is seeking to supervise policy judgments of agency
heads in order to ensure that they decide specific questions as he thinks best, he is
unquestionably violating the INM standard. To that extent, the President lacks su-
pervisory power as a matter of statutory interpretation. (As before, it is possible
to mount a constitutional objection.) But what if he is attempting to ensure that
agency heads actually do their jobs in some minimal sense? What if he is attempt-
ing
to ensure that they are efficient? That they do not neglect their duty? That
they do not engage in malfeasance? The President is acting within appropriate
bounds, whether he is engaging in supervision in some broad sense or actually
directing the agency heads what to do.
In our
view, inefficiency and malfeasance are relatively straightforward.
109
As
we have noted, a dictionary definition of inefficiency could be taken to be quite
broad, but that interpretation was rejected in Seila Law by reference to what the
Court saw as the implicit holding of Humphrey’s Executor.
110
A broad under-
standing
of inefficiency would also be inconsistent with the context that gave rise
to the IMN standard.
111
The standard does authorize the President to tell commis-
sioners to show up for work, to do their jobs expeditiously, and to produce in ac-
cordance with a respectable schedule. He can also set out standards to combat
wrongdoing and corruption. The harder issues lie elsewhere. With respect to pro-
cedural
requirements, broad policy guidance, and specific direction, the relevant
question is clear: is the assertion of presidential authority an effort to ensure
against some kind of neglect of duty? But the answer to that question is less clear.
Various procedural
requirements, actual and imagined, are easy to justify on
neglect-of-duty grounds. Suppose that the President is demanding some kind of
explanation of agency action—for example, with a requirement that agencies
identify the legal authority that justifies their decisions.
112
A somewhat harder
109. If we were to embrace a broad understanding of “efficiency,” the analysis would be more
complicated, and the President would have more expansive authority.
110. See supra note 68 (noting the Seila Law Court’s plain rejection of the broad view of
inefficiency, which was described in Judge Griffith’s concurring opinion in PHH Corp.).
111. See Manners & Menand, supra note 20 (manuscript at 74–75) (describing the “inefficient
[railroad] commissioner” as one “unable to comprehend the technical reports and complicated account
books of the railroads he oversaw, or dishonest and unscrupulous in the exercise of his duties”).
112. See, e.g., Exec. Order No. 12,866, 3 C.F.R. § 638 (1994), reprinted in 5 U.S.C. § 601, at 802–06
(2012).
660 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
case would arise if he requires independent agencies to catalogue the costs and
benefits of what they are planning to do.
113
If the President’s goal here is to ensure
that agencies follow his preferred policies, he has exceeded the bounds of his
authority. But if he is attempting to ensure that they are acting within the bounds
of their legal duty,
114
any such mandate is lawful and all the more clearly so inso-
far as the requirement is merely procedural. Something similar can be said for
other exercises of substantive supervisory power, such as actual and hypothetical
supervisory power. The President is entitled to direct independent agencies to fol-
low general policies and principles insofar as their failure to do so would count as
neglect of duty.
115
In this light, we could imagine both easy and hard cases. Suppose, for example,
that the President wants the SEC to offer an account of the legal foundation of its
regulations. If that is all he wants, the President can fairly say that he is trying to
ensure against neglect of duty; a requirement that the SEC provide an account of
the legal foundations of its regulations would be consistent with the INM
stand-
ard.
What about a general requirement that independent agencies catalogue the
costs and benefits of what they do? If the President can plausibly argue that he is
imposing that requirement to ensure against neglect of duty, the INM standard
allows him to proceed. To know whether an argument to that effect is plausible,
we would have to parse the underlying statutes to see if a failure to consider costs
and benefits counts, in fact, as a neglect of legal duty.
If, by contrast, the President says that the SEC must consider the impacts of its
decisions on climate
change, and if the SEC (rightly) concludes that it lacks the
authority to do that, then the President cannot order the SEC to do as he wishes. A
harder case would arise if the President directs the SEC to consider the impacts of
its decisions on climate change, and if the President and the SEC have a reasona-
ble difference of opinion about whether the SEC has the authority to do so. If the
SEC reasonably believes that it lacks that authority, it would probably be hard to
argue that its failure to do as the President wishes represents a neglect of duty.
What about a requirement that, to the extent permitted by law, agencies make
cost–benefit
analysis the rule of decision? That would be harder. The question is
whether such a requirement can be seen as an effort to ensure that agencies do not
neglect their duty. For present purposes, it should be sufficient to say that some
cases clearly fall inside and outside the line. The President is certainly entitled to
take steps to ensure against plainly unlawful rulemaking or plainly arbitrary poli-
cymaking,
where arbitrariness is understood in the general terms of the APA. The
President is forbidden from requiring independent agencies to follow policies
113. See, e.g., Exec. Order No. 13,563, 3 C.F.R. § 215 (2012), reprinted in 5 U.S.C. § 601, at 816–17
(2012).
114. See Michigan v. EPA, 576 U.S. 743, 759–60 (2015) (holding that agencies must “consider”
costs as well as benefits under a provision of the Clean Air Act). We do not mean to express a view on
when, if ever, agencies must engage in cost–benefit analysis under the Administrative Procedure Act.
115. See Exec. Order No. 13,579, 3 C.F.R. § 256 (2011), reprinted in 5 U.S.C. § 601, at 817–18
(2012).
2021] PRESIDENTIAL REVIEW 661
that the agencies reasonably believe to be wrong or mistaken (and that agencies
could therefore decline to follow without neglecting their legal duty). When cases
are hard, it is because of a lack of clarity about whether the President’s judgment
reflects a decision to impose his own policy choice when others are reasonable, or
whether the President is genuinely ruling legally irresponsible choices, violative
of an agency’s legal duty, out of bounds. The central point is that whatever the
President does, it must fit under the general idea of preventing neglect of duty as
substantive law understands it.
What about the directive power? Here again, the answer depends on whether
the
President is seeking to prevent violations of the INM standard. It follows that
the President could not direct a member of, say, the Federal Trade Commission to
vote in accordance with the President’s own policy preferences. But if that
mem-
ber
seeks to go on vacation for two years or to take a bribe, the President can
direct otherwise. And if a member of an independent agency seeks to act in a way
that self-evidently and patently violates the law, the President is entitled to say: I
direct you not to vote that way.
C. ON POLICING LINES
With respect to neglect of duty, the line that we are drawing might well be dif-
ficult
to police—a concern that maps onto that in Seila Law, where the Court was
troubled by the difficulty of producing a workable standard.
116
We hope that the
line is clear enough, even though we acknowledge that under our approach, hard
intermediate cases cannot be avoided. We have discussed several such cases. Is
the existence of such cases a decisive objection?
In principle, it might well be better to have a hard-and-fast line. But under any
approach,
it is difficult to avoid hard cases. One way to do so would be to adopt
maximalism and to allow the President to control independent agencies
essen-
tially
as he wishes. But that approach was decisively rejected in Seila Law and
Humphrey’s Executor.
117
Another way to avoid or at least to reduce hard or inter-
mediate cases would be to adopt minimalism, which would not allow anything
like policymaking control by the President. But even under minimalism, it would
be necessary to decide what kinds of inefficiencies and malfeasance are legiti-
mate bases for removal. In any case, minimalism would violate the natural mean-
ing of the statutory words— neglect of duty—and severely compromise the
President’s authority under the Take Care Clause. In our view, hard intermediate
cases are not too high a price to pay for preventing violations of both the govern-
ing statute and the Constitution. And if the analysis here is correct, most cases
fall on one or the other side of the line.
There is a practical problem. Any President might well be tempted to say that
commissioners are neglecting their
duties if they do not do what he thinks best. If
the Chair of the Federal Reserve Board does not cut interest rates when the
116. See Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2206 (2020).
117. See id. at 2192; Humphrey’s Ex’r v. United States, 295 U.S. 602, 629 (1935).
662 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
economy is struggling, the President might insist that the Chair has not merely made
a mistake but instead committed an egregious wrong, amounting to neglect of
duty.
118
On multiple occasions, President Donald Trump has signaled his view that the Federal Reserve has
been egregiously wrong. See, e.g., Heather Long, Federal Reserve Makes Small Interest Rate Cut. Trump Slams
Central Bank for Having ‘No Guts.,’ W
ASH. POST (Sept. 18, 2019, 4:45 PM), https://www.washingtonpost.com/
business/2019/09/18/federal-reserve-cuts-interest-rates-quarter-point-counter-trumps-trade-war.
If a member of the Nuclear Regulatory Commission does not take an
aggressive stand on safety issues, the President might urge that the member has not
merely erred but instead defaulted on his legal responsibility. If a member of the
National Labor Relations Board does not do what the President thinks should be
done to protect labor unions, the President might conclude that he has neglected his
duty. In many imaginable cases, a presidential claim to this effect might be sincere
but altogether wrong, turning our proposed approach into a close sibling of maxim-
alism, which we have rejected. The problem might be heightened by another practi-
cal reality: any member of an independent agency would know that a president
might try to overreach, and for that reason, might do his anticipated bidding.
What is the remedy? There are two institutional answers, and there is one
response on a conceptual
level. The first institutional answer involves the Executive
Branch itself. To a greater or lesser extent, Presidents are restrained by their lawyers
in the White House and the Department of Justice.
119
When things are working as
they should (which is admittedly not always the case), those constraints are real, and
they deter the President from crossing legal lines. Deliberative processes are usually
not disclosed to the public, but we can report, from personal experience and from
personal accounts of those we trust, that when White House Counsel or the Attorney
General tell a President that he lacks the legal authority to do what he wishes, the
President ordinarily yields. The second institutional answer involves the federal
courts. As Humphrey’s Executor shows, a removed member of an independent
com-
mission
has an ability to go to court to contest the removal.
120
It is true that any such
proceeding is time-consuming, expensive, and after the fact. No one should doubt
that power can be abused, and the remedies for abused power can be inadequate.
More importantly, however, it is a misconception to think that the occasional
blurriness of legal lines
is a unique objection to our suggested reading of INM
clauses. One more time: It is an objection to any rule other than perhaps the most
extreme possible rule such as minimalism or maximalism (an approach that the
Court has never adopted, even in Myers). Any plausible rule will establish some
line or some imperfectly precise standard for evaluating presidential action, and
it can then be fairly argued that the presidency can test the limits.
121
Potentially
118.
119. For a sustained argument to this effect, see generally J
ACK GOLDSMITH, POWER AND
CONSTRAINT: THE ACCOUNTABLE PRESIDENCY AFTER 9/11 (2012).
120. Humphrey’s Ex’r, 295 U.S. 602, 618–19.
121. It is true that the minimalist position, forbidding the President from exercising any kind of
control over policy judgments, might seem to create a hard-edged constraint on overreaching. But for
reasons that we have offered, that position runs up against the best understanding of the INM standard
especially when it is read against the constitutional background.
2021] PRESIDENTIAL REVIEW 663
hard cases and the risk of abuse, by themselves, are not decisive objections to tak-
ing the neglect of duty standard seriously.
C
ONCLUSION
The extent of presidential authority over the independent regulatory commis-
sion
remains one of the great unanswered questions in American public law.
Most of the proposed answers attempt to specify what it means to say that the
Executive Branch is “unitary.” On both originalist and non-originalist grounds,
there are reasonable arguments in favor of the view that, as a matter of
constitu-
tional
right, the President must have substantial ability to remove and supervise
all those who execute federal law.
122
There are also reasonable counterarguments,
pointing to ambiguities in the historical record, changed circumstances, and the
importance of diffusing power and ensuring against White House control of
cer-
tain
subject areas, such as monetary policy and communications.
123
For our pur-
poses, the most important conclusion is that if agencies are genuinely insulated
from presidential oversight, there are serious constitutional questions.
Our emphasis here has been on long-neglected words in the INM standard: neglect
of
duty. Those words fit remarkably well with a core responsibility of the President,
which is to take care that the laws be faithfully executed. If the President believes that
commissioners have neglected their (legal) duties, he would seem to have authority to
remove them under the clear language of the INM standard. At the same time, serious
questions remain. Under the INM standard, it would be implausible to say that the
President can remove, say, a member of the National Labor Relations Board on the
ground that he voted for a rule that a federal court invalidated. It would be even more
implausible to say that the President can remove a member for voting for a rule that,
in the President’s view, a court should invalidate. Although that interpretation would
not adopt the maximalist position, it would come far too close to it. The more sensible
understanding is that “neglect” of duty refers to something egregious, such as frequent
or repeated disrespect for legal standards or egregious disrespect on especially
promi-
nent occasions.
That
interpretation of the INM standard, based on a conception of presidential
review, greatly softens the constitutional objection. To be sure, those who believe in
a strongly unitary presidency would not be entirely satisfied. In their view, all
agency heads must be at-will employees, and the President must have plenary
powers of supervision and perhaps direction over them.
124
In our view, that conclu-
sion
is difficult to defend. But our goal here has not been to resolve the constitutional
issue. It has been to suggest that the prior issue is statutory, and that existing statutes
grant the President a significant degree of authority over the supposedly independent
agencies—authority that includes much, and on a reasonable view all, of what he is
entitled to have under the Take Care Clause.
122. See, e.g., CALABRESI & YOO, supra note 4, at 3–4; Lessig & Sunstein, supra note 36.
123. See, e.g., Gerhard Casper, An Essay in Separation of Powers: Some Early Versions and
Practices, 30 W
M. & MARY L. REV. 211, 211 (1989).
124. See C
ALABRESI & YOO, supra note 4, at 3–4.
664 THE GEORGETOWN LAW JOURNAL [Vol. 109:637
