United States Court of Appeals
         FOR THE DISTRICT OF COLUMBIA CIRCUIT



Argued June 12, 2020                   Decided June 24, 2020

                         No. 20-5143

                  IN RE: MICHAEL T. FLYNN,
                         PETITIONER


       On Emergency Petition for a Writ of Mandamus


     Sidney Powell argued the cause for petitioner. With her on
the petition for a writ of mandamus were Molly McCann and
Jesse R. Binnall.

     Jeffrey B. Wall, Deputy Solicitor General, U.S.
Department of Justice, argued the cause for respondent United
States of America. With him on the response to the petition for
a writ of mandamus were Noel J. Francisco, Solicitor General,
Eric J. Feigin, Deputy Solicitor General, Frederick Liu,
Assistant to the Solicitor General, Kenneth C. Kohl, Acting
Principal Assistant U.S. Attorney, and Jocelyn Ballantine,
Assistant U.S. Attorney.

    David Yost, Attorney General, Office of the Attorney
General for the State of Ohio, Benjamin M. Flowers, Solicitor
General, Steve Marshall, Attorney General, Office of the
Attorney General for the State of Alabama, Kevin G. Clarkson,
Attorney General, Office of the Attorney General for the State
of Alaska, Leslie Rutledge, Attorney General, Office of the
Attorney General for the State of Arkansas, Ashley Moody,
Attorney General, Office of the Attorney General for the State
                              2

of Florida, Christopher M. Carr, Attorney General, Office of
the Attorney General for the State of Georgia, Jeff Landry,
Attorney General, Office of the Attorney General for the State
of Louisiana, Lynn Fitch, Attorney General, Office of the
Attorney General for the State of Mississippi, Eric Schmitt,
Attorney General, Office of the Attorney General for the State
of Missouri, Timothy C. Fox, Attorney General, Office of the
Attorney General for the State of Montana, Mike Hunter,
Attorney General, Office of the Attorney General for the State
of Oklahoma, Alan Wilson, Attorney General, Office of the
Attorney General for the State of South Carolina, Ken Paxton,
Attorney General, Office of the Attorney General for the State
of Texas, Sean D. Reyes, Attorney General, Office of the
Attorney General for the State of Utah, and Patrick Morrisey,
Attorney General, Office of the Attorney General for the State
of West Virginia, were on the brief for amici curiae the States
in support of petitioner.

     William J. Olson, Jeremiah L. Morgan, Herbert W. Titus,
and Robert J. Olson were on the brief for amici curiae Former
United States Attorney General Edwin Meese III and
Conservative Legal Defense and Education Fund in support of
petitioner.

     Jerome M. Marcus was on the brief for amici curiae
Eleven Members of the United States House of Representatives
in support of petitioner.

    John Reeves, pro se, was on the brief for amicus curiae
John M. Reeves in support of petitioner and the United States.

   Michael H. McGinley was on the brief for amici curiae
Majority Leader Mitch McConnell and Senators Tom Cotton,
                              3

Mike Braun, Kevin Cramer, Ted Cruz, Charles E. Grassley,
and Rick Scott in support of the United States.

     Leslie McAdoo Gordon was on the brief for amicus curiae
Federal Practitioners in support of petitioner and the United
States.

    Eric B. Rasmusen, pro se, was on the brief for amicus
curiae Professor Eric Rasmusen in support of petitioner.

    Beth A. Wilkinson argued the cause for respondent Judge
Emmet G. Sullivan. With her on the response to the petition for
a writ of mandamus were Kosta S. Stojilkovic and Rakesh
Kilaru.

    Eugene R. Fidell, Stanley J. Marcus, and Gershon M.
Ratner were on the brief for amicus curiae Lawyers Defending
American Democracy, Inc. in support of respondent.

    Lawrence Robbins, Alan E. Untereiner, D. Hunter Smith,
and William W. Taylor III were on the brief for amicus curiae
Watergate Prosecutors in support of respondent.

     Daniel E. Jackson and John W. Keker were on the brief for
amicus curiae Former Federal District Court Jurists in support
of respondent.

    Gregory S. Smith was on the brief for amicus curiae New
York City Bar Association in support of respondent.

    Before: HENDERSON, WILKINS, and RAO, Circuit Judges.

    Opinion for the Court filed by Circuit Judge RAO.
                                4

    Opinion dissenting in part filed by Circuit Judge WILKINS.

     RAO, Circuit Judge: Michael Flynn, former National
Security Advisor to President Donald J. Trump, pleaded guilty
to making false statements under 18 U.S.C. § 1001. Before
sentencing, Flynn moved to withdraw his plea, alleging that the
government failed to produce material exculpatory evidence
and breached the plea agreement. Several months later, the
U.S. Attorney for the District of Columbia filed a motion to
dismiss all charges. See FED. R. CRIM. P. 48(a) (“The
government may, with leave of court, dismiss an indictment,
information, or complaint.”). In its motion, the government
explains that in light of newly discovered evidence of
misconduct by the Federal Bureau of Investigation, the
prosecution can no longer prove beyond a reasonable doubt
that any false statements made by Flynn were material to a
legitimate investigation—an element the government contends
is necessary under Section 1001. See United States v. Gaudin,

, 509 (1995). The government’s motion to dismiss
also explains that “continued prosecution of the charged crime
does not serve a substantial federal interest.” Gov’t Mot.
Dismiss Criminal Information, No. 1:17-cr-232, ECF No. 198,
at 2 (May 7, 2020). The district judge currently presiding over
the case has yet to decide the government’s motion. Instead, he
has appointed an amicus to present arguments in opposition to
the government’s motion and to address whether Flynn should
be held in criminal contempt for perjury. The district judge has
also scheduled a hearing on these questions for July 16, 2020.

    Flynn petitioned for a writ of mandamus before this court
pursuant to the All Writs Act, 28 U.S.C. § 1651, seeking three
forms of relief: (1) an order directing the district court to grant
the motion to dismiss; (2) an order vacating the amicus
appointment; and (3) an order reassigning the case to a different
                                 5

district judge. For this court to grant a writ of mandamus, “the
right to relief must be ‘clear and indisputable’; there must be
‘no other adequate means to attain the relief’; and ‘the issuing
court, in the exercise of its discretion, must be satisfied that the
writ is appropriate under the circumstances.’” In re Cheney,

, 312–13 (D.C. Cir. 2008) (quoting Cheney v. U.S.
Dist. Court, 

, 380–81 (2004)). Applying these
standards, we grant Flynn’s petition in part.

     Although Rule 48 requires “leave of court” before
dismissing charges, “decisions to dismiss pending criminal
charges—no less than decisions to initiate charges and to
identify which charges to bring—lie squarely within the ken of
prosecutorial discretion.” United States v. Fokker Servs. B.V.,

, 742 (D.C. Cir. 2016). “To that end, the Supreme
Court has declined to construe Rule 48(a)’s ‘leave of court’
requirement to confer any substantial role for courts in the
determination whether to dismiss charges.” Id.; see also
Newman v. United States, 

, 480 (D.C. Cir. 1967)
(“Few subjects are less adapted to judicial review than the
exercise by the Executive of his discretion in deciding …
whether to dismiss a proceeding once brought.”). The
Judiciary’s role under Rule 48 is thus confined to “extremely
limited circumstances in extraordinary cases.” United States v.
Hamm, 

, 629 (5th Cir. 1981); United States v.
Ammidown, 

, 621 (D.C. Cir. 1973) (emphasizing
that Rule 48 motions must be granted “in the overwhelming
number of cases”). More specifically, “[t]he principal object of
the ‘leave of court’ requirement is … to protect a defendant
against prosecutorial harassment … when the Government
moves to dismiss an indictment over the defendant’s
objection.” Rinaldi v. United States, 

, 29 n.15
(1977). Rule 48 thus “gives no power to a district court to deny
a prosecutor’s … motion to dismiss charges based on a
                                   6

disagreement with the prosecution’s exercise of charging
authority.” Fokker 

.1

     Whatever the precise scope of Rule 48’s “leave of court”
requirement, this is plainly not the rare case where further
judicial inquiry is warranted. To begin with, Flynn agrees with
the government’s motion to dismiss, and there has been no
allegation that the motion reflects prosecutorial harassment.
Additionally, the government’s motion includes an extensive
discussion of newly discovered evidence casting Flynn’s guilt
into doubt. See Mot. to Dismiss, ECF No. 198. Specifically, the
government points to evidence that the FBI interview at which
Flynn allegedly made false statements was “untethered to, and
unjustified by, the FBI’s counterintelligence investigation into
Mr. Flynn.”

In light of this evidence, the government
maintains it cannot “prove either the relevant false statements
or their materiality beyond a reasonable doubt.”

is a quintessential justification for dismissing
charges. See 

(explaining that a
motion to dismiss should be granted “if it is explained to the
judge that there was … an insufficiency of evidence … or other
similar consideration”).

     The government’s representations about the insufficiency
of the evidence are entitled to a “presumption of regularity …

1
  In response to the petition, the district judge argues greater judicial
scrutiny is warranted when a Rule 48(a) motion is filed after a guilty
plea because formal judicial action has already been taken. Yet this
claim conflicts with black letter law: “Rule 48(a) continues to apply
even after conviction and sentencing while the case is on direct
appeal, and the same standard applies to a government request for
dismissal at that stage as applies if the request came prior to trial.”
3B WRIGHT & MILLER, FED. PRAC. & PROC. CRIM. § 802 (4th ed.
2013).
                                7

in the absence of clear evidence to the contrary.” United States
v. Armstrong, 

, 464 (1996) (quotation marks
omitted). On the record before the district court, there is no
clear evidence contrary to the government’s representations.
The justifications the district court offers in support of further
inquiry—for instance, that only the U.S. Attorney signed the
motion, without any line prosecutors, and that the motion is
longer than most Rule 48(a) motions—are insufficient to rebut
the presumption of regularity to which the government is
entitled.

     These clearly established legal principles and the
Executive’s “long-settled primacy over charging decisions,”
Fokker 

, foreclose the district court’s
proposed scrutiny of the government’s motion to dismiss the
Flynn prosecution. Before this court, the district judge explains
that he plans to “question the bona fides of the government’s
motion,” Sullivan Response 29 (quotation marks omitted),
“inquire about the government’s motions and representations,”
Sullivan Reply 26, “illuminat[e] the full circumstances
surrounding the proposed dismissal,”

and probe
“whether the presumption of regularity for prosecutorial
decisions is overcome” in “the unusual facts of this case,”
Sullivan Response 3. A hearing may sometimes be appropriate
before granting leave of court under Rule 48; however, a
hearing cannot be used as an occasion to superintend the
prosecution’s charging decisions, because “authority over
criminal charging decisions resides fundamentally with the
Executive, without the involvement of—and without oversight
power in—the Judiciary.” Fokker 

. The
district court’s orders appointing an amicus, see infra 8–10, and
scheduling the proposed hearing therefore constitute clear legal
error.
                              8

     Because legal errors ordinarily may be corrected on
appeal, a writ of mandamus is proper only if there is “no other
adequate means to attain … relief.” 

.
Although “an abstract concern with the separation of powers,”
does not rise to the level of an irreparable injury, In re
Al Nashiri, 

, 79–81 (D.C. Cir. 2015), we have
found the requisite harm as a matter of course when a party
alleges the district court’s action usurps a specific executive
power. See In re Cheney, 

, 731 (D.C. Cir. 2005)
(en banc); Fokker 

; Cobell v. Norton,

, 1139–40 (D.C. Cir. 2003); In re Sealed Case
No. 98-3077, 

, 1065–66 (D.C. Cir. 1998).

     In this case, the district court’s actions will result in
specific harms to the exercise of the Executive Branch’s
exclusive prosecutorial power. The contemplated proceedings
would likely require the Executive to reveal the internal
deliberative process behind its exercise of prosecutorial
discretion, interfering with the Article II charging authority.

(citing United States v. Cox, 

, 171 (5th Cir. 1965)). Thus, the district court’s
appointment of the amicus and demonstrated intent to
scrutinize the reasoning and motives of the Department of
Justice constitute irreparable harms that cannot be remedied on
appeal. See 

(“[I]nterference with the
internal deliberations of a Department of the Government of
the United States … cannot be remedied by an appeal from the
final judgment.”); see also 

.

    We must also assure ourselves that issuance of the writ “is
appropriate under the circumstances.” 

.
The circumstances of this case demonstrate that mandamus is
appropriate to prevent the judicial usurpation of executive
power. The first troubling indication of the district court’s
                              9

mistaken understanding of its role in ruling on an unopposed
Rule 48(a) motion was the appointment of John Gleeson to
“present arguments in opposition to the government’s Motion.”
Order Appointing Amicus Curiae, No. 1:17-cr-232, ECF No.
205, at 1 (May 13, 2020) (emphasis added). Whatever the
extent of the district court’s “narrow” role under Rule 48(a),
see Fokker 

, that role does not include
designating an advocate to defend Flynn’s continued
prosecution. The district court’s order put two “coequal
branches of the Government … on a collision course.” 

. The district court chose an amicus who had
publicly advocated for a full adversarial process. Based on the
record before us, the contemplated hearing could require the
government to defend its charging decision on two fronts—
answering the district court’s inquiries as well as combatting
Gleeson’s arguments. Moreover, the district court’s invitation
to members of the general public to appear as amici suggests
anything but a circumscribed review. See May 12, 2020,
Minute Order, No. 1:17-cr-232. This sort of broadside inquiry
would rewrite Rule 48(a)’s narrow “leave of court” provision.

     And we need not guess if this irregular and searching
scrutiny will continue; it already has. On May 15, Gleeson
moved for permission to file a brief addressing, among other
things, “any additional factual development [he] may need
before finalizing [his] argument” and suggesting a briefing and
argument schedule. Mot. to File Amicus Br., No. 1:17-cr-232,
ECF No. 209, at 1–2 (May 15, 2020). The district court granted
the motion and then set a lengthy briefing schedule and a July
16, 2020, hearing. See May 19, 2020, Minute Order, No. 1:17-
cr-232. In his brief opposing the government’s motion, Gleeson
asserted the government’s reasons for dismissal were “pretext”
and accused the government of “gross prosecutorial abuse.”
Amicus Br., No. 1:17-cr-232, ECF No. 225, at 38–59 (June 10,
                                  10

2020). He relied on news stories, tweets, and other facts outside
the record to contrast the government’s grounds for dismissal
here with its rationales for prosecution in other cases. See

46–47, 57–59.

     These actions foretell not only that the scrutiny will
continue but that it may intensify. Among other things, the
government may be required to justify its charging decisions,
not only in this case, but also in the past or pending cases cited
in Gleeson’s brief. Moreover, Gleeson encouraged the district
court to scrutinize the government’s view of the strength of its
case—a core aspect of the Executive’s charging authority. See
In re United States, 

, 453 (7th Cir. 2003)
(condemning district court’s failure to dismiss criminal charges
based on its view that “the government has exaggerated the risk
of losing at trial”). As explained above, our cases are crystal
clear that the district court is without authority to do so. See
Fokker 

; 

.

     Finally, each of our three coequal branches should be
encouraged to self-correct when it errs. If evidence comes to
light calling into question the integrity or purpose of an
underlying criminal investigation, the Executive Branch must
have the authority to decide that further prosecution is not in
the interest of justice.2 As the Supreme Court has explained,
“the capacity of prosecutorial discretion to provide
individualized justice is firmly entrenched in American law. …

2
  See U.S. Dep’t of Justice, Justice Manual § 9-27.200 cmt. (2020)
(“[A]s a matter of fundamental fairness and in the interest of the
efficient administration of justice, no prosecution should be initiated
against any person unless the attorney for the government believes
that the admissible evidence is sufficient to obtain and sustain a
guilty verdict by an unbiased trier of fact.”).
                                11

[A] system that did not allow for discretionary acts of leniency
would be totally alien to our notions of criminal justice.”
McCleskey v. Kemp, 

, 311–12 (1987) (quotation
marks omitted); see also United States v. Smith, 

,
160 (4th Cir. 1995) (“[T]he duty of the United States Attorney
[is] not simply to prosecute but to do justice.”) (quotation
marks omitted). In the third branch, when a district court
oversteps, the mandamus remedy allows the court of appeals to
prevent encroachment on a coequal department. See 

(“Accepted mandamus standards are broad
enough to allow a court of appeals to prevent a lower court
from interfering with a coequal branch’s ability to discharge its
constitutional responsibilities.”).

     Because this is not the unusual case where a more
searching inquiry is justified, and because there is no adequate
remedy for the intrusion on “the Executive’s long-settled
primacy over charging decisions,” Fokker 

, we grant the petition for mandamus in part and order the
district court to grant the government’s Rule 48(a) motion to
dismiss the charges against Flynn.

     We deny Flynn’s petition to the extent he seeks
reassignment of the district judge. This case does not meet the
“high bar” for reassignment,

which would be
appropriate only if the district judge’s conduct was “so extreme
as to display clear inability to render fair judgment,” Liteky v.
United States, 

, 551 (1994). Flynn focuses
primarily on comments the district judge made at sentencing,
but “judicial remarks during the course of a trial that are critical
or disapproving of, or even hostile to, counsel, the parties, or
their cases, ordinarily do not support a bias or partiality
challenge.”

see also In re Barry, 

, 914
(D.C. Cir. 1991) (“[A] trial judge is entitled to form his own
                                12

judgment as to the conduct of a defendant and to take that
judgment into account in sentencing.”). In light of these
precedents, the district judge’s conduct did not indicate a clear
inability to decide this case fairly. We decline to reassign the
case to a new judge simply to grant the government’s Rule
48(a) motion to dismiss.

                              ***

    We include the following responses to the dissenting
opinion in order to clarify the extent to which the dissent’s
arguments are foreclosed by longstanding precedent and
fundamental separation of powers principles.

     First, the dissent glosses over the presumption of
regularity to which the Executive is entitled in the exercise of
its prosecutorial discretion. While the district court is not a
“rubber stamp” when considering whether to grant “leave of
court” under Rule 48, 

, any judicial
inquiry is strictly bounded by the presumption of regularity,
and the presumption is overcome only in extraordinary cases
and by “clear evidence to the contrary,” 

. The dissent fundamentally misstates our opinion by
insisting we hold “a district court may not even consider [a
Rule 48 motion] before giving its ‘leave.’” Dissenting Op. 3.
Perhaps that is to distract from the simple fact that neither the
dissent nor the district judge has identified a legitimate basis to
rebut the presumption here. Our precedents emphatically leave
prosecutorial charging decisions to the Executive Branch and
hold that a court may scrutinize a motion to dismiss only on the
extraordinary showing of harassment of the defendant or
malfeasance such as bribery—neither of which is manifest in
the record before the district court. 

6–7. The dissent
argues the presumption is overcome in this case because of the
                                13

government’s “complete reversal” in moving to dismiss
charges it previously supported. Dissenting Op. 17–18. Yet the
government necessarily reverses its position whenever it
moves to dismiss charges under Rule 48(a), and so the reversal
standing alone cannot provide the irregularity. Given the
absence of any legitimate basis to question the presumption of
regularity, there is no justification to appoint a private citizen
to oppose the government’s motion to dismiss Flynn’s
prosecution. See 

.3

     Second, the dissent undermines our recent decision in
Fokker Services by recasting its necessary and well-considered
reasoning as dicta. In that case, we relied on “settled
constitutional understandings” to determine that Rule 48(a)’s
“leave of court” requirement “confers no new power in the
courts to scrutinize and countermand the prosecution’s exercise
of its traditional authority over charging and enforcement
decisions.” Fokker 

, 743. This part of
the opinion is binding because the narrow interpretation of
Rule 48(a) was essential to our interpretation of the Speedy
Trial Act’s parallel “approval of the court” requirement.

see also Seminole Tribe of Fla. v. Florida, 

,
67 (1996) (explaining that courts are bound by “those portions
of the opinion necessary to th[e] result”). The dissent attempts
to limit the holding of Fokker Services to its “particular legal

3
  The dissent’s only remaining argument relating to the presumption
of regularity is that the government is purportedly misinterpreting
Section 1001’s materiality element. Dissenting Op. 18. The question
before us, however, is not whether the Executive’s legal theory is
ultimately correct, but whether the Executive has the constitutional
discretion to end this prosecution—which for the reasons already
discussed, it plainly does. In the initiation and cessation of a
prosecution, it is the Executive’s view of the law that matters, not
ours, because of its authority over criminal charging decisions.
                                 14

context,” Dissenting Op. 4, but the decision is fundamentally
about the scope of the Executive’s constitutional charging
authority. As such, it is directly controlling here.

     Third, the dissent asserts our adherence to Fokker Services
creates a circuit split. Dissenting Op. 4. Yet if a split exists, a
proposition we dispute, it would be a result of Fokker Services’
binding interpretation of Rule 48(a). Even on the dissent’s view
that we should limit our reading of the case, the dissent’s
proposed free-ranging public interest inquiry reaches beyond
anything in our precedents. The level of scrutiny “left open” by
Rule 48(a), Dissenting Op. 6, must be understood in light of the
respective constitutional powers assigned to the Judiciary and
the Executive, which suggests a natural limit on the appropriate
judicial inquiry. The district court has no mechanism by which
it can maintain a prosecution in the absence of the Executive
Branch moving forward. Therefore, while the district court
may check for regularity and ensure that a defendant is not
being harassed by a motion to dismiss, its decision cannot turn
on “what the judge independently consider[s] best in the public
interest.” 

. 4 A court should not

4
  The dissent maintains that the Supreme Court in Rinaldi and
Thompson v. United States conducted an “independent evaluation”
of the record to determine whether an unopposed Rule 48(a) motion
should be granted. Dissenting Op. 10 

; 

, 250 (1980)). Yet the references to an “independent
evaluation” reflect the unremarkable proposition that, unlike a
district court, a court of review like this court or the Supreme Court
is limited to reviewing the facts included in the record. See, e.g.,
United States v. Bell, 

, 225 (D.C. Cir. 2013) (“[O]n
appeal we are limited to the record in the district court.”); FED. R.
APP. P. 10. The Supreme Court’s mention of an “independent
evaluation” or “independent examination” does not support the
authority of a district court—which is ordinarily not limited to a
                                 15

second-guess the prosecutor’s evaluation of the evidence
except in the “blatant and extreme case” where “the dismissal
has no basis in fact” whatsoever.

622. The dissent
would sanction an expansive judicial inquiry under Rule 48(a)
that runs afoul of Supreme Court precedent, this court’s
precedent, and out of circuit precedent. See ICC v. Bhd. of
Locomotive Engineers, 

, 283 (1987) (“[I]t is
entirely clear that the refusal to prosecute cannot be the subject
of judicial review.”); 

; In re United

.5

     Fourth, although our dissenting colleague apparently
agrees there is no need to cite a case on all fours before granting
mandamus relief, he faults the majority for failing to identify a
case in which mandamus was granted in nearly identical
circumstances. Dissenting Op. 1, 7, 13. Fokker Services,
however, directly rejects such a requirement: “[W]e have never
required the existence of a prior opinion addressing the precise

closed record—under Rule 48(a) to develop its own record of the
prosecution’s charging decisions through the appointment of a
private amicus or otherwise.
5
  The dissent relies upon the Fifth Circuit’s “betrayal of the public
trust” language to authorize courts to impose their independent
conception of the public interest upon the Executive. Dissenting Op.
5 (quoting 

). As our sister circuits have
explained, such a “betrayal” arises only where there is clear evidence
of misconduct such as “bribery, animus towards the victim, or a
desire to attend a social event,” United States v. HSBC Bank USA,

, 141 (2d Cir. 2017), or a federal prosecutor “acting
alone rather than at the direction or with the approval of the Justice
Department,” In re United 

. See also 

–30 (noting “extremely limited circumstances” for
judicial oversight such as when a prosecutor accepts a bribe). Such
circumstances are not alleged here.
                               16

factual circumstances or statutory provision at issue in order to
find clear error justifying mandamus 

–
50. In any event, as explained above, Fokker Services directly
forecloses the district court’s actions in these circumstances.
Some questions regarding Rule 48(a) may be “unsettled,” as
the dissent maintains, but there is nothing unsettled about the
principle that a district court cannot second-guess prosecutorial
discretion where there has been no indication of irregularity.
The novelty of the district court’s usurpation of power
heightens rather than lessens the need for mandamus. Fokker

.

     Fifth, the dissent minimizes the import of the district
court’s orders, claiming that we are granting mandamus
“before the district court has acted.” Dissenting Op. 2
(quotation marks omitted). Yet the district court has acted here.
It has ordered briefing and scheduled a hearing in order to
provide a court-appointed amicus the opportunity “to present
arguments in opposition to the government’s Motion to
Dismiss.” Order Appointing Amicus Curiae, ECF No. 205, at
1. In other words, the court has appointed one private citizen to
argue that another citizen should be deprived of his liberty
regardless of whether the Executive Branch is willing to pursue
the charges. Although no decision has yet been made on the
motion to dismiss, the district court’s judicial supervision,

7–8, “threatens to chill law enforcement by
subjecting the prosecutor’s motives and decisionmaking to
outside inquiry.” Wayte v. United States, 

, 607–08
(1985). As explained above, we have held such “interference
with the internal deliberations” of the Executive Branch to be
a quintessential irreparable injury giving rise to mandamus.

–43.
                              17

     Sixth, the dissent maintains that mandamus must focus
exclusively on the petitioner’s harms, not the government’s,
because the government has not formally petitioned for
mandamus. Dissenting Op. 7–10 & n.2. Our court has squarely
rejected this argument. In Cobell, we granted mandamus on the
basis of irreparable injury to the Executive Branch in the
absence of a “separate petition for mandamus” filed by the
government. See 

n.*; see also Ex
parte Peru, 

, 587–88 (1943) (relying on harm to
the Executive, namely “embarrass[ing] the executive arm of
the Government in conducting foreign relations,” to grant a
mandamus petition to a foreign sovereign without requiring a
separate government petition). Here, we invited the
government to respond, and at argument the government
offered to file a petition at the court’s request. Oral Arg. Tr.
46–47. The Executive Branch is not just any “different party,”
Dissenting Op. 8 (emphasis omitted), but a coequal branch of
government responsible for prosecutorial decisions. See

(“As this case implicates the
separation of powers, the Court of Appeals must also ask, as
part of th[e mandamus] inquiry, whether the District Court’s
actions constituted an unwarranted impairment of another
branch in the performance of its constitutional duties.”).
Therefore, we must consider the irreparable harms to the
Executive in a case such as this, in which the government has
moved to dismiss. Cf. Doe v. Exxon Mobil Corp., 

,
356–57 (D.C. Cir. 2007) (noting it would be appropriate to
examine separation of powers harms to the Executive if it had
joined the petitioner’s mandamus request or “requested the
dismissal of the action” below).

    Seventh, the dissent swings for the fences—and misses—
by analogizing a Rule 48(a) motion to dismiss with a selective
prosecution claim. Of course, the Executive may not
                                18

“selectively prosecute[] someone based on impermissible
considerations.” In re Aiken Cty., 

, 264 n.7 (D.C.
Cir. 2013). But “the equal protection remedy is to dismiss the
prosecution, not to compel the Executive to bring another
prosecution.”

517 U.S. at 459, 463).
When the court scrutinizes a selective prosecution claim, it
exercises the core Article III power by protecting individuals
from improper and unconstitutional prosecutions. Young v.
United States ex rel. Vuitton et Fils S.A., 

, 816
(1987) (Scalia, J., concurring in the judgment) (“The judicial
power … includes the power to serve as a neutral adjudicator
in a criminal case.”). Yet unwarranted judicial scrutiny of a
prosecutor’s motion to dismiss places the court in an entirely
different position. Rather than allow the Executive Branch to
dismiss a problematic prosecution, the court assumes the role
of inquisitor, prolonging a prosecution deemed illegitimate by
the Executive. Judges assume that role in some countries, but
Article III gives no prosecutorial or inquisitorial power to
federal judges. See

power … does not include
the power to seek out law violators in order to punish them.”);
United States v. Sineneng-Smith, 

, 1579 (2020)
(“[C]ourts … do not, or should not, sally forth each day looking
for wrongs to right. They wait for cases to come to them.”)
(citation and quotation marks omitted). To suggest that judicial
dismissal of an improper prosecution is analogous to the
judicial continuation of an improper prosecution turns the
separation of powers on its head.

      Ultimately, the dissent fails to justify the district court’s
unprecedented intrusions on individual liberty and the
Executive’s charging authority. This is not a case about
whether “a district judge may even hold a hearing on a Rule
48(a) motion.” Dissenting Op. 11 (emphasis omitted). Rather,
it is about whether, after the government has explained why a
                                19

prosecution is no longer in the public interest, the district judge
may prolong the prosecution by appointing an amicus,
encouraging public participation, and probing the
government’s motives. On that, both the Constitution and cases
are clear: he may not.

                              ***

     For the foregoing reasons, we grant Flynn’s petition for a
writ of mandamus in part and direct the district court to grant
the government’s Rule 48(a) motion to dismiss. In light of that
grant, we vacate the district court’s order appointing an amicus
as moot. See Great W. Sugar Co. v. Nelson, 

, 94
(1979).

                                                      So ordered.
     WILKINS, Circuit Judge, dissenting in part: It is a great
irony that, in finding the District Court to have exceeded its
jurisdiction, this Court so grievously oversteps its own. This
appears to be the first time that we have issued a writ of
mandamus to compel a district court to rule in a particular
manner on a motion without first giving the lower court a
reasonable opportunity to issue its own ruling; the first time
any court has held that a district court must grant “leave of
court” pursuant to Federal Rule of Criminal Procedure 48(a)
without even holding a hearing on the merits of the motion; and
the first time we have issued the writ even though the petitioner
has an adequate alternative remedy, on the theory that another
party would not have had an adequate alternate remedy if it had
filed a petition as well. Any one of these is sufficient reason to
exercise our discretion to deny the petition; together, they
compel its rejection. I therefore respectfully dissent from the
majority’s grant of the writ.

                                I.

     Mandamus is a “drastic and extraordinary remedy,”
Cheney v. U.S. Dist. Court for D.C., 

, 380 (2004)
(quoting Ex parte Fahey, 

, 259–60 (1947)), and
its “three threshold requirements are jurisdictional,” such that
the absence of any one compels denial of the writ and dismissal
of the petition for want of jurisdiction, Am. Hosp. Ass’n v.
Burwell, 

, 189 (D.C. Cir. 2016); see also In re
Kellogg Brown & Root, Inc., 

, 760 (D.C. Cir.
2014) (“[T]hree conditions must be satisfied before a court
grants a writ of mandamus: (1) the mandamus petitioner must
have ‘no other adequate means to attain the relief he desires,’
(2) the mandamus petitioner must show that his right to the
issuance of the writ is ‘clear and indisputable,’ and (3) the
court, ‘in the exercise of its discretion, must be satisfied that
the writ is appropriate under the circumstances.’” (quoting

–81)). In issuing a writ of mandamus
compelling the District Court to immediately grant the
                                  2
Government’s motion to dismiss the information against
Flynn, the majority concludes that each of these prerequisites
is satisfied. The majority is in each respect mistaken.

                                 A.

     One of the most striking conclusions of the majority is
virtually buried: It finds the issuance of mandamus to be
appropriate under the circumstances, where those
circumstances include the absence of any district-court ruling
on the motion at issue. Our Court is “particularly disinclined”
to issue this extraordinary writ “before the district court has
acted.” Republic of Venezuela v. Philip Morris Inc., 

, 198 (D.C. Cir. 2002); see also United States v. Hubbard,

, 309 (D.C. Cir. 1980) (observing that mandamus
would be inappropriate prior to district-court action because
“[i]t is the trial court and not this court that should engage in
the initial consideration”).1 This “particular[]” reluctance is the
natural consequence of our considered recognition of our
proper role in the federal judicial system: “As an appellate
court, ‘we are a court of review, not of first view[.]’” Capitol
Servs. Mgmt., Inc. v. Vesta Corp., 

, 789 (D.C. Cir.
2019) (alteration in original) (quoting Cutter v. Wilkinson, 

, 718 n.7 (2005)). The majority, which cites to no case
in which we have granted mandamus so precipitously, is

1
  The majority argues that the district has “acted,” because by
appointing amicus and scheduling a hearing, it has threatened to
exercise “judicial supervision” over prosecutorial decisionmaking.
Majority Op. 16. By this logic, mandamus becomes an ordinary
means to pretermit district-court consideration and have this court
decide whether to grant or deny a motion to dismiss once the lower
court “acts” in a manner that threatens to scrutinize the prosecutor’s
discretion. Relying as it does on diluting mandamus and overriding
long-settled procedure, this reasoning collapses under its own
weight.
                               3
apparently willing to set aside this Court’s well-established and
well-founded concern for the maintenance of the ordinary
course in order to proceed in an unprecedented manner. But I
view our historical disinclination to act out of turn as a
compelling, independent ground for declining to mandate the
immediate grant of the Government’s Rule 48(a) motion.

                               B.

     In considering whether Flynn’s right to relief is “clear and
indisputable,” it serves to remember that the question at hand
is not whether or under what circumstances a district court may
deny a Rule 48(a) motion, but whether it may give
consideration to such a motion before ruling on it. It should
come as no surprise that, before today, neither we nor any other
Court of Appeals has ever read Rule 48(a)’s “leave of court”
provision to mean that a district court may not even consider
such a motion before giving its “leave.” Cf. United States v.
Ammidown, 

, 622 (D.C. Cir. 1973) (“[W]e do not
think Rule 48(a) intends the trial court to serve merely as a
rubber stamp for the prosecutor’s decision.”). In fact, some of
our case law clearly points in the opposite direction: “The
requirement of judicial approval entitles the judge to obtain and
evaluate the prosecutor’s reasons.”

     In contending that its trailblazing result is somehow
compelled by precedent, the majority transforms dicta into
dogma. “[A] statement not necessary to a court’s holding is
dictum,” In re Grand Jury Investigation, 

, 1053
(D.C. Cir. 2019), and “[d]icta is never binding on any court,”
Murray Energy Corp. v. EPA, 

, 627 (D.C. Cir.
2019) (citing Glus v. Brooklyn E. Dist. Terminal, 

,
235 (1959)). In United States v. Fokker Services B.V., 

(D.C. Cir. 2016), this Court addressed the district court’s
denial of a joint motion to exclude time under the Speedy Trial
                               4
Act, 18 U.S.C. § 3161 et seq., pursuant to a deferred
prosecution 

–38. The Court there
had no occasion to make any decision about Rule 48(a), yet the
majority invokes Fokker as “foreclos[ing] the district court’s
proposed scrutiny” of such a motion. Majority Op. 7. Nor are
we bound by Fokker’s sweeping elucidations on the scope of
executive power, issued as they were in Fokker’s particular
legal context. Cohens v. Virginia, 

, 399 (1821) (“It
is a maxim not to be disregarded, that general expressions, in
every opinion, are to be taken in connection with the case in
which those expressions are used. If they go beyond the case,
they may be respected, but ought not to control the judgment
in a subsequent suit when the very point is presented for
decision.”).

     In addition to being improper, the majority’s reliance on
the Rule 48(a) dicta from Fokker creates a split with our sister
Courts of Appeals. Fokker’s Rule 48(a) commentary fails to
take due cognizance of federal appellate authority establishing
that an important impetus behind the Supreme Court’s insertion
into Rule 48(a) of the “leave of court” requirement was the
protection of the public interest, not simply the prevention of
abuse of the defendant. See, e.g., In re Richards, 

,
786–87 (3d Cir. 2000) (“Rule 48(a) . . . also permits courts
faced with dismissal motions to consider the public interest in
the fair administration of criminal justice and the need to
preserve the integrity of the courts.”); United States v. Cowan,

, 509–13 (5th Cir. 1975) (concluding that Rule
48(a)’s “history . . . belies the notion that [the Rule’s] only
scope and purpose is the protection of the defendant. . . . [I]t
[is] manifestly clear that the Supreme Court intended to clothe
the federal courts with a discretion broad enough to protect the
public interest in the fair administration of criminal justice”).
In the same vein, numerous federal appellate courts have
recognized that a court in receipt of an unopposed Rule 48(a)
                               5
motion may consider the public interest in ruling thereon. See,
e.g., United States v. Romero, 

, 1251 (10th Cir.
2004) (“[A] court is generally required to grant a prosecutor’s
Rule 48(a) motion to dismiss unless dismissal is clearly
contrary to manifest public interest.” (quoting United States v.
Carrigan, 

, 1463 (10th Cir. 1985)); United States
v. Pimentel, 

, 1033 n.5 (2d Cir. 1991) (same);
United States v. Hamm, 

, 629 (5th Cir. 1981)
(“[E]ven when the defendant consents to the motion to dismiss,
the trial court, in extremely limited circumstances in
extraordinary cases, may deny the motion when the
prosecutor’s actions clearly indicate a ‘betrayal of the public
interest.’” (quoting 

)); 

(concluding it is “appropriate” for a trial judge to
consider the “protection of the public interest” “in considering
whether to deny approval [] to dismissals of cases”); see also
Rinaldi v. United States, 

, 31–32 (1977) (per
curiam) (in reviewing a district court’s denial of an unopposed
Rule 48(a) motion, “agree[ing] with the Solicitor General that
. . . no societal interest would be vindicated” by continuing the
prosecution (internal quotation marks omitted)); cf. Young v.
United States, 

, 259 (1942) (“The public interest
that a result be reached which promotes a well-ordered society
is foremost in every criminal proceeding. That interest is
entrusted to our consideration and protection as well as that of
the enforcing officers.”).

     Fokker’s oversight is understandable, since the analogy to
Rule 48(a) was raised and addressed by only one of the three
parties during briefing, such that the Court did not have the
benefit of adverse presentation of the issue before expounding
on it in dictum. But the majority has used Fokker functionally
to constrict the parameters of the District Court’s Rule 48(a)
inquiry to include only the prevention of prosecutorial
harassment of the defendant. No binding authority establishes
                                6
this, and indeed the Supreme Court itself recognizes appellate
authority to the contrary. See 

n.15. In
fact, the Rinaldi Court expressly reserved—found it
“unnecessary to decide”—whether a district court had
discretion “to deny a Government dismissal motion to which
the defendant has consented if the motion is prompted by
considerations clearly contrary to the public interest.”

as Fokker did in language the majority now quotes, see
Majority Op. 5, that “the Supreme Court has declined to
construe Rule 48(a)’s ‘leave of court’ requirement to confer
any substantial role for courts in the determination whether to
dismiss 

(citing 

n.15), is to intimate that the Supreme Court had occasion to
construe Rule 48(a) one way or another on this issue, which it
did not.

     The fact is that “[t]he Supreme Court has left open whether
the court can ever refuse leave to dismiss if the government’s
request to dismiss is consented to by the defendant.” 3B
CHARLES ALAN WRIGHT, ARTHUR R. MILLER, & PETER J.
HENNING, FEDERAL PRACTICE AND PROCEDURE § 802 (4th ed.
2013) (citing 

, and Watts v. United
States, 

(1975) (mem.)). This Court has twice
opined, in dicta, on the factors that might properly guide a
court’s consideration and disposition of such a motion. See

–46; 

–22.
While the majority is bound by neither of these decisions, its
election to treat Fokker as controlling is particularly unsettling
in view of its willful neglect of our Court’s countervailing
views expressed in Ammidown. There, the Court envisioned a
district court in receipt of an unopposed Rule 48(a) motion as
being “entitle[d] . . . to obtain and evaluate the prosecutor’s
reasons,” and not “serv[ing] merely as a rubber stamp for the
prosecutor’s 

. According to
Ammidown, “the judge should be satisfied that the agreement
                               7
adequately protects the public interest” and “may withhold
approval if he finds that the prosecutor” committed “such a
departure from sound prosecutorial principle as to [constitute]
an abuse of prosecutorial discretion.”

     None of the authority on which the majority relies supports
its novel premise that “further judicial inquiry”—by which it
appears to mean any judicial inquiry—is proper only in “rare”
or “extraordinary” cases. See Majority Op. 5–6. And although
it invokes a host of non-binding authority regarding
permissible grounds for denying a Rule 48(a) motion, see

14–15 & n.5, that question is simply not before us.

     In NetCoalition v. SEC, 

, 354 (D.C. Cir.
2013), we noted that “[m]andamus does not lie when our
precedent no longer, at least in part, binds,” and we have
declined to find “a ‘clear and indisputable’ right to mandamus
relief” where the parties present countervailing “substantial
argument[s]” on unsettled legal issues, In re Khadr, 

, 100 (D.C. Cir. 2016). The majority declares it to be “clear
and indisputable” that no consideration of the Rule 48(a)
motion may be had by the very court whose leave is required
to effectuate that motion, but the majority’s need to engage in
such contortions to reach that conclusion is reason enough to
deny the writ.

                              C.

     Next, the majority addresses whether “other adequate
means to attain the relief” exist, concluding that “the district
court’s actions will result in specific harms to the exercise of
the Executive Branch’s exclusive prosecutorial power.”
Majority Op. 8. Here, too, the majority errs, this time by
shifting its focus away from the adequate means by which the
party seeking mandamus may obtain relief and by erecting an
indefensibly high wall between the branches of government.
                                  8
     The majority states that “there must be ‘no other adequate
means to attain the relief.’” Majority Op. 5. But in Cheney v.
U.S. District Court for the District of Columbia, the source of
this proposition, the Supreme Court stated that “the party
seeking issuance of the writ [must] have no other adequate
means to attain the relief he 

(alteration in original) (emphasis added). Indeed, the Supreme
Court has consistently framed this element around the
petitioner, not around other parties who have not moved for
mandamus relief. See, e.g., Kerr v. U. S. Dist. Court for N.
Dist. of Cal., 

, 403 (1976) (“Among these
[conditions for issuance of mandamus] are that the party
seeking issuance of the writ have no other adequate means to
attain the relief he desires.” (emphasis added)). The majority
cites no cases in which a court has concluded that a petitioner
has satisfied her burden to show no other adequate means to
attain her sought-after relief based on the absence of alternative
avenues of relief for a different party that did not petition the
court for the writ. 2 Neither Flynn nor the Government cited

2
 Contrary to “squarely reject[ing]” the notion that a court must focus
on the adequate relief of the party seeking mandamus, Majority Op.
17, Cobell rejected the notion that the party seeking interlocutory
appeal and, in the alternative, mandamus relief, must file two
different petitions in order to abide by our procedural rules. See
Cobell v. Norton, 

, 1140 n.* (D.C. Cir. 2003) (citing
FED. R. APP. P. 21 and CIR. R. 21). In Ex parte Republic of Peru, in
which the Supreme Court considered a petition for mandamus sought
by Peru, the Supreme Court analyzed whether Peru had an adequate
remedy. It concluded that Peru, as “a friendly sovereign state,” was
entitled to have its claims “presented and settled in the course of the
conduct of foreign affairs by [the Executive Branch],” and that Peru
should not have to endure “the delay and inconvenience of a
prolonged litigation” when the relief it seeks should be resolved in a
non-judicial forum. 

, 586–87 (1943). Here, and unlike
                                   9
such a case. The inconvenient reality is that the petitioner—
Flynn—has an adequate means, via a traditional appeal, to
attain relief should the District Court deny the Government’s
Rule 48(a) motion. See In re al-Nashiri, 

, 78 (D.C.
Cir. 2015) (“Mandamus is inappropriate in the presence of an
obvious means of review: direct appeal from final judgment.”);

–81 (noting that the requirement of
absence of adequate alternative remedies is “designed to ensure
that the writ will not be used as a substitute for the regular
appeals process”). This fact alone defeats our jurisdiction and
requires the Court to dismiss Flynn’s petition for a writ of
mandamus.

   And even if it were proper for the Court to consider the
Government’s argument that it will suffer an irreparable injury,

in Ex parte Republic of Peru, the majority grants mandamus solely
on the basis of a non-petitioner’s interests.

     The majority goes on to say that we “must” consider the
irreparable harms of the non-moving party because that party filed
the Rule 48(a) motion that got this ball rolling in the district court,
but one need look no further than the majority’s immediately
succeeding citation to find contradictory authority. See Majority Op.
17 (citing Doe v. Exxon Mobile Corp., 

, 356–57 (D.C.
Cir. 2007)). “[W]e need not resolve the political question issue on
the merits at this time,” because “[n]one of the cases cited by our
[dissenting] colleague stand for the proposition that we should grant
a mandamus for which the executive has not prayed.” Exxon Mobile

; see also

(noting that, had the
Executive “requested dismissal of the action,” the separation-of-
powers issue “would have been before the district court” in the first
instance). And while the majority points out that the Government
offered at oral argument to file a mandamus petition, Majority Op.
17, the fact remains that no such petition has ever been filed, and thus
none is before us.
                               10
the majority overstates the separation-of-powers dilemma
presented by Rule 48(a) motions. It is an unremarkable
proposition that, pursuant to his constitutional duty to “take
Care that the Laws be faithfully executed,” U.S. CONST. art II,
§ 3, the Executive has “broad discretion” over prosecutorial
matters, and that a “presumption of regularity” attaches to the
exercise of that discretion, United States v. Armstrong, 

, 464 (1996) (citations omitted). It is (or should be)
equally uncontroversial that such discretion is not absolute and
that the presumption of regularity does not shield such
discretion from review. Indeed, the Supreme Court said as
much in Rinaldi, where the Court “[did] not presume[] bad
faith on the part of the Government at the time it sought leave
to dismiss the 

, yet nonetheless
conducted an “independent evaluation” of the record,

see also

to determine whether the unopposed Rule
48(a) motion should have been granted. See also Thompson v.
United States, 

, 250 (1980) (performing
“independent examination of the record” and remanding, rather
than granting, government’s motion to dismiss indictment).

     The same principle holds true in the selective-prosecution
context, where the district court may dismiss an indictment if
it finds “the decision to prosecute [was] ‘deliberately based
upon an unjustifiable standard such as race, religion, or other
arbitrary classification.’” Wayte v. United States, 

, 608 (1985) (quoting Bordenkircher v. Hayes, 

, 364 (1978)). While the Supreme Court has iterated some
separation-of-powers concerns around the conduct of
discovery pursuant to such a motion, see 

, it has never cited the separation of powers as prohibiting
the district court from holding a hearing on such a motion. To
the contrary, courts have allowed not just plain hearings (or
oral arguments) on such motions to dismiss, but evidentiary
hearings where the prosecutor is subject to questioning under
                               11
oath. See United States v. Falk, 

, 623 (7th Cir.
1973) (en banc) (remanding and reassigning case for a hearing
in which the defendant could question the Assistant United
States Attorney as to specific issues related to the defendant’s
motion to dismiss for selective prosecution); see also United
States v. Berrios, 

, 1209, 1212 (2d Cir. 1974)
(finding no abuse of discretion in the district court’s order that
the government turn over, for in camera review, a
memorandum sent by the prosecutor to the U.S. Department of
Justice recommending prosecution of the defendant).

     In other words, it is not inconsistent with the separation of
powers for a district court to conduct regular proceedings and
afford consideration to a motion, even if the eventual grant or
denial of the motion might intrude on the Executive’s exercise
of his prosecutorial discretion. Again, this is not a case where
we are being asked to decide whether the district judge may
call the prosecutor to the stand or whether a Rule 48(a) motion
may lead to an evidentiary hearing. This is a case about
whether a district judge may even hold a hearing on a Rule
48(a) motion. While the selective-prosecution context is
admittedly different than Rule 48(a), these cases nonetheless
contradict the majority’s conclusion that holding a hearing, in
and of itself, is a per se improper intrusion upon executive
power. If the presumption of regularity does not prevent
holding a hearing or independently examining prosecutorial
discretion in the selective-prosecution context, there is no good
reason why the presumption of regularity precludes a hearing
on a motion to dismiss under Rule 48(a), disallowing any
consideration whatsoever and forbidding the district court from
expressing its views on the record with respect to its previous
findings on materiality and guilt—all in the name of the
separation of powers.
                               12
     None of this is to say that a district court, in considering
an unopposed Rule 48(a) motion, necessarily has carte blanche
to do whatever it likes.         One can certainly imagine
circumstances in which a district court employed devices that
would be sufficiently intrusive into the Executive’s purview as
to violate the separation of powers and counsel in favor of
immediate relief from the related orders. See, e.g., Matter of
Commodity Futures Trading Comm’n, 

, 872, 874
(7th Cir. 2019) (granting mandamus to vacate order of district
court requiring executive personnel “to appear and reveal what
lies behind their published words”); 

,
391 (indicating mandamus might be appropriate where district
court approved discovery requests to the Executive that
“ask[ed] for everything under the sky”). An appellate court
faced with such conditions might conclude that “exceptional
circumstances amounting to a judicial ‘usurpation of power’”
existed, such that the “extraordinary remedy” of mandamus
might be justified. Will v. United States, 

, 95
(1967). But that case is not before us. Not only has the District
Court not issued subpoenas or ordered discovery, it has given
no indication of its intent to do so. Instead, it is the District
Court’s intention to consider the motion in the ordinary course,
with briefing and argument, and the majority now labels that a
“judicial usurpation of power.” See

concluding, the
majority has improperly siloed the different branches of
government in an effort to prevent the possibility of intrusive
inquiry that may or may not occur. To the extent that any of
the majority’s fears of intrusive inquiries are well-founded,
they could easily be resolved by providing instruction to the
District Court as to the metes and bounds of a permissible Rule
48(a) hearing, with this panel (if it wished) retaining
jurisdiction in the event those bounds were overstepped. Cf.

(“Against the possibility that, upon
remand, a difference might arise between the court and the
government regarding the [extent to which matters of
                               13
prosecutorial decisionmaking remain confidential], we retain
jurisdiction for the purpose of review of the court’s ruling.”).

     “[A] petitioner’s right to relief is ‘clear and indisputable’
where he or she can point to ‘cases in which a federal court has
held that’ relief is warranted ‘in a matter involving like issues
and comparable circumstances.’ Accordingly, we will deny
mandamus even if a petitioner’s argument, though ‘pack[ing]
substantial force,’ is not clearly mandated by statutory
authority or case law.” In re Al Baluchi, 

, 369
(D.C. Cir. 2020) (second alteration in original) (quoting Doe v.
Exxon Mobil Corp., 

, 355 (D.C. Cir. 2007), and In
re 

–100). No one contends that Flynn
must have a case “on all fours,” Majority Op. 15, but the
required showing—one of legal certainty—is a “high hurdle.”

. Said another way, “[l]egal
aporias are the antithesis of the ‘clear and indisputable’ right
needed for mandamus relief.” In re 

.
In issuing the writ compelling the District Court to grant the
pending motion without considering it, the majority shuts its
eyes to the unsettled state of the law on the relevant questions:
the import of Rule 48(a)’s “leave of court” provision, the size
and shape of a district court’s discretion in considering an
unopposed Rule 48(a) motion, and the interplay between the
Executive’s prosecutorial discretion and the Judiciary’s
adjudicative power in these circumstances. Flynn has adequate
means to attain the relief he seeks, and he has pointed to no
authority mandating his preferred outcome here. As such,
Flynn fails to carry his burden, and especially given that the
District Court has yet to rule on the motion to dismiss, the writ
should not issue to compel the District Court to grant the
motion.
                                 14
                                 II.

     The majority also concludes that the writ should issue to
compel the vacation of the District Court’s order appointing
amicus (1) to address whether Flynn should be held in criminal
contempt for perjury, and (2) to present arguments in
opposition to the Government’s otherwise-unopposed Rule
48(a) motion. In neither respect has Flynn carried his burden
to establish that his right to relief is “clear and indisputable.”

                                 A.

     “The power to punish for contempts is inherent in all
courts; its existence is essential . . . to the due administration of
justice.” Ex parte Robinson, 

, 510 (1873); accord
Michaelson v. United States, 

, 65 (1924) (referring
to this premise as “settled law”). Federal Rule of Criminal
Procedure 42 provides a procedure by which a district court
may appoint an attorney to prosecute contempt, should the
government decline to do so. FED. R. CRIM. P. 42(a)(2). This
Rule reflects the fact that “it is long settled that courts possess
inherent authority to initiate contempt proceedings for
disobedience to their orders, authority which necessarily
encompasses the ability to appoint a private attorney to
prosecute the contempt.” Young v. United States ex rel. Vuitton
et Fils S.A., 

, 793 (1987). “Moreover, a court has
the power to conduct an independent investigation in order to
determine whether it has been the victim of fraud.” Chambers
v. NASCO, Inc., 

, 44 (1991). Far from establishing
his clear and indisputable right to relief, neither Flynn, nor the
majority in his stead, engages this precedent or forwards any
legal arguments as to why a district court that may undeniably
appoint a private attorney to prosecute contempt lacks the
lesser power to appoint amicus to advise it regarding whether
it ought to do so. Nor does the majority explain why directing
                               15
the District Court to grant the motion to dismiss renders moot
the District Court’s appointment of amicus to advise it on the
legally separate issue of contempt.

                               B.

     Similarly, Flynn fails to establish that it is clear and
indisputable that the District Court erred in its appointment of
amicus to present arguments in opposition to the Government’s
Rule 48(a) motion. Flynn’s only legal argument on the topic is
addressed to the District Court’s Local Rules, the ambiguity of
which falls well short of bringing the issue beyond dispute.
(Notably, the majority’s solution to Flynn’s failure to carry his
burden on this point is to ignore the issue altogether. The
majority uses its passing conclusion that amicus participation
is improper as a basis for ordering the separate relief of
mandamus on the Rule 48(a) motion—and then, having done
so, it declares the amicus issue moot.)

     Local Civil Rule 1.1(a) provides that “[t]hese Rules govern
all proceedings in the United States District Court for the
District of Columbia,” and Local Civil Rule 7(o) contemplates
the submission of amicus briefs. But Local Civil Rule 7,
entitled “Motions,” is duplicated to a limited extent in Local
Criminal Rule 47, also entitled “Motions,” and the latter
includes no mention of amicus briefs. And, as Flynn argues,
the Supreme Court observed in Hollingsworth v. Perry, 

(2010), that local rules “have the force of law,”

(internal quotation marks omitted). Flynn’s contention
that the Local Rules should be read as prohibiting the
participation of amici in criminal cases is therefore a plausible
one, as the Civil Rules provide for amici while the Criminal
Rules do not.
                               16
     But this Court has never held that an arguable proposition
entitles a petitioner to the extraordinary writ of mandamus. See
In re Al 

(“[W]e will deny mandamus
even if a petitioner’s argument, though ‘pack[ing] substantial
force,’ is not clearly mandated by statutory authority or case
law.” (second alteration in original)). And relying entirely as
he does on his construction of the Local Rules, Flynn fails to
engage with the fact that, in the absence of countervailing
authority, courts have “inherent power to provide themselves
with appropriate instruments required for the performance of
their duties. This power includes authority to appoint persons
unconnected with the court to aid judges in the performance of
specific judicial duties, as they may arise in the progress of a
cause.” Ex parte Peterson, 

, 312 (1920) (citation
omitted). The character and confines of such inherent
authority, as concerns a district court’s appointment or
allowance of amici, are simply not well developed. Cf. United
States ex rel. Gudur v. Deloitte Consulting LLP, 

920, 927 (S.D. Tex. 2007) (observing that “[n]o statute, rule,
or controlling case defines a federal district court’s power to
grant or deny leave to file an amicus brief”); Jin v. Ministry of
State Sec., 

(D.D.C. 2008) (asserting, on
the basis of a number of district-court cases, that “[d]istrict
courts have inherent authority to appoint or deny amici[,]
which is derived from Rule 29 of the Federal Rules of
Appellate Procedure” (citation omitted)).
     Both this Court and the Supreme Court regularly permit
the participation of amici in the criminal context, however, and
there is no readily apparent reason why, in appropriate
circumstances, a district court might not exercise its inherent
power to do the same—especially in the absence of any
authority expressly prohibiting it. Flynn’s counsel even
conceded at oral argument that district courts have the authority
to accept amicus briefs in some criminal cases. Oral Arg. Tr.
19–21. To the extent the majority’s sub silencio holding on the
                               17
propriety of amicus participation rests on the absence of such
authority in this particular case, the distinction has no legal
basis, and certainly not a clear and indisputable one. And if the
majority’s unuttered reasoning were premised on the absence
of a case or controversy, its mandate that the District Court
grant the Government’s Rule 48(a) motion to dismiss the
information with prejudice would be wholly inappropriate.
See, e.g., Flynt v. Weinberger, 

, 135–36 (D.C. Cir.
1985) (per curiam); Murray v. Conseco, Inc., 

, 605
(7th Cir. 2006). What is indisputable is that adverse
presentation of the relevant issues aids courts in their
decisionmaking—indeed, this is one of the foundational
premises of our judicial system. United States v. Nobles, 

, 230 (1975). Yet the majority gives this bedrock
principle of our legal system no weight or consideration. As
such, I must dissent from the majority’s functional ruling that
the appointment of amicus violated a clear and indisputable
right held by Flynn.

                             ***

     The majority opinion effectively transforms the
presumption of regularity into an impenetrable shield. In 2017,
the then-Acting Attorney General told the Vice President that
Flynn’s false statements “posed a potential compromise
situation for Flynn” with the Russians, Gov’t Mot. Dismiss
Crim. Info. Ex. 3 at 8, No. 1:17-cr-232, ECF No. 198-4 (May
7, 2020), and just a few months ago, the prosecution said that
Flynn’s false statements to the FBI “went to the heart” of a
valid counterintelligence inquiry and “were absolutely
material,” Gov’t Surreply Mot. Compel Produc. Brady Mat. at
10–11, No. 1:17-cr-232, ECF No. 132 (Nov. 1, 2019). Now, in
a complete reversal, the Government says none of this is
true. Gov’t Mot. Dismiss Crim. Info. at 13–16, No. 1:17-cr-
232, ECF No. 198. The Government doubles down by
                               18
asserting in its motion to dismiss that Flynn’s statements could
not have been “material” within the meaning of 18 U.S.C.
§ 1001 because the FBI had no grounds for any “viable”
investigation of Flynn at the time he made those statements,

even though that contention appears squarely belied by
our precedent, see United States v. Moore, 

, 701
(D.C. Cir. 2010) (“We . . . hold[] a statement is material if it
has a natural tendency to influence, or is capable of influencing,
either a discrete decision or any other function of the agency to
which it was addressed.”) (emphasis added); United States v.
Hansen, 

, 949 (D.C. Cir. 1985) (Scalia, J.) (“A lie
influencing the possibility that an investigation might
commence stands in no better posture under § 1001 than a lie
distorting an investigation already in progress.”). This is no
mere about-face; it is more akin to turning around an aircraft
carrier.

      The Government asserted to us that it has no duty to inform
the court in a motion to dismiss pursuant to Rule 48(a) of all of
its reasons for seeking dismissal. Oral Arg. Tr. 33. Today the
majority declares that nevertheless—in spite of the
Government’s abrupt reversal on the facts and the law, and
although the Government declares itself entitled not to be
forthcoming with the District Court—these circumstances
merit no further examination to determine whether there may
be additional reasons for the prosecutor’s actions, and if so, if
any such reasons are impermissible. Under the majority’s
interpretation of Rule 48(a), so long as the defendant consents
to the dismissal, “leave of court” is a dead letter.

    The Government may be entitled to “leave of court” under
Rule 48(a) to dismiss the criminal information to which Flynn
pled guilty, but that is not for us, as a Court of Appeals, to
decide in the first instance. Rather, the District Court must be
given a reasonable opportunity to consider and hold a hearing
                            19
on the Government’s request to ensure that it is not clearly
contrary to the public interest. I therefore dissent.