J.G.G. v. Donald Trump (EN BANC ORDER)

United States Court of Appeals

FOR THE DISTRICT OF COLUMBIA CIRCUIT

____________

No. 25-5124 September Term, 2025

1:25-cv-00766-JEB

Filed On: November 14, 2025

J.G.G., et al.,

Appellees

v.

Donald J. Trump, in his official capacity as

President of the United States, et al.,

Appellants

BEFORE: Srinivasan, Chief Judge, and Henderson, Millett**, Pillard*, Wilkins*,

Katsas, Rao, Walker, Childs***, Pan***, and Garcia*, Circuit Judges

ORDER

Appellees’ petition for rehearing en banc and the response thereto were circulated to

the full court, and a vote was requested. Thereafter, a majority of the judges eligible to

participate did not vote in favor of the petition. Upon consideration of the foregoing, it is

ORDERED that the petition be denied.

Per Curiam

FOR THE COURT:

Clifton B. Cislak, Clerk

BY: /s/

Daniel J. Reidy

Deputy Clerk

* A statement by Circuit Judges Pillard, Wilkins, and Garcia respecting the denial of rehearing

en banc is attached.

** A statement by Circuit Judge Millett, dissenting from the denial of rehearing en banc is

attached.

*** A statement by Circuit Judge Pan, joined by Circuit Judge Childs, dissenting from the denial

of rehearing en banc is attached.

PILLARD, WILKINS, and GARCIA, Circuit Judges, respecting the

denial of rehearing en banc:

On Saturday, March 15, 2025, Executive Branch officials

were implementing a plan to make novel use of a long-dormant

wartime statute—the Alien Enemies Act of 1798, invoked only

three times in our nation’s history—to summarily remove

hundreds of people from immigration detention in the United

States to indefinite confinement in a prison in El Salvador. See

50 U.S.C. § 21; Invocation of the Alien Enemies Act Regarding

the Invasion of the United States by Tren de Aragua, 90 Fed.

Reg. 13,033 (Mar. 20, 2025). The plaintiffs hastened to

challenge their removal in district court. They sought a

temporary restraining order (TRO) to allow the court to make

a preliminary assessment of their claims.

The district court convened an emergency hearing. Acting

under extraordinary time pressure while planes were already

departing, the court issued the TRO. It directed the government

to “prevent the removal” of all noncitizens in U.S. custody

subject to the proclamation. Hr’g Tr. 42:16–21, J.G.G. v.

Trump, No. 25-cv-766 (JEB), Dkt. No. 20 (D.D.C. Mar. 16,

2025). The court made the TRO’s meaning clear. It instructed

government counsel that “any plane containing these folks that

is going to take off or is in the air needs to be returned to the

United States,” and that counsel must “inform your clients . . .

immediately.” Hr’g Tr. 43:11–19. The court promptly

memorialized its TRO in a minute order stating that, “[a]s

discussed in today’s hearing,” the government was enjoined

from removing class members for 14 days. 7:25 p.m. Minute

Order, J.G.G. v. Trump, No. 25-cv-766 (JEB) (D.D.C. Mar. 15,

2025). Yet government officials pressed ahead, flying the

plaintiffs to El Salvador and delivering them into Salvadoran

custody for indefinite imprisonment. See J.G.G. v. Trump, 778

F. Supp. 3d 24, 33–35 (D.D.C. 2025).

In the days and weeks that followed, the district court tried

to understand what had transpired over that weekend despite

2

the government’s repeated refusals, citing shifting reasons, to

answer basic questions about when the flights had taken off and

when the plaintiffs were handed over to Salvadoran custody.

Id. at 35–37. Based on the facts available to it, the district court

issued a decision showing “probable cause exists to find the

Government in criminal contempt.” Id. at 30. As the district

court put it, it “appears obvious” that the government had

“deliberately flouted” the TRO. Id. at 37. The court took no

action on that probable-cause determination beyond ordering

defendants to “file by April 23, 2025, declaration(s) identifying

the individual(s) who, with knowledge of the Court’s classwide

Temporary Restraining Order, made the decision not to halt the

transfer of class members out of U.S. custody on March 15 and

16, 2025.” Order, J.G.G. v. Trump, No. 25-cv-766 (JEB), Dkt.

No. 80 (D.D.C. Apr. 16, 2025). The court did not refer the

matter for prosecution or order any other action from the

defendants. The government has never asserted any immunity

or privilege to withhold the identities of the decision makers.

Instead of complying, the government immediately

appealed that facially non-final order and sought a stay pending

appeal. All three members of the original panel agreed that

such an interlocutory order is not subject to appellate review.

J.G.G. v. Trump, 147 F.4th 1044, 1045 (D.C. Cir. 2025) (per

curiam). Ordinarily, criminal contempt is only appealable

following conviction and sentence. See In re Cys, 362 A.2d

726, 728–29 (D.C. 1976); see also 3A Charles Alan Wright &

Arthur R. Miller, Federal Practice & Procedure—Criminal

§ 714 (4th ed. Sep. 2025 update) (explaining that “a criminal

contempt order is a final and appealable judgment, but the

defendant cannot appeal before the court imposes sentence”).

Here, the district court had not even decided whether to make

a referral for prosecution. A majority of the panel nonetheless

granted an administrative stay pending appeal and then, on

August 8, announced that “the government has satisfied the

stringent requirements for a writ of mandamus” and issued the

3

writ to vacate the district court’s order. J.G.G., 147 F.4th at

1045. The government has yet to provide the requested

information.

The facts the district court recounted present grave rule-

of-law concerns. Obedience to court orders is vital to the

ability of the judiciary to fulfill its constitutionally appointed

role. Judicial orders are not suggestions; they are binding

commands that the Executive Branch, no less than any other

party, must obey. See Walker v. City of Birmingham, 388 U.S.

307, 314, 320–21 (1967). Disagreements with judicial

decisions must be resolved through motions, stays, and

appeals, not through unilateral noncompliance. The power to

hold a party in contempt of court enforces that principle and is

accordingly foundational to the rule of law and an essential

safeguard against defiance and lawlessness. See, e.g.,

Michaelson v. United States, 266 U.S. 42, 65–66 (1924).

The district court’s order here was a measured and

essential response to what it reasonably perceived as shocking

Executive Branch conduct. The first step to judicial exercise

of contempt authority may require factual inquiry regarding the

seeming defiance, including finding out who was

responsible—precisely what the district court sought here.

A midstream intervention by an appellate court into a

district court’s ongoing effort to identify potential contemnors

must surely be reserved for truly exceptional circumstances.

We doubt this order presented such circumstances and met the

demanding standard for mandamus relief.

That said, any errors in the panel’s analysis do not warrant

the further step of en banc review. Despite the seriousness of

the underlying facts, the product of the panel’s intervention—

the actual order it has issued—has no further practical effect,

nor does it establish any binding precedent.

4

As for practical effect, although Judge Katsas would have

barred the contempt proceeding entirely, Judge Rao ruled only

that the district court could not allow the defendants to avoid

contempt proceedings by belatedly complying with the TRO

they appear to have violated. That was an error, in Judge Rao’s

opinion, because that TRO had been vacated for want of venue

by the time of the contempt proceeding. J.G.G., 147 F.4th at

1064–68 (Rao, J., concurring); see Trump v. J.G.G., 604 U.S.

670, 671–72 (2025). The contempt-avoidance option in the

district court’s order no longer has real-world impact given

ensuing events: The affected individuals were released from

the Salvadoran prison and transferred to Venezuela, and their

ability to challenge their removal is now being litigated

separately. See J.G.G. v. Trump, 2025 WL 2317650, at *1, *3

(D.C. Cir. Aug. 8, 2025). The petition for rehearing en banc

itself acknowledges that this aspect of the district court’s order

has been “overtaken by events” and is no longer of any

practical significance. Pet. for Reh’g En Banc 14; see id. at

14–16. Indeed, all of this may have rendered moot any

perceived defect in the contempt-avoidance option, further

undercutting the case for the full court to decide whether to set

aside the panel disposition.

In the end, two judges on the panel voted to deny the

primary relief the government sought—to bar the criminal

contempt proceeding. Those judges agreed that the district

court retains discretion to proceed to consider whether the facts

support a criminal contempt referral, including by requiring the

government to identify the potential contemnors. Judge Rao,

notably, declined to “grant the government’s request to

terminate the criminal contempt proceedings.” J.G.G., 147

F.4th at 1072 (Rao, J., concurring). Her opinion expressly

acknowledges that “[w]hether to proceed with criminal

contempt is a choice left to the district court.” Id. at 1073. And

Judge Pillard sees no basis to have intervened in the first place.

See id. at 1098 (Pillard, J., dissenting). To dispel any doubt,

5

we observe that no member of this court has taken the position

that the panel’s disposition stands in the way of the district

court proceeding just as it intended to do in April, only without

the voluntary contempt-avoidance option that has now been

overtaken by events.

Nor will the panel order have precedential effect on any

other case. The order granting the writ of mandamus contains

no substantive reasoning of any kind as to when mandamus

relief is warranted in general, or why mandamus relief was

warranted in this case. See J.G.G., 147 F.4th at 1044–45 (per

curiam). So—as before—mandamus in this circuit remains a

“drastic” remedy, available only in “extraordinary situations.”

Illinois v. Ferriero, 60 F.4th 704, 714 (D.C. Cir. 2023) (citation

omitted). The party seeking the writ must establish a “clear

and indisputable” right to relief, “no other adequate means” to

obtain it, and that issuance is “appropriate under the

circumstances.” In re Cheney, 544 F.3d 311, 312–13 (D.C. Cir.

2008). And the grant of mandamus in this case cannot serve as

a guidepost for future panels.

Neither of the separate concurrences has precedential

effect either. Those opinions have no meaningful points of

agreement. Judge Rao, as noted, objected to a purportedly

improper commingling of civil and criminal contempt in the

unique circumstances presented here. J.G.G., 147 F.4th at

1064–68 (Rao, J., concurring). Judge Katsas focused on a

perceived ambiguity in the district court’s TRO itself. See id.

at 1051–58 (Katsas, J., concurring). The Supreme Court has

instructed that, when its own decisions lack a majority

rationale, the controlling opinion is the position taken by “those

Members who concurred in the judgments on the narrowest

grounds.” Marks v. United States, 430 U.S. 188, 193 (1977)

(citation omitted). Even if one were to apply that approach to

fragmented circuit panel decisions—which is an open question

in this court—it governs only “when one opinion is a logical

6

subset of other, broader opinions.” King v. Palmer, 950 F.2d

771, 781 (D.C. Cir. 1991). Here, neither concurrence is a

logical subset of the other—they address entirely different

putative defects in the district court’s order. Neither

concurrence, therefore, can meaningfully inform the resolution

of any future case. This, too, is not disputed by any member of

this court.

Those considerations counsel against en banc review.

This court reserves en banc proceedings for questions of

exceptional importance, or for orders that establish precedent

in conflict with decisions of this court or the Supreme Court.

Fed. R. App. P. 40(b)(2).

These proceedings unquestionably involve at least two

exceptionally important principles: the foundational

importance of a district court’s ability to enforce its orders and

to consider criminal sanctions for willful disobedience, and the

need for appellate courts to avoid premature intervention into

district court proceedings absent the most exceptional

circumstances.

But the peculiar circumstances of this case make en banc

review an unnecessary and inapposite vehicle to vindicate

either principle. As to the first, as just explained, the panel’s

order does not prevent the district court from exercising its

contempt authority. The district court remains free to require

the government to identify the decision makers who directed

the potentially contemptuous actions and to carefully consider

next steps. As to the second, en banc review could not undo

the time that our court has devoted to this case at the expense

of the district court’s ability to proceed; further review could

only prolong that delay. And because the panel’s order has no

precedential effect, en banc review is unnecessary to clarify the

law and reaffirm the demanding requirements for issuing a writ

of mandamus.

7

In short, despite the gravity of the underlying facts and the

understandable concern about the government’s compliance

with court orders, the writ of mandamus issued by the panel has

no ongoing practical or precedential effect.

The vote not to review the panel’s order en banc should

not be misinterpreted. We fully appreciate the seriousness of

the governmental conduct the district court observed, and we

do not endorse the panel’s premature intervention in the district

court’s decision-making process. Nor should this vote be

understood as judicial acquiescence to any disregard for court

orders. It most assuredly should not be read to diminish the

judiciary’s proper use of its contempt power or to excuse non-

compliance by government officials. To the contrary, respect

for judicial orders, and courts’ independence, are essential to

the rule of law. The district court has every ability to set a new

deadline for production of the information it sought regarding

the probable contempt it identified or to proceed otherwise

within his sound discretion. For its part, the government will

have the full opportunity at the appropriate time to raise any

defenses it may have. If it is dissatisfied with any appealable

order the district court may enter, it will have the opportunity

to seek review through the ordinary process. That is how our

system of justice is designed to work.

MILLETT, Circuit Judge, dissenting from the denial of

rehearing en banc:

I dissent from the denial of rehearing en banc. The

unreasoned per curiam order granting the writ and the

splintered opinions concurring in its issuance contravene both

(1) settled precedent requiring an exceptional and demanding

justification for mandamus relief, Cheney v. United States

District Court for D.C., 542 U.S. 367, 380 (2004), and (2) our

en banc decision in In re Flynn, 973 F.3d 74 (D.C. Cir. 2020),

which ruled that it is improper to use mandamus to cut off

lawful district court proceedings before that court has a chance

to rule. That combined precedent should apply with its fullest

force when, as here, the district court exercises its most

fundamental and inherent authority to obtain non-privileged

information necessary to police the integrity of proceedings in

its own courtroom.

PAN, Circuit Judge, joined by CHILDS, Circuit Judge,

dissenting from the denial of rehearing en banc:

In our constitutional system of checks and balances, it is

the duty of the Judicial Branch to “say what the law is,” and to

restrain the Executive and Legislative Branches when they

exceed the bounds of their authority. Marbury v. Madison, 5

U.S. (1 Cranch) 137, 177 (1803). The Framers designed the

tripartite structure of our government to ensure that no single

branch could accumulate excessive power and thereby threaten

the liberty that they fought a war to secure. The case before

this court implicates those foundational principles of our

democracy. When the political branches cease to honor the

judgments and decisions of the judiciary, the carefully

calibrated structure of our government collapses, and we are no

longer a society that is protected by the rule of law.

In this case, the district court was called upon to assess the

legality of an action taken by the Executive Branch. To

maintain the status quo while it performed its constitutional

function, the district court ordered the government to

temporarily pause the allegedly unlawful conduct. But the

government apparently defied the court’s order. Faced with a

serious transgression that went to the heart of the judicial

process, the district court properly proceeded to investigate

whether the government had committed a contempt of court

and who should be held accountable for that violation if it

occurred. Although the district court’s proceedings were

ongoing, and it therefore had not yet done anything that was

appealable, the government filed an obviously meritless appeal

and asked this court to stop the district court in its tracks. The

government also asked for extraordinary relief, in the form of

a writ of mandamus, asserting that it had a clear and

indisputable right to relief and that the district court should be

compelled to terminate the contempt proceedings that were

underway.

2

A panel of this court improvidently intervened. The panel

majority immediately halted the district court proceedings by

entering an administrative stay. Almost four months later, the

panel majority granted the government’s request for a writ of

mandamus and vacated the district court’s order that required

the government to provide information about its apparent

contempt of court.

The case now comes before the full appellate court on a

call for rehearing en banc. The question before us is whether

the full court should review and vacate the panel’s order

granting the government’s petition for a writ of mandamus. As

relevant here, we must decide whether this case raises a

question of exceptional importance, such that we should

correct any error by the panel. Fed. R. App. P. 40(b)(2)(D).

Applying that standard, a majority of the court votes to deny en

banc review, even though a separate majority agrees that the

writ of mandamus was issued in error.1

Some of my colleagues who vote against en banc review

disavow the issuance of the writ: They “doubt” that the panel’s

order “met the demanding standard for mandamus relief.”

Statement of Judges Pillard, Wilkins, & Garcia at 3. But they

1

Six of the eleven active judges on this court opine that the

district court did not err. See Statement of Judge Millett at 1

(“[H]ere, the district court exercise[d] its most fundamental and

inherent authority to obtain non-privileged information necessary to

police the integrity of proceedings in its own courtroom.”);

Statement of Judges Pillard, Wilkins, & Garcia at 3 (“The district

court’s order here was a measured and essential response to what it

reasonably perceived as shocking Executive Branch conduct.”);

Statement of Judge Pan & Judge Childs, infra at 11–12 (“We should

vacate the erroneous mandamus order to make clear that the district

court acted properly when it investigated the government’s apparent

violation of a court order, and to communicate that we unequivocally

reject the government’s unsupported claim of judicial overreach.”).

3

reason that the standard for en banc review is not met because

they believe that the mandamus order has no “practical effect.”

Id. at 3–4. They opine that the mandamus order sets no

precedent and does not impede the district court from issuing a

new order to reinitiate its contempt proceedings. See id. at 4–

5. They decline to grant en banc review just to vacate an order

that they think has no impact on the logistics of the case before

us, or on future cases. And yet, they emphasize that they fully

appreciate the seriousness of the government’s conduct, that

they do not endorse the panel’s premature interjection into the

district court’s decision-making process, and that their votes

should not be understood to either acquiesce in the disregard of

court orders or diminish the judiciary’s proper use of its

contempt power. See id. at 7.

In my view, those colleagues acknowledge but do not

properly address the context in which this case arose and the

overarching constitutional principles at stake. While my

colleagues narrow the aperture through which we view this

case and see it as a mistake that has no real-world

consequences, I take a broader view of “practical effects.” As

I see it, an attempt by the government to prevent the district

court from performing its constitutional function has a real and

detrimental impact on the ability of the judiciary to uphold the

rule of law. The mandamus order undercuts the district court’s

legitimate efforts to investigate an apparent contempt of court,

and we should vacate it. I also believe that the panel’s

erroneous order dilutes our rigorous standard for granting

mandamus relief and implies, through its sheer novelty, that the

government is not subject to the same rules and standards as

other litigants who seek justice before courts of law. I therefore

believe that this case raises multiple issues of exceptional

importance. I would grant en banc review and vacate the

panel’s mandamus order.

4

I.

On March 15, 2025, the government relied on a new

interpretation of the Alien Enemies Act of 1798 (AEA) to

commence deporting Venezuelan nationals based on their

alleged membership in Tren de Aragua, a Venezuelan

transnational gang. J.G.G. v. Trump, 778 F. Supp. 3d 24, 31

(D.D.C. 2025). Five individuals taken to the airport for

deportation filed suit to challenge the government’s novel legal

theory. Id. at 32. At an emergency hearing, the district court

provisionally certified a class of detainees subject to removal

under the AEA and entered a temporary restraining order

(TRO) to preserve the status quo while it considered the

legality of the deportations. Id. at 34. The TRO required the

government to refrain from deporting any class members for

the time being. Id. The court orally instructed the government

that if class members were on a plane “that is going to take off

or is in the air, . . . those people need to be returned to the United

States. However that’s accomplished, whether turning around

a plane or not [dis]embarking anyone on the plane . . . , I leave

to you. But this is something that you need to make sure is

complied with immediately.” Hr’g Tr. 43:13–19, J.G.G., 778

F. Supp. 3d 24 (No. 25-cv-766), ECF No. 20. Shortly after the

hearing, the court entered a minute order: “As discussed in

today’s hearing, . . . [t]he Government is ENJOINED from

removing [class members] . . . for 14 days.” Mar. 15, 2025,

7:25 PM Min. Order, J.G.G., 778 F. Supp. 3d 24 (No. 25-cv-

766).

During and after the emergency hearing, the government

transported detainees to El Salvador, apparently in violation of

the TRO. Public “boasts” by government officials “intimated

that they had defied the Court’s Order deliberately and

gleefully.” J.G.G., 778 F. Supp. 3d at 35 (citing the Secretary

of State’s retweet of a post stating that return of removed class

members was “Oopsie . . . Too late .”). Moreover, in

5

response to the TRO, the President of the United States

publicly called for the impeachment of the district court judge.

Donald J. Trump (@realDonaldTrump), Truth Social (Mar. 18,

2025, at 9:05 ET), https://perma.cc/XZ43-TTFQ.

Recognizing the gravity of the government’s apparent

violation of the TRO, the district court commenced

proceedings to determine what had happened. The district

court gave the government multiple opportunities to explain or

to justify its conduct, in written filings or at court hearings. But

the government sought to evade the court’s inquiries by

attempting to cancel hearings, refusing to answer basic

questions, and claiming that providing information about the

deportations would “jeopardize state secrets.” J.G.G., 778 F.

Supp. 3d at 35–36. The government also offered “imaginative”

and “hyper-technical” arguments for why it had complied with

the district court’s TRO. Id. at 35, 38. The district court

ultimately issued a memorandum opinion finding probable

cause that the government had committed criminal contempt

by violating the TRO, based on a well-supported finding that

the government’s conduct “manifest[ed] a willful disregard of

the Court’s legally binding proscriptions.” Id. at 38.

Contemporaneously with the probable-cause finding, the

district court issued an order requiring the government to

“identify[] the individual(s) who, with knowledge of the

Court’s classwide Temporary Restraining Order, made the

decision not to halt the transfer of class members out of U.S.

custody on March 15 and 16, 2025.” Order, J.G.G., 778 F.

Supp. 3d 24 (No. 25-cv-766), ECF No. 80.

Although the district court proceedings were ongoing —

and only final orders of the district court are appealable — the

government filed an appeal to stop the district court from

getting to the bottom of its contempt of court. The government

also petitioned this court for a writ of mandamus, claiming that

6

it had a clear and indisputable right to have the contempt

proceedings terminated.

A writ of mandamus is a “drastic” remedy “reserved for

really extraordinary causes.” Cheney v. U.S. Dist. Ct. for D.C.,

542 U.S. 367, 380 (2004) (quoting Ex parte Fahey, 332 U.S.

258, 259–60 (1947)). It is an order that “compel[s] the

performance of a clear nondiscretionary duty,” Pittston Coal

Grp. v. Sebben, 488 U.S. 105, 121 (1988) (cleaned up), or

remedies “a clear abuse of discretion,” Cheney, 542 U.S. at 380

(cleaned up). The writ is available only when (1) the petitioner

lacks “other adequate means to attain the relief he desires”; (2)

the petitioner’s “right to issuance of the writ is clear and

indisputable”; and (3) “the issuing court, in the exercise of its

discretion, [is] satisfied that the writ is appropriate under the

circumstances.” Id. at 380–81 (cleaned up).

After the government filed its appeal and mandamus

petition, a divided panel of this court entered an administrative

stay that temporarily halted the district court proceedings.

Administrative stays are not adjudications on the merits.

Because they are not based on any legal analysis, they are

intended to be brief and merely to give courts time to consider

emergency petitions — such as a motion for a stay pending

appeal, which requires satisfaction of a demanding four-part

test. See D.C. Circuit Handbook of Practice and Internal

Procedures 33 (2025); see also United States v. Texas, 144 S.

Ct. 797, 799 (2024) (mem.) (Barrett, J., concurring in denial of

applications to vacate stay) (describing an administrative stay

as “a short-lived prelude to the main event . . . [that] should last

no longer than necessary to make an intelligent decision on the

motion for a stay pending appeal”). The administrative stay in

this case remained in place for almost four months.

On August 8, 2025, the panel vacated the administrative

stay and issued its ruling. The panel unanimously agreed that

7

it lacked jurisdiction to hear the government’s appeal of the

district court’s investigative order and dismissed the appeal.

J.G.G. v. Trump, 147 F.4th 1044, 1045 (D.C. Cir. 2025). But

two of the panel members voted to grant the government’s

petition for a writ of mandamus, under unusual circumstances:

Those two judges relied on completely different grounds to

find a “clear and indisputable” right to relief. One panel

member focused on the facts, reasoning that the district court’s

TRO prohibiting the removal of class members was so plainly

ambiguous that the government could not have willfully

violated it. Id. at 1051–58 (Katsas, J., concurring). That judge

would have terminated the criminal-contempt proceedings

entirely. The second panel member, relying on a new legal

theory that had not been pressed by the government, posited

that the district court erred by giving the government the option

of “purging” the contempt — i.e., allowing the government to

avoid further contempt proceedings by providing habeas

hearings to class members who had been wrongfully deported.

Id. at 1064–70 (Rao, J., concurring); see also J.G.G., 778 F.

Supp. 3d at 54.

The third judge on the panel dissented from granting the

mandamus petition. She strongly disagreed with the factual

and legal analyses of her colleagues and argued that neither of

their alternative theories was correct, much less indisputably

so. J.G.G., 147 F.4th at 1076–77, 1086–98 (Pillard, J.,

dissenting). She also argued that mandamus — which is

premised on the violation of an “indisputable” right — was

inappropriate where the court could not agree on a single

rationale for granting the writ. Id. at 1074. Ultimately, the

panel majority’s mandamus order vacated the district court’s

investigative order, but it did not terminate the criminal-

contempt proceedings or prevent the district court from

reissuing a similar order.

8

II.

Lawyers representing the wrongfully deported parties

petitioned for rehearing en banc, and a member of this court

sua sponte called for rehearing en banc. See D.C. Circuit

Handbook of Practice and Internal Procedures 59 (2025). En

banc review is reserved for decisions that conflict with a

decision of this court, the Supreme Court, or another court of

appeals, and for cases that raise questions of exceptional

importance. Fed. R. App. P. 40(b)(2).

In my view, this case warrants en banc review because it

implicates principles that are vital to maintaining the structure

of our government and our commitment to the rule of law. The

government’s apparent “willful disregard of . . . legally binding

proscriptions” was a blow to our constitutional system that the

district court was duty-bound to address. J.G.G., 778 F. Supp.

3d at 38. We should vacate the panel’s mandamus order

because it is exceptionally important (1) to correct a serious

error that dilutes our stringent standard for granting mandamus

relief; (2) to repudiate the implication that the government is

not subject to the same rules and standards as other litigants

who turn to the courts to seek relief; and (3) to dispel any

suggestion that the district court acted improperly when it took

decisive steps to enforce a court order.

I start with the simple fact that a majority of the en banc

court believes that the panel majority erred when it issued the

writ of mandamus. See supra n.1. Indeed, it is difficult to see

how the government had an “indisputable” right to relief when

the panel members themselves disputed the basis for granting

the writ, and neither member of the panel majority could point

to any cases in which mandamus was granted based on “like

issues and comparable circumstances.” J.G.G., 147 F.4th at

1074 (Pillard, J., dissenting) (cleaned up). Although en banc

review is not to be deployed for the correction of one-off errors,

9

the panel’s mandamus order sets a dubious precedent for

granting the extraordinary relief of mandamus when no panel

majority agrees on the basis for issuing the writ. With the

increased volume of emergency petitions that have been filed

in our court of late, it is possible that what happened in this case

will recur. Although some of my colleagues opine that the

mandamus order has no precedential value, I do not see why a

future panel could not rely on it to issue a writ of mandamus

despite the failure of the panel members to agree on a single

legal theory to justify that “drastic” action.

I do not question the good faith of my colleagues in the

panel majority, who each undoubtedly cast their vote based on

a sincere belief that the mandamus standard was met. To my

mind, the legal issue that warrants review is whether

mandamus may be granted when two judges essentially concur

in the judgment, with each believing in a different clear and

indisputable right to relief. The panel majority assumed that

such an outcome was permissible without explaining why. But

mandamus is a “drastic” remedy, available only in

“extraordinary situations,” and is “hardly ever granted.”

Illinois v. Ferriero, 60 F.4th 704, 714 (D.C. Cir. 2023) (cleaned

up). Meeting such an “exacting” standard surely must require

at least a majority of the panel to agree on what entitles the

petitioner to relief. The writ may be issued only if the court

determines that the petitioner has a “clear and indisputable”

entitlement to relief. Cheney, 542 U.S. at 381 (cleaned up).

Individual judges who express idiosyncratic opinions in

separate statements do not speak for the court. And, in my

view, they cannot combine non-overlapping theories to satisfy

a legal standard premised on the existence of a “clear and

indisputable right,” especially when the separate statements do

not rely on well-established principles of law. Cf. In re Al

Baluchi, 952 F.3d 363, 369 (D.C. Cir. 2020) (“[W]e will deny

mandamus [whenever] a petitioner’s argument . . . is not clearly

mandated by statutory authority or case law.”).

10

Our failure to vacate the mandamus order suggests that if

two judges rely on disparate, novel legal theories to vote in

favor of granting mandamus, those legal theories may be

insulated from scrutiny because they individually are

nonprecedential, even though their combined effect is to cause

the “drastic” consequence of issuing the writ. See Statement of

Judges Pillard, Wilkins, & Garcia at 5 (declining to vacate the

mandamus order in part because “[n]either of the separate

concurrences has precedential effect”). We should grant en

banc review and vacate the mandamus order to make clear that

our mandamus standard remains high and difficult to satisfy:

It cannot be met by employing the panel majority’s

unprecedented, splintered approach.

The mandamus order also creates the unintended but

unfortunate impression that the government is not subject to the

same rules and standards as other parties who come before our

court. It is beyond debate that we must apply the law

evenhandedly and consistently, regardless of the identity of the

parties that stand before us. Otherwise, we open ourselves to

charges of favoritism and bias, and we may erode public

confidence in the judicial system. See Glob. Health Council v.

Trump, 153 F.4th 1, 35–36 (D.C. Cir. 2025) (Pan, J.,

dissenting). Here, the court’s unjustified intervention in the

district court proceedings, culminating in the issuance of an

unusual split-decision mandamus order, may have been

perceived by the public as bending the rules to benefit one of

the parties to this case: the government. As perhaps the only

mandamus precedent of its kind — relying on two different

theories of purportedly “indisputable” entitlements to relief —

the mandamus order unintentionally suggests that the

government is entitled to special treatment in the form of a

bespoke legal standard. We should dispel that implication to

the extent that we can — by vacating the mandamus order.

11

Finally, and most importantly, the mandamus order casts

unjustified doubt on the propriety of the district court’s

contempt proceedings: It wrongfully holds that the district

court clearly abused its discretion and that the government had

a clear and indisputable right to be relieved from complying

with the district court’s investigative order. The granting of

mandamus signified that “drastic” relief was necessary because

the district court either failed to perform a “clear

nondiscretionary duty” or committed “a clear abuse of

discretion.” And yet, a majority of the en banc court seems to

believe that the mandamus order should not have issued at all.

See supra n.1.

We should not leave in place a ruling that says that the

district court erred when it did not. We live in a world in which

judges face threats and harassment because of their rulings and

have been called “rogue” by government officials who disagree

with them. And we cannot overlook that, in response to this

very case, the President of the United States called for the

district judge’s impeachment. Some lawyers and interested

observers might read the statements accompanying our denial

of en banc review and realize that a majority of the full court

does not believe that the district court was in the wrong. See

supra n.1. But the meaning of this vote is likely to be

misunderstood by members of the broader public who just read

headlines, or who choose to focus only on the bottom line —

they will see only that we let stand the mandamus order, which

took the drastic step of vacating the district court’s

investigative order, based on a determination that the district

court clearly abused its discretion. That misperception will

harm public confidence in the judiciary and in the

righteousness of the district court’s contempt proceedings.

We should vacate the erroneous mandamus order to make

clear that the district court acted properly when it investigated

the government’s apparent violation of a court order, and to

12

communicate that we unequivocally reject the government’s

unsupported claim of judicial overreach. The district court

performed its constitutional duty with unwavering integrity and

courage, in the face of undue public criticism from the most

powerful official in our nation. Although the government and

the panel majority have deviated from norms and applicable

legal standards in this case, the district court most certainly has

not. In my view, it is unacceptable to leave in place an

erroneous mandamus order that holds that the district court

exceeded its authority when it did nothing but its duty. The

mandamus order baselessly impugns the district court’s

handling of this case, and it should not stand.2

* * *

The context in which this case was brought and litigated

makes it exceptionally important. Our constitutional system

was functioning as designed until a panel of this court

improvidently intervened. The district court was called upon

to check an allegedly unlawful policy implemented by the

Executive Branch. When the government apparently defied a

court order, the district court properly investigated. In merely

seeking information about the government’s apparent

contempt of court, the district court did not violate any

nondiscretionary duty nor clearly abuse its discretion. The

panel’s erroneous issuance of a writ of mandamus vacating the

district court’s investigative order undercuts the district court’s

2

In a closing nod to “practical effects,” I note that we can grant

en banc review and vacate the mandamus order without additional

oral argument or briefing. Thus, granting en banc review would not

necessarily delay the proceedings in the district court any more than

what we are doing now. Either way, the process can be boiled down

to the writing of statements by members of the court who wish to

explain their votes.

13

efforts to hold the government accountable for a serious

transgression and to uphold the rule of law.

We must set the record straight: It is the panel majority

that erred by granting the writ of mandamus, and the district

court did nothing wrong. As a court of law, the coin of our

realm is reasoned decision-making, based on faithful

application of the law to the facts, without fear or favor. That

is what allows us to play our role within the constitutional

system that the Framers envisioned. The district court upheld

those principles, while the mandamus order undermines them.

I would grant the petition for rehearing en banc and vacate the

mandamus order.