FOR PUBLICATION
UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT
EAST BAY SANCTUARY COVENANT; Nos. 18-17274
AL OTRO LADO; INNOVATION LAW 18-17436
LAB; CENTRAL AMERICAN
RESOURCE CENTER, D.C. No.
Plaintiffs-Appellees, 4:18-cv-06810-
JST
v.
JOSEPH R. BIDEN, President of the ORDER AND
United States; MERRICK B. AMENDED
GARLAND, Attorney General; JEAN OPINION
KING, Acting Director, Executive
Office for Immigration Review
(EOIR); ALEJANDRO MAYORKAS,
Secretary, U.S. Department of
Homeland Security; TRACY RENAUD,
Senior Official Performing the
Duties of the Director, U.S.
Citizenship and Immigration
Services; TROY MILLER, Senior
Official Performing the Duties of the
Commissioner, U.S. Customs and
Border Protection; TAE D. JOHNSON,
Acting Director, U.S. Immigration
and Customs Enforcement,
Defendants-Appellants.
2 EAST BAY SANCTUARY COVENANT V. BIDEN
Appeal from the United States District Court
for the Northern District of California
Jon S. Tigar, District Judge, Presiding
Argued and Submitted October 1, 2019
San Francisco, California
Filed February 28, 2020
Amended March 24, 2021
Before: Ferdinand F. Fernandez, William A. Fletcher,
and Richard A. Paez, Circuit Judges.
Order;
Opinion by Judge Paez;
Concurrence by Judge Fernandez;
Concurrence in Denial of Rehearing En Banc by
Judge Paez;
Dissent from Denial of Rehearing En Banc by
Judge Bumatay;
Dissent from Denial of Rehearing En Banc by
Judge VanDyke
EAST BAY SANCTUARY COVENANT V. BIDEN 3
SUMMARY *
Immigration / Preliminary Injunctions
The panel filed: 1) an order denying on behalf of the
court a petition for rehearing en banc; 2) an amended opinion
affirming the district court’s grant of a temporary restraining
order and a subsequent grant of a preliminary injunction
enjoining enforcement of a rule and presidential
proclamation that, together, strip asylum eligibility from
every migrant who crosses into the United States along the
southern border of Mexico between designated ports of
entry; and 3) an amended concurrence.
Addressing briefing on the President’s revocation of the
proclamation at issue, the panel agreed with the parties that
this appeal was not moot, but declined to hold the case in
abeyance while the government reviews the interim final
rule at issue. The panel noted that the parties may address
further developments and whether any such developments
render the case moot on remand.
In the amended opinion, the panel explained that the
Department of Justice and Department of Homeland
Security adopted an interim final rule in November 2018
(“the Rule”) that makes migrants who enter the United States
in violation of a “presidential proclamation or other
presidential order suspending or limiting the entry of aliens
along the southern border with Mexico” categorically
ineligible for asylum. The same day, President Trump
issued a presidential proclamation (“the Proclamation”) that
*
This summary constitutes no part of the opinion of the court. It
has been prepared by court staff for the convenience of the reader.
4 EAST BAY SANCTUARY COVENANT V. BIDEN
suspended the entry of all migrants along the southern border
of the United States for ninety days, except for any migrant
who enters at a port of entry and properly presents for
inspection.
Legal services organizations representing asylum-
seekers (“the Organizations”) sued to prevent enforcement
of the Rule. The district court entered a temporary
restraining order enjoining the Rule, and the government
appealed, seeking a stay in this court of the district court’s
order pending appeal. In a published order, a motions panel
denied the stay, and the Supreme Court denied a stay as well.
The district court issued an injunction barring enforcement
of the Rule, the government appealed, and this court
consolidated the two appeals.
First, addressing the effect of the motions panel’s order
on the present panel’s decision, the panel concluded that a
published motions panel order may be binding as precedent
for other panels deciding the same issue, but it is not binding
here. The panel explained that this is because the issues are
different: in deciding whether to stay a preliminary
injunction pending appeal, the motions panel was predicting
the likelihood of success of the appeal, meaning it was
predicting rather than deciding what the merits panel will
decide; however, in resolving the merits of a preliminary
injunction appeal, the merits panel was deciding the
likelihood of success of the actual litigation. The panel
explained that such a predictive analysis should not, and
does not, forever decide the merits.
The panel noted that there may be circumstances where
a motions panel does answer the same legal question that is
presented to the merits panel, observing that this court
addressed such a circumstance in Lair v. Bullock, 697 F.3d
1200 (9th Cir. 2012), in which the merits panel held that it
EAST BAY SANCTUARY COVENANT V. BIDEN 5
was bound by the motions panel’s published decision on a
particular issue, where the motions panel answered precisely
the same question that was before the merits panel.
Accordingly, the panel noted that, to the extent the issues
share predictive similarity here, the motions panel may be
persuasive but not binding.
Next, the panel considered the government’s challenge
to the court’s jurisdiction. First, the panel concluded that the
Organizations had established organizational standing by
showing that the Rule perceptibly impaired their ability to
perform their services. Second, the panel rejected the
government’s argument that the court should avoid
interfering with the Rule on the ground that the power to
expel or exclude aliens is a fundamental sovereign attribute
exercised by the government’s political departments largely
immune from judicial control. The panel explained it was
responsible for reviewing whether the government has
overstepped its delegated authority under the Immigration
and Nationality Act (“INA”) and encroached upon
Congress’s legislative prerogative. Third, the panel rejected
the government’s argument that three statutory provisions,
8 U.S.C. §§ 1252(e)(3), 1252(a)(5), and 1252(b)(9), divested
this court of jurisdiction. The panel explained that none of
these provisions have any bearing on the Rule because they
govern judicial review of removal orders or challenges
inextricably linked with actions taken to remove migrants
from the country. Finally, the panel concluded that the
Organizations fell within the zones of interests of the INA.
The panel next addressed the Organizations’ likelihood
of success on the merits of their claims under the
Administrative Procedure Act (“APA”). Applying the
Chevron framework, the panel held that the Rule conflicts
with the INA’s section on asylum, which states that a
6 EAST BAY SANCTUARY COVENANT V. BIDEN
migrant may apply for asylum when she is “physically
present in the United States” or “arrives in the United States
(whether or not at a designated port of arrival…)[.]”
8 U.S.C. § 1158(a)(1). Because the Rule requires migrants
to enter at ports of entry to preserve their eligibility for
asylum, the panel explained that it is effectively a categorical
ban on migrants who use a method of entry explicitly
authorized by Congress in § 1158(a).
The panel further concluded that, even if the text of
§ 1158(a) were ambiguous, the Rule fails at the second step
of Chevron because it is an arbitrary and capricious
interpretation of the statute. The panel explained that the
BIA and this court have long recognized that a refugee’s
method of entry is a discretionary factor in determining
whether the migrant should be granted relief, but that the
method of entry should be carefully evaluated in light of the
harsh consequences that may result. Thus, the panel
concluded that, given the Rule’s effect of conditioning
asylum eligibility on a factor that has long been understood
as worth little if any weight in adjudicating asylum
applications, it is an arbitrary and capricious interpretation
of § 1158(a).
The panel also concluded that the Rule is unreasonable
in light of the United States’s treaty obligations under the
1951 United Nations Convention Relating to the Status of
Refugees (“1951 Convention”) and the 1967 United Nations
Protocol Relating to the Status of Refugees.
The panel briefly addressed the procedural arguments
raised by the parties regarding whether the Rule was invalid
because it was issued without public notice and comment or
complying with the thirty-day grace period required by the
APA. The panel concluded that the Rule likely does not
EAST BAY SANCTUARY COVENANT V. BIDEN 7
properly fall under the good-cause exception or the foreign-
affairs exception to these procedural requirements.
Next, the panel concluded that the Organizations had
demonstrated a sufficient likelihood of irreparable injury to
warrant injunctive relief, explaining that the Organizations
had shown that they will suffer a significant change in their
programs and a concomitant loss of funding absent a
preliminary injunction. The panel also concluded that the
public interest weighs sharply in the Organizations’ favor.
Finally, addressing the scope of the remedy, the panel
concluded that the district court did not abuse its discretion
in issuing an injunction preventing any action to implement
the Rule. The panel noted that the Organizations do not limit
their potential clients to refugees who enter only at the
Mexican border with California and Arizona, and that the
government had not proposed an alternative form of the
injunction that accounts for the scope of the harms, but
applies only within the Ninth Circuit.
Concurring in the result, Judge Fernandez wrote that he
concurred in the majority opinion because, and for the most
part only because, he believes that this panel is bound by the
motions panel’s published decision in this case. Judge
Fernandez wrote that the panel is bound by the law of the
circuit as well as the law of the case doctrine. Judge
Fernandez noted that the majority had now taken steps, in its
amended opinion, to obscure its attacks on the doctrines of
law of the case and law of the circuit, which were set forth
in its original opinion. Writing that his concurrence
nevertheless remained the same and was based on the same
reasoning, Judge Fernandez stated that the majority’s
attempt to pull an invisibility cloak over the mainspring of
its attacks cannot hide damage wrought by them.
8 EAST BAY SANCTUARY COVENANT V. BIDEN
Concurring in the denial of rehearing en banc, Judge
Paez, joined by Judge W.Fletcher, wrote to respond to
several of the arguments in Judge Bumatay’s and Judge
VanDyke’s dissents. Responding to Judge Bumatay’s
contention that the amended majority opinion’s application
of precedent relating to organizational standing deviated
from John Marshall’s and James Madison’s respective
visions of Article III, Judge Paez wrote that the majority
opinion’s organizational standing holding was consistent
with Supreme Court precedent, Article III, and decisions of
this and other courts.
Judge Paez also responded to Judge VanDyke’s attacks
on the integrity of the majority opinion. First, Judge Paez
wrote that Judge VanDyke accused the majority of
impropriety in addressing the effect of the motions panel’s
published opinion on the merits panel’s subsequent
consideration of the appeal. Noting that this potentially
dispositive issue was expressly raised by the parties’ briefs,
Judge Paez concluded that it was clearly appropriate for the
majority to address the issue. Second, Judge Paez wrote that
Judge VanDyke complained that the majority engaged in
“mischief” by amending its opinion during the en banc
process. Judge Paez described the court’s en banc process.
Judge Paez wrote that, consistent with this process, the panel
majority here considered the debate about law-of-the-case
where a motions panel has published an order, the panel
majority was persuaded to revise its opinion by the
discussions in the memorandum exchange, and the revisions,
which were proposed to the court during the memorandum
exchange period, were reflected in the amended majority
opinion.
Dissenting from the denial of rehearing en banc, Judge
Bumatay, joined by Judges Ikuta, Bennet, R. Nelson, Lee,
EAST BAY SANCTUARY COVENANT V. BIDEN 9
and VanDyke, wrote that judges are not “Platonic
Guardians” of the nation’s public policies; judges have no
business standing athwart the choices of the political
branches no matter how misguided the judges believe them
to be. That fundamental limitation, Judge Bumatay wrote, is
even more pronounced in the immigration context.
Judge Bumatay concluded that the majority opinion
ignored constitutional limits on jurisdiction by stretching
organizational standing doctrine beyond Article III’s reach,
inaugurating a new standard of organizational standing—
one that allows an organization to challenge a disfavored
government policy by merely asserting that the change could
result in “one less client” or cause it to shift resources to
better support its mission.
Judge Bumatay also wrote that the majority re-wrote the
asylum statute to add a prohibition on the Executive’s
authority not found anywhere in the legislative text. Judge
Bumatay concluded that the Rule easily fit within the asylum
scheme enacted by Congress, explaining that the Rule does
nothing to limit the ability of aliens to apply for asylum, as
it only restricts the parameters for a successful asylum
petition. Judge Bumatay wrote that the panel’s reading
essentially conflates applying for and receiving asylum, and
the result is that the court has substituted its version of the
asylum law for the one actually passed by Congress.
Dissenting from the denial of rehearing en banc, Judge
VanDyke wrote separately to emphasize it was never
necessary in this case for the panel majority to wade into the
issue of the binding effect of motions panels’ published
opinions and purport to overrule Lair v. Bullock. The panel
majority only did so because of Innovation Law Lab v. Wolf,
the other case it decided the same day it decided this
10 EAST BAY SANCTUARY COVENANT V. BIDEN
case. That was the panel majority’s “mischief”—not the
panel amending its opinion during the en banc process.
Further, Judge VanDyke wrote that while it is true that
the amended rationale in the East Bay majority’s opinion
does not represent the ham-fisted upheaval its original
opinion did, it merely trades one evil for another. Instead of
rudely casting aside precedent, it distinguishes—rather than
overrules—Lair, creating a new standardless standard that
will allow any merits panel to disregard a motions panel’s
published decision resolving virtually identical claims.
EAST BAY SANCTUARY COVENANT V. BIDEN 11
COUNSEL
Scott Grant Stewart (argued) and Sarah Harrington, Deputy
Assistant Attorneys General; Francesca Genova and T.
Benton York, Trial Attorneys; Erez Reuveni, Assistant
Director; William C. Peachey, Director; August E. Flentje,
Special Counsel; Joseph H. Hunt, Assistant Attorney
General; Office of Immigration Litigation, United States
Department of Justice, Washington, D.C.; for Defendants-
Appellants.
Lee P. Gelernt (argued), Judy Rabinovitz, Omar C. Jadwat,
Anand Balakrishnan, Daniel Galindo, and Celso Perez,
American Civil Liberties Union Foundation, Immigrants’
Rights Project, New York, New York; Katrine Eiland,
Jennifer Chang Newell, Cody Wofsy, and Spencer Amdur,
American Civil Liberties Union Foundation, Immigrants’
Rights Project, San Francisco, California; Melissa Crow,
Southern Poverty Law Center, Washington, D.C.; Mary
Bauer, Southern Poverty Law Center, Charlottesville,
Virginia; Gracie Willis, Southern Poverty Law Center,
Decatur, Georgia; Baher Azmy, Angelo Guisado, and Ghita
Schwarz, Center for Constitutional Rights, New York, New
York; Christine P. Sun and Vasudha Talla, American Civil
Liberties Union Foundation of Northern California Inc., San
Francisco, California; for Plaintiffs-Appellees.
Lawrence J. Joseph, Washington, D.C.; Christopher J.
Hajec, Director of Litigation, Immigration Reform Law
Institute, Washington, D.C.; for Amicus Curiae Immigration
Reform Law Institute.
Patrick W. Pearsall, Karthik P. Reddy, and Vaishalee V.
Yeldandi, Jenner & Block LLP, Washington, D.C.; Alice
Farmer, Office of the United Nations High Commissioner
12 EAST BAY SANCTUARY COVENANT V. BIDEN
for Refugees, Washington, D.C.; for Amicus Curiae Office
of the United Nations High Commissioner for Refugees.
Richard D. Bernstein, Washington, D.C.; Richard Mancino,
Willkie Farr & Gallagher LLP, New York, New York; for
Amici Curiae Peter Keisler, Stuart Gerson, Carter Phillips,
John Bellinger III, Samuel Witten, Ray Lahood, Brackett
Denniston, Stanley Twardy, and Richard Bernstein.
Alan E. Schoenfeld, Wilmer Cutler Pickering Hale and Dorr
LLP, New York, New York; Alex Gazikas, Wilmer Cutler
Pickering Hale and Dorr LLP, Washington, D.C.; Peter S.
Margulies, Bristol, Rhode Island; Shoba Sivaprasad Wadhia,
University Park, Pennsylvania; for Amici Curiae Professors
of Immigration Law.
Margaret L. Carter, Dmitiri D. Portnoi, and Daniel R. Suvor,
O’Melveny & Myers LLP, Los Angeles, California; Barbara
J. Parker, City Attorney; Maria Bee, Erin Bernstein, Malia
McPherson, Zarah Rahman, and Suzanne Dershowitz;
Office of the City Attorney Oakland, California; Edward N.
Siskel, Corporation Counsel, City of Chicago Department of
Law, Chicago, Illinois; Zachary W. Carter, Corporation
Counsel, New York City Law Department, New York, New
York; for Amici Curiae 21 Counties, Cities, and Local
Officials.
Xavier Becerra, Attorney General; Michael L. Newman,
Senior Assistant Attorney General; Christine Chuang,
Supervising Deputy Attorney General; Shubhra Shivpuri
and James F. Zahradka II, Deputy Attorneys General; Office
of the Attorney General, Oakland, California; Philip J.
Weiser, Attorney General, Denver, Colorado; William
Tong, Attorney General, Hartford, Connecticut; Kathleen
Jennings, Attorney General, Wilmington, Delaware; Karl A.
EAST BAY SANCTUARY COVENANT V. BIDEN 13
Racine, Attorney General, Washington, D.C.; Clare E.
Connors, Attorney General, Honolulu, Hawaii; Kwame
Raoul, Attorney General, Chicago, Illinois; Tom Miller,
Attorney General, Des Moines, Iowa; Brian E. Frosh,
Attorney General, Baltimore, Maryland; Maura Healey,
Attorney General, Boston, Massachusetts; Dana Nessel,
Attorney General, Lansing, Michigan; Keith Ellison,
Attorney General, St. Paul, Minnesota; Aaron D. Ford,
Attorney General, Carson City, Nevada; Gurbir S. Grewal,
Attorney General, Trenton, New Jersey; Hector Balderas,
Attorney General, Santa Fe, New Mexico; Letitia James,
Attorney General, New York, New York; Ellen F.
Rosenblum, Attorney General, Salem, Oregon; Josh
Shapiro, Attorney General, Harrisburg, Pennsylvania;
Peter F. Neronha, Attorney General, Providence, Rhode
Island; Thomas J. Donovan, Jr., Attorney General,
Montpelier, Vermont; Mark R. Herring, Attorney General,
Richmond, Virginia; Robert W. Ferguson, Attorney
General, Olympia, Washington; for Amici Curiae
California, Colorado, Connecticut, Delaware, District of
Columbia, Hawaii, Illinois, Iowa, Maryland, Massachusetts,
Michigan, Minnesota, Nevada, New Jersey, New Mexico,
New York, Oregon, Pennsylvania, Rhode Island, Vermont,
Virginia, and Washington.
14 EAST BAY SANCTUARY COVENANT V. BIDEN
ORDER
1. The full court was advised of the petition for
rehearing en banc. A judge of the court called for rehearing
en banc. The matter failed to receive a majority of the votes
of the non-recused active judges in favor of en banc
consideration. Fed. R. App. P. 35. Rehearing en banc is
DENIED.
2. Attached are Judge Paez’s concurrence to and Judge
Bumatay’s and Judge VanDyke’s dissents from the denial of
rehearing en banc.
3. The opinion filed on February 28, 2020 is amended
as follows:
On page 7, strike the sentence starting, “How strictly the
order binds this court,” up and through the sentence on
page 13 starting, “We discuss the merits of a stay request[.]”
Replace it with:
The Organizations contend we must
affirm the preliminary injunction because the
published motions panel order denying a stay
of the injunction pending appeal controls our
decision on the merits of the preliminary
injunction appeal. They argue that reversal of
the injunction would amount to overruling
the motions panel order, which we cannot do.
See Miller v. Gammie, 335 F.3d 889, 900 (9th
Cir. 2003) (holding that a published opinion
may only be overruled when it is clearly
irreconcilable with an intervening higher
authority). The published motions panel
order may be binding as precedent for other
EAST BAY SANCTUARY COVENANT V. BIDEN 15
panels deciding the same issue, but it is not
binding here. This is because the issues are
different. In deciding whether the court
should stay the grant or denial of a
preliminary injunction pending appeal, the
motions panel is predicting the likelihood of
success of the appeal. That is, the motions
panel is predicting rather than deciding what
our merits panel will decide. In resolving the
merits of a preliminary injunction appeal, our
merits panel is deciding the likelihood of
success of the actual litigation.
The Supreme Court has recognized this
distinction and made clear that it leads to
different analyses of the equities. See Trump
v. Int’l Refugee Assistance Project, 137 S. Ct.
2080, 2087 (2017). There is more judicial
discretion with respect to a stay. See id.; see
also Doe #1 v. Trump, 957 F.3d 1050, 1058
(9th Cir. 2020) (“A request for a stay pending
appeal is committed to the exercise of judicial
discretion.”); 11 WRIGHT & MILLER,
FEDERAL PRACTICE & PROCEDURE
§ 2904, at 702–03 (3d ed. 2012).
In Nken v. Holder the Supreme Court
emphasized that “[a]n injunction and a stay
have typically been understood to serve
different purposes” and a stay halting “the
conduct or progress of litigation before the
court, ordinarily is not considered an
injunction.” 556 U.S. 418, 428, 430 (2009)
(internal quotations omitted). Indeed, our
court has relied on Nken for the proposition
16 EAST BAY SANCTUARY COVENANT V. BIDEN
that “there are important differences between
a preliminary injunction and a stay pending
review…” Leiva-Perez v. Holder, 640 F.3d
962, 966 (9th Cir. 2011) (citing Nken,
129 S. Ct. at 1756–59). A stay “operates
upon the judicial proceeding itself,” while a
preliminary injunction “direct[s] an actor’s
conduct.” Nken, 556 U.S. at 428, 429.
In the government’s appeal, we are
charged with determining whether the district
court abused its discretion in granting the
preliminary injunction, see All. for the Wild
Rockies v. Cottrell, 632 F.3d 1127, 1131 (9th
Cir. 2011); the motions panel, instead,
considered whether the government raised
serious questions relating to the propriety of
the district court’s preliminary injunction and
whether the government would likely prevail
on appeal, see Leiva-Perez, 640 F.3d at 965–
66. The question presented to the motions
panel is an additional step removed from the
underlying merits of the district court’s
preliminary injunction. We discuss the
merits in “likelihood terms” and exercise
restraint in assessing the merits of either
question, see Sierra Club, 929 F.3d at 688,
but particularly so when considering the
“extraordinary request” to stay a preliminary
injunction granted by a district court. Barr v.
E. Bay Sanctuary Covenant, No. 19A230,
2019 WL 4292781, at *1 (Sept. 11, 2019)
(Sotomayor, J., dissenting from grant of a
stay).”
EAST BAY SANCTUARY COVENANT V. BIDEN 17
On page 14, insert the following footnote after the
citation to Leiva-Perez v. Holder, 640 F.3d 962, 967 (9th Cir.
2011):
There may be circumstances where a
motions panel does answer the same legal
question that is presented to the merits panel.
See Lair v. Bullock, 798 F.3d 736, 747 (9th
Cir. 2015) (“Lair II”). In Lair I, the motions
panel, while conducting a probabilistic and
discretionary analysis of the Nken factors,
addressed a pure question of law—whether
the Supreme Court had abrogated a relevant
circuit precedent. Lair v. Bullock, 697 F.3d
1200, 1206 (9th Cir. 2012) (“Lair I”). The
motions panel held that the very question at
issue had already been dispositively
answered in prior circuit precedent. Id. The
motions panel’s specific holding on this pure
question of law was neither based on an
assessment of probability nor an exercise of
discretion—it was necessarily compelled by
preexisting binding precedent.
Therefore, the merits panel in Lair II held
that it was bound by the motions panel’s
published decision in Lair I on this particular
issue where the motions panel answered
precisely the same question that was before
the merits panel. 798 F.3d at 747 (“The
motions panel in Lair I explicitly held that
Randall did not contain a majority opinion
capable of abrogating Eddleman.”) (internal
citations omitted). Lair II’s statement in this
context that “a motions panel’s published
18 EAST BAY SANCTUARY COVENANT V. BIDEN
opinion binds future panels the same as does
a merits panel’s published opinion,” id.,
simply confirms our circuit’s stare decisis
principle that a question already answered in
binding precedent will be controlled by that
answer when the same question is presented
in the future. See also Miller v. Gammie,
335 F.3d 889, 893 (9th Cir. 2003) (en banc).
But here, unlike in Lair II, none of the
questions answered by the motions panel
were pure questions of law for which
preexisting binding authority necessarily
compelled the answer.
On page 14, strike the sentence starting, “The question
before us now,” up and through the sentence on page 15
starting with, “Given the preliminary stage of the appellate
process[.]”
Replace it with:
The inquiry with respect to the stay
differs from the inquiry as to the preliminary
injunction. To the extent the issues share
predictive similarity, the motions panel may
be persuasive but not binding.
On page 15, strike “re-evaluate,” and replace it with
“consider.”
On page 16, strike “, as it did previously before the
district court and before the motions panel.”
On page 16, strike “renews,” and replace it with
“makes.”
EAST BAY SANCTUARY COVENANT V. BIDEN 19
On page 16, strike “before this court.”
On page 18, strike “We agree with the motions panel and
the district court,” and replace it with “We conclude.”
On page 27, strike “again” in the sentence before the
beginning of subsection B.
On page 28, strike “continue to” from the last full
sentence of the page.
On page 34, strike “again” from the sentence starting
with, “The government again suggests that the existence of
these eligibility bars[.]”
On page 37, strike “the motions panel” from the sentence
starting with, “The Attorney General’s interpretation,” and
the citation to EBSC II, 932 F.3d at 772–773, that appears
after the sentence.
On page 50, strike “we agree with the motions panel
that.”
4. An amended opinion is filed concurrently with this
order. An amended opinion concurring in the result, which
adds a new footnote 1 on page 1 and renumbers the ensuing
footnotes, is also filed concurrently with this order. No
further petitions for rehearing or rehearing en banc may be
filed.
5. In light of Appellants’ February 16, 2021 letter
informing the court that the President signed Executive
Order 14010 revoking “Proclamation 9880 of May 9, 2019,”
Dkt. No. 87, the panel directed the parties to simultaneously
file letter briefs addressing whether all or any aspect of this
appeal has been rendered moot. We have considered the
20 EAST BAY SANCTUARY COVENANT V. BIDEN
parties’ responses and agree with them that this appeal is not
moot. We decline to accept the parties’ suggestion to hold
the case in abeyance while Appellants review the interim
final rule at issue. The parties may address future
developments related to Appellants’ review of the interim
final rule and whether any such developments render the
case moot in the district court on remand.
OPINION
PAEZ, Circuit Judge:
Forty years ago, Congress recognized that refugees
fleeing imminent persecution do not have the luxury of
choosing their escape route into the United States. It
mandated equity in its treatment of all refugees, however
they arrived. 1
This principle is embedded in the Refugee Act of 1980,
which established an asylum procedure available to any
migrant, “irrespective of such alien’s status,” and
irrespective of whether the migrant arrived “at a land border
or port of entry.” Pub. L. No. 96-212, § 208(a), 94 Stat. 102,
105 (1980). Today’s Immigration and Nationality Act
(“INA”) preserves that principle. It states that a migrant who
arrives in the United States—“whether or not at a designated
port of arrival”—may apply for asylum. See 8 U.S.C.
§ 1158(a).
1
See 125 Cong. Rec. 35,813–14 (1979) (statement of Rep.
Holtzman).
EAST BAY SANCTUARY COVENANT V. BIDEN 21
In November 2018, the Departments of Justice and
Homeland Security jointly adopted an interim final rule (“the
Rule”) which, coupled with a presidential proclamation
issued the same day (“the Proclamation”), strips asylum
eligibility from every migrant who crosses into the United
States between designated ports of entry. In this appeal, we
consider whether, among other matters, the Rule unlawfully
conflicts with the text and congressional purpose of the INA.
We conclude that it does.
I.
The Rule announces a new bar to asylum eligibility. It
makes migrants who enter the United States in violation of
“a presidential proclamation or other presidential order
suspending or limiting the entry of aliens along the southern
border with Mexico” categorically ineligible for asylum.
See Aliens Subject to a Bar on Entry Under Certain
Presidential Proclamations; Procedures for Protection
Claims, 83 Fed. Reg. 55,934, 55,952 (Nov. 9, 2018)
(codified at 8 C.F.R. §§ 208.13, 208.30). Migrants who are
ineligible for asylum under the Rule will also automatically
receive negative credible-fear determinations in expedited-
removal proceedings. See id. at 55,935, 55,952. Typically,
a migrant in expedited-removal proceedings who
demonstrates a “credible fear” of persecution must be
allowed to present her asylum claim before an immigration
judge. See 8 U.S.C. § 1225(b)(1)(A)(ii), (B)(v). A migrant
who enters the United States in contravention of a
proclamation will instead need to demonstrate a “reasonable
fear” of persecution or torture—which is more difficult than
establishing a credible fear of persecution—to obtain other
forms of relief. See 83 Fed. Reg. at 55,936, 55,952; see also
8 C.F.R. § 208.31(c); 8 U.S.C. § 1225(b)(1)(B)(v).
22 EAST BAY SANCTUARY COVENANT V. BIDEN
The same day the Departments of Justice (“DHS”) and
Homeland Security (“DHS”) adopted the Rule, President
Trump issued the Proclamation. The Proclamation suspends
the entry of all migrants along the southern border of the
United States for ninety days, except for any migrant who
“enters the United States at a port of entry and properly
presents for inspection.” See Presidential Proclamation No.
9,822, Addressing Mass Migration Through the Southern
Border of the United States, 83 Fed. Reg. 57,661, 57,663
(Nov. 9, 2018).
Individually, the Rule and Proclamation have little
effect. The Proclamation does not have the force of law, and
the Rule only effectuates proclamations. But together, the
Rule and Proclamation make asylum entirely unavailable to
migrants who enter the country between ports of entry. The
magnitude of the Rule’s effect is staggering: its most direct
consequence falls on “the more than approximately 70,000
aliens a year (as of FY 2018) estimated to enter between the
ports of entry [who] then assert a credible fear in expedited-
removal proceedings.” 83 Fed. Reg. at 55,948. These
migrants would typically proceed to an asylum hearing
before an immigration judge but will now be unable to do so
because they have entered the country at a place other than a
port of entry.
The day the Proclamation and Rule issued, four legal
services organizations that represent current and future
asylum-seekers sued to prevent enforcement of the Rule.
East Bay Sanctuary Covenant, Al Otro Lado, Innovation
Law Lab, and Central American Resource Center of Los
Angeles (collectively, “the Organizations”) argued that the
Rule was likely unlawful because it was issued without
public notice and comment or complying with the thirty-day
grace period required by the Administrative Procedure Act
EAST BAY SANCTUARY COVENANT V. BIDEN 23
(“APA”), see 5 U.S.C. § 553(b)–(d). The Organizations also
argued that the Rule conflicts with the plain text of the INA
and is arbitrary and capricious because it constitutes a severe
departure from the Board of Immigration Appeals’s and the
Ninth Circuit’s interpretation of asylum practices in the
United States.
The district court agreed that the Rule “irreconcilably
conflicts with the INA and the expressed intent of Congress”
and entered a temporary restraining order enjoining the
Rule’s enforcement and ordering the government “to return
to the pre-Rule practices for processing asylum
applications.” See E. Bay Sanctuary Covenant v. Trump
(EBSC I), 349 F. Supp. 3d 838, 844, 868–69 (N.D. Cal.
2018). Eight days after the court’s order, the government
filed an appeal and an emergency motion in the district court
to stay the temporary restraining order pending appeal. The
court denied the stay motion three days later.
The following day, the government sought an immediate
stay in our court of the district court’s order pending appeal.
In a lengthy published order, a motions panel of this court
denied the government’s request to stay enforcement of the
court’s order. See E. Bay Sanctuary Covenant v. Trump
(EBSC II), 932 F.3d 742, 755, 762 (9th Cir. 2018). Although
temporary restraining orders are typically not appealable, the
panel concluded that appellate jurisdiction existed under
28 U.S.C. § 1292(a)(1) because the temporary restraining
order was effective for thirty days, well beyond the fourteen-
day limit imposed by Federal Rule of Civil Procedure 65(b).
Id. at 762–63. The government’s application for a stay from
the Supreme Court was also denied. See Trump v. E. Bay
Sanctuary Covenant, 139 S. Ct. 782 (2018).
While the government’s stay application was pending
before the Supreme Court, the Organizations filed a motion
24 EAST BAY SANCTUARY COVENANT V. BIDEN
for a preliminary injunction in the district court. The
arguments presented during the second round of litigation
were “nearly identical” to those made during the first. See
E. Bay Sanctuary Covenant v. Trump (EBSC III), 354 F.
Supp. 3d 1094, 1102 (N.D. Cal. 2018). Relying heavily on
the motions panel’s published order, the district court again
issued an injunction barring enforcement of the Rule. See
id. at 1121.
The government again appeals, arguing that the district
court erred when it entered the injunction or that the
injunction should at least be narrowed. We consolidated the
government’s appeal from the temporary restraining order
with the appeal from the preliminary injunction. 2 For the
reasons explained below, we agree with the district court that
the Rule is inconsistent with the INA, and we affirm the
district court’s orders granting preliminary injunctive relief.
II.
We first consider the effect of the motion panel’s order
on the present panel’s decision. The Organizations contend
we must affirm the preliminary injunction because the
published motions panel order denying a stay of the
injunction pending appeal controls our decision on the merits
2
Although the Proclamation expired by its terms in February 2019,
the President issued a new Proclamation, which did not substantially
change the terms of the original Proclamation and extended its effect for
an additional ninety days. When that Proclamation expired in May, the
President again re-issued it and extended the effect of the initial
Proclamation “for an additional 90 days beyond the date when the United
States obtains relief from the preliminary injunction of the interim final
rule[.]” See Presidential Proclamation No. 9,880, Addressing Mass
Migration Through the Southern Border of the United States, 84 Fed.
Reg. 21,229, 21,229 (May 8, 2019).
EAST BAY SANCTUARY COVENANT V. BIDEN 25
of the preliminary injunction appeal. They argue that
reversal of the injunction would amount to overruling the
motions panel order, which we cannot do. See Miller v.
Gammie, 335 F.3d 889, 900 (9th Cir. 2003) (holding that a
published opinion may only be overruled when it is clearly
irreconcilable with an intervening higher authority).
The published motions panel order may be binding as
precedent for other panels deciding the same issue, but it is
not binding here. This is because the issues are different. In
deciding whether the court should stay the grant or denial of
a preliminary injunction pending appeal, the motions panel
is predicting the likelihood of success of the appeal. That is,
the motions panel is predicting rather than deciding what our
merits panel will decide. In resolving the merits of a
preliminary injunction appeal, our merits panel is deciding
the likelihood of success of the actual litigation.
The Supreme Court has recognized this distinction and
made clear that it leads to different analyses of the equities.
See Trump v. Int’l Refugee Assistance Project, 137 S. Ct.
2080, 2087 (2017). There is more judicial discretion with
respect to a stay. See id.; see also Doe #1 v. Trump, 957 F.3d
1050, 1058 (9th Cir. 2020) (“A request for a stay pending
appeal is committed to the exercise of judicial discretion.”);
11 WRIGHT & MILLER, FEDERAL PRACTICE &
PROCEDURE § 2904, at 702–03 (3d ed. 2012).
In Nken v. Holder the Supreme Court emphasized that
“[a]n injunction and a stay have typically been understood to
serve different purposes” and a stay halting “the conduct or
progress of litigation before the court, ordinarily is not
considered an injunction.” 556 U.S. 418, 428, 430 (2009)
(internal quotations omitted). Indeed, our court has relied on
Nken for the proposition that “there are important differences
between a preliminary injunction and a stay pending
26 EAST BAY SANCTUARY COVENANT V. BIDEN
review…” Leiva-Perez v. Holder, 640 F.3d 962, 966 (9th
Cir. 2011) (citing Nken, 129 S. Ct. at 1756–59). A stay
“operates upon the judicial proceeding itself,” while a
preliminary injunction “direct[s] an actor’s conduct.” Nken,
556 U.S. at 428, 429.
In the government’s appeal, we are charged with
determining whether the district court abused its discretion
in granting the preliminary injunction, see All. for the Wild
Rockies v. Cottrell, 632 F.3d 1127, 1131 (9th Cir. 2011); the
motions panel, instead, considered whether the government
raised serious questions relating to the propriety of the
district court’s preliminary injunction and whether the
government would likely prevail on appeal, see Leiva-Perez,
640 F.3d at 965–66. The question presented to the motions
panel is an additional step removed from the underlying
merits of the district court’s preliminary injunction. We
discuss the merits in “likelihood terms” and exercise
restraint in assessing the merits of either question, see Sierra
Club, 929 F.3d at 688, but particularly so when considering
the “extraordinary request” to stay a preliminary injunction
granted by a district court. Barr v. E. Bay Sanctuary
Covenant, No. 19A230, 2019 WL 4292781, at *1 (Sept. 11,
2019) (Sotomayor, J., dissenting from grant of a stay). Such
a predictive analysis should not, and does not, forever decide
the merits of the parties’ claims. This sort of “pre-
adjudication adjudication would defeat the purpose of a stay,
which is to give the reviewing court the time to ‘act
responsibly,’ rather than doling out ‘justice on the fly.’”
Leiva-Perez v. Holder, 640 F.3d 962, 967 (9th Cir. 2011)
(quoting Nken v. Holder, 556 U.S. 418, 427 (2009)). 3
3
There may be circumstances where a motions panel does answer
the same legal question that is presented to the merits panel. See Lair v.
EAST BAY SANCTUARY COVENANT V. BIDEN 27
Notably, when acting on the government’s stay motion
in this case, the motions panel acknowledged the preliminary
nature of the stay proceedings. The panel issued a lengthy
opinion with detailed analysis, but repeatedly “stress[ed]”
that the case was still “at a very preliminary stage of the
proceedings,” and expected that “[f]urther development of
the record as the case progresses may alter [their]
conclusions.” EBSC II, 932 F.3d at 780. The panel also left
open various mixed questions of law and fact for a later
court—pointing out, for example, that if “facts develop in
the district court that cast doubt on the Organizations’
Bullock, 798 F.3d 736, 747 (9th Cir. 2015) (“Lair II”). In Lair I, the
motions panel, while conducting a probabilistic and discretionary
analysis of the Nken factors, addressed a pure question of law—whether
the Supreme Court had abrogated a relevant circuit precedent. Lair v.
Bullock, 697 F.3d 1200, 1206 (9th Cir. 2012) (“Lair I”). The motions
panel held that the very question at issue had already been dispositively
answered in prior circuit precedent. Id. The motions panel’s specific
holding on this pure question of law was neither based on an assessment
of probability nor an exercise of discretion—it was necessarily
compelled by preexisting binding precedent.
Therefore, the merits panel in Lair II held that it was bound by the
motions panel’s published decision in Lair I on this particular issue
where the motions panel answered precisely the same question that was
before the merits panel. 798 F.3d at 747 (“The motions panel in Lair I
explicitly held that Randall did not contain a majority opinion capable
of abrogating Eddleman.”) (internal citations omitted). Lair II’s
statement in this context that “a motions panel’s published opinion binds
future panels the same as does a merits panel’s published opinion,” id.,
simply confirms our circuit’s stare decisis principle that a question
already answered in binding precedent will be controlled by that answer
when the same question is presented in the future. See also Miller v.
Gammie, 335 F.3d 889, 893 (9th Cir. 2003) (en banc).
But here, unlike in Lair II, none of the questions answered by the
motions panel were pure questions of law for which preexisting binding
authority necessarily compelled the answer.
28 EAST BAY SANCTUARY COVENANT V. BIDEN
standing, the district court is, of course, free to revisit this
question,” id. at 763 n.6, and reiterating that its conclusions
were reached “at [the current] stage of the proceedings,” see
id. at 763, 767, 778, 779.
The inquiry with respect to the stay differs from the
inquiry as to the preliminary injunction. To the extent the
issues share predictive similarity, the motions panel may be
persuasive but not binding.
III.
We next consider the government’s challenge to our
jurisdiction. The government argues that the Organizations
lack Article III standing because they have not suffered a
cognizable injury and are outside the zone of interests
protected by the INA. The government also makes three
arguments before this court: (1) the Organizations lack a
“legally protected interest in maintaining their current
organizational structure or in the [R]ule’s application to third
parties,” Op. Br. of Gov’t at 29, 4 (2) the “immigration
context” of the Rule counsels against judicial intrusion, and
(3) various portions of the INA divest this court of
jurisdiction to entertain this appeal. We address each
argument in turn.
A.
The Article III standing inquiry serves a single purpose:
to maintain the limited role of courts by ensuring they protect
against only concrete, non-speculative injuries. See Lujan v.
Defenders of Wildlife, 504 U.S. 555, 583 (1992). Parties
4
We refer to the government’s opening brief as “Op. Br. of Gov’t,”
and to the government’s reply brief as “Reply Br. of Gov’t.”
EAST BAY SANCTUARY COVENANT V. BIDEN 29
must have a “personal stake in the outcome” sufficient to
ensure the court that, absent judicial review, they will suffer
or have suffered some direct injury. See id.
Organizations can assert standing on behalf of their own
members, see Friends of the Earth, Inc. v. Laidlaw Envtl.
Servs. (TOC), Inc., 528 U.S. 167, 181 (2000), or in their own
right, Havens Realty Corp. v. Coleman, 455 U.S. 363, 378–
79 (1982). To determine whether organizational standing
requirements have been satisfied, we “conduct the same
inquiry as in the case of an individual: Has the plaintiff
‘alleged such a personal stake in the outcome of the
controversy as to warrant his invocation of federal-court
jurisdiction?’” Havens, 455 U.S. at 378–79. The
Organizations therefore have the burden of demonstrating
that (1) they have suffered an injury-in-fact, meaning an
injury that is “concrete and particularized” and “actual and
imminent,” (2) the alleged injury is “fairly traceable” to the
defendants’ conduct, and (3) it is “more than speculative”
that the injury is judicially redressable. Lujan, 504 U.S.
at 560–61.
In Havens, the Supreme Court held that a fair housing
organization had standing under the Fair Housing Act where
the defendants’ allegedly racial steering practices had
frustrated the organization’s ability to assist equal access to
housing, and it had to devote “significant resources” to
identify and counteract those practices. 455 U.S. at 379.
Because the defendants’ practices had “perceptibly
impaired” the organization’s ability to provide its services,
the Court explained, “there can be no question that the
organization has suffered injury in fact.” Id.
We have read Havens to hold that an organization has
direct standing to sue where it establishes that the
defendant’s behavior has frustrated its mission and caused it
30 EAST BAY SANCTUARY COVENANT V. BIDEN
to divert resources in response to that frustration of purpose.
See Fair Hous. of Marin v. Combs, 285 F.3d 899, 905 (9th
Cir. 2002). Of course, organizations cannot “manufacture
the injury by incurring litigation costs or simply choosing to
spend money fixing a problem that otherwise would not
affect the organization at all,” but they can show they “would
have suffered some other injury” had they “not diverted
resources to counteracting the problem.” La Asociacion de
Trabajadores de Lake Forest v. Lake Forest, 624 F.3d 1083,
1088 (9th Cir. 2010); see also, e.g., El Rescate Legal Servs.,
Inc. v. Exec. Office of Immigration Review, 959 F.2d 742,
745, 748 (9th Cir. 1991).
We conclude that the Organizations have established that
the Rule has “perceptibly impaired” their ability to perform
the services they were formed to provide. EBSC II, 932 F.3d
at 765. This is sufficient for organizational standing. See
Combs, 285 F.3d at 904–05.
The Organizations share the same mission of assisting
migrants seeking asylum. “[B]ecause the Rule significantly
discourages a large number of [asylum-seekers] from
seeking asylum given their ineligibility,” the Rule frustrates
their mission. EBSC II, 932 F.3d at 766. The Rule has also
caused the Organizations to divert their already limited
resources in response to the collateral obstacles it introduces
for asylum-seekers. East Bay Sanctuary Covenant
(“EBSC”) and Innovation Law Lab (“ILL”), for example,
are located near Berkeley, California, and in Oregon,
respectively, and because most asylum-seekers who enter at
a designated port of entry will “remain detained in detention
facilities near the border hundreds of miles away,” EBSC III,
354 F. Supp. 3d at 1109 (internal quotation marks omitted),
those organizations “cannot represent asylum seekers.”
Decl. of Michael Smith at ¶ 6. Unaccompanied minors are
EAST BAY SANCTUARY COVENANT V. BIDEN 31
now often unable to seek asylum alone, and “[s]ince the new
rule was announced, Al Otro Lado [(“AOL”)] has been
overwhelmed with children who traveled to the southern
border of the United States to apply for asylum but now
cannot do so.” Supp. Decl. of Erika Pinheiro at ¶¶ 4, 15.
Caring for the often nonlegal needs of these unaccompanied
children is not part of AOL’s core mission and is “causing a
near complete diversion of [AOL’s] resources.” Id. ¶ 16. It
has “expended significant resources to send staff to the
border as it attempts to shift its programs.” EBSC III, 354 F.
Supp. 3d at 1109.
The funding on which the Organizations critically
depend is also jeopardized by the Rule. EBSC only “rarely”
represents people in removal proceedings. Decl. of Michael
Smith at ¶ 8. Because 80 percent of its clients have entered
without stopping at a port of entry in the past, EBSC stands
to “lose a significant amount of business and suffer a
concomitant loss of funding” if these individuals are deemed
categorically ineligible for asylum. EBSC III, 354 F. Supp.
3d at 1109 (citing EBSC II, 932 F.3d at 767). AOL and
CARECEN explain that the Rule decreases the funding they
stand to receive from the California Department of Social
Services. AOL often represents detained immigrants in their
bond proceedings, and “[s]ince the [R]ule went into effect,”
AOL has “not received a single referral for a bond case, as
persons who enter without inspection are ostensibly being
put into ‘Withholding-only’ proceedings and no longer
initially eligible for bond.” Supp. Decl. of Erika Pinheiro
at ¶ 22. CARECEN receives from the Department a flat
amount of funding per client it assists, and because more of
its clients are being put into more time- and resource-
intensive withholding proceedings, it will assist less clients
and receive less funding. Decl. of Daniel Sharp at ¶ 7.
32 EAST BAY SANCTUARY COVENANT V. BIDEN
Each organization would have lost clients seeking refuge
in the United States had it not diverted resources toward
counteracting the effect of the Rule. La Asociacion de
Trabajadores de Lake Forest, 624 F.3d at 1088. The
Organizations are not required to demonstrate some
threshold magnitude of their injuries; 5 one less client that
they may have had but-for the Rule’s issuance is enough. In
other words, plaintiffs who suffer concrete, redressable
harms that amount to pennies are still entitled to relief.
The government advances three additional justiciability
arguments. First, the government argues that the
Organizations have “no legally protected interest in
maintaining their current organizational structure or in the
Rule’s application to third parties, which the motions panel
did not consider in its analysis.” Op. Br. of Gov’t at 28. This
position misunderstands the injury-in-fact inquiry and
conflates organizational standing with third-party standing,
5
The government notes that “East Bay Sanctuary Covenant has only
‘around 35 clients who have entered without inspection and [who] expect
to file for affirmative asylum in the upcoming months,’” while, “[b]y
comparison, the ‘current backlog of asylum cases exceeds 200,000’ and
more than 200,000 inadmissible aliens present themselves for inspection
at ports of entry annually (even without the additional incentive to do so
that the Rule will create).” Op. Br. of Gov’t at 27 n.4.
The comparative magnitude of the harms alleged by the parties,
however, is not relevant for standing purposes; “a loss of even a small
amount of money is ordinarily an ‘injury.’” Czyzewski v. Jevic Holding
Corp., 137 S. Ct. 973, 983 (2017); see also Wallace v. ConAgra Foods,
Inc., 747 F.3d 1025, 1029 (8th Cir. 2014) (“The consumers’ alleged
economic harm—even if only a few pennies each—is a concrete, non-
speculative injury.”); Carpenters Indus. Council v. Zinke, 854 F.3d 1, 5
(D.C. Cir. 2017) (“A dollar of economic harm is still an injury-in-fact for
standing purposes.”).
EAST BAY SANCTUARY COVENANT V. BIDEN 33
which the Organizations have conceded is not at issue. 6 An
injury-in-fact is “an invasion of a legally protected interest,”
see Lujan, 504 U.S. at 560, but this means an interest that is
only concrete and particularized and actual or imminent—
not an interest protected by statute. This distinction prevents
Article III standing requirements from collapsing into the
merits of a plaintiff’s claim; “a petitioner’s ‘legally protected
interest’ need not be a statutorily created interest,” Ass’n of
Pub. Agency Customers v. Bonneville Power Admin.,
733 F.3d 939, 950 (9th Cir. 2013), and a plaintiff can have
standing despite losing on the merits. See also In re Special
Grand Jury 89-2, 450 F.3d 1159, 1172 (10th Cir. 2006)
(citing Warth v. Seldin, 422 U.S. 490, 500 (1975)
(“[S]tanding in no way depends on the merits of the
plaintiff’s contention that particular conduct is
illegal….”)).
More recent Supreme Court opinions have described
injury-in-fact as “a judicially cognizable interest”—
implying that “an interest can support standing even if it is
not protected by law…so long as it is the sort of interest that
courts think to be of sufficient moment to justify judicial
intervention.” In re Special Grand Jury 89-2, 450 F.3d
at 1172 (citing Bennett v. Spear, 520 U.S. 154, 167 (1997));
6
Many of the cases cited by the government in support of this
proposition do not concern organizational standing under Article III. In
O’Bannon v. Town Court Nursing Center, 447 U.S. 773 (1980), the
Court addressed whether nursing home residents have a right to an
administrative hearing before a state or federal agency hearing before the
agency revokes the home’s authority to provide them with nursing care
at government expense. Linda R.S. v. Richard D., 410 U.S. 614 (1973),
DaimlerChrysler Corp. v. Cuno, 547 U.S. 332 (2006), Arpaio v. Obama,
797 F.3d 11 (D.C. Cir. 2015), and Sure-Tan, Inc. v. N.L.R.B., 467 U.S.
883 (1984) all describe limitations on third-party, not organizational,
standing.
34 EAST BAY SANCTUARY COVENANT V. BIDEN
see also, e.g., Hollingsworth v. Perry, 570 U.S. 693, 707
(2013). Whether the Organizations have a sufficient
statutory or otherwise legal basis for their claims is irrelevant
at this threshold stage.
The government next argues that we should avoid
interfering with DOJ’s and DHS’s decision to adopt the Rule
because “[t]he Supreme Court has ‘long recognized the
power to expel or exclude aliens as a fundamental sovereign
attribute exercised by the Government’s political
departments largely immune from judicial control.’” See
Op. Br. of Gov’t at 30 (quoting Fiallo v. Bell, 430 U.S. 787,
792 (1977)).
We do not conduct independent policy analyses of
executive decisions. But we do “police the separation of
powers in litigation involving the executive[.]” In re
Cheney, 334 F.3d 1096, 1106 (D.C. Cir. 2003), vacated and
remanded on other grounds, 542 U.S. 367 (2004). For this
reason, there is a strong presumption favoring judicial
review of administrative action, see Bowen v. Mich. Acad. of
Family Phys., 476 U.S. 667, 670 (1986); non-reviewability
is an exception that must be clearly evidenced in the statute,
see Barlow v. Collins, 397 U.S. 159, 166–67 (1970).
Without such review, “statutes would in effect be blank
checks drawn to the credit of some administrative officer or
board.” Bowen, 476 U.S. at 671 (citing S. Rep. No. 752,
79th Cong., 1st Sess., 26 (1945)). Efficient agency
administration always requires some authority and
responsibility to resolve questions left unanswered by
Congress. It does not include the “power to revise clear
EAST BAY SANCTUARY COVENANT V. BIDEN 35
statutory terms.” 7 Utility Air Reg. Grp. v. E.P.A., 573 U.S.
302, 327 (2014).
We are therefore responsible for reviewing whether the
government has overstepped its delegated authority under
the INA and encroached upon Congress’s legislative
prerogative. See 5 U.S.C. § 706(2)(A).
Finally, the government argues that three provisions of
the Illegal Immigration Reform and Immigration
Responsibility Act (“IIRIRA”), 8 U.S.C. §§ 1252(e)(3),
1252(a)(5), and 1252(b)(9), divest this court of jurisdiction
to entertain this appeal. These statutes, in the government’s
view, require the Organizations to bring their claims in
individual-removal proceedings or in the District Court for
the District of Columbia.
Section 1252(e)(3) authorizes a limited court review of
expedited-removal proceedings. The statute requires that
judicial review of such administrative decisions be initiated
in the District Court for the District of Columbia, and limits
review to “determinations of (i) whether such section, or any
regulation issued to implement such section, is
constitutional; or (ii) whether such a regulation…is not
consistent with applicable provisions of this subchapter or is
7
The irony of the government’s position is that
section 1158(b)(2)(C)—the INA rule-making delegation upon which it
relies—is based on a congressional mandate that was intended, at least
in part, to curtail “unfettered executive discretion” and assure Congress’s
“proper and substantial role in refugee admissions, given [its] plenary
power over immigration.” 125 Cong. Rec. 35,814–15 (1979) (statement
of Rep. Holtzman) (emphasis added). “[I]f there is a separation-of-
powers concern here, it is between the President and Congress.”
EBSC II, 932 F.3d at 774.
36 EAST BAY SANCTUARY COVENANT V. BIDEN
otherwise in violation of law.” 8 Section 1252(e)(3), in short,
limits jurisdiction over challenges to regulations
implementing expedited-removal orders. See Barajas-
Alvarado, 655 F.3d at 1086 n.10.
Section 1252(a)(5) operates in conjunction with section
1252(e). It limits review of expedited-removal orders to
habeas review under 1252(e) and further restricts any
appellate habeas review to considering only whether the
migrant is lawfully in the country. See id. at 1082; 8 U.S.C.
§ 1252(e)(2). Section 1252(b)(9) also applies only to
removal orders, but instead channels “[j]udicial review of all
questions of law and fact…arising from any action taken or
proceeding brought to remove an alien from the United
States[,]” to the courts of appeals. 8 U.S.C. § 1252(b)(9);
see also I.N.S. v. St. Cyr, 533 U.S. 289, 313 (2001).
In the APA context, these provisions prohibit “a claim
by an alien, however it is framed, [that] challenges the
procedure and substance of an agency determination that is
‘inextricably linked’ to the order of removal[.]” Martinez v.
Napolitano, 704 F.3d 620, 623 (9th Cir. 2012). “[C]laims
that are independent of or collateral to the removal process”
are not actions taken to “remove an alien from the United
States.” J.E.F.M. v. Lynch, 837 F.3d 1026, 1032 (9th Cir.
2016); 8 U.S.C. § 1252(b)(9). The purpose of these claim-
channeling provisions is to “limit all aliens to one bite of the
8
Migrants can be placed in expedited removal proceedings when
they arrive at ports of entry without documents, misrepresent their
identities, or present fraudulent documents. See United States v.
Barajas-Alvarado, 655 F.3d 1077, 1081 (9th Cir. 2011). Undocumented
migrants who receive removal orders but indicate an intention to apply
for asylum or a fear of persecution may still be considered for asylum.
See id. (citing 8 U.S.C. § 1225(b)(1)(A)(i)).
EAST BAY SANCTUARY COVENANT V. BIDEN 37
apple with regard to challenging an order of removal.”
Martinez, 704 F.3d at 622.
None of these provisions have any bearing on the Rule.
Sections 1252(a)(5), (b)(9), and (e)(3) govern judicial
review of removal orders or challenges inextricably linked
with actions taken to remove migrants from the country. The
Rule “governs eligibility for asylum and screening
procedures for aliens subject to a presidential proclamation
or order restricting entry[.]” 83 Fed. Reg. at 55,934
(emphasis added). Bars to asylum eligibility may eventually
be relevant to removal proceedings, but they are not
“regulation[s]…to implement [removal orders]” or
otherwise entirely linked with removal orders. 9 8 U.S.C.
§ 1252(e)(3); see also Martinez, 704 F.3d at 623; O.A. v.
Trump, 404 F. Supp. 3d 109, 141 (D.D.C. 2019)
(Ҥ 1252(e)(3) is about challenges to expedited removal
orders and the implementation of the expedited removal
provisions that Congress enacted in IIRIRA.”). This is
consistent with the purposes of these jurisdictional
limitations: allowing collateral APA challenges to an
9
In another, strikingly similar context, the government appears to
agree with this interpretation of section 1252(e)(3). Before the District
Court for the District of Columbia, the government argued that section
1252(e)(3) divested the court of jurisdiction to hear an APA challenge to
an immigration decision issued by the Attorney General. See Grace v.
Whitaker, 344 F. Supp. 3d 96, 105 (D.D.C. 2018). The Attorney
General’s decision, and a policy memorandum that adopted the standards
in the decision, invoked the expedited-removal statute and required that
“claims based on membership in a putative particular social group
defined by the members’ vulnerability to harm…will not establish the
basis for asylum, refugee status, or a credible or reasonable fear of
persecution.” Id. at 110. The government there argued that the policy
memorandum and the Attorney General’s decision did not “implement”
section 1225(b) because it “was a decision about petitions for asylum
under section 1158.” Id. at 115–16 (emphasis added).
38 EAST BAY SANCTUARY COVENANT V. BIDEN
asylum-eligibility rule does not undermine Congress’s
desire to “limit all aliens to one bite of the apple with regard
to challenging” their removal orders. See Martinez, 704 F.3d
at 622.
At best, the law governing asylum is collateral to the
process of removal. Migrants in the country who file
affirmatively for asylum, or who are otherwise lawfully in
the country—such as those who have a valid visa, maintain
Temporary Protected Status, or are given parole, for
example—can apply and be eligible for asylum and never
encounter any of the statutory provisions governing
removal. See 8 C.F.R. § 208.4(a)(5)(iv). Other subsections
of the INA explicitly grant this court jurisdiction to review
denials of individual asylum applications, further reinforcing
that the jurisdiction-stripping provisions cited by the
government were not intended to apply at all to challenges
to asylum eligibility rules. See 8 U.S.C.
§§ 1252(a)(2)(B)(ii), (a)(2)(D); see also Morales v.
Gonzales, 478 F.3d 972, 978–79 (9th Cir. 2007).
We hold that the Organizations’ claims are justiciable
and they have otherwise satisfied the Article III standing
requirements.
B.
We generally also require that plaintiffs fall within the
“zone of interests” protected by the statute in question to
bring their claims in federal court. Lexmark Int’l, Inc., v.
Static Control Components, Inc., 572 U.S. 118, 129 (2014).
The breadth of the zone-of-interests test varies, depending
on the provisions of law at issue. Id. Under the APA, the
test is not “especially demanding.” Id. at 130 (quotations
and citations omitted). The zone-of-interests analysis
forecloses suit “only when a plaintiff’s interests are so
EAST BAY SANCTUARY COVENANT V. BIDEN 39
marginally related to or inconsistent with the purposes
implicit in the statute that it cannot reasonably be assumed
that Congress authorized that plaintiff to sue.” Id.
(quotations and citations omitted).
The Organizations bring their claims under the APA,
but because the APA provides a cause of action only to
those “suffering legal wrong because of agency
action…within the meaning of a relevant statute,” 5 U.S.C.
§ 702, the relevant zone of interest is that of the INA. EBSC
II, 932 F.3d at 767–68. And the relevant purpose is not that
of the entire INA; it is “by reference to the particular
provision of law upon on which the plaintiff relies.”
Bennett, 520 U.S. at 175–76.
In our review, we are “not limited to considering the
[specific] statute under which [plaintiffs] sued, but may
consider any provision that helps us to understand Congress’
overall purposes” in enacting the statute. Clarke v. Sec.
Indus. Ass’n, 479 U.S. 388, 401 (1987); see also EBSC II,
932 F.3d at 768. This inquiry is intended only to help clarify
the act’s scope—not determine whether Congress intended a
cause of action to arise for the plaintiff in question. See
Match-E-Be-Nash-She-Wish Band of Pottawatomi Indians
v. Patchak, 567 U.S. 209, 225 (2012) (“We do not require
any indication of congressional purpose to benefit the
would-be plaintiff.”) (internal quotation marks omitted).
The Organizations’ claims fall within the zone of
interests of the INA and of the regulatory amendments
implemented by the Rule. The Rule, much like the scope of
section 1158(b) of the INA, shapes asylum eligibility
requirements for migrants. The Organizations’ purpose is to
help individuals apply for and obtain asylum, provide low-
cost immigration services, and carry out community
education programs with respect to those services. EBSC III,
40 EAST BAY SANCTUARY COVENANT V. BIDEN
354 F. Supp. 3d at 1108–10. This is sufficient for the Court’s
lenient APA test: at the very least, the Organizations’
interests are “marginally related to” and “arguably within”
the scope of the statute. See Patchak, 567 U.S. at 224, 225.
IV.
We turn to the merits of the preliminary injunction 10
entered by the district court. A plaintiff seeking a
preliminary injunction must establish that (1) he is likely to
succeed on the merits, (2) he is likely to suffer irreparable
harm in the absence of preliminary relief, (3) the balance of
equities tips in his favor, and (4) an injunction is in the public
interest. All. for the Wild Rockies, 632 F.3d at 1131 (citing
Winter v. National Res. Def. Council, 555 U.S. 7, 20 (2008)).
When the government is a party, the last two factors (equities
and public interest) merge. Nken, 556 U.S. at 435. These
factors are evaluated on a sliding scale. All. for the Wild
Rockies, 632 F.3d. at 1131–34.
We review for abuse of discretion the district court’s
grant of a preliminary injunction. Arc of Cal. v. Douglas,
757 F.3d 975, 983 (9th Cir. 2014). District courts abuse their
discretion when they rely on an erroneous legal standard or
clearly erroneous finding of fact. Id. (internal quotations
omitted).
10
The terms of the temporary restraining order entered by the district
court in EBSC I technically differ from the terms of the preliminary
injunction entered by the court in EBSC III, but the difference has no
practical effect: both injunctions prevent enforcement of the Rule and are
identical in scope. Therefore, we review them together.
EAST BAY SANCTUARY COVENANT V. BIDEN 41
A.
The likelihood of the Organizations’ success on the
merits depends on the substantive and procedural validity of
the Rule. See EBSC III, 354 F. Supp. 3d at 1111–12. They
must establish a likelihood that the Rule is either
substantively or procedurally invalid. See EBSC II, 932 F.3d
at 770. Because the record on appeal is now “fully
developed,” and the substantive validity of the Rule “rest[s]
primarily on interpretations of law, not the resolution of
factual issues, we may consider the merits of the case and
enter a final judgment to the extent appropriate.” Beno v.
Shalala, 30 F.3d 1057, 1063 (9th Cir. 1994) (internal
quotation marks omitted).
1.
The APA requires that we “hold unlawful and set aside
agency action, findings, and conclusions found to be…an
abuse of discretion, or otherwise not in accordance with
law.” 5 U.S.C. § 706(2)(A). Presidential action is not
ordinarily “agency action,” and is typically unreviewable
under the APA. Franklin v. Massachusetts, 505 U.S. 788,
796 (1992). But the Proclamation and Rule together create
an “operative rule of decision” for asylum eligibility that is
reviewable by this court. EBSC II, 932 F.3d at 770; see also
City of Carmel-by-the-Sea v. U.S. Dep’t of Transp., 123 F.3d
1142, 1166 (9th Cir. 1997) (holding that executive orders
with “specific statutory foundation[s]” that do not expressly
preclude judicial review are treated as agency action and
reviewed under the APA); Public Citizen v. U.S. Trade Rep.,
5 F.3d 549, 552 (D.C. Cir. 1993) (“Franklin is limited to
those cases in which the President has final constitutional or
statutory responsibility for the final step necessary for the
agency action directly to affect the parties.”).
42 EAST BAY SANCTUARY COVENANT V. BIDEN
To determine whether the Rule is “not in accordance
with law,” we apply the framework established in Chevron,
U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842
(1984). Under Chevron, we first consider “whether
Congress has directly spoken to the precise question at issue.
If the intent of Congress is clear, that is the end of the
matter.” Campos-Hernandez v. Sessions, 889 F.3d 564, 568
(9th Cir. 2018) (quoting Chevron, 467 U.S. at 842). Federal
courts are “the final authority on issues of statutory
construction and must reject administrative constructions
which are contrary to clear congressional intent.” Chevron,
467 U.S. at 843 n.9.
a.
We consider, then, whether the Rule conflicts with
Congress’s intent. The only section of the INA implicated
by the Rule is section 1158 (“Asylum”). That section begins
by stating that an undocumented migrant may apply for
asylum when she is “physically present in the United States”
or “arrives in the United States (whether or not at a
designated port of arrival…)[.]” 8 U.S.C. § 1158(a)(1).
DOJ and DHS adopted the Rule under section
1158(b)(2)(C)’s grant of authority to the Attorney General
to “establish additional limitations and conditions, consistent
with this section, under which an alien shall be ineligible for
asylum[.]” 11
11
A separate subsection of section 1158, 1158(d)(5)(B), grants the
Attorney General authority to impose “conditions or limitations on the
consideration of an application for asylum not inconsistent with this
chapter.” As the motions panel observed, had the Rule explicitly
conditioned applications for asylum (instead of eligibility for asylum) on
arriving at a designated point of entry, the Rule would be “quite
obviously, ‘not in accordance with law,’” EBSC II, 932 F.3d at 770
EAST BAY SANCTUARY COVENANT V. BIDEN 43
We agree with the district court that the Rule is “not in
accordance with law.” 5 U.S.C. § 706(2)(A). Section
1158(a) provides that migrants arriving anywhere along the
United States’s borders may apply for asylum. The Rule
requires migrants to enter the United States at ports of entry
to preserve their eligibility for asylum. It is effectively a
categorical ban on migrants who use a method of entry
explicitly authorized by Congress in section 1158(a). As the
district court stated, “[i]t would be hard to imagine a more
direct conflict” than the one presented here. EBSC III,
354 F. Supp. 3d at 1112.
The government argues that the structure of section 1158
mandates a different result. Critical to the government’s
argument is that section 1158 splits asylum applications
(§ 1158(a)) and eligibility (§ 1158(b)) into two different
subsections; therefore, the government explains, Congress
intended to allow DOJ to promulgate limitations on asylum
eligibility without regard to the procedures and
authorizations governing asylum applications. The text in
section 1158(a) requires only that migrants arriving between
ports of entry be permitted to “apply for asylum,” and the
Rule does not prevent migrants from submitting futile
asylum applications. (emphasis added).
This argument is unconvincing. We avoid absurd results
when interpreting statutes. Rowland v. Cal. Men’s Colony,
Unit II Men’s Adv. Council, 506 U.S. 194, 200–01 (1993).
Explicitly authorizing a refugee to file an asylum application
because he arrived between ports of entry and then
summarily denying the application for the same reason
borders on absurdity. The consequences of denial at the
(quoting 5 U.S.C. § 706(2)(A)), because 1158(a) directs migrants to
“apply for asylum” in accordance with section 1158.
44 EAST BAY SANCTUARY COVENANT V. BIDEN
application or eligibility stage are, to a refugee, the same.
See EBSC II, 932 F.3d at 771. Had Congress intended to
allow DOJ and DHS to override this provision, it could have
said so in its delegation of authority to the Attorney General
or in the statutory provisions governing asylum applications.
And Congress signaled its desire that any eligibility
limitations be consistent with application requirements;
limitations promulgated under the eligibility subsection of
the statute must be “consistent with this section”—meaning
the entirety of section 1158—not just consistent with this
subsection.
The other categorical bars to asylum in section 1158(b)
of the INA do not meaningfully inform our reading of the
statute and the Rule. See EBSC II, 932 F.3d at 771 n.12. The
INA contains various provisions making ineligible asylum
applicants who committed a serious, nonpolitical crime
outside the United States prior to arrival (8 U.S.C.
§ 1158(b)(2)(A)(iii)), assisted or otherwise participated in
the persecution of another person (8 U.S.C.
§ 1158(b)(2)(A(i)), or were firmly resettled in another
country prior to arriving in the United States (8 U.S.C.
§ 1158(b)(2)(A)(vi)), among other things. The government
suggests that the existence of these eligibility bars in the INA
demonstrates that Congress intended certain categories of
migrants to be permitted to apply for asylum even though
they are categorically ineligible. A migrant who was firmly
resettled in another country, for example, is still free to
complete an asylum application, even though she will be
barred from seeking asylum under section
1158(b)(2)(A)(vi).
But—unlike the eligibility bar effected by the Rule—the
statutory asylum bars in the INA do not separately conflict
with explicit text in section 1158(a). There is no provision
EAST BAY SANCTUARY COVENANT V. BIDEN 45
in section 1158(a), for example, that affirmatively requires
that migrants who were firmly resettled in another country
be permitted to apply for asylum. The Rule creates the only
bar to eligibility under section 1158(b) that directly conflicts
with language in section 1158(a). The statutory eligibility
bars noted above do not suggest Congress intended that
migrants who are subject to them be permitted to apply for
asylum. See also EBSC II, 932 F.3d at 772 (“‘[t]o say that
one may apply for something that one has no right to receive
is to render the right to apply a dead letter.’”) (quoting EBSC
I, 349 F. Supp. 3d at 857). The district court correctly
concluded that the Rule is substantively invalid because it
conflicts with the plain congressional intent instilled in
8 U.S.C. § 1158(a), and is therefore “not in accordance with
law,” 5 U.S.C. § 706(2)(A).
b.
But even if the text of section 1158(a) were ambiguous,
the Rule fails at the second step of Chevron because it is an
arbitrary and capricious interpretation of that statutory
provision. If the statute is “silent or ambiguous with respect
to the specific issue, the question for the court is whether the
agency’s answer is based on a permissible construction of
the statute.” Campos-Hernandez, 889 F.3d at 568 (quoting
Chevron, 467 U.S. at 843). Under this standard, we must
give effect to an agency’s reasonable interpretation of a
statute, unless the interpretation is inconsistent with clearly
expressed congressional intent. See United States v. Fulton,
475 U.S. 657, 666–67 (1986).
The Board of Immigration Appeals (“BIA”) and this
court have long recognized that a refugee’s method of
entering the country is a discretionary factor in determining
whether the migrant should be granted humanitarian relief.
EBSC II, 932 F.3d at 772. More than thirty years ago, the
46 EAST BAY SANCTUARY COVENANT V. BIDEN
BIA stated that “an alien’s manner of entry or attempted
entry is a proper and relevant discretionary factor” to
adjudicating asylum applications under section 1158(a), but
“it should not be considered in such a way that the practical
effect is to deny relief in virtually all cases.” 12 Matter of
Pula, 19 I. & N. Dec. 467, 473 (B.I.A. 1987), superseded in
part by statute on other grounds as stated in Andriasian v.
I.N.S., 180 F.3d 1033, 1043–44 (9th Cir. 1999); see also
EBSC II, 932 F.3d at 772. The court explained that it would
instead evaluate “the totality of the circumstances and
actions of an alien in his flight from the country where he
fears persecution,” rather than deny asylum outright because
of a single procedural flaw in the migrant’s application.
Matter of Pula, 19 I. & N. Dec. at 473–74.
Especially where a migrant may be eligible only for
asylum and cannot establish the more stringent criteria for
withholding-of-removal, the discretionary factors—
including method of entry—should be “carefully evaluated
in light of the unusually harsh consequences which may
befall an alien[.]” Id. at 474. Indeed, “the danger of
persecution should generally outweigh all but the most
egregious of adverse factors.” Id.
We have supported the BIA’s understanding of section
1158(a). The most vulnerable refugees are perhaps those
fleeing across the border through the point physically closest
to them. That a refugee crosses a land border instead of a
12
The BIA in Pula interpreted section 1158(a) before it was
amended to include the particular phrase at issue (“whether or not at a
designated port of arrival”). At the time, the relevant sentence stated
“The Attorney General shall establish a procedure for an alien physically
present in the United States or at a land border or port of entry,
irrespective of such alien’s status, to apply for asylum[.]” 8 U.S.C.
§ 1158(a) (1980).
EAST BAY SANCTUARY COVENANT V. BIDEN 47
port-of-entry says little about the ultimate merits of her
asylum application; “if illegal manner of flight and entry
were enough independently to support a denial of
asylum,…virtually no persecuted refugee would obtain
asylum.” EBSC II, 932 F.3d at 773 (quoting Huang v. I.N.S.,
436 F.3d 89, 100 (2d Cir. 2006)). Given the Rule’s effect of
conditioning asylum eligibility on a factor that has long been
understood as “worth little if any weight,” see Mamouzian v.
Ashcroft, 390 F.3d 1129, 1138 (9th Cir. 2004), in
adjudicating whether a migrant should be granted asylum, it
is an arbitrary and capricious interpretation of section
1158(a).
The Attorney General’s interpretation of section 1158(a)
is also unreasonable, as the district court discussed, in light
of the United States’s treaty obligations. See EBSC III,
354 F. Supp. 3d at 1112–13. The United States agreed to
comply with Articles 2 through 34 of the 1951 United
Nations Convention Relating to the Status of Refugees
(“1951 Convention”) and the 1967 United Nations Protocol
Relating to the Status of Refugees (“1967 Protocol”) in
1968. H.R. Rep. 96-781 (Conf. Rep.), at 19–20 (1980), as
reprinted in 1980 U.S.C.C.A.N. 160, 160–62; see also I.N.S.
v. Cardoza-Fonseca, 480 U.S. 421, 429, 436–37 (1987). To
streamline the United States’s refugee procedures and
implement the country’s new treaty commitments, Congress
passed the Refugee Act of 1980, which amended the INA
and created the country’s first codified rules governing
asylum. S. Rep. No. 96-256, at 1 (1979), as reprinted in
1980 U.S.C.C.A.N. 141, 141–42, 144; H.R. Doc. No. 96-
608, at 17–18 (1979); see also Negusie v. Holder, 555 U.S.
511, 535–36 (2009).
48 EAST BAY SANCTUARY COVENANT V. BIDEN
As the United Nations High Commissioner of Refugees
(“UNHCR”) explains, 13 the Rule runs afoul of three of these
codified rules: the right to seek asylum, the prohibition
against penalties for irregular entry, and the principle of non-
refoulement embodied in Article 31(1) of the 1951
Convention. Neither the 1967 Protocol nor the 1951
Convention require countries to accept refugees, but they do
ensure that refugees at each signatory’s borders have legal
and political rights and protections. See Cong. Research
Serv. S522-10, Review of U.S. Refugee Resettlement
Programs and Policies 15–16 (1980).
The definition of “refugee” used in the 1951 Convention
is “virtually identical” to the one adopted by Congress in the
INA. Cardoza-Fonseca, 480 U.S. at 437. Under both the
INA and the 1951 Convention, refugees are all individuals
who—because of a “well-founded fear of being persecuted
for reasons of race, religion, nationality, membership in a
particular social group or political opinion”—are “unable,”
or, because of such fear, “unwilling to return” to their home
countries. See 8 U.S.C. § 1101(a)(42); 1951 Convention,
Art. 1(A)(2). Once individuals meet the statutory definition
of a “refugee,” they may be granted asylum under the INA.
See 8 U.S.C. § 1158(b)(1)(A).
13
The arguments presented by the United Nations in its amicus brief
on how the 1951 Convention and 1967 Protocol should be construed are
not binding on this court. See Cardoza-Fonseca, 480 U.S. at 439 n.22.
But they do “provide[] significant guidance in construing the [1967]
Protocol, to which Congress sought to conform[,]” and are “useful in
giving content to the obligations that the Protocol establishes.” Id; see
also Miguel-Miguel v. Gonzales, 500 F.3d 941, 949 (9th Cir. 2007) (“We
view the UNHCR Handbook as persuasive authority in interpreting the
scope of refugee status under domestic asylum law.”) (internal quotation
marks and citation omitted).
EAST BAY SANCTUARY COVENANT V. BIDEN 49
Both the INA and the 1951 Convention acknowledge
that individuals may be stripped of their refugee status even
when they meet the other eligibility criteria for asylum. The
refugee provisions of the 1951 Convention “shall not apply”
to “any person with respect to whom there are serious
reasons for considering” that such a person has committed a
crime against peace, a war crime, a crime against humanity,
a non-political crime outside of the country of refuge, prior
to their admission as a refugee, or has been “guilty of acts
contrary to the purposes and principles of the United
Nations.” 1951 Convention, Art. 1(F)(a)–(c). The statutory
bars for eligibility in the INA are similarly severe.
Individuals who are otherwise refugees may not apply for
asylum if the Attorney General determines that they
“ordered, incited, assisted, or otherwise participated” in the
persecution of another, based on a trait protected by the INA;
“constitute[] a danger to the community of the United
States”; committed a “serious nonpolitical crime” outside
the country; are a “danger to the security” of the country;
have engaged in terrorist activities; or were “firmly resettled
in another country prior to arriving in the United States.”
8 U.S.C. § 1158(b)(2)(A)(i)–(vi).
The exceptions listed in the 1951 Convention “require
individualized assessments and ‘must be [interpreted]
restrictive[ly].” Br. for UNHCR as Amicus Curiae at 14 n.6
(quoting Office of the United Nations High Commissioner
for Refugees, Handbook on Procedures and Criteria for
Determining Refugee Status ¶ 149 (Geneva, 1979)). So too
the categorical bars on eligibility in the INA are interpreted
with lenience toward migrants to avoid infringing on the
commitments set forth in the 1951 Convention and 1967
Protocol. See, e.g., Ali v. Ashcroft, 394 F.3d 780, 790 (9th
Cir. 2005) (A “narrow interpretation of the firm resettlement
bar would limit asylum to refugees from nations contiguous
50 EAST BAY SANCTUARY COVENANT V. BIDEN
to the United States or to those wealthy enough to afford to
fly here in search of refuge. The international obligation our
nation agreed to share when we enacted the Refugee
Convention into law knows no such limits.”); Cardoza-
Fonseca, 480 U.S. at 449.
The asylum bars in the INA and in the 1951 Convention
appear to serve either the safety of those already in the
United States or, in the case of the firm-resettlement bar, the
safety of refugees. The Rule ensures neither. Even a broad
interpretation of these eligibility bars does not naturally
encompass a refugee’s method of entry. Illegal entry is not
ordinarily considered a “serious crime.” See Pena-
Cabanillas v. United States, 394 F.2d 785, 788 (9th Cir.
1968) (stating that the statute criminalizing entry into the
United States “is not based on any common law crime, but
is a regulatory statute enacted to assist in the control of
unlawful immigration by aliens” and “is a typical mala
prohibita offense”). Nor does a migrant’s method of entry
per se create a danger to the United States, serve as a useful
proxy for terrorist activity, or suggest the persecution of
another.
And the Rule surely does not suggest that the migrant has
received protection in a third country. Many migrants enter
between ports of entry out of necessity: they “cannot satisfy
regular exit and entry requirements and have no choice but
to cross into a safe country irregularly prior to making an
asylum claim.” Br. for UNHCR as Amicus Curiae at 15
(citing Memorandum by the Secretary-General, Ad Hoc
Comm. on Statelessness, Status of Refugees & Stateless
Persons, at Annex Art. 24, cmt. ¶ 2, U.N. Doc. E/AC.32/2
(Jan. 3, 1950); UNHCR Executive Committee Conclusion
No. 58 (XL) ¶ (i) (Oct. 13, 1989)). This was well recognized
when the Refugee Act of 1980 was drafted. See Pub. L. No.
EAST BAY SANCTUARY COVENANT V. BIDEN 51
96-212, § 208(a), 94 Stat. 102, 105 (1980). Prior to the
passage of the Act, migrants who arrived at a port of entry
were “given an opportunity to have their [asylum]
applications heard in a hearing before an immigration
judge,” but refugees arriving “at a land border of the United
States [we]re not given this right.” Refugee Act of 1979:
Hearing on H.R. 2816 before the Subcomm. on Immigration,
Refugees, and Int’l Law of the Comm. on the Judiciary, 96th
Cong. 190 (1979) (testimony of David Carliner, American
Civil Liberties Union). In its attempt to streamline the
country’s refugee and asylum laws, Congress was urged to
consider that “persons who seek any benefits under [the
INA] should be entitled to a uniform procedure.” Id.
Congress heeded this consideration during the drafting of the
Refugee Act, eventually describing it as “establish[ing] a
more uniform basis for the provision of assistance to
refugees, and [] other purposes.” Refugee Act of 1980, Pub.
L. No. 96-212, 94 Stat. 102 (1980). The Rule defies this
desire for uniformity and denies refuge to those crossing a
land border. The effects of the Rule contravene the United
States’s commitments in the 1951 Convention.
Article 31(1) of the 1951 Convention also explains that
signatories “shall not impose penalties” on account of
refugees’ “illegal entry or presence,” 1951 Convention
Art 31(1). Notwithstanding the government’s
interpretations otherwise, “deportation is an integral part—
indeed, sometimes the most important part—of the penalty
that may be imposed” on migrants who are found guilty of
specified crimes, or for other reasons are barred from
seeking asylum. 14 See Padilla v. Kentucky, 559 U.S. 356,
14
The UNHCR’s view is that “penalties” in Article 31(1)
“encompasses civil or administrative penalties as well as criminal ones.”
Br. for UNHCR as Amicus Curiae at 20.
52 EAST BAY SANCTUARY COVENANT V. BIDEN
364 (2010) (footnote omitted). The Rule imposes an
additional penalty on refugees because of their “illegal
entry” by risking the deportation of migrants who enter the
country at a land border. 1951 Convention Art. 31(1).
And by categorically denying refugees an opportunity to
seek asylum only because of their method of entry, the Rule
is also in tension with the United States’s commitment to
avoid refouling individuals to countries where their lives are
threatened. Article 31(1) of the 1951 Convention prohibits
signatories from “expel[ling] or return[ing] (‘refouler’) a
refugee in any manner whatsoever to the frontiers of
territories where his life or freedom would be threatened[.]”
The INA’s withholding-of-removal, 8 U.S.C. § 1231(b)(1),
and Convention Against Torture (“CAT”) protections,
8 C.F.R. § 1208.16–18, are not as great as those conferred
by the INA’s asylum provisions. The evidentiary standard
that applicants must meet for either withholding-of-removal
or CAT relief is higher than the evidentiary standard for
asylum. See, e.g., Ling Huang v. Holder, 744 F.3d 1149,
1152 (9th Cir. 2014). Applicants for withholding-of-
removal and CAT relief must establish a “clear probability”
that they would be persecuted or tortured, respectively, if
they were removed to their home countries. See Korablina
v. I.N.S., 158 F.3d 1038, 1045–46 (9th Cir. 1998); Wakkary
v. Holder, 558 F.3d 1049, 1053 (9th Cir. 2009). A “clear
probability” of persecution or torture means that it is “more
likely than not” that applicants will be persecuted upon their
removal. I.N.S. v. Stevic, 467 U.S. 407, 424, 429–30 (1984).
Applicants for asylum instead must demonstrate only
that they are “unable or unwilling” to return to their home
countries “because of persecution or a well-founded fear of
persecution[.]” 8 U.S.C. § 1101(a)(42)(A). “One can
certainly have a well-founded fear of an event happening
EAST BAY SANCTUARY COVENANT V. BIDEN 53
when there is less than a 50% chance of the occurrence
taking place”; it would only be “too apparent,” for example,
for a refugee to have a “well-founded fear of being
persecuted” where “every tenth adult male person is either
put to death or sent to some remote labor camp” in the
applicant’s home country. Cardoza-Fonseca, 480 U.S.
at 431 (citing 1 A. Grahl-Madsen, The Status of Refugees in
International Law 180 (1966)). The Rule, then, risks the
removal of individuals with meritorious asylum claims who
cannot petition for withholding of removal or CAT relief.
By doing so, it is inconsistent with our treaty commitment to
non-refoulement.
The Rule is “arbitrary, capricious, or manifestly contrary
to the statute,” Chevron, 467 U.S. at 844, both because it is
contrary to plain congressional intent, and because it is an
arbitrary and capricious interpretation of section 1158(a).
Even if we agreed that the text of section 1158(a) is
ambiguous, the Rule flouts this court’s and the BIA’s
discretionary, individualized treatment of refugees’ methods
of entry, and infringes upon treaty commitments we have
stood by for over fifty years.
2.
Because we conclude that the Rule is substantively
invalid, we only briefly address the procedural arguments
raised by the parties. The APA requires public notice and
comment and a thirty-day grace period before a proposed
rule takes effect. 5 U.S.C. § 553(b)–(d). The notice-and-
comment requirements are exempted when “there is
involved a military or foreign affairs function of the United
States[,]” id. § 553(a), or when “the agency for good cause
finds…that notice and public procedure thereon are
impracticable, unnecessary or contrary to the public
interest.” Id. § 553(b)(B). The thirty-day lag in publication
54 EAST BAY SANCTUARY COVENANT V. BIDEN
can be waived where “good cause [is] found.” 15 Id.
§ 553(d)(3).
The Rule was issued without notice and comment or the
grace period. The government argues that the Rule was
properly issued because it falls under either the good-cause
or the foreign-affairs exceptions to these procedural
requirements.
a.
Proper invocation of the good-cause exception is
“sensitive to the totality of the factors at play.” United States
v. Valverde, 628 F.3d 1159, 1164 (9th Cir. 2010). The
exception is a “high bar” because it is “essentially an
emergency procedure.” Id. at 1164, 1165. The government
must make a sufficient showing that “‘delay would do real
harm’ to life, property, or public safety,” EBSC II, 932 F.3d
at 777 (quoting Valverde, 628 F.3d at 1164–65), or that
“some exigency” interferes with its ability to carry out its
mission. Nat. Res. Def. Council, Inc. v. Evans, 316 F.3d 904,
911 (9th Cir. 2003).
In support of its reliance on the exception, the
government now cites a Washington Post article indicating
that when the United States stopped its policy of separating
migrant parents from their children, smugglers told asylum-
15
“Different policies” underlie the good-cause exception for the
thirty-day grace period and the good-cause exception for the notice-and-
comment requirement, and “they can be invoked for different reasons.”
Riverbend Farms, Inc. v. Madigan, 958 F.2d 1479, 1485 (9th Cir. 1992).
Notice-and-comment requirements are intended to ensure public
participation in rulemaking, and the thirty-day waiting period is
“intended to give affected parties time to adjust their behavior before the
final rule takes effect.” Id.
EAST BAY SANCTUARY COVENANT V. BIDEN 55
seekers that “the Americans do not jail parents who bring
children—and to hurry up before they might start doing so
again.” See Nick Miroff and Carolyn Van Houten, The
Border is Tougher to Cross Than Ever. But There’s Still One
Way into America, Wash. Post (Oct. 24, 2018). The district
court concluded that the article “at least supports the
inference” that the Rule might result in similar changes in
immigration policy, and held that the government had
“identified a ‘rational connection between the facts found
and the choice made’ to promulgate the interim Rule on an
emergency basis.” EBSC III, 354 F. Supp. 3d at 1115
(quoting Valverde, 628 F.3d at 1168).
A citation to this single article is not sufficient to
demonstrate that the delay caused by notice-and-comment or
the grace period might do harm to life, property, or public
safety. See EBSC II, 932 F.3d at 777. The government’s
reasoning continues to be largely speculative, see id. at 778;
no evidence has been offered to suggest that any of its
predictions are rationally likely to be true. The article does
not directly relate to the Rule, the consequences of the Rule,
or anything related to asylum eligibility.
Even if it did, that “the very announcement of [the]
proposed rule itself can be expected to precipitate activity by
affected parties that would harm the public welfare,” Reply
Br. of Gov’t at 21, is likely often, or even always true. The
lag period before any regulation, statute, or proposed piece
of legislation allows parties to change their behavior in
response. If we were to agree with the government’s
assertion that notice-and-comment procedures increase the
potential harm the Rule is intended to regulate, these
procedures would often cede to the good-cause exception.
Because the government has failed to demonstrate the
existence of an exigency justifying good cause, we hold that
56 EAST BAY SANCTUARY COVENANT V. BIDEN
the Rule likely does not properly fall under the good cause
exception.
b.
For the foreign affairs exception to apply, “the public
rulemaking provisions should provoke definitely
undesirable international consequences.” Yassini v.
Crosland, 618 F.2d 1356, 1360 n.4 (9th Cir. 1980).
Otherwise, the exception “would become distended if
applied to INS actions generally, even though immigration
matters typically implicate foreign affairs.” Id. Use of the
exception is generally permissible where the international
consequences of the rule-making requirements are obvious
or thoroughly explained. We have rejected its use where the
government has failed to substantiate its reliance on the
exception or explain the detrimental effects of compliance
with the APA’s requirements. See EBSC II, 932 F.3d at 776–
77.
The government cites to four documents in support of its
renewed argument that the foreign-affairs exception is
justified: a Memorandum of Understanding (“MOU”)
between DHS and the Mexican government, the Washington
Post article, credible-fear origin data published by the
Executive Office of Immigration Review (“EOIR”), and a
speech by President Trump. The four documents appear to
demonstrate that the Rule and Proclamation are related to
ongoing changes in the national immigration landscape, but
still fail to establish that adhering to notice and comment and
a thirty-day grace period will “provoke definitely
undesirable international consequences.” Yassini, 618 F.3d
at 1360 n.4.
We agree with the government that the cited MOU does
broadly “show[] that [immigration] negotiations have
EAST BAY SANCTUARY COVENANT V. BIDEN 57
happened in the past,” Op. Br. of Gov’t at 49, but this is
insufficient to demonstrate that notice and comment will
provoke undesirable international consequences. Indeed,
the MOU’s substance seems to undermine the “broader
diplomatic program involving sensitive and ongoing
negotiations with Mexico.” Op. Br. of Gov’t at 47 (internal
quotations omitted). Article 3 of the MOU states that
“[l]ocal repatriation agreements should conform to mutually
established criteria and principles for the repatriation of
Mexican nationals being repatriated from the United States
to Mexico.” The unilateral repatriation of Mexican nationals
set forth by the Rule—without requesting public
participation—undermines these terms.
The cited Washington Post page discusses an increase in
the proportion of families that seek asylum and the EOIR
data lists the country of origin of credible-fear cases and
summarizes the number of people that attempt to enter the
United States with an asylum application, the number of
cases completed in 2018, and the outcome of credible fear
cases. It is unclear how these data “reflect[] motivations for
crossing the border illegally,” Op. Br. of Gov’t at 49, and
even less clear how they demonstrate the consequences of
requesting public notice-and-comment on foreign policy.
And the speech by President Trump, as the district court
noted, discusses the domestic consequences of foreign
immigration, not the foreign policy consequences of
immigration into the United States. See EBSC III, 354 F.
Supp. 3d at 1114. The speech—like the MOU, the article,
and the EOIR data—does not suggest that the APA’s
rulemaking provisions might trigger or even shape
immediate consequences in foreign affairs.
The evidence relied on by the government here is largely
the same as the evidence previously before the motions panel
58 EAST BAY SANCTUARY COVENANT V. BIDEN
and the district court. While we remain “sensitive to the fact
that the President has access to information not available to
the public, and…[are] cautious about demanding
confidential information,” the connection between
negotiations with Mexico and the immediate
implementation of the Rule is still “not apparent.” EBSC II,
932 F.3d at 776. Broadly citing to the Rule’s immigration
context is insufficient to invoke the foreign-affairs
exception. See Yassini, 618 F.2d at 1360 n.4. The
government has not made a “sufficient showing” that “the
public rulemaking provisions should provoke definitely
undesirable international consequences.” Id.; see also
Evans, 316 F.3d at 912.
In sum, the government has not established that DOJ and
DHS properly invoked the foreign-affairs exception to the
notice-and-comment requirement and thirty-day grace
period.
B.
We next consider whether the Organizations have
established that, in the absence of a preliminary injunction,
they are likely to suffer irreparable harm. See Arizona
Dream Act Coalition v. Brewer, 757 F.3d 1053, 1068 (9th
Cir. 2014). Irreparable harm is “harm for which there is no
adequate legal remedy, such as an award for damages.” Id.
For this reason, economic harm is not generally considered
irreparable. But where parties cannot typically recover
monetary damages flowing from their injury—as is often the
case in APA cases—economic harm can be considered
irreparable. See California v. Azar, 911 F.3d 558, 581 (9th
Cir. 2018). Intangible injuries may also qualify as
irreparable harm, because such injuries “generally lack an
adequate legal remedy.” Brewer, 757 F.3d at 1068.
EAST BAY SANCTUARY COVENANT V. BIDEN 59
We agree with the district court that the Organizations
have established that they will suffer a significant change in
their programs and a concomitant loss of funding absent a
preliminary injunction enjoining enforcement of the Rule.
EBSC II, 932 F.3d at 767. Both constitute irreparable
injuries: the first is an intangible injury, and the second is
economic harm for which the Organizations have no vehicle
for recovery.
The Rule has already prompted the Organizations to
change their core missions. Since the Rule issued, ILL has
placed programmatic expansions on hold and has “had to
lessen its caseload[.]” Supp. Decl. of Stephen W. Manning
at ¶ 14. CARECEN notes that it will “divert significant
resources,” including “staff time and organizational
resources” to respond to the Rule. Decl. of Daniel Sharp at
¶¶ 11–13. EBSC has had to “divert resources away from its
core programs to address the new policy.” Decl. of Michael
Smith at ¶ 15. And, as discussed in Part III, supra, the
Organizations each stand to lose funding because of their
core changes in mission.
Importantly, the Organizations also filed suit the same
day that the Rule and the first proclamation issued; while not
dispositive, this suggests urgency and impending irreparable
harm. See Oakland Tribune, Inc. v. Chronicle Pub. Co., Inc.,
762 F.2d 1374, 1377 (9th Cir. 1985). We agree with the
district court that the Organizations have demonstrated a
sufficient likelihood of irreparable injury to warrant
injunctive relief. EBSC III, 354 F. Supp. 3d at 1116.
C.
The government next argues that the harms it will suffer
because of the preliminary injunction—namely, the harm
caused by the injunction “undermin[ing] the Executive
60 EAST BAY SANCTUARY COVENANT V. BIDEN
Branch’s constitutional and statutory authority to secure the
Nation’s borders,” and the “entry of illegal aliens”—
outweigh the benefit to the public and the Organizations
conferred by the injunction. Op. Br. of Gov’t at 51–52.
Relevant equitable factors include the value of complying
with the APA, the public interest in preventing the deaths
and wrongful removal of asylum-seekers, preserving
congressional intent, and promoting the efficient
administration of our immigration laws at the border.
First, “[t]he public interest is served by compliance with
the APA.” Azar, 911 F.3d at 581. Indeed, it “does not matter
that notice and comment could have changed the substantive
result; the public interest is served from the proper process
itself.” Id. at 581–82. The Organizations and various Amici
informed the district court that they would have submitted
comments explaining why the Rule disrupts their
organizational missions and fails to meet its intended
purpose, had they had the opportunity. The APA’s
requirements reflect “a judgment by Congress that the public
interest is served by a careful and open review of proposed
administrative rules and regulations.” Alcaraz v. Block,
746 F.2d 593, 610 (9th Cir. 1984) (citing Phil. Citizens in
Action v. Schweiker, 669 F.2d 877, 881 (9th Cir. 1982)). The
government’s failure to comply with the APA—particularly
given the strength of the Organizations’ procedural attack on
the Rule—weighs in favor of granting injunctive relief.
Second, the public has an interest in “ensuring that we
do not deliver aliens into the hands of their persecutors,”
Leiva-Perez, 640 F.3d at 971, and “preventing aliens from
being wrongfully removed, particularly to countries where
they are likely to face substantial harm,” Nken, 556 U.S. at
436. The Rule will likely result in some migrants being
wrongfully denied refugee status in this country. For
EAST BAY SANCTUARY COVENANT V. BIDEN 61
migrants affected by the Rule, withholding of removal and
CAT protection are the only forms of relief available. As
discussed, these forms of relief demand a higher burden of
proof than an asylum claim. At the initial screening
interview with an asylum officer, an applicant seeking
asylum need only present a “credible fear” of persecution,
while an applicant seeking withholding of removal of CAT
protection must demonstrate the higher “reasonable fear” of
persecution or torture.
The government’s opening brief notes that 17 percent of
the 34,158 migrants whose cases were completed in 2018
received asylum. See Op. Br. of Gov’t at 52. Assuming the
number of migrants remains constant, if even just 25 percent
of asylum-seekers with meritorious claims are denied
asylum because of their method of entry, over 1,000 people
will either be returned to home countries where they face
“persecution based on ‘race, religion, nationality,
membership in a political social group, or political
opinion,’” EBSC III, 354 F. Supp. 3d at 1117 n.15 (quoting
8 U.S.C. §§ 1101(a)(42), 1158(b)(1)), or forced to proceed
on limited-relief claims that demand more stringent
showings. If the rate of migration and the rate of migrants
claiming fear during the expedited removal process
continues to increase, see 84 Fed. Reg. at 21,229, the scale
of this wrongful removal will only worsen.
Third, the public has an interest in ensuring that the
“statutes enacted by [their] representatives are not imperiled
by executive fiat.” EBSC II, 932 F.3d at 779 (internal
quotation marks omitted). The INA, and the United States’s
signatory status to the 1951 Convention, “reflect the balance
Congress struck between the public interests in rendering
aliens who enter illegally inadmissible and subject to
criminal and civil penalties, and…preserving their ability to
62 EAST BAY SANCTUARY COVENANT V. BIDEN
seek asylum.” EBSC III, 354 F. Supp. 3d at 1117–18
(citations omitted). The Rule and Proclamation disrupt that
balance by overriding plain congressional intent.
Finally, the government and the public have an interest
in the “efficient administration of the immigration laws at
the border.” EBSC II, 932 F.3d at 779 (internal quotation
marks omitted). This interest is “weighty.” Landon v.
Plasencia, 459 U.S. 21, 34 (1982). “[C]ontrol over matters
of immigration is a sovereign prerogative, largely within the
control of the executive and the legislature.” Id. The
government has a compelling interest in ensuring that
injunctions—such as the one granted here—do not
undermine separation of powers by blocking the Executive’s
lawful ability to regulate immigration and rely on its
rulemaking to aid diplomacy.
The role of the judiciary in reviewing such policies is
narrow. It is merely to ensure that executive procedures do
not violate principles of due process or “displace
congressional choices of policy.” Id. at 35. This executive
deference, then, is closely linked with our determination on
the substantive validity of the Rule. Essentially, the weight
we ascribe to this factor depends on the extent to which we
agree that the Rule overrides plain congressional intent.
Because the Organizations have established that the Rule is
invalid, we do not place much weight on this factor. As the
motions panel noted: “[t]here surely are enforcement
measures that the President and the Attorney General can
take to ameliorate the [immigration] crisis, but continued
inaction by Congress is not a sufficient basis under our
Constitution for the Executive to rewrite our immigration
laws.” EBSC II, 932 F.3d at 774.
In sum, we agree with the district court that there is a
significant basis for concluding that the public interest
EAST BAY SANCTUARY COVENANT V. BIDEN 63
weighs “sharply” in the Organizations’ favor. See EBSC III,
354 F. Supp. 3d at 1111.
V.
Finally, we turn to the remedy entered by the district
court: an injunction preventing enforcement of the Rule.
The injunction enjoins the part of the Rule that removes
asylum eligibility from migrants who fail to follow a
presidential proclamation. EBSC III, 354 F. Supp. 3d
at 1121. It does not enjoin the credible-fear amendments,
but “they have no independent effect,” so they are effectively
enjoined as well. Id. at 1121 n.22. We conclude that the
district court did not abuse its discretion in enjoining
enforcement of the Rule.
Injunctive relief should be “no more burdensome to the
defendant than necessary to provide complete relief to the
plaintiffs before the court.” Univ. of Cal. v. U.S. Dep’t of
Homeland Sec., 908 F.3d 476, 511 (9th Cir. 2018) (internal
quotations omitted). “Where relief can be structured on an
individual basis, it must be narrowly tailored to remedy the
specific harm shown,” but there is “no general requirement
that an injunction affect only the parties in the suit.” Bresgal
v. Brock, 843 F.2d 1163, 1169–1170 (9th Cir. 1987). The
equitable relief granted by the district court is acceptable
where it is “necessary to give prevailing parties the relief to
which they are entitled.” Id. at 1170–71. District courts
have “considerable discretion” in crafting suitable equitable
relief; correspondingly, appellate review is “narrow.”
Lamb-Weston, Inc. v. McCain Foods, Ltd., 941 F.2d 970,
974 (9th Cir. 1991).
As discussed, the harms caused to the Organizations as a
result of the Rule include a (1) loss of funding and
(2) disruption of organizational purpose. Adequate
64 EAST BAY SANCTUARY COVENANT V. BIDEN
equitable relief must remedy both harms. Bresgal, 843 F.2d
at 1170–71. Both harms are due, in part, to the Rule’s likely
consequence of preventing asylum-seekers with meritorious
claims from entering the country along our southern border
and successfully obtaining asylum. The stymied flow of
refugees will result in less funding for the Organizations, and
a shift (sometimes wholesale) in their organizational
missions.
The Organizations do not limit their potential clients to
refugees that enter the United States only at the California-
Mexico or Arizona-Mexico border; they represent “asylum
seekers” broadly. Unlike the plaintiffs in California v.
Azar—individual states seeking affirmance of an injunction
that applied past their borders—the Organizations here “do
not operate in a fashion that permits neat geographic
boundaries.” EBSC III, 354 F. Supp. 3d at 1120–21; see also
Califano v. Yamasaki, 442 U.S. 682, 702 (1979) (The “scope
of injunctive relief is dictated by the extent of the violation
established, not by the geographical extent of the plaintiff
class.”). An injunction that, for example, limits the
application of the Rule to California, would not address the
harm that one of the Organizations suffers from losing
clients entering through the Texas-Mexico border. One
fewer asylum client, regardless of where the client entered
the United States, results in a frustration of purpose (by
preventing the organization from continuing to aid asylum
applicants who seek relief), and a loss of funding (by
decreasing the money it receives for completed cases).
The government suggests that plaintiffs “identify actual
aliens in the United States who would otherwise be subject
to the Rule,” Op. Br. of Gov’t at 57, but this suggestion fails
to redress the scope of the Organizations’ harms. Part of the
harm the Organizations have alleged is the difficulty posed
EAST BAY SANCTUARY COVENANT V. BIDEN 65
by the Rule in helping them reach migrants who will cross
the border; their missions are not limited to helping
individuals currently present in the United States. Even if
their missions were so limited, asking the Organizations to
seek and list every person in the country they might help in
the coming months is infeasible and impracticable. The
“Government has not proposed a workable alternative form
of the [injunction] that accounts” for the harm at issue but
“nevertheless appl[ies] only within the [] borders” of the
Ninth Circuit. Washington v. Trump, 847 F.3d 1151, 1167
(9th Cir. 2017); see also EBSC II, 932 F.3d at 779; EBSC III,
354 F. Supp. 3d at 1121.
Two other factors support the district court’s decision to
enjoin Defendants from taking any action to implement the
Rule. First, “[w]hen a reviewing court determines that
agency regulations are unlawful, the ordinary result is that
the rules are vacated—not that their application to the
individual petitioners is proscribed.” Univ. of Cal., 908 F.3d
at 511 (internal quotation marks omitted). Singular
equitable relief is “commonplace” in APA cases, and is often
“necessary to provide the plaintiffs” with “complete
redress.” Id. at 512. Our “typical response is to vacate the
rule and remand to the agency”; we “ordinarily do not
attempt, even with the assistance of agency counsel, to
fashion a valid regulation from the remnants of the old rule.”
Harmon v. Thornburgh, 878 F.2d 484, 494 (D.C. Cir. 1989).
Because of the broad equitable relief available in APA
challenges, a successful APA claim by a single individual
can affect an “entire” regulatory program. Lujan v. Nat’l
Wildlife Fed’n, 497 U.S. 871, 890 n.2 (1990).
Second, as the district court noted, there is an important
“need for uniformity in immigration policy.” Id. at 511; see
also EBSC III, 354 F. Supp. 3d at 1120–21. We previously
66 EAST BAY SANCTUARY COVENANT V. BIDEN
have recognized that the “Constitution requires a uniform
Rule of Naturalization; Congress has instructed that the
immigration laws of the United States should be enforced
vigorously and uniformly; and the Supreme Court has
described immigration policy as a comprehensive and
unified system.” Univ. of Cal., 908 F.3d at 511 (quoting
United States v. Texas, 809 F.3d 134, 187–88 (5th Cir. 2014)
(emphases in original)). The INA itself “was designed to
implement a uniform federal policy, and the meaning of
concepts important to its application are not to be determined
according to the law of the forum, but rather require[] a
uniform federal definition.” Kahn v. I.N.S., 36 F.3d 1412,
1414 (9th Cir. 1994) (internal quotation marks omitted).
Different interpretations of executive policy across circuit or
state lines will needlessly complicate agency and individual
action in response to the United States’s changing
immigration requirements. For these reasons, in
immigration cases, we “consistently recognize[] the
authority of district courts to enjoin unlawful policies on a
universal basis.” EBSC II, 932 F.3d at 779 (citing Univ. of
Cal., 908 F.3d at 511).
The government again “raises no grounds on which to
distinguish this case from our uncontroverted line of
precedent.” Id. Given the context of this case and the harm
the district court sought to address, we find no error or abuse
of discretion in the terms or scope of the preliminary
injunction.
VI.
For the reasons discussed, the district court’s orders
granting preliminary injunctions are AFFIRMED.
EAST BAY SANCTUARY COVENANT V. BIDEN 67
FERNANDEZ, Circuit Judge, concurring in the result:
I concur in the majority opinion because, and for the
most part only because, I believe that we are bound by the
published decision in East Bay Sanctuary Covenant v.
Trump (East Bay I), 932 F.3d 742 (9th Cir. 2018).
More specifically, we are bound by both the law of the
circuit and the law of the case. 1 Of course, the rules that
animate the former doctrine are not the same as those that
animate the latter. See Gonzalez v. Arizona, 677 F.3d 383,
389 n.4 (9th Cir. 2012) (en banc).
As we have said: “Circuit law…binds all courts within a
particular circuit, including the court of appeals itself. Thus,
the first panel to consider an issue sets the law not only for
all the inferior courts in the circuit, but also future panels of
the court of appeals.” Hart v. Massanari, 266 F.3d 1155,
1171 (9th Cir. 2001). Moreover: “Once a panel resolves an
issue in a precedential opinion, the matter is deemed
resolved, unless overruled by the court itself sitting en banc,
or by the Supreme Court.” Id. (footnote omitted). Published
opinions are precedential. See id. at 1177; see also
Gonzalez, 667 F.3d at 389 n.4. That remains true, even if
some later panel is satisfied that “arguments have been
characterized differently or more persuasively by a new
1
The majority has now taken steps to obscure its attacks on our
doctrines of law of the case and law of the circuit, which were set forth
in its original opinion, East Bay Sanctuary Covenant v. Trump, 950 F.3d
1242, 1261–65 (9th Cir. 2020). Nevertheless, my concurrence remains
the same and is based upon the same reasoning. See id. at 1284–86
(Fernandez, J., concurring in the result). In short, the majority’s attempt
to pull an invisibility cloak over the mainspring of its attacks cannot hide
damage wrought by them.
68 EAST BAY SANCTUARY COVENANT V. BIDEN
litigant,” 2 or even if a later panel is convinced that the earlier
decision was “incorrectly decided” and “needs
reexamination.” 3 And those rules are not mere formalities
to be nodded to and avoided. Rather, “[i]nsofar as there may
be factual differences between the current case and the
earlier one, the court must determine whether those
differences are material to the application of the rule or allow
the precedent to be distinguished on a principled basis.”
Hart, 266 F.3d at 1172. In this case, there are no material
differences—in fact, the situation before this panel is in
every material way the same as that before the motions
panel. Furthermore, there is no doubt that motions panels
can publish their opinions, 4 even though they do not
generally do so. 5 Once published, there is no difference
between motions panel opinions and other opinions; all are
entitled to be considered with the same principles of
deference by ensuing panels. Thus, any hesitation about
whether they should be precedential must necessarily come
before the panel decides to publish, not after. As we held in
Lair v. Bullock, 798 F.3d 736 (9th Cir. 2015):
Lair contended at oral argument that a
motions panel’s decision cannot bind a merits
panel, and as a result we are not bound by the
motions panel’s analysis in this case. Not so.
We have held that motions panels can issue
published decisions.…[W]e are bound by a
2
United States v. Ramos-Medina, 706 F.3d 932, 939 (9th Cir. 2013).
3
Naruto v. Slater, 888 F.3d 418, 425 n.7 (9th Cir. 2018).
4
See 9th Cir. Gen. Order 6.3(g)(3)(ii); see also id. at 6.4(b).
5
See Haggard v. Curry, 631 F.3d 931, 933 n.1 (9th Cir. 2010) (per
curiam).
EAST BAY SANCTUARY COVENANT V. BIDEN 69
prior three-judge panel’s published opinions,
and a motions panel’s published opinion
binds future panels the same as does a merits
panel’s published opinion.
Id. at 747 (citations omitted). 6 Therefore, the legal
determinations in East Bay I are the law of the circuit.
We have explained the law of the case doctrine as “a
jurisprudential doctrine under which an appellate court does
not reconsider matters resolved on a prior appeal.” Jeffries
v. Wood, 114 F.3d 1484, 1488–89 (9th Cir. 1997) (en banc),
overruled on other grounds by Gonzalez, 677 F.3d at 389
n.4. While we do have discretion to decline application of
the doctrine, “[t]he prior decision should be followed unless:
(1) the decision is clearly erroneous and its enforcement
would work a manifest injustice, (2) intervening controlling
authority makes reconsideration appropriate, or
(3) substantially different evidence was adduced at a
subsequent trial.” Id. at 1489 (internal quotation marks and
footnote omitted). 7 We have also indicated that, in general,
6
The majority opines that in this respect Lair’s holding is dicta. Not
so. The court’s first basis for rejecting Lair’s contention was the basis
just quoted. Its second basis was then set forth. Id. It gave both of those
alternatives weight and attention. See Woods v. Interstate Realty Co.,
337 U.S. 535, 537, 69 S. Ct. 1235, 1237, 93 L. Ed. 1524 (1949) (holding
“where a decision rests on two or more grounds, none can be relegated
to the category of obiter dictum.”); see also United States v. Vidal-
Mendoza, 705 F.3d 1012, 1016 n.5 (9th Cir. 2013); Guadalupe-Cruz v.
INS, 240 F.3d 1209, 1211 & n.5 (9th Cir.), corrected, 250 F.3d 1271 (9th
Cir. 2001).
7
The majority seems to add a fourth exception, that is, motions
panel decisions never constitute the law of the case. That would be
strange if those decisions can constitute the law of the circuit, which they
can. Moreover, the case primarily cited for that proposition did not
70 EAST BAY SANCTUARY COVENANT V. BIDEN
“our decisions at the preliminary injunction phase do not
constitute the law of the case,” 8 but that is principally
because the matter is at the preliminary injunction stage and
a further development of the factual record as the case
progresses to its conclusion may well require a change in the
result. 9 Even so, decisions “on pure issues of law…are
binding.” Ranchers Cattlemen, 499 F.3d at 1114. Of course,
the case at hand has not progressed beyond the preliminary
injunction stage. It is still at that stage, and the factual record
has not significantly changed between the record at the time
of the decision regarding the stay motion and the current
record. Therefore, as I see it, absent one of the listed
exceptions, which I do not perceive to be involved here, the
law of the case doctrine would also direct that we are bound
by much of the motions panel’s decision in East Bay I.
Applying those doctrines:
(1) The Organizations have standing. East Bay I,
932 F.3d at 765–69.
(2) The Organizations are likely to succeed on the
substantive merits. See id. at 770–74. As to procedural
validity regarding adoption of the regulation, the motions
indicate it was dealing with a published motions panel decision or one
that set forth its reasoning. See United States v. Lopez-Armenta,
400 F.3d 1173, 1175 (9th Cir. 2005). It also dealt with the unique area
of jurisdiction. See id.
8
Ranchers Cattlemen Action Legal Fund United Stockgrowers of
Am. v. U.S. Dep’t of Agric., 499 F.3d 1108, 1114 (9th Cir. 2007); see
also Stormans, Inc. v. Wiesman, 794 F.3d 1064, 1074, 1076 n.5 (9th Cir.
2015); Ctr. for Biological Diversity v. Salazar, 706 F.3d 1085, 1090 (9th
Cir. 2013).
9
See Ctr. for Biological Diversity, 706 F.3d at 1090.
EAST BAY SANCTUARY COVENANT V. BIDEN 71
panel decision that the foreign affairs exception to the notice
and comment procedures does not apply is binding. Id.
at 775–77. In addition, while the motions panel decision
regarding the good cause exceptions is not fully binding,
what it did determine was that the information then brought
to the attention of the panel and the district court did not
suffice. Id. at 777–78. In light of that, I agree with the
majority that merely adding the twenty-five-word sentence
from a Washington Post article was insufficient to justify
changing the motions panel result.
(3) The decisions made by the motions panel regarding
harm to the Organizations and balance of hardships are also
binding decisions regarding the propriety of the preliminary
injunction. Id. at 767, 778–79.
(4) The scope of the injunction is not overly broad. Id.
at 779–80.
Thus, I respectfully concur in the result of the majority
opinion.
PAEZ, Circuit Judge, concurring in the denial of rehearing
en banc, joined by FLETCHER, Circuit Judge.
We concur fully in the court’s denial of rehearing en
banc. We respond to several of the arguments in Judge
Bumatay’s and Judge VanDyke’s dissents.
I.
In his dissent from denial of rehearing en banc, Judge
Bumatay takes issue with the entire doctrinal category of
“organizational standing.” He contends that the amended
72 EAST BAY SANCTUARY COVENANT V. BIDEN
majority opinion’s 1 application of Supreme Court and circuit
precedent, concluding that the plaintiff Organizations
established standing, deviates from John Marshall’s and
James Madison’s respective visions of Article III. PJB
Dissent 83–84, 91. The majority opinion, however, does not
revise or deviate from organizational standing doctrine.
Instead, it applies well-settled, binding authority. Its
organizational standing holding fits squarely within the
Supreme Court’s analysis in Havens Realty Corp. v.
Coleman, 455 U.S. 363 (1982) and related cases in our court.
Although the majority opinion thoroughly addresses the
Organizations’ standing, we respond briefly to several of the
points raised by Judge Bumatay.
The majority opinion makes clear that an organization
suffers a cognizable Havens injury only if the challenged
action “perceptibly impaired,” id. at 379, the organization by
frustrating its mission and causing a diversion of resources
in response to that frustration, Fair Hous. of Marin v. Combs,
285 F.3d 899, 905 (9th Cir. 2002). The majority opinion
summarizes the record evidence demonstrating the various
ways that the Rule’s extensive restrictions on asylum
frustrated the missions of the Organizations and caused them
to divert significant resources away from their mission in
response. Amend. Op. 28–38.
Judge Bumatay’s dissent ignores the record evidence
indicating the injuries suffered by the Organizations and
substitutes his own conclusion that the Organizations simply
refuse to see the Rule as a new opportunity to pursue their
missions. PJB Dissent 89. Judge Bumatay quizzically asks,
“But how does spending money assisting migrants seeking
1
For simplicity, hereinafter we refer to the amended majority
opinion as the “majority opinion.”
EAST BAY SANCTUARY COVENANT V. BIDEN 73
asylum ‘frustrate’ the organizations’ mission to help
migrants seeking asylum?” Id. The record evidence, as set
forth in the Supplemental Excerpts of Record (“SER”),
thoroughly documents how the Rule harmed the
Organizations: 2
• “Since the new rule was announced, Al Otro Lado
has been overwhelmed with children who traveled to
the southern border of the United States to apply for
asylum but now cannot do so. These children are
very vulnerable, as they are unaccompanied and so
have no one to look out for them in Tijuana, a city
currently seeing record levels of violence…Caring
for these children is incredibly time consuming and
causing a near complete diversion of Al Otro Lado’s
resources away from its core mission, which is
providing legal services. In addition to providing
these children with legal advice, attending to the
many nonlegal needs of Al Otro Lado’s
unaccompanied children clients is forcing Al Otro
Lado to divert resources away from advising the
many other clients who need the organization’s legal
help, thus undermining our ability to achieve our
core mission. For example, staff members are going
with the children to try to get them on the list to cross
at the port of entry. Staff have also had to negotiate
with Mexican officials so that they will not take these
children into custody and instead allow them to stay
at a local youth shelter while Al Otro Lado
determines the best way to ensure that the children
have access to a legal means to seek asylum in the
2
Instead, Judge Bumatay answers with his own evidence-free
musing. PJB Dissent 89 (“This sounds more like fulfilling their mission,
rather than ‘frustrating’ it.”).
74 EAST BAY SANCTUARY COVENANT V. BIDEN
United States.” SER 4 (Declaration of Erika
Pinheiro).
• “Immediately after the interim final rule was
announced, up to 15 children at a time were waiting
in our office around the clock while we arranged
their accommodations. [Al Otro Lado’s] office often
felt more like a daycare center than a legal services
organization. Because our office was being used to
shelter our unaccompanied minor clients, we could
not spend as much time with our other clients.” SER
22 (Declaration of Nicole Ramos).
• “[East Bay Sanctuary Covenant]’s mission is to serve
poor individuals who have few resources and cannot
afford attorneys…EBSC is neither staffed nor
funded to provide representation for asylum
applicants in removal proceedings. To do so would
involve a fundamental change in our program and
mission…[I]f the new policy remains in effect,
EBSC stands to lose nearly all of our funding for our
affirmative asylum program, because 80% of our
clients in that program entered without inspection.
But under the new policy, if they enter without
inspection, they will no longer be able to access
asylum at all.” SER 26–27 (Declaration of Michael
Smith).
• “The new Proclamation would also put a financial
strain on this organization…[F]or much of its work
in removal proceedings, [Central American
Resource Center] received a flat fee regardless of the
time spent on a particular case. Because withholding
and CAT cases are more resource-intensive than
asylum cases, precluding certain CARECEN clients
EAST BAY SANCTUARY COVENANT V. BIDEN 75
from applying for asylum would create a greater
financial loss in these cases. In other words, we
would likely receive the same funding, but have to
devote more hours per case.” SER 58 (Declaration of
Daniel Sharp).
Judge Bumatay’s dissent ignores the record evidence,
glosses over the majority opinion’s analysis of the evidence,
and fails to cite a single case holding that the injuries which
the Organizations suffered fall outside the Supreme Court’s
decision in Havens. 3 His dissent also overlooks circuit
precedent that addresses a variety of analogous situations in
which organizations demonstrated they were “perceptibly
impaired” by being forced to divert “significant resources”
to establish standing under Havens. See 455 U.S. at 379;
Fair Hous. of Marin v. Combs, 285 F.3d 899, 905 (9th Cir.
2002); La Asociacion de Trabajadores de Lake Forest v.
Lake Forest, 624 F.3d 1083, 1088 (9th Cir. 2010); El Rescate
Legal Servs., Inc. v. Exec. Office of Immigration Review, 959
F.2d 742, 745, 748 (9th Cir. 1991). Finally, his dissent
ignores the fact that the opinion is consistent with the
interpretation of organizational standing by other circuits. 4
3
Judge Bumatay’s attempt to reduce the majority’s analysis to,
“Any loss of clients caused by the Rule, no matter how minimal, was
sufficient to confer Article III standing,” PJB Dissent 90, is misguided
and similarly ignores the record evidence.
4
See e.g., Centro de la Comunidad Hispana de Locust Valley v.
Town of Oyster Bay, 868 F.3d 104, 109–111 (2d Cir. 2017) (holding that
plaintiff organization dedicated to ending discrimination against
immigrant workers had standing to challenge enforcement of ordinance
that prohibited persons from stopping a vehicle to solicit employment
because the organization had committed attention, time, money and
personnel to prepare a response to the ordinance and because
enforcement of the ordinance might deter the organization from
76 EAST BAY SANCTUARY COVENANT V. BIDEN
In short, the majority opinion’s organizational standing
holding is consistent with Supreme Court precedent, Article
III, the decisions of our court and the decisions of other
courts.
II.
We respond briefly to Judge VanDyke’s attacks on the
integrity of the majority opinion.
First, Judge VanDyke accuses the majority of
impropriety in addressing the effect of the motions panel’s
published opinion on the merits panel’s subsequent
consideration of the appeal. This issue was expressly raised
in the parties’ briefs, and the majority opinion does nothing
more than address it. The government’s opening brief argued
that the merits panel was not bound by the motions panel’s
published opinion. The Organizations answered by arguing
that we were bound. The government replied to bolster their
position that we were not bound.
Under the circumstances here, where the parties
presented and fully briefed a potentially dispositive issue, it
counseling clients at a worksite where they congregated); OCA–Greater
Houston v. Texas, 867 F.3d 604, 610–612 (5th Cir. 2017) (holding that a
plaintiff organization that promoted voting among Asian-Americans had
standing to challenge voting laws that restricted reliance on assistants to
translate for voters with limited English-language abilities because the
organization had redirected resources towards educating members and
the public on how to comply with the law); Arcia v. Florida Secretary of
State, 772 F.3d 1335, 1341–1342 (11th Cir. 2014) (holding that plaintiff
organizations engaged in voter registration efforts had standing to
challenge a program designed to remove individuals from voting rolls
because the organizations diverted resources to address the program); see
also 13A Charles Alan Wright et al., Federal Practice & Procedure
§ 3531.9.5 n.15 (3d ed.) (collecting cases).
EAST BAY SANCTUARY COVENANT V. BIDEN 77
was clearly appropriate for the majority to address the issue.
This is how the process of adjudication is designed to
function. The majority opinion abides by the Supreme
Court’s unambiguous explanation that “in our adversarial
system of adjudication, we follow the principle of party
presentation…[I]n both civil and criminal cases, in the first
instance and on appeal, we rely on the parties to frame the
issues for decision and assign to courts the role of neutral
arbiter of matters the parties present.” United States v.
Sineneng-Smith, 140 S. Ct. 1575, 1579 (2020) (internal
citations and quotation marks omitted). There is nothing
remarkable about the majority opinion addressing an issue
presented and framed by the parties.
Second, Judge VanDyke complains that the majority
engages in “mischief,” LVD Dissent 104, by amending its
opinion during the en banc process. The majority did,
indeed, revise its opinion as a result of the en banc process.
There was, however, no “mischief” in that decision. A brief
explanation of our court’s en banc process provides the
backdrop for the amended opinion.
After a judge of our court calls for a vote to determine
whether a case should be reheard en banc under Fed. R. App.
P. 35(f), our General Orders require the calling judge to
“forward a memorandum” to all members of the court
“setting forth reasons” for the call. G.O. 5.4.c.1. Thereafter,
“any judge may circulate memoranda in response to [the] en
banc call,” to facilitate robust discussion about the issues.
G.O. 5.5.a. The opportunity to exchange views about the
arguments in the en banc call includes the judges who
authored the majority opinion. Id. The panel majority may
propose revisions to address concerns raised by off-panel
judges. Id. If, after the close of the memorandum exchange
period the en banc call fails to obtain the votes of a majority,
78 EAST BAY SANCTUARY COVENANT V. BIDEN
the panel resumes control of the case and may amend its
opinion. G.O. 5.5.c.
Consistent with this process, the panel majority here
considered the debate about law-of-the-case where a
motions panel has published an order. The panel majority
was persuaded to revise its opinion by the discussions in the
memorandum exchange. The revisions, which were
proposed to the court during the memorandum exchange
period, are reflected in the amended majority opinion. The
process by which the panel majority reached this conclusion
is nothing new. The role this process plays to collaboratively
strengthen the opinions of our court is acknowledged in
several opinions. See e.g., Carver v. Lehman, 558 F.3d 869,
878–79 (9th Cir. 2009) (Smith, J.) (“Until the mandate has
issued, opinions can be, and regularly are, amended or
withdrawn, by the merits panel at the request of the parties
pursuant to a petition for panel rehearing, in response to an
internal memorandum from another member of the court
who believes that some part of the published opinion is in
error, or sua sponte by the panel itself…[This] collaborative
process strengthens, not weakens, the final quality of those
opinions, thereby better enabling them to stand the test of
time, and engender the respect of thoughtful citizens for both
the opinion, and the court that produced it.”); Perez v. City
of Roseville, 926 F.3d 511, 526 (9th Cir. 2019) (Ikuta, J.)
(“Like all three-judge panels, we must resolve the case
before us to the best of our abilities, which may include
reconsidering and revising an opinion that has not yet
mandated.”).
In sum, there was nothing improper in the majority’s
revision of its opinion to accommodate the views of our
colleagues.
EAST BAY SANCTUARY COVENANT V. BIDEN 79
* * *
For all the above reasons, in addition to those in the
amended opinion, we concur in the court’s denial of
rehearing en banc.
BUMATAY, Circuit Judge, joined by IKUTA, BENNETT,
R. NELSON, LEE, and VANDYKE, Circuit Judges,
dissenting from the denial of rehearing en banc:
We are not “Platonic Guardians” of our nation’s public
policies. See L. Hand, The Bill of Rights 73 (1958). As
judges, we have no business standing athwart the choices of
the political branches no matter how misguided we believe
them to be. That fundamental limitation on our role is even
more pronounced in the immigration context, where it is
long settled that “the admission and exclusion of foreign
nationals is a fundamental sovereign attribute exercised by
the Government’s political departments largely immune
from judicial control.” See Trump v. Hawaii, 138 S. Ct.
2392, 2418 (2018) (simplified). The Supreme Court has
repeatedly warned us we overstepped our bounds when we
tried to curtail immigration policies in the recent past. See,
e.g., id.; Dep’t of Homeland Sec. v. Thuraissigiam, 140
S. Ct. 1959 (2020); Nielsen v. Preap, 139 S. Ct. 954 (2019);
Barr v. E. Bay Sanctuary Covenant, 140 S. Ct. 3 (2019);
Jennings v. Rodriguez, 138 S. Ct. 830 (2018).
Unfortunately, we have not learned from our mistakes.
Today, we once again second-guess the Executive’s
immigration policies.
This time, we enjoin an immigration regulation
temporarily limiting asylum eligibility to those who enter the
country at a port of entry, deeming the policy “absurd.” E.
80 EAST BAY SANCTUARY COVENANT V. BIDEN
Bay Sanctuary Covenant v. Trump (“East Bay”), 950 F.3d
1242, 1272 (9th Cir. 2020). To get there, we disregard two
central precepts of the judicial role. First, we ignore
constitutional limits on our jurisdiction by stretching
organizational standing doctrine beyond Article III’s reach.
Second, we re-write the asylum statute to add a prohibition
on the Executive’s authority not found anywhere in the
legislative text.
Whatever we may think of the prior administration’s
immigration policies, we are not “free-wheeling enforcers of
the Constitution and laws.” Initiative & Referendum Inst. v.
Walker, 450 F.3d 1082, 1087 (10th Cir. 2006) (en banc). We
lack any constitutional power to offer meaningful relief in
the absence of an injured party. With no injured party here,
we had no role to play in this dispute. And even if a proper
party were before the court, our role would be a limited
one—to determine the law as written by Congress. We have
no authority to look beyond the plain text of a statute to
manufacture a legislative prohibition against disfavored
policies.
The proper forum for policy disputes is the ballot box,
not the courthouse. Yet, by failing to revisit our decision
here, we again ordain ourselves a super-legislature—
presiding over our nation’s immigration policies. Because
we donned the cloaks of Platonic Guardians rather than the
robes of impartial judges in this case, I respectfully dissent
from the denial of rehearing en banc.
I.
In late 2018, the Department of Justice and the
Department of Homeland Security jointly adopted an interim
final rule restricting eligibility for asylum. See Aliens
Subject to a Bar on Entry Under Certain Presidential
EAST BAY SANCTUARY COVENANT V. BIDEN 81
Proclamations; Procedures for Protection Claims, 83 Fed.
Reg. 55,934 (Nov. 9, 2018) (“Rule”). The Rule didn’t curb
who may apply for asylum; rather, it would have made aliens
who enter the United States “contrary to the terms of the
proclamation or order” ineligible for asylum. Id. at 55,952.
That same day, former President Trump issued a
proclamation suspending the entry of non-permanent-
resident aliens crossing from Mexico outside of a designated
port of entry. See Proclamation No. 9822, Addressing Mass
Migration Through the Southern Border of the United States,
83 Fed. Reg. 57,661 §§ 1, 2 (Nov. 9, 2018)
(“Proclamation”). The suspension on entry was set to expire
after 90 days. Id. at 57,663. Together, the Rule and
Proclamation would have had the effect of barring asylum
for any alien who illegally enters the United States between
ports of entry on the southern border during the 90-day
period.
The Rule never went into effect. We enjoined it
immediately. East Bay Sanctuary Covenant v. Trump (“East
Bay II”), 932 F.3d 742, 755 (9th Cir. 2018). Yet, the Rule
was not challenged by any alien subject to the new asylum
bar. Instead, the Rule was challenged by four lawyers’
organizations that represent current and future asylum-
seekers. Id. On review of the merits, we concluded that the
organizations had standing to bring suit and held that the
Rule was “unlawful[]” because it conflicted with “the text
and congressional purpose” of the asylum statute, 8 U.S.C.
§ 1158. East Bay, 950 F.3d at 1259. Both conclusions are
wrong and should have been reviewed en banc.
82 EAST BAY SANCTUARY COVENANT V. BIDEN
II.
A.
The Constitution limits our jurisdiction to “Cases” and
“Controversies.” U.S. Const. art. III, § 2. This limitation is
grounded in the Framers’ “concern about the proper—and
properly limited—role of the courts in a democratic society.”
Warth v. Seldin, 422 U.S. 490, 498 (1975). Outside of our
jurisdiction, we lack the “power to declare the law.” Steel
Co. v. Citizens for a Better Env’t, 523 U.S. 83, 94 (1998).
To appreciate the jurisdictional bounds of the judicial power,
“we must refer directly to the traditional, fundamental
limitations upon the powers of common-law courts.”
Spokeo, Inc. v. Robins, 136 S. Ct. 1540, 1550–51 (2016)
(Thomas, J., concurring) (simplified).
At the time of the Founding, “the requirements of public
control over public rights and private control over private
rights predominated in American law.” Ann Woolhandler &
Caleb Nelson, Does History Defeat Standing Doctrine?, 102
Mich. L. Rev. 689, 695 (2004). 1 Indeed, when it came to
standing, “eighteenth- and nineteenth-century courts were
well aware of the need for proper parties, and they linked
1
“Private rights” are those “belonging to individuals, considered as
individuals.” 3 W. Blackstone, Commentaries on the Laws of England
*2 (1768). These rights traditionally included the rights of personal
security, property rights, and contract rights. Spokeo, 136 S. Ct. at 1550
(Thomas, J., concurring). By contrast, “only the government had the
authority to vindicate a harm borne by the public at large[.]” Id. at 1551;
see Lujan v. Defs. of Wildlife, 504 U.S. 555, 576 (1992) (“Vindicating
the public interest (including the public interest in Government
observance of the Constitution and laws) is the function of Congress and
the Chief Executive.”).
EAST BAY SANCTUARY COVENANT V. BIDEN 83
that issue to the distinction between public and private
rights.” Id. at 691. 2
From the earliest days of our Republic, “federal courts
did not…entertain mandamus actions initiated by private
relators who lacked private injury.” Id. at 707. When asked
to command the Secretary of State to deliver Mr. Marbury’s
judicial commission, the Supreme Court in Marbury v.
Madison commented on the limits of the judicial power:
“The province of the court is, solely, to decide on the rights
of individuals, not to enquire how the executive, or executive
officers, perform duties in which they have…discretion.”
5 U.S. (1 Cranch) 137, 170 (1803) (emphasis added). Chief
Justice Marshall was careful to identify that the Court was
adjudicating an individual right—that of Mr. Marbury to his
commission—and not a “political” subject. Marbury, 5 U.S.
at 166. Political subjects, by contrast, “respect the nation,
not individual rights, and being entrusted to the executive,
the decision of the executive is conclusive.” Id. 3
2
See, e.g., In re Wellington, 33 Mass. 87, 105 (1834)
(“Undoubtedly,…a private individual can apply for a writ of mandamus
only in a case where he has some private or particular interest to be
subserved, or some particular right to be pursued or protected by the aid
of this process, independent of that which he holds in common with the
public at large; and it is for the public officers exclusively to apply, where
public rights are to be subserved.”).
3
Chief Justice Marshall would later caution that “[i]f the judicial
power extended to every question under the laws of the United States[,]
[t]he division of power [among the branches of government] could exist
no longer, and the other departments would be swallowed up by the
judiciary.” Jonathan Adler, God, Gaia, the Taxpayer, and the Lorax:
Standing, Justiciability, and Separation of Powers After Massachusetts
and Hein, 20 Regent U. L. Rev. 175, 180–81 (2008) (quoting Chief
Justice Marshall’s papers).
84 EAST BAY SANCTUARY COVENANT V. BIDEN
That view of jurisdiction remained constant. Nearly a
century and half later, the Supreme Court similarly observed
that Article III authorizes federal courts to supervise the
proper functioning of the political bodies “only to the extent
necessary to protect justiciable individual rights against
administrative action fairly beyond the granted powers.”
Stark v. Wickard, 321 U.S. 288, 310 (1944).
But in the latter half of the 20th Century, a “sea
change…occurred in the judicial attitude towards the
doctrine of standing—particularly as it affects judicial
intrusion into the operations of the other two branches.” See
Antonin Scalia, The Doctrine of Standing as an Essential
Element of the Separation of Powers, 17 Suffolk U. L. Rev.
881, 882–83 (1983); see, e.g., Flast v. Cohen, 392 U.S. 83
(1968) (holding federal taxpayers had standing to challenge
federal expenditures that would assist religious schools
under the Establishment Clause). Courts began venturing
beyond protecting individual rights to serve the growing
belief that the judicial branch should adjudicate the public
law claims of public interest organizations. See, e.g., Scenic
Hudson Pres. Conference v. FPC, 354 F.2d 608, 616 (2d Cir.
1965) (“In order to insure that the Federal Power
Commission will adequately protect the public interest in the
aesthetic, conservational, and recreational aspects of power
development, those who by their activities and conduct have
exhibited a special interest in such areas, must be held to be
included in the class of ‘aggrieved’ parties[.]”).
In the last three decades, however, the Supreme Court
has deliberately reined in Article III’s standing analysis.
See, e.g., Lujan, 504 U.S. at 562–67 (holding that
environmental organizations did not have standing to
challenge environmental regulations when they suffered no
direct injury); Spokeo, 136 S. Ct. at 1549 (holding that a
EAST BAY SANCTUARY COVENANT V. BIDEN 85
violation of a statutorily granted right does not automatically
confer standing); Lexmark Int’l, Inc. v. Static Control
Components, Inc., 572 U.S. 118, 129 n.3 (2014) (calling into
question prudential standing doctrine).
Nonetheless, lower courts still at times exhibit a flippant
attitude towards Article III’s injury-in-fact requirements.
Nowhere is this more evident today than in the muddled
doctrine of organizational standing. See, e.g., People for the
Ethical Treatment of Animals v. U.S. Dep’t of Agric., 797
F.3d 1087 (D.C. Cir. 2015) (granting standing to public
interest groups that voluntarily expended resources to
counteract government inaction with respect to
environmental regulation); see id. at 1099 (Millett, J.,
dubitante) (declaring that this “ruling is in grave tension with
Article III precedent and principles”).
But the era of judicial adventurism has passed. It is time
we returned to our proper province—which “is, solely, to
decide on the rights of individuals.” Marbury, 5 U.S. at 170.
B.
Against this backdrop, it is difficult to see how the legal
services groups here could bring their suit under the
traditional notions of Article III. Arguably the most
important of Article III’s jurisdictional requirements is that
plaintiffs have standing to bring their case to federal court.
For either an individual or an organization to establish
Article III standing, the party must show (1) an injury in fact
that is concrete, particularized, and actual or imminent (not
conjectural or hypothetical); (2) that the injury is fairly
traceable to the challenged conduct; and (3) that the injury
will be redressed by a favorable decision. Lujan, 504 U.S.
at 561.
86 EAST BAY SANCTUARY COVENANT V. BIDEN
Here, the plaintiff organizations cannot assert a proper
injury-in-fact. The four organizations “share the same
mission of assisting migrants seeking asylum.” East Bay,
950 F.3d at 1266. But the organizations didn’t bring the suit
on behalf of client asylum-seekers harmed by the new
Rule—as one might expect. Rather, they sued in their own
right on behalf of their own interests. Id. As laudable as the
organizations’ work may be, the Rule didn’t regulate their
attorneys’ ability to provide legal services in any way; it only
conditioned a successful asylum application on an alien’s
manner of entry. The Rule therefore didn’t represent a direct
harm to the organizations themselves. As the historical
record shows, it is “substantially more difficult” to establish
standing when a plaintiff asserts an injury arising from the
government’s regulation of “someone else.” Lujan, 504 U.S.
at 562.
So how, then, were the organizations able to meet Article
III standing? First, the panel majority overcame the injury-
in-fact hurdle by holding that the organizations were injured
because the Rule amorphously “frustrate[d] their mission”
by causing them to redirect resources to assist asylum-
seekers in different ways. East Bay, 950 F.3d at 1266.
Second, the panel majority proclaimed that the lawyers’
organizations were injured because the Rule would diminish
their client base—announcing a shockingly broad rule that
standing is conferred if a policy change would result in even
“one less client” for an organization. Id. at 1267.
East Bay inaugurates a new standard of organizational
standing—one that allows an organization to challenge a
disfavored government policy by merely asserting that the
change could result in “one less client” or cause it to shift
resources to better support its mission. Id. at 1266–67. Such
EAST BAY SANCTUARY COVENANT V. BIDEN 87
a broad and malleable standard is an end-run around Article
III.
1.
Under traditional standing principles, the organizations
could not have established an injury-in-fact. As a general
rule, a plaintiff “must allege facts showing that he is himself
adversely affected” in order to establish standing. Sierra
Club v. Morton, 405 U.S. 727, 740 (1972) (emphasis added).
A plaintiff does not suffer injury by voluntarily expending
resources to counteract a governmental action that only
indirectly affects the plaintiff. Clapper v. Amnesty Int’l
USA, 133 S. Ct. 1138, 1148–51 (2013); see Linda R.S. v.
Richard D., 410 U.S. 614, 619 (1973) (stating that a party’s
interest in having the law properly enforced against others is
not a cognizable injury under Article III). While vindicating
one’s rights is central to our American legal tradition, it must
be the party’s rights being vindicated. In this case, none of
the organizations had a right that was violated by the Rule.
As such, we had no right to interfere with the political
branches.
The Rule imposed no new restriction or obligation on
attorneys of asylum seekers. It only sought to add a
condition for the successful grant of asylum. While it is easy
to see how the Rule would have cut against the
organizations’ goal to “assist[] migrants seeking asylum” as
a policy matter, East Bay, 950 F.3d at 1266, “a setback to
the organization’s abstract social interests” is not a concrete
injury, Havens Realty Corp. v. Coleman, 455 U.S. 363, 379
(1982) (holding that injury occurs when a policy
“perceptibly impair[s] [the organization’s] ability to
provide…services for [its clients]”). “[A] mere interest in a
problem, no matter how longstanding the interest and no
matter how qualified the organization is in evaluating the
88 EAST BAY SANCTUARY COVENANT V. BIDEN
problem, is not sufficient by itself to” confer standing.
Sierra Club, 405 U.S. at 739 (simplified). Praiseworthy as
the organizations’ mission may be, “Article III requires more
than a desire to vindicate value interests.” Diamond v.
Charles, 476 U.S. 54, 66 (1986). This principle is meant, in
part, to ensure that courts do not wade into “political or
ideological disputes about the performance of government.”
United States v. Richardson, 418 U.S. 166, 192 (1974)
(Powell, J., concurring).
To be sure, under our existing precedent, an organization
can sue “on its own behalf…when it suffered both a
diversion of its resources and a frustration of its mission.”
La Asociacion de Trabajadores de Lake Forest v. City of
Lake Forest, 624 F.3d 1083, 1088 (9th Cir. 2010)
(simplified) (emphasis added). But that’s not enough. It
must also “show that it would have suffered some other
injury if it had not diverted resources to counteracting the
problem.” Sierra Club v. Trump, 963 F.3d 874, 884 (9th Cir.
2020) (simplified). An organization is not injured by
choosing, of its own volition, to spend more money or to
shift resources from one mission goal to another. To be
injured, the organization must expend resources outside of
its mission. See Havens Realty Corp., 455 U.S. at 378–79
(recognizing injury when the organization devoted resources
to identifying and counteracting racially discriminatory
steering practices when its mission was to provide
counseling and referral services to low- and moderate-
income home-seekers). As recently as 2019, we found that
Second Amendment rights groups had no standing to
challenge the enforcement of a gun-confiscation law because
implementation of the law did not “impede[] their ability to
carry out their mission or require[] them to divert substantial
resources away from the organizations’ preferred uses.”
EAST BAY SANCTUARY COVENANT V. BIDEN 89
Rodriguez v. City of San Jose, 930 F.3d 1123, 1135 (9th Cir.
2019).
But we lifted these restrictions on organizational
expenditures in East Bay. Instead, we held that any change
in resources—even if consistent with its mission—is enough
to confer organizational injury. See East Bay, 950 F.3d at
1266 (holding the organizations were injured because the
Rule “caused the Organizations to divert their already
limited resources in response to the collateral obstacles it
introduces for asylum-seekers”). For example, we noted
how one organization had shifted resources to send staff to
the border to be closer to migrant detention facilities. Id.
But how does spending money assisting migrants seeking
asylum “frustrate” the organizations’ mission to help
migrants seeking asylum? This sounds more like fulfilling
their mission than “frustrating” it. 4 See Fair Hous. Council
of San Fernando Valley v. Roommate.com, LLC, 666 F.3d
1216, 1224 (9th Cir. 2012) (Ikuta, J., concurring in part and
dissenting in part) (“How can an organization have a legally
protected interest in not spending money to advance its core
mission?”). Given that the Rule did not directly “impair[]
[the organizations’] ability to” assist asylum-seekers,
Havens Realty Corp., 455 U.S. at 379, we should have
declined to find jurisdiction here.
2.
East Bay’s second ground of injury is even more
problematic than the first. In addition to the diversion of
4
The panel seemingly concedes the deficiency of its opinion by
relying now on extra facts from the supplemental record to support its
analysis. See Concurrence in Order at 73. But future litigants should not
have to pore through the supplemental record in order to reconstruct what
the panel actually meant.
90 EAST BAY SANCTUARY COVENANT V. BIDEN
resources, we recognized an injury to the organizations
simply because the Rule would have reduced their client
base. See 950 F.3d at 1266 (“Because the Rule significantly
discourages a large number of asylum-seekers from seeking
asylum given their ineligibility, the Rule frustrates [the
organizations’] mission.”) (simplified). Any “los[s of]
clients” caused by the Rule, no matter how minimal, was
sufficient to confer Article III standing. Id. We even went
so far as to announce that “one less client that [the
organizations] may have had but-for the Rule’s issuance is
enough” for constitutional injury. Id. at 1267.
As a result of our ruling, the distinction between injuries
to an organization and the injuries to third parties served by
the organization becomes meaningless. Under East Bay, a
group has standing to petition our court simply by asserting
that a government policy impacts third parties assisted by the
organization. Such a low threshold, I predict, will lead to
significant unintended consequences.
Under the “one less client” theory of injury, any group
could challenge any governmental policy in its own right if
the policy would have any de minimis effect on its client
base. Law firms, for example, could directly challenge
statutes and regulations without bothering to obtain an
injured client. Public interest groups could seek to enjoin
any regulation that targets its area of interest. We have
moved well beyond requiring particularized and concrete
injury and have embraced a “general grievance” theory of
jurisdiction by construing organizational standing so
broadly.
Jurisdiction is not like the Big Bang Theory, with an
ever-expanding universe. Yet, East Bay’s sweeping
pronouncement marks a new outer bound in our inflation of
organizational standing.
EAST BAY SANCTUARY COVENANT V. BIDEN 91
C.
While the expansion of organizational standing doctrine
ostensibly redounds to the benefit of public-interest groups,
the real issue here is the aggrandizement of our own
authority. Article III’s “limitations preserve separation of
powers by preventing the judiciary’s entanglement in
disputes that are primarily political in nature.” Spokeo, 136
S. Ct. at 1551 (Thomas, J., concurring). By ignoring these
fundamental principles, we risk our courts becoming
“publicly funded forums for the ventilation of public
grievances.” Valley Forge Christian Coll. v. Ams. United
for Separation of Church & State, Inc., 454 U.S. 464, 473
(1982).
By loosening organizational standing requirements, we
only increase our own authority to adjudicate policy
disputes. We no longer need to wait for a rule or regulation
to actually injure a party. Now, we can skip ahead and
immediately superintend any policy disagreement from the
get-go by entertaining the bevy of public interest
organizations willing to challenge the disfavored policy du
jour. That makes us a super-legislature, not a court. Cf. The
Federalist No. 47 (James Madison) (quoting 1 Montesquieu,
Spirit of the Laws 181 (1750)) (“Were the power of judging
joined with the legislative, the life and liberty of the subject
would be exposed to arbitrary control, for the judge would
then be the legislator.”). We should have fixed this problem
by reviewing this case en banc and articulating a clear
organizational standing doctrine grounded in Article III and
the standing principles respected by our courts since the
Founding.
92 EAST BAY SANCTUARY COVENANT V. BIDEN
III.
Compounding the error of improperly expanding
organizational standing, our court also substantially
misreads the asylum law’s text. The prior administration
promulgated the Rule under the grant of authority Congress
conferred on the Executive Branch to adopt limitations on
asylum eligibility. See 8 U.S.C. § 1158(b)(2)(C). But
instead of acceding to the express will of Congress and the
Executive, we enjoin the Rule by manufacturing
prohibitions in the text to avoid what we believe would be
an “absurd” policy.
We start, as always, with the text of the statute. United
States v. Graf, 610 F.3d 1148, 1166 (9th Cir. 2010). The
asylum statute is divided into two relevant parts:
• Section 1158(a) – “Authority to apply for asylum”
– provides that “any alien” who arrives in the United
States may apply for asylum “whether or not [the
alien arrives] at a designated port of arrival.” 8
U.S.C. § 1158(a)(1).
• Section 1158(b) – “Conditions for granting
asylum” – permits the grant of asylum to an alien
who meets the definition of a “refugee” and
authorizes the Attorney General to “establish” by
regulation “additional limitations and conditions,
consistent with this section, under which an alien
shall be ineligible for asylum.” Id. § 1158(b)(2)(C).
By these provisions, Congress endowed the Executive
with broad discretion to regulate eligibility for asylum. The
plain text and structure of the asylum statute indicate that
Congress sought to separate the requirements and
procedures for applying for asylum from granting asylum.
EAST BAY SANCTUARY COVENANT V. BIDEN 93
In subsection (a), Congress tells us who can apply for
asylum. Subject to some exception, “any alien” has
authority to do so, regardless of the alien’s manner of entry
into the country. See id. § 1158(a)(1)–(2). In subsection (b),
Congress tells us who is eligible to be granted asylum. See
id. § 1158(b)(1)–(2). On that front, § 1158(b) broadly
authorizes the Attorney General to promulgate regulations
establishing conditions on eligibility for asylum. Id.
§ 1158(b)(2)(C). Accordingly, while allowing “any alien”
to apply for asylum with little limitation, Congress also
affords the Executive extensive authority to regulate who
receives asylum. Importantly, the sole textual limitation on
Congress’s grant of authority is that any regulation be
“consistent with” the rest of § 1158.
The Rule easily fits within the asylum scheme enacted
by Congress. The Rule does nothing to limit the ability of
aliens to apply for asylum and only restricts the parameters
for a successful asylum petition. Under a straightforward
reading of § 1158, the Executive was permitted to do so.
Contrary to the panel majority’s decision here, the Rule does
not conflict with any provision of § 1158. The bifurcation
of the statute indicates that the two authorities—applying for
asylum and receiving asylum—are purposefully distinct.
The Executive cannot interfere with the former but can
expressly alter the latter. And it is perfectly “consistent” to
allow an alien to apply for asylum generally and to restrict
asylum eligibility to only a subset of those who apply. This
is especially so where, as here, the restriction is of limited
time and scope. 5 See 83 Fed. Reg. 57,661 §§ 1, 2 (confining
5
Contrary to the panel’s opinion here—and fatal to its analysis—
the Rule doesn’t serve as a complete or categorical ban on eligibility for
all aliens arriving outside of a port of entry. See East Bay, 950 F.3d at
1272 (describing the Rule as a “categorical ban”).
94 EAST BAY SANCTUARY COVENANT V. BIDEN
the Rule to a 90-day period along the southern border). And
the temporary restriction was promulgated to respond to a
specific crisis at the border. See East Bay II, 932 F.3d at 754
(describing the asylum system as “overburdened” and
acknowledging that courts were receiving a “staggering
increase in asylum applications,” with a backlog
“exceed[ing] 200,000,” resulting in “thousands of [asylum]
applicants who had been detained by immigration authorities
[being] released into the United States”). Accordingly,
nothing in the plain language or structure of § 1158
precludes temporarily conditioning eligibility for asylum on
the alien’s manner of entry—especially given the
Executive’s concern for the situation at the border.
Despite the clear grant of discretionary authority to the
Executive, our court strikes down the Rule by inventing a
conflict with Congress’s supposed intent. See East Bay, 950
F.3d at 1272–73 (holding the Rule “unlawful” because “it
conflicts with the plain congressional intent”). As explained
above, the Rule is wholly consistent with § 1158(a); yet the
panel professes that it “would be hard to imagine a more
direct conflict than the one presented here.” Id. at 1272
(simplified).
The Supreme Court has warned us against going beyond
the “plain language” of the statute and “read[ing] into the
text additional limitations designed to narrow the scope” of
the statute. Millbrook v. United States, 569 U.S. 50, 55
(2013). By grafting § 1158(a)’s broad scope of who can
apply for asylum onto § 1158(b)’s necessarily narrower
scope of who may (or may not) receive asylum, we do just
that.
The panel’s reading essentially conflates applying for
and receiving asylum. But just because an alien can apply
for asylum does not entitle him or her to receive asylum.
EAST BAY SANCTUARY COVENANT V. BIDEN 95
Instead, whether an alien is ultimately granted asylum is
purely a matter of executive grace. See 8 U.S.C.
§ 1158(b)(1)(A) (establishing that the Executive “may grant
asylum”). 6 Indeed, the government is authorized to deny a
bid for asylum even if the “alien is a refugee within the
meaning of [the statute].” Id. § 1158(b)(1)(A), (b)(2)(A).
All the more baffling, the panel even acknowledges that the
Executive may consider an alien’s manner of entry when
adjudicating an individual asylum case. See East Bay, 950
F.3d at 1273 (“[We] have long recognized that a refugee’s
method of entering the country is a discretionary factor in
determining whether the migrant should be granted
[asylum].”). If the Executive can consider manner of entry
in individual cases, it stands to reason it can do so more
broadly as well. Cf. Lopez v. Davis, 531 U.S. 230, 243–44
(2001) (rejecting argument that the Bureau of Prisons was
required to make “case-by-case assessments” of eligibility
for sentence reductions and explaining that an agency “is not
required continually to revisit ‘issues that may be established
fairly and efficiently in a single rulemaking’”) (simplified).
The panel also justified its departure from the plain text
by arguing that this reading would lead to “absurd results.”
950 F.3d at 1272 (“Explicitly authorizing a refugee to file an
asylum application because he arrived between ports of
entry and then summarily denying the application for the
6
See Moncrieffe v. Holder, 569 U.S. 184, 187 (2013) (describing
grants of asylum as a “form[] of discretionary relief”); INS v. Aguirre–
Aguirre, 526 U.S. 415, 420 (1999) (explaining that the “decision whether
asylum should be granted to an eligible alien is committed to the
Attorney General’s [and the Secretary’s] discretion” under the INA); INS
v. Cardoza-Fonseca, 480 U.S. 421, 443 (1987) (holding that an alien
who satisfies asylum requirements under the INA “does not have a right
to remain in the United States; he or she is simply eligible for asylum, if
the Attorney General, in his discretion, chooses to grant it”).
96 EAST BAY SANCTUARY COVENANT V. BIDEN
same reason borders on absurdity.”). The “absurdity canon
isn’t a license for us to disregard statutory text where it
conflicts with our policy preferences.” Tamm v. UST–U.S.
Trustee, Honolulu (In re Hokulani Square, Inc.), 776 F.3d
1083, 1088 (9th Cir. 2015). And frankly, there is no
absurdity here at all. The Rule says that everyone who
arrives outside of a port of entry is able to apply for asylum,
but because of the identified migrant crisis, for the 90-day
period at the southern border, applicants must come in
through a port of entry to successfully gain asylum. While
the panel majority may disagree with that policy decision,
there is nothing absurd about it. Indeed, § 1158 sets
numerous categorical exclusions from asylum eligibility for
aliens who are statutorily authorized to apply for asylum.
See 8 U.S.C. § 1158(b)(2).
If Congress wanted to preclude the Executive from
basing asylum decisions on the manner of the alien’s arrival,
it could have easily said so. Simmons v. Himmelreich, 136
S. Ct. 1843, 1848 (2016) (“[W]e presume Congress says
what it means and means what it says.”). But here, we have
supplied the policy preference. Although the panel majority
purports to divine Congress’s intent, that is beside the point.
Based on the text of the statute, the asylum-application
provision expressly authorizes aliens who enter between
ports of entry to file asylum applications, while the asylum-
granting provision contains no such guarantee. Instead of
trying to extract the intent of some 535 legislators in enacting
§ 1158, we should have simply adhered to the plain meaning
of the text.
We should have been especially reluctant to exercise
judicial power to circumvent the political branches’
implementation of immigration law. Just last term, the
Supreme Court explained,
EAST BAY SANCTUARY COVENANT V. BIDEN 97
[T]he power to admit or exclude aliens is a
sovereign prerogative[.] [T]he Constitution
gives the political department of the
government plenary authority to decide
which aliens to admit, and a concomitant of
that power is the power to set the procedures
to be followed in determining whether an
alien should be admitted.
Thuraissigiam, 140 S. Ct. at 1982 (simplified).
The Rule, therefore, rests on solid statutory ground.
Congress permits the Attorney General to establish by
regulation “additional limitations and conditions” on asylum
eligibility. 8 U.S.C. § 1158(b)(2)(C). The Attorney General
did just that. That should have been the end of the analysis.
Nevertheless, we substitute our version of the asylum law for
the one actually passed by Congress.
IV.
For these reasons, I respectfully dissent from the denial
of rehearing en banc.
VANDYKE, Circuit Judge, dissenting from denial of
rehearing en banc:
I agree with Judge Bumatay’s analysis and join his
dissent in full. I write separately to address the East Bay
panel majority’s amended decision—specifically its past and
present treatment of the binding effect of motions panels’
published opinions—and the tortured path that got us here.
It was never necessary in this case for the panel majority to
wade into this issue at all (more on that later), but both the
98 EAST BAY SANCTUARY COVENANT V. BIDEN
panel’s original and now-amended analyses are equally
unsatisfying and will undoubtedly lead to more mischief on
an important issue that we will wrestle with for years to
come. Although the revised East Bay panel opinion now no
longer purports to overrule Lair v. Bullock, 798 F.3d 736
(9th Cir. 2015)—which, of course, it never had the authority
to do in the first place—the replacement rationale it provides
is no less troublesome, and can only serve as an ever-ready
escape valve for merits panels who wish to disregard a
published motions panel decision that decided effectively
identical issues.
I.
A.
Some table-setting is necessary. First, the law. Before
this case, our law-of-the-circuit rule vis-à-vis a motions
panel’s published opinion was well-established and
straightforward. In Lair v. Bullock, a panel of our court held
as a matter of first impression that “a motions panel’s
published opinion binds future panels the same as does a
merits panel’s published opinion.” 798 F.3d at 747. There’s
not much ambiguity in that statement.
But what about our law-of-the-case jurisprudence? Does
that separate doctrine somehow undermine Lair’s clear
ruling by allowing a merits panel to avoid being bound by a
motions panel’s published opinion in the same case? That
question was resolved by our court years before Lair. “No,”
we said, en banc. See Gonzalez v. Arizona, 677 F.3d 383,
389 n.4 (9th Cir. 2012) (en banc) (“We now hold that the
exceptions to the law of the case doctrine are not exceptions
to our general ‘law of the circuit’ rule, i.e., the rule that a
published decision of this court constitutes binding authority
which ‘must be followed unless and until overruled by a
EAST BAY SANCTUARY COVENANT V. BIDEN 99
body competent to do so.’”) (quoting Hart v. Massanari,
266 F.3d 1155, 1170 (9th Cir. 2001)).
Taken together, Lair and Gonzalez were therefore clear
and categorical: a motions panel’s published opinion is the
law of the circuit—like any other published panel opinion—
for all subsequent panels, even the merits panel in the same
appeal. 1
B.
Next, this case. On October 1, 2019, the same merits
panel heard and submitted this and another high-profile case,
both challenging Trump Administration immigration
policies. See E. Bay Sanctuary Covenant v. Trump (“East
Bay”), 950 F.3d 1242 (9th Cir. 2020); Innovation Law Lab
v. Wolf (“ILL”), 951 F.3d 1073 (9th Cir. 2020), cert. granted,
141 S. Ct. 617 (2020). In both cases, district courts had
enjoined the government from carrying out the
Administration’s policies, and the government applied to
this court for emergency stays of those injunctions pending
appeal. In this case (East Bay), a motions panel of our court
denied the government’s stay request in a published opinion.
See E. Bay Sanctuary Covenant v. Trump, 932 F.3d 742, 755
(9th Cir. 2018). But in ILL, a different motions panel granted
the government’s stay request, again in a published opinion.
See Innovation Law Lab v. McAleenan (“ILL (motions
panel)”), 924 F.3d 503, 510 (9th Cir. 2019).
Both cases proceeded to be heard by the same panel
considering the merits of the district courts’ preliminary
1
It’s worth emphasizing up front that I’m somewhat ambivalent as
to whether the Lair-Gonzalez rule was the correct rule. But right or
wrong, it clearly was the rule in our circuit, and if we wanted to change
it, we should have done so en banc.
100 EAST BAY SANCTUARY COVENANT V. BIDEN
injunctions. That panel published decisions in both cases on
the same day: February 28, 2020. In East Bay, a two-judge
majority on the merits panel agreed with the motions panel
and affirmed the district court’s injunction. In ILL, the same
merits panel majority also affirmed the district court’s
injunction. 2 But the merits panel in ILL had a Lair problem:
the motions panel’s published decision had reached a
directly contrary result on at least one major issue in the case.
Compare, e.g., ILL (motions panel), 924 F.3d at 509 (“[W]e
conclude that DHS is likely to prevail on its contention that
§ 1225(b)(1)” is consistent with the Administration’s
policy.), with ILL, 951 F.3d at 1081 (“We conclude that
plaintiffs have shown a likelihood of success on their claim
that [the policy] is inconsistent with 8 U.S.C. § 1225(b).”).
So the panel majority resourcefully applied some good
old-fashioned judge-jitsu, admirable perhaps for its audacity
if not its propriety or subtlety. The majority first turned to
East Bay—where Lair was both not at issue (because the
motions and merits panels in East Bay reached the same
conclusion) and, for that reason, not particularly relevant—
to create a new rule effectively toppling Lair and
neutralizing its thorny imposition on the panel majority’s
desired outcome in ILL. In East Bay, the majority’s opinion
announced a new rule: “we treat the motions panel’s
decision as persuasive, but not binding.” East Bay, 950 F.3d
at 1265. The decision simply asserted that Lair’s holding
was dicta, and that even if it were precedential, there were
“good policy and practical reasons for departing from
2
Judge Fernandez, the third judge on both merits panels, wrote
separately in both cases because, applying Lair, he believed the merits
panel in both cases was bound by the earlier published motions panel
opinions. East Bay, 950 F.3d at 1284 (Fernandez, J., concurring in the
result); ILL, 951 F.3d at 1095 (Fernandez, J., dissenting).
EAST BAY SANCTUARY COVENANT V. BIDEN 101
Lair[],” including law-of-the-case considerations. See id.
at 1262–63.
That alone was remarkable for its chutzpah, considering
it casually disregarded one of our court’s foundational rules
undergirding the rule of law itself: that three-judge panels
cannot overturn prior panel opinions unless they are “clearly
irreconcilable with the reasoning or theory of intervening
higher authority.” Miller v. Gammie, 335 F.3d 889, 893 (9th
Cir. 2003) (en banc); see also United States v. Contreras,
593 F.3d 1135, 1136 (9th Cir. 2010) (en banc). But a full-
blown critique of this rationale is unnecessary now because
the majority has amended its opinion and thereby withdrawn
some more offensive aspects of its prior analysis. To
summarize, though, the East Bay majority’s curt analysis in
its initial opinion rather gruffly dismissed Lair and merely
paid lip service to Gonzalez, which is why it was later forced
to amend its opinion.
But the real head-scratcher for the original East Bay
decision was why the majority promulgated its new rule at
all. If you were looking only at East Bay itself, you would
wonder what prompted the majority there to even delve into
the Lair discussion. Again, in East Bay, the motions panel
and the merits panel were in vigorous agreement that the
government was unlikely to succeed on the merits. So why
drag Lair into it?
The reason is obvious once you zoom out: because of
ILL, the other case decided by the same panel on the same
day. In ILL, the same two-judge majority needed to navigate
around a published motions panel decision that went the
other way. See ILL, 951 F.3d at 1081 (“A preliminary
question is whether a merits panel is bound by the analysis
of a motions panel on a question of law, performed in the
course of deciding an emergency request for a stay pending
102 EAST BAY SANCTUARY COVENANT V. BIDEN
appeal.”). And how did the ILL majority answer that
question? By dutifully following precedent: “On that
question, we follow East Bay…in which we held that a
motions panel’s legal analysis, performed during the course
of deciding an emergency motion for a stay, is not binding
on later merits panels.” Id. The only thing missing from the
majority’s ILL decision is a solemn citation to Miller v.
Gammie so everyone can know how serious the majority was
about faithfully adhering to binding precedent—its own, in
East Bay, that is. 3
To summarize ILL: the motions panel in a published
opinion addressed the same legal questions presented to the
merits panel, and under Lair and Gonzalez, the motions
panel’s resolution of those questions should have controlled.
But the merits panel majority didn’t like the result reached
by the motions panel. It needed that published opinion to be
3
The panel majority defends its original East Bay decision rejecting
Lair by reaching for the oldest excuse in the Book (literally): “But we
were asked to do it!” See Genesis 3:12 (NIV) (“The man [Adam] said,
‘The woman you put here with me—she gave me some fruit from the
tree, and I ate it.’”). The obvious response is no less antediluvian, and
will be familiar to any parent: just because you were asked to do
something doesn’t mean you should have. Indeed, here the panel
majority has implicitly acknowledged it shouldn’t have done what it
originally did, by amending its opinion to remove the more offensive
aspects of its original rationale. So even assuming the parties had asked
the panel to do what it did, that wouldn’t excuse it. But, in fact, the panel
majority can’t fairly blame the parties. The panel in this case only
needed to address Lair if it departed from the motions panel’s
conclusions, which it didn’t. The government didn’t even cite Lair,
much less argue Lair’s rule was dicta, until a footnote in its reply brief.
The panel majority purported to nullify Lair in its original opinion only
because it needed Lair’s rule gone for ILL—not because Lair stood in
the way of the panel’s conclusion in this case (which the panel’s
amended opinion demonstrates, by reaching the same conclusion but
without dissing Lair).
EAST BAY SANCTUARY COVENANT V. BIDEN 103
“persuasive,…not binding.” East Bay, 950 F.3d at 1265. So
the majority overruled Lair in East Bay—where it was
completely unnecessary to do so—so that it could
immediately apply East Bay’s ILL-conceived rule in a
decision where it could do some real work. 4
Back to East Bay: after the government’s request for en
banc rehearing, the majority agreed to amend its opinion; a
majority of our court was unwilling to vote in favor of en
banc rehearing; and we arrived at our current juncture. The
East Bay panel majority’s amended opinion has now nixed
its categorical rejection of Lair. It replaces its prior
repudiation of Lair with one of the oldest common law
axioms: a “published motions panel order may be binding as
precedent for other panels deciding the same issue.” It
vaguely concedes that “[t]here may be circumstances where
a motions panel does answer the same legal question that is
presented to the merits panel.” But, as you might have
guessed, the majority still concludes that in East Bay itself,
the motions panel opinion “is not binding…because the
issues are different.” That way, you see, East Bay can still
purport to serve as precedent for ILL and subsequent
decisions that have involved Trump Administration policies
4
Along with the same policy arguments outlined in East Bay, the
ILL majority noted that its disregard of the motions panel’s earlier
published decision was especially appropriate in ILL because, in that
case, one of the ILL motions panel judges had “expressed the hope that
the merits panel, with the benefit of full briefing and argument, would
decide the legal questions differently.” ILL, 951 F.3d at 1081. For those
not tracking the judicial composition of the various panels, that same
motions panel judge was also part of the two-judge majority on the East
Bay and ILL merits panels, thus triggering the venerable “I-told-you-so”
rationale for disregarding precedent.
104 EAST BAY SANCTUARY COVENANT V. BIDEN
where the merits panel disagreed with the published motions
panel opinion. 5
In just a moment, I’ll discuss the deficiencies of the now-
amended East Bay opinion’s seemingly innocuous and banal
reasoning, and why we still should have taken this case en
banc despite the panel’s retrenchment. But first, it’s
important to recap what happened here. Originally, the East
Bay majority adopted a clearly indefensible rationale to
fabricate a new rule it could apply in a different case, where
it needed to circumvent a motions panel opinion that decided
the very same issue of law. See ILL, 951 F.3d at 1081
(reaching a directly contrary legal conclusion to that of the
motions panel on whether the executive branch’s policy was
consistent with the relevant statute). After being called to
the carpet by a petition for en banc rehearing, the majority
stepped back from that blatantly inappropriate power-grab
behind its original East Bay rule. Of course, that’s easy
enough to do now. The original East Bay rule has already
served its purpose: the majority got to its preferred outcome
in ILL by applying the precedent-disregarding “rule” it
manufactured in East Bay. Discarding the old East Bay rule
now, after the fact, and in a case where the “rule” was
complete dicta to begin with, is pretty much costless.
Mischief managed.
5
At least one other merits panel of our court followed ILL’s lead
when it declined to be bound by a published motions panel opinion. That
case involved the Public Charge Rule, where, as discussed further below,
the motions panel reached legal conclusions contrary to those of the
merits panel majority on virtually every question in the case. Compare
generally City & County of San Francisco v. USCIS, 981 F.3d 742 (9th
Cir. 2020) (merits panel), with City & County of San Francisco v. USCIS,
944 F.3d 773 (9th Cir. 2019) (motions panel).
EAST BAY SANCTUARY COVENANT V. BIDEN 105
Or maybe not. In the real world, the lingering effects of
pulling a stunt like this don’t just vanish. Even putting aside
the problems identified in Judge Bumatay’s dissent, and the
problems with East Bay’s new replacement rationale
discussed in the next section, this looks bad for our court—
because it is. Even if none of the other reasons for taking
this case en banc had merit (which they do), we should have
done so if only to demonstrate to the public that we don’t
approve of the barely disguised shenanigans that happened
in East Bay and ILL. It would have helped promote public
confidence in the integrity and impartiality of our court. 6
II.
It is true that the amended rationale in the East Bay
majority’s opinion doesn’t represent the ham-fisted
upheaval its original opinion did. But it merely trades one
evil for another. Instead of rudely casting aside precedent, it
distinguishes—rather than overrules—Lair, creating a new
standardless standard that, as a practical matter, will allow
any merits panel to disregard a motions panel’s published
decision resolving virtually identical claims. In addition to
the several compelling reasons given in Judge Bumatay’s
dissent as to why we should have taken this case en banc,
6
The panel majority misconstrues what I characterize as its
“mischief.” I agree “there was nothing improper in the majority’s
revision of its opinion to accommodate the views of our colleagues.”
The “mischief” I reference was the panel’s “ILL-conceived” rule from
its original opinion—not the panel “amending its opinion during the en
banc process.” My point is that while the panel majority (and apparently
a majority of our court) may think that the panel has “managed” its
original mischief by amending its opinion, it hasn’t—especially when
you consider the indiscernible mess left by the panel’s new but still
unsatisfying rationale, combined with our other recent decisions that I
discuss in detail below.
106 EAST BAY SANCTUARY COVENANT V. BIDEN
there are at least two more considerations—related to each
other—that further counseled in favor of reviewing the
revised East Bay opinion.
First, the East Bay majority’s new standard is
indeterminate at best, retreating to the truism that two things
are different, except when they’re not: “The published
motions panel order may be binding as precedent for other
panels deciding the same issue, but it is not binding here.”
This nebulous standard is completely unhelpful to future
panels attempting to apply our precedent. It is only helpful
if we want to maximize merits panels’ discretion to do
whatever they want with published motions panels’ opinions
in the same case. It is terrible for predictability and the rule
of law, and should have been considered en banc for this
reason alone.
Second, and relatedly, although there might be one
logically satisfying way to reconcile East Bay’s new
“sometimes-it’s-binding, sometimes-it’s-not” rule with Lair,
that specific reconciliatory interpretation of the East Bay
majority’s rule obviously cannot be what the East Bay
majority actually meant. How do we know? Because of
what the judges from the East Bay majority have done in
subsequent cases. Their clear actions prevent their mushy
words in East Bay from meaning the only thing that might
have been meaningful.
So after East Bay we really are left with a truly
ambiguous standard giving merits panels maximum
discretion and minimal guidance. To explain requires me to
first elaborate on how one could try to reconcile Lair and the
new East Bay, and then describe how the panel majority’s
actions in subsequent cases make clear that that can’t be
what East Bay means.
EAST BAY SANCTUARY COVENANT V. BIDEN 107
Here is how an intrepid law professor might try to
reconcile Lair and East Bay: The amended East Bay opinion
no longer rejects Lair out-of-hand, but nods to Lair while
distinguishing it by retaining a lengthy but abstract
discussion about the differences between adjudicating stay
applications and preliminary injunctions. The opinion
engages in a superficial and formalistic analysis of the
technical distinctions between stays and preliminary
injunctions, but never pins down what those differences
might mean here in the real world. Once we dip below
30,000 feet, however, we observe that whether a motions
panel grants or denies a stay should have very different
implications for a subsequent panel considering the merits of
the preliminary injunction itself. And therein lies a
theoretical key to reconciling Lair and East Bay.
To see how the stay and preliminary injunction standards
interact differently depending on whether a stay is granted
or denied requires looking closely at the standards
themselves. A party seeking a preliminary injunction before
a district court bears the burden of establishing:
[1] that he is likely to succeed on the merits,
[2] that he is likely to suffer irreparable harm
in the absence of preliminary relief, [3] that
the balance of equities tips in his favor, and
[4] that an injunction is in the public interest.
Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 20 (2008).
By comparison, after a district court issues a preliminary
injunction, the losing party seeking a stay of that injunction
bears the considerably heavier burden of establishing:
(1) [that] the stay applicant has made a strong
showing that he is likely to succeed on the
merits; (2) [that] the applicant will be
108 EAST BAY SANCTUARY COVENANT V. BIDEN
irreparably injured absent a stay; (3) [that]
issuance of the stay will [not] substantially
injure the other parties interested in the
proceeding; and (4) [that] the public interest
lies [with the issuance of a stay].
Nken v. Holder, 556 U.S. 418, 426 (2009) (emphases added)
(citation omitted). And the stay applicant doesn’t present its
arguments in a vacuum, but must overcome a presumptively
valid lower court decision that went the other way. See Sw.
Voter Registration Educ. Project v. Shelley, 344 F.3d 914,
918 (9th Cir. 2003) (explaining that appellate review of
district court preliminary injunctions is “limited[,] and
deferential”). On top of that, the stay applicant bears the
substantial burden of strongly showing that it is the party
who is likely to succeed on the merits. And the applicant
must establish it “will be irreparably injured absent a stay.”
Nken, 556 U.S. at 434 (emphasis changed). Compare those
showings to the less demanding “likely to succeed on the
merits” and “likely to suffer irreparable harm” of the
preliminary injunction inquiry. Given the higher standards
and inverted burdens, it is no surprise that stay applications
are (and should be) denied most of the time; they are granted
even more infrequently than are preliminary injunctions,
which are themselves an “extraordinary remedy.” Winter,
555 U.S. at 22.
Because of that, losing a stay motion (even in a published
opinion) 7 logically should not usually foreordain that party’s
fate on the merits of the preliminary injunction. A stay
applicant unable to meet the onerous stay standard could still
7
Most motions panel decisions aren’t published, and therefore
cannot tie the hands of later merits panels. See Haggard v. Curry,
631 F.3d 931, 933 n.1 (9th Cir. 2010) (per curiam); 9th Cir. R. 36-3(a).
EAST BAY SANCTUARY COVENANT V. BIDEN 109
prevail in reversing a district court preliminary injunction,
since the burden before the merits panel flips back to the
original injunction proponent. See id. at 20. In most cases,
like East Bay, the interplay of the shifting stay and injunction
standards means that a motions panel’s denial of a stay won’t
control the outcome of the later panel just considering the
merits of the preliminary injunction itself. So, if it could be
limited to that context, the amended East Bay opinion’s
observation that published stay opinions are not always
binding on later merits panels could make some sense. It
also might not be inconsistent with Lair, because unlike in
East Bay, Lair involved a situation where a stay was granted
by the motions panel.
In a Lair-type circumstance—that is, in the doubly rare
case where the motions panel granted a stay of the
preliminary injunction in a published opinion—the stay
applicant is necessarily very likely to succeed on all the legal
factors relevant to the preliminary injunction itself. If a
motions panel concludes that an applicant has demonstrated
(1) a strong likelihood of success on the merits, (2) that it
will be irreparably harmed absent a stay, (3) that other parties
won’t be substantially harmed by the stay, and (4) that the
public interest is served by a stay, then a fortiori the other
party will almost always be prevented from meeting the
considerably less burdensome preliminary injunction
standard—unless the merits panel is allowed to just
disregard the motions panel’s conclusions. Both parties
can’t likely succeed on the merits. Both can’t show that the
equities and public interest tip in their favor. That’s simply
not how the standards logically interact. If, in a published
decision, a motions panel concludes that the stay applicant
surmounted its Sisyphean burden—despite the odds stacked
against it—that’s generally a conclusive legal determination
110 EAST BAY SANCTUARY COVENANT V. BIDEN
that the applicant should also prevail on the merits, and so a
preliminary injunction is unwarranted. 8
One might conclude, therefore, that, because Lair
involved precisely this rare situation of a merits panel faced
with a published motions panel decision granting a stay, this
is the circumstance Lair had in view for its categorical
motions-panels-opinions-bind-later-merits-panels rule.
And, in contrast, East Bay’s new observation that motions
panel decisions don’t always bind later merits panels is
simply a reflection of the reality in that case—that the earlier
motions panel had denied the stay. Accordingly, one could
argue that the East Bay panel majority’s focus on the
distinctions between stays and preliminary injunctions was
intended to serve as precisely this basis for reconciling East
Bay and Lair—i.e., that East Bay applies if the motions panel
denied the stay, but Lair applies if the motions panel granted
the stay.
But here in the real world, whoever makes this otherwise
possibly satisfying academic argument is delusional. The
East Bay panel majority can’t have meant to reconcile Lair
and East Bay this way because the very same two-judge
panel majority immediately cited and relied on East Bay to
ignore a motions panel opinion granting a stay in ILL, and
has never retreated from that. In fact, one of those two
judges did the same thing again months later as part of the
8
The exception might arise in a more traditional case where, after a
stay is granted, the parties further develop the record in front of the
district court before having the merits of their preliminary injunction
considered on appeal. In that case, a merits panel might actually be
reviewing a meaningfully different case than the motions panel did. But
as the procedural histories of the various cases recounted herein
illustrate, the development of our caselaw in this area does not seem to
be driven much by “traditional” cases—at least not recently.
EAST BAY SANCTUARY COVENANT V. BIDEN 111
merits panel majority in the Public Charge cases, which also
involved a published motions panel opinion granting a stay
that was subsequently ignored by the two-judge merits panel
majority. So that theoretically reconciliatory reading of Lair
and East Bay, as nice as it might be, just isn’t on the table—
unless we want to ignore reality.
As if ILL were not bad enough, our merits panel decision
in the Public Charge cases highlights just how unstructured
our treatment of published motions panel stay opinions has
become. In three recent cases argued and decided together
by this court, challengers to the Trump Administration’s
revamped “Public Charge” rule obtained preliminary
injunctions from two district courts in our circuit. The
government sought an emergency stay of these injunctions,
which a motions panel of our court granted in a lengthy
published opinion. According to the motions panel, the
government “mustered a strong showing of likelihood of
success on the merits.” City & County of San Francisco, 944
F.3d at 807. More specifically, the motions panel concluded:
• The new rule was not contrary to law. See id. at 798
(“Congress has not spoken directly to the
interpretation of ‘public charge’ in the INA. Nor did
it unambiguously foreclose the interpretation
articulated in the Final Rule.”).
• The new rule was not arbitrary and capricious. See
id. at 790.
• The government would suffer irreparable harm
absent a stay of the injunctions. See id. at 806.
• A stay would serve the public interest. See id. at 807.
112 EAST BAY SANCTUARY COVENANT V. BIDEN
Contrast these clear holdings with those of the two-judge
merits panel majority, which entirely reversed course on the
same record and issues. City & County of San Francisco,
981 F.3d at 763. The merits panel majority concluded:
• “[T]he plaintiffs have demonstrated a high likelihood
of success in showing that the Rule is inconsistent
with any reasonable interpretation of the statut[e].”
Id. at 758.
• “We must conclude that the Rule’s promulgation was
arbitrary and capricious.” Id. at 762.
• Plaintiffs established that they will likely suffer
irreparable harm absent the preliminary injunctions.
See id.
• The public interest and equities favor the injunctions.
See id.
Same case, same legal issues, no interstitial factual
developments—diametrically opposed legal conclusions. 9
The merits panel majority in the Public Charge appeal
explained that it could depart from the carefully-reasoned
holdings of the motions panel because (a) it had the benefit
of more briefing on the same issues (including amicus
briefs); and (b) some other circuits had since reached
conclusions contrary to those of the motions panel. See id.
at 753–54. Because our Public Charge decision never
9
As a result, we now have on-point, circuit precedent in two
different published opinions (from the motions panel and the merits
panel) that reach opposite conclusions in the Public Charge Rule cases.
Which binding precedent should the lower courts, and panels in
subsequent cases, follow? Addressing this problem is another reason our
court should have taken this case en banc.
EAST BAY SANCTUARY COVENANT V. BIDEN 113
mentioned East Bay or ILL (or Lair or Gonzalez, or, for that
matter, any precedent for disregarding the motions panel’s
published legal conclusions), it isn’t clear whether the panel
majority was implicitly applying the East Bay/ILL rule, 10 or
whether it was creating yet another novel rule that motions
panels’ opinions are not binding whenever there is a lot more
merits briefing—including some really good amicus
briefs—and other circuits have reached a different
conclusion than did our motions panel. 11
As these cases illustrate, the undeniable reality is that
after Gonzalez, Lair, East Bay, ILL, and City & County of
San Francisco have all been thrown in the mixer, our circuit
doesn’t have anything close to a cognizable rule about how
merits panels should treat motions panels’ earlier published
stay opinions. Perhaps, sitting en banc, we could adopt the
approach described above reconciling Lair and East Bay. Or
we could make a rule that motions panels’ decisions granting
a stay should always be unpublished, and thus not binding.
Or, amending our earlier Gonzalez decision, we could
conclude that motions panels’ decisions—published or
not—are not binding on later panels in the same appeal
(including merits panels), and address how our court and the
10
As already noted, the Public Charge merits panel majority
included an overlapping judge from the East Bay and ILL panel majority,
so the Public Charge panel majority could hardly have been unaware of
the East Bay rule, or Lair.
11
Of course, if other circuits’ intervening contrary decisions were
actually a legitimate consideration in deciding whether to apply binding
circuit precedent, then so might be the fact that the U.S. Supreme Court
had stayed those very same contrary decisions. See Dep’t of Homeland
Sec. v. New York, 140 S. Ct. 599, 599 (2020) (staying Second Circuit’s
preliminary injunction); Wolf v. Cook County, 140 S. Ct. 681, 681 (2020)
(same for Seventh Circuit).
114 EAST BAY SANCTUARY COVENANT V. BIDEN
district courts in other cases should treat inconsistent
published motions and merits opinions from the same case.
Our court sitting en banc might come up with any number of
good solutions to the problems presented in cases like this.
But until we do, we and those who litigate before us must
content ourselves with the slop at the intersection of East
Bay, ILL, City & County of San Francisco, Lair, Gonzalez—
and whatever future panel opinions applying those divergent
precedents deliver. It’s a pity that a majority of our court
could not be persuaded to take this case en banc to untangle
the muddle we’ve made of the Lair–Gonzalez rule.
III.
Recognizing the amended East Bay rationale for what it
is—a plastic truism that, without now saying so, still
subversively displaces Lair’s rule—it’s hard to see how
anyone could think this is a good thing for the state of the
law in our circuit. Going forward, the hyper-elasticity of the
new Lair–East Bay standard will undoubtedly be useful for
merits panels seeking flexibility, but obviously less helpful
to litigants seeking clarity and predictability. This problem
is not going to go away. Given our circuit’s propensity to be
the venue of choice to overrule certain administrations’
policies, we are sure to see this problem manifest again (and
again) before too long. Until we fix it, everybody will know
that whether a motions panel’s published stay opinion has
any binding force just depends on which merits panel you
draw—not the law. Cite Lair. Cite East Bay. Cite ILL or
City & County of San Francisco. Trust us to pick which one
fits best.
That breaks faith with our most fundamental obligation,
which is to apply the law, not our preferences. Panel
majorities should not be allowed to simply change the rules
when they find them inconvenient obstacles to preferred
EAST BAY SANCTUARY COVENANT V. BIDEN 115
substantive outcomes. In the end, this type of overt results-
oriented judging can’t help but discredit our court and the
rule of law generally.
Any of these considerations should individually have
been important enough to correct East Bay en banc. But
apparently, even in toto, they were not. I must therefore
respectfully express my disappointment.