Manuel Olivas-Motta v. Matthew Whitaker

FOR PUBLICATION

UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT

MANUEL JESUS OLIVAS-MOTTA, No. 14-70543
AKA Manuel Jesus Olivas-Notta,
Petitioner, Agency No.
A021-179-705
v.

MATTHEW G. WHITAKER, Acting OPINION
Attorney General,
Respondent.

On Petition for Review of an Order of the
Board of Immigration Appeals

Argued and Submitted September 10, 2018
San Francisco, California

Filed December 19, 2018

Before: J. Clifford Wallace, Johnnie B. Rawlinson,
and Paul J. Watford, Circuit Judges.

Opinion by Judge Wallace;
Dissent by Judge Watford
2 OLIVAS-MOTTA V. WHITAKER

SUMMARY *

Immigration

The panel denied Manuel Jesus Olivas-Motta’s petition
for review of a decision of the Board of Immigration Appeals
that found him removable for having committed two crimes
involving moral turpitude.

Olivas-Motta, a lawful permanent resident, was placed
in removal proceedings based on his convictions for felony
endangerment under Arizona Revised Statutes § 13-1201
and facilitation to commit unlawful possession of marijuana
for sale. The immigration judge concluded (and the parties
did not dispute before this court) that the facilitation offense
was a crime involving moral turpitude. The immigration
judge and Board determined that felony endangerment was
neither categorically a crime involving moral turpitude nor a
crime involving moral turpitude under the modified
categorical approach, but examined evidence beyond the
record of conviction and found the offense involved moral
turpitude.

While Olivas-Motta’s petition for review was pending
before this court, the Board published In re Leal, 26 I. & N.
Dec. 20 (B.I.A. 2012) (Leal I), which held that felony
endangerment under Arizona Revised Statutes § 13-1201
was categorically a crime involving moral turpitude, and this
court upheld that determination in Leal v. Holder, 771 F.3d
1140 (9th Cir. 2014) (Leal II). Because the Board had not

*
This summary constitutes no part of the opinion of the court. It
has been prepared by court staff for the convenience of the reader.
 OLIVAS-MOTTA V. WHITAKER 3

decided Olivas-Motta’s appeal on the ground that his offense
was categorically a crime involving moral turpitude, this
court did not consider Leal I’s relevance to Olivas-Motta’s
petition, but granted the petition and remanded because an
immigration judge and the Board are confined to the record
of conviction in this context. On remand, the Board
dismissed Olivas-Motta’s appeal, applying Leal I to
conclude that felony endangerment was categorically a
CIMT.

The panel rejected Olivas-Motta’s argument that Leal II
was wrongly decided, explaining that the panel has no power
to overrule circuit precedent.

The panel also rejected Olivas-Motta’s argument that the
Board’s application of Leal I was impermissibly retroactive.
Concluding that a change in law must have occurred before
this court’s retroactivity analysis from Montgomery Ward &
Co., Inc. v. FTC, 691 F.2d 1322 (9th Cir. 1982), is
implicated, the panel held that Montgomery Ward is only
applicable when an agency consciously overrules or
otherwise alters its own rule or regulation, or expressly
considers and openly departs from a circuit court decision.

Applying this standard to Olivas-Motta’s case, the panel
concluded that there was no change in law raising
retroactivity concerns: before Olivas-Motta pleaded guilty,
the Board had never determined in a precedential opinion
whether felony endangerment under Arizona Revised
Statutes § 13-1201 was a crime involving moral turpitude
and, therefore, the application of the statute was simply
unclear until Leal I.

Olivas-Motta also contended that the Attorney General’s
decision in In re Silva-Trevino, 24 I. & N. 687 (A.G. 2008),
4 OLIVAS-MOTTA V. WHITAKER

abolished the requirement that a crime with a mens rea of
recklessness could not constitute a crime involving moral
turpitude unless the offense presented an aggravating factor.
The panel disagreed, explaining that the aggravating-factor
analysis from earlier cases is harmonious with the later
approach in Silva-Trevino, and concluding that, as to
Arizona felony endangerment, Silva-Trevino did not change
the law.

Next, the panel rejected Olivas-Motta’s argument that,
due to both claim and issue preclusion, the Board could not
revisit on remand whether his offense was categorically a
crime involving moral turpitude. Although Olivas-Motta
had not exhausted the issue of preclusion, the panel
concluded it had jurisdiction to consider the claim because
Olivas-Motta did not have an opportunity to argue it until the
Board issued its second decision. On the merits, the panel
held that there was no error because both forms of preclusion
require the existence of a separate action, but the Board on
remand was acting within the same, single proceeding. The
panel also concluded that the rule of mandate, which
prohibits an agency from deviating from a court’s remand
order, did not foreclose the Board’s consideration of whether
the statute was categorically a crime involving moral
turpitude because nothing in the remand restricted the Board
from considering that issue.

Finally, the panel rejected Olivas-Motta’s argument that
8 U.S.C. § 1227(a)(2)(A)(ii), under which Olivas-Motta was
found to be removable for having committed crimes
involving moral turpitude, is unconstitutionally vague,
explaining that the Supreme Court and this court have
repeatedly rejected that vagueness challenge and that this
panel lacked authority to reconsider this court’s prior
decisions.
 OLIVAS-MOTTA V. WHITAKER 5

Dissenting, Judge Watford disagreed with the panel’s
conclusion that Leal I was not a change in law necessary to
trigger retroactivity analysis. Judge Watford wrote that,
under the Board’s standards for determining which
recklessness offenses were crimes involving moral turpitude
in effect at the time Olivas-Motta pleaded guilty, there was
at least a realistic chance that his endangerment offense
would not be classified as a crime involving moral turpitude.
However, after the Attorney General’s decision in Silva-
Trevino, it was nearly certain that his offense would be
classified as a crime involving moral turpitude, and the
Board’s decision in Leal I eliminated what little uncertainty
remained. According to Judge Watford, this was a change
in the governing standard that attached new legal
consequences to events completed before its enactment.
Applying the retroactivity test from Montgomery Ward,
Judge Watford concluded that the balance of factors weighs
in favor of Olivas-Motta. Judge Watford would grant the
petition for review and remand so that the agency could
conduct the analysis in the first instance.

COUNSEL

K. Lee Hartzler (argued), San Diego, California, for
Petitioner.

Sarah A. Byrd (argued) and Keith I. McManus, Senior
Litigation Counselors; Cindy S. Ferrier, Assistant Director;
Civil Division, United States Department of Justice,
Washington, D.C.; for Respondent.
6 OLIVAS-MOTTA V. WHITAKER

OPINION

WALLACE, Circuit Judge:

An immigration judge (IJ) ordered Manuel Jesus Olivas-
Motta’s removal because he had been convicted of two
crimes involving moral turpitude (CIMTs). The Board of
Immigration Appeals (Board) dismissed Olivas-Motta’s
appeal from the IJ’s order. Olivas-Motta now petitions for
review of the Board’s dismissal. We have jurisdiction under
8 U.S.C. § 1252, and we deny the petition.

I.

Olivas-Motta is a citizen of Mexico who was admitted to
the United States of America as a lawful permanent resident
on or about October 12, 1976. He has since been convicted
of two felonies. On August 11, 2003, he was convicted of
facilitation to commit unlawful possession of marijuana for
sale in violation of Arizona Revised Statutes §§ 13-1004, 13-
3405. On November 26, 2007, he was convicted of felony
endangerment under Arizona Revised Statutes § 13-1201.

On April 2, 2009, the Department of Homeland Security
initiated removal proceedings against Olivas-Motta under
8 U.S.C. § 1227(a)(2)(A)(ii) as an alien convicted of two
CIMTs. The IJ determined, and the parties no longer dispute,
that the facilitation offense was a CIMT. As to the
endangerment offense, the IJ determined that it was neither
categorically a CIMT nor a CIMT under the modified
categorical approach. However, the IJ examined evidence
beyond the record of conviction, including police reports,
and determined that the offense involved moral turpitude.
The IJ then sustained the charge of removal. The Board
relied on the same grounds to conclude that the
 OLIVAS-MOTTA V. WHITAKER 7

endangerment offense was a CIMT and dismissed Olivas-
Motta’s appeal.

Olivas-Motta petitioned for review of the Board’s
decision. While the petition was pending, the Board
published an opinion holding that felony endangerment
under Arizona Revised Statutes § 13-1201 was categorically
a CIMT. In re Leal, 26 I. & N. Dec. 20, 27 (B.I.A. 2012)
(Leal I). We upheld that determination. Leal v. Holder,
771 F.3d 1140, 1148–49 (9th Cir. 2014) (Leal II). But we
declined to consider Leal I’s relevance to Olivas-Motta in
his first petition because the Board had not originally
decided his appeal on the ground that felony endangerment
was categorically a CIMT. Olivas-Motta v. Holder, 746 F.3d
907, 917 (9th Cir. 2013), as amended (April 1, 2014); see
also Ali v. Holder, 637 F.3d 1025, 1029 (9th Cir. 2011)
(confining our review to grounds relied upon by the Board).
Instead, we granted the petition and remanded because “an
IJ and the [Board] are confined to the record of conviction in
determining whether an alien has been convicted of a
CIMT.” Olivas-Motta, 746 F.3d at 908. On remand, the
Board applied Leal I to conclude that felony endangerment
was categorically a CIMT and dismissed Olivas-Motta’s
appeal.

Olivas-Motta again petitions for review of the Board’s
dismissal. He argues that the Board’s application of Leal I
was impermissibly retroactive, that preclusion bars the
Board from reconsidering whether felony endangerment was
categorically a CIMT, and that the phrase CIMT is
unconstitutionally vague.

Olivas-Motta also argues that we are not bound by Leal
II because it was wrongly decided. But this panel has no
power to overrule circuit precedent. Miller v. Gammie,
335 F.3d 889, 900 (9th Cir. 2003) (en banc) (holding that
8 OLIVAS-MOTTA V. WHITAKER

circuit precedent may be overturned only en banc, subject to
exceptions not applicable here).

II.

We review constitutional claims and questions of law de
novo. Latter-Singh v. Holder, 668 F.3d 1156, 1159 (9th Cir.
2012); see also 8 U.S.C. § 1252(a)(2)(C), (D). Whether a
new agency interpretation may be applied retroactively is a
question of law. See Garfias-Rodriguez v. Holder, 702 F.3d
504, 514–15 (9th Cir. 2012) (en banc). Whether preclusion
is available is also a question of law. Oyeniran v. Holder,
672 F.3d 800, 806 (9th Cir. 2012), as amended (May 3,
2012).

III.

When an agency decides to create a new rule through
adjudicatory action, that new rule may apply retroactively to
regulated entities. SEC v. Chenery Corp., 332 U.S. 194, 203
(1947). “[R]etroactivity must be balanced against the
mischief of producing a result which is contrary to a
statutory design or to legal and equitable principles.” Id. “If
that mischief is greater than the ill effect of the retroactive
application of a new standard, it is not the type of
retroactivity which is condemned by law.” Id.

We have applied this rule in the immigration context to
determine whether Board decisions may apply retroactively.
See, e.g., Garfias-Rodriguez, 702 F.3d at 515–23; Miguel-
Miguel v. Gonzales, 500 F.3d 941, 950–53 (9th Cir. 2007).
In such cases, we have relied on the five-factor test set forth
in Montgomery Ward & Co., Inc. v. FTC, 691 F.2d 1322,
1333 (9th Cir. 1982). Olivas-Motta argues that, in this case,
the Montgomery Ward factors strongly counsel against
retroactively applying Leal I to his case, and that the Board
 OLIVAS-MOTTA V. WHITAKER 9

accordingly erred in concluding that Arizona felony
endangerment is categorically a CIMT.

A.

As a threshold matter, we must address whether
retroactivity is implicated by Leal I. The government argues
that a change in law is a prerequisite to Montgomery Ward
balancing, and that we should not conduct a retroactivity
analysis because no change in law occurred. Olivas-Motta
argues that the Montgomery Ward factors themselves
account for whether a change in law has occurred, and that
Montgomery Ward balancing is therefore appropriate
because Leal I was decided after his guilty plea.

We conclude that a change in law must have occurred
before Montgomery Ward is implicated. The requirement
that the law have changed in some way is generally a settled
principle of retroactivity analysis. See James B. Beam
Distilling Co. v. Georgia, 501 U.S. 529, 534 (1991) (“It is
only when the law changes in some respect that an assertion
of nonretroactivity may be entertained”); Morales-Izquierdo
v. Dep’t of Homeland Sec., 600 F.3d 1076, 1090 (9th Cir.
2010), overruled in part on other grounds by Garfias-
Rodriguez, 702 F.3d at 516 (“Montgomery Ward and its
progeny deal with the problems of retroactivity created when
an agency, acting in an adjudicative capacity, so alters an
existing agency-promulgated rule that it deprives a regulated
party of the advance notice to conform its conduct to the
rule”). It would be incongruous to apply a different rule here
because the principles animating a statute’s retroactivity —
“fair notice, reasonable reliance, and settled expectations”
— are equally animating in Olivas-Motta’s immigration
proceedings. See Vartelas v. Holder, 566 U.S. 257, 273
(2012) (quoting Landgraf v. USI Film Prod., 511 U.S. 244,
270 (1994)). Moreover, were we to adopt the rule that
10 OLIVAS-MOTTA V. WHITAKER

Montgomery Ward balancing is required regardless of
whether a change in law has occurred, the mere existence of
a new published decision on an issue would always trigger
retroactivity analysis. This too is contrary to settled law on
this issue. See Manhattan Gen. Equip. Co. v. Comm’r of
Internal Revenue, 297 U.S. 129, 135 (1936) (holding that a
tax regulation elaborating on a standard governed by statute
“is no more retroactive in its operation than is a judicial
determination construing and applying a statute to a case in
hand”). We therefore hold that Montgomery Ward
retroactivity analysis is only applicable when “an agency
consciously overrules or otherwise alters its own rule or
regulation,” or “expressly considers and openly departs from
a circuit court decision.” 1 Garfias-Rodriguez, 702 F.3d at
518–19.

Olivas-Motta’s primary argument against this
conclusion is the language of the Montgomery Ward factors.
It is true that the second Montgomery Ward factor is
“whether the new rule represents an abrupt departure from
well established practice or merely attempts to fill a void in
an unsettled area of law.” Montgomery Ward, 691 F.2d at
1333 (emphasis added) (quoting Retail, Wholesale and
Dep’t Store Union v. NLRB, 466 F.2d 380, 390 (D.C. Cir.
1972)). This language suggests that a change in law can

1
Judge Watford disagrees with our analysis and would conclude that
a change in law occurs when the Board’s decision was not “clearly
foreshadowed.” Diss. at 22. It is true that the Supreme Court has stated
that a new principle of law can be established by “deciding an issue
whose resolution was not clearly foreshadowed.” Chevron Oil Co. v.
Huson, 404 U.S. 97, 106 (1991). But “agency decisions are not
analogous to court decisions.” Garfias-Rodriguez, 702 F.3d at 520.
Chevron Oil, which dealt with court decisions, is not apposite when an
agency makes an adjudicatory decision that clarifies the scope of a
statute it is charged with executing.
 OLIVAS-MOTTA V. WHITAKER 11

occur when the rule was previously unclear, and an
adjudicatory decision brings clarity to the issue. But we must
consider the Montgomery Ward factors in light of the general
rules of retroactivity, which require a change of law. In
addition, the other Montgomery Ward factors themselves
contemplate a change from a “former rule” or “old
standard.” See Montgomery Ward, 691 F.2d at 1333 (quoting
Retail, 466 F.2d at 390). We therefore distinguish between
cases where a rule, such as 8 U.S.C. § 1227(a)(2)(A)(ii),
already exists, and an administrative decision simply
clarifies the rule’s application, and cases where the decision
itself would “take away or impair vested rights acquired
under existing laws, or create a new obligation, impose a
new duty, or attach a new disability, in respect to
transactions or considerations already past.” Vartelas, 566
U.S. at 266 (alterations omitted) (quoting Soc’y for
Propagation of Gospel v. Wheeler, 22 F. Cas. 756, 767 (No.
13,156) (CCNH 1814)). In the latter cases, the
administrative decision has altered the legal consequences
flowing from events and “considerations already past,” and
thus changed the law. See id. (quoting Soc’y for Propagation
of Gospel, 22 F. Cas. at 767). But in the former, where the
adjudicatory decision does not trigger a new obligation,
impair a previously vested right, or attach new harm, no new
legal consequences flow from the decision, and retroactivity
is not implicated. The second Montgomery Ward factor is
therefore better understood as evaluating the character of a
change in law, once such a change has occurred, rather than
evaluating whether the change occurred in the first instance.

Olivas-Motta points to language in Garfias-Rodriguez
suggesting that a change in law is not a prerequisite to
Montgomery Ward balancing. See Garfias-Rodriguez,
702 F.3d at 516 (“Chief Judge Kozinski . . . applies
retroactivity principles to conclude that retroactivity analysis
12 OLIVAS-MOTTA V. WHITAKER

does not apply, effectively resolving the retroactivity
question against Garfias”). We do not think Garfias-
Rodriguez stands for the proposition Olivas-Motta believes
it does. There was no dispute in that case that the law had
changed; rather, the issue was how we should treat the
unquestionable change of law of this circuit when it was
prompted by a decision of the Board. See id. at 515–20.
Garfias-Rodriguez did not hold that Montgomery Ward
balancing is required when no change in law has taken place.

B

Applying this standard to this case, there was no change
in law. Before Olivas-Motta’s 2007 guilty plea, the Board
had never determined in a precedential opinion whether
felony endangerment in Arizona was a CIMT. The Board
had only issued unpublished decisions on the issue. See, e.g,
In Re Carlos Mario Almeraz-Hernandez, 2006 WL
3203649, at *2 (B.I.A. Sept. 6, 2006) (holding § 13-1201 is
not categorically a CIMT). Unpublished decisions are not
precedential and “do not bind future parties.” Marmolejo-
Campos v. Holder, 558 F.3d 903, 909 (9th Cir. 2009).
Olivas-Motta therefore cannot argue that Leal I “attach[ed]
a new disability” to his guilty plea that did not exist at the
time he entered it. See Vartelas, 566 U.S. at 266 (quoting
Soc’y for Propagation, 22 F. Cas. at 767). Rather, 8 U.S.C.
§ 1227(a)(2)(A)(ii) had already created the legal
consequences of his plea, and it was merely unclear whether
it would apply. Leal I’s settling of that ambiguity did not
change the law any more than “a judicial determination
construing and applying a statute to a case in hand” would
have. See Manhattan Gen. Equip., 297 U.S. at 135.

Olivas-Motta counters that, notwithstanding the lack of
a precedential opinion on Arizona felony endangerment,
Leal I still constituted a change in law because of broader
 OLIVAS-MOTTA V. WHITAKER 13

changes in the law of CIMTs. According to Olivas-Motta,
the law before 2008 was that a crime with a mens rea of
recklessness could not constitute a CIMT unless the offense
presented an “aggravating factor,” thus preventing Arizona
endangerment from qualifying. But after the Attorney
General’s decision in In re Silva-Trevino, 24 I. & N. 687
(A.G. 2008), argues Olivas-Motta, the aggravating-factor
requirement was abolished, thus leading to the decision in
Leal I.

We are not persuaded that Silva-Trevino created the
change in law identified by Olivas-Motta. As we explained
in Leal II, the aggravating-factor requirement “[wa]s not due
to the reckless mens rea involved, but rather because of the
underlying conduct; both this court and the Board have
repeatedly stated that simple assault is, in general, not a
CIMT.” 771 F.3d at 1148. Thus, in Olivas-Motta’s cited
cases, the aggravating-factor analysis is harmonious with the
Attorney General’s later approach in Silva-Trevino.
Compare In re Fualaau, 21 I. & N. Dec. 475, 478 (B.I.A.
1996) (“In order for an assault of the nature at issue in this
case to be deemed a crime involving moral turpitude, the
element of a reckless state of mind must be coupled with an
offense involving the infliction of serious bodily injury”
(emphasis added)), with Silva-Trevino, 24 I. & N. Dec. at
689 n.1 (“a crime must involve both reprehensible conduct
and some degree of scienter, whether specific intent,
deliberateness, willfulness, or recklessness” (emphasis
added)). The earlier Board cases and Silva-Trevino did not
apply the aggravating-factor requirement to all recklessness
crimes, and Silva-Trevino did not purport to overrule
decisions holding that simple assault is not a CIMT. See Leal
II, 771 F.3d at 1148 (stating after Silva-Trevino: “It thus
follows that, in order for an assault to be considered a CIMT,
there must be some additional factor involved in the specific
14 OLIVAS-MOTTA V. WHITAKER

offense to distinguish it from generic simple assault”). As to
Arizona felony endangerment then, Silva-Trevino did not
change the law.

Olivas-Motta’s argument to the contrary relies on
unpublished Board decisions on this matter. Olivas-Motta is
correct that unpublished Board decisions predating Silva-
Trevino relied on Fualaau to conclude that Arizona
endangerment was not a CIMT. See, e.g., Almeraz-
Hernandez, 2006 WL 3203649, at *2. Olivas-Motta is also
correct that Leal I cited Silva-Trevino as the controlling
framework before concluding that Arizona felony
endangerment was categorically a CIMT. 26 I. & N. Dec. at
21, 27. But once more, unpublished decisions “do not bind
future parties.” Marmolejo-Campos, 558 F.3d at 909.
Olivas-Motta’s attorney may have made a calculation that
Arizona felony endangerment would not be considered a
CIMT based on unpublished decisions, but Fualaau did not
foreclose the conclusion that it was a CIMT before Silva-
Trevino, nor did Silva-Trevino require the Board to conclude
that it was a CIMT afterwards. The application of the statute
was simply unclear until Leal I, at which point the published
Board opinion resolved the issue. Put differently, when
Olivas-Motta pleaded guilty in 2007, it was possible that his
conviction would not be adjudicated a CIMT, but no law
guaranteed that. Leal I’s conclusive resolution of this
uncertainty did not create a new legal harm to Olivas-Motta
that did not already exist.

Because there was no change in the law raising
retroactivity concerns, the Board did not err by applying Leal
I to conclude that Arizona endangerment is a CIMT.
 OLIVAS-MOTTA V. WHITAKER 15

IV.

Preclusion prevents parties “from contesting matters that
they have had a full and fair opportunity to litigate,” thus
protecting “against ‘the expense and vexation attending
multiple lawsuits, conserv[ing] judicial resources, and
foster[ing] reliance on judicial action by minimizing the
possibility of inconsistent decisions.’” Taylor v. Sturgell,
553 U.S. 880, 892 (2008) (quoting Montana v. United States,
440 U.S. 147, 153–54 (1979)). “Under the doctrine of claim
preclusion, a final judgment forecloses ‘successive litigation
of the very same claim, whether or not relitigation of the
claim raises the same issues as the earlier suit.’” Id. (quoting
New Hampshire v. Maine, 532 U.S. 742, 748 (2001)). Under
the doctrine of issue preclusion, parties may not relitigate
“‘an issue of fact or law actually litigated and resolved in a
valid court determination essential to the prior judgment,’
even if the issue recurs in the context of a different claim.”
Id. (quoting New Hampshire, 532 U.S. at 748–49). Olivas-
Motta contends that, due to both types of preclusion, the
Board could not revisit on remand whether felony
endangerment was categorically a CIMT, after determining
initially that it was not.

A.

Before we can evaluate Olivas-Motta’s argument, we
must address whether we have jurisdiction to consider it. The
government argues that Olivas-Motta failed to exhaust his
administrative remedies because he did not argue preclusion
before the Board. Olivas-Motta responds that he could not
raise preclusion because it was not implicated until the
Board applied Leal I to his appeal.

We conclude that we have jurisdiction. 8 U.S.C.
§ 1252(d)(1) provides that a court may review a final order
16 OLIVAS-MOTTA V. WHITAKER

of removal only if “the alien has exhausted all administrative
remedies available to the alien as of right.” We have held
that “1252(d)(1) mandates exhaustion and therefore
generally bars us, for lack of subject-matter jurisdiction,
from reaching the merits of a legal claim not presented in
administrative proceedings below.” Barron v. Ashcroft,
358 F.3d 674, 678 (9th Cir. 2004). But section 1252(d)(1) by
its terms limits the petitioner’s duty to “remedies available
to the alien as of right.” We have thus held that we retain
jurisdiction over petitions where the challenged agency
action was committed by the Board after briefing was
completed, because the only remaining administrative
remedies for such an action were not available “as of right.”
Alcaraz v. INS, 384 F.3d 1150, 1159–60 (9th Cir. 2004).

In this case, after we granted Olivas-Motta’s first petition
and remanded to the Board, Olivas-Motta was never
provided an opportunity to argue preclusion until the Board
issued its second decision. At that point, his only remedies
were discretionary, and there was no higher administrative
authority to correct the supposed error. See id. A petition for
review to this court was therefore proper, and section
1252(d)(1) does not divest us of jurisdiction.

B.

On the merits of Olivas-Motta’s preclusion argument,
we hold there was no error. Claim preclusion requires a final
judgment on the merits in a separate action. Valencia-
Alvarez v. Gonzales, 469 F.3d 1319, 1323–24 (9th Cir.
2006). By granting Olivas-Motta’s petition for review in
2013, his original action continued, and no separate action
commenced. See id. at 1324. Similarly, issue preclusion only
applies when issues are “litigated and decided in the prior
proceedings.” Oyeniran, 672 F.3d at 806 (emphasis added).
Multiple proceedings are a prerequisite before issue
 OLIVAS-MOTTA V. WHITAKER 17

preclusion can apply. See id. Because the Board on remand
was acting within the same proceedings as in Olivas-Motta’s
original appeal, preclusion does not apply.

Olivas-Motta counters this argument by citing an
unpublished decision of this court relating to the rule of
mandate and making preclusion arguments by analogy. This
was also the argument that Olivas-Motta made to the Board
on remand. We consider this argument to be a rule of
mandate argument, rather than one of claim preclusion or
issue preclusion. Olivas-Motta has not argued that the Board
could not reconsider this issue because of law of the case.

The rule of mandate is related to, but distinct from, claim
preclusion and issue preclusion. Under the rule of mandate,
an administrative agency may not deviate from a supervising
court’s remand order, and the reviewing court may review
the agency’s decision on remand “to assure that its prior
mandate is effectuated.” Sullivan v. Hudson, 490 U.S. 877,
886 (1989); see also Mendez-Gutierrez v. Gonzales,
444 F.3d 1168, 1172–73 (9th Cir. 2006) (holding that the
rule of mandate applies to decisions of the Board on remand
from this court). Thus, as with claim preclusion and issue
preclusion, the rule of mandate can prevent parties from
relitigating issues already decided. But the scope of the rule
is limited to that which is before the court “and disposed of
by its decree.” United States v. Thrasher, 483 F.3d 977, 981
(9th Cir. 2007) (quoting In re Sanford Fork & Tool Co.,
160 U.S. 247, 255 (1895)). An administrative agency may
therefore consider on remand “any issue not expressly or
impliedly disposed of on appeal.” Stacy v. Colvin, 825 F.3d
563, 568 (9th Cir. 2016) (quoting Odima v. Westin Tucscon
Hotel, 53 F.3d 1484, 1497 (9th Cir. 1995)).

Our mandate in Olivas-Motta’s first petition did not
conclude that felony endangerment was not a CIMT, or that
18 OLIVAS-MOTTA V. WHITAKER

Leal I was wrongly decided. 746 F.3d at 916–17. Instead, we
“h[e]ld only that that we [could] not deny Olivas-Motta’s
petition based on a conclusion reached by the [Board] in a
separate case decided two years after it decided the appeal
now before us.” Id. at 917. Nothing in our remand restricted
the Board from considering the import of Leal I on Olivas-
Motta’s appeal. Accordingly, the rule of mandate did not
foreclose the Board’s reconsideration of the issue.

V.

The void-for-vagueness doctrine stems from the Fifth
Amendment’s guarantee of due process. Johnson v. United
States, 135 S. Ct. 2551, 2556 (2015). “[T]he Government
violates this guarantee by taking away someone’s life,
liberty, or property under a criminal law so vague that it fails
to give ordinary people fair notice of the conduct it punishes,
or so standardless that it invites arbitrary enforcement.” Id.
Because “deportation is ‘a particularly severe penalty,’
which may be of greater concern to a convicted alien than
‘any potential jail sentence,’” a provision of immigration law
making an alien deportable is subject to the void-for-
vagueness doctrine. Sessions v. Dimaya, 138 S. Ct. 1204,
1213 (2018) (quoting Jae Lee v. United States, 137 S. Ct.
1958, 1968 (2017)).

Olivas-Motta argues that, even if applying Leal I to his
appeal was not impermissibly retroactive or precluded, we
should nonetheless grant the petition because 8 U.S.C.
§ 1227(a)(2)(A)(ii) itself is unconstitutionally vague. While
he recognizes that both the Supreme Court and this court
have repeatedly rejected that argument, see Jordan v. De
George, 341 U.S. 223, 232 (1951); Martinez-De Ryan v.
Sessions, 895 F.3d 1191, 1194 (9th Cir. 2018), Olivas-Motta
contends that the Board’s interpretation of the statute has
 OLIVAS-MOTTA V. WHITAKER 19

expanded the meaning of “moral turpitude” to the point that
there is no meaningful standard guiding aliens’ conduct.

We are not persuaded that this argument is
distinguishable from those rejected in past cases. As we
explained in Leal II, a crime is morally turpitudinous if it
involves a conscious decision and a resulting harm, where
“more serious resulting harm is required” “as the level of
conscious behavior decreases, i.e., from intentional to
reckless conduct.” 771 F.3d at 1146 (quoting Ceron v.
Holder, 747 F.3d 773, 783 (9th Cir. 2014) (en banc)). That
is the standard the Board applied to evaluate Arizona felony
endangerment, id. at 1147, and that standard is sufficiently
meaningful to provide fair notice under our precedent.
Martinez-De Ryan, 895 F.3d at 1193–94. To the extent
Olivas-Motta asks us to reconsider those decisions, that is
beyond this panel’s authority. Miller, 335 F.3d at 900.

VI.

The Board did not commit any of the raised legal errors
by concluding that Olivas-Motta’s conviction for reckless
endangerment was a crime involving moral turpitude. We
therefore deny the petition.

PETITION DENIED.
20 OLIVAS-MOTTA V. WHITAKER

WATFORD, Circuit Judge, dissenting:

When a non-citizen is charged with a crime and deciding
whether to plead guilty, the immigration consequences of a
conviction are often a major consideration. For some
defendants, preserving the chance to remain in the United
States is more important than the length of any prison
sentence that might be imposed. Padilla v. Kentucky,
559 U.S. 356, 368 (2010). With that in mind, competent
defense counsel “may be able to plea bargain creatively with
the prosecutor in order to craft a conviction and sentence that
reduce the likelihood of deportation, as by avoiding a
conviction for an offense that automatically triggers the
removal consequence.” Id. at 373. Such plea bargains are
mutually beneficial for the prosecution: A defendant who
might otherwise have proceeded to trial may be persuaded to
forgo that right in exchange for a deal that allows him to
plead guilty to an offense that reduces the risk of removal.
Id.

An assessment of the immigration consequences
attending a guilty plea must, of course, be based on the law
as it exists at the time of the plea. If the law on that subject
changes after a defendant pleads guilty, he usually cannot go
back and undo his conviction, even if the conviction now
carries far more serious immigration consequences than
before. For that reason, when there is an intervening change
in the law, we are required to assess whether the new rule
may be applied retroactively in subsequent removal
proceedings.

The majority refuses to engage in that analysis because
it concludes that no “new rule” was adopted after Manuel
Olivas-Motta pleaded guilty. I respectfully disagree.
 OLIVAS-MOTTA V. WHITAKER 21

The sole issue in Olivas-Motta’s removal proceedings is
whether reckless endangerment under Arizona Revised
Statutes § 13-1201 constitutes a crime involving moral
turpitude. When Olivas-Motta pleaded guilty to that offense
in 2007, the Board of Immigration Appeals (BIA) had not
decided in a precedential opinion whether reckless
endangerment should be classified as a crime involving
moral turpitude. But in 2012, long after Olivas-Motta
pleaded guilty, the BIA held for the first time that reckless
endangerment under § 13-1201 is a crime involving moral
turpitude. Matter of Leal, 26 I. & N. Dec. 20, 27 (BIA
2012), aff’d sub nom. Leal v. Holder, 771 F.3d 1140 (9th Cir.
2014).

The holding in Matter of Leal represents a “new rule”
under any definition of that term. The Supreme Court has
said that a decision can establish a new rule “either by
overruling clear past precedent on which litigants may have
relied, or by deciding an issue of first impression whose
resolution was not clearly foreshadowed.” Chevron Oil Co.
v. Huson, 404 U.S. 97, 106 (1971) (citations omitted). The
BIA did not overrule past precedent in Matter of Leal, but it
did resolve an issue of first impression—whether reckless
endangerment qualifies as a crime involving moral turpitude.
The BIA’s resolution of that issue was not clearly
foreshadowed by precedent existing at the time Olivas-
Motta pleaded guilty. In fact, as discussed below, the BIA’s
precedent in 2007 suggested that reckless endangerment
under § 13-1201 would not be classified as a crime involving
moral turpitude. Thus, Matter of Leal plainly constitutes the
“change in law” that the majority identifies as necessary to
trigger retroactivity analysis. Maj. op. at 9.

The majority suggests that our case is analogous to one
in which a statutory provision is on the books when a
22 OLIVAS-MOTTA V. WHITAKER

defendant pleads guilty, and a court later does nothing more
than construe and apply that statute in the case at hand. Maj.
op. at 12. In that scenario, the majority asserts, we would
not regard the judicial interpretation as a “new rule” subject
to retroactivity analysis.

The majority’s assertion would be correct if the court’s
decision were “dictate[d] by the plain language of the
statute.” Harper v. Virginia Department of Taxation,
509 U.S. 86, 111 (1993) (Kennedy, J., concurring in part and
concurring in the judgment) (internal quotation marks
omitted). But that is certainly not the case here. The
governing statutory standard is supplied by 8 U.S.C.
§ 1227(a)(2)(A)(ii), which renders a non-citizen removable
if he’s been convicted of two or more “crimes involving
moral turpitude.” The quoted phrase has no intelligible
meaning; it creates what Justice Jackson rightly labeled “an
undefined and undefinable standard.” Jordan v. De George,
341 U.S. 223, 235 (1951) (Jackson, J., dissenting). Neither
our court nor the BIA has been able to come up with “any
coherent criteria for determining which crimes fall within
that classification and which crimes do not.” Nunez v.
Holder, 594 F.3d 1124, 1130 (9th Cir. 2010). The BIA has
been able to give the statutory standard concrete meaning
mainly by declaring, through case-by-case adjudications,
which specific offenses are covered and which are not. See
Marmolejo-Campos v. Holder, 558 F.3d 903, 910–11 (9th
Cir. 2009) (en banc).

Against that backdrop, each BIA decision that designates
a new offense (or class of offenses) as a crime involving
moral turpitude potentially creates a “new rule” for
retroactivity purposes—at least where, as here, the decision
was not clearly foreshadowed by prior precedent. That does
not mean retroactive application of all such decisions is
 OLIVAS-MOTTA V. WHITAKER 23

prohibited; it just means that the decisions must be analyzed
under the framework we’ve established for assessing
whether retroactive application is permissible.

This case is a prime example of one in which retroactive
application of a new rule is impermissible. Olivas-Motta
was originally charged in 2007 with attempted murder and
aggravated assault with a deadly weapon, offenses that
would clearly render him removable if he were convicted.
He had already been convicted of one crime involving moral
turpitude; he would be subject to removal if convicted of a
second, and the BIA had already classified attempted murder
and aggravated assault with a deadly weapon as crimes
involving moral turpitude. See Matter of Sanchez-Linn, 20 I.
& N. Dec. 362, 366 (BIA 1991); Matter of Medina, 15 I. &
N. Dec. 611, 614 (BIA 1976).

Minimizing the likelihood of removal was of paramount
concern to Olivas-Motta. He was born in Mexico, but his
parents brought him to the United States in 1976 when he
was only ten days old. He has lived his entire life in this
country as a lawful permanent resident. He is married to a
U.S. citizen, and both of his children are U.S. citizens. Most
of his family members are also either U.S. citizens or lawful
permanent residents. For him, being removed to Mexico
would truly be “the equivalent of banishment or exile.”
Padilla, 559 U.S. at 373 (quoting Delgadillo v. Carmichael,
332 U.S. 388, 390–91 (1947)).

Although Olivas-Motta believed himself innocent of the
charges he faced, he was no doubt “acutely aware” of the
severe immigration consequences a conviction would
trigger. INS v. St. Cyr, 533 U.S. 289, 322 (2001). He
therefore had his attorney explore the possibility of pleading
guilty to a lesser offense. Olivas-Motta’s defense counsel
consulted with an experienced immigration lawyer, who
24 OLIVAS-MOTTA V. WHITAKER

surveyed the law as it then stood. She advised that having
Olivas-Motta plead guilty to reckless endangerment under
§ 13-1201 would minimize the risk of deportation because
that offense in all likelihood would not be regarded as a
crime involving moral turpitude. Olivas-Motta relied on that
advice in deciding to plead guilty and forgo his right to a
trial.

The advice Olivas-Motta received was sound at the time.
The BIA had long held that crimes involving moral turpitude
require some form of corrupt or evil intent. See, e.g., Matter
of P—, 3 I. & N. Dec. 56, 59 (BIA 1947). The BIA retreated
from that position in 1976, when it held that certain offenses
committed with a mens rea of recklessness could qualify as
well. Medina, 15 I. & N. Dec. at 614. But the Board also
made clear that a crime involving reckless conduct is not per
se a crime involving moral turpitude. In re Fualaau, 21 I. &
N. Dec. 475, 478 (BIA 1996). Something more was
required, although exactly what that something more
consisted of remained open to debate. In predicting the
likely classification of reckless endangerment, the best
guidance came from a series of cases involving
manslaughter and assault offenses, which held that a crime
committed with a mens rea of recklessness had to include as
an element some sort of aggravating circumstance, such as
the infliction of death or serious bodily injury. See id.
(serious bodily injury); Matter of Wojtkow, 18 I. & N. Dec.
111, 113 (BIA 1981) (death); Medina, 15 I. & N. Dec. at 614
(use of a deadly weapon).

As of 2007, the BIA had not issued a precedential
decision involving a reckless endangerment offense.
Nonetheless, reckless endangerment under Arizona law did
not appear to qualify as a crime involving moral turpitude,
for although it requires a mens rea of recklessness, it does
 OLIVAS-MOTTA V. WHITAKER 25

not require proof of any of the aggravating circumstances
found in past cases. The felony version of the offense, to
which Olivas-Motta pleaded guilty, simply requires
“recklessly endangering another person with a substantial
risk of imminent death.” Ariz. Rev. Stat. § 13-1201(A). To
be sure, there was ongoing debate about whether placing
someone in grave risk of death or serious bodily injury could
itself be deemed an aggravating circumstance, see Knapik v.
Ashcroft, 384 F.3d 84, 90 (3d Cir. 2004); In re Braimllari,
2006 WL 729794, at *1 (BIA Feb. 14, 2006), but the BIA
had rejected that view in two non-precedential decisions,
both of which expressly held that reckless endangerment
under § 13-1201 did not qualify as a crime involving moral
turpitude. In re Almeraz-Hernandez, 2006 WL 3203649, at
*2–3 (BIA Sept. 6, 2006); In re Valles-Moreno, 2006 WL
3922279, at *2–3 (BIA Dec. 27, 2006).

In 2008, however, the Attorney General replaced the
BIA’s former standard for determining which recklessness
offenses qualify as crimes involving moral turpitude with a
new standard. In Matter of Silva-Trevino, 24 I. & N. Dec.
687 (A.G. 2008), the Attorney General declared that, to
qualify as a crime involving moral turpitude, an offense need
involve only “reprehensible conduct and some degree of
scienter.” Id. at 689 n.1. The effect of this change was to
eliminate the aggravating-circumstance requirement for
offenses with a mens rea of recklessness. 1 Under the new

1
The government argues, and the majority appears to agree, that the
aggravating-circumstance requirement was limited to assault offenses,
and thus did not apply to offenses like reckless endangerment. Maj. op.
at 13–14. But none of the BIA’s cases in this area held that the
aggravating-circumstance requirement was limited to assault offenses
alone, and there is no logical reason why it would not extend to a
comparably serious offense such as reckless endangerment. Indeed, in
each of the pre-Silva-Trevino cases in which the BIA held that a reckless
26 OLIVAS-MOTTA V. WHITAKER

standard, it was now far more likely that reckless
endangerment under § 13-1201 would be classified as a
crime involving moral turpitude. After all, placing someone
in “substantial risk of imminent death” would certainly seem
to qualify as reprehensible conduct. And indeed, in 2012,
that is exactly what the BIA concluded in Matter of Leal,
where the agency held for the first time that reckless
endangerment under § 13-1201 constitutes a crime involving
moral turpitude. 26 I. & N. Dec. at 27.

The question thus becomes whether Matter of Leal may
be applied retroactively to Olivas-Motta’s case—in other
words, whether the immigration consequences of his
conviction should be assessed under the law as it stood in
2007, when he pleaded guilty, or under the law as it stood in
2014, when the BIA adjudicated his appeal. To answer that
question, we apply the test from Montgomery Ward & Co. v.
FTC, 691 F.2d 1322 (9th Cir. 1982), which requires us to
balance five factors: “(1) whether the particular case is one
of first impression, (2) whether the new rule represents an
abrupt departure from well established practice or merely
attempts to fill a void in an unsettled area of law, (3) the
extent to which the party against whom the new rule is
applied relied on the former rule, (4) the degree of the burden
which a retroactive order imposes on a party, and (5) the
statutory interest in applying a new rule despite the reliance
of a party on the old standard.” Id. at 1333 (internal
quotation marks omitted).

endangerment offense qualified as a crime involving moral turpitude, no
one disputed that the aggravating-circumstance requirement applied.
The BIA simply concluded in those cases, involving statutes from other
States, that the requirement was satisfied. See Knapik, 384 F.3d at 90;
Braimllari, 2006 WL 729794, at *1.
 OLIVAS-MOTTA V. WHITAKER 27

The weight to be accorded the first, fourth, and fifth
factors has already been settled. We have held that the first
factor does not favor either party in the immigration context.
Garfias-Rodriguez v. Holder, 702 F.3d 504, 520–21 (9th Cir.
2012) (en banc). The fourth factor strongly favors Olivas-
Motta, as the burden imposed by retroactively applying the
law in effect in 2014 is severe: Under the rule adopted in
Matter of Leal, his conviction for reckless endangerment
would be regarded as a crime involving moral turpitude,
subjecting him to removal from the United States and
separation from his family. See id. at 523. The fifth factor
points in the government’s favor, since “non-retroactivity
impairs the uniformity of a statutory scheme, and the
importance of uniformity in immigration law is well
established.” Id.

The second and third factors, then, are dispositive, and
they tip the balance in Olivas-Motta’s favor. When he
pleaded guilty to reckless endangerment in 2007, the BIA
had an established standard for determining which
recklessness offenses constitute crimes involving moral
turpitude. The Attorney General’s subsequent decision in
Silva-Trevino represented an “abrupt departure” from that
standard, Montgomery Ward, 691 F.2d at 1333 (internal
quotation marks omitted), in the sense that it replaced the
aggravating-circumstance requirement with a new, more
expansive standard. That change in the governing standard
was outcome determinative with respect to certain offenses,
as we know from the way the BIA classified § 13-1201
before and after Silva-Trevino. Before the Attorney
General’s decision, the BIA had held (in non-precedential
decisions) that reckless endangerment under § 13-1201 does
not constitute a crime involving moral turpitude; afterward
the BIA definitively held exactly the opposite. Because
Olivas-Motta had no reason to anticipate elimination of the
28 OLIVAS-MOTTA V. WHITAKER

aggravating-circumstance requirement, his reliance on the
pre-Silva-Trevino standard when deciding to plead guilty
was eminently reasonable. Cf. Garfias-Rodriguez, 702 F.3d
at 521 (second and third factors weigh in favor of retroactive
application when the petitioner “could reasonably have
anticipated the change in the law such that the new
‘requirement would not be a complete surprise’”) (quoting
Montgomery Ward, 691 F.2d at 1333–34).

It’s true, as the government argues, that the status of
§ 13-1201 had not been settled definitively in Olivas-
Motta’s favor prior to 2008. So this is not a case in which it
is 100% clear that Olivas-Motta would have prevailed under
the pre-Silva-Trevino standard. But, contrary to the
majority’s apparent assumption, see Maj. op. at 14, that is
far from fatal under the second and third Montgomery Ward
factors.

In Miguel-Miguel v. Gonzales, 500 F.3d 941 (9th Cir.
2007), we ruled for the petitioner in circumstances quite
similar to those present here. When the petitioner in that
case pleaded guilty to selling a small amount of cocaine, a
relatively minor drug-trafficking offense like his would be
classified as a “particularly serious crime” on a case-by-case
basis using a multi-factor test. Id. at 945–46, 950. After he
pleaded guilty, the Attorney General created a new standard
that presumed all drug-trafficking offenses to be particularly
serious crimes, with the presumption rebuttable only in very
narrow circumstances. Id. at 946–47. We held that the
second and third factors favored the petitioner because at the
time he pleaded guilty, there was a “realistic chance” that the
BIA would find that his crime was not particularly serious,
whereas under the Attorney General’s new standard there
was “a near (if not total) certainty” that his crime would be
 OLIVAS-MOTTA V. WHITAKER 29

classified as particularly serious, thereby resulting in his
removal. Id. at 952; see also St. Cyr, 533 U.S. at 321.

Olivas-Motta’s situation is no different. At the time he
pleaded guilty, there was at least a realistic chance that his
reckless endangerment offense would not be classified as a
crime involving moral turpitude; the BIA had already so held
in two non-precedential decisions. After the Attorney
General’s decision in Silva-Trevino, however, it was nearly
certain that his offense would be classified as a crime
involving moral turpitude, and the BIA’s decision in Matter
of Leal soon eliminated what little uncertainty remained on
that score. To the same extent as in Miguel-Miguel, the
change in the governing standard “attaches new legal
consequences to events completed before its enactment,”
Vartelas v. Holder, 566 U.S. 257, 273 (2012) (internal
quotation marks omitted), and therefore may not be applied
retroactively.

Under the Montgomery Ward test, the balance of factors
weighs in favor of Olivas-Motta. Three of the factors favor
him—one strongly so—while only one of the factors points
in the government’s favor. That means the status of his
conviction for reckless endangerment should be analyzed
under the law as it stood in 2007, applying the standard that
prevailed before the Attorney General’s decision in Silva-
Trevino. See id. at 261. I would grant Olivas-Motta’s
petition for review and remand so that the agency can
conduct that analysis in the first instance.