Summary
lamenting that "the Ninth Circuit's abysmal and indefensible immigration precedents are the gifts that keep on taking"
Summary of this case from Molina v. GarlandOpinion
No. 17-70127
08-27-2021
Jorge Andres Reyes AFANADOR, aka Jorge Alberto Reyes, Petitioner, v. Merrick B. GARLAND, Attorney General, Respondent.
Saad Ahmad (argued), Fremont, California; Raul Ray, Law Offices of Raul Ray, San Jose, California; for Petitioner. Jonathan K. Ross (argued), Trial Attorney; Erica B. Miles, Senior Litigation Counsel; Office of Immigration Litigation, Civil Division, United States Department of Justice, Washington, D.C.; for Respondent.
Saad Ahmad (argued), Fremont, California; Raul Ray, Law Offices of Raul Ray, San Jose, California; for Petitioner.
Jonathan K. Ross (argued), Trial Attorney; Erica B. Miles, Senior Litigation Counsel; Office of Immigration Litigation, Civil Division, United States Department of Justice, Washington, D.C.; for Respondent.
Before: William A. Fletcher, Sandra S. Ikuta, and Lawrence VanDyke, Circuit Judges.
Dissent by Judge VanDyke
IKUTA, Circuit Judge:
Jorge Andres Reyes Afanador, a native of Colombia, petitions for review of a ruling by the Board of Immigration Appeals (BIA) that he was removable as an alien convicted of two crimes involving moral turpitude under 8 U.S.C. § 1227(a)(2)(A)(ii). We hold that the BIA erred in applying its decision in Matter of Cortes Medina , 26 I. & N. Dec. 79, 81 (BIA 2013), retroactively to classify Reyes's 2011 conviction for indecent exposure as a crime involving moral turpitude. Therefore, we grant the petition for review.
I
Reyes is a native of Colombia who entered the United States on a visitor's visa in 1989. Before adjusting his status to lawful permanent resident, Reyes had numerous criminal arrests and convictions, including two indecent exposure convictions, one in 2007 under California Penal Code section 314.1 and the other in 2008 under California Penal Code section 647(a). In 2009, Reyes married a United States citizen and successfully adjusted his status through an application that his wife filed on his behalf.
California Penal Code section 314.1 provides, in pertinent part, that "[e]very person who willfully and lewdly ... [e]xposes his person, or the private parts thereof, in any public place, or in any place where there are present other persons to be offended or annoyed thereby" is guilty of a crime.
In 2011, Reyes was again charged with indecent exposure under section 314.1 and pleaded no contest to a felony. He was sentenced to eight months in jail and three years probation for this offense. In 2014, he pleaded no contest to an additional felony for violation of section 314.1, for which he was sentenced to 16 months in prison and three years of parole upon release. In 2015, the government initiated removal proceedings. Relying on the 2011 and 2014 convictions, the government issued a notice to appear (NTA) charging Reyes with being subject to removal from the United States under 8 U.S.C. § 1227(a)(2)(A)(ii) as an alien convicted of two crimes involving moral turpitude not arising out of a single scheme of criminal misconduct.
8 U.S.C. 1227(a)(2)(A)(ii) provides:
Any alien who at any time after admission is convicted of two or more crimes involving moral turpitude, not arising out of a single scheme of criminal misconduct, regardless of whether confined therefor and regardless of whether the convictions were in a single trial, is deportable.
At his July 2016 hearing before the immigration judge (IJ), Reyes admitted the factual allegations in the NTA, but challenged his removability on the ground that his convictions under section 314.1 were not categorically crimes involving moral turpitude. He also submitted applications for relief from removal. Relying on a precedential BIA opinion holding that a violation of section 314.1 was a crime involving moral turpitude, see Cortes Medina , 26 I. & N. Dec. at 81, the IJ concluded that Reyes was removable as charged. The IJ also denied Reyes's cancellation of removal application as a matter of discretion and denied his applications for asylum, withholding of removal, and relief under the Convention Against Torture (CAT).
Reyes submitted applications for cancellation of removal, asylum, withholding of removal, and protection under the Convention Against Torture (CAT). He based these applications on his fear of returning to Colombia because a drug cartel harassed his father in the 1980s.
On appeal to the BIA, Reyes argued that the IJ erred by ignoring a Ninth Circuit decision holding that an offense under section 314.1 was not categorically a crime involving moral turpitude, see Nunez v. Holder , 594 F.3d 1124, 1133 (9th Cir. 2010). The BIA rejected this argument on the ground that its precedential opinion in Cortes Medina superseded our decision in Nunez . See Nat'l Cable & Telecomms. Ass'n v. Brand X Internet Servs. , 545 U.S. 967, 982, 125 S.Ct. 2688, 162 L.Ed.2d 820 (2005) (holding that federal courts must defer to certain agency decisions even when they overrule prior judicial opinions). The BIA rejected Reyes's other claims. This petition for review followed.
II
Reyes argues that Cortes Medina cannot be applied retroactively to his 2011 conviction, and therefore the BIA erred in treating that conviction as a crime involving moral turpitude. If Reyes is correct, then only his 2014 conviction is a crime involving moral turpitude; this means that Reyes was not removable under section 1227(a)(2)(A)(ii), which applies to an alien convicted of two or more crimes involving moral turpitude.
We begin with the background principles for applying new legal requirements retroactively. It has long been established that legislation does not apply retroactively absent a clear indication that Congress intended to make the statute retroactive. Reynolds v. McArthur , 27 U.S. 417, 434, 2 Pet. 417, 7 L.Ed. 470 (1829). This general rule is based on "deeply rooted" principles of equity and due process. Landgraf v. USI Film Prods. , 511 U.S. 244, 265, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994). "Elementary considerations of fairness dictate that individuals should have an opportunity to know what the law is and to conform their conduct accordingly; settled expectations should not be lightly disrupted." Id.
It is not always clear whether new legislation has a retroactive effect, however, and the Supreme Court has acknowledged that "[a]ny test of retroactivity will leave room for disagreement in hard cases." Id. at 270, 114 S.Ct. 1483. As a general rule, legislation is deemed retroactive (and therefore impermissible unless expressly sanctioned by Congress) if "it changes the legal consequences of acts completed before its effective date," or if it "takes away or impairs vested rights acquired under existing laws, or creates a new obligation, imposes a new duty, or attaches a new disability." Id. at 269, 114 S.Ct. 1483 n.23 (cleaned up). In determining whether a statute attaches new legal consequences to events completed before its enactment, courts "should be informed and guided by familiar considerations of fair notice, reasonable reliance, and settled expectations." INS v. St. Cyr , 533 U.S. 289, 321, 121 S.Ct. 2271, 150 L.Ed.2d 347 (2001) (cleaned up).
For instance, the Supreme Court has explained that a rule does not have an impermissible retroactive effect "merely because it is applied in a case arising from conduct antedating the statute's enactment," or "merely because it draws upon antecedent facts for its operation." Landgraf , 511 U.S. at 269 & n.24, 114 S.Ct. 1483. And "[e]ven uncontroversially prospective statutes may unsettle expectations and impose burdens on past conduct: a new property tax or zoning regulation may upset the reasonable expectations that prompted those affected to acquire property; a new law banning gambling harms the person who had begun to construct a casino before the law's enactment or spent his life learning to count cards." Id. at 269 n.24, 114 S.Ct. 1483.
The Supreme Court has applied these principles in several immigration cases. In St. Cyr , the Supreme Court considered the retroactivity of new legislation which stripped the government of discretion to waive deportation for aliens who had committed certain crimes. Id. at 321–22, 121 S.Ct. 2271. The Court held this legislation was retroactive to the extent it applied to aliens who had entered guilty pleas before the legislation was enacted. Id. at 326, 121 S.Ct. 2271. The Court reached a similar conclusion in Vartelas v. Holder when it considered new legislation that rendered lawful permanent residents convicted of certain crimes inadmissible if they left the United States for a brief period. 566 U.S. 257, 262–63, 132 S.Ct. 1479, 182 L.Ed.2d 473 (2012). The Court held this legislation had a congressionally unauthorized retroactive effect to the extent it applied to lawful permanent residents who had previously pleaded guilty to such crimes. Id. at 263, 275, 132 S.Ct. 1479. In both cases, the new statutory provision changed the legal consequences of plea agreements entered into before the statute's effective date.
By contrast to legislation, judicial decisions have been governed by a "fundamental rule of retrospective operation" for "near a thousand years." Harper v. Va. Dep't of Tax'n , 509 U.S. 86, 94, 113 S.Ct. 2510, 125 L.Ed.2d 74 (1993) (cleaned up). Courts must apply judicial decisions announcing new interpretations of criminal procedural rules "retroactively to all cases, state or federal, pending on direct review or not yet final." Griffith v. Kentucky , 479 U.S. 314, 328, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987). Federal courts must also retroactively apply new rules announced in civil cases, except in narrow circumstances. Harper , 509 U.S. at 97–98, 113 S.Ct. 2510.
Agency determinations may be legislative or judicial, because agencies engage in both rulemaking and adjudication. Agency determinations are judicial in nature when an agency's adjudicatory decisions apply preexisting rules to new factual circumstances. De Niz Robles v. Lynch , 803 F.3d 1165, 1172 (10th Cir. 2015) (Gorsuch, J.). In these circumstances, an agency's determinations apply retroactively, like other judicial decisions. See id. ; see also SEC v. Chenery Corp ., 332 U.S. 194, 203, 67 S.Ct. 1760, 91 L.Ed. 1995 (1947) ( Chenery II ) ("Every case of first impression has a retroactive effect, whether the new principle is announced by a court or by an administrative agency.").
An agency's determinations may also be legislative in nature if Congress has delegated legislative authority to the agency. An agency may exercise its legislative authority in two different ways: it may proceed either through formal notice-and-comment rulemaking, or it may proceed through agency adjudication. Chenery II , 332 U.S. at 202, 67 S.Ct. 1760 ; see also Montgomery Ward & Co. v. FTC , 691 F.2d 1322, 1328 (9th Cir. 1982) ("It is well settled that the decision whether to proceed by adjudication or rule-making lies in the first instance within the [agency's] discretion." (cleaned up)).
When an agency engages in formal rulemaking, the rules it promulgates are analogous to legislation and are construed to apply only prospectively (unless Congress has expressly authorized it to promulgate a retroactively applicable rule). See Bowen v. Georgetown Univ. Hosp. , 488 U.S. 204, 208, 109 S.Ct. 468, 102 L.Ed.2d 493 (1988) ; De Niz Robles , 803 F.3d at 1172 ("[A]bsent express congressional approval, newly promulgated agency rules should apply only prospectively because of their affinity to legislation.") (citing Bowen , 488 U.S. at 208, 109 S.Ct. 468 ).
An agency may also exercise its congressionally delegated legislative authority through adjudicatory proceedings, where "new administrative policy [is] announced and implemented through adjudication." Montgomery Ward , 691 F.2d at 1328. When an agency issues new rules of general applicability through adjudication, such rules are analogous to legislation, because the agency's interpretation of a statute "is not a once-and-for-always definition of what the statute means, but an act of interpretation in light of its policymaking responsibilities that may be reconsidered on a continuing basis." Garfias-Rodriguez v. Holder , 702 F.3d 504, 515–16 (9th Cir. 2012) (en banc) (cleaned up). But while rules promulgated through legislation or formal rulemaking generally apply prospectively (unless Congress has sanctioned retroactive application), adjudicatory rules may have a permissible retroactive effect, even without authorization from Congress, in some circumstances. See Chenery II , 332 U.S. at 203, 67 S.Ct. 1760. We determine whether a rule's retroactive effect is permissible by engaging in a case-by-case analysis to balance "a regulated party's interest in being able to rely on the terms of a rule as it is written, against an agency's interest in retroactive application of an adjudicatory decision." Montgomery Ward , 691 F.2d at 1333. We have adopted the following five-factor balancing test, which we apply on a case-by-case basis:
Chenery II explains that rules made through agency adjudication must be allowed to have a retroactive effect in certain situations because administrative agencies face problems they "could not reasonably foresee" or "problem[s] [that are] so specialized and varying in nature as to be impossible of capture within the boundaries of a general rule." 332 U.S. at 202–03, 67 S.Ct. 1760. "In those situations, the agency must retain power to deal with the problems on a case-to-case basis if the administrative process is to be effective." Id. at 203, 67 S.Ct. 1760.
"[A] change in law must have occurred before Montgomery Ward is implicated." Olivas-Motta v. Whitaker , 910 F.3d 1271, 1276 (9th Cir. 2018).
Among the considerations that enter into a resolution of the problem are (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. (quoting Retail, Wholesale & Dep't Store Union v. NLRB , 466 F.2d 380, 390 (D.C. Cir. 1972) ).
If the agency's adjudicatory rule does not have a retroactive effect, the court need not apply the Montgomery Ward factors. See Singh v. Napolitano , 649 F.3d 899, 901 n.1 (9th Cir. 2011) (per curiam).
The first factor (whether a case is one of first impression) does not apply in the immigration context. See Garfias-Rodriguez , 702 F.3d at 520–21. In that context, the government is the plaintiff and there is no concern "that a party seeking to overturn an administrative rule" would not get the benefit of its efforts. De Niz Robles , 803 F.3d at 1177 n.9.
The second factor considers whether the adjudicatory rule "represents an abrupt departure from well established practice," Garfias-Rodriguez , 702 F.3d at 518 (quoting Montgomery Ward , 691 F.2d at 1333 ), or is "an abrupt break with well-settled policy," id. (quoting ARA Servs., Inc. v. NLRB , 71 F.3d 129, 135 (4th Cir. 1995) ). Because the BIA functions as a court of appellate review within a larger agency framework, see 8 C.F.R. § 1003.1, it establishes policy only through its precedential decisions, which are binding on third parties. Marmolejo-Campos v. Holder , 558 F.3d 903, 909 n.9 (9th Cir. 2009) (en banc). The BIA's unpublished decisions, which do not bind the BIA or future parties, see id. , do not establish the practice or policy of the agency, see 8 C.F.R. § 1003.1(d)(1) (stating that the BIA, "through precedent decisions , shall provide clear and uniform guidance to the [agency], the immigration judges, and the general public on the proper interpretation and administration of the Act and its implementing regulations" (emphasis added)); cf. Indep. Ins. Agents of Am., Inc. v. Ludwig , 997 F.2d 958, 962 (D.C. Cir. 1993) (holding that "in the real world of agency practice, informal unpublished letters should not engender reliance" (cleaned up)).
Next, the third factor (a party's reliance interests) directs "our attention to the question whether the petitioner can claim reasonable reliance on some past rule or decision, a due process concern always at the heart of retroactivity analysis." De Niz Robles , 803 F.3d at 1177 (assessing the Tenth Circuit's version of the Montgomery Ward factors, known as the Stewart Capital factors). "The second and the third factors are closely intertwined," Garfias-Rodriguez , 702 F.3d at 521, and therefore are analyzed together. "If a new rule ‘represents an abrupt departure from well established practice,’ a party's reliance on the prior rule is likely to be reasonable, whereas if the rule ‘merely attempts to fill a void in an unsettled area of law,’ reliance is less likely to be reasonable." Id. (quoting Retail, Wholesale & Dep't Store Union , 466 F.2d at 390–91 ).
The fourth and fifth factors balance the costs to the regulated party against the benefits of allowing the agency to give a retroactive effect to a new rule. De Niz Robles , 803 F.3d at 1177.
Whether an agency announces a new interpretation of an ambiguous statute through formal rulemaking or through adjudication, a court must defer to the agency's decision so long as it is reasonable. Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc. , 467 U.S. 837, 842–43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984) ; see also INS v. Aguirre-Aguirre , 526 U.S. 415, 424, 119 S.Ct. 1439, 143 L.Ed.2d 590 (1999) (holding that the court should have applied Chevron deference to the BIA's "construction of the statute which it administers" in an adjudicatory proceeding). This is true even if the agency's interpretation overrides a court's prior judicial construction of the statute. See Brand X , 545 U.S. at 982, 125 S.Ct. 2688. And such a rule may even apply retroactively, if the rule has a permissible retroactive effect in a particular case. See Garfias-Rodriguez , 702 F.3d at 520.
III
We now consider whether Reyes's 2011 conviction under section 314.1 can be considered a crime involving moral turpitude under the BIA's decision in Cortes Medina .
A
The BIA deemed Reyes to be removable as an "alien who at any time after admission is convicted of two or more crimes involving moral turpitude." 8 U.S.C. § 1227(a)(2)(A)(ii). The phrase "crime involving moral turpitude" is inherently ambiguous, and neither we nor the BIA have established any clear-cut criteria "for determining which crimes fall within that classification and which crimes do not." Nunez , 594 F.3d at 1130. Because the phrase "crimes involving moral turpitude" refers to a category of crimes rather than a specific offense with identifiable elements, cf . 8 U.S.C. § 1101(a)(43), "the BIA has sensibly moved from trying to define the phrase itself to instead giving examples of the types of offenses that qualify as crimes involving moral turpitude." Ortega-Lopez v. Barr , 978 F.3d 680, 685 (9th Cir. 2020) (cleaned up). We have deferred under Chevron to this offense-by-offense approach when articulated by the BIA in a published opinion. See Marmolejo-Campos , 558 F.3d at 910–11 (deference for defining categories of crimes involving moral turpitude); Ceron v. Holder , 747 F.3d 773, 778 (9th Cir. 2014) (en banc) (deference under Chevron ).
The question whether a conviction under the statute at issue in this case, section 314.1, is a conviction for a crime involving moral turpitude has proven problematic. Before 2010, the BIA held in nonprecedential decisions that convictions for indecent exposure under section 314.1 were crimes involving moral turpitude. See Nunez , 594 F.3d at 1133. In 2010, however, Nunez held that a conviction under section 314.1 was not a crime involving moral turpitude because the full range of conduct that section 314.1 prohibits is not morally turpitudinous. See id. Specifically, we held that "[b]ecause nude dancers at bars and partially exposed purveyors of ‘sexual’ insults have been convicted under section 314.1, there is a ‘realistic probability, not a theoretical possibility, that [California] would apply [the indecent exposure] statute to conduct that falls outside the generic definition of [moral turpitude].’ " Id. at 1138 (quoting Gonzales v. Duenas-Alvarez , 549 U.S. 183, 193, 127 S.Ct. 815, 166 L.Ed.2d 683 (2007) ).
But in 2013, the BIA once again held that section 314.1 was a crime involving moral turpitude, this time in a precedential opinion. See Cortes Medina , 26 I. & N. Dec. at 81. The BIA held that an indecent exposure offense is a crime involving moral turpitude when it involves an element of lewd intent. Id. at 83. Applying this principle, the BIA held that section 314.1 was a crime involving moral turpitude because a finding of lewdness is necessary for a conviction under the statute. Id. at 84. Further, and contrary to Nunez , the BIA held that there was "no ‘realistic probability’ of a conviction in California under [section 314.1] for nude dancing or other conduct that does not involve moral turpitude." Id. at 86.
In Betansos v. Barr , we assumed that Cortes Medina enunciated a new agency rule promulgated through adjudication, and we reiterated the principle that "[u]nder Brand X , we must defer to the BIA's interpretation of [crime involving moral turpitude] in Cortes Medina unless its conclusion is unreasonable." 928 F.3d 1133, 1141 (9th Cir. 2019). After determining that BIA did not misrepresent any authorities, engaged in reasoned and thorough analysis, and relied on published BIA authority, we held that the BIA's interpretation of section 314.1 in Cortes Medina was reasonable, and therefore superseded Nunez . Id. at 1142.
Betansos next assumed that Cortes Medina had a retroactive effect, and turned to the question whether it was permissible to apply Cortes Medina retroactively to a petitioner who incurred a conviction under section 314.1 in 2002. Id. at 1143, 1145. To answer this question, Betansos applied the Montgomery Ward factors pertinent to immigration decisions to the petitioner's case. See id. at 1143 ; see also Garfias-Rodriguez , 702 F.3d at 520 (citing Montgomery Ward , 691 F.2d at 1334 ). Given our rule that the second and third Montgomery Ward factors are "closely intertwined," Betansos , 928 F.3d at 1143 (quoting Garfias-Rodriguez , 702 F.3d at 521 ), we analyzed them together. We held that the second factor favored the petitioner because Cortes Medina represented an abrupt departure from Nunez , "the first precedential opinion on the issue." Id. at 1143–44. Turning to the third factor (a party's reliance interest), we held that the petitioner could not take advantage of the rule that "reliance is presumed if the former, favorable rule was in place at the time the petitioner pleaded guilty or was convicted," because the petitioner pleaded guilty before Nunez had been decided. Id . at 1144–45. Without a presumption of reliance, we considered whether the petitioner had nevertheless relied on Nunez by expending fees, making strategic decisions, or choosing not to apply for other forms of relief while Nunez was operative. Id. at 1145. We concluded that the petitioner failed "to identify a specific event or action that he took (or failed to take) in the past in reliance on Nunez that now carries new consequences or burdens under Cortes Medina ," and therefore the third factor weighed against the petitioner. Id. While the fourth factor weighed in the petitioner's favor because deportation creates a substantial burden, the fifth factor tipped toward the government based on the need for uniformity in immigration law. Id. After weighing these factors in light of each other, we held that it was permissible to apply Cortes Medina retroactively to the petitioner in that case. Id. Given that the petitioner's section 314.1 conviction in 2002 constituted a crime involving moral turpitude under Cortes Medina , we denied the petition for review. Id. at 1145–46.
B
We now turn to the question whether Cortes Medina ’s interpretation of "crime involving moral turpitude" as including convictions under section 314.1 has an impermissible retroactive effect on Reyes's 2011 conviction.
Reyes does not argue that his 2014 conviction is not a crime involving moral turpitude. Nor could he, because Cortes Medina was decided in 2013, and Reyes could no longer reasonably rely on Nunez . See Garfias-Rodriguez , 702 F.3d at 522.
1
We first consider whether applying Cortes Medina to Reyes's guilty plea in 2011 would have any retroactive effect at all. Both the government and the dissent argue that it would not. In their view, after Cortes Medina was decided in 2013, Reyes was on notice that if he incurred another section 314.1 conviction, he would be deemed to have committed two crimes involving moral turpitude. Dissent at 1004-05. Therefore, they reason, Reyes cannot identify any act he took in reliance on Nunez that now creates a new burden or disability under Cortes Medina . Dissent at 1004-05. Accordingly, the government and the dissent claim that Cortes Medina is not retroactive as to Reyes.
We disagree, because this argument is contrary to the Supreme Court's definition of retroactivity. Landgraf made clear that a law is retroactive "if it changes the legal consequences of acts completed before its effective date," gives "a quality or effect to acts or conduct which they did not have or did not contemplate when they were performed," or "attaches a new disability" to prior conduct. 511 U.S. at 269–70 & n.23, 114 S.Ct. 1483 (cleaned up).
In this case, Cortes Medina changed the legal consequences of Reyes's 2011 conviction under section 314.1 in two ways. First, as a result of the interaction of California criminal law and federal immigration law, Reyes became removable under 8 U.S.C. § 1227(a)(2)(A)(i) after Cortes Medina classified his section 314.1 offense as a crime involving moral turpitude. Specifically, Reyes had already been convicted of a violation of section 314.1 before 2011. Under California law, a second offense under section 314.1 is a felony punishable by a prison term of at least 16 months, see Cal. Penal Code §§ 314.1, 1170(h)(1). Under immigration law, an alien who is convicted of a crime involving moral turpitude for which a sentence of one year or longer may be imposed, is removable. 8 U.S.C. § 1227(a)(2)(A)(i). Such an alien is also ineligible for cancellation of removal. See 8 U.S.C. § 1229b(b)(1)(C). Therefore, Cortes Medina attached new legal consequences to Reyes's 2011 conviction of a section 314.1 offense.
Second, after Cortes Medina was decided, Reyes's 2011 conviction of a section 314.1 offense became a potential ground for removal under § 1227(a)(2)(A)(ii) (an alien convicted of two crimes involving moral turpitude). Although the new legal consequences of pleading guilty to a crime involving moral turpitude did not affect Reyes until he pleaded to a further violation in 2014, it is often the case that the burden imposed by a retroactive rule becomes apparent only after a party engages in additional conduct. In Vartelas , for instance, the new legislation which reclassified certain offenses as crimes involving moral turpitude did not affect the lawful permanent aliens before the court until they made a brief departure from the United States. 566 U.S. at 274–75, 132 S.Ct. 1479 ; see also Camins v. Gonzales , 500 F.3d 872, 884–85 (9th Cir. 2007) (same). Because Cortes Medina "attache[d] a new disability" to Reyes's guilty plea in 2011, it has a retroactive effect regardless whether Reyes realized the consequences immediately. See Landgraf , 511 U.S. at 269 n.23, 114 S.Ct. 1483.
2
Although the BIA's adjudicatory rule in Cortes Medina has a retroactive effect on Reyes's 2011 conviction, it is not necessarily an impermissible retroactive effect that is "fatal to its validity." See Chenery II , 332 U.S. at 203, 67 S.Ct. 1760. We must still determine whether it is permissible to apply Cortes Medina retroactively to Reyes. We do so by applying the Montgomery Ward factors to weigh the burdens and benefits of retroactivity in this case. In conducting this analysis, we are guided by Betansos.
The dissent argues that Cortes Medina does not have an impermissible retroactive effect because it is like a recidivist statute, which can permissibly increase the consequences of a decision to reoffend in the future. Dissent at 1005-06 (quoting Rummel v. Estelle , 445 U.S. 263, 278, 100 S.Ct. 1133, 63 L.Ed.2d 382 (1980) ). But the cases analyzing recidivist statutes are not relevant here. Because Congress enacts recidivist statutes in order to impose enhanced penalties on defendants based on their criminal history, Congress necessarily authorizes the retroactive effect of such statutes. See Landgraf , 511 U.S. at 280, 114 S.Ct. 1483 (holding that a statute may apply retroactively when "clear congressional intent favor[s] such a result"). Because the "presumption against retroactive legislation," id. at 265, 114 S.Ct. 1483, does not apply to recidivist statutes, challenges to recidivist statutes necessarily focus on "contentions that [the statute] violated ‘constitutional strictures dealing with double jeopardy, ex post facto laws, cruel and unusual punishment, due process, equal protection, and privileges and immunities’ " rather than on the question whether the statute had a retroactive effect not authorized by Congress, Rummel , 445 U.S. at 268, 100 S.Ct. 1133 (quoting Spencer v. Texas , 385 U.S. 554, 560, 87 S.Ct. 648, 17 L.Ed.2d 606 (1967) ). By contrast, the retroactive effect of a rule made through agency adjudication is impermissible unless it passes muster under Montgomery Ward .
As indicated in Betansos , the first factor (whether the case is one of first impression) does not apply in the immigration context.
Moving to the second factor, "whether the new rule represents an abrupt departure from well established practice," Betansos already decided that Cortes Medina abruptly departed from Nunez , and so the second factor favors Reyes. See Betansos , 928 F.3d at 1143–44.
The third factor, the extent to which Reyes relied on Nunez , also favors Reyes. Although the reliance analysis "is highly fact dependent and conducted on a case-by-case basis," id. at 1146 n.6, the Supreme Court has indicated that "as a general matter, alien defendants considering whether to enter into a plea agreement are acutely aware of the immigration consequences of their convictions," St. Cyr , 533 U.S. at 322, 121 S.Ct. 2271 ; see also Garcia-Martinez v. Sessions , 886 F.3d 1291, 1295 (9th Cir. 2018) (relying on this presumption in applying the Montgomery Ward factors). Therefore, although Reyes "has not shown that he in fact relied on Nunez ," see Betansos , 928 F.3d at 1145 (emphasis added), there is a rebuttable presumption that Reyes knew about and relied on Nunez when he pleaded guilty to a section 314.1 offense in 2011. The government has not rebutted this presumption.
The dissent (but not the government) argues that the presumption that Reyes relied on Nunez is rebutted because unpublished BIA decisions issued before Nunez held that convictions under section 314.1 were crimes involving moral turpitude. Dissent at 1002, 1003 n.4. This is incorrect, because we have ruled that aliens are entitled to rely on our published decisions even when earlier BIA opinions "should have enabled noncitizens ... to predict" that our precedent "would not survive." Acosta-Olivarria v. Lynch , 799 F.3d 1271, 1276 (9th Cir. 2015). Our rule on this point is equally applicable when the earlier BIA opinions are unpublished, and therefore do not establish binding policies or practices, see 8 C.F.R. § 1003.1(d)(1) (providing that the BIA issues guidance, policies, and practices "through precedent decisions"). Because aliens within the Ninth Circuit can "rely on our opinions in making decisions," Acosta-Olivarria , 799 F.3d at 1276, despite the existence of prior BIA opinions, Reyes was entitled to rely on Nunez before the BIA issued its precedential decision in Cortes Medina .
In applying the fourth and the fifth factors (the degree of burden on Reyes and the statutory interest in applying a new rule), we are again guided by Betansos . See id. at 1145–46. Cortes Medina significantly burdens Reyes because it would make his 2011 conviction a crime involving moral turpitude, which carries unfavorable immigration consequences. See id. at 1145. As in Betansos , the fifth factor tips toward the government because "non-retroactivity impairs the uniformity of a statutory scheme, and the importance of uniformity in immigration law is well established." See id. (quoting Garfias-Rodriguez , 702 F.3d at 523 ).
The government has not attempted to demonstrate that uniformity in this context has some special importance or weight.
Taking these factors together, the second, third, and fourth factors all weigh against applying Cortes Medina to Reyes's 2011 conviction, and the fifth factor, which weighs in favor of retroactivity, is not dispositive. Because imposing new legal consequences on Reyes's decision to plead guilty to a section 314.1 offense would conflict with principles of "fair notice, reasonable reliance and settled expectations," St. Cyr , 533 U.S. at 323, 121 S.Ct. 2271 (quoting Landgraf , 511 U.S. at 270, 114 S.Ct. 1483 ), and the government has not shown any significant countervailing concerns beyond the general interest in uniformity in the statutory scheme, we conclude that it would be impermissible to apply Cortes Medina to Reyes's 2011 conviction.
The dissent's central reason for claiming that Cortes Medina is not impermissibly retroactive amounts to little more than a disagreement with Betansos . First, the dissent argues, we should ignore Betansos ’s conclusion on the second Montgomery Ward factor (that Cortes Medina abruptly departed from Nunez ) because the third factor (reliance) was controlling in Betansos , and therefore the second factor was irrelevant to the facts of the case and "obviously dictum." Dissent at 1001-02. This argument evinces a misunderstanding of the Montgomery Ward factors. In determining whether a new rule has an impermissible retroactive effect, we must engage in a balancing test that weighs all factors together, with the second and third factors "closely intertwined," and no one factor being dispositive. See Garfias-Rodriguez , 702 F.3d at 523. Because no single Montgomery Ward factor is controlling, Betansos ’s analysis of the second factor was "an issue germane to the eventual resolution of the case" that was resolved "after reasoned consideration" and is therefore not dicta. See Miranda B. v. Kitzhaber , 328 F.3d 1181, 1186 (9th Cir. 2003) (per curiam) (quoting United States v. Johnson , 256 F.3d 895, 914 (9th Cir. 2001) (en banc)). Accordingly, "we are bound by our precedent." Hammad v. Holder , 603 F.3d 536, 544 n.9 (9th Cir. 2010) (citing Miller v. Gammie , 335 F.3d 889, 900 (9th Cir. 2003) (en banc)).
Moreover, the dissent errs in asserting that Betansos did not have to address the second factor (whether Cortes Medina was an abrupt departure from Nunez ) for the petitioner in that case. Dissent at 1005 n.6; see also id. at 1001-02. Although the petitioner in Betansos was convicted before Nunez was decided, which eliminates the presumption of reliance that exists when "the former, favorable rule was in place at the time the petitioner pleaded guilty or was convicted," Betansos , 928 F.3d at 1144, the petitioner argued that he relied on Nunez by citing its determination that section 314.1 is not a crime involving moral turpitude during his legal proceedings, id. at 1145. Betansos rejected the petitioner's argument as not being "the type of specific reliance interest we have generally held sufficient," but indicated that petitioner could have demonstrated reliance by "identify[ing] a specific event or action that he took (or failed to take) in the past in reliance on Nunez that now carries new consequences or burdens under Cortes Medina. " Id.
Second, the dissent contends that the third factor (reliance) does not favor Reyes because the state of the law in 2011 when Reyes pleaded guilty is irrelevant; rather, the dissent argues, all that matters is the state of the law in 2014 when Reyes was subject to the conviction that triggered Reyes's eligibility for removal. Dissent at 1006-07. This view is erroneous, because the focus of the retroactivity analysis is on whether Cortes Medina changed the legal consequences of Reyes's 2011 conviction, not on how it affected Reyes's later actions.
We conclude that the BIA improperly applied Cortes Medina to Reyes's 2011 conviction under section 314.1, and therefore that conviction may not be deemed a crime involving moral turpitude. Because Reyes has only one conviction for a crime involving moral turpitude, and the government sought to remove him for having two convictions for crimes involving moral turpitude, § 1227(a)(2)(A)(ii), we grant the petition for review on this ground. On remand, the agency may consider additional evidence from the parties, including evidence rebutting the presumption that Reyes relied on Nunez when he pleaded guilty under section 314.1 in 2011.
Because we hold the BIA erred in ruling that Reyes was removable under 8 U.S.C. § 1227(a)(2)(A)(ii), we do not address Reyes's additional claims that the BIA erred in denying him relief from removal.
PETITION GRANTED.
VANDYKE, Circuit Judge, dissenting:
Yet again we are presented with a case study in how the Ninth Circuit's abysmal and indefensible immigration precedents are the gifts that keep on taking—even where, as here, my colleagues in the majority are simply trying to faithfully apply our faithless precedent. My colleagues feel cabined by a chain of errors from our past, initiated when Judge Reinhardt disdainfully dismissed California jury verdicts and the BIA, pronouncing that, in his view, "lewdly ... [e]xpos[ing] ... private parts ... in any public place" is neither "base, vile, and depraved," nor does it "shock the conscience." Nunez v. Holder , 594 F.3d 1124, 1130, 1138 (9th Cir. 2010) (citation omitted). After the BIA rushed to correct our grossly wrong precedent, see Matter of Cortes Medina , 26 I. & N. Dec. 79, 83 (BIA 2013), we begrudgingly acknowledged that the BIA was right after all, and that "lewd intent" in "willful exposure" to unwilling victims indeed constitutes a crime involving moral turpitude (CIMT). See Betansos v. Barr , 928 F.3d 1133, 1142 (9th Cir. 2019). But like a politician who can't even manage a good mea culpa, our Betansos decision somehow concluded that the BIA's Cortes Medina decision was a "complete surprise," even though the BIA had reached exactly that conclusion in case after case for well over a decade by that point—including in the Nunez decision reversed by Judge Reinhardt. Id. at 1144. One wonders how our court can be so bad at immigration law.
Presented with this mess, the majority apparently believes it must hold its nose and propagate the errors of Nunez and Betansos due to the timing of one of the petitioner's crimes in this case. Jorge Reyes Afanador committed several sexual assaults and received multiple convictions for indecent exposure. Congress created an obligation to deport those who commit multiple CIMTs, 8 U.S.C. § 1227(a)(2)(A)(ii), and Cortes Medina and Betansos authoritatively hold that convictions such as Reyes's are CIMTs. Reyes seems clearly deportable, but must we now nevertheless wring our hands because our court incorrectly second-guessed the BIA in Nunez , and the first of Reyes's two crimes occurred in the interstitial period between Nunez ’s publication in 2010 and the BIA's correction of that error in Cortes Medina in 2013? I don't think so. Due to the multiple crimes Reyes committed—the key conviction occurring after Cortes Medina was decided—I don't believe we are forced to extend in this case our past distortion of the "fluid boundaries" of "vile" conduct constituting a CIMT. Nunez , 594 F.3d at 1132. We should have denied Reyes's petition.
1. The Errors of Nunez
The collateral consequences of our reprehensible Nunez decision reverberate through the majority opinion. The petitioner in Nunez had convictions for petty theft (a CIMT) and indecent exposure under California Penal Code § 314.1. Nunez , 594 F.3d at 1128–29. He argued before our court that the BIA erred in concluding that his indecent exposure conviction was also categorically a CIMT. Id. at 1129. After acknowledging that "[m]orality is not a concept that courts can define by judicial decrees," id. at 1127, the Nunez majority proceeded to do exactly that. It concluded the conduct criminalized by section 314.1 was not categorically a CIMT because it simply "encompass[ed] mere acts of provocation, bad taste, and failed humor." Id. at 1138. The Nunez majority determined that the unpublished BIA opinion under review improperly categorized section 314.1 convictions as crimes involving moral turpitude because, in the majority's view, section 314.1 clearly encompassed non-criminal conduct and could realistically apply to such conduct. Id.
In reaching this conclusion, the Nunez majority minimized the "turpitudinous" nature of a section 314.1 crime, allowing that while "California's indecent exposure statute criminalizes a range of conduct that offends the sensibilities of many, and perhaps most, people," id. at 1133, the panel theorized that perhaps the victims of intentional lewd exposure of a defendant's private parts were only "in theory ... offended by the conduct, even if, in actuality, they [were] not." Id. at 1134. Citing nude dancing, a man who "in a fit of ‘road rage’ exposed his penis and yelled" a vulgar remark at a female driver, and a boy who intentionally engaged the attention of his "two female classmates" and exposed himself, id. at 1137, the majority declared such behavior "relatively harmless." Id. at 1138. Discounting section 314.1's requirement of lewd intent—found satisfied by a jury in the road rage case and the state court of appeals in the classroom exposure case—the Nunez majority determined these examples were simply instances of "transitory nudity" and included only "brief reference[s] to sex." Id. Ultimately, the majority concluded such acts of exposure with sexual intent to unwilling victims were "not ‘base, vile, and depraved,’ nor do they shock the conscience," and were thus not CIMTs. Id.
Nunez reflects a teachable moment for the judiciary: judges should be more circumspect in establishing themselves as the arbiters of sexual morality in society. See Nunez , 594 F.3d at 1133, 1138 (opinion of Reinhardt, J.) (opining that "society is past the point where ... a brief reference to sex necessarily transforms an otherwise de minimus provocation into a morally turpitudinous offense," particularly for "mere acts of provocation, bad taste, and failed humor" that do not, in the judge's view, "shock the conscience").
The court in Nunez reversed the BIA by a bare two-judge majority. The dissent argued that the judiciary's perception of what should be defined as a "tasteless prank" rather than sexual assault does not, in and of itself, change the law or the facts. Id. at 1139 (Bybee, J., dissenting). The dissent observed first that the nude dancing case cited by the majority as evidence for "a realistic probability ... that [California] would apply its statute to conduct" falling outside of the definition of a CIMT was both "expressly disapproved" by the California Supreme Court and had "not been cited" since 1982. Id. (citation omitted). As to the road rage and classroom exposure cases relied on by the majority, the dissent derided the majority's "collateral attack on these convictions" as "wholly inappropriate ... and, at best, revisionist history." Id. In the road rage case, the California Court of Appeal determined that the defendant's exposure "for the purpose of sexually insulting ... the other person" demonstrated sufficient lewd intent to violate section 314.1. Id. at 1147 (citation omitted). In the classroom case, the California Court of Appeal expressly found "facts suggesting lewd intent" because the student who exposed himself took "deliberate action directed at two young girls." Id. at 1148 (citation omitted). The dissent in Nunez concluded that, as "a collateral attack, [the majority opinion] is sorely misplaced," explaining "[i]f ... the majority means to remove lewd conduct from the category of crimes involving moral turpitude, its discussion is a wholesale assault on sex crimes as crimes involving moral turpitude." Id.
2. The Restoration by Cortes Medina
The BIA quickly "remed[ied the] deficiency" exploited by the Nunez majority and published a decision defining crimes under section 314.1 as CIMTs, unsurprisingly adopting the analysis from the Nunez dissent. Cortes Medina , 26 I. & N. Dec. at 81 n.3. The BIA applied the categorical approach to determine whether section 314.1 demonstrated a "realistic probability" of criminalizing actions not involving moral turpitude. Id. at 82 (citation omitted). Because section 314.1 required lewd intent, the BIA surveyed the caselaw and concluded that the element of intent brought "the offense of indecent exposure within the definition of a crime involving moral turpitude." Id. at 83.
Analyzing California court decisions involving section 314.1, the BIA noted that the defendants all "had the requisite obscene or indecent intent at the time of the offense" and determined that the Nunez majority created a "definition of moral turpitude [that was] too narrow." Id. at 83–84. Adopting the Nunez dissent's examination of the road rage and classroom exposure cases, the BIA concluded that neither the Nunez majority nor the respondent in the case before it "presented any evidence that California applied the statute" to conduct that was "not morally turpitudinous." Id. at 86. As such, the BIA decided that a conviction under section 314.1 was "categorically a crime involving moral turpitude for purposes of the immigration laws." Id.
Post- Cortes Medina , our court in Betansos reluctantly admitted that the BIA had overruled Nunez . Although "hesitat[ing] to defer to the BIA's general understanding of the term ‘moral turpitude,’ " Betansos , 928 F.3d at 1139 (citation omitted), this court ultimately deferred to Cortes Medina because the panel could "[ ]not say that the BIA's decision [was] unreasonable" and thus had to defer under Chevron v. Natural Resources Defense Council, Inc. , 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). Betansos , 928 F.3d at 1142.
But because Betansos committed his section 314.1 crime in 2002, the panel also analyzed whether it could apply Cortes Medina retroactively pursuant to Montgomery Ward & Co., Inc. v. FTC , 691 F.2d 1322 (9th Cir. 1982). Betansos , 928 F.3d at 1143–46. The Betansos panel analyzed the third Montgomery Ward factor—the petitioner's reliance on a prior rule—and concluded that Betansos could not demonstrate a reliance interest on our court's wrong Nunez decision because Betansos's last section 314.1 conviction was in 2002, long before Nunez was decided. Betansos , 928 F.3d at 1145. This holding on the third factor was dispositive because "Betansos [could] not show[ ] that he in fact relied on Nunez (under Montgomery Ward factor three)." Id.
Notwithstanding that the third factor controlled Betansos's appeal, the panel also asserted, despite its irrelevance to the facts of the case, that the panel believed "it would have been reasonable to rely on Nunez between February 2010 and January 2013 (under Montgomery Ward factor two)." Id. This statement was obviously dictum. And as is often the case with dicta, because it was untethered from any real facts actually presented in that case, in hindsight it was also obviously wrong. The panel characterized Cortes Medina as a "complete surprise" because it "represent[ed] an ‘abrupt departure’ from Nunez ." Id. at 1143–44 (quoting Garfias-Rodriguez v. Holder , 702 F.3d 504, 521 (9th Cir. 2012) (en banc)). But this ignored years of unpublished BIA decisions prior to Nunez —including the BIA's decision in Nunez itself—that consistently concluded that convictions under section 314.1 are CIMTs. By the time our court in Nunez concluded that a section 314.1 violation was not a CIMT, the BIA had treated crimes under section 314.1 as CIMTs for years, and previously had gone back and forth with our court over that definition in unpublished decisions. See, e.g. , Pannu v. Holder , 639 F.3d 1225, 1227 (9th Cir. 2011) (noting that the BIA in 2004 held that Pannu's convictions under section 314.1 were CIMTs, but in 2006 this court reversed the BIA in a memorandum disposition because "indecent exposure convictions were not categorically CIMTs," causing the BIA to change its grounds for denying Pannu relief); In re David Coronado Orozco , No. AXXXXX3077-SAN, 2008 WL 4722691, at *1 (BIA Oct. 3, 2008) (unpublished) ("[W]e agree with the Immigration Judge that the respondent's conviction under California Penal Code § 314.1 is a crime involving moral turpitude."); In re Amir Gholamali Khaksarian , No. AXXXX4241-SAN, 2007 WL 1676924, at *1 n.1 (BIA May 18, 2007) (per curiam) (unpublished) (acknowledging that, given the state of the law before the BIA, "[t]he respondent concede[d] that his conviction for indecent exposure [under section 314.1 was] a CIMT"). The Betansos panel considered none of this, but instead cursorily asserted that "[t]his is not a case where there was an ongoing conversation or back-and-forth between this Court and the BIA about the proper interpretation." Betansos , 928 F.3d at 1144.
Even though Betansos pled guilty in 2002 and therefore could not have relied on our circuit's 2010 Nunez decision when he pled , the majority here contends that Betansos ’s statement that "it would have been reasonable to rely on Nunez between February 2010 and January 2013" was not dictum because Betansos "could have demonstrated reliance" in other ways by "identifying a specific event or action he took ... in reliance on Nunez " (quoting Betansos ). But Betansos did not actually assert any reliance on Nunez —whether in why he pled guilty in 2002 or in any of the other ways referenced by the majority—so Montgomery Ward ’s third factor controlled everything in that decision, including any hypothetical way that Betansos theoretically "could have" relied on Nunez (but didn't). The majority's attempt to rely on counter-factual possibilities that Betansos "d[id] not assert," 928 F.3d at 1145, is as nonsensical as saying that the Betansos ’s second-factor discussion was not dicta because, even though Betansos pled guilty in 2002, he "could have" pled guilty in 2011 (in some parallel universe).
The majority also objects that "Betansos ’s analysis of the second factor was ‘an issue germane’ " to the case's resolution and thus could not be dicta, "[b]ecause no single Montgomery Ward factor is controlling." Perhaps no single factor is controlling, but obviously one or more factors can be irrelevant in a particular case. See, e.g. , Great W. Bank v. Off. of Thrift Supervision , 916 F.2d 1421, 1432 (9th Cir. 1990) ("considering the relevant [Montgomery Ward ] factors," the court determined that "two elements weigh decisively in favor of the Bank Board," highlighting that there was no "well-established practice" and "the strong interest in applying a rule that corresponds to the plain language of the statute," and did not discuss the other three factors). Indeed, the majority here acknowledges this by observing that the first factor "does not apply in the immigration context" at all. In Betansos , the second factor—i.e., whether Cortes Medina was a "complete surprise" in light of Nunez —was just as irrelevant as the first, given that "Betansos could not have relied on Nunez when he pleaded guilty in 2002 because Nunez had not yet been decided." 928 F.3d at 1145 (emphasis added). The entire discussion of the second factor made no difference to the outcome in Betansos and was wholly unnecessary. Id. at 1143–45.
And while the BIA's decision in Cortes Medina obviously was "an ‘abrupt departure’ from Nunez ," id. at 1143, that is not the test under Montgomery Ward ’s second factor. That factor asks whether the BIA's decision was "abrupt departure from well established practice ." Montgomery Ward , 691 F.2d at 1333 (emphasis added). The BIA's "well established practice" dating back to at least 2004 was exactly the same as its ruling in Cortes Medina : that a violation of section 314.1 is a CIMT. Given the ongoing "back-and-forth between this Court and the BIA" since at least 2004, the most that could be fairly said was that Cortes Medina authoritatively "fill[ed] a void in an unsettled area of law," Betansos , 928 F.3d at 1144 (citation omitted)—"unsettled" only because, until our 2019 Betansos decision, we refused for a decade-and-a-half to defer to the BIA's consistent position.
The majority in this case makes two unsuccessful attempts to defend the substance of Betansos ’s indefensible conclusion that Cortez Medina represented a "complete surprise." Id. First, the majority tries to subtly refashion Montgomery Ward ’s second-factor inquiry from whether the agency's "new rule represents an abrupt departure from well established practice ," Montgomery Ward , 691 F.2d at 1333 (emphasis added), into asking instead whether the agency had any "adjudicatory rule" formally promulgated through the agency's precedential decision-making. If the agency didn't have any such formal "adjudicatory rule" arising from a precedential decision, then, per the majority, none of the agency's past practice matters. But of course, if all that Montgomery Ward ’s second factor cared about was whether the agency's new rule was a departure from an old "rule" (adjudicatory or otherwise), it could have easily said that—and would have. Instead, the Montgomery Ward test focuses on changes from past agency "practice" or "policy," not just an agency's past formal "rules."
In a footnote, the majority makes the same novel argument in a different way: that unpublished BIA decisions "do not establish binding policies or practices" because our court's unpublished judicial decisions are not precedential. Putting aside whether the comparison between an agency's and a court's decisions is an apt one for this context, the majority is again trying to refashion the Montgomery Ward test from considering the agency's past "practice" to considering only its "binding ... practices." The second factor of Montgomery Ward has never been so constrained, focusing only on an agency's past "practices" that were technically "binding" on other parties—i.e., formal "rules." Rather, it broadly asks whether the new rule is inconsistent with the agency's "well established practice ."
This is clearly demonstrated by the test's namesake case itself. Montgomery Ward was a case addressing FTC rules regarding the placement of warranty information around Montgomery Ward's department stores, and the FTC argued that the binders containing warranty information should have been available on every floor. 691 F.2d at 1326–27. In analyzing whether the FTC's current interpretation of its rule demonstrated "an abrupt departure from well established practice," the court in Montgomery Ward observed that the plaintiff's prior "involvement in the rule-making process would indicate that [the new agency interpretation of the rule in its order] would not be a complete surprise," id. at 1333–34, given that certain aspects of the new interpretation were "presaged in the commentary published in the Federal Register upon the promulgation of the presale rule," id. at 1330. Montgomery Ward ’s discussion of the activity prior to the promulgation of the agency's formal rule demonstrates the second factor's inquiry does not stop at simply surveying past formal "rules." "Well established practice" is a broader inquiry.
This is further demonstrated by the two examples contrasted in Montgomery Ward ’s articulation of the second factor. On one hand, "an abrupt departure from well established practice" by the agency cuts against retroactive application of a new rule. Id. at 1333. On the other hand, where an agency is "merely attempt[ing] to fill a void in an unsettled area of law," that cuts in favor of retroactive application. Id. Here, both of Montgomery Ward ’s paradigmatic examples favor the government. The BIA never "depart[ed] from its well established practice," and because, prior to Cortez Medina , it had not issued a formal rule, it was also "fill[ing] a void in an unsettled area of law."
To the extent the majority implies that the BIA's nonprecedential opinions do not put aliens on notice, there is no basis for that conclusion. The point is not simply that unpublished decisions, in isolation, would put an alien on notice; the point is that an alien could observe through Nunez that there was a clear "back and forth" between this court and the BIA, which would prompt an analysis of the BIA's position—which would be clear via the years of unpublished decisions. Any alien who read Nunez itself would know both that (1) it was inconsistent with the BIA's position, and (2) the BIA has the last word on such questions. Nunez , 594 F.3d at 1129. It is clear other petitioners were well aware of the BIA's consistent practice, see, e.g. , In re Amir Gholamali Khaksarian , 2007 WL 1676924, at *1 n.1, and, as discussed, it is the agency's "practice" that matters.
Which leads directly into the majority's second failed attempt to bolster Betansos ’s unnecessary dicta. The majority strangely concludes that, while years of the BIA's consistent unpublished decisions did "not establish the practice or policy of the agency," somehow our court's Nunez decision did. See also Betansos , 928 F.3d at 1144 (" Cortez Medina did not ‘fill a void.’ Nunez had already filled the void ...."). This is directly at odds with our en banc court's recognition that the agency gets to tell us what ambiguous statutory terms like "moral turpitude" mean, not the other way around. See Garfias-Rodriguez , 702 F.3d at 523 ("any reliance ... placed on our decisions h[o]ld[s] some risk because our decisions [a]re subject to revision by the BIA under Chevron and [ National Cable & Telecommunications Association v. Brand X Internet Services (Brand X) , 545 U.S. 967, 125 S.Ct. 2688, 162 L.Ed.2d 820 (2005) ]"). Betansos ’s statement that our court's Nunez decision constituted the well established practice of the BIA is obviously not very well thought through, and the majority's decision to credit it flips the normal deference we give agencies under Chevron and Brand X .
Both because the BIA's Cortes Medina decision (1) was not an "abrupt departure from well established practice," and because it (2) settled an "ongoing ... back-and-forth between this Court and the BIA about the proper interpretation" of ambiguous language in INA, Montgomery Ward ’s second factor obviously cuts strongly in favor of the government. Betansos , 928 F.3d at 1143–44. The Betansos panel reached the opposite conclusion only because it addressed the issue as dicta, "casually and without analysis" and as "merely a prelude to" Montgomery Ward ’s third factor—the "legal issue that command[ed] the panel's full attention" because it controlled the outcome in Betansos . United States v. Johnson , 256 F.3d 895, 915 (9th Cir. 2001) (en banc) (Kozinski, J., concurring).
3. Our Court's Embarrassing Nunez Blunder Does Not Benefit Reyes
Even putting all that aside, and assuming we were still operating in the vestiges of the shadow cast by our court's errors in Nunez and erroneous dicta in Betansos , we still would not need to remand this case. While the majority concludes the lingering presence of Nunez forces its hand, I arrive at a different conclusion on how Montgomery Ward factors two and three apply to Reyes's multiple convictions.
The majority inexplicably mischaracterizes my dissent as "little more than a disagreement with Betansos ." While I do argue that Betansos ’s dicta was wrong and that we need not follow it, this section of my dissent explains why, even applying that dicta, the majority reaches the wrong conclusion.
Under the third factor, while the majority argues that Cortes Medina had a retroactive effect on Reyes's 2011 section 314.1 indecent exposure conviction, Reyes's removability was only triggered by the commission of his second CIMT in 2014—after the BIA had decided Cortes Medina . Reyes was on notice by that point that he could not rely on Nunez and that his conviction of a second CIMT under California Penal Code section 314.1 would make him removable. Thus, Montgomery Ward ’s third factor weighs against Reyes in this case.
As our court has repeatedly recognized, the "second and the third factors are closely intertwined," so it's not surprising that the second factor also goes against Reyes. Betansos , 928 F.3d at 1143 (citation omitted). Like the third factor, since Cortes Medina had already issued before his second section 314.1 indecent exposure conviction, Reyes can hardly argue that the immigration consequences of his second CIMT conviction were a "complete surprise" or "abrupt departure from well established practice" by the BIA. Id. (citation omitted). Once Cortes Medina issued, Reyes could not justifiably rely on Nunez in such a conflict because our court is not the final arbiter of CIMT definitions—the Attorney General is, as our court was forced to acknowledge in Betansos . See also Garfias-Rodriguez , 702 F.3d at 523 ("any reliance ... placed on our decisions h[o]ld[s] some risk because our decisions [a]re subject to revision by the BIA under Chevron and Brand X "). And even if, until Betansos , Reyes had reason to hope our court would ultimately refuse to defer to the BIA, the most he could argue was that his convictions occurred when the immigration consequences of those convictions were "an unsettled area of law." Betansos , 928 F.3d at 1143 (citation omitted). Again, that would mean Montgomery Ward ’s second factor goes against him. Cf. 928 F.3d at 1143–44.
The majority seems to assume that because Montgomery Ward ’s second and third factors are often intertwined, the panel in Betansos had to decide the second factor, even when the third factor was decisive. But that position ignores the reason the two factors are often intertwined: because it is only when an agency's action arises out of well established practice that a petitioner can establish reliance on that practice. Where there was no reliance, as in Betansos , it makes no difference whether the practice was well established. That the factors are often related doesn't mean the court always must address both. Indeed, as in Betansos , precisely because they are related sometimes one factor will control, rendering the other superfluous.
A. Montgomery Ward Factor 3
The majority focuses solely on Reyes's 2011 crime and determines that, based on caselaw where a change in the applicable law changed the impact of a single crime on the petitioner's life, Cortes Medina "attache[d] a new disability" to Reyes's 2011 conviction for indecent exposure under California Penal Code section 314.1 by defining it as a CIMT. This analysis overlooks a key point: had Reyes only been convicted that single time in 2011, Cortes Medina would have no impact whatsoever on his immigration status—Reyes needed two convictions for indecent exposure to trigger his deportation based on CIMT convictions. See 8 U.S.C. § 1227(a)(2)(A)(ii). It was the last domino to fall—the 2014 indecent exposure conviction—that triggered Reyes's eligibility for deportation.
Following an unfortunately well-traveled path for our court, the majority supports this part of its decision by advocating for the petitioner, making arguments on his behalf that he never made himself. See United States v. Sineneng-Smith , ––– U.S. ––––, 140 S. Ct. 1575, 1579, 206 L.Ed.2d 866 (2020) (chastising our court for ignoring the party presentation principle). Citing 8 U.S.C. § 1227(a)(2)(A)(i), the majority defends its blinkered focus on Reyes's 2011 conviction by arguing that that conviction alone might have created other hypothetical immigration consequences. But Reyes presumably didn't make that argument because it's irrelevant. The government has not sought to remove Reyes under § 1227(a)(2)(A)(i). What matters for purposes of reliance in this case is whether Reyes could have relied on Nunez as giving him a free pass under the only provision at issue in this case : § 1227(a)(2)(A)(ii). As explained, when he committed his second CIMT in 2014 after Nunez was gone, he could not. The majority's speculative invocation of § 1227(a)(2)(A)(i) is strange. Conjecture that the government's hypothetical use of a past conviction could create retroactivity problems in an entirely different imaginary case involving an entirely different statutory provision says nothing about the retroactivity or reliance issue in this case , and I am unaware of any caselaw to the contrary.
The cases the majority cites highlight this distinction, as Justice Ginsburg specifically distinguished the retroactive application of the law at issue in Vartelas from "prosecutions depend[ing] on criminal activity ... occurring after the provision's effective date." Vartelas v. Holder , 566 U.S. 257, 270, 132 S.Ct. 1479, 182 L.Ed.2d 473 (2012). Reyes's case was not a situation in which he continued acting lawfully and the law changed in a way that would impact his life based on a "crime he was ‘helpless to undo.’ " Id. (citation omitted). Immediately after Cortes Medina was decided, Reyes was and would have remained unaffected—as long as he did not commit another indecent exposure.
Instead, Reyes recidivated and was again convicted of indecent exposure in 2014. As such, the effect of Cortes Medina is more like an "[e]nhanced punishment imposed for the later offense[, which] is not to be viewed as an additional penalty for the earlier crimes, but instead, as a stiffened penalty for the latest crime, which is considered to be an aggravated offense because [it is] a repetitive one." Id. at 271–72, 132 S.Ct. 1479 (cleaned up) (quoting Witte v. United States , 515 U.S. 389, 400, 115 S.Ct. 2199, 132 L.Ed.2d 351 (1995) ); cf. Landgraf v. USI Film Prods. , 511 U.S. 244, 280, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994) (observing that statutes have an impermissible retroactive effect when they "impose new duties with respect to transactions already completed " (emphasis added)). Increasing punishment for repetitive criminal convictions after an individual commits his first crime, and imposing that enhancement only after the individual is convicted for a subsequent crime, does not constitute an impermissible retroactive application. See Rummel v. Estelle , 445 U.S. 263, 278, 100 S.Ct. 1133, 63 L.Ed.2d 382 (1980) (affirming the trial court's application of Texas's recidivist statute, passed after Rummel's first two felonies, to Rummel's third felony conviction to incarcerate him for life, because he had been "informed of the consequences of lawlessness and given an opportunity to reform, all to no avail. [The recidivist statute] thus is nothing more than a societal decision that when such a person commits yet another felony, he should be subjected to the admittedly serious penalty of incarceration for life ...."); People v. Murillo , 39 Cal.App.4th 1298, 46 Cal. Rptr. 2d 403, 407–08 (1995) (although the defendant argued that "because he suffered his prior conviction before the Three Strikes Law became effective, his prior [crime] cannot qualify as a ‘strike,’ " the court "disagree[d]" because such a "view is patently inconsistent with the purpose of the statute and the companion initiative and, in our view, constitutes an absurd result"). All Cortes Medina did to Reyes was increase the potential consequences if he decided to reoffend in the future .
The majority strives mightily to avoid any consideration of how courts have treated recidivism laws, asserting that "cases analyzing recidivist statutes are not relevant here" because there are differences between such statutes and immigration statutes, including that retroactivity is more permissible in the recidivism context. No doubt there are differences. But none of those differences have any relevance with respect to why the analysis from those cases fits perfectly here. I don't reference those cases to argue that retroactive application of Cortez Medina should be allowed here. I reference them to show that applying Cortez Medina here is not retroactive at all . The rationale that applying an enhanced penalty is not retroactive where the offender's last crime in a series was committed after the enhanced penalty was enacted is directly on point with this case, and the majority's attempt to distinguish those cases for entirely irrelevant reasons is telling.
Whether Reyes would have thought in 2011 that a conviction under section 314.1 was a CIMT is simply not relevant in this case—his first section 314.1 conviction in 2011 did not trigger his deportation under 8 U.S.C. § 1227(a)(2)(A)(ii), which requires "two or more crimes." The relevant question is whether Reyes, in the moments before he violated section 314.1 for a second time in 2014 by publicly masturbating in front of a store, could have constructively relied on Nunez to protect his imminent crime under section 314.1 from being defined as a second —and therefore removable—CIMT. Given the intervening Cortes Medina decision before his second offense, the answer is no. Cortes Medina made clear that Reyes's earlier 2011 section 314.1 offense was, in fact, a CIMT, and that if he committed and was convicted of a second section 314.1 offense, he would be removable under § 1227(a)(2)(A)(ii).
As the Betansos panel explained, we may presume reliance if the "former, favorable rule was in place at the time the petitioner pleaded guilty or was convicted." Betansos , 928 F.3d at 1144. But because the BIA's authoritative Cortes Medina decision had already superseded the Nunez decision when Reyes pled in 2014, Reyes cannot demonstrate reliance on Nunez under factor three. He "knew" (either constructively or in reality) that he already had one CIMT under Cortes Medina , and that by pleading to a second, he would become deportable. Reyes therefore could not have "relied" on Nunez when it mattered: when he pled to his second CIMT in 2014.
The majority argues that the government in this case did not rebut the presumption that Reyes relied on Nunez . There was no presumption to rebut, because Reyes's 2014 conviction—after the "former, favorable rule was in place" even under Betansos ’s faulty dicta—was what triggered his removability under § 1227(a)(2)(A)(ii). Moreover, as the government observed, Reyes provided "no evidence that he actually relied on Nunez in 2011."
B. Montgomery Ward Factor Two
The majority here accepts the Betansos panel's factor two dicta, but as discussed, our panel is not bound by that obviously wrong and unreasoned dicta and should not reflexively apply it here to conclude that the second factor weighs in favor of Reyes. But even if we were required to follow Betansos ’s dicta, the key facts of this case are different from Betansos ’s hypothetical because Reyes committed his critical second offense and pled guilty in 2014, after "it would have been reasonable to rely on Nunez between February 2010 and January 2013." Betansos , 928 F.3d at 1145. And even if the second factor did weigh in favor of Reyes, here, as in Betansos , the fact that the third (and fifth) factors weigh against Reyes would result in denying his petition. See Betansos , 928 F.3d at 1145–46 ("[A]lthough the second factor arguably favors Betansos, we have held that factors two and three are ‘intertwined.’ Because factor three weighs against Betansos in this case, we hold that overall the factors support retroactive application against Betansos.").
* * *
Today's decision demonstrates what we all know from hard experience: fixing mistakes is usually much harder than making them in the first place. Unfortunately, the majority's opinion perpetuates rather than acknowledges and addresses some of our more blatant recent immigration gaffes, and in so doing misses an opportunity to right our circuit's badly listing immigration ship—at least a little. Reyes continued to commit crimes of moral turpitude even after he was on notice that if he committed another one, he would suffer immigration consequences. I would deny his petition, and so must respectfully dissent.
Reyes's other arguments on appeal relating to his application for cancellation of removal and an alleged due process violation are similarly unavailing. We lack jurisdiction to review the merits of a discretionary decision to deny cancellation of removal. 8 U.S.C. § 1252(a)(2)(B)(i) ; Vilchez v. Holder , 682 F.3d 1195, 1198 (9th Cir. 2012). And although Reyes claims he was not afforded the opportunity to testify regarding his asylum claim, the IJ held multiple hearings to address the merits of his arguments for relief from removal, where he received "a full and fair opportunity to be represented by counsel ... and to present testimony and other evidence in support of [his asylum] application." Vargas-Hernandez v. Gonzales , 497 F.3d 919, 927 (9th Cir. 2007).