DULCE M. ZARAGOZA, Petitioner, v. MERRICK B. GARLAND, Attorney General of the United States, Respondent.
Nos. 19-3437 & 20-1591
United States Court of Appeals for the Seventh Circuit
ARGUED DECEMBER 3, 2020 — DECIDED NOVEMBER 8, 2022
SYKES, Chief Judge. Dulce Zaragoza, a native and citizen of Mexico and a lawful permanent resident of the United States, pleaded guilty to the Indiana offense of criminal neglect of a dependent after locking her six-year-old son in a
Zaragoza fought removal on several grounds, with her arguments expanding as the proceedings progressed. Before the immigration judge, she argued that the Indiana neglect offense does not qualify as a crime involving moral turpitude. The judge disagreed and entered a removal order, and Zaragoza appealed to the Board of Immigration Appeals (“BIA” or “the Board“). In the meantime, she petitioned the state court to modify her sentence. Her purpose was to bring herself within the so-called “petty offense” exception to inadmissibility, which is available to first-time offenders sentenced to six months or less.
The BIA rejected both arguments, agreeing with the immigration judge that the Indiana offense is categorically a crime involving moral turpitude, and further holding that the sentence-modification order was not effective to establish Zaragoza‘s eligibility for the petty-offense exception. For the latter conclusion, the Board relied on a recent decision of
Zaragoza sought reconsideration, this time adding two more arguments: (1) the phrase “crime involving moral turpitude” is unconstitutionally vague; and (2) the Attorney General‘s decision in Thomas is impermissibly retroactive as applied to her. The BIA disagreed on both counts. Zaragoza petitioned for review in this court, reprising the entire array of arguments she presented to the Board.
We agree with the BIA‘s resolution of all issues but one: applying Thomas in Zaragoza‘s case is an impermissibly retroactive application of a new rule. We therefore remand to the BIA for further proceedings consistent with this opinion.
I. Background
After emigrating from her native Mexico, Zaragoza settled in Indiana with her three children and in August 2011 became a lawful permanent resident. On October 9, 2013, she punished her six-year-old son by barricading him in a closet while she was at work. She left him with nothing except a cup of water, a hot-dog bun with ketchup on it, and a bowl to urinate in if needed. After instructing her older son not to release the younger boy from the closet, she left the house. The boy remained in confinement for six hours.
Zaragoza was charged in state court with neglect of a dependent in violation of
A few months later, Zaragoza traveled abroad. On July 7, 2015, she returned through Chicago and presented herself for inspection as a returning lawful permanent resident. Customs officials discovered her neglect conviction and paroled her into the United States in anticipation of removal proceedings. On August 6 DHS initiated removal proceedings based on her neglect conviction, which the agency classified as a “crime involving moral turpitude,” making her inadmissible under
Zaragoza moved to terminate the proceedings, arguing that neglect of a dependent is not a crime involving moral turpitude. An immigration judge disagreed, concluding that Indiana‘s neglect offense is a crime involving moral turpitude under the categorical approach as explained in the
Zaragoza sought review in the BIA. While her appeal was pending, she petitioned the state court to modify her sentence to 179 days in prison. That was an odd request on the surface, not least because Zaragoza had long since completed her sentence. But her purpose was apparent in light of the removal peril she faced. As a first-time offender, if her sentence was not “in excess of 6 months,” she would qualify for the petty-offense exception to inadmissibility under
Back before the BIA, Zaragoza reiterated her position that the Indiana neglect offense is not a crime involving moral turpitude, but she now also claimed that the petty-offense exception lifted the inadmissibility bar. The BIA rejected both arguments. In a decision issued on November 14, 2019, the Board first agreed with the immigration judge‘s ruling that the neglect offense is categorically a crime of moral turpitude. Turning to the petty-offense exception, the Board explained that under the Attorney General‘s recent decision in Thomas, issued just a few weeks earlier, the state court‘s sentence-modification order had no effect for immigration purposes because it was not based on a procedural or substantive defect in the underlying criminal proceeding. Zaragoza‘s eligibility thus turned on her original sentence, not her sentence as modified. Because she was originally sentenced to one year in prison, she did not
In the meantime, she asked the BIA to reconsider its decision. Her motion added two new arguments. She now claimed that the statutory phrase “crime involving moral turpitude” is unconstitutionally vague. She also argued that applying the Attorney General‘s decision in Thomas to her amounted to an impermissibly retroactive application of a new rule.
The Board denied the reconsideration motion, standing by its decision that the Indiana neglect offense is a crime involving moral turpitude and rejecting the new vagueness challenge to the statute. The Board also rejected Zaragoza‘s claim that applying Thomas in her case is an impermissibly retroactive application of a new rule. Zaragoza petitioned for review of the BIA‘s second order, and we consolidated the two petitions. See
II. Discussion
Zaragoza reprises the full assortment of legal challenges that she raised before the agency. Some of the issues are complex, and two have attracted support from amici curiae. For ease of presentation, we separate them into two groups.
In the first group are arguments pertaining to the meaning and application of
Because these are legal issues, our standard of review is de novo, Meraz-Saucedo v. Rosen, 986 F.3d 676, 684 (7th Cir. 2021), with one important qualifier. We defer to the agency‘s reasonable interpretation of the immigration laws in its precedential decisions and also its “[n]on-precedential decisions that rely on applicable Board precedent.” Cano-Oyarzabal v. Holder, 774 F.3d 914, 916 (7th Cir. 2014).
A. Crime Involving Moral Turpitude
1. Unconstitutional Vagueness
Zaragoza begins with the argument that the statutory phrase “crime involving moral turpitude” is unconstitutionally vague.2 She primarily relies on a trio of recent Supreme Court decisions addressing vagueness challenges to the definitions of “crime of violence” and “violent felony” in statutes that use these terms to denote certain convictions that carry sentencing and immigration consequences. See United States v. Davis, 139 S. Ct. 2319 (2019) (residual clause defining “crime of violence,”
Zaragoza responds that Jordan is no longer authoritative because it did not consider the categorical way in which the BIA and the courts now classify convictions for immigration and sentencing purposes. But she challenges the language of the statute, not the decision method courts use to classify convictions for these purposes. And Jordan squarely holds that the statutory phrase “crime involving moral turpitude” as used in immigration law is not unconstitutionally vague.
Zaragoza also argues that Jordan may be disregarded because it is out of sync with the Court‘s intervening decisions in Johnson, Dimaya, and Davis. That argument cannot succeed in the court of appeals. Jordan is binding on us until the Supreme Court says otherwise. State Oil Co. v. Kahn, 522 U.S. 3, 20 (1997) (“[I]t is this Court‘s prerogative alone to overrule one of its precedents.“). The Court has repeatedly reaffirmed this point: “If a precedent of this Court has direct application in a case, yet appears to rest on reasons rejected in some other line of decisions, the Court of Appeals should follow the case which directly controls, leaving to this Court the prerogative of overruling its own decisions.” Agostini v. Felton, 521 U.S. 203, 237–38 (1997) (quotation marks and alteration omitted).
Accordingly, we and other courts have already rejected post-Johnson vagueness challenges to the phrase “crime involving moral turpitude.” Dominguez-Pulido v. Lynch, 821 F.3d 837, 842–43 (7th Cir. 2016); see also Islas-Veloz v. Whitaker, 914 F.3d 1249, 1250 (9th Cir. 2019) (“The Court‘s more recent decisions in Johnson and Dimaya did not reopen inquiry into the constitutionality of the phrase.“); Moreno v. Att‘y Gen., 887 F.3d 160, 166 (3d Cir. 2018); Boggala v. Sessions, 866 F.3d 563, 570 (4th Cir. 2017). We do so again here.
2. Neglect of a Dependent Is a Crime Involving Moral Turpitude
Zaragoza next challenges the Board‘s conclusion that the Indiana neglect offense qualifies as a crime involving moral turpitude. Like other statutory contexts in which the agency must classify convictions for immigration purposes, the categorical approach applies to this inquiry. See Silva-Trevino, 26 I. & N. Dec. at 830; see also Garcia-Martinez v. Barr, 921 F.3d 674, 679 (7th Cir. 2019) (“Both Chevron deference and the soundness of the Board‘s reasoning in Silva-Trevino ... thus lead us to adopt that framework for characterizing crimes of moral turpitude in immigration cases.“).
Applying that framework, we examine whether the statutory definition of the offense fits within the “generic” definition of a crime involving moral turpitude. Silva-Trevino, 26 I. & N. Dec. at 831. The comparison focuses on “the minimum conduct that has a realistic probability of being prosecuted under the statute of conviction, rather than on the facts underlying the [noncitizen‘s] particular violation of that statute.” Id. (citing Moncrieffe v. Holder, 569 U.S. 184, 190–91 (2013)); see also Garcia-Martinez, 921 F.3d at 679. If Zaragoza can show that the state courts have applied the neglect statute to conduct that does not come within the generic definition, then the offense is not categorically a crime involving moral turpitude. See Moncrieffe, 569 U.S. at 206
We begin with the generic definition of a “crime involving moral turpitude,” acknowledging (as we must) that the “moral turpitude label” is “an odd match for the categorical approach.” Garcia-Martinez, 921 F.3d at 679. Though the phrase is not defined in statute, we give Chevron deference to decisions of the BIA reasonably interpreting that term. Cano-Oyarzabal, 774 F.3d at 916. In Silva-Trevino the Board held that a crime involving moral turpitude has “two essential elements: reprehensible conduct and a culpable mental state.” 26 I. & N. Dec. at 834. For conduct to be “reprehensible,” it must be “inherently base, vile, or depraved, and contrary to the accepted rules of morality and the duties owed between persons or to society in general.” Id. at 833 (quotation marks omitted). A culpable mental state means “some degree of scienter, either specific intent, deliberateness, willfulness, or recklessness.” Matter of Ortega-Lopez, 27 I. & N. Dec. 382, 385 (B.I.A. 2018) (quoting Matter of Louissaint, 24 I. & N. Dec. 754, 757 (B.I.A. 2009)).
We compare that definition with the Indiana neglect offense as defined by statute and as applied by the Indiana courts. Garcia-Martinez, 921 F.3d at 680. The relevant part of Indiana‘s neglect statute provides: “A person having the care of a dependent ... who knowingly or intentionally ... abandons or cruelly confines the dependent ... commits neglect of a dependent.”
So defined, the Indiana neglect offense categorically matches both elements of the generic definition of a “crime involving moral turpitude.” First, and more straightforwardly, the offense requires a sufficiently culpable mental state. The Indiana statute requires intentional or knowing conduct, and the generic crime involving moral turpitude can be established by “specific intent, deliberateness, willfulness, or recklessness.” Ortega-Lopez, 27 I. & N. Dec. at 385.
Second, the neglect offense requires “reprehensible conduct.” Abandoning or cruelly confining a dependent, as the Indiana courts interpret this offense, qualifies as “inherently base, vile, or depraved, and contrary to the accepted rules of morality and the duties owed between persons or to society in general.” Silva-Trevino, 26 I. & N. Dec. at 833 (quotation marks omitted).
Zaragoza emphasizes that the neglect offense does not require proof of conduct that endangers the dependent‘s life or health, see Hartbarger, 555 N.E.2d at 487, which she believes makes the Indiana offense broader than the generic definition of a crime involving moral turpitude. We disagree. Under the Hartbarger standard, the offender‘s conduct must be “likely to result in a harm such as disfigurement, mental distress, extreme pain or hurt, or gross degradation.” Id. A person who exposes a dependent child to that degree of risk by abandoning or cruelly confining him has engaged in reprehensible conduct.
Additionally, as the BIA correctly recognized, Zaragoza has failed to demonstrate that there is a “realistic probability” that the neglect statute will be applied to actions causing only minor mental distress, rather than conduct that is “inherently base, vile, or depraved.” Silva-Trevino, 26 I. & N. Dec. at 831, 833 (quotation marks omitted). She points to her own conviction, which she claims did not involve conduct that is inherently base, vile, or depraved. We disagree. Zaragoza barricaded her six-year-old son in a closet using large items of furniture and leaving him nothing but water, a hot-dog bun with ketchup, and a bowl to urinate in, and she then left her house for six hours. This is undoubtedly “contrary to the accepted rules of morality and the duties owed between persons.” Id. at 833 (quotation marks omitted).3
The better reading of Silva-Trevino is that the realistic-probability principle considers the minimum conduct that realistically could be successfully prosecuted under the statute in question. Silva-Trevino relied on the Supreme Court‘s decision in Moncrieffe, which explained that the categorical approach examines “the minimum conduct criminalized by the state statute” and requires a showing of “a realistic probability ... that the State would apply its statute” to the specified conduct. 569 U.S. at 191 (emphases added) (quotation marks omitted); see also Matter of Chairez-Castrejon, 26 I. & N. Dec. 349, 356 (B.I.A. 2014) (applying the Moncrieffe rule only to successful prosecutions). In other words, the Supreme Court‘s realistic-probability test, which Silva-Trevino incorporates, considers how state law is applied by state courts, not prosecutors. Gonzales v. Duenas-Alvarez, 549 U.S. 183, 193 (2007) (explaining that to show a realistic probability, an offender “must at least point to his own case or other cases in which the state courts in fact did apply the statute in the special (nongeneric) manner for which he argues“).
This understanding of Moncrieffe and Silva-Trevino comports with well-established background norms. Prosecutors are not expositors of law—courts are, which is why we defer to state courts in understanding the content of state law. Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78 (1938). Because the state appellate court held that the specific conduct in Hartbarger and Scruggs did not violate the neglect statute, those cases do not help Zaragoza here.
Zaragoza also likens the Indiana neglect statute to other state statutes that have been held not to constitute crimes involving moral turpitude. Her comparators, however, are inapt. For example, she points to a Fifth Circuit case holding that a Texas child-abandonment statute is not a crime involving moral turpitude. Rodriguez-Castro v. Gonzales, 427 F.3d 316, 324 (5th Cir. 2005). But the Texas statute penalized mere negligence, which is not a sufficient mental state of culpability to qualify as a crime involving moral turpitude. Id. at 322–23. As we‘ve explained, the Indiana neglect statute requires intentional or knowing wrongdoing, which squarely falls within the generic definition of a crime involving moral turpitude.
In another of Zaragoza‘s examples, the BIA held that simple battery under California law is not a crime involving moral turpitude because it requires no more than an intentional “touching” of another without consent. In re Sanudo, 23 I. & N. Dec. 968, 972 (B.I.A. 2006). Indiana‘s neglect statute, by contrast, requires a likelihood of “disfigurement,
In sum, as interpreted and applied by the state courts, the Indiana neglect statute requires proof that the defendant intentionally or knowingly abandoned or cruelly confined a dependent in such a way that will likely result in “disfigurement, mental distress, extreme pain or hurt, or gross degradation.” Id. We agree with the BIA that this offense is categorically a crime involving moral turpitude. Accord Hernandez-Perez v. Holder, 569 F.3d 345, 348 (8th Cir. 2009) (concluding that a similar child-endangerment statute under Iowa law is a crime involving moral turpitude).
B. Petty-Offense Exception
Although the Indiana neglect conviction qualifies as a crime involving moral turpitude, the inadmissibility bar is lifted for first-time offenders like Zaragoza if the crime in question was punishable by one year or less and the sentence did not exceed six months. More specifically, Zaragoza
It‘s undisputed that Zaragoza satisfies the first requirement. The neglect offense normally carries a maximum penalty of up to three years in prison, but she pleaded guilty pursuant to a statute that permitted the court to enter judgment for a Class A misdemeanor, which carries a maximum penalty of one year of imprisonment.
The dispute here centers on the second requirement. The exception applies only if the offender was sentenced to a term of six months or less. Based on Zaragoza‘s original sentence, she is clearly ineligible. She was sentenced to one year of imprisonment suspended to time served plus 30 days. The suspension has no effect on the analysis. As defined in the Immigration and Nationality Act (“INA” or “the Act“), “[a]ny reference to a term of imprisonment or a sentence ... is deemed to include the period of incarceration or confinement ordered by a court of law regardless of any suspension of the imposition or execution of that imprisonment or sentence in whole or in part.”
But what about the sentence-modification order? Recall that in February 2019 while her appeal was pending before the BIA, Zaragoza sought and obtained an order from a state judge reducing her one-year sentence to 179 days. At the time BIA precedent recognized state-court sentence-modification orders as effective for immigration purposes. See Matter of Cota-Vargas, 23 I. & N. Dec. 849, 852 (B.I.A. 2005) (holding that an immigration court must give full faith and credit to a state-court decision modifying a sentence). Accordingly, Zaragoza asked the BIA to evaluate her eligibility for the petty-offense exception based on her sentence as modified.
The BIA did not rule on her appeal until nine months after she had obtained the sentence-modification order. By then the Attorney General had issued his decision in Thomas, overruling Cota-Vargas and holding that state-court sentence-modification orders are effective for immigration purposes only if based on a procedural or substantive defect in the underlying criminal proceeding. Thomas, 27 I. & N. Dec. at 674. Zaragoza‘s sentence modification was not based on such a defect. Applying Thomas, the Board declined to give it effect for purposes of evaluating her eligibility for the petty-offense exception. Because Zaragoza‘s original sentence exceeded six months, the Board found her ineligible for the exception.
1. Thomas Is Entitled to Deference
The pre-Thomas legal landscape was a patchwork of inconsistent rules regarding the immigration consequences of state-court orders altering a criminal sentence. If a state court vacated a conviction for reasons other than a defect in the criminal proceeding, then the immigration consequences remained fixed to the original conviction and sentence. Matter of Pickering, 23 I. & N. Dec. 621, 624 (B.I.A. 2003). But if a state court modified a sentence, then the immigration consequences were fixed to the new order. Cota-Vargas, 23 I. & N. Dec. at 852; In re Song, 23 I. & N. Dec. 173, 174 (B.I.A. 2001). If a state court clarified a sentence, then the immigration judge was to consider several characteristics of the state-court order before deciding whether immigration consequences should attach to the original sentence or the clarification order. Matter of Estrada, 26 I. & N. Dec. 749, 755–56 (B.I.A. 2016).
These inconsistencies, coupled with the perception that state courts were using sentence modifications to circumvent federal immigration law, led the Attorney General to step in. In May 2019 he directed the BIA to refer two pending cases to him for a clarifying opinion. Thomas, 27 I. & N. Dec. at 674.
While their removal proceedings were pending, Thomas and Thompson returned to state court and obtained orders reducing their sentences to slightly under 12 months. Id. at 678–79. The modification orders were not based on any defect in the underlying criminal proceedings, but Thomas and Thompson argued in their removal proceedings that because their sentences had been reduced to less than one year, they were no longer removable as aggravated felons. After directing the BIA to refer the cases, the Attorney General invited the parties and any interested amici to submit briefs regarding the effect of state-court sentence-modification orders for immigration purposes. Id. at 674.
On October 25, 2019, the Attorney General issued his decision overruling Cota-Vargas, Song, and Estrada and holding that state-court sentence-modification orders “have no effect for immigration purposes if based on reasons unrelated to the merits of the underlying criminal proceeding, such as rehabilitation or the avoidance of immigration consequences.” Id.
(A) The term “conviction” means, with respect to an alien, a formal judgment of guilt of the alien entered by a court or, if adjudication has been withheld, where—
(i) a judge or jury has found the alien guilty or the alien has entered a plea of guilty or nolo contendere or has admitted sufficient facts to warrant a finding of guilt, and
(ii) the judge has ordered some form of punishment, penalty, or restraint on the alien‘s liberty to be imposed.
(B) Any reference to a term of imprisonment or a sentence with respect to an offense is deemed to include the period of incarceration or confinement ordered by a court of law regardless of any suspension of the imposition or execution of that imprisonment or sentence in whole or in part.
Anchoring his analysis in the text of these definitions, the Attorney General began with an uncontroversial observation: “An alien plainly has been convicted under the INA when a court has entered ‘a formal judgment of guilt,’ and he has received a sentence when the court orders a ‘period of incarceration or confinement,’ no matter whether the sentence [has been] executed.” Thomas, 27 I. & N. Dec. at 680–81. In other words, by virtue of the “regardless” clause in the definition, the terms “conviction” and “term of imprison-
The Attorney General then looked to the statutory history to confirm this implication, noting that
Based on this review of the statutory text and history, the Attorney General concluded that “the phrase ‘term of imprisonment or a sentence’ in paragraph (B) is best read to concern an alien‘s original criminal sentence, without regard to post-sentencing alterations that, like a suspension, merely alleviate the impact of that sentence.” Id. But he carved out an exception based on the reasoning in Pickering: “If the original sentence was altered because of a legal defect, then the sentence was not legally effective, and there is no valid
The Attorney General thus extended the Pickering vacatur framework to all sentence alterations, including vacaturs, modifications, and clarifications. Id. at 683–85. He accordingly held that “state-court orders that modify, clarify, or otherwise alter a criminal alien‘s sentence ... will be given effect for immigration purposes only if based on a procedural or substantive defect in the underlying criminal proceeding.” Id. at 690 (emphasis added).
Decisions of the Attorney General interpreting the federal immigration statutes are entitled to Chevron deference. I.N.S. v. Aguirre-Aguirre, 526 U.S. 415, 424 (1999); see also
Neither the petty-offense exception nor the definitions in
We begin with the language of the petty-offense exception. Zaragoza qualifies for relief from inadmissibility if she “was not sentenced to a term of imprisonment in excess of 6 months (regardless of the extent to which the sentence was ultimately executed).”
Next, as Thomas explains, the definition of the phrase “term of imprisonment” in
Were there any doubt about this analysis, circuit precedent confirms our conclusion. After the BIA held in Pickering that immigration consequences remain fixed to a conviction even after it is vacated (with an important exception for vacaturs based on a legal defect), we concluded that the agency‘s decision was entitled to Chevron deference. Ali v. Ashcroft, 395 F.3d 722, 728–29 (7th Cir. 2005). Because the definition of “conviction” in
The same result follows here. As we‘ve explained, in Thomas the Attorney General extended the Pickering vacatur rule to all sentence alterations—including, as relevant here, sentence modifications. True, resolving the immigration effect of vacaturs and sentence modifications turns on separate statutory definitions—for the former, it is the definition of “conviction” in
Zaragoza responds that deferring to Thomas is inconsistent with our obligation to give full faith and credit to a state court‘s modification of a sentence. See
In sum, the Attorney General‘s decision in Thomas is entitled to deference as a permissible construction of the terms “conviction” and “term of imprisonment or a sentence” as defined in
Zaragoza argued in her reconsideration motion that applying Thomas to her is an impermissibly retroactive application of a new rule. The Board rejected that argument. We review retroactivity questions independently, owing no deference to the agency‘s ruling. Velásquez-García v. Holder, 760 F.3d 571, 578–79 (7th Cir. 2014)
The law generally disfavors the retroactive application of new legal rules. See id. at 579. “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.” Landgraf v. USI Film Prods., 511 U.S. 244, 265 (1994). But deciding when a new rule operates retroactively “is not always a simple or mechanical task.” Id. at 268. A new rule is not necessarily “retroactive” in effect simply because it is applied in a case arising from conduct predating its adoption. Id. at 269. Rather, a determination that a particular rule operates retroactively “comes at the end of a process of judgment concerning the nature and extent of the change in the law and the degree of connection between the operation of the new rule and a relevant past event.” Id. at 270. The inquiry, in other words, “demands a commonsense, functional judgment” and is “informed and guided by ‘familiar considerations of fair notice, reasonable reliance, and settled expectations.‘” Martin v. Hadix, 527 U.S. 343, 357–58 (1999) (quoting Landgraf, 511 U.S. at 270).
As a general matter, “[a] rule is considered to be retroactive” in effect “when it ‘attaches new legal consequences to events completed before its enactment.‘” Velásquez-García, 760 F.3d at 579 (quoting Landgraf, 511 U.S. at 270). A rule
A vexing problem in retroactivity cases is to identify the appropriate reference point for analysis—the event or moment in time by which to judge whether an intervening legal development imposes new consequences or disabilities. See Vartelas v. Holder, 566 U.S. 257, 269–70 (2012); id. at 277 (Scalia, J., dissenting). Zaragoza and the government offer competing possibilities.7 The government contends that the correct point of reference is March 31, 2014, when Zaragoza was convicted and sentenced for the neglect offense. Because she had no fixed entitlement to the petty-offense exception at that point in time, the government argues that Thomas does not have retroactive effect as applied to her. Zaragoza counters with two possible reference points—either March 31,
We conclude that the proper reference point for the retroactivity inquiry is the February 2019 sentence-modification order. As we‘ve noted, retroactivity analysis is concerned with the degree of connection between the new rule and the relevant past event. More specifically, the analysis asks whether a new rule “impairs vested rights acquired under existing laws.” St. Cyr, 533 U.S. at 321 (quotation marks omitted). When the state court entered the sentence-modification order, Zaragoza acquired a legal entitlement to the petty-offense exception under existing immigration law—namely, the BIA‘s decision in Cota-Vargas. As of that event, she had a complete defense to removal. The Attorney General‘s decision in Thomas overruled Cota-Vargas, eliminating the defense. So the “transaction[] or consideration[]” to which Thomas attached legal consequences was the sentence-modification order. Id. The rule of Thomas therefore has retroactive effect as applied in Zaragoza‘s removal proceedings.
The government resists this conclusion, relying on the Supreme Court‘s decisions in Vartelas and St. Cyr, both of which used the immigrant‘s underlying conviction to evaluate the retroactive effect of later legal developments. But neither of those cases involved sentence modifications, so the Court had no occasion to consider the problem presented here.
Instead, this case is closer to our decision in Jeudy v. Holder, 768 F.3d 595 (7th Cir. 2014). There, an immigrant
We disagreed, instead holding that the stop-time rule was impermissibly retroactive as applied in Jeudy‘s case because he “was actually eligible for discretionary relief before [the new stop-time rule] took effect.” Id. at 603. In other words, in Jeudy there wasn‘t a retroactivity problem simply because his conviction predated the new law; rather, a postconviction event made Jeudy unquestionably eligible for relief from removal, only to then be foreclosed by a later legal development.
The same is true here. A postconviction event—Zaragoza‘s sentence modification—gave her a right to relief from removal, only to be taken away by Thomas. Because the state court‘s sentence-modification order predated Thomas, the decision is retroactive as applied to her.
Our next question is whether retroactive application is impermissible in Zaragoza‘s case. As we explained in Velásquez-García, when an agency interprets a statute “as an incident of its adjudicatory function,” it may permissibly apply the new interpretation in the case in which it is announced. 760 F.3d at 581 (quotation marks omitted). But “a
(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 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 NLRB v. Wayne Transp., 776 F.2d 745, 751 n.8 (7th Cir. 1985)). “Like most such unweighted multi-factor lists, this one serves best as a heuristic; no one consideration trumps the others.” Id.
The first factor—whether the case is one of “first impression“—requires some clarification. As we‘ve noted, when an agency announces a new rule in the exercise of its adjudicative function, it may apply the rule in the proceeding before it; that is the case of “first impression.” Id. This case, however, is one of “second impression“: the BIA applied Thomas retroactively to Zaragoza—a stranger to the case in which the new rule was announced—even though she had already acquired a right to relief from removal by operation of the prior rule of Cota-Vargas. The government concedes, and we agree, that this factor tips against retroactive application.
The third factor is the extent of Zaragoza‘s reliance interests. Though no one factor in the Velásquez-García list is decisive, this one has a significant role to play. See Vartelas, 566 U.S. at 274 (“[T]he likelihood of reliance on prior law strengthens the case for reading a newly enacted law prospectively.“); Landgraf, 511 U.S. at 270 (retroactivity doctrine is guided by “familiar considerations of fair notice, reasonable reliance, and settled expectations“). “Importantly, the critical question is not whether a party actually relied on the old law, but whether such reliance would have been reasonable.” Velásquez-García, 760 F.3d at 582; see also Jeudy, 768 F.3d at 604.
Reliance on the old law was objectively reasonable in the circumstances here. Under the rule of Cota-Vargas, Zaragoza had clear right to relief under the petty-offense exception when the state court modified her sentence. That is, Cota-Vargas gave her a complete defense to removal once she obtained the order from the state court reducing her sentence to six months or less. When she did so, she reasonably relied on then-existing law, which lifted the inadmissibility bar and eliminated the basis for her removal.
The reliance interests here are arguably stronger than those at issue in Velásquez-García. There, the immigrant took preliminary steps toward acquiring permanent-resident status but did not file an application within one year of the date when his visa number became available, as required by
If it was reasonable for Velásquez-García to rely on non-precedential BIA decisions generously interpreting a statutory deadline to apply for immigration benefits, then it was reasonable for Zaragoza to rely on the BIA‘s precedential decision in Cota-Vargas, which gave her a complete defense to removability based on her sentence modification.
The fourth factor is the degree of burden that the retroactive rule imposes. “Courts have long recognized the obvious hardship imposed by removal.” Id. at 584. This factor clearly favors Zaragoza. The final factor is the statutory interest in
In sum, all but one of the Velásquez-García factors weigh against retroactive application of Thomas in Zaragoza‘s case. Accordingly, we conclude that applying the new rule to her would work a manifest injustice. We GRANT the petitions for review and REMAND to the BIA for further proceedings consistent with this opinion.