Argument Preview: Nijhawan v. Holder

Brian Goldman previews Monday’s first argument.  Additional information on the case is available on SCOTUSwiki, here.

Much like Flores-Figueroa v. United States (No. 08-108), argued earlier this Term, Nijhawan v. Holder (08-495) presents the Court with conflicting interpretations of a statute that sits at the intersection of federal criminal law and immigration law and policy. A provision of the Immigration and Nationality Act, 8 U.S.C. § 1227(a)(2)(A)(iii), makes “[a]ny alien who is convicted of an aggravated felony . . . deportable.” As relevant here, Section 1101(a)(43) defines an aggravated felony to include, among others, “an offense that . . . involves fraud or deceit in which the loss to the victim or victims exceeds $10,000” (Subparagraph (M)(i)), and a conspiracy to commit such an offense (Subparagraph (U)). On Monday, April 27, in Nijhawan v. Holder, the Court will consider whether, for an alien to be deportable on account of an aggravated felony, the amount of loss must actually have been found beyond a reasonable doubt by a jury, or whether an alien’s stipulation for sentencing purposes that the loss exceeded $100 million is sufficient for the conviction to qualify as an aggravated felony.

In 2002, petitioner Manoj Nijhawan, an Indian citizen and lawful permanent resident in the United States, was arrested for his involvement in a conspiracy to commit fraud against major banks by obtaining loans for a metal trading company under false pretenses. As the assistant to the CFO, Nijhawan was one of fifteen defendants indicted and one of four to be convicted on conspiracy, fraud, and money laundering charges. The amount of money lost or laundered was not an element of any of the offenses, and the district court in New York explicitly instructed the jury that no specific finding of loss was needed to convict. To resolve disputed issues, including the potential effect of the Court’s decision in Booker v. United States, which was then pending before the Court, Nijhawan and federal prosecutors agreed to set his Sentencing Guidelines level based upon a total loss in excess of $100 million for sentencing purposes, while allowing him to seek a large downward departure to put his sentencing range at forty-one to fifty-one months. Aware of the possible immigration consequences of such an agreement, everyone involved – the court, the prosecutor, and Nijhawan – agreed that the stipulation was only to be used for sentencing purposes.

While Nijhawan was incarcerated, the Government began removal proceedings against him. A removal order was entered, and the Board of Immigration Appeals (BIA) affirmed, finding that even if the amount of loss was not an element of Nijhawan’s offenses, it could look beyond the conviction itself to determine the amount of loss. The BIA rejected Nijhawan’s argument that it should adopt a “categorical” approach, which would require the amount of loss to be either an element of the convicted offenses or explicitly found by a jury. On appeal, the Third Circuit affirmed, in a divided opinion that acknowledged contrary authority from other circuits. In that court’s view, the amount of loss needs only to be sufficiently “tethered” to the conviction. Because most fraud statutes do not contain loss as an element of the offense, the court reasoned, Nijhawan’s construction “would render [Subparagraph (M)(i)] largely inoperative, for rarely will a defendant be convicted of a fraud offense with loss as an element found by the jury or explicitly admitted to in a guilty plea.” Additionally, the court expressed concern that adopting the “categorical” approach would result in fewer deportations.

Nijhawan filed a petition for certiorari, in which he advanced three main arguments. First, he highlighted a 3-3 split among the circuit courts on the question whether removability under the INA could be established by facts not found beyond a reasonable doubt during criminal proceedings. The Second, Ninth, and Eleventh Circuits had adopted a “modified categorical” approach, which rejected reliance upon unproven facts such as those in sentencing determinations, while the First and Fifth Circuits had previously come down on the side of the Third.

Second, Nijhawan argued that split was untenable because it allowed the Government to engage in forum shopping, given its power to transfer aliens in removal proceedings to detention centers in other circuits with more favorable law. Moreover, the split violated Article I, § 8 of the Constitution, which requires a uniform rule of naturalization. Because anyone convicted of an aggravated felony would not be able to naturalize, aliens with convictions like Nijhawan’s would not be able to naturalize in three circuits but would be eligible in three others.

Finally, Nijhawan contended that the Third Circuit was wrong on the merits, advancing arguments about Subparagraph (M)(i)’s plain text, administrability, and the rule of lenity that he would later repeat in his brief on the merits (see below).

In its brief in opposition, the Government acknowledged that the Third Circuit’s “analysis diverges from decisions of other courts of appeals,” but it nonetheless concluded that “[f]urther review is not warranted.” The Government began by responding to Nijhawan’s arguments on the merits and defending the decision below, again in a preview of its later merits briefing.

Next, the Government addressed the circuit split. It cast doubt on the continuing validity of the Second, Ninth, and Eleventh Circuit decisions in light of an intervening decision by the BIA, which had clarified that when a “removal charge requires proof of some fact that is not an element of the predicate offense,” the immigration courts may consider “any evidence[] otherwise admissible in removal proceedings”—regardless whether such facts were found at trial. The Second and Eleventh Circuit decisions predated the BIA ruling, and although the Ninth Circuit issued its decision after the BIA clarification, the court did not consider the ruling in its opinion. Thus, the Government recommended, certiorari should be denied to allow these circuits to revisit the issue in light of the BIA ruling and, in particular, accord the agency’s determination the deference required by N.C.T.A. v. Brand X Internet Services.

In his reply brief, Nijhawan dismissed the possibility that the circuit split would resolve itself as “wholly speculative.” He rejected the Government’s argument that any Brand X deference would apply anyway, given the plain meaning of the statutory text, the rule of lenity, and the fact that Subparagraph (M)(i) is related to federal criminal statutes, agency interpretations of which are not accorded Chevron deference.

The Court granted certiorari on January 16, 2009, but it limited its grant to the question formulated by the Government in its brief in opposition. Rather than address Nijhawan’s broader questions of whether Subparagraph (M)(i) requires the amount of loss to be an element of the conviction, and whether the rule of lenity should apply, the Court agreed to consider only whether Nijhawan’s conviction qualified as an aggravated felony “where petitioner stipulated for sentencing purposes that the victim loss associated with his fraud offense exceeded $100 million . . . .”

In his brief on the merits, Nijhawan focuses heavily on the plain language of Subparagraph (M)(i) to argue that Congress codified a categorical approach. After reviewing the history of the categorical approach and legislation authorizing deportation, Nijhawan begins by emphasizing the words “convicted of” in the statute. Removability is premised on what the alien has actually been convicted of, he suggests, not merely what he has done. Nijhawan points to other immigration provisions that are triggered by an alien’s conduct, rather than a formal conviction, to argue that deportability permitted only by a conviction is to be understood on its plain terms.

Next, Nijhawan analyzes the grammatical structure of the “aggravated felony” definition in Subparagraph (M)(i). He suggests that the phrase “an offense that . . . involves fraud or deceit in which the loss to the victim or victims exceeds $10,000” contains a “restrictive clause,” beginning with “that,” which defines an offense to require both (1) fraud or deceit and (2) a minimum loss. Without elaboration, he explains that excluding the “in which” phrase from the restrictive clause would render the word “that” a nullity. Looking to the structure of other types of aggravated felonies defined in Subparagraph (M), Nijhawan then observes that when Congress intended to use a “qualifier” to describe a circumstance outside of the conviction, it uses the language “for which.” In his view, “for which”/“in which” distinction is apparent on the face of the text; by contrast, the Third Circuit’s alternative concept of “tethering” is invented out of whole cloth.

Nijhawan next argues that post-conviction sentencing determinations or restitution orders are inappropriate substitutes for jury findings of the amount of loss, because the burden of proof in such proceedings (preponderance of the evidence) is lower than in the criminal trial (beyond a reasonable doubt) or in immigration court (clear and convincing evidence). Nijhawan argues that his stipulation at sentencing did not alter the jury’s verdict; moreover, he made it after agreeing with prosecutors and the district court that the amount of loss issue would be reserved for immigration proceedings.

Nijhawan moves on to describe the categorical approach, as applied to removal cases generally and as applied here. He begins with the approach’s history, and then describes its two forms: the “formal categorical” approach, in which the elements giving rise to deportability must be found among the statutory elements of the offense itself, and the “modified categorical” approach, which permits a carefully circumscribed inquiry into whether what was found in the adjudication of guilt would satisfy the grounds of removability. He notes that either approach would mandate reversal: None of the statutes under which he was convicted have an element of the offense the amount of loss, so the formal categorical approach could not be satisfied. And the jury did not make any finding of the amount of loss—indeed, Nijhawan’s express request that it be instructed to do so was denied—so the modified categorical approach could not be satisfied either. To show that the formal categorical approach would not set an impossibly high bar to deportation, Nijhawan provides an appendix of federal and state statutes that do require the amount loss as an element.

With regard to the practical consequences of his position, Nijhawan briefly addresses the Third Circuit’s view that it would lead to fewer deportations, dismissing that as “a policy matter for Congress to address.” He also notes that while a categorical approach would establish a clear, bright-line test that would guide government prosecutors’ actions (e.g., by prompting them to secure defined loss amounts in plea agreements or jury verdicts), the Third Circuit’s “tethering” test would shunt critical fact-finding into ill-equipped immigration courts.

Finally, Nijhawan argues that to the extent Subparagraph (M)(i) is ambiguous, the rule of lenity or narrow construction should apply. The BIA’s precedential decision is not due any deference because “aggravated felony” issues implicate federal criminal law and it thus does not resolve the ambiguity. Moreover, he suggests that the BIA decision rests on flawed reasoning, does not purport to be retroactive, and would lead to unconstitutional non-uniformity of naturalization standards by allowing Immigration Judges to reach different results in similar cases.

In its brief on the merits, the Government begins with the plain text of Subparagraph (M)(i) and concludes that the amount of loss need not be an element of a qualifying offense. First, it argues that the amount-of-loss phrase beginning with “in which” is itself a restrictive clause, independent of the phrase containing the requisite elements of a qualifying “offense.” Second, it observes that other types of aggravated felonies defined in § 1101(a)(43) specify that certain factors be found “as an element” of the offense, but such specification is absent in Subparagraph (M)(i). Relying in part on this Term’s decision in United States v. Hayes, it suggests that this omission is meaningful. Third, it dismisses Nijhawan’s argument that the deportation provision’s use of the phrase “convicted of” offers any assistance, because it simply modifies “an aggravated felony,” the definition of which is in dispute. Fourth, it notes that Subparagraph (M)(ii), which immediately follows, contains an almost identical loss criterion but just a single enumerated offense—federal tax evasion—which does not have as an element any specific amount of loss. Because a formal categorical approach would render the definition a nullity, the Government argues it must not be the approach Congress intended in either Subparagraph (M)(ii) or Subparagraph (M)(i), which must be construed in pari materia. The Government rejects the modified categorical approach as well, and concludes that the amount of loss is a mere qualifying condition, like others found in § 1101(a)(43).

The Government next criticizes Nijhawan’s construction of Subparagraph (M)(i) as “frustrat[ing] Congress’s intent to remove criminal aliens” by precluding its application to the most common federal criminal fraud offenses and most state-law fraud offenses, except the “scattered,” “odd patchwork of federal and state fraud offenses” Nijhawan collected in his brief. Consequently, the Government concludes, the Third Circuit was correct to hold that Nijhawan’s rule would render Subparagraph (M)(i) “largely inoperative.” The government also lambastes Nijhawan’s suggestion that prosecutors simply ensure they secure jury findings of (or plea agreements with) loss amounts, finding it to be at odds with Congress’s intent to prevent criminal proceedings from being manipulated to affect immigration consequences, particularly where federal government pressures on state prosecutors would create commandeering problems.

Turning to Nijhawan’s affirmative case for the categorical approach, the Government notes that the INA simply does not require it, and the Court’s immigration case law has remained silent on the question. Moreover, the Government distinguishes the Court’s sentencing enhancement precedents that established the categorical approach, finding them to be rooted in Sixth Amendment concerns about the role of jurors as factfinders—concerns not present in civil immigration proceedings. To the extent that those prior decisions rested on concerns that sentencing courts would retry criminal defendants to determine relevant facts, the Government argues, those practical difficulties are not present here because the BIA and the Attorney General have said that immigration courts’ de novo factual inquiries are worthwhile. And in response to Nijhawan’s concerns about the burden of proof, the Government notes that the burden in immigration proceedings is unaffected, so the immigration court would need to find clear and convincing facts in outside evidence (such as a sentencing report).

Whatever ambiguity remains, the Government insists, must be resolved by deferring to the Attorney General’s interpretation. And principles of lenity should apply only after deferring to that interpretation; otherwise, the Attorney General’s authority to interpret the INA would be meaningless. Moreover, deference is not precluded by Subparagraph (M)(i)’s connection to criminal law; the Subparagraph is purely internal to the INA, as opposed to Title 18 of the U.S. Code, and most of its applications occur in the civil immigration context.

In his reply brief, Nijhawan again makes his case for a categorical approach and rebuts some of the Government’s stronger arguments. First, he rejects the Government’s reading of the “that” and “in which” clauses, suggesting instead that the last antecedent rule means the “in which” clause, containing the amount-of-loss requirement, should limit the “that” clause containing the elements. Alternatively, he argues that even if the Government is correct that the amount-of-loss requirement modifies “an offense” directly, a defendant must nonetheless be “convicted of” that amount of loss. He also attempts to distinguish Hayes, cited by the Government, as addressing a structurally different statute and posing problems not present here. Next, he cites the U.S. Attorney’s Manual and the DOJ Criminal Tax Manual to argue that because Subparagraph (M)(ii) clearly “contemplate[s] application of the modified categorical approach,” applying the approach to Subparagraph (M)(i) would pose no conflict.

Nijhawan next explains why his reading would not frustrate Congress’s purposes. First, he again emphasizes that Congress “legislated against the backdrop of the categorical approach” and dismisses legislative history concerning deporting criminal aliens that post-dates the enactment of Subparagraph (M)(i). Second, he argues that his interpretation would not preclude the Subparagraph’s application to most criminal fraud statutes, because “in accordance with past practice,” prosecutors could include amount-of-loss provisions in plea agreements. This would pose no more commandeering problems, he suggests, than the Government’s ongoing efforts to foster local government support for federal immigration enforcement. Moreover, in addition to the “majority of states” he identified where loss is an element of a fraud offense, he notes that the modified categorical approach would further include states that follow the Model Penal Code approach to fraud offenses: requiring a jury finding of the amount of loss to determine the grade of the offense.

Next, Nijhawan responds to some of the Government’s major points regarding the categorical approach. He argues that the Sixth Amendment concerns from the Court’s sentencing enhancement cases are implicated here, because the “aggravated felony” definition applies also to a criminal immigration statute. On the question of what burden of proof the amount of loss would be held to, he suggests that “the government gets this issue backward” because by requiring a conviction as the predicate for deportation, the statute requires proof beyond a reasonable doubt.

Nijhawan then expands his previous argument for the rule of lenity. He emphasizes the connection to the criminal immigration statute – an offense that is frequently prosecuted – to suggest that the traditional rule of lenity should apply, and no deference should be due to the BIA, because the Court’s interpretation of Subparagraph (M)(i) will greatly impact future criminal prosecutions in which the provision would be construed uniformly. The Court’s prior cases interpreting other aspects of the definition of “aggravated felony” have applied the rule of lenity and have not mentioned agency deference, he notes. As an alternative, he argues again for the rule of lenity or narrow construction in the immigration context, challenging as inaccurate the Government’s claim that lenity comes into play only after deference has been accorded.