Farhane v. United States

20-1666
Farhane v. United States

In the
United States Court of Appeals
For the Second Circuit
________

AUGUST TERM 2021

ARGUED: NOVEMBER 15, 2021
DECIDED: AUGUST 11, 2023

No. 20-1666

ABDERRAHMANE FARHANE,
Petitioner-Appellant,

v.

UNITED STATES OF AMERICA,
Respondent-Appellee,
________

Appeal from the United States District Court
for the Southern District of New York.
________

Before: WALKER, WESLEY, and CARNEY, Circuit Judges.
________
Abderrahmane Farhane appeals from the denial in the United
States District Court for the Southern District of New York (Loretta A.
Preska, J.) of his habeas petition to vacate his 2006 guilty plea,
conviction, and sentence. Farhane asserts that he received ineffective
assistance of counsel because his lawyer did not warn him of the risks
of denaturalization and possible subsequent deportation arising from
2 No. 20-1666

his guilty plea. For the reasons that follow, we affirm the district
court.

Judge Walker writes for the majority and concurs in a separate
opinion.

Judge Carney dissents in a separate opinion.

________

RAMZI KASSEM (Naz Ahmad, Princess
Masilungan, Mudassar Toppa, on the brief), CLEAR
Clinic, Main Street Legal Services, Inc., CUNY
School of Law, Long Island City, NY; Alan E.
Schoenfeld, Margaret T. Artz, Paloma Naderi, on
the brief, Wilmer Cutler Pickering Hale and Dorr
LLP, New York, NY, for Petitioner-Appellant
Abderrahmane Farhane.

JUN XIANG (Karl Metzner, on the brief), Assistant
United States Attorneys, for Audrey Strauss,
United States Attorney for the Southern District of
New York, New York, NY, for Respondent-Appellee
United States of America.

Joel B. Rudin, Matthew A. Wasserman, Haran Tae,
on the brief, National Association of Criminal
Defense Lawyers, New York, NY, and Law Offices
of Joel B. Rudin, P.C., New York, NY; Timothy P.
Murphy, on the brief, New York State Association
of Criminal Defense Lawyers, Albany, NY, for
amici curiae National Association of Criminal Defense
Lawyers and New York State Association of Criminal
Defense Lawyers.

Andrew Z. Michaelson, Ana V.B. Daily, Kathryn
R. Barry, on the brief, King & Spalding LLP, New
York, NY; John C. Yang, Niyati Shah, Marita
3 No. 20-1666

Etcubañez, on the brief, Asian Americans
Advancing Justice, Washington, DC, for amicus
curiae Asian Americans Advancing Justice.

Andrew D. Silverman, Daniel A. Rubens, Alyssa
Barnard-Yanni, Lauren Weber, on the brief, Orrick,
Herrington & Sutcliffe LLP, New York, NY,
Seattle, WA, Washington, DC, for amici curiae
Professors of Criminal Law, Criminal Procedure,
Immigration Law, and Legal Ethics.

Manuel D. Vargas, Marie Mark, Nabilah
Siddiquee, Leila Kang, on the brief, Immigrant
Defense Project, New York, NY, for amicus curiae
Immigrant Defense Project.

________

JOHN M. WALKER, JR., Circuit Judge:

Abderrahmane Farhane appeals from the denial in the United
States District Court for the Southern District of New York (Loretta A.
Preska, J.) of his habeas petition to vacate his 2006 guilty plea,
conviction, and sentence. Farhane asserts that he received ineffective
assistance of counsel because his lawyer did not warn him of the risks
of denaturalization and possible subsequent deportation arising from
his guilty plea. For the reasons that follow, we AFFIRM the district
court.

BACKGROUND

In 2006, Appellant Farhane, a naturalized American citizen
born in Morocco, pleaded guilty to providing false statements to
federal law enforcement and conspiring to violate a money
laundering statute, 18 U.S.C. § 1956. In his allocution, he stated that,
in 2001, he conspired to transfer money to mujahideen in Afghanistan
and Chechnya. He was sentenced in the Southern District of New
4 No. 20-1666

York to 156 months of imprisonment and two years of supervised
release. In 2011, after a lengthy appeal process, his conviction became
final.

In 2017, Farhane was released from custody and returned to his
home in Brooklyn. In August 2018, the government filed a complaint
in the Eastern District of New York seeking to revoke Farhane’s
citizenship under 8 U.S.C. § 1451(a). Section 1451(a) provides for the
civil denaturalization of individuals whose naturalization orders and
certificates were “illegally procured or were procured by concealment
of a material fact or by willful misrepresentation.” 1

Farhane had been naturalized in 2002. During that process, he
told the government, on two forms and in one interview under oath,
that he had never knowingly committed a crime for which he had not
been arrested. This was a lie. In fact, just a few months earlier,
Farhane had conspired with two others (one of whom was an FBI
informant) to send money to fighters engaged in jihad in Afghanistan
and Chechnya. They discussed the topic multiple times in person and
over the phone, and Farhane gave advice as to how to avoid law
enforcement detection of the money transfer. Relying on Farhane’s
2006 guilty plea to these crimes, the denaturalization complaint
alleged that Farhane had been unlawfully naturalized because he had:
(a) joined a money laundering conspiracy and (b) concealed it from
naturalization authorities.

In December 2018, while still on supervised release after
serving his prison sentence, Farhane filed a 28 U.S.C. § 2255 habeas
corpus petition in the Southern District of New York to vacate his
guilty plea, conviction, and sentence. He asserted that he had
received ineffective assistance of counsel at the time of his 2006 guilty

1 8 U.S.C. § 1451(a).
5 No. 20-1666

plea because his lawyer had not warned him of the risk of
denaturalization and deportation before he pleaded guilty. 2 The
district court denied the petition, concluding that his counsel’s failure
to warn him of the denaturalization risk was not objectively
unreasonable. We granted Farhane’s motion for a certificate of
appealability. The denaturalization proceeding in the Eastern District
has been stayed pending the resolution of Farhane’s habeas petition.

DISCUSSION

On appeal, Farhane claims that he received ineffective
assistance of counsel when his lawyer failed to tell him that
denaturalization and deportation could be consequences of his
pleading guilty. The Supreme Court set forth the test for ineffective
assistance in Strickland v. Washington, requiring a defendant to
establish both his trial counsel’s deficient performance and the
defendant’s resulting prejudice. 3 Farhane argues that the Sixth
Amendment required his lawyer to warn him of the possible
naturalization and immigration consequences of his guilty plea.
Farhane says that, if he had known of these risks, he would not have
pleaded guilty.

The government responds by arguing, as a threshold matter,
that the Sixth Amendment does not require attorneys to warn of the
risk of denaturalization. It also asserts that Farhane cannot establish
either element of the Strickland test. 4 Because civil denaturalization is

2 The government’s complaint only seeks Farhane’s denaturalization,
but, if it is granted, he anticipates the government will then move to deport
him.
3 See 466 U.S. 668, 687 (1984).

The government also offers another argument on this appeal—that
4

Farhane is advancing a “new rule” that, under Teague v. Lane, 489 U.S. 288,
6 No. 20-1666

a collateral and not a direct consequence of a conviction, we agree that
the Sixth Amendment does not require attorneys to warn of that risk.
Thus, we affirm the district court.

I. The Sixth Amendment and Denaturalization

The Sixth Amendment guarantees criminal defendants the
effective assistance of counsel during plea negotiations. 5 Effective
assistance includes warning defendants of the “direct” consequences
of pleading guilty, such as the offense’s maximum prison term and
the likely sentence as set forth in a plea agreement. 6 We have long
held, however, that an attorney need not warn of every possible
“collateral consequence of conviction.” 7 Such collateral consequences

316 (1989), cannot be established on collateral review. The government did
not make this argument in the district court, however, and has offered no
justification for the omission. Accordingly, the argument is waived. See
Greene v. United States, 13 F.3d 577, 586 (2d Cir. 1994) (noting that “it is a
well-established general rule that an appellate court will not consider an
issue raised for the first time on appeal” when “[n]o reason is offered by the
government for the failure to raise it below”); see also Danforth v. Minnesota,
552 U.S. 264, 289 (2008) (stating that a litigant “can waive a Teague defense,
during the course of litigation, . . . by failing to raise it in a timely manner”).
5 See Lafler v. Cooper, 566 U.S. 156, 162 (2012).

See, e.g., Brady v. United States, 397 U.S. 742, 755 (1970) (defining direct
6

consequences in the closely analogous Fifth Amendment context as
including “the actual value of any commitments made to him by the court,
prosecutor, or his own counsel”); United States v. Del Rosario, 902 F.2d 55, 59
(D.C. Cir. 1990), abrogated on other grounds by Padilla v. Kentucky, 559 U.S. 356
(2010); United States v. Salmon, 944 F.2d 1106, 1130 (3d Cir. 1991), abrogated
on other grounds by United States v. Caraballo-Rodriguez, 726 F.3d 418 (3d Cir.
2013).

United States v. Parrino, 212 F.2d 919, 921 (2d Cir. 1954); Santiago v.
7

Laclair, 588 F. App’x 1, 4 (2d Cir. 2014); see also Chaidez v. United States, 655
F.3d 684, 690 (7th Cir. 2011), aff’d, 568 U.S. 342 (2013) (noting that “the lower
7 No. 20-1666

are “categorically removed from the scope of the Sixth Amendment.” 8
A defendant can only establish an ineffective assistance claim as to a
collateral consequence if his attorney affirmatively misadvises him. 9
Failing to warn of the collateral risk alone is not enough.

The instant appeal is resolved by the straightforward
application of this direct/collateral framework. Farhane and the
dissent suggest, however, that this framework may not survive the
Supreme Court’s decision in Padilla v. Kentucky. 10 Accordingly, we
take this opportunity to reaffirm the direct/collateral distinction and
apply it in the post-Padilla context. In doing so, we hold that the
distinction remains valid, that it applies to civil denaturalization, and
that such denaturalization is a collateral consequence of conviction
and so is not covered by the Sixth Amendment’s right to effective
assistance of counsel.

A. The Direct/Collateral Framework

The distinction between direct consequences, as to which a
constitutionally competent attorney must advise her client before he
enters a guilty plea, and collateral consequences, about which she

federal courts, including at least nine Courts of Appeals, had uniformly
held that the Sixth Amendment did not require counsel to provide advice
concerning any collateral (as opposed to direct) consequences of a guilty
plea”); Torrey v. Estelle, 842 F.2d 234, 237 (9th Cir. 1988) (holding that
“[f]ailure to advise [the defendant] of a collateral penalty cannot be held to
be below an objective standard of reasonableness” and thus evidence of
ineffective assistance).
8 Chaidez v. United States, 568 U.S. 342, 349 (2013) (quotation marks
omitted); see id. at 350-52 (describing the “almost unanimous[]” consensus
of state and federal courts).
9See, e.g., United States v. Santelises, 509 F.2d 703, 704 (2d Cir. 1975) (per
curiam).
10 559 U.S. 356 (2010).
8 No. 20-1666

need not, is grounded in the text of the Sixth Amendment. The
amendment guarantees a defendant “the [a]ssistance of [c]ounsel for
his defen[s]e” in his “criminal prosecution[].” 11 The amendment’s
scope is thus textually limited to the direct consequences of the
prosecution and does not require “sound advice about the collateral
consequences of conviction.” 12

This language accords with the practical limitations of law
practice. Criminal lawyers are “not expected to possess—and very
often do not possess—expertise in other areas of the law, and it is
unrealistic to expect them to provide expert advice on matters that lie
outside their area of training and experience.” 13 Indeed, a
conviction’s potential collateral consequences are numerous and
varied. They include “civil commitment, civil forfeiture, the loss of
the right to vote, disqualification from public benefits, ineligibility to
possess firearms, dishonorable discharge from the Armed Forces, and
loss of business or professional licenses.” 14 Some limiting principle is
thus required. Otherwise, as Farhane is attempting to do here, guilty
pleas could be overturned years or decades later due to events that a
competent defense counsel could not have reasonably foreseen.

It is unsurprising, then, that courts have nearly uniformly
applied the direct/collateral framework in the Sixth Amendment
context. In Hill v. Lockhart, the Supreme Court explicitly avoided the
question of whether the Sixth Amendment applies to the collateral

11 U.S. Const. amend. VI.
12 See Padilla, 559 U.S. at 388 (Scalia, J., dissenting).
13 Id. at 376 (Alito, J., concurring).
14 Id.
9 No. 20-1666

consequences of a guilty plea. 15 Since then, at least ten circuit courts
and thirty state appellate courts have held that “counsel’s failure to
inform a defendant of the collateral consequences of a guilty plea is
never a violation of the Sixth Amendment,” with only a handful of
state courts concluding otherwise. 16 Indeed, the framework is so
uniformly applied that, despite lacking explicit Supreme Court
sanction, it has been described as “one of the most widely recognized
rules of American law.” 17

Our dissenting colleague suggests that this substantive and
precedential firmament was undermined by the Supreme Court’s
opinion in Padilla v. Kentucky. In Padilla, the Court held that the Sixth
Amendment required counsel to “inform her client whether his plea
carries a risk of deportation.” 18 This holding contravened the “almost
unanimous[]” circuit court consensus that deportation was a
collateral consequence of conviction. 19 In so holding, the Court noted
that it had “never applied a distinction between direct and collateral
consequences to define the scope” of effective counsel. 20

Thus, Padillia “breach[ed] the previously chink-free wall
between direct and collateral consequences.” 21 But, in doing so, it did

15474 U.S. 52, 60 (1985); see Chaidez, 568 U.S. at 349 (characterizing the
decision).
16 Chaidez, 568 U.S. at 350 (quotation marks omitted).
17 Id. at 351 (quotation marks omitted).
18559 U.S. at 374. At times, the opinion refers generally to “immigration
consequences.” See, e.g., id. at 369. The holding and reasoning, however,
are clearly limited to deportation arising from a criminal conviction.
19 Chaidez, 568 U.S. at 350.
20 Padilla, 559 U.S. at 365-66.
21 Chaidez, 568 U.S. at 352-53.
10 No. 20-1666

not reject the direct/collateral framework altogether. Instead, it held
that the framework was “ill suited to evaluating” whether the risk of
deportation was covered by the Sixth Amendment duty to warn
because of deportation’s “particularly severe” nature and near-
automatic relationship to criminal conviction. 22 These characteristics
made it “uniquely difficult to classify as either a direct or a collateral
consequence.” 23 Subsequently, Justices have emphasized that Padilla
“did not eschew the direct-collateral divide across the board” 24 and
that deportation is “the most difficult penalty to classify as either a
collateral or direct consequence.” 25

Our court has also recognized Padilla’s narrow reasoning,
noting that it was “limited to . . . deportation,” 26 as has the Seventh
Circuit, which stated that “Padilla is rife with indications that the
Supreme Court meant to limit its scope to the context of deportation
only.” 27 In the face of the Sixth Amendment’s text, the practical
necessity, and our well-established precedent, Padilla’s “chink” in the
otherwise solid wall of precedent does not justify the wholesale
abandonment of the direct/collateral distinction. Instead, we align
ourselves with the Seventh Circuit in applying our pre-Padilla
precedents and affirming the distinction’s threshold applicability to
the Sixth Amendment. 28

22 Padilla, 559 U.S. at 365-66.
23 Id. at 357.
24 Chaidez, 568 U.S. at 355.
25 Id. at 366 (Sotomayor, J., dissenting) (quotation marks omitted).
26 United States v. Youngs, 687 F.3d 56, 62 (2d Cir. 2012).
27 United States v. Reeves, 695 F.3d 637, 640 (7th Cir. 2012).
28 Id.
11 No. 20-1666

The dissent argues that “many courts have responded [to
Padilla] by closely reexamining their Sixth Amendment precedents,”
but offers in support only a handful of state supreme and
intermediate court cases and one non-precedential Tenth Circuit
summary order addressing Padilla briefly in a footnote. 29 Meanwhile,
several state supreme courts and a circuit court have reaffirmed the
pre-Padilla framework. 30 The cases cited by the dissent do not amount
to a widespread reconsideration of the previously settled law, which
is unsurprising given Padilla’s explicitly limited scope. The few state
court cases cited by the dissent do not justify ignoring our
longstanding precedent. 31

Our dissenting colleague acknowledges that Padilla did not
“eradicate[] entirely the direct/collateral framework in the Sixth
Amendment context” and that “[t]he dichotomy will continue to
apply usefully in some—even many or most—cases.” 32 She argues,
however, that the applicability of the framework should be addressed
consequence by consequence. But the utility of the framework

See, e.g., Diss. Op. at 9, 16 n.9 (citing United States v. Tuakalau, 562 F.
29

App’x 604, 609 n.4 (10th Cir. 2014) (summary order); Commonwealth v.
Thompson, 548 S.W.3d 881, 891-94 (Ky. 2018); Alexander v. State, 772 S.E.2d
655, 659 (Ga. 2015); People v. Hughes, 983 N.E.2d 439, 454-56 (Ill. 2012);
Commonwealth v. Pridham, 394 S.W.3d 867, 879 (Ky. 2012); Taylor v. State, 698
S.E.2d 384, 388 (Ga. Ct. App. 2010)).

See, e.g., Reeves, 695 F.3d 637; Taylor v. State, 887 N.W.2d 821, 823 (Minn.
30

2016); State v. Trotter, 330 P.3d 1267, 1269 (Utah 2014).

See In re Arab Bank, PLC Alien Tort Statute Litig., 808 F.3d 144, 155 (2d
31

Cir. 2015), as amended (Dec. 17, 2015), aff’d sub nom. Jesner v. Arab Bank, PLC,
138 S. Ct. 1386 (2018) (noting that we are bound to follow our precedent
unless there is “conflict, incompatibility, or inconsistency” between it and a
intervening Supreme Court decision (quotation marks and alterations
omitted)).
32 Diss. Op. at 14.
12 No. 20-1666

necessarily lies in it being a guide to defense counsel when advising
a client in advance of a guilty plea. The framework as it exists allows
defense lawyers to classify the many different types of consequences
of conviction as either direct or collateral. Artificially dismantling the
well-established categorical direct/collateral distinction and confining
our reasoning to a case-by-case approach invites incoherence and
confusion. Of course, we cannot rule out that there may be other
apparently collateral consequences that are so severe and automatic
that they are, like deportation, “ill suited” to the framework. As we
discuss in the following section, however, civil denaturalization is not
such a consequence. And as a general matter we adhere to our pre-
Padilla precedents in which the collateral/direct distinction is
presumed to apply.

B. The Framework’s Applicability to Denaturalization

Accepting the direct/collateral framework’s general validity,
we now look to whether it applies to civil denaturalization. Farhane
and the dissent argue that it does not. They analogize
denaturalization to deportation, and, citing Padilla, argue that it is
similarly “ill suited” to the direct/collateral framework.

The Padilla Court emphasized two factors when concluding
that the distinction did not apply to deportation: deportation’s
severity and its automatic character. 33 In Padilla and Chaidez, the
Court emphasized deportation’s “particularly severe” character, 34
and the government does not dispute that, like deportation,

See Chaidez, 568 U.S. at 352. Farhane also argues that the Court
33

considered the professional recognition of the duty to warn of the risk of
deportation. The Court’s discussion of professional consensus, however,
was limited to the Strickland factors, which are not relevant in this analysis.
See Padilla, 559 U.S. at 366-67.
34 Padilla, 559 U.S. at 365; Chaidez, 568 U.S. at 352.
13 No. 20-1666

denaturalization is serious. 35 Severity alone, however, is not enough
to equate the two proceedings. Padilla emphasized that it was
deportation’s “close connection to the criminal process” that made it
“uniquely difficult to classify as either a direct or a collateral
consequence.” 36 So, with the severity of denaturalization undisputed,
we turn to the Padilla Court’s analysis of deportation’s nearly
automatic relationship to conviction.

Padilla concluded that the direct/collateral framework was not
applicable because deportation, while technically a civil proceeding,
was “nevertheless intimately related to the criminal process.” 37 This
intimate connection was the result of “recent changes in our
immigration law” that made “removal nearly an automatic result for
a broad class of noncitizen offenders.” 38 The Court noted that an alien
was rendered deportable by the fact of his conviction. 39 Thus, it was
the guilty plea itself that rendered Padilla deportable. This automatic
relationship between conviction and deportation made it “uniquely
difficult to classify [deportation] as either a direct or a collateral
consequence,” and so the Court concluded that the Sixth Amendment
applied to warning of the risk of deportation without regard to the
direct/collateral threshold distinction. 40

See Knauer v. United States, 328 U.S. 654, 659 (1946) (stating that
35

“denaturalization, like deportation, may result in the loss of all that makes
life worth living” (quotation marks omitted)).
36 Padilla, 559 U.S. at 366.
37 Id. at 365.
38 Id. at 366.
39 See id. at 368 (citing 8 U.S.C. § 1227(a)(2)).
40 Id. at 366.
14 No. 20-1666

Civil denaturalization, by contrast, lacks this “automatic”
relationship to the guilty plea. It depends not on the fact of conviction
but on the individual’s actions before and at the time of
naturalization. 41 It can occur with or without a criminal conviction.
While Farhane’s conviction might alleviate the government’s
evidentiary burden, the government could have sought his
denaturalization by proving his substantive conduct, without waiting
for, relying on, or even referencing his criminal conviction. Civil
denaturalization is therefore not “intimately related to the criminal
process,” and applying the direct/collateral framework presents no
particular difficulty. 42

C. Denaturalization as a Collateral Consequence

Having recognized the direct/collateral framework’s ongoing
vitality and applicability to this case, we now reach the “threshold
question” for an ineffective assistance claim based on a failure to
warn. That is, whether the unwarned of event—here,
denaturalization—was a direct or collateral consequence of the guilty
plea. 43

This is true of denaturalization under 8 U.S.C. § 1451(a), commonly
41

referred to as “civil denaturalization.” In contrast, “criminal
denaturalization,” governed by § 1451(e), provides that when a person is
convicted of unlawfully naturalizing in violation of 18 U.S.C. § 1425, the
“the court in which such conviction is had shall thereupon revoke, set aside,
and declare void the final order admitting such person to citizenship, and
shall declare the certificate of naturalization of such person to be canceled.”
See Cassandra Burke Robertson & Irina D. Manta, (Un)civil Denaturalization,
94 N.Y.U. L. Rev. 402, 407 (2019) (describing the two kinds of
denaturalization). Our holding in this case is limited to civil
denaturalization.
42 Padilla, 559 U.S. at 365.
43 Chaidez, 568 U.S. at 349.
15 No. 20-1666

A consequence is collateral if it does not “directly flow[] from
the judgment,” even if it “depend[s] on a conviction of [a] crime.” 44
We have described as collateral a variety of results not directly
connected to criminal proceedings, such as losing certain civil rights,
including the right to vote; becoming ineligible to serve in the armed
forces; and being collaterally estopped from raising defenses in civil
cases. 45 At the same time, the Supreme Court, while never
“attempt[ing] to delineate the world of ‘collateral consequences,’” has
noted that “effects of a conviction commonly viewed as collateral
include civil commitment, civil forfeiture, sex offender registration,
disqualification from public benefits, and disfranchisement.” 46 In the
closely related Fifth Amendment context, we have defined direct
consequences as those that “have a definite, immediate and largely
automatic effect on the range of the defendant’s punishment,” with
all other consequences being collateral. 47

In light of this precedent, Farhane’s pending denaturalization
is plainly a collateral consequence of his guilty plea. Civil
denaturalization is a separate proceeding that may or may not occur
following the plea. 48 The government exercises considerable
discretion in bringing denaturalization cases, as does the district court

44 Parrino, 212 F.2d at 921.
45 See id. at 922; Del Rosario, 902 F.2d at 59.
46 Chaidez, 568 U.S. at 349 n.5.

Youngs, 687 F.3d at 60 (quotation marks omitted); cf. id. at 62 (noting
47

that the “Sixth Amendment responsibilities of counsel to advise of the
advantages and disadvantages of a guilty plea are greater than the
responsibilities of a court under the Fifth Amendment”).

See 8 U.S.C. § 1451(a) (empowering the government to “institute
48

proceedings in any district court of the United States in the judicial district
in which the naturalized citizen may reside . . . for the purpose of revoking
and setting aside the order admitting such person to citizenship”).
16 No. 20-1666

in evaluating the evidence. 49 As noted earlier, denaturalization also
does not require a conviction; indeed, the government could seek to
denaturalize Farhane without relying on his guilty plea. Instead, it
turns on the defendant’s actual conduct, which the government must
establish by “clear, unequivocal, and convincing evidence.” 50 Thus,
denaturalization does not “directly flow[] from” a conviction. 51 And
while Farhane’s guilty plea likely prevents him from disputing his
eligibility for denaturalization, collateral estoppel alone does not
transform a collateral consequence into a direct one. 52

As a collateral consequence, denaturalization is “categorically
removed from the scope of the Sixth Amendment.” 53 Accordingly, as
a categorical matter, Farhane’s counsel was not ineffective for failing
to warn him of the risk of denaturalization in advising him on his
guilty plea. For the same reason, Farhane’s counsel was not
ineffective for failing to warn him of the risk of deportation. Farhane
can only be deported if he is first denaturalized. As a possible
consequence of a collateral proceeding, the risk of deportation in this
case is even further removed from the direct consequences of
Farhane’s conviction. Therefore, Farhane cannot pass the threshold
to establish that he suffered ineffective assistance of counsel, and the

See Amber Qureshi, The Denaturalization Consequences of Guilty Pleas,
49

130 Yale L.J. F. 166, 169-71 (2020) (describing the history of denaturalization
and noting its historical rarity); see also Youngs, 687 F.3d at 63 (emphasizing
the importance of the court’s power to independently evaluate the evidence
in the direct/collateral analysis).

Kungys v. United States, 485 U.S. 759, 781 (1988) (quotation marks
50

omitted).
51 Parrino, 212 F.2d at 921.
52 See id. at 922.
53 Chaidez, 568 U.S. at 349 (quotation marks omitted).
17 No. 20-1666

district court’s denial of his habeas petition to withdraw his guilty
plea should not be disturbed.

Farhane’s counterarguments are not persuasive. He contends
that his denaturalization and subsequent deportation are direct
consequences of his guilty plea. They are direct, he claims, because
his guilty plea prevents him from contesting the grounds for his
denaturalization. As discussed above, however, collateral estoppel
guaranteeing the outcome of a particular proceeding does not make
that proceeding “direct.” Holding otherwise would render a wide
variety of civil proceedings direct, vastly expanding the Sixth
Amendment’s scope. 54 Farhane attempts to limit this expansion by
emphasizing the severity of denaturalization. But many collateral
consequences are severe, including disenfranchisement, the loss of
the right to travel abroad, the revocation of a driver’s license, civil
commitment, and sex offender registration. 55 Once the
direct/collateral framework is applied, whether or not a consequence
is collateral does not turn on its severity. 56

Farhane also argues that Padilla requires warning in his case
because the government’s ultimate intention is to deport him. He is
wrong. The government’s motive, whatever it may be, does not
transform denaturalization into a direct consequence of his plea.
Because denaturalization is collateral, everything thereafter to which
it is a precondition is also collateral. Thus Padilla, which requires
warnings of deportations arising automatically from guilty pleas, has

54 See Parrino, 212 F.2d at 922.

See Chaidez, 568 U.S. at 349 n.5 (listing examples); Del Rosario, 902 F.2d
55

at 59 (collecting cases).
56 See Youngs, 687 F.3d at 60 (describing the relevant considerations).
18 No. 20-1666

no purchase here, where the possibility of deportation is predicated
on a collateral consequence.

Finally, Farhane and the dissent both suggest that it would be
incongruous to hold that the Sixth Amendment does not apply to
warnings about denaturalization because it would mean that
naturalized citizens, who are not entitled to a warning about a plea
that could lead to deportation down the road, are less protected than
noncitizens, who must be warned if their plea exposes them to
deportation. But this compares apples to oranges: Congress chose to
treat the civil denaturalization of citizens differently from the removal
of aliens by not tying denaturalization to a criminal conviction. It is
unsurprising that a competent counsel’s obligations would differ
under these different statutory schemes.

Farhane’s possible denaturalization and possible subsequent
deportation are collateral consequences of his guilty plea. The Sixth
Amendment does not require competent counsel to warn Farhane of
these risks. Thus, Farhane’s ineffective assistance of counsel claim
flounders at the threshold question of the Sixth Amendment’s
applicability. Accordingly, we affirm the district court without
needing to apply Strickland.

CONCLUSION

For the foregoing reasons, we AFFIRM the judgment of the
district court.
20-1666
Farhane v. United States

WALKER, Circuit Judge, concurring:

I write separately, in response to Judge Carney’s dissent, to
explain that Farhane’s ineffective assistance claim would also fail
under the Strickland analysis. To establish ineffective assistance, a
defendant must show that his “counsel’s representation fell below an
objective standard of reasonableness” and that he suffered prejudice
as a result. 1 Judge Carney would hold that defense counsel acted
unreasonably in failing to warn Farhane of the risk of
denaturalization when Farhane pleaded guilty in 2006. I disagree.

The first prong—whether counsel acted unreasonably—“is
necessarily linked to the practice and expectations of the legal
community.” 2 This is because “[t]he proper measure of attorney
performance remains simply reasonableness under prevailing
professional norms.” 3 Therefore, “[p]revailing norms of practice as
reflected in American Bar Association standards and the like . . . are
guides to determining what is reasonable.” 4 In interpreting such
standards, however, we must recognize that some of them “may
represent only the aspirations of a bar group rather than an empirical
assessment of actual practice.” 5 Ultimately, the “courts must judge
the reasonableness of counsel’s challenged conduct on the facts of the

1 Strickland, 466 U.S. at 688, 692; see United States v. Pinhasov, 762 F. App’x
43, 46 (2d Cir. 2019) (quoting the Strickland test).
2 Padilla, 559 U.S. at 366.
3 Strickland, 466 U.S. at 688.
4 Id. at 689.
5 Padilla, 559 U.S. at 377 (Alito, J., concurring).
2 No. 20-1666

particular case, viewed as of the time of counsel’s conduct,” 6 and
without the benefit of hindsight.

In this case, Farhane’s lawyer did not act unreasonably. The
record suggests that, while the lawyer knew that Farhane was
naturalized, he did not know (1) when Farhane was naturalized in
relation to the crimes to which Farhane later pleaded guilty and (2)
whether Farhane thus might have lied during his naturalization.
What is more, in 2006, when Farhane entered his plea, civil
denaturalization was extremely rare. The government brought fewer
than 150 civil denaturalization cases in the forty-four years between
1967 and 2012—just over three cases per year across the entire
country. 7 And Farhane has not cited any authority that admonished
lawyers to advise clients of the risk of denaturalization. Instead, the
contemporaneous ABA standards spoke generally of “basic
immigration consequences,” which would usually only apply to non-
citizens, without mentioning denaturalization. 8 General invocations
of “immigration consequences” are not enough to establish that, in
2006, competent counsel should have warned a client he knew to be
naturalized, but not when or under what circumstances, of the rare
possibility of civil denaturalization, much less any possibility of
deportation.

6 Roe v. Flores-Ortega, 528 U.S. 470, 477 (2000).
7 Qureshi, 130 Yale L.J. F. at 170.
8Standards for Criminal Justice, Pleas of Guilty 14-3.2, 14-3.2 cmt. (Am.
Bar Ass’n, 3d ed. 1999) (stating that “counsel should be familiar with basic
immigration consequences that flow from different types of guilty pleas”);
see also N.Y. State Defs. Ass’n, Client Advisory Bd., Client-Centered
Representation Standards 17 (July 25, 2005).
3 No. 20-1666

Moreover, because in 2006 the direct/collateral distinction
presented a “chink-free wall,” 9 a lawyer at that time would have
believed that constitutionally competent representation did not
require warnings of collateral consequences. Thus, it is even less
likely that in 2006 competent criminal defense counsel would have
felt obliged to investigate the obscure and remote possible collateral
consequences of the guilty plea. Of course, Farhane’s judgment upon
his 2006 guilty plea did not become final until 2011, so Padilla, decided
in 2010, is applicable law in this case. Nevertheless, at the time of the
guilty plea, his attorney could not have reasonably foreseen the
possible second-order consequences of a decision that would be
issued four years later. Strickland has never required an attorney to
gaze into a crystal ball.

Farhane’s lawyer did not know at the time that Farhane was at
risk of denaturalization; he was not put on notice of the risk by the
legal community’s standards; and, if it had somehow occurred to him,
despite denaturalization being a collateral consequence of a guilty
plea, he would have reasonably believed that no further investigation
or warning was constitutionally required. In this context, I cannot
conclude that his actions were objectively unreasonable.
Accordingly, I would also affirm the district court on the basis that
counsel was not ineffective under Strickland’s unreasonableness
prong.

9 Chaidez, 568 U.S. at 352.
20-1666
Farhane v. United States

CARNEY, Circuit Judge, dissenting:

Abderrahmane Farhane is a naturalized U.S. citizen who immigrated to this

country nearly three decades ago, settling here and raising a family. Because in 2006 he

pleaded guilty to criminal charges—charges for which he then duly served over ten

years in prison—Farhane can expect to lose his U.S. citizenship in denaturalization

proceedings that the government has already filed against him. When those

proceedings end, it is a “virtual certainty” that he will be deported. Sessions v. Dimaya,

138 S. Ct. 1204, 1211 (2018).

In a petition brought under 28 U.S.C. § 2255, he now plausibly maintains that in

the plea-bargaining process, his defense counsel did not advise him that his guilty plea

could carry any adverse immigration consequences at all. Farhane urges that the legal

representation he received was thus ineffective under the Sixth Amendment, relying on

Padilla v. Kentucky, 559 U.S. 356 (2010). In Padilla, the Supreme Court held that Jose

Padilla’s right to effective assistance required that counsel warn him of any deportation

risk arising from his 2002 guilty plea. 559 U.S. at 374; see Padilla v. Commonwealth, 381

S.W.3d 322, 324 (Ky. Ct. App. 2012) (setting forth the timeline of Padilla’s plea and

conviction). Accordingly, Farhane seeks vacatur of his guilty plea, sentence, and

conviction. A panel of this Court granted a certificate of appealability.

The Majority rejects Farhane’s petition, concluding that his expected

denaturalization—which will in turn lead to deportation—is a “collateral” consequence

of his conviction, and so his counsel bore no obligation to alert him to this risk of

pleading guilty. As I understand the Majority, even today a defendant has no right to be

warned by counsel that a guilty plea could lead to any adverse immigration

consequences, unless the client’s circumstances exactly match those in Padilla—that is,

the client must be a noncitizen whose guilty plea will lead to immediate removability.

In reaching this conclusion, the Majority embraces what in my view is an erroneously
restrictive interpretation of Padilla and incorrectly circumscribes the advice that a

criminal defendant is constitutionally entitled to receive.

In Padilla, as just summarized, the Supreme Court ruled that a noncitizen

received constitutionally ineffective assistance when his criminal defense counsel did

not advise him that his guilty plea to a state drug offense made him subject to automatic

deportation. 559 U.S. at 360. The Kentucky Supreme Court, having labeled deportation

a “collateral” consequence of a guilty plea, recognized no basis for an ineffective

assistance claim and rejected Padilla’s habeas petition. Id. at 364–65. Identifying error in

the state court’s analysis, the Supreme Court acknowledged that lower courts had long

used a distinction between “direct” and “collateral” consequences to categorize the

array of results flowing from a criminal conviction, but in the context of Padilla’s Sixth

Amendment claim it declined to apply the distinction to deportation. Id. Deportation is

a “particularly severe penalty,” it stressed, and observed that developments in

immigration law since 1990 made deportation “nearly an automatic result for a broad

class of noncitizen offenders.” Id. at 365–66 (internal quotation marks omitted). It

cautioned that deportation’s “close connection to the criminal process” made it

“uniquely difficult to classify as either a direct or a collateral consequence,” and

instructed that the dichotomy was “ill suited” to assessing the validity of a criminal

defendant’s Sixth Amendment challenge involving the risk of deportation. Id. at 366.

While declining to decide whether the direct/collateral distinction can, in other cases,

“define the scope of constitutionally ‘reasonable professional assistance’ required”

under Strickland v. Washington, 466 U.S. 668, 689 (1984), it proceeded to apply Strickland

to determine that Padilla’s counsel had fallen below an objectively reasonable standard

in 2002 when he failed to advise Padilla of the risk of deportation triggered by his guilty

plea. Padilla, 559 U.S. at 365. Criminal defendants, the Court said, may care deeply

about remaining in the United States, and Padilla was constitutionally entitled to be

advised of the severe risk of being removed. Id. at 368. It remanded for a determination

2
whether Padilla met Strickland’s second prong by showing prejudice from the

ineffective assistance he received. Id. at 375.

In my view, the direct/collateral dichotomy that the Padilla Court found inapt in

the context of deportation is similarly ill-suited to evaluating Farhane’s Sixth

Amendment claim, which rests on a risk of denaturalization of which he was never

informed. Regardless of the direct/collateral dichotomy’s intuitive appeal and practical

utility with respect to the many other types of consequences that a conviction may

carry, denaturalization shares with deportation the severity and entanglement with the

criminal process that the Padilla Court relied on to conclude that risk of deportation was

a poor fit for consideration under that framework. What’s more, denaturalization itself

risks a related deportation, underscoring the close parallels between Farhane’s and

Padilla’s Sixth Amendment claims.

The Supreme Court long ago recognized that stripping an individual of

citizenship “is an extraordinarily severe penalty.” Klapprott v. United States, 335 U.S. 601,

612 (1949) (plurality opinion). In line with that recognition, it has cautioned that

“denaturalization, like deportation, may result in the loss of all that makes life worth

living.” Knauer v. United States, 328 U.S. 654, 659 (1946) (internal quotation marks

omitted). And every naturalized citizen who, like Farhane, pleads guilty to pre-

naturalization criminal conduct has created for himself a serious risk of

denaturalization based on the theory that he acquired his citizenship illegally or by

willful misrepresentation. Indeed, the facts established through a guilty plea will be

treated as conclusive in a subsequent denaturalization proceeding. See Maietta v. Artuz,

84 F.3d 100, 102 n.1 (2d Cir. 1996). Courts in such denaturalization proceedings have

“no discretion to excuse the conduct” in question, Fedorenko v. United States, 449 U.S.

490, 517 (1981), and so the subjects of these proceedings should be prepared for the

prospect of deportation after denaturalization. And deportation, the Supreme Court has

observed, is “the equivalent of banishment or exile.” Delgadillo v. Carmichael, 332 U.S.

3
388, 390–91 (1947); see also Jordan v. De George, 341 U.S. 223, 232 (1951) (Jackson, J.,

dissenting) (describing deportation as a “life sentence of banishment”).

Farhane’s guilty plea thus exposed him, and continues to expose him, to serious

risks of denaturalization and of deportation. In light of the Supreme Court’s reasoning

in Padilla and in view of these additional considerations, I would conclude that the

direct/collateral dichotomy does not bar Farhane’s Sixth Amendment challenge. I would

reach the merits of his § 2255 claim. Applying the two-pronged Strickland test, I would

rule that the failure of Farhane’s counsel to advise him as to the risks of

denaturalization and deportation in connection with his guilty plea was objectively

unreasonable. Accordingly, I would vacate the district court’s judgment dismissing

Farhane’s claim and remand to allow the district court to address in the first instance

whether he was prejudiced by his counsel’s deficient performance.

By too swiftly dismissing Farhane’s arguments and too broadly reembracing the

direct/collateral dichotomy’s application to Sixth Amendment challenges, the Majority

mistakenly constricts Padilla. I agree with the Majority that in neither the Sixth nor the

Fifth Amendment contexts does Padilla require a wholesale abandonment of the

direct/collateral dichotomy. In my view, however, the Court’s reasoning and

commentary in Padilla all but mandate that we reach the same result as to

denaturalization as it did regarding deportation. A key result of the Majority’s decision,

I fear, is that many naturalized citizen criminal defendants will be left ignorant now, as

Farhane was in 2006, of the grave immigration risks that attach to their guilty pleas.

Under the Majority’s approach, those defendants will have no recourse. In my view,

and as I believe Padilla requires, the Sixth Amendment demands more.

For these reasons and others discussed below, I respectfully dissent.

4
I. Denaturalization cannot be dismissed as a merely “collateral” consequence in
assessing a Sixth Amendment ineffective assistance claim.

An individual serving a sentence for a federal crime may move to vacate the

sentence on the ground that “the sentence was imposed in violation of the Constitution

or laws of the United States.” 28 U.S.C. § 2255(a). 1 Under the Sixth Amendment, a

criminal defendant is entitled to the effective assistance of counsel “at ‘critical stages of

a criminal proceeding,’ including when he enters a guilty plea.” Jae Lee v. United States,

582 US. 357, 363 (2017) (quoting Lafler v. Cooper, 566 U.S. 156, 165 (2012)). The Supreme

Court instructed in Strickland v. Washington that a defendant’s failure to receive such

assistance, to his prejudice, is grounds for awarding habeas relief. See 466 U.S. at 697–98.

Ordinarily, we evaluate a defendant’s ineffective assistance claim by applying

the two-pronged approach set forth in Strickland: the court evaluates, first, whether

counsel’s performance fell below an objective standard of reasonableness, and second,

whether the defendant was prejudiced as a result. See id. at 687–88. The Majority does

not apply Strickland, however, because it holds that Farhane’s claim challenges a

“collateral” consequence of his conviction, and that the Sixth Amendment’s guarantee

of effective counsel therefore has no role to play. That holding is incorrect for several

reasons.

Padilla requires reconsidering whether the direct/collateral dichotomy
applies to denaturalization.

This case appears to be the first in the thirteen years since Padilla in which our

Court has, in a published decision, considered the direct/collateral dichotomy in the

Relief under this section is available to a “prisoner in custody under sentence of a court,”
1

but a petitioner who is subject to supervised release when he files a petition may be considered
“in custody” for purposes of section 2255(a). See Scanio v. United States, 37 F.3d 858, 860 (2d Cir.
1994). Farhane was on supervised release when he filed his section 2255(a) motion and is thus
eligible for relief.

5
Sixth Amendment context. 2 The distinction itself has been subject to varying

interpretations, but broadly has turned on the consequence’s causal proximity to the

conviction. In the context of a due process challenge to a guilty plea’s voluntariness, for

example, we explained that “direct consequences [are] those that have a definite,

immediate and largely automatic effect on the range of the defendant’s punishment,

and any other consequence is merely collateral.” United States v. Youngs, 687 F.3d 56, 60

(2d Cir. 2012) (internal quotation marks omitted). In a pre-Padilla case that the Majority

cites, meanwhile, we called collateral any consequence that “does not directly flow from

the judgment, even if it depends on a conviction of a crime.” Maj. Op. at 15 (internal

quotation marks omitted) (quoting United States v. Parrino, 212 F.2d 919, 921 (2d Cir.

1954)). No matter the approach taken, 3 neither the Majority nor the government

2 Not long ago, in a summary order, we expressly declined to address whether
“denaturalization is a ‘collateral’ consequence of a criminal conviction that falls outside the
ambit of the Sixth Amendment right to counsel”; we instead affirmed on the merits a district
court’s rejection of an ineffective assistance claim that was based on exposure to
denaturalization. See United States v. Nunez, 844 F. App’x 443, 444 (2d Cir. 2021). In addition, in
Santiago v. Laclair, we affirmed the denial of habeas relief to a state prisoner who claimed, based
on Padilla, that his counsel was ineffective for failing to advise that a sentence for a new crime
would run consecutively to the non-discharged portion of a sentence for a prior felony. 588 F.
App’x 1, 2–4 (2d Cir. 2014). In dictum, the Santiago Court characterized Padilla’s holding as
“narrow” and “limited specifically to the unique penalty of deportation,” based on
deportation’s severity, its impact on families, and the observation that “preserving the client’s
right to remain in the United States may be more important to the client than any potential jail
sentence.” Id. at 3–4 (quoting Padilla, 559 U.S. at 368) (internal quotation marks and alterations
omitted). Neither decision is precedential nor speaks to the applicability of the direct/collateral
framework in this case.

3 The Supreme Court has acknowledged—and declined to resolve—the “disagreement
among the courts over how to distinguish between direct and collateral consequences.” See
Padilla, 559 U.S. at 364 n.8. Some courts seem to consider “direct” only those consequences
“within the sentencing authority of the . . . trial court.” Id. at 364. On the flip side of the coin,
consequences that require the actions of “another agency over which the trial judge has no
control,” or those that automatically apply by operation of law, might be deemed “collateral.”
See, e.g., United States v. Delgado-Ramos, 635 F.3d 1237, 1239 (9th Cir. 2011) (noting the two
descriptions of “collateral” consequences and rejecting the “operation of law” standard). In the

6
disputes that consequences that courts have labeled “collateral” can be just as

devastating as those called “direct.” 4 Further, even those consequences that courts have

Fifth Amendment context, some courts have concluded that, even when a consequence is
“direct,” a court’s failure to warn of such a consequence before accepting a plea does not violate
due process if that particular consequence is “remedial and civil rather than punitive.” See
Mitschke v. State, 129 S.W.3d 130, 135 (Tex. Crim. App. 2004).

Indeed, courts have drawn differing boundaries between direct and collateral consequences
for use in assessing the Fifth Amendment’s due process requirements (that the trial court ensure
a defendant’s guilty plea is “intelligent and voluntary,” Youngs, 687 F.3d at 59) and in assessing
the Sixth Amendment’s guarantee of effective assistance (where counsel’s role is, in part, to
advise in the “negotiation of a plea bargain,” Padilla, 559 U.S. at 373). See infra Part I.A.3 and
note 9. Some commentators have traced the roots of the direct/collateral distinction back fifty
years to the Supreme Court’s Fifth Amendment decision in Brady v. United States, 397 U.S. 742
(1970). See, e.g., Jenny Roberts, The Mythical Divide Between Collateral and Direct Consequences of
Criminal Convictions: Involuntary Commitment of “Sexually Violent Predators”, 93 MINN. L. REV.
670, 684–85 (2008). In considering there how to assess the voluntariness of guilty pleas, the
Brady Court adopted Judge Tuttle’s language about direct consequences:

A plea of guilty entered by one fully aware of the direct consequences, including
the actual value of any commitments made to him by the court, prosecutor, or his
own counsel, must stand unless induced by threats (or promises to discontinue
improper harassment), misrepresentation (including unfulfilled or unfulfillable
promises), or perhaps by promises that are by their nature improper as having
no proper relationship to the prosecutor’s business (e.g. bribes).

Brady, 397 U.S. at 755 (emphasis added) (alteration omitted) (quoting Shelton v. United States, 246
F.2d 571, 572 n.2 (5th Cir. 1957) (en banc)). The Court did not otherwise address the idea of a
“direct consequence” or explore the notion of “collateral” consequences, but the array of cases
discussed in this dissent demonstrates the weighty legal significance with which these words
have since been imbued.

4 Consequences that courts have deemed “collateral” include revocation of parole, civil
commitment, consecutive rather than concurrent sentencing, disenfranchisement,
disqualification from public benefits, deportation, dishonorable discharge from the armed
services, and loss of business or professional licenses. Gabriel J. Chin & Richard W. Holmes, Jr.,
Effective Assistance of Counsel and the Consequences of Guilty Pleas, 87 CORNELL L. REV. 697, 705–06
(2002).

7
described as “collateral”—such as deportation—can be unavoidably “enmeshed [in]

criminal convictions.” Padilla, 559 U.S. at 365–66.

As described above, in Padilla the Supreme Court declined to categorize the “risk

of deportation” as either direct or collateral before proceeding to assess the petitioner’s

ineffective assistance claim under Strickland. 559 U.S. at 364–66. Rather, stressing that

deportation was a “severe penalty” and that it was “nearly an automatic result for a

broad class of noncitizen offenders,” the Court held that “[t]he collateral versus direct

distinction is . . . ill suited to evaluating a Strickland claim concerning the specific risk of

deportation.” Id. at 365–66 (internal quotation marks omitted). It ultimately concluded

that Padilla had sufficiently alleged that the counsel he received was constitutionally

deficient. Id. at 374.

In so finding—as it soon after confirmed in Chaidez v. United States—the Court

disclaimed strict allegiance to the proposition that “the Sixth Amendment does not

require attorneys to inform their clients of a conviction’s collateral consequences,

including deportation.” 568 U.S. 342, 350 (2013). Rather, the Padilla Court observed that,

in contrast to the lower courts, which were “almost unanimous[]” on the issue, id., the

Supreme Court itself had “never applied a distinction between direct and collateral

consequences to define the scope of constitutionally reasonable professional assistance

required under Strickland,” Padilla, 559 U.S. at 365 (internal quotation marks omitted).

And it expressly declined to consider “[w]hether that distinction is appropriate” in

cases involving consequences other than deportation. Id. It also did not explore the

boundaries of the dichotomy, offering no view about which consequences should be

considered “collateral” and which “direct.”

In declining to apply the direct/collateral framework to determine the availability

of an ineffective assistance claim, Padilla “breach[ed] the previously chink-free wall

between direct and collateral consequences,” Chaidez, 568 U.S. at 352–53, and thus

marked a “major upheaval in Sixth Amendment law,” Padilla, 559 U.S. at 383 (Alito, J.,

8
concurring). Commentators described the decision’s effects as “seismic.” See McGregor

Smyth, From “Collateral” to “Integral”: The Seismic Evolution of Padilla v. Kentucky and Its

Impact on Penalties Beyond Deportation, 54 HOW. L. J. 795, 798 (2011).

In the decade since, many courts have responded by closely reexamining their

Sixth Amendment precedents and sometimes limiting application of the

direct/collateral distinction in the context of a criminal defendant’s ineffective assistance

claim. See, e.g., Alexander v. State, 772 S.E.2d 655, 659 (Ga. 2015) (overruling its pre-Padilla

law and requiring that ineffective assistance claims be evaluated under Strickland

regardless “whether a guilty plea gives rise to a direct or collateral consequence”);

Commonwealth v. Thompson, 548 S.W.3d 881, 890 (Ky. 2018) (recognizing that, after

Padilla, “severe and definite consequences implicating effective assistance of counsel

may be ill-suited to classification as either direct or collateral but should be addressed in

a Padilla-type analysis”); Commonwealth v. Pridham, 394 S.W.3d 867, 879 (Ky. 2012)

(rejecting the Commonwealth’s “minimalist reading of Padilla” that would “implicate[]

no collateral consequence but deportation”); cf. United States v. Tuakalau, 562 F. App’x

604, 609 n.4 (10th Cir. 2014) (summary order) (suggesting that “Padilla may have called

the distinction between direct and collateral consequences into doubt”); Taylor v. State,

698 S.E.2d 384, 387 (Ga. Ct. App. 2010) (“Padilla . . . calls into question the application of

the direct versus collateral consequences distinction in the context of ineffective

assistance claims.”). 5

5 Other courts, to be sure, have after examination decided to maintain the distinction’s
application at least in some particular contexts. See, e.g., Commonwealth v. Thompson, 548 S.W.3d
881, 893 n.10 (Ky. 2018) (noting that “some states have concluded that counsel’s failure to
inform the client of sex offender registration cannot be ineffective assistance” (citing State v.
Trotter, 330 P.3d 1267, 1269 (Utah 2014) and Taylor v. State, 887 N.W.2d 821, 826 (Minn. 2016)));
Kennedy v. Kohnle, 810 S.E.2d 543, 548 n.4 (Ga. 2018) (collecting cases, and observing “that there
is a split among other jurisdictions as to whether to extend Padilla to advice about parole
eligibility”); State v. LeMere, 879 N.W.2d 580, 588–99 (Wis. 2016) (evaluating Wisconsin’s civil
commitment statute in light of Padilla and concluding that the possibility of civil commitment

9
 Unlike those courts, the Majority accepts with only a summary analysis the

notion that the direct/collateral framework applies to the particular consequence of

denaturalization. Indeed, the Majority here doubles down on the direct/collateral

distinction, declaring that it continues to apply uniformly, subject only to the adequacy

of counsel’s advice on the particular consequence addressed in Padilla: the risk of

deportation. Setting aside its resistance to examining whether the framework should

continue to apply more generally, see supra note 5, the Majority’s analysis here as to

Farhane’s submissions is flawed in several respects. First, it misreads the Supreme

Court’s ruling in Chaidez; second, it too readily accepts pre-Padilla precedent as still

applicable; and third, it obscures the important distinction between our Fifth

Amendment decisions on the court’s obligation to ensure a voluntary and knowing

plea, and our Sixth Amendment decisions on counsel’s obligation to provide

meaningful advice.

1. Chaidez is limited to Padilla’s retroactivity.

Two years after handing down Padilla, the Supreme Court held in Chaidez that

Padilla “announced a new rule” that was not previously “apparent to all reasonable

jurists”; the rule therefore did not apply retroactively to convictions that had already

become final on direct review. 568 U.S. at 354 (internal quotation marks omitted). Now

invoking Chaidez, the Majority suggests that the Supreme Court has endorsed the

application of the direct/collateral distinction in all cases not involving deportation. In

this, they err.

for people convicted of sexually violent offenses is a collateral consequence for which counsel
has no duty to advise). The lack of uniformity in response, however, only underscores that, after
Padilla, courts should seriously examine the direct/collateral framework’s applicability to the
consequences at issue in a given case.

10
 The Majority cites Chaidez for the sweeping proposition that “collateral

consequences are ‘categorically removed from the scope of the Sixth Amendment.’”

Maj. Op. at 6–7 (quoting Chaidez, 568 U.S. at 349). But in doing so, it takes the quoted

language out of context: Chaidez described the analysis in Padilla as “consider[ing] a

threshold question: Was advice about deportation ‘categorically removed’ from the

scope of the Sixth Amendment right to counsel[?]” Chaidez, 568 U.S. at 349 (quoting

Padilla, 559 U.S. at 366). The Chaidez Court then described Padilla as holding that advice

about deportation was not categorically removed from Sixth Amendment protection. Id.

at 353 (commenting that “Padilla . . . rejected that categorical approach—and so made

the Strickland test operative—when a criminal lawyer gives (or fails to give) advice

about immigration consequences”).

The Majority also relies on Chaidez for the proposition that the direct/collateral

dichotomy is “one of the most widely recognized rules of American law.” Maj. Op. at 9

(quoting Chaidez, 568 U.S. at 351). The Chaidez Court, however, offered this

characterization only to establish what the legal landscape looked like pre-Padilla. It

explained that when it decided Padilla—abandoning application of the dichotomy in the

context of deportation—it thus “altered the law of most jurisdictions.” Chaidez, 568 at

352. Indeed, Padilla announced a “new rule”—a new Sixth Amendment rule that did not

qualify for retroactive application. Id. at 348–50. In neither Padilla nor Chaidez did the

Court give a general endorsement to applying the direct/collateral distinction as to

consequences other than deportation. Nor did it otherwise “delineate the world of

‘collateral consequences’” as to which the Sixth Amendment would have no

application. Id. at 349 n.5. The Majority thus reads Chaidez incorrectly when it suggests

that Chaidez endorses the direct/collateral distinction here, or in any other particular

circumstance.

11
 2. The Majority erroneously relies on our pre-Padilla Sixth Amendment cases
and ignores developments since Padilla.

Second, the Majority relies heavily on caselaw of this Court that either preceded

Padilla, see, e.g., United States v. Santelises, 509 F.2d 703, 704 (2d Cir. 1975) (per curiam),

abrogated by Padilla, 559 U.S. 356; Parrino, 212 F.2d at 921, 6 or caselaw from other courts

that, while more recent, does not address denaturalization and provides, at best, limited

support to the Majority’s position, see, e.g., United States v. Reeves, 695 F.3d 637, 640 (7th

Cir. 2012) (considering whether failure to advise on enhanced sentencing for future

criminal conduct can give rise to an ineffective assistance of counsel claim).

In Parrino, a 1954 decision cited by the Majority, we observed that criminal

convictions can carry consequences like deportation that have “terrific impact” but that

do not “directly flow[]” from a judgment, and we suggested that defendants have no

constitutional right to be apprised by counsel of such collateral consequences before

entering a guilty plea, no matter how “surprised” they may be by those consequences.

212 F.2d at 921–22. There, a panel majority denied the vacatur motion, seeing no basis to

hold that “defendants are subjected to manifest injustice, if held to their plea, merely

because they did not understand or foresee such collateral consequences.” 7 But the

Supreme Court’s ruling in Padilla surely supervenes the Parrino majority’s ruling;

indeed, the facts are nearly identical, and as discussed above, Padilla was unequivocal in

holding that deportation cannot be classified as either a “direct” or “collateral”

consequence of a conviction. Padilla, 559 U.S. at 365–66. The Majority’s reasoning from

6 The Majority also cites an out-of-circuit decision that spoke broadly about different types
of collateral consequences, but that decision, too, was abrogated by Padilla. See United States v.
Del Rosario, 902 F.2d 55, 59 (D.C. Cir. 1990), abrogated by Padilla, 559 U.S. 356.

7 The dissenting judge in Parrino would have granted relief, and he described the sentence as
follows: “For all practical purposes, the court sentenced [Parrino] to serve (a) two years in jail
and (b) the rest of his life in exile.” 212 F.2d at 924 (Frank, J., dissenting).

12
our Circuit’s dated precedent therefore rests on a much shakier foundation than it

perhaps acknowledges.

Further, while the Majority takes issue with my observation that many courts

have responded to Padilla by “reexamining their Sixth Amendment precedents,” the

decisions by “several state supreme courts and a circuit court” that the Majority relies

on to undermine this observation, Maj. Op. at 11, serve only to reinforce the point that,

while the outcome may not be foretold, careful examination is now appropriate. Citing

Reeves, 695 F.3d at 640, for example, the Majority asserts that it “align[s] [itself] with the

Seventh Circuit in . . . affirming the [direct/collateral] distinction’s threshold

applicability to the Sixth Amendment.” Maj. Op. at 10. But in Reeves, the Seventh Circuit

carefully considered, and rejected, a defendant’s ineffective assistance claim based on

his counsel’s failure to advise him that a guilty plea to one crime could trigger statutory

sentencing enhancements after conviction for a second, later crime. 695 F.3d at 639–41.

Before applying the direct/collateral framework, the court—prompted by Padilla—

considered whether the framework applied at all: it compared the characteristics of

sentencing enhancement to the “unique” and “automatic consequence” of deportation

and concluded that its precedent addressing the counsel’s duty to advise on an issue of

future punishment for a second independent crime was not disturbed by Padilla. Id. at

639–40 (citing Lewis v. United States, 902 F.2d 576, 577 (7th Cir. 1990)). The circumstances

and consequences presented there are far removed from those faced by Farhane.

The decision in Taylor v. State, 887 N.W.2d 821 (Minn. 2016), also cited by the

Majority, similarly gives scant support for its view. In Taylor, the Minnesota Supreme

Court considered whether warning of the post-conviction requirement to register as a

predatory offender was part of the effective assistance required by the Sixth

Amendment. Id. at 822. Much as the Seventh Circuit did in Reeves, the Taylor court first

reexamined its pre-Padilla precedent in which it had “held . . . that the requirement to

register as a predatory offender is a collateral consequence” not encompassed by the

13
duty to advise. Id. at 823–24 (citing Kaiser v. State, 641 N.W.2d 900, 907 (Minn. 2002)).

After methodically comparing the features of predatory-offender registration to

deportation, it concluded that, “because deportation is a more severe consequence than

predatory-offender registration, . . . failure to advise a defendant about predatory-

offender-registration requirements” does not violate a defendant’s Sixth Amendment

right to effective assistance. Id. at 824–26. Tellingly, it did not simply label the

registration requirement as collateral and dismiss it as therefore unworthy of a

freestanding Sixth Amendment analysis.

Finally, in its 2014 decision in State v. Trotter, the Supreme Court of Utah

carefully considered whether, in light of Padilla, “Utah’s sex offender registration

requirement is sufficiently akin to deportation such that the direct-collateral divide is

‘ill-suited’ to dispose of Mr. Trotter’s claims.” 330 P.3d 1267, 1272 (Utah 2014). Just as

the Reeves and Taylor courts had done, the Trotter court carefully compared a sex

offender registration requirement to deportation, both in its severity and in its

relationship to the criminal justice system. Id. at 1272–75. After extensively analyzing

the effects of those requirements under the Padilla Court’s reasoning, it concluded that

the state’s registration requirement was a separate, civil penalty; that it should not be

removed from the “generally applicable direct-collateral dichotomy”; and that it did not

carry with it an independent constitutional obligation to advise. Id.

To be clear, I do not claim that Padilla or Chaidez, seismic though they may have

been, eradicated entirely the direct/collateral framework in the Sixth Amendment

context. The dichotomy will continue to apply usefully in some—even many or most—

cases, as demonstrated by the differing outcomes in the cases discussed in this section.

But, regardless of their outcome, what all of these decisions show is that after Padilla,

courts have abandoned automatic application of the direct/collateral framework to

resolve ineffective assistance claims. Especially in the absence of any precedent from

our court that squarely places denaturalization in the “collateral consequences”

14
category, I cannot join the Majority in so summarily applying the framework to dismiss

Farhane’s claim here. Nor, after studying the Majority’s analysis, can I join in its result.

3. The Sixth Amendment effective assistance obligations of counsel are broader
than the Fifth Amendment Due Process obligations of courts accepting a
guilty plea.

Third, the Majority errs by relying on inapt precedents concerning a district

court’s obligation to ensure that a guilty plea is voluntary and intelligent under the Fifth

Amendment’s Due Process Clause and Federal Rule of Criminal Procedure 11. My

colleagues suggest that in United States v. Youngs, 687 F.3d 56 (2d Cir. 2012), we squarely

held that the blunt categorical distinction between direct and collateral consequences

remains good law for all purposes, Fifth and Sixth Amendment duties alike, even after

Padilla, except in the sole case of deportation risk. I disagree. That was not the question

before the Youngs panel, and it could not have so held. Nor, contrary to the Majority’s

implication, have courts generally found to be coextensive the Fifth Amendment

obligations of a court accepting a guilty plea and the Sixth Amendment obligations of

counsel to give effective assistance.

The Youngs Court considered a Fifth Amendment challenge to the court’s actions

when accepting a guilty plea that entailed a possibility of eventual civil commitment. 8

See Youngs, 687 F.3d at 58–63. The Court “conclude[d] that advising of the possibility of

8 Youngs pleaded guilty in 2008 to production and possession of child pornography. Youngs,
687 F.3d at 58. During the plea hearing, the court reviewed details of the plea agreement,
including possible prison terms and supervised release provisions. Id. On direct appeal, Youngs
argued that his due process rights were violated because the court failed to inform him that his
guilty plea subjected him to possible civil commitment under the Adam Walsh Child Protection
and Safety Act of 2006, 18 U.S.C. § 4248(a). Youngs, 687 F.3d at 58–59. Under the Act, the
Attorney General or the Director of the Bureau of Prisons (“BOP”) could certify an individual in
BOP custody as a “sexually dangerous person.” Id. If the government then was able to show by
clear and convincing evidence, at a subsequent district court hearing, that the person was
“sexually dangerous,” the person would be committed to further custody until the court
determined otherwise. Id.

15
civil commitment . . . does not fall within the scope of a district court’s [Fifth

Amendment] due process obligations because the concerns expressed by the Supreme

Court in Padilla as to deportation in the context of adequate counsel under the Sixth

Amendment do not apply to such a remote and uncertain consequence as civil

commitment.” Id. at 62. In so concluding, we went to some lengths to distinguish the

court’s Fifth Amendment Due Process obligations from counsel’s Sixth Amendment

obligations of counsel, explaining, “These Sixth Amendment responsibilities of counsel

to advise of the advantages and disadvantages of a guilty plea are greater than the

responsibilities of a court under the Fifth Amendment.” Id. While discussing the general

categories of collateral and direct consequences, the panel further “recognize[d] that

Padilla may create some uncertainty as to the usefulness of categorizing certain

consequences as either ‘direct’ or ‘collateral,’” even in the Fifth Amendment context. Id.

(emphasis in original). 9

9 We are not alone in recognizing that counsel’s Sixth Amendment obligations are more
extensive than the court’s Fifth Amendment due process obligations. See, e.g., People v. Hughes,
983 N.E.2d 439, 450–57 (Ill. 2012) (determining that a court’s failure to inform a criminal
defendant of the possibility of civil commitment could not give rise to a Fifth Amendment due
process claim, but counsel’s failure to advise of the same possibility might give rise to a Sixth
Amendment claim). In Hughes, the Illinois Supreme Court interpreted Padilla to require
consideration of the severity of an otherwise “collateral consequence” before “categorically
exclud[ing it] from a cognizable claim of ineffective assistance of counsel and a defendant’s
sixth amendment rights.” Id. at 455. In concluding that involuntary commitment was a
sufficiently severe consequence, the court highlighted scholarship arguing for a broader reading
of counsel’s role in advising on consequences of a guilty plea than that ordinarily assigned to
the district court under either Rule 11 or the Fifth Amendment. Id. at 453–54 (endorsing the
observation of several commentators that “counsel’s role encompasses a broader range of
considerations” than the court’s role in advising about collateral consequences in the plea
process); see, e.g., Chin & Holmes, supra note 4, at 730 (“There is good reason to doubt that the
duties and conduct of courts and defense lawyers should be regarded as identical in this [duty
of advisement] context.”); Margaret Colgate Love, Collateral Consequences After Padilla v.
Kentucky: From Punishment to Regulation, 31 ST. LOUIS U. PUB. L. REV. 87, 100 (2011) (“The
considerations that make the direct/collateral distinction sensible from the standpoint of
institutional competence when applied to a court, do not apply to criminal defense lawyers’
relationships with their clients.”); Roberts, supra note 3, at 696–97 (explaining that “importation

16
 The Majority implicitly acknowledges this limitation on Youngs’s holding, see

Maj. Op. at 15 n.47, but still asserts generally that, in its view, the Fifth Amendment due

process obligation of the court and Sixth Amendment effective assistance of counsel

contexts are “closely analogous,” id. at 6 n.6, and “closely related,” id. at 15. And

doubtless there is important overlap between the court’s Fifth Amendment obligations

and counsel’s Sixth Amendment obligations; the analysis of one might inform the other.

It strikes me as incorrect, however, as it did the Youngs court, to act on an apparent

assumption that in truth an identical standard governs both constitutional protections.

Cf. Michel v. United States, 507 F.2d 461, 466 (2d Cir. 1974) (“Defense counsel is in a much

better position to ascertain the personal circumstances of his client so as to determine

what indirect consequences the guilty plea may trigger.”). Thus, in light of the different

governing standards, we held in Youngs that “the Padilla Court’s unwillingness to apply

the direct/collateral distinction in the Sixth Amendment context does not demonstrate

the Court’s intention to do away with that distinction entirely in the Fifth Amendment

context.” Youngs, 687 F.3d at 62; see also United States v. Delgado-Ramos, 635 F.3d 1237,

1241 (9th Cir. 2011) (“While Padilla’s holding is directly applicable to our Sixth

Amendment analysis . . . , it sheds no light on the obligations a district court may have

under Rule 11 and due process.”). Accordingly, I cannot agree with the Majority’s

contention that Youngs directs us to apply the direct/collateral distinction in the Sixth

Amendment context, or to do so in all cases except those on all fours with Padilla.

* * *

While I worry that the Majority’s broad language risks foreclosing future Sixth

Amendment challenges based on failures to advise as to other assertedly “collateral”

of the due process-based collateral-consequences rule into the ineffective assistance realm is
highly problematic because it treats the roles of defense counsel and the trial judge as identical”
even though “[t]he judge and defense counsel play very different roles with respect to a person
pleading guilty in a criminal case”).

17
consequences, 10 in my view the Majority’s primary error lies in its reluctance to apply

Padilla’s reasoning step-by-step, methodically, to the specific risk of denaturalization,

deferring instead to a broad-brush application of a traditional direct/collateral

distinction. As detailed above, the Majority has not identified any authority—from our

Court or otherwise—that compels continued application of the distinction in all cases

not leading directly to immediate and automatic deportation. Contrary to the Majority’s

contention, cases since Padilla—including those the Majority relies on—do, indeed,

reflect “a widespread reconsideration of the previously settled law.” See Maj. Op. at 11.

In light of Padilla, we should carefully evaluate whether the direct/collateral distinction

properly applies to denaturalization in the context of Sixth Amendment claims of

ineffective assistance. In doing so, we should also consider how developments in the

criminal justice system—including the well-known, widespread, and dominant use of

plea bargaining 11 and the growing prevalence and severity of consequences of a

10These other “collateral” consequences, too, can be harsh—and more severe than the
“direct” punishment imposed for a conviction. See Chin & Holmes, supra note 4, at 699–700
(describing how “the imposition of collateral consequences has become an increasingly central
purpose of the modern criminal process,” in many cases “[t]he real work of the conviction is
performed by the collateral consequences,” and “traditional sanctions such as fine[s] or
imprisonment are comparatively insignificant”).

11 As of 2009—shortly after Farhane pleaded guilty—guilty pleas represented ninety-seven
percent of federal convictions and ninety-four percent of state convictions. See Missouri v. Frye,
566 U.S. 134, 143 (2012). In contrast, until the 1970s, “15% to 20% of federal defendants went to
trial. But as soon as mandatory minimums and mandatory guidelines took effect in the late
1970s and early 1980s, the percentage began to rapidly decrease: by 2000 only 5% of all federal
defendants (reportedly even a smaller percentage of state defendants) went to trial. In 2015,
only 2.9% of federal defendants went to trial . . . .” Jed S. Rakoff, Why Prosecutors Rule the
Criminal Justice System—and What Can Be Done About It, 111 NW. U. L. REV. 1429, 1432 (2017). The
Supreme Court thus recognized, “[t]he reality is that plea bargains have become so central to
the administration of the criminal justice system that defense counsel have responsibilities in
the plea bargain process, responsibilities that must be met to render the adequate assistance of
counsel that the Sixth Amendment requires in the criminal process at critical stages.” Frye, 566
U.S. at 143.

18
conviction, 12 including denaturalization—bear on counsel’s constitutional

responsibilities to advise or seek advice on such consequences and to discuss them with

his client before the client enters a guilty plea.

Under Padilla, the direct/collateral dichotomy is ill suited to
denaturalization.

This Court should follow the courts that have responded to Padilla by

reexamining the automatic application of the direct/collateral dichotomy to resolve

Sixth Amendment ineffective assistance claims. At the very least, we should now

consider carefully whether denaturalization is sufficiently severe and “intimately

related to the criminal process” such that Padilla’s reasoning precludes the summary

conclusion that denaturalization is a merely “collateral” consequence of Farhane’s

guilty plea as to which no pre-plea advice was due. See 559 U.S. at 365.

As emphasized above, in Padilla, the Court held that “[t]he collateral versus

direct distinction is . . . ill suited to evaluating a Strickland claim concerning the specific

risk of deportation.” 559 U.S. at 366 (emphasis added). The Court’s holding rested on

deportation’s “particularly severe” nature, and developments in immigration law that

had made removal “nearly an automatic result” of guilty pleas to certain offenses “for a

12 See, e.g., Lynn Adelman, Criminal Justice Reform: The Present Moment, 2015 WIS. L. REV. 181,
183 (2015) (describing how, in the 1960s, “[c]ollateral consequences were less numerous, less
severe, and not as difficult to avoid or mitigate”); Roberts, supra note 3, at 700 (noting that “[t]he
Brady decision [in 1970] came well before the current reality of widespread, harsh collateral
consequences,” and arguing that “[t]he collateral-consequences rule is outdated for three
interrelated reasons: (1) the rise in the percentage of criminal prosecutions that are resolved by
guilty plea; (2) increased prosecution of minor offenses; and (3) the rise in the number and
severity of collateral consequences of criminal convictions”); ABA Standards for Criminal Justice:
Pleas of Guilty, Commentary to Standard 14-3.2(f), at 126 (3d ed. 1999) (“An increasing burden
must fall to defense counsel by virtue of the growing number and range of consequences of
conviction.”).

19
broad class of noncitizen offenders,” highlighting how “enmeshed criminal convictions

and the penalty of deportation” had become. Id. at 365–66.

Considering the same factors, it is apparent that the direct/collateral dichotomy is

likewise a poor fit for evaluating a Sixth Amendment claim concerning the risk of

denaturalization: that is, complete loss of U.S. citizenship. First, both the government

and the Majority agree with Farhane that denaturalization is a severe penalty, with

effects comparable to those arising from deportation. Maj. Op. at 12–13. 13 The Supreme

Court “has long recognized the plain fact that to deprive a person of his American

citizenship is an extraordinarily severe penalty.” Klapprott, 335 U.S. at 612. And, for

naturalized citizens like Farhane, denaturalization typically presages deportation with

the separation from family and other deep losses that entails.

Second, for naturalized citizens who plead guilty to criminal conduct predating

their naturalization, that plea carries with it an automatic and heightened risk of

denaturalization. Denaturalization, like deportation, is thus “enmeshed” in the criminal

justice system. See Padilla, 559 U.S. at 365–66. The civil denaturalization statute directs

revocation of citizenship on the ground that the naturalization was “illegally procured”

or “procured by concealment of a material fact or by willful misrepresentation.” 8

U.S.C. § 1451(a). 14 A naturalized citizen has illegally procured citizenship and concealed

13 Denaturalization revokes citizenship conferred by a court and returns the individual to
the immigration status held pre-naturalization effective as of the original naturalization date.
U.S. Citizenship & Immigration Servs., 12 Policy Manual, Part L—Revocation of Naturalization,
Ch. 3—Effects of Revocation of Naturalization, available at https://www.uscis.gov/policy-
manual/volume-12-part-l-chapter-3 (current as of July 31, 2023). It entails loss of the right to
vote and a new, extreme vulnerability to deportation. As noted further below, infra note 17, it
can also mean the loss of citizenship for children who derive citizenship from the denaturalized
parent. Indeed, in Farhane’s case, those two of his children whose citizenship is derived from
his naturalization will “be deemed to have lost and to lose [their] citizenship” as of Farhane’s
denaturalization. 8 U.S.C. § 1451(d).

14Further highlighting denaturalization’s relationship to the criminal justice system,
Farhane’s plea made him vulnerable to a charge of naturalization fraud under 18 U.S.C. § 1425.

20
a material fact when he does not disclose illegal conduct occurring before his citizenship

application but naturalizes despite that nondisclosure. 15 See Kungys v. United States, 485

U.S. 759, 767 (1988). And, as described above, a defendant in a denaturalization

proceeding may not contest the factual underpinnings of the government’s case if it is

based on conduct to which the defendant has already pleaded guilty. See Maietta, 84

F.3d at 102 n.1. Then, because courts “lack equitable discretion to refrain from entering

Upon a conviction under that statute, 8 U.S.C. § 1451 provides that the court of conviction “shall
thereupon revoke, set aside, and declare void the final order admitting such person to citizenship,
and shall declare the certificate of naturalization of such person to be canceled.” See 8 U.S.C.
§ 1451(a), (e) (emphasis added).

Notwithstanding this automatic relationship between criminal penalty and criminal
denaturalization, at least some government entities have recommended civil denaturalization
under section 1451(a) as “the most effective remedy” for the government after a naturalized
citizen’s criminal conviction because of “the broader scope of actions warranting [civil]
denaturalization,” and the “constitutional and statutory limitations inherent in [criminal]
actions.” Anthony D. Bianco et al., Civil Denaturalization: Safeguarding the Integrity of U.S.
Citizenship, 65 U.S. Attys’ Bull. 5, 8 (July 2017). In civil denaturalization proceedings, for
example, the government need only show “clear, unequivocal, and convincing” proof of the
alleged conduct—a lighter burden than “establish[ing] the offense and its elements beyond a
reasonable doubt,” as required in a criminal prosecution. Id. The government may also prefer
civil proceedings because “many of the due process protections afforded in a criminal
proceeding, such as a jury trial and a right to counsel, are not mandated,” and because no
statute of limitations applies. Id. The overarching point is that a guilty plea like Farhane’s
triggers serious exposure to denaturalization under at least two different approaches available
to the government.

15 Accordingly, the standard citizenship application form asks all applicants, “Have you
EVER committed . . . a crime or offense for which you were NOT arrested?” Maslenjak v. United
States, 582 U.S. 335, 346 (2017) (bold in original). An admission to having engaged in many
types of criminal conduct automatically precludes an applicant from demonstrating the “good
moral character” required for naturalization. See 8 C.F.R. § 316.10(b)(2)(iv), (b)(3)(iii). Thus, a
naturalized citizen who pleads guilty to any such pre-naturalization conduct will necessarily
have illegally procured his citizenship and failed to disclose that conduct when applying for
citizenship. See 8 U.S.C. §§ 1427(a), 1451(a). That guilty plea therefore makes the citizen
vulnerable to the denaturalization statute, which provides, “It shall be the duty of” the
government “to institute proceedings . . . for the purpose of revoking and setting aside”
naturalization in such circumstances. See id. § 1451(a).

21
a judgment of denaturalization against a naturalized citizen whose citizenship was

procured illegally or by willful misrepresentation of material facts,” Fedorenko, 449 U.S.

at 517, a serious risk of denaturalization is virtually automatic once the citizen has

pleaded guilty to pre-naturalization illegal conduct.

Such severe and nearly automatic consequences of a guilty plea are exemplified

by Farhane’s circumstances. Through his guilty plea, Farhane admitted that beginning

in November 2001, he conspired to commit money laundering in violation of 18 U.S.C.

§ 371. 16 Accordingly, in a denaturalization proceeding, he will have effectively admitted

that he illegally procured his citizenship based on his criminal conduct occurring

shortly before his 2002 naturalization, and that he concealed a material fact or made a

willful misrepresentation regarding that criminal conduct. See 8 U.S.C. § 1451(a). And,

as already noted, the court adjudicating the government’s denaturalization petition will

have “no discretion to excuse the conduct.” Fedorenko, 449 U.S. at 517. Farhane will

therefore almost inevitably lose his U.S. citizenship, and two of his children will then

lose theirs. 17 See 8 U.S.C. § 1451(d).

16 So far as the parties have advised and our examination has revealed, the record contains
no evidence that the conspiracy Farhane pleaded guilty to—while serious and deserving of the
carceral sentence he received—resulted in the actual transfer of any funds. At sentencing,
Farhane’s counsel stated, “eventually this whole thing petered out. Nothing, the government
concedes, was transferred, no money or equipment was transferred to anyone. Then the
government kept tabs on Mr. Farhane for four years. They sent other people to investigate him,
other [confidential informants,] . . . to see if he would be willing to engage in conduct, and
nothing happened . . . .” J. App’x at 283. And, although he was sentenced to thirteen years’
imprisonment, Farhane was released three years early in light of his record of good behavior.

17 The governing statute provides that where an individual is denaturalized based on
concealment of a material fact or willful misrepresentation in procuring naturalization—as
would be the case with Farhane—any person claiming citizenship through such denaturalized
person “shall be deemed to have lost . . . citizenship.” 8 U.S.C. 1451(d); see also Bianco et al.,
supra note 14, at 16 (“If the defendant’s spouse or children obtained citizenship based on the
defendant’s naturalization, the denaturalization judgment revokes the spouse’s and children’s
naturalization as a matter of law.”).

22
 The Majority’s stated reasons for finding that Farhane received effective

assistance of counsel despite the absence of a warning as to these highly probable

results are not persuasive. The Majority does not dispute that denaturalization is a

likely and severe consequence of Farhane’s guilty plea, nor that Farhane’s deportation is

highly probable after his denaturalization. Instead, the Majority states that

denaturalization lacks the “automatic” relationship to the guilty plea that the Padilla

Court attributed to deportation, and that this attenuation relieves counsel of the

obligation to call the risk of the consequence to a defendant’s attention. Maj. Op. at 14.

But the Majority’s argument that denaturalization is not sufficiently automatic is

not convincing. The Majority observes that “[c]ivil denaturalization is a separate

proceeding that may or may not occur following the plea.” Maj. Op. at 15. But the same

is true, of course, of deportation proceedings that follow entry of a guilty plea. Padilla,

559 U.S. at 365 (acknowledging that removal is “civil in nature” and “not, in a strict

sense, a criminal sanction”). Similarly, the Majority points out that denaturalization can

occur without a criminal conviction. Again, the same is true of deportation. See, e.g., 8

U.S.C § 1227(a)(1), (3)–(6) (providing numerous grounds for removal not predicated on

a criminal conviction). 18 In focusing on these mistaken distinctions, the Majority

overlooks a fundamental similarity: many convictions create a “nearly . . . automatic”

risk of denaturalization, just as they do for deportation, because the applicable statutory

and regulatory provisions establish that an applicant is ineligible for naturalization

based on the criminal conduct admitted to in a guilty plea. See Padilla, 559 U.S. at 366. 19

18As earlier noted, supra note 14, denaturalization is closely connected to the criminal
process also because the government may pursue criminal denaturalization charges against
defendants who unlawfully procure their naturalizations. See 18 U.S.C. § 1425.

19 As discussed above, specific types of criminal conduct render an individual ineligible for
naturalization. See supra at 21 & n.15. The fact that one must consult other statutory provisions
to determine whether prior criminal conduct renders naturalization “illegally procured,” 8
U.S.C. § 1451(a), does not make the operation of these provisions any less automatic following a

23
 The Majority also asserts that “[t]he government exercises considerable

discretion in bringing denaturalization cases,” implying that the absence of certainty,

too, excuses counsel’s failure to advise of the risk. Maj. Op. at 15. But the Supreme

Court decided Padilla as it did while recognizing “the equitable discretion vested in the

Attorney General to cancel removal . . . .” See 559 U.S. at 364. Moreover, pathways to

avoid removal remain even for noncitizens convicted of offenses like Padilla’s,

including through prosecutorial restraint. See, e.g., Ortiz v. Lynch, 640 F. App’x 42, 44–45

(2d Cir. 2016) (summary order) (referencing memorandum from Department of

Homeland Security directing prosecutorial discretion in pursuing removals, including

of certain noncitizens convicted of aggravated felonies); State v. Shata, 868 N.W.2d 93,

108–10 (Wis. 2015) (recognizing the executive branch’s “essentially unreviewable

prosecutorial discretion with respect to commencing deportation proceedings”). It is for

those reasons, perhaps, that the Court in Padilla did not speak in absolute terms: it

found that constitutionally effective counsel had an obligation to warn that a “risk of

deportation” flowed “nearly” automatically from a guilty plea. 559 U.S. at 366 (emphasis

added). The defendant there was only “subject to automatic deportation,” id. at 360

guilty plea to such conduct. Further, there may be room for argument—and litigation—
regarding whether a particular criminal conviction provides a basis for denaturalization. But the
same is true, of course, for deportation. We need only look to a sampling of the cases this court
has adjudicated involving complex applications of the “categorical approach” to recognize that
whether a particular conviction renders a person deportable is often not cut-and-dried. See, e.g.,
Debique v. Garland, 58 F.4th 676, 680–85 (2d Cir. 2023) (applying categorical approach to
determine whether sexual abuse of a minor was an “aggravated felony” and “a crime of child
abuse” that made petitioner removable); Jack v. Barr, 966 F.3d 95, 97–99 (2d Cir. 2020) (assessing
whether New York convictions for possession and sale of firearms rendered petitioner
removable); Hylton v. Sessions, 897 F.3d 57, 60–63 (2d Cir. 2018) (same for state conviction for
sale of marijuana). The potential for debate about whether a particular conviction actually
renders a noncitizen deportable does not affect whether, under Padilla, counsel is required to
advise that a guilty plea carries a “risk of deportation.” 559 U.S. at 374. The same is true with
respect to whether a particular guilty plea carries with it a risk of denaturalization.

24
(emphasis added); in other words, he would be deportable, but it did not necessarily

follow that he would automatically be deported. In the same way, a defendant whose

guilty plea exposes him to denaturalization is subject to denaturalization; that does not

necessarily mean that he will automatically be denaturalized.

Finally, the Majority asserts that the district court has “considerable discretion

. . . in evaluating the evidence” in a denaturalization proceeding. Maj. Op. at 15–16. As

already emphasized, however, courts lack discretion to reevaluate evidence established

by an earlier guilty plea, Maietta, 84 F.3d at 102 n.1, and they lack any equitable

discretion to deny the government’s application to revoke the citizenship of a citizen

subject to denaturalization, Fedorenko, 449 U.S. at 517. Similarly, courts are generally

understood to lack equitable discretion to deny a government application to remove a

noncitizen who is subject to deportation. See Padilla, 559 U.S. at 362–64. As Padilla

recognized, it is this lack of judicial discretion that creates a nearly automatic risk of

deportation—and the same is true for denaturalization proceedings based on a prior

criminal conduct.

In sum, the Majority’s reasoning does not justify its view that denaturalization

lacks the “‘automatic’ relationship to the guilty plea” that the Padilla Court saw in

deportation. Maj. Op. at 14; see 559 U.S. at 366. Considering both denaturalization’s

severity and its similarly automatic nature in cases such as Farhane’s, it is my view that

the “collateral versus direct distinction” is “ill suited to evaluating a Strickland claim

concerning the specific risk of” denaturalization. Padilla, 559 U.S. at 366.

The direct/collateral dichotomy does not apply for the additional reason
that Farhane’s plea exposes him to a substantial risk of deportation.

The Majority’s conclusion that the direct/collateral framework applies and bars

consideration of Farhane’s Sixth Amendment claim is also wrong for a separate, albeit

25
related, reason: Farhane faces a substantial “risk of deportation” based on his guilty

plea. Id.

As detailed above, Farhane’s guilty plea exposes him to a tremendous risk of

denaturalization: The government has already initiated denaturalization proceedings

against him, using the admissions in his guilty plea as the foundation of its complaint.

But once his citizenship is revoked, Farhane will be subject to removal as a noncitizen

convicted of an aggravated felony. See 8 U.S.C. §§ 1101(a)(43)(D), (U), 1227(a)(2)(A)(iii);

Dimaya, 138 S. Ct. at 1211 (“[R]emoval is a virtual certainty for an alien found to have an

aggravated felony conviction, no matter how long he has previously resided here.”).

The record leaves no room for doubt that the government intends to pursue such

removal proceedings. As a 2017 Department of Justice bulletin explained, “Typically,

the government does not expend resources on civil denaturalization actions unless the

ultimate goal is the removal of the defendant from the United States. [DOJ] attorneys

confirm that goal before filing the complaint.” Anthony D. Bianco et al., Civil

Denaturalization: Safeguarding the Integrity of U.S. Citizenship, 65 U.S. Attys’ Bull. 5, 17

(July 2017). 20 While the government may offer—as it did to Farhane—to “negotiate

terms of a settlement of a denaturalization case,” those terms “will not include any

promise of relief from removal.” Id. Therefore, although Farhane must be denaturalized

before he can be deported, he nonetheless plainly faces a substantial “risk of

deportation” based on his guilty plea. Padilla, 559 U.S. at 366. His counsel failed to

advise him of that risk in addition to the risk of denaturalization. Accordingly, under

Padilla, Farhane’s claim is not subject to the direct/collateral dichotomy and cannot be

20 I note in passing that Anthony D. Bianco is the lead author of the quoted article in the DOJ
bulletin, which advises of likely deportation after denaturalization. Mr. Bianco has appeared as
government counsel on the pending denaturalization complaint against Farhane.

26
barred as a merely collateral consequence of his plea; rather, it is governed by Strickland.

Id.

The Majority counters that “[b]ecause denaturalization is collateral, everything

thereafter . . . is also collateral.” Maj. Op. at 17. The Majority’s assertion cannot be

squared with Padilla’s holding, which—the Court there emphasized time and again—

applies to the need for advice regarding a client’s “risk of deportation.” Padilla, 559 U.S.

at 366–67, 374; accord Chaidez, 568 U.S. at 344 (“In [Padilla], this Court held that the Sixth

Amendment requires an attorney for a criminal defendant to provide advice about the

risk of deportation arising from a guilty plea.”). I do not mean to suggest that effective

counsel must warn of every consequence that could eventually lead to deportation, no

matter how attenuated the connection between that consequence and deportation. But

as described above, the connection between denaturalization and deportation is

obvious and explicit. The Majority’s argument might be more persuasive if Padilla

applied only to cases in which a defendant faces a certainty of immediate deportation,

but that was not Padilla’s holding. Nor did Padilla limit its application to claims brought

by noncitizen defendants; it described its conclusion as applying broadly to any

“client”—without differentiation—whose plea carries a “risk of deportation.” 559 U.S.

at 374 (“[W]e now hold that counsel must inform her client whether his plea carries a

risk of deportation.”).

Indeed, it seems paradoxical to construe Padilla to provide stronger protection to

a noncitizen at risk of deportation (like Padilla) than to a U.S. citizen at risk of

denaturalization followed by deportation (like Farhane). Had Farhane been a lawful

permanent resident when he pleaded guilty, rather than a naturalized citizen, it would

be settled by Padilla that Strickland governed his Sixth Amendment ineffective assistance

claim. It is difficult to fault the argument put forward by Farhane and amici that it

makes little sense for the Sixth Amendment to “provide more protection to a noncitizen

who was given misadvice, or no advice at all, about the deportation consequences of a

27
guilty plea, than to a naturalized citizen who was similarly not warned about the

possible loss of citizenship, to be followed by deportation.” Appellant’s Reply Br. at 12;

see Br. of Amicus Curiae Immigrant Defense Project at 22 (“[T]he logic of Padilla must

apply equally, if not more forcefully, to defendants who face the risk of losing their

citizenship and being removed from this country.”); Br. of Amici Curiae National

Association of Criminal Defense Lawyers et al. at 13 (“It hardly makes sense to say that

counsel has a Sixth Amendment duty to inform her client whether his plea carries a risk

of deportation, except when the government first has to strip the client of citizenship.”

(internal quotation marks omitted)).

The Majority dismisses this argument as “compar[ing] apples to oranges,” Maj.

Op. at 18. They point out, correctly, that Congress created different statutory schemes

for denaturalization and deportation. But no statute compels the result that the Majority

reaches. The Sixth Amendment interpretation and the direct/collateral framework that

the Majority invokes represent judge-made doctrines. Particularly in light of the

Supreme Court’s directive to consider the merits of a Sixth Amendment claim when a

guilty plea carries a “risk of deportation,” Padilla, 559 U.S. at 366, 374, we should take a

more flexible, less rote approach.

***

For the foregoing reasons, I would decline to apply the direct/collateral

distinction here and would not dismiss Farhane’s claim as concerning a merely

collateral consequence of his plea. Instead, I would proceed to address the merits of his

claim that he received ineffective assistance of counsel under the Sixth Amendment.

II. Farhane received objectively unreasonable representation.

Turning, then, to the merits, I conclude that Farhane received objectively

unreasonable assistance from his trial counsel. The failure to advise him as to the risks

28
of denaturalization and deportation were objectively unreasonable in 2006, when he

was counseled to plead guilty, and it would be objectively unreasonably now, in 2023.

To establish a claim for ineffective assistance of counsel, a defendant must show

(1) “that counsel’s representation fell below an objective standard of reasonableness,”

and (2) that “any deficiencies in counsel’s performance [were] prejudicial to the

defense.” Strickland, 466 U.S. at 688, 692. A defendant satisfies the first prong by

showing that his counsel’s performance “falls outside the ‘wide range of professionally

competent assistance.’” Kovacs v. United States, 744 F.3d 44, 50 (2d Cir. 2014) (quoting

Strickland, 466 U.S. at 690). To determine objective reasonableness, the Supreme Court

instructs us to evaluate “prevailing professional norms” at the time of the

representation, using “American Bar Association [“ABA”] standards and the like [as]

guides to determining what is reasonable.” Padilla, 559 U.S. at 366 (internal quotation

marks omitted).

Prevailing professional norms in 2006 required attorneys to advise clients
about the risk of denaturalization and deportation.

Farhane argues that the performance of his trial counsel (an individual not

representing him in this appeal) was objectively unreasonable because counsel failed to

advise him of the likely denaturalization and deportation resulting from his guilty plea.

I agree. Padilla may have “mark[ed] a major upheaval in Sixth Amendment law” by

applying Strickland to the particular ineffective assistance claim at issue there, id. at 383

(Alito, J., concurring). But under prevailing professional norms in 2006—just as in 2002,

when Padilla received advice on his plea, id. at 367–68—counsel’s failure to advise

Farhane about the immigration consequences of his guilty plea, or indeed even to flag

any possible adverse immigration consequences, was objectively unreasonable, in two

ways.

29
 First, as reviewed above, Farhane’s guilty plea exposes him to a substantial risk

of deportation. As the Court detailed in Padilla, “[t]he weight of prevailing professional

norms supports the view that counsel must advise her client”—naturalized citizen or

noncitizen—“regarding the risk of deportation.” Id. at 367. The Court emphasized that

“[f]or at least the past 15 years”—that is, since at least 1995—“professional norms have

generally imposed an obligation on counsel to provide advice on the deportation

consequences of a client’s plea.” 21 Id. at 372. The Court therefore held that “counsel

must inform her client whether his plea carries a risk of deportation,” and that failure to

do so constitutes objectively unreasonable assistance. Id. at 374. Although for Farhane,

denaturalization is a predicate step to deportation, the two are tightly linked, as

demonstrated above: when a naturalized citizen faces a risk of denaturalization, he will

almost always face an accompanying risk of deportation. Indeed, because of his guilty

plea, Farhane is now exposed to a substantial risk of deportation. Accepting Farhane’s

allegations as true, his trial counsel did not warn him of this potential consequence.

Accordingly, by the time Farhane pleaded guilty in 2006, counsel’s performance fell

outside the “wide range of professionally competent assistance.” Strickland, 466 U.S. at

690. As already observed, to hold otherwise creates the nonsensical result that the Sixth

Amendment provides less protection to naturalized citizens at risk of deportation than

to noncitizens facing the same potential consequence.

21 In Chaidez, the Supreme Court held that the rule announced in Padilla does not apply
retroactively to claims brought by defendants whose convictions became final before Padilla was
decided. Chaidez, 568 U.S. at 347. There is no dispute that Farhane may rely on Padilla in his
habeas petition, however, because his conviction did not become final until May 2011, after
Padilla was decided. Additionally, although the Supreme Court decided Padilla in 2010, the
defendant in the underlying criminal case, Jose Padilla, was counseled and entered his guilty
plea in 2002—four years before Farhane pleaded guilty. See Padilla v. Commonwealth, 381 S.W.3d at
324. The Court’s analysis of the prevailing professional norms with respect to immigration
consequences therefore applies with as much force in the context of Farhane’s representation as
Padilla’s.

30
 Second, and more broadly, professional norms prevailing in 2006 when Farhane

pleaded guilty required defense counsel to advise clients about any substantial

immigration consequences that follow guilty pleas, including denaturalization. When

Farhane entered his guilty plea, the ABA Standards for Criminal Justice provided—in

the commentary to the same section and version referenced by the Supreme Court in

Padilla, 559 U.S. at 367:

[C]ounsel should interview the client to determine what collateral
consequences are likely to be important to a client given the client’s
particular personal circumstances and the charges the client faces. . . . [I]t
may well be that many clients’ greatest potential difficulty, and greatest
priority, will be the immigration consequences of a conviction. To reflect
this reality, counsel should be familiar with the basic immigration
consequences that flow from different types of guilty pleas, and should
keep this in mind in investigating law and fact and advising the client.

ABA Standards for Criminal Justice: Pleas of Guilty, Commentary to Standard 14-3.2(f), at

127 (3d ed. 1999).

Likewise, as of 2005, the New York State Bar Association stated that defense

counsel should “[o]btain[] all available information concerning the client’s background

and circumstances for purposes of . . . avoiding, if at all possible, collateral

consequences including but not limited to deportation” and should also “[p]rovid[e] the

client with full information concerning such matters as . . . immigration . . . and other

collateral consequences under all possible eventualities.” NYSBA Standards for Providing

Mandated Representation, Standard I-7(a), (e), at 16–17 (2005).

Another resource cited by the Supreme Court in Padilla, 559 U.S. at 367, similarly

explained that, even as of 1997, “defense counsel should advise the defendant of . . . [a]ll

of the consequences and ramifications of a particular plea, including . . . effects on . . .

immigration status.” G. Nicholas Herman, Plea Bargaining § 3:03, at 20–21 (1997). And a

2005 publication by the New York State Defenders Association’s Immigrant Defense

Project warned defense counsel to be cognizant that “the government can attempt to

31
take away the citizenship of a naturalized citizen” who then “may again be vulnerable

to deportation” based on a criminal conviction. New York State Defenders Ass’n

Immigrant Defense Project et al., Deportation 101, at 21 (Feb. 2005); see also Br. of Amicus

Curiae Immigrant Defense Project at 13 (“At the time of Mr. Farhane’s case, long-

established professional norms required defense counsel to advise a client regarding the

immigration consequences of a plea.”).

The government does not meaningfully dispute that, by 2006, professional norms

required defense counsel to provide advice on immigration consequences of guilty

pleas. Instead, the government argues that because several of the resources cited above

refer to “immigration consequences generally,” without “specifically discuss[ing]

denaturalization,” those resources do not support the contention that a defense attorney

attuned to such “general” consequences would be “aware of the possibility of civil

denaturalization, which, unlike every other immigration consequence, affects United

States citizens.” Appellee’s Br. at 32–33.

The government’s argument is unpersuasive. Denaturalization is undoubtedly a

severe immigration consequence, as the government itself recognizes; it is also an

important element of the government’s program of enforcing immigration laws

generally. See Bianco et al., 65 U.S. Attys’ Bull. at 5 (quoting Attorney General Jefferson

B. Sessions III’s statement that DOJ “will aggressively pursue denaturalization . . . to

strategically enforce the nation’s immigration laws” (omission in original)); id. at 6

(“Actions to revoke naturalization unlawfully obtained or obtained by fraud are an

integral part of the government’s arsenal of remedies to enforce the immigration

laws . . . .”). It affects only people who immigrated to the United States, and it is

intended to lead to removal from the country, making it at least as severe as

deportation. See Klapprott, 335 U.S. at 616–17 (Rutledge, J., concurring) (commenting

that “[t]o take away a man’s citizenship deprives him of a right no less precious than

life or liberty” and that “in its wake may follow the most cruel penalty of banishment”).

32
 Further, the denaturalization laws are neither obscure nor difficult to

understand. By the time Farhane faced criminal charges, the statute authorizing

denaturalization had not been materially amended in more than forty years. Compare 8

U.S.C. § 1451(a) (as most recently amended in 1994) with Immigration and Nationality

Act, Pub. L. No. 414, § 340, 66 Stat. 163, 260 (1952) and Act of Sept. 26, 1961, Pub. L. No.

87-301, § 18, 75 Stat. 650, 656. With respect to the grounds for denaturalization, the

statute is “succinct, clear, and explicit.” 22 Padilla, 559 U.S. at 368. And the statute was at

issue in numerous published decisions involving civil denaturalization, see, e.g., Kungys,

485 U.S. at 763–64, including decisions of our Court in the period shortly before Farhane

pleaded guilty, see, e.g., United States v. Reimer, 356 F.3d 456, 457 (2d Cir. 2004). Indeed,

the government’s briefing implicitly recognizes as much. See Appellee’s Br. at 14 (citing

statutory provisions and caselaw preceding Farhane’s plea related to civil

denaturalization actions).

22 Section 1451(a) provides in relevant part:

It shall be the duty of the United States attorneys for the respective districts,
upon affidavit showing good cause therefor, to institute proceedings in any
district court of the United States in the judicial district in which the
naturalized citizen may reside at the time of bringing suit, for the purpose of
revoking and setting aside the order admitting such person to citizenship and
canceling the certificate of naturalization on the ground that such order and
certificate of naturalization were illegally procured or were procured by concealment
of a material fact or by willful misrepresentation, and such revocation and setting
aside of the order admitting such person to citizenship and such canceling of
certificate of naturalization shall be effective as of the original date of the order
and certificate, respectively.

8 U.S.C. § 1451(a) (emphasis added). Even minimal research into the grounds for
denaturalization identified in the statute and Farhane’s circumstances would reveal the several
potential bases for deportation that Farhane would be exposed to based on his guilty plea, as
discussed elsewhere. See supra note 15.

33
 As these authorities demonstrate, the potential consequences of a guilty plea for

a naturalized American’s citizenship status and ability to remain in the country were

well known when counsel advised Farhane to plead guilty to criminal conduct

occurring before his naturalization. That some contemporaneous practice guides do not

specifically mention denaturalization does not absolve counsel of the general

responsibility to, at the very least, advise the client “that pending criminal charges may

carry a risk of adverse immigration consequences.” Padilla, 559 U.S. at 369. This duty

exists even if defense counsel, who need not be “well versed” in immigration law, sees

the potential consequences as “unclear or uncertain.” Id. Accordingly, I find

unpersuasive the government’s contention that the professional norms prevailing in

2007 and requiring defense counsel to advise clients of potential adverse immigration

consequences did not apply to the consequence of denaturalization.

This broader interpretation of defense counsel’s duty is consistent with how the

Supreme Court and our Court, both, have characterized Padilla’s holding. See, e.g.,

Chaidez, 568 U.S. at 353 (“Padilla . . . made the Strickland test operative . . . when a

criminal lawyer gives (or fails to give) advice about immigration consequences.”); Doe v.

United States, 915 F.3d 905, 910 (2d Cir. 2019) (characterizing Padilla as holding “that

attorneys must affirmatively warn their clients of the immigration consequences of their

potential convictions”); Sutherland v. Holder, 769 F.3d 144, 147 (2d Cir. 2014) (per curiam)

(“Padilla . . . held that an attorney is ineffective for failing to advise a client of the

immigration consequences of a guilty plea.”). My interpretation also accords with other

decisions suggesting that denaturalization is among the potential “immigration

consequences” of which defendants should be made aware before entering a guilty

plea. See Rodriguez v. United States, 730 F. App’x 39, 42 (2d Cir. 2018) (summary order)

(finding that “counsel’s apparent advice . . . that [a client] did not have to worry about

the immigration consequences of a plea ignored the possibility of denaturalization” and

34
therefore fell below an objective standard of reasonableness); 23 cf. United States v. Ataya,

884 F.3d 318, 326 (6th Cir. 2018) (vacating a guilty plea, on plain error review, in part

because “[a] plea colloquy that does not put the defendant on notice that pleading

guilty will expose him to the loss of his American citizenship harms the fundamental

purpose of the judicial proceeding”).

In short, by the time Farhane entered his guilty plea in 2006, and as recognized

by the Padilla Court as to advice given in 2002, well-established prevailing professional

norms guided that defense counsel should advise a client that a plea agreement could

carry substantial immigration consequences generally—consequences not limited to

deportation. These norms weigh strongly in favor of the conclusion that Farhane

received objectively unreasonable representation. 24

23 The government and the district court attempt to distinguish Rodriguez on the ground that
Rodriguez involved affirmative misadvice from an attorney. That distinction is unavailing,
however, for the reasons explained by the Supreme Court in rejecting the same argument in
Padilla:

A holding limited to affirmative misadvice would invite two absurd results. First,
it would give counsel an incentive to remain silent on matters of great importance,
even when answers are readily available. Silence under these circumstances would
be fundamentally at odds with the critical obligation of counsel to advise the client
of the advantages and disadvantages of a plea agreement. . . . Second, it would
deny a class of clients least able to represent themselves the most rudimentary
advice on deportation even when it is readily available.

559 U.S. at 370–71 (internal quotation marks omitted). These considerations weigh just as
strongly with respect to advice regarding denaturalization, and the government does not offer a
persuasive argument in favor of upholding a distinction between misadvice and failure to
advise in the denaturalization context.

24Nor are these norm-setting materials properly disregarded as merely “aspirations of a bar
group.” Contra Conc. Op. at 1. The Supreme Court in Padilla, while acknowledging that such
guides are not “inexorable commands,” understood them as “valuable measures of the
prevailing professional norms of effective representation” in 2002, Padilla, 559 U.S. at 367; I see
no reason not to do the same here.

35
 Defense counsel should have known that Farhane’s guilty plea put him at
risk of denaturalization and deportation.

The government also advances the view that, as the district court held, “nothing

in the record suggests his lawyer knew, or should have known, about the circumstances

giving rise to Mr. Farhane’s denaturalization exposure,” suggesting a concomitant

absence of a duty to advise. Appellee’s Br. at 32 (quoting United States v. Farhane, No.

05-cr-673-4 (LAP), No. 18-cv-11973 (LAP), 2020 WL 1527768, at *2 (S.D.N.Y. Mar. 31,

2020)); see also id. at 29 (“As the District Court found, Farhane’s counsel had no reason to

believe that Farhane faced a denaturalization risk.”). The government’s argument does

not hold up, however, on the facts and under the prevailing professional norms

discussed above.

For the reasons I have discussed, prevailing norms in 2006 called for trial counsel

to advise Farhane of the “basic immigration consequences” that could flow from his

criminal charges. ABA Standards for Criminal Justice: Pleas of Guilty, Commentary to

Standard 14-3.2(f), at 127. As the ABA explained in 1999, “counsel has a duty to conduct

a sufficient investigation to understand the unique issues that confront each client”;

counsel should “interview the client to determine” which “consequences are likely to be

important . . . given the client’s particular personal circumstances and the charges the

client faces.” Id. at 120, 127; see Br. of Amicus Curiae Immigrant Defense Project at 14–16

(collecting authorities suggesting that defense counsel has “an affirmative duty to

investigate the immigration consequences of a criminal case for all clients who are not

born in the United States”).

Here, the record reflects that trial counsel had all the information necessary to

prompt an inquiry into whether severe immigration consequences could result from

Farhane’s guilty plea. As of at least the initial detention hearing on November 2, 2005—

about one year before Farhane entered his guilty plea—counsel’s own statements

reflected awareness that Farhane was an immigrant and a naturalized citizen. See J.

36
App’x at 128 (counsel stating that Farhane is “an immigrant. Yes. He’s a naturalized

United States citizen.”). Judge Walker explains in his concurrence that, in his view,

counsel did not act unreasonably in failing to advise Farhane about the risks of

denaturalization (and likely deportation) because the record does not establish his

knowledge as to “when Farhane was naturalized in relation to the crimes to which

Farhane later pleaded guilty.” Conc. Op. at 2. But if counsel was unaware of the timing,

he could have easily learned this information by asking a single question of his client—a

question he likely should have asked, given his knowledge of Farhane’s status as a

naturalized citizen. See NYSBA Standards for Providing Mandated Representation, Standard

I-7(a), (e), at 16–17 (2005) (stating that defense counsel should inquire into the client’s

background to advise regarding potential collateral consequences, “including but not

limited to deportation”). Upon inquiring into the timing of Farhane’s naturalization,

even superficial research into the statutes and caselaw regarding denaturalization

would have revealed the denaturalization and deportation risks associated with entry

of a guilty plea. 25

Counsel had another reason, too, to be on notice of the potential denaturalization

consequences of Farhane’s guilty plea: one amicus represents, and the government does

not appear to contest, that “[a]n above-average number of civil denaturalization cases

were filed in 2001 and 2002 and were attributable to prosecution trends in the wake of

the events of September 11.” Br. of Amicus Curiae Immigrant Defense Project at 7–8. 26

There can be no doubt that, by the start of the sentencing phase, counsel should have
25

known that the conduct underlying Farhane’s guilty plea predated his naturalization: counsel’s
sentencing memorandum references the paragraph of the presentence investigation report
including the statement that Farhane became a naturalized citizen on April 19, 2002, and further
notes that part of Farhane’s offense conduct dated to “late 2001.” J. App’x at 206.

26But see Irina D. Manta & Cassandra B. Robertson, Inalienable Citizenship, 99 N.C. L. REV.
1425, 1438 (2021) (asserting that “even the period after 9/11 did not involve a spike [in
denaturalization proceedings] after leaders of both political parties opposed proposals to use
denaturalization in the fight against terrorism”). The authors also note, however, in discussing

37
The potential risk of denaturalization was thus particularly acute for Farhane, who was

indicted in the Southern District of New York on money-laundering charges arising out

of a terrorism investigation begun in the wake of the September 11 attacks.

Judge Walker and the government also urge that, “[g]iven the infrequency with

which the Government sought civil denaturalization, a criminal defense attorney who

was unaware of such a risk in 2006 cannot be said to have acted unreasonably.”

Appellee’s Br. at 34; see also Conc. Op. at 2 (describing denaturalization as “extremely

rare” in 2006). But neither cites any authority to support the proposition that

representation is not constitutionally deficient simply because a severe outcome occurs

in a relatively small fraction of all proceedings. 27 To the contrary, counsel has an

Farhane’s criminal case, that “in the months after 9/11, the government ramped up its law
enforcement activities in Muslim communities to identify individuals who might be plotting
terrorist attacks or assisting those interested in doing so.” Id. at 1451.

27 Although data show that the government pursued relatively few denaturalization actions
between 1968 and 2012, “[o]ver the last decade, the federal government has mounted a new
concerted campaign to increase the use of denaturalization to revoke the citizenship of foreign-
born U.S. citizens . . . .” Amber Qureshi, The Denaturalization Consequences of Guilty Pleas, 130
YALE L.J.F. 166, 170, 173 (2020); see also Cassandra B. Robertson & Irina D. Manta, (Un)civil
Denaturalization, 94 N.Y.U. L. REV. 402, 409–14 (2019) (remarking on the trend toward increased
civil denaturalization proceedings for foreign-born U.S. citizens). In the first two years of
President Trump’s administration, the government “filed twice as many denaturalization cases
in each of [those] years as the average number of denaturalization cases for the prior twelve
years.” Qureshi, supra, at 173; see Amanda Frost, Alienating Citizens, 114 NW. U. L. REV. 241, 245
(2019) (describing how the Obama administration’s “investigation into a limited number of
naturalization files, which it dubbed ‘Operation Janus,’” was “escalated” by the Trump
administration “into an investigation of hundreds of thousands of naturalized citizens for errors
in the naturalization process”). An Amicus advises that, as denaturalization proceedings have
become increasingly frequent, they have also begun to be “instituted against individuals who
committed less serious crimes” than those giving rise to the typical denaturalization proceeding
in the past. Br. for Amicus Curiae Asian Americans Advancing Justice at 23. These recent trends
do not affect the analysis of whether trial counsel provided objectively reasonable
representation to Farhane in 2006, true. They do underscore, however, the need for defense
counsel today to advise clients of the risk of denaturalization. This need renders even more

38
obligation to understand and advise on the unique issues that confront each client. See

ABA Standards for Criminal Justice: Pleas of Guilty, Commentary to Standard 14-3.2(f), at

120, 127. The authorities discussed above support the conclusion that, aware of

Farhane’s status as a naturalized citizen, trial counsel was obligated to conduct at least a

minimal investigation of the relevant facts and to advise Farhane about the potential

denaturalization consequences of his guilty plea.

For these reasons, the government is incorrect in asserting that counsel had no

reason to know that the plea he advised Farhane to take would create a substantial risk

of denaturalization and deportation.

Farhane received objectively unreasonable assistance.

In light of the foregoing, defense counsel was obligated at the very least to advise

Farhane that the “pending criminal charges [against him] may carry a risk of adverse

immigration consequences,” including denaturalization and deportation. Padilla, 559

U.S. at 369. Farhane maintains—and the government does not contest—that his counsel

did not provide any such warning or other advice related to immigration consequences,

despite knowledge of Farhane’s status as a naturalized citizen. Farhane plausibly avers

that he would not have entered a guilty plea had he known that he could lose his U.S.

citizenship or face deportation as a result, explaining that a plea agreement that

“opened the door to loss of citizenship and deportation, and that exposed two of [his]

children to a similar risk, was contrary to [his] priorities.” J. App’x at 299. As the

Supreme Court observed in Padilla, avoiding those permanent and drastic consequences

“may be more important to the client than any potential jail sentence.” 559 U.S. at 368

(internal quotation marks omitted).

concerning the broad effects that the Majority’s categorical ruling—which appears to apply to
all denaturalization risks, past, present, and future—may have.

39
 Judge Walker adds that because Padilla was not decided until 2010—four years

after Farhane’s guilty plea—finding counsel’s conduct objectively unreasonable is to

“require[] an attorney to gaze into a crystal ball.” Conc. Op. at 3. But this misstates

Farhane’s burden in this appeal; counsel need not have predicted Padilla nor its

outcome to satisfy his Sixth Amendment obligations. The only relevant considerations

are first, that “Padilla . . . is applicable law in this case,” and second, that under such

law, for the reasons above, counsel’s conduct was objectively unreasonable when he

counseled Farhane to enter his guilty plea without advising that the plea could result in

loss of citizenship, and removal. Id.; see also supra note 21. Under these circumstances,

counsel’s failure to advise Farhane of the immigration consequences of his guilty plea

was objectively unreasonable. The district court therefore erred when it concluded that

Farhane failed to meet the first prong of the Strickland analysis.

To receive relief, however, Farhane must also satisfy Strickland’s second prong by

demonstrating that he was prejudiced by counsel’s performance. See 466 U.S. at 692.

The district court did not address this factor. Accordingly, in light of my conclusion that

counsel’s performance was objectively unreasonable, I would vacate the district court’s

denial of Farhane’s motion and remand to allow the district court to address the issue of

prejudice in the first instance. See, e.g., Padilla, 559 U.S. at 369 (“Whether Padilla is

entitled to relief on his claim will depend on whether he can satisfy Strickland’s second

prong, prejudice, a matter we leave to the Kentucky courts to consider in the first

instance.”); Rodriguez, 730 F. App’x at 44 (vacating the district court’s denial of relief and

“remand[ing] for the district court to develop a fuller record concerning the issue of

prejudice”).

CONCLUSION

Farhane’s guilty plea makes it virtually inevitable that he will lose his U.S.

citizenship and be removed from this country—“banish[ed]” from his home of nearly

40
thirty years. See DeGeorge, 341 U.S. at 232 (Jackson, J., dissenting). His children, too, face

the loss of their own citizenship. His trial counsel’s failure to advise him that his plea

agreement risked triggering these consequences constituted objectively unreasonable

performance under prevailing professional norms in 2006, as it would now.

Accordingly, the district court’s judgment should be vacated and the case remanded for

a determination, under Strickland, of whether Farhane was prejudiced by his counsel’s

constitutionally deficient performance. The Majority’s contrary conclusion cannot be

reconciled with the Supreme Court’s decision in Padilla.

I respectfully dissent.

41