People v. Peque

Chief Judge Lippman

(dissenting). I respond to the opinion subscribed to by three Judges, whom I refer to as the plurality, because that is the only writing offering reasons for the results announced in the above-captioned appeals. Although I would join a writing finding a due process entitlement on the part of a noncitizen defendant to be advised by the court of the possible immigration consequences of pleading guilty and making relief available when that entitlement is not honored, the plurality opinion does not meet the latter condition and I accordingly do not join it. I do, however, agree with the Judges who have signed the plurality opinion and with Judge Rivera, that deportation is such an important plea consequence that “it must be mentioned by the trial court to a defendant as a matter of fundamental fairness” (plurality op at 193).

The United States Supreme Court acknowledged in Padilla v Kentucky ( 559 US 356 [2010]) that “[d]eportation as a consequence of a criminal conviction is, because of its close connection to the criminal process, uniquely difficult to classify as either a direct or a collateral consequence” (id. at 366). The Court, accordingly, declined to use the direct/collateral distinction to ascertain whether deportation was a conviction consequence of which a pleading noncitizen defendant was required to be advised. Instead, the Court took note of certain realities whose crucial bearing upon a noncitizen’s decision whether to enter a plea of guilty were, by the time of the Court’s decision, undeniable. Prominent among these was that deportation had, since the mid-1990s, become for noncitizen defendants a virtually automatic consequence of convictions falling within several very broad penal categories, and that deportation was a particularly harsh superadded exaction—one that the Court did not shrink from referring to as a “penalty” ( 559 US at 364 ). Indeed, the Court had already recognized that deportation was a conviction consequence often more dreaded by noncitizen defendants than any prison sentence that might be imposed, either pursuant to a plea agreement or after trial (see 559 US at *208 368 , citing INS v St. Cyr, 533 US 289, 322 [2001]). In holding then that the constitutionally effective representation of a noncitizen contemplating the entry of a guilty plea required the provision of accurate advice as to the immigration consequences of the conviction that would ensue, the Court was driven by the recognition that a plea entailing deportation very often will be impossible to characterize as voluntary where that uniquely important consequence has not been disclosed to the defendant—that such pleas are categorically different from “the vast majority ... [in which] the overwhelming consideration for the defendant is whether he will be imprisoned and for how long” (People v Gravino, 14 NY3d 546, 559 [2010]).

The question now presented is whether, after Padilla , the description of deportation as a direct or a collateral plea consequence retains viability as a means of defining, not counsel’s, but the court’s duty in assuring the voluntariness of a plea. The plain answer to this question must be that it does not. If deportation is “uniquely difficult to classify as either a direct or a collateral consequence,” logically it is so for all purposes, not simply for the purpose of determining what advice counsel must give in satisfaction of the Sixth Amendment requirement of effective representation.

Once it is settled that the relevant inquiry is not whether deportation may be formally categorized as a direct or collateral consequence, but whether it is, as the Padilla Court observed, a consequence so certain, potentially pivotal and prevalent as to make its disclosure essential to assuring that the guilty plea of a noncitizen is knowing, intelligent and voluntary, it should be clear that the court’s allocutional obligations in taking a noncitizen’s plea are fully implicated. The realities shaping the court’s obligations, with respect to the conviction consequence of deportation, are not essentially different from those to which counsel must be responsive in advising a noncitizen defendant.

It is by now practically self-evident that the judicial obligation in taking a plea—i.e., assuring on the record that the defendant fully understands what the plea connotes and its consequences (see Boykin v Alabama, 395 US 238, 244 [1969]), or, in other words, that “the plea represents a voluntary and intelligent choice among the alternative courses of action open to the defendant” (No rth Carolina v Alford, 400 US 25, 31 [1970], citing Boykin, 395 US at 242 )—cannot realistically be met in the case of a noncitizen defendant unless the court’s canvass extends to ascertaining that the plea is made with the awareness that it may well result in the pleader’s deportation.

*209 The plurality, wisely, does not avoid this conclusion; to do so would, in a very large number of cases, be to reduce to a painfully obvious fiction the notion so favored by the law that the taking of a plea in open court serves as an effective procedural bulwark against an uninformed and thus involuntary surrender of basic constitutional protections to which the defendant would otherwise be entitled prior to any adjudication of guilt (see e.g. Brady v United States, 397 US 742 , 747 n 4 [1970] [“the record must affirmatively disclose that a defendant who pleaded guilty entered his plea understandingly and voluntarily”]; Boykin, 395 US at 242 [“prosecution (must) spread on the record the prerequisites of a valid waiver”]; Carnley v Cochran, 369 US 506, 516 [1962] [“The record must show . . . that an accused . . . intelligently and understandingly rejected (a constitutional right). Anything less is not waiver”]; and see People v Cornell, 16 NY3d 801, 802 [2011] [“due process requires that the record must be clear that the plea represents a voluntary and intelligent choice among . . . alternative courses of action” (internal quotation marks and citations omitted)]; People v Louree, 8 NY3d 541, 544-545 [2007]; see also plurality op at 184). Yet, while recognizing that the court has an independent due process obligation to notify a noncitizen defendant that his or her plea may result in deportation (plurality op at 176 [“We . . . hold that due process compels a trial court to apprise a defendant that, if the defendant is not an American citizen, he or she may be deported as a consequence of a guilty plea to a felony”])—a proposition with which I certainly agree—the plurality affords no remedy where that condition of due process has not been met, and in fact not one of the present appellants will in the end obtain relief.

If a plea proceeding fails of its essential purpose—if it does not create a record from which the knowing and voluntary nature of the defendant’s waiver and concomitant choice between available alternative courses of action may be readily understood—the plea is infirm. And, in that case, the appropriate response is to permit the plea’s withdrawal, not to cast about for a means of deeming the infirmity harmless (see McCarthy v United States, 394 US 459, 466 [1969] [“if a defendant’s guilty plea is not equally voluntary and knowing, it has been obtained in violation of due process and is therefore void”]). We have, in fact, permitted withdrawal as a matter of course where the defect in the plea amounts to a due process violation. In People v Catu ( 4 NY3d 242 [2005]), for example, we said:

*210 “Because a defendant pleading guilty to a determinate sentence must be aware of the postrelease supervision component of that sentence in order to knowingly, voluntarily and intelligently choose among alternative courses of action, the failure of a court to advise of postrelease supervision requires reversal of the conviction. The refusal of the trial court and Appellate Division to vacate defendant’s plea on the ground that he did not establish that he would have declined to plead guilty had he known of the postrelease supervision was therefore error (see also People v Coles, 62 NY2d 908, 910 [1984] [‘harmless error rules were designed to review trial verdicts and are difficult to apply to guilty pleas’]).

“In light of this result, we do not reach defendant’s alternative claim of ineffective assistance of counsel” (id. at 245 [emphasis supplied]).

The Court’s 1 present approach to dealing with a due process violation identical in kind to that addressed by Catu , although practically far more consequential, is precisely contrary to that deemed “require[d]” in Catu . The defendant’s remedy now is said to lie in a postconviction motion in which it will be up to him or her—often without the aid of counsel and in a non-native tongue 2 —to navigate the postconviction relief maze in order to prove a circumstance that should have been, but was not, negated by the accepted plea—namely, that the plea was entered in ignorance of its deportation consequence, which, if disclosed, would, with reasonable probability, have caused its rejection. In short, having demonstrably been denied due process, a defendant is, under today’s decision, relegated to a claim that reduces to one for ineffective assistance—a claim that would, in the vast majority of cases, have been obviated by a constitutionally adequate plea. It was, of course, in recognition of the primacy of the plea court’s due process obligation, that Catu premised the right to plea withdrawal exclusively on the plea court’s default, and consequently did not reach Catu’s ineffective assistance claim. While the plurality stresses that the judicial obligation in *211 taking a plea is independent of the obligation of counsel to provide accurate advice as to a plea’s immigration consequence (plurality op at 193), the net effect of its decision is remedially to telescope the two, so that a due process claim based on a judicial default will not occasion relief except where there is also an attendant meritorious Padilla claim. The plurality acknowledges that this is so but says that it is appropriate since in People v Harnett ( 16 NY3d 200 [2011]) it was observed in a purely theoretical aside that “the issue of whether the plea was voluntary may be closely linked to the question of whether a defendant received the effective assistance of counsel” (id. at 207). But the issue of whether a plea is actually voluntary, appropriately implicated in determining whether a plea should in fairness be vacated where the plea is not facially deficient—the circumstance to which the above-quoted language from Harnett speaks—is not the issue presented here. The issue posed in the present appeals is instead whether the plea itself comports with due process when its canvass does not extend to its immigration consequence. Having evidently held that it does not, it makes no sense at all to then require, as a condition of relief, that a defendant whose plea was facially deficient prove a negative— namely, that the due process denial was not harmless. Due process violations are presumptively prejudicial—that is why they are so classified. The accommodation of the contrary, illogical premise, could not have been within Harnett’s contemplation.

The delicacy with which the plurality treats People v Ford ( 86 NY2d 397 [1995])—a decision which, after Padilla , is in its two principal holdings, if not in its ratio decidendi, no longer viable—stands in strikingly awkward contrast to its abandonment of the remedial course charted in and required by Catu . Perhaps the plurality reasons that because deportation does not precisely fit the description of a direct conviction consequence and is, in its view “technically” a collateral consequence (plurality op at 192), that it is not governed by Catu . But this simply revives the direct/collateral distinction as a meaningful tool in characterizing deportation as a plea consequence. Not only is this use of the distinction demonstrably inapt after Padilla , it is utterly inconsistent with the plurality’s correct conclusion that due process requires the plea to establish that a noncitizen defendant was advised of its possible deportation consequence. If, in fact, it continues to be material—even after Padilla—that deportation is not, strictly speaking, a “direct” conviction consequence within the meaning of Ford , it should follow that a plea court’s nondisclosure of that consequence does not rise to *212 the level of a due process defect. But, that is a conclusion that the plurality, with ample empirical and legal justification, rightly eschews.

The plurality does not, however, eschew the remedial path hypothetically sketched in Gravino ( 14 NY3d at 559 ) and Harnett (16 NY3d at 207). Traveling it, however, is, as noted, inappropriate where the plea is affected by a due process deficiency such as the one the plurality identifies today. Plainly, the address of a due process violation was not what was intended when it was suggested in Gravino that a court might, as an “exercise [of] discretion” ( 14 NY3d at 559 ) vacate a plea if various conditions were met, among them that the defendant proved that, but for the nondisclosure of a consequence “of such great importance to him” (id.; and see Harnett, 16 NY3d at 207), he would not have pleaded guilty. The relief adverted to in Harnett and Gravino did not depend upon or respond to a default by the court in establishing the voluntariness of the plea; its purpose was rather to allow for a remedy precisely in those situations where the defendant was materially uninformed as to a plea consequence which, although of “great importance to him,” was not one about which the plea court was obliged to warn. In the cases before us, by contrast, we deal with judicial omissions incompatible with due process and bearing critically upon the very basis of the plea. The remedy in that latter circumstance is not “discretionary” as per Gravino’’s dicta, it is “required” as per Catu’s holding.

Today’s plurality decision speaks eloquently of the severity of deportation as a conviction consequence (plurality op at 192-193), but in the end treats removal as just another collateral consequence that may be of “great importance” to a defendant, leaving the defendant to prove to the satisfaction of the court that took the plea, that the plea was uninformed as to the important consequence, and that, had that consequence been disclosed, the plea would not have been entered—or, at least, that the plea’s rejection would have been reasonably probable. Thus, although the Court now roots the judicial obligation to inform a pleading noncitizen of immigration consequences in due process, as a practical matter judges and defendants remain just as they were—a judge’s default in informing a noncitizen defendant that he may be deported will only be rectified in the context of a claim for what is essentially ineffective assistance of counsel, which is to say in the context of a claim that, of course, already exists, but is extraordinarily difficult to make *213 out (see e.g. People v Hernandez, 22 NY3d 972 [2013] [no reasonable probability that a defendant with six young children in this country would have rejected a plea to preserve a possibility of avoiding deportation]). The disjunction between the right recognized and the remedy offered is palpable. If due process requires a warning “to protect the rights of the large number of noncitizen defendants pleading guilty to felonies in New York” (plurality op at 197), it must be that the failure to give the warning is at least presumptively prejudicial. Here, however, the plurality illogically and unfairly places upon the demonstrably unwarned members of the vulnerable noncitizen class the formidable burden of proving individual prejudice.

In advocating the conceptually straightforward and until now legally uncontroversial notion, that a guilty plea unequal to the basic due process purpose of demonstrating that its entry was knowing and voluntary should be permitted to be withdrawn, I acknowledge the inevitable concern that its embrace in the present context would provoke a stampede to the courthouse. That concern, rationally assessed, I believe is exaggerated. New York has required by statute, now for some 18 years, that judges warn noncitizens of their pleas’ potential immigration consequences (see CPL 220.50 [7]). It cannot be presumed that the statute has been pervasively ignored (see Padilla, 559 US at 372 [“For at least the past 15 years, professional norms have generally imposed an obligation on counsel to provide advice on the deportation consequences of a client’s plea . . . We should, therefore, presume that counsel satisfied their obligation to render competent advice at the time their clients considered pleading guilty”]). But, if it has been, that is all the more reason to doubt the efficacy of substituting one toothless command 3 for another, as the plurality today proposes. Nor is there reason to believe that noncitizen defendants will rush to scuttle pleas that were genuinely advantageous, notwithstanding an unallocuted deportation consequence (see Padilla, 559 US at 372-373 ). Moreover, inasmuch as Padilla broke new ground “by *214 breaching the previously chink-free wall between direct and collateral consequences” (Chaidez v United States, 568 US —, —, 133 S Ct 1103, 1110 [2013]), there is strong reason to suppose that any remedy stemming from the demise of that “chink-free wall” would be limited to cases still on direct appeal (see id.). Finally, in the long term, affording noncitizens prompt and effective relief from pleas that manifestly fail to provide the assurance of voluntariness that due process requires, will reduce rather than increase postconviction claims and thus protect rather than subvert the finality of plea-based judgments of conviction.

The conscientious provision of the already statutorily prescribed judicial warning—which all of the present appellants agree is adequate—would itself obviate the overwhelming majority of postconviction claims relating to undisclosed immigration consequences. And, in those presumably rare cases where, despite the remedy of plea withdrawal, there was a judicial default, all of the concerned parties would be spared complicated and prolonged motion practice; the defendant would simply, logically, fairly and expeditiously be given his or her plea back and proceed to trial on the indictment. I note that several jurisdictions have such a rule (see RI Gen Laws § 12-12-22 [c]; Cal Penal Code § 1016.5 [b]; Conn Gen Stat Ann § 54-1j [c]; DC Code § 16-713 [b]; Mass Gen Laws ch 278, § 29D; Ohio Rev Code Ann § 2943.031 [D]; Vt Stat Ann tit 13, § 6565 [c] [2]; Wash Rev Code § 10.40.200 [2]; Wis Stat Ann § 971.08 [2]); the sky has not fallen as a result.

The literal-minded application of the direct/collateral distinction, Padilla notwithstanding, has given rise to a state of affairs where a court must, on pain of reversal, inform a pleading defendant of a term of postrelease supervision (PRS) but may, without consequence, fail to disclose to the same defendant that the plea will result in deportation, an outcome not merely overshadowing but usually nullifying the term of postrelease supervision. 4 An analytic paradigm that would yield such an objectively skewed ordering of interests and corresponding *215 judicial concerns cannot and will not be viewed except as unmoored from the considerations of fundamental fairness that ought to animate our jurisprudence in passing upon pleas, the means by which guilt is established in the vast majority of criminal cases. Nothing in today’s very long plurality decision functions to diminish this signal anomaly one whit. Calling the court’s failure to advise of an immigration consequence a due process denial without affording the defendant a remedy for that denial amounts to no more than a verbal gesture. While the plurality insists that our precedents do not allow more, that is transparently incorrect. As noted, this Court has been clear as to the remedy required when a plea court fails to establish on the record, to the extent that due process requires, that a plea is a knowing and intelligent choice between available alternative courses of action. The notion, then, that the plurality is somehow constrained to withhold relief for the nonperformance of the “distinct” and “independent” judicial due process obligation it has postulated is altogether puzzling. It would be one thing if, like Judge Pigott, the plurality simply found that, Padilla notwithstanding, Ford remained good law for the proposition that judges have no due process duty to advise pleading noncitizens of immigration consequences. But, having found to the contrary, the failure to afford any logically and legally responsive remedy to noncitizen defendants left unwarned by the court as to the possible immigration consequences of their pleas represents a perplexing election—one that is in no way explained post -Padilla by clinging, practically as an article of faith, to an orthodoxy that, as the plurality opinion acknowledges at length, time and circumstance have overwhelmed, at least with respect to the characterization of immigration consequences for plea purposes.

Given the plurality’s indisposition to navigate the not so complicated route from its understanding of what due process requires of a court taking a plea, to a logical and efficacious remedy when the standard it has set has not been met—indeed, its evident determination instead to follow a tortuous path influenced by what is, in the present context, a thoroughly discredited formalism—a legislatively prescribed remedy will be necessary to untie the Gordian knot now fashioned and protect the adjudicative rights of noncitizen criminal defendants.

I would reverse in each of the cases before us.

*216 In Peque, although there was some fairly random mention of deportation at the sentencing proceeding (see n 4, supra), there was no judicial advisement at either plea or sentence as to the prospect of deportation, and Peque was manifestly confused as to what his plea involved. I do not, moreover, believe it reasonable to require preservation in this context. The purpose of the judicial advisement here at issue is to assure that the defendant is aware of the plea consequence. A preservation requirement presumes knowledge that would make the advisement unnecessary—a classic “Catch-22,” particularly inappropriate when dealing with the class of defendants “least able to represent themselves” )(Padilla, 559 US at 370-371 ) and where a meritorious claim for plea withdrawal—at least under the plurality formulation—presupposes that the defendant has been ineffectively represented. 5

The advisement provided in connection with defendant Diaz’s plea was, I believe, affirmatively misleading as to the likelihood of any immigration consequence and, on that ground, Diaz should be permitted to withdraw his plea and face trial on the indictment charging an A-I drug felony; if that is a risk he wishes to take to preserve the possibility of remaining in this country where he has resided legally for most of his life and has an infant child, he should be permitted to do so.

I would note in passing that Diaz’s case illustrates the extreme procedural difficulty of obtaining relief by the means now prescribed. Although the plurality acknowledges that the “trial court clearly failed to tell [Diaz] that he might be deported” (plurality op at 200), and purports to afford him the possibility of relief, it logically precludes him from prevailing in any ensuing litigation, since the showing of prejudice it requires has already been made and found wanting; Diaz’s CPL 440.10 motion was denied on the ground that he failed to satisfy the *217 Strickland prejudice prong, and leave to appeal was thereafter denied by an Appellate Division Justice. Like Diaz, all defendants alleging a due process violation by reason of an inadequate plea, in order to obtain relief, would, under today’s plurality decision, be compelled to split their claim between a direct appeal and a separate 440.10 proceeding—a complication that is pointless, since a defendant under current law, which the plurality does not alter in any practical respect, can in the end only obtain relief via a 440.10 claim for ineffective representation. Rather than temporize, I would afford Diaz actual relief from a plea that was not demonstrably knowing and voluntary.

Finally, as to defendant Thomas, inasmuch as his case is on direct appeal, I believe he is entitled to the benefit of our current jurisprudence. His postplea fraud upon the court logically has no bearing upon whether his plea was knowing, intelligent and voluntary, and there is no ground advanced by the plurality or the People to except from the rule that, ordinarily, a direct appeal from a judgment of conviction will be governed by the law as it exists at the time the appeal is decided (see People v Jean-Baptiste, 11 NY3d 539, 542 [2008])—a bright line demarcation we have adhered to, even where there has been lengthy delay attributable to the appellant (see e.g. People v Martinez, 20 NY3d 971 [2012]).

The People, I note, really do not identify any relevant prejudice traceable to defendant’s fabrication of his demise. Thomas has already served his enhanced sentence. Even if his plea withdrawal motion is granted, and the People are unable to reprosecute him for lack of witnesses or physical evidence, 6 he will have been amply punished for his criminal conduct and for his chicanery. There is nonetheless a real and persisting issue as to the validity the plea upon which this punishment was based, and, in that connection, the People can claim no vested interest in the application of outdated precedent, or, in other words, the retention of the pre-Padilla legal context. This is especially so since Padilla was remedial; it responded to circumstances existing long before its issue in 2010 (see Padilla, 559 US at 362-363 ). Indeed, by 1992, the year of defendant’s plea, deportation had become mandatory for noncitizens convicted of crimes falling into several broadly defined categories, one of which was for drug offenses; with a few closely drawn exceptions not applicable *218 cable to defendant, virtually all drug convictions by that time entailed automatic removal. 7

The People, whose interest properly lies not simply in winning this appeal but doing justice, can claim no prejudice from Padilla’s application to Thomas’s case. To the extent that the decision’s “new rule” retroactively applied may unduly impair the finality of convictions, that has been dealt with by the Supreme Court in Chaidez , which limits Padilla’s backward reach to cases that have not become final, i.e., those, like defendant’s, still on direct appeal (568 US at —, 133 S Ct at 1113 ).

In my view, Thomas’s right to relief is made out by the record of his plea proceeding at which, five days after his alleged wrongdoing and before being indicted, Thomas, then a 21-year-old novice to the criminal justice system, entered a plea to attempted criminal sale of a controlled substance in the third degree in exchange for a disarmingly attractive 30-day jail sentence without being advised by the court, or indeed by anyone present, that, upon his release, he would be deported. It is, I believe, clear that Thomas’s was not a knowing and voluntary plea.

. I refer here to the approach shared by the plurality and the dissenters for whom Judge Pigott has written, for as Judge Pigott has noted, those approaches are in their resolution practically indistinguishable. Neither affords noncitizen defendants relief from pleas that fail to establish the defendants’ awareness of their deportation consequence.

. We speak here of what Padilla , with doubtless accuracy, described as the “class . . . least able to represent themselves” ( 559 US at 370-371 ).

. After requiring that noncitizen defendants be warned as to the possible immigration consequences of their contemplated pleas, CPL 220.50 (7) adds the proviso that the failure to give the prescribed warning “shall not be deemed to affect the voluntariness of a plea of guilty.” That the plurality ultimately finds this proviso compatible with its notion of what due process avails a pleading defendant (plurality op at 199 n 12) is strikingly indicative of how very narrow its decision is.

. The nullifying effect of deportation upon a PRS term was, of course, the circumstance about which defendant Peque’s attorney wondered aloud, when the deportation issue surfaced at Peque’s sentencing. He said, “Mr. [Peque] is subject to deportation following the completion of his sentence. I’m not sure how that’s going to impact, assuming the Court imposes the sentence that’s been agreed upon, I’m not sure how that will affect the post-release supervision aspect of it.”

. In view of this latter circumstance, the utility of the plurality’s advice that a noncitizen defendant seeking plea withdrawal for nonadvisement as to an immigration consequence “should make every effort to develop an adequate record of the circumstances surrounding the plea at sentencing” (plurality op at 199) is dubious. If counsel has, by hypothesis, been ineffective it does not seem reasonable to expect the same attorney to make a record as to the very matter as to which the representation was deficient. If, as the plurality points out, it is not generally prudent to assume that “defense counsel ‘will’ do something simply because it is required of effective counsel” (Moncrieffe v Holder, 569 US —, —, 133 S Ct 1678, 1692 [2013]) (plurality op at 193), surely it cannot be prudent to suppose that ineffective counsel will do something because it is required of effective counsel.

. It is noted that while the People raise these impediments to reprosecution on appeal in a general way, they have never made any concrete allegation that they would be unable to proceed against defendant on the sale counts with which he was initially charged.

. As defendant observes, the immigration consequences of his plea to an attempted drug sale were dictated by 8 USC § 1251 (a) (2) (B) (i), a statute materially identical to its successor, 8 USC § 1227 (a) (2) (B) (i), the provision that applied to Padilla, and which was described by the Supreme Court as “succinct, clear, and explicit” ( 559 US at 368 ).