Argued January 2, 2013; decided February 12, 2013

People v Belliard, 89 AD3d 1443, affirmed.

Graffeo, J.

Penal Law § 70.25 (2-a) requires that a prison term imposed upon a second felony offender run consecutively to a previously imposed undischarged sentence. We hold that the consecutive nature of the new sentence is a collateral rather than direct consequence of a conviction in determining the adequacy of a plea allocution. [*2]

Following a buy and bust operation in 2006, defendant Rafael Belliard was arrested for possessing cocaine and a loaded firearm. He was subsequently charged with criminal possession of a controlled substance in the first and third degrees and criminal possession of a weapon in the second degree. In July 2007, defendant pleaded guilty to all three counts of the indictment. At the time of his guilty plea, defendant had a prior undischarged state sentence stemming from an earlier state felony drug conviction. And in committing the 2006 crimes, defendant was also facing a federal violation of supervised release in connection with two prior federal felonies.

During the July 2007 plea colloquy, the trial court explained that, as a second felony drug offender, defendant would receive a prison sentence of 12 years, followed by five years of postrelease supervision (PRS). Defense counsel asked the court to delay sentencing until the federal court had resolved defendant's violations of his supervised release, thereby permitting the new state sentence to be served concurrently with the federal sentence. The court agreed. No mention was made by defense counsel or the court as to whether the negotiated 12-year prison term would run concurrently or consecutively with the prior undischarged state sentence arising from his previous state drug conviction. Defendant allocuted to the charges and the court accepted his plea and adjourned his sentencing.

About two months later, defendant was sentenced to a determinate prison term of 12 years plus five years of PRS on{**20 NY3d at 384} the first-degree drug possession conviction, to be served concurrently with determinate terms of five years and five years of PRS on the third-degree drug possession and second-degree weapon possession convictions. The sentencing court further stated that the 12-year sentence would be concurrent to the federal sentence, as permitted by Penal Law § 70.25 (4).^[FN1] As a second felony drug offender, Penal Law § 70.25 (2-a) required that defendant's 12-year prison term run consecutively to his prior undischarged state sentence.^[FN2] The [*3]court, however, was silent on this subject.

Defendant appealed from the judgment of conviction, asserting that his guilty plea was involuntary and should be vacated because the trial court did not advise him of a consequence of his plea, namely, that the 12-year term of imprisonment would run consecutively to his prior undischarged state sentence. The Appellate Division rejected this contention and affirmed (89 AD3d 1443 [4th Dept 2011]). A Judge of this Court granted defendant leave to appeal (18 NY3d 955 [2012]), and we now affirm.

Defendant maintains—and the dissent agrees—that his plea must be vacated because the trial court neglected to inform him that the determinate term of 12 years imposed as a result of his plea bargain was to run consecutively to the undischarged portion of the sentence relating to the earlier state drug conviction. He seeks to analogize his situation to that in People v Catu (4 NY3d 242 [2005]), where we held that a mandatory PRS term is a direct consequence of a conviction and must be addressed by the court during a plea allocution. The People counter that the consecutive nature of defendant's sentence is a collateral{**20 NY3d at 385} consequence and the trial judge's failure to raise it does not undermine the validity of the plea.

A trial court is constitutionally required to ensure that a defendant, before entering a guilty plea, has a full understanding of what the plea entails and its consequences. The court is not obligated to engage in any particular litany during the plea colloquy, "but due process requires that the record must be clear that the plea represents a voluntary and intelligent choice among the alternative courses of action open to the defendant" (People v Ford, 86 NY2d 397, 403 [1995] [internal quotation marks and citations omitted]). Nevertheless, because it would be unfeasible for a court to advise a defendant of all the possible ramifications of a guilty plea, our cases have drawn a distinction between the direct and collateral consequences of a plea.

The significance of this distinction is that a trial court has a duty to advise a defendant about direct consequences, which are defined as having "a definite, immediate and largely automatic effect on defendant's punishment" (id.). A court's failure to comply with this mandate is not subject to harmless error review and requires reversal (see People v Harnett, 16 NY3d 200, 205 [2011]). In contrast, a court may, but need not, refer to collateral consequences during a plea allocution. A court's silence regarding collateral consequences "will not warrant vacating a plea because they are peculiar to the individual and generally result from the actions taken by agencies the court does not control" (Ford, 86 NY2d at 403). [*4]

In Catu, defendant pleaded guilty to various crimes and his sentence, by virtue of his status as a second felony offender, was required by statute to include a five-year period of PRS (see Penal Law § 70.45). The trial court, however, did not inform defendant of the period of supervision during the plea colloquy. We held that PRS is a direct consequence of a conviction as it constitutes a "component of [a] sentence" (Catu, 4 NY3d at 245) and, consequently, a defendant must be advised of the PRS term associated with the subject crimes. We further noted that, "[w]hereas the term of supervision to be imposed may vary depending on the degree of the crime and the defendant's criminal record, imposition of supervision is mandatory and thus has a definite, immediate and largely automatic effect on defendant's punishment" (id. at 244 [internal quotation marks and citation omitted]). As a result, we concluded that defendant was entitled to have his plea vacated.{**20 NY3d at 386}

Next, in People v Gravino (14 NY3d 546 [2010]) and its companion case, People v Ellsworth, we addressed whether mandatory registration under the Sex Offender Registration Act (SORA) and terms of probation are direct or collateral consequences of a plea. In Gravino, the trial judge did not notify defendant that she would have to register as a sex offender under SORA because of her rape conviction. And in Ellsworth, defendant was not told about the conditions of probation, including noncontact with his own children, before entering a guilty plea to a sex crime. On appeal, both defendants relied on Catu, asserting that SORA registration and probation conditions, like PRS, were direct consequences requiring that their pleas be set aside.

We rejected the proposed analogy between SORA and PRS in Gravino, stressing that PRS "is, by statute, a component element of a sentence, which is why a judge must pronounce the period of postrelease supervision at sentencing; it is thus an integral part of the punishment meted out upon a defendant's conviction of a crime" (id. at 556 [citation omitted]). In contrast, we noted that SORA is not a penal statute nor is the registration requirement a part of the sentence; rather, SORA is a remedial statute designed to prevent future criminal behavior. Although SORA unquestionably imposes significant burdens on a registrant, we determined that it is not a " 'direct consequence' of a conviction within the meaning of Ford as interpreted in Catu" (id. at 557). Similarly, in Ellsworth, we held that the conditions of probation are collateral consequences, reasoning that "courts taking guilty pleas cannot be expected to predict any and every potential condition of probation that might be recommended in the presentence report—an impossible task given the individualized nature of probation supervision" (id. at 558). The end result was that neither of these defendants was awarded relief.

We most recently dealt with the direct versus collateral framework in Harnett, where the trial judge did not apprise defendant that he might be subject to the Sex Offender Management and Treatment Act (SOMTA) as a result of his guilty plea to first-degree sexual abuse. Subsequently, defendant argued that the possibility of being held as a detained sex offender past his release date amounted to a direct consequence that he should have been notified [*5]about at his plea proceeding. We disagreed and affirmed the conviction, concluding that SOMTA consequences, while serious, are collateral. In so holding, we observed:{**20 NY3d at 387}

"The direct consequences of a plea—those whose omission from a plea colloquy makes the plea per se invalid—are essentially the core components of a defendant's sentence: a term of probation or imprisonment, a term of postrelease supervision, a fine. Our cases have identified no others" (Harnett, 16 NY3d at 205).

Finally, although not a plea vacatur case, People ex rel. Gill v Greene (12 NY3d 1 [2009], cert denied sub nom. Gill v Rock, 558 US 837 [2009]) also bears on our analysis of defendant's claim. In Gill, petitioner pleaded guilty as a second felony offender and was required by Penal Law § 70.25 (2-a)—the same provision at issue here—to have his sentence run consecutively with his prior undischarged sentences. The trial judge did not address this topic at the plea or sentencing stages of the proceedings and, lacking a court directive, the Department of Correctional Services (DOCS) administratively calculated the consecutive sentence. Relying on Matter of Garner v New York State Dept. of Correctional Servs. (10 NY3d 358 [2008]), where we held that only a court, not DOCS, may impose a term of PRS, petitioner brought a habeas corpus petition seeking to annul DOCS's determination that the sentences ran consecutively. We squarely rejected petitioner's contention and stated that, when a court is required by statute to impose a consecutive sentence but does not address the matter, it is deemed to have imposed the consecutive sentence.

In reaching this conclusion, we explained that the analogy petitioner proposed between consecutive sentencing and PRS was "flawed" because the problem in Garner "was that a part of the sentence—the PRS term—was never imposed" (Gill, 12 NY3d at 5-6). By comparison, we held that the Gill court did impose the sentence (an indeterminate term of 2½ to 5 years); all that was omitted was the characterization of the sentence as concurrent or consecutive. In language that is particularly relevant here, we remarked:

"Nothing in [Penal Law § 70.25 (2-a)] and nothing in the Constitution requires the sentencing court to say the word 'consecutive,' either orally or in writing. Nothing in the statute even requires that the sentencing court be made aware that the prior sentences are undischarged. Unlike the petitioners in Garner and Earley [v Murray (451 F3d 71 [2d Cir 2006])], who were told nothing about PRS by the{**20 NY3d at 388} courts that sentenced them, Gill was told in plain terms that he was being sentenced to 2½ to 5 years in prison. He was never given any reason to think that part or all of that sentence would be effectively nullified, by running simultaneously with sentences he had already [*6]received. Indeed, nothing in the record here shows the court knew that previous undischarged sentences existed" (id. at 6).

In this case, defendant similarly seeks to draw a parallel between PRS and consecutive sentencing, claiming that, like the mandatory PRS term in Catu, a sentence that must run consecutively to a prior undischarged sentence is a direct consequence of a plea. But this contention is undermined by Gravino and Harnett, where we clarified that the PRS term in Catu constituted a direct consequence because it was "a component element of a sentence" (Gravino, 14 NY3d at 556) or, stated otherwise, one of "the core components of a defendant's sentence" (Harnett, 16 NY3d at 205).^[FN3] Unlike PRS, the consecutive nature of a prison term pursuant to Penal Law § 70.25 (2-a) is not a component of the sentence itself, a point that was critical to our holding in Gill (see also Wilson v McGinnis, 413 F3d 196, 200 [2d Cir 2005] [rejecting the related argument that "mandatory consecutive state sentences . . . ha(ve) the effect of lengthening the actual state sentence and punishment that was imposed pursuant to the guilty plea" (internal quotation marks and emphasis omitted)]). And here, during defendant's plea colloquy, it is undisputed that he was informed of the core elements of his sentence when the trial judge stated that, as a second felony drug offender, he would receive 12 years' imprisonment to be followed by five years of PRS. In other words, he was fully advised of the direct consequences of his guilty plea.

Furthermore, as we emphasized in Gill, nothing in the statutory scheme requires a trial court to characterize a term of imprisonment as "consecutive," particularly since the judge may not even be aware that a defendant has time remaining on a previously-imposed sentence. Indeed, as in Gill, nothing in the record before us demonstrates that the trial judge knew defendant had a prior undischarged sentence. It also cannot be overlooked that defendant had no reasonable basis to believe that the 12-year term would run concurrently with his prior{**20 NY3d at 389} unrelated sentence, "effectively nullif[ying]" the new sentence (Gill, 12 NY3d at 6). Although the trial judge stated that the newly imposed sentence would be concurrent to the federal sentence—the sole issue of concern voiced by his counsel—no reference was made to the prior undischarged state sentence.

In sum, our holdings in Gravino, Harnett and Gill lead us to conclude that the consecutive nature of defendant's sentence pursuant to Penal Law § 70.25 (2-a) is a collateral consequence of his conviction. For the reasons stated, the failure of the trial court to address the impact of Penal Law § 70.25 (2-a) during the plea colloquy does not require vacatur of the plea. [*7]

Accordingly, the order of the Appellate Division should be affirmed.

Chief Judge Lippman (dissenting). The problematic quality of today's decision is evident from its opening sentences. The Court begins by acknowledging, as it must, that Penal Law § 70.25 (2-a) mandates, in the case of a second felony offender, the imposition of a prison term consecutive to one previously imposed but undischarged, yet then announces that a consecutive sentence imposed pursuant to Penal Law § 70.25 (2-a) is not a direct consequence of a second felony offender's plea. Nothing that follows in the decision's lengthy discussion of this Court's cases explains how it is that the subject legally mandated, automatically attaching, penalty enhancing sentence attribute may be characterized as anything other than a "direct" consequence of an underlying plea. It is, in fact, no less direct than the sentence itself—so direct that the sentencing court need not even mention it for it to take effect, or so this Court has held (People ex rel. Gill v Greene, 12 NY3d 1, 6 [2009] ["We read the words of Penal Law § 70.25 (2-a)—'the court must impose a sentence to run consecutively with respect to such undischarged sentence'—to mean that any sentence imposed by the court shall run consecutively to the undischarged sentence, whether the sentencing court says so or not"]). If it is the case, as the majority allows, that the court must advise a pleading defendant of his or her sentence because the sentence is a direct penal consequence of the plea, the question naturally arises as to how it is that the court is not also obliged to advise the defendant of the highly material circumstance that the sentence imposed must run consecutively to any undischarged sentence—a penal consequence entailed by and logically no less immediate than the sentence itself. The only answer offered by the Court is that the consecutive running of the sentence is not a "component" of the sentence.

{**20 NY3d at 390}It is true that in Gill the Court said that the consecutive running of a sentence pursuant to Penal Law § 70.25 (2-a) is not a "part of" the sentence, but merely a statutorily supplied sentence "characterization" (12 NY3d at 6). However, while this very fine semantic distinction may have been pertinent in defining the extent of the judicial obligation in pronouncing a sentence—the issue in Gill—it should count for very little in defining the prior obligation of the court to assure, as a condition of a plea's validity, that the defendant has been informed, not just of the sentence that will be pronounced in exchange for the plea, but of any other similarly direct plea consequence.

In People v Ford (86 NY2d 397, 403 [1995]) this Court described a direct consequence of a plea as one with "a definite, immediate and largely automatic effect on defendant's punishment." It is plain that this formulation, which we have employed repeatedly since Ford (see People v Catu, 4 NY3d 242, 244 [2005]; People v Gravino, 14 NY3d 546, 553-554 [2010]; People v Harnett, 16 NY3d 200, 205 [2011]), and to which the majority again purports to adhere, may encompass more than the "component" or "core elements" of the sentence. That the Catu court said that postrelease supervision should be deemed a direct [*8]consequence of a plea because it is a significant component of the sentence, is not logically equivalent to the much broader proposition the majority now embraces, that only a sentence component will qualify as a direct plea consequence. Nor does the Court's observation in Harnett, that our five decisions in the relevant area had not identified any direct plea consequences other than the "core components" of a defendant's sentence (16 NY3d at 205), amount to a reasoned argument that the "core components" of a sentence are exhaustive of the "definite, immediate and largely automatic effects" upon punishment to which Ford referred. Indeed, the very case from which Ford's formulation is borrowed, Cuthrell v Director, Patuxent Inst. (475 F2d 1364 [4th Cir 1973], cert denied 414 US 1005 [1973]), observes that the distinction between direct and collateral plea consequences is "sometimes shaded" (id. at 1366), and itself adverts to a plea consequence—i.e., ineligibility for parole—which had been widely viewed as a direct consequence even though it was not the sentence per se but, like the consequence here at issue, an entailed attribute of the sentence.

In any case, it does not seem arguable that the mandatory consecutive running of a sentence pursuant to Penal Law § 70.25 (2-a) corresponds precisely to Ford's description of a{**20 NY3d at 391} direct plea consequence. The effect of the statute's consecutive mandate is both punitive and definite. Penal Law § 70.25 (2-a) is patently directed at augmenting the length of time a defendant will spend behind bars; the consecutive "characteristic" of the sentence is nothing if not "an integral part of the punishment meted out upon a defendant's conviction of a crime" (People v Gravino, 14 NY3d at 556). And, although the effect of the statute's application will vary from case to case (see Catu, 4 NY3d at 245), it is a virtual certainty that in every case governed by Penal Law § 70.25 (2-a) there will be some penal consequence beyond that pronounced as part of the sentence proper. Nor can there be any question that the provision's effect on punishment is immediate and automatic. As noted, we have held that the statute will apply by operation of law in the case of every second felony offender with a prior undischarged sentence (Gill, 12 NY3d at 6).

Of course there is legitimate concern with requiring judges to advise defendants of consequences that will depend on their individual circumstances or that emanate from non-judicial agencies—circumstances of which a court may not be aware—and that is the reason for distinguishing between direct and collateral plea consequences; it is not that collateral plea consequences are necessarily less important to a defendant's decision as to whether to enter a plea. Indeed, it is now clear that a defendant's understanding of what would under the Ford dichotomy be deemed a collateral plea consequence may yet be highly material to a plea's constitutional validity (see Padilla v Kentucky, 559 US 356, —, 130 S Ct 1473, 1486 [2010]). The dichotomy's legitimate utility, then, lies not in gratuitously narrowing the court's scope of required inquiry, but simply in tailoring a judge's plea-vetting obligation to what he or she can reasonably be expected to know respecting a plea's material punitive consequences for the [*9]defendant.

Even while acknowledging this practically necessary limitation, it is exceedingly difficult to understand why it should operate to relieve a judge from informing a second felony offender tendering a plea that his or her sentence must run consecutively to any undischarged term previously imposed—in other words, that the period of incarceration resulting from the plea may actually be significantly longer than the stated term in the sentence. In the ordinary course of preplea proceedings, a judge will learn of a defendant's predicate status, and this case is not in that respect exceptional, as the majority suggests. Moreover,{**20 NY3d at 392} inasmuch as a defendant's predicate status is extant, documented information that is easily made available to the court—and is routinely furnished by the prosecution prior to sentencing—it is plain that there exists no considerable impediment to its production in connection with a plea.

It is the core object of every defendant entering a plea to minimize the punitive consequence of his or her conviction. Where, as here, a plea entails incarceration beyond that expressly agreed to as part of the bargained for sentence, it cannot be ascertainably knowing, intelligent and voluntary unless that entailment is first disclosed by the court. This conclusion I do not believe is soundly or prudently avoided by the analysis the Court advances today.

Judges Read, Smith and Pigott concur with Judge Graffeo; Chief Judge Lippman dissents in an opinion; Judge Rivera taking no part.

Order affirmed.

"When a person, who is subject to any undischarged term of imprisonment imposed at a previous time by a court of another jurisdiction, is sentenced to an additional term or terms of imprisonment by a court of this state, the sentence or sentences imposed by the court of this state . . . shall run either concurrently or consecutively with respect to such undischarged term in such manner as the court directs at the time of sentence."

Footnote 2: Penal Law § 70.25 (2-a) states in relevant part:

"When an indeterminate or determinate sentence of imprisonment is imposed pursuant to . . . subdivision three . . . of section 70.71 . . . and such person is subject to an undischarged indeterminate or determinate sentence of imprisonment imposed prior to the date on which the present crime was committed, the court must impose a sentence to run consecutively with respect to such undischarged sentence."

Footnote 3: The dissent disregards these passages from Gravino and Harnett.