Wesley, Circuit Judge

Robert Nowakowski , pro se, Brooklyn, NY.

Victor Barall , Assistant District Attorney (Leonard Joblove, Assistant District Attorney, on the brief), for Kenneth P. Thomson, District Attorney for Kings County, Brooklyn, NY, for Respondent–Appellee.

Before: Straub, Wesley, and Livingston, Circuit Judges.

Judge Livingston dissents in a separate opinion.

Wesley, Circuit Judge:

Petitioner–Appellant Robert Nowakowski was convicted of harassment in the second degree, an offense classified as a violation under state law, and sentenced to one year's conditional discharge, requiring one day of community service. Before completing this sentence, Nowakowski filed a petition for a writ of habeas corpus under 28 U.S.C. § 2254 in the United States District Court for the Eastern District of New York (Vitaliano, J. ). Because Nowakowski fulfilled the requirements of his sentence during the pendency of the habeas proceeding, the District Court concluded that Nowakowski's case presented no live case or controversy sufficient to establish Article III standing under Spencer v. Kemna , 523 U.S. 1, 118 S.Ct. 978, 140 L.Ed.2d 43 (1998).

We granted a certificate of appealability with instructions to brief two questions of first impression we now answer: First , whether a sentence of conditional discharge and one day's community service, unfulfilled as of the time of filing the habeas petition, satisfies the “in custody” requirement of § 2254. And second , whether a presumption of continuing collateral consequences applies to Nowakowski's conviction, thus presenting a live case or controversy under Article III despite the expiration of his sentence. Because we answer both questions in the affirmative, we VACATE the District Court's dismissal of Nowakowski's petition and REMAND for further proceedings consistent with this opinion. BACKGROUND

Robert Nowakowski was arrested on October 31, 2006, on charges of assault against another tenant in his building. He contends that these charges were fabricated by his landlord and his neighbors, including the now-deceased victim. After a bench trial, he was convicted of harassment in the second degree—which is classified as a violation under New York state law—in the Criminal Court of the City of New York on September 18, 2008, and ordered to pay a fine of $100. The sentence was stayed for over four years during post-conviction appeals and collateral proceedings. On May 14, 2013, the Criminal Court vacated the fine and sentenced Nowakowski to a one-year conditional discharge, requiring him to complete one day of community service within that time. This change in sentence occurred at Nowakowski's request because he could not afford the fine and administrative charges, which totaled $195.

Pursuant to the amended sentence, the Kings County District Attorney's Office sent Nowakowski a “Notice of C.S. Obligation,” dated June 6, 2013. This notice informed Nowakowski that he had been referred for community service on July 2, 2013, with the Parks Department. It informed him that he was required to appear on that date in a specific location, that the date would “NOT be rescheduled,” and that if he failed to appear or complete the required service, “a warrant may be issued for [his] arrest.” Appellant App. 18.

On July 1, 2013, Nowakowski filed a petition under 28 U.S.C. § 2254 for habeas relief. He then appeared and completed his community service before appearing in the Criminal Court on July 9, 2015, with proof of completion. Prior to filing his federal habeas petition, Nowakowski had filed a pro se civil complaint in the United States District Court for the Eastern District of New York, alleging violations of 42 U.S.C. § 1983 by arrest and imprisonment without probable cause, assault and excessive force, and the state tort of malicious abuse of process. See Second Am. Compl., Nowakowski v. City of New York et al. , No. 1:08–cv–00399–RJD–LB (E.D.N.Y. filed May 19, 2008), ECF No. 17. His civil action remains stayed pending resolution of his federal habeas proceeding.

On November 7, 2013, the District Court initially dismissed Nowakowski's petition without prejudice, because it contained an unexhausted claim of ineffective assistance of appellate counsel. Nowakowski both moved to vacate the dismissal, stating he wished to delete his unexhausted claim and proceed only on his exhausted claims, and filed a motion in our Court for a certificate of appealability. We construed the motion as one for remand to the District Court for consideration of his Rule 59(e) motion and granted it. On April 2, 2014, the District Court reopened Nowakowski's case and granted Nowakowski's motion for reconsideration, allowing his exhausted claims to proceed.

Following briefing, the District Court concluded that the expiration of Nowakowski's conditional discharge on May 14, 2014, rendered his petition moot, unless Nowakowski could demonstrate a continuing collateral consequence under Spencer , supra . Nowakowski argued that his conviction would preclude his § 1983 action from proceeding under Heck v. Humphrey , 512 U.S. 477, 114 S.Ct. 2364, 129 L.Ed.2d 383 (1994) ; the District Court held that Spencer rejected this argument against mootness as an insufficient collateral consequence. Thus, the District Court entered a Memorandum and Order on May 30, 2014, dismissing the petition as moot and denying a certificate of appealability. Nowakowski moved for a certificate of appealability in this Court, which we granted. See Order, Nowakowski v. New York , No. 14–1964 (2d Cir. Dec. 8, 2014), ECF No. 21.

DISCUSSION

We review de novo a district court's denial of a § 2254 petition, including whether a petitioner was in custody at the time of filing, see Carvajal v. Artus , 633 F.3d 95, 104 (2d Cir. 2011), and whether his petition is moot, Marrero Pichardo v. Ashcroft , 374 F.3d 46, 50–51 (2d Cir. 2004). As Nowakowski briefed and argued this case pro se , we construe his “appellate briefs and submissions liberally and interpret them to raise the strongest arguments they suggest.” Wright v. Comm'r , 381 F.3d 41, 44 (2d Cir. 2004).

I.

The first question we must decide is whether Nowakowski was “in custody” and thus able to seek federal habeas relief. In order for a federal court to have jurisdiction over a habeas petition, the petitioner must be “in custody pursuant to the judgment of a State court” at the time the petition is filed. 28 U.S.C. § 2254(a) ; Maleng v. Cook , 490 U.S. 488, 490–91, 109 S.Ct. 1923, 104 L.Ed.2d 540 (1989).

Despite the “chief use of habeas” being “the release of persons held in actual, physical custody in prison or jail,” the Supreme Court has affirmed “that, besides physical imprisonment, there are other restraints on a man's liberty, restraints not shared by the public generally, which have been thought sufficient in the English-speaking world to support the issuance of habeas corpus.” Jones v. Cunningham , 371 U.S. 236, 240, 83 S.Ct. 373, 9 L.Ed.2d 285 (1963) ; see also id. at 238, 83 S.Ct. 373. The Jones Court found jurisdiction where an individual was released from imprisonment on parole subject to explicit conditions—for example, regular reporting to his parole officer; remaining in a particular community, residence, and job; and refraining from certain activities. Id. at 242, 83 S.Ct. 373. The Supreme Court has likewise found jurisdiction where a petitioner was released on his own recognizance prior to trial but had to appear in criminal court when ordered and where failure to do so would result in issuance of an arrest warrant. See Hensley v. Mun. Court, San Jose–Milpitas Judicial Dist. , 411 U.S. 345, 351, 93 S.Ct. 1571, 36 L.Ed.2d 294 (1973) ; see also Justices of Bos. Mun. Court v. Lydon , 466 U.S. 294, 300–01, 104 S.Ct. 1805, 80 L.Ed.2d 311 (1984) (finding jurisdiction over petitioner released on his own recognizance prior to trial “subject to the conditions that he would appear when ordered by the court, that he would waive extradition if he was apprehended outside the State, and that a court could revoke the order of release and require that he be returned to confinement or post bail”).

The Courts of Appeals, including ours, have recognized that a variety of nonconfinement restraints on liberty satisfy the custodial requirement. See, e.g. , Earley v. Murray , 451 F.3d 71, 75 (2d Cir. 2006) (post-release supervision); Barry v. Bergen Cty. Prob. Dep't , 128 F.3d 152, 160–62 (3d Cir. 1997) (500 hours of community service); Poodry v. Tonawanda Band of Seneca Indians , 85 F.3d 874, 894–95 (2d Cir. 1996) (banishment from tribal land); Dow v. Circuit Court of First Circuit Through Huddy , 995 F.2d 922, 923 (9th Cir. 1993) (per curiam) (mandatory fourteen-hour alcohol rehabilitation program); Sammons v. Rodgers , 785 F.2d 1343, 1345 (5th Cir. 1986) (per curiam) (unexpired suspended sentence); United States ex rel. B. v. Shelly , 430 F.2d 215, 217 n.3 (2d Cir. 1970) (probation). Those cases where courts have declined to find the petitioners sufficiently “in custody” have typically involved the imposition of fines or civil disabilities, such as suspension of licenses. See, e.g. , Barnickel v. United States , 113 F.3d 704, 706 (7th Cir. 1997) (restitution); United States v. Michaud , 901 F.2d 5, 7 (1st Cir. 1990) (per curiam) (monetary fine); Lefkowitz v. Fair , 816 F.2d 17, 20 (1st Cir. 1987) (suspension of medical license); Lillios v. New Hampshire , 788 F.2d 60, 61 (1st Cir. 1986) (per curiam) (fine and temporary suspension of driver's license); Ginsberg v. Abrams , 702 F.2d 48, 49 (2d Cir. 1983) (per curiam) (revocation of law, real estate, and insurance licenses); see also Maleng , 490 U.S. at 492, 109 S.Ct. 1923 (collateral consequences of a completed sentence do not constitute custody).

The custody inquiry therefore “requires a court to judge the ‘severity’ of an actual or potential restraint on liberty.” Poodry , 85 F.3d at 894. Though the language of habeas cases often refers to “severe restraints on individual liberty” or “cases of special urgency,” Hensley , 411 U.S. at 351, 93 S.Ct. 1571, these terms describe the nature, rather than the duration, of the restraint. It is evident that a single day of incarceration would be sufficient custody for jurisdiction if the petitioner filed while subject to such a sentence. Similarly, courts have considered even restraints on liberty that might appear short in duration or less burdensome than probation or supervised release severe enough because they required petitioners to appear in certain places at certain times, thus preventing them from exercising the free movement and autonomy available to the unrestricted public, or exposed them to future adverse consequences on discretion of the supervising court. See id. (custody satisfied where petitioner was required to appear in court when ordered and subject to revocation of release); Barry , 128 F.3d at 161 (despite flexibility in scheduling, requirements “to be in a certain place—or in one of several places—to attend meetings or to perform services” are clearly “restraints on [petitioner's] liberty not shared by the public generally”); Dow , 995 F.2d at 923 (mandatory class attendance and “physical presence at a particular place” constituted custody, despite lasting only fourteen hours over three days); Sammons , 785 F.2d at 1345 (potential of revocation of suspended sentence or other adverse action during term sufficient for custody).

With these examples in mind, we turn to the facts before us, which are uncontested. Although the Criminal Court initially imposed a fine of $100, it vacated this sentence and replaced it with a one-year conditional discharge and an order that Nowakowski perform one day of community service. At the time of the petition's filing, therefore, the Criminal Court required Nowakowski (1) to complete a day of community service (2) by a particular date and (3) to report to the Criminal Court upon completion of that service. Pursuant to these requirements, Nowakowski received a “Notice of C.S. Obligation” from the Kings County District Attorney. The Notice commanded that Nowakowski appear at a particular location at a particular time on a particular day, and informed him that he had no opportunity to reschedule the date and that failure to appear could result in the issuance of a bench warrant. In addition, during the one-year term of his sentence—which lasted for nine months after his petition was filed—the Criminal Court retained jurisdiction to modify or enlarge the conditions of, or to revoke entirely, the conditional discharge. See N.Y. PENAL LAW § 65.05[2].

Nowakowski's sentence falls within the category of restraints that satisfy the statutory requirement of custody. These restrictions are “not shared by the public generally,” Jones , 371 U.S. at 240, 83 S.Ct. 373, require Nowakowski's physical presence at particular times and locations, both for community service and court appearances, see Barry , 128 F.3d at 161 ; Dow , 995 F.2d at 923, and carry with them the potential for future adverse consequences during the term of the sentence, including arrest for noncompliance and modification or revocation of the conditional discharge, see Hensley , 411 U.S. at 351, 93 S.Ct. 1571 ; Sammons , 785 F.2d at 1345. They are wholly unlike the economic penalties suffered in fine-only or license-revocation sentences, where the punishments “implicate only property, not liberty.” Barry , 128 F.3d at 161. Consequently, we conclude that, at the time the petition was filed, Nowakowski was “in custody” within the meaning of § 2254.

II.

The second question we must answer is whether Nowakowski's case is moot. Unlike the “in custody” requirement, mootness is not fixed at the time of filing but must be considered at every stage of the habeas proceeding. See Carafas v. LaVallee , 391 U.S. 234, 237, 88 S.Ct. 1556, 20 L.Ed.2d 554 (1968). Once, as here, a petitioner's sentence has expired, “some concrete and continuing injury other than the now-ended incarceration or parole—some ‘collateral consequence’ of the conviction—must exist if the suit is to be maintained.” Spencer , 523 U.S. at 7, 118 S.Ct. 978. In Spencer , the Court conducted a two-step analysis: First , the Court determined whether a presumption of “continuing collateral consequences” should apply. Id. at 8, 118 S.Ct. 978. And second , the Court determined whether, applying the presumption or not, there was sufficient evidence that such consequences in fact existed. Id. at 14, 118 S.Ct. 978 ; accord United States v. Mercurris , 192 F.3d 290, 293 (2d Cir. 1999). We conduct each inquiry in turn.

A.

Spencer traced more than forty years of Supreme Court precedent to explain the development of the presumption of continuing collateral consequences after its first articulation in Sibron v. New York , 392 U.S. 40, 88 S.Ct. 1889, 20 L.Ed.2d 917 (1968). The Court observed that it applied the presumption only to “criminal convictions” and expressly declined to extend it outside of that context to parole revocation. Spencer , 523 U.S. at 9–13, 118 S.Ct. 978 ; see also Lane v. Williams , 455 U.S. 624, 632–33, 102 S.Ct. 1322, 71 L.Ed.2d 508 (1982). In so doing, it candidly acknowledged that such a presumption “sits uncomfortably beside the long-settled principle” that Article III standing cannot be inferred and that the proponent of jurisdiction bears the burden of demonstrating it. Spencer , 523 U.S. at 10–11, 118 S.Ct. 978 (internal quotation marks omitted). It further considered “of particular relevance” that, with criminal convictions, “the presumption of significant collateral consequences is likely to comport with reality,” calling this observation “ ‘an obvious fact of life.’ ” Id. at 12, 118 S.Ct. 978 (quoting Sibron , 392 U.S. at 55, 88 S.Ct. 1889 ). The Court's clear reluctance to extend the Sibron presumption outside of this narrow category has guided our Court in declining to apply it when the defendant does not challenge a criminal conviction. See Mercurris , 192 F.3d at 293 (sentencing enhancement); United States v. Probber , 170 F.3d 345, 348 (2d Cir. 1999) (revocation of supervised release).

Spencer —as well as our opinions in Mercurris and Probber —may be fairly characterized as declining to apply the presumption to cases in which something ancillary to a conviction was challenged, even if of a criminal nature. Here, by contrast, it is evident that Nowakowski challenges a conviction. What is disputed is whether this conviction is criminal, for reasons that will shortly become clear. Our Circuit's precedent has never answered this question, nor has the Supreme Court spoken on the subject in the context of the Sibron presumption. Therefore, we proceed cautiously, examining the nature of the offense of which Nowakowski was convicted and drawing on principles in other areas of law where the Supreme Court has addressed similar considerations.

Before commencing our analysis, we think it necessary to explain briefly the New York scheme of penal offenses. The New York Penal Law defines an “offense” as “conduct for which a sentence to a term of imprisonment or to a fine is provided” by a state or local law, ordinance, or regulation. N.Y. PENAL LAW § 10.00[1]. The Penal Law categorizes each offense as one of four types, listed here in descending order of seriousness: a felony, a misdemeanor, a violation, and a traffic infraction. Id. § 10.00[2]–[5]. Of these four, only a felony and a misdemeanor are labeled “[c]rime[s].” Id. § 10.00[6]. Violations are offenses, other than traffic infractions, “for which a sentence to a term of imprisonment in excess of fifteen days cannot be imposed.” Id. § 10.00[3]. Misdemeanors permit incarceratory sentences of up to one year, while felonies permit incarceratory sentences over one year. Id. § 10.00[4]–[5]. Traffic infractions are violations of the Vehicle and Traffic Law, “which [are] not declared by this chapter or other law of this state to be a misdemeanor or a felony.” N.Y. Veh. & Traf. Law § 155.

In the case before us, Nowakowski was convicted of harassment in the second degree, which New York classifies as a violation. See N.Y. PENAL LAW § 240.26. Thus, the critical question is whether, taking into account New York's decision not to label violations as crimes, Nowakowski's conviction is nonetheless “criminal” for purposes of the Sibron presumption.

We start from first principles. The Sibron presumption is a judicial doctrine concerning mootness under Article III. See Liner v. Jafco, Inc. , 375 U.S. 301, 304, 84 S.Ct. 391, 11 L.Ed.2d 347 (1964). Whether a case is moot is a question of federal, not state, law. See id. As a result, we must determine whether Nowakowski's conviction is civil or criminal in nature by reference to federal principles—state law provides the necessary facts underlying the question, but federal law provides the rule of decision. See, e.g. , United States v. Juvenile Male , 564 U.S. 932, 131 S.Ct. 2860, 2864, 180 L.Ed.2d 811 (2011) (per curiam) (analyzing a certified question of Montana law to determine whether the doctrine of continuing collateral consequences was satisfied). For several reasons, we think the Supreme Court's cases determining the applicability of federal constitutional protections are the most relevant precedents upon which to draw. See, e.g. , Kennedy v. Mendoza–Martinez , 372 U.S. 144, 165–68, 83 S.Ct. 554, 9 L.Ed.2d 644 (1963) (concluding constitutional criminal protections apply where Congress applies a punitive sanction); see also, e.g. , Allen v. Illinois , 478 U.S. 364, 368, 106 S.Ct. 2988, 92 L.Ed.2d 296 (1986) (conducting same analysis with respect to sanctions imposed by state law).

First, the inquiries are identical in the question presented: whether a particular proceeding is civil or criminal in nature under federal law. See Hicks ex rel. Feiock v. Feiock , 485 U.S. 624, 630, 108 S.Ct. 1423, 99 L.Ed.2d 721 (1988) (endorsing “the characterization of this proceeding and the relief given as civil or criminal in nature, for purposes of determining the proper applicability of federal constitutional protections” as raising “a question of federal law rather than state law”). And second, both the civil-criminal analysis and the Sibron presumption are judicially created doctrines that give effect to constitutional requirements. See Kennedy , 372 U.S. at 167–68, 83 S.Ct. 554 ; Sibron , 392 U.S. at 50, 88 S.Ct. 1889. Consequently, we see no reason why, if Nowakowski's conviction is “criminal” for purposes of federal constitutional protections, it should not be criminal for purposes of the Sibron presumption, and so we apply the former analysis here.

The Supreme Court has observed that “[t]he categorization of a particular proceeding as civil or criminal is first of all a question of statutory construction.” Kansas v. Hendricks , 521 U.S. 346, 361, 117 S.Ct. 2072, 138 L.Ed.2d 501 (1997) (internal quotation marks omitted). In construing statutes, we endeavor to “determine the legislative objective”—i.e. , to establish either a civil regulatory penalty or a criminal punishment. Smith v. Doe , 538 U.S. 84, 92, 123 S.Ct. 1140, 155 L.Ed.2d 164 (2003). We start with the fact that New York does not formally classify violations as “crimes.” See N.Y. PENAL LAW § 10.00[6]. Where a legislature has expressly designated a sanction as “civil,” that may in some cases suffice to demonstrate evidence of intent. See Allen , 478 U.S. at 368, 106 S.Ct. 2988 (describing an express label of a proceeding as “civil” as indicating the state's intent “to proceed in a nonpunitive, noncriminal manner”); United States v. Ward , 448 U.S. 242, 249, 100 S.Ct. 2636, 65 L.Ed.2d 742 (1980) (observing Congress labeled a particular sanction as a “civil penalty”). Here, rather than identifying violations or their attendant sanctions as civil in nature, New York has merely excluded them from the classification of “crime.” In a case concerning the Ex Post Facto Clause, where Alaska's sex offender registration statute was neither explicitly denoted as civil nor criminal, the Supreme Court looked to the purposes of the law as articulated in its text, as well as “[o]ther formal attributes of a legislative enactment, such as the manner of its codification or the enforcement procedures it establishes.” Smith , 538 U.S. at 93–94, 123 S.Ct. 1140.

The reasons for applying a functional approach are clear. States have widely varying designations of offenses under their penal codes, and these designations subject defendants to different ranges of punishment. For example, elsewhere in our Circuit, Connecticut declines to designate “violations” as crimes, while Vermont has no violations and instead labels all offenses in its penal code as felonies or misdemeanors. Compare CONN. GEN. STAT. § 53a–24(a), with VT. STAT. ANN. tit. 13, § 1. While New York violations can result in fifteen-day imprisonment, Connecticut violations result only in fines. Compare N.Y. PENAL LAW § 10.00[3], with CONN. GEN. STAT. § 53a–27(a). The maximum incarceratory sentence for misdemeanors is one year in New York and Connecticut but two years in Vermont. Compare N.Y. PENAL LAW § 10.00[4], and CONN. GEN. STAT. § 53a–26(a), with VT. STAT. ANN. tit. 13, § 1. Such variances counsel against adopting a purely labels-dependent approach to our analysis. If we were to do otherwise, federal jurisdiction over habeas petitions arising from similar or identical conduct and punishment would be controlled by vagaries of nomenclature, not substance.

Adopting the approach of the Smith Court, we examine New York's penal code and laws regarding Nowakowski's conviction as a whole, giving due weight to the State's legislative judgments. We conclude that New York punishes violations such as Nowakowski's under its criminal, not civil, authority. An action to prosecute a violation is designated a “criminal action” under New York law and may be commenced by filing an information or prosecutor's information, see N.Y. CRIM. PROC. LAW § 1.20[4], [6], [16], which are instruments that “constitute[ ] an accusation on behalf of the state as plaintiff,” id. § 1.20[1]. Such actions are governed by the New York Criminal Procedure Law, see id. § 1.10[1](a), which requires the State to prove guilt beyond a reasonable doubt, see id. § 70.20. These are all customary indicia of the State's exercise of criminal jurisdiction.

Our understanding that Nowakowski's conviction was secured pursuant to New York's criminal authority accords with traditional conceptions of the distinction between criminal and civil jurisdiction:

The distinction of public wrongs from private, of crimes and misdemeanors from civil injuries, seems principally to consist in this: that private wrongs, or civil injuries, are an infringement or privation of the civil rights which belong to individuals, considered merely as individuals; public wrongs, or crimes and misdemeanors, are a breach and violation of the public rights and duties, due to the whole community, considered as community, in its social aggregate capacity.

2 WILLIAM BLACKSTONE, COMMENTARIES ON THE LAWS OF ENGLAND , bk. 4, ch. 1, at 5 (1st ed. 1769). Harassment in the second degree is conduct that New York seems to view as a public wrong and wishes to punish in its social aggregate capacity. See N.Y. PENAL LAW , art. 240 (designated “Offenses Against Public Order” and including harassment in the second degree); cf. Smith , 538 U.S. at 94, 123 S.Ct. 1140 (observing that codification in the criminal code may constitute some evidence of punitive intent). Thus, the New York statutory scheme evinces an intent to treat a violation as criminally sanctionable conduct, notwithstanding the formal designation of only misdemeanors and felonies as “crimes.”

In the context of federal constitutional protections, a conclusion that the state legislature intended a criminal punishment ordinarily “ends the inquiry.” Smith , 538 U.S. at 92, 123 S.Ct. 1140. However, the Supreme Court has also developed a set of factors that are “neither exhaustive nor dispositive, but are useful guideposts” for determining whether a statutory penalty is criminal or civil in nature, even where a legislature's intent is to impose civil penalties. Id. at 97, 123 S.Ct. 1140 (citations and internal quotation marks omitted). These factors, first announced in Kennedy v. Mendoza–Martinez , supra , are as follows:

[1] Whether the sanction involves an affirmative disability or restraint, [2] whether it has historically been regarded as a punishment, [3] whether it comes into play only on a finding of scienter, [4] whether its operation will promote the traditional aims of punishment—retribution and deterrence, [5] whether the behavior to which it applies is already a crime, [6] whether an alternative purpose to which it may rationally be connected is assignable for it, and [7] whether it appears excessive in relation to the alternative purpose assigned.

372 U.S. at 168–69, 83 S.Ct. 554 (footnotes omitted). The Supreme Court has instructed that “[a]bsent conclusive evidence of [legislative] intent as to the penal nature of a statute, these factors must be considered in relation to the statute on its face.” Id. at 169, 83 S.Ct. 554.

However, here, as in Kennedy , “the objective manifestations of [legislative] purpose indicate conclusively that the provisions in question can only be interpreted as punitive,” and therefore, “a detailed examination along such lines is unnecessary.” Id. ; see also Smith , 538 U.S. at 92–93, 123 S.Ct. 1140. Nonetheless, for the sake of thoroughness and to assure ourselves of the soundness of our ultimate conclusion, we briefly address these factors in the context before us.

The first three factors easily weigh in favor of finding this violation criminal in nature. As discussed above, a conviction exposes a defendant to an incarceratory sentence of up to fifteen days. Of course, imprisonment is “the paradigmatic affirmative disability or restraint,” Smith , 538 U.S. at 100, 123 S.Ct. 1140, and thus incontrovertibly considered punishment. In addition, conviction requires proof of scienter beyond a reasonable doubt—namely, “intent to harass, annoy or alarm another person.” N.Y. PENAL LAW § 240.26 ; see also N.Y. CRIM. PROC. LAW § 70.20. The fourth factor similarly weighs in favor of considering the conviction criminal: the statute clearly operates to deter bad conduct, and the possible punishments serve retribution on offenders. The fifth factor is not particularly appropriate to this context—the criminality, or not, of Nowakowski's conduct is established by this statute in its own right.

The final two factors—the existence of a rational alternative purpose and whether the sanction is excessive in relation to it—do not appear to weigh in favor of finding the sanction civil. As the analysis of factors one, three, and four show, the principal effect of the sanctions here is punitive—imprisonment, monetary fines, or both—and do not appear to have an alternate purpose that contributes to a regulatory or civil interest of the state. For example, the Supreme Court concluded in Smith that sex offender registration served a nonpunitive civil purpose: “public safety, which is advanced by alerting the public to the risk of sex offenders in their community.” 538 U.S. at 103, 123 S.Ct. 1140 (alteration and internal quotation marks omitted). No analogous civil interest immediately rises to mind on the law before us, but we need not conclusively decide the question in light of the weight of the other factors.

In summary, we conclude that Nowakowski's conviction is criminal in nature for the purposes of invoking the Sibron presumption. We do so primarily because New York has evinced a legislative intent to treat such convictions as criminal, and such intent is supported by our consideration of the Kennedy factors.

B.

Having concluded the presumption of continuing collateral consequences should apply, we now turn to whether sufficient collateral consequences to Nowakowski's conviction have been demonstrated. Because Nowakowski's conviction was based on one of the lowest level offenses under state law, we think it is likely that he will suffer fewer collateral consequences than if convicted of a felony or even a misdemeanor. Thus, as his case may end up being on the margins, we think it useful to examine how the presumption will functionally affect our inquiry.

First, Spencer defined the Supreme Court's application of the principle as being “willing to presume that a wrongful criminal conviction has continuing collateral consequences (or, what is effectively the same, to count collateral consequences that are remote and unlikely to occur).” 523 U.S. at 8, 118 S.Ct. 978. This approach has led the Court “to accept the most generalized and hypothetical of consequences as sufficient to avoid mootness in challenges to conviction.” Id. at 10, 118 S.Ct. 978. Spencer used two hypothetical possibilities that a previous decision had deemed sufficient to avoid mootness: “the possibility that the conviction would be used to impeach testimony [a defendant] might give in a future proceeding and the possibility that it would be used to subject him to persistent felony offender prosecution if he should go to trial on any other felony charges in the future.” Id. (quoting Evitts v. Lucey , 469 U.S. 387, 391 n.4, 105 S.Ct. 830, 83 L.Ed.2d 821 (1985) ). These two possibilities span the progression of the Sibron presumption. The Court has consistently repeated the idea that a conviction subjects a criminal defendant to the potential for an increased sentence for a subsequent conviction resulting from a not-yet-extant criminal prosecution. See, e.g. , Minnesota v. Dickerson , 508 U.S. 366, 371 n.2, 113 S.Ct. 2130, 124 L.Ed.2d 334 (1993) ; Pennsylvania v. Mimms , 434 U.S. 106, 108 n.3, 98 S.Ct. 330, 54 L.Ed.2d 331 (1977) ; Benton v. Maryland , 395 U.S. 784, 790–91, 89 S.Ct. 2056, 23 L.Ed.2d 707 (1969) ; Sibron , 392 U.S. at 55–56, 88 S.Ct. 1889. In these cases, the Court has also referred to the potential impeachment to which a defendant may be subject in future proceedings. See, e.g. , Benton , 395 U.S. at 791, 89 S.Ct. 2056 ; Sibron , 392 U.S. at 55–56, 88 S.Ct. 1889 ; see also Spencer , 523 U.S. at 10, 118 S.Ct. 978 ; Evitts , 469 U.S. at 391 n.4, 105 S.Ct. 830. Notably, these two consequences both require uncertain future proceedings, the first of which would occur, if at all, by virtue of the defendant's subsequent criminal conduct—i.e. , circumstances of his own making. Thus, the first effect of the presumption is to accept a broader category of consequences as sufficient for purposes of avoiding mootness.

Next, we turn to how the presumption affects the parties' obligations to present and prove the existence of collateral consequences. Although Spencer marked the Supreme Court's most complete discussion of the presumption, the Court never explicitly identified the nature or operation of the presumption. Lower federal courts have nonetheless overwhelmingly treated the Sibron presumption as rebuttable and placed the burden on the state to prove that no collateral consequences will result. See United States v. Quezada–Enriquez , 567 F.3d 1228, 1232 n.2 (10th Cir. 2009) ; D.S.A. v. Circuit Court Branch 1 , 942 F.2d 1143, 1146 n.3 (7th Cir. 1991) ; Malloy v. Purvis , 681 F.2d 736, 739 (11th Cir. 1982) ; Felton v. Mazzuca , No. 98 Civ. 4567(RJS), 2012 WL 4462009, at *6 n.4 (S.D.N.Y. Sept. 27, 2012) ; United States v. Hill , 171 F.Supp.2d 1032, 1037–38 (D.S.D. 2001). But see Chaker v. Crogan , 428 F.3d 1215, 1219 (9th Cir. 2005) (recognizing conclusive presumption).

We think the majority approach is correct. Sibron held “that a criminal case is moot only if it is shown that there is no possibility that any collateral legal consequences will be imposed on the basis of the challenged conviction.” 392 U.S. at 57, 88 S.Ct. 1889. This language suggests that the presumption may be rebutted and provides a standard of proof for that rebuttal—“no possibility” of collateral consequences. Further, Spencer 's observation that the presumption is an anomaly in Article III standing weighs against making its application conclusive even where the proof shows no collateral consequences. See 523 U.S. at 12, 118 S.Ct. 978.

Where the burden of proof rests is a more difficult question. The burden of proof is commonly understood to encompass both the obligation to produce some evidence on an issue—i.e. , the burden of production—and the obligation to persuade the decision maker that the standard of proof has been met in one's favor—i.e. , the burden of persuasion. See Burden of Proof , BLACK'S LAW DICTIONARY (10th ed. 2014). Were we to conclude that the Sibron presumption imposes both obligations on the state, we would be requiring the state both to present potential collateral consequences to the reviewing court and then disprove them. There are numerous problems with such an approach, not the least of which is the difficulty inherent in proving a universal statement —that no collateral consequences exist. Further, it creates a situation in which—through human error or otherwise—a state may fail to present an extant collateral consequence to the court, which would have established that the habeas petition was not moot. Finally, forcing a state to argue against itself would go even further afield from the traditional application of standing, in which the proponent of jurisdiction bears the burden of “alleg[ing] facts demonstrating that he is a proper party to invoke judicial resolution of the dispute.” Spencer , 523 U.S. at 11, 118 S.Ct. 978 (internal quotation marks omitted).

By contrast, requiring habeas petitioners to identify at least some collateral consequence that threatens them balances practical considerations and operation of the presumption. In the ordinary case, petitioners—and the court—may look to the consequences regularly appearing in this context: e.g. , future sentence enhancement, impeachment, or civil disabilities. Where those readily identifiable consequences are not present, however, the state should not bear the burden of both identifying and refuting every possible alleged consequence in its laws. Mindful of the Supreme Court's caution in this area, we conclude that a petitioner seeking habeas review must identify some continuing collateral consequences that may flow from his criminal conviction—including those that, as discussed above, are merely hypothetical and speculative. Once a petitioner does so, however, the state bears the burden to prove by sufficient evidence that there is “no possibility” such consequences will attach to his conviction. See Sibron , 392 U.S. at 57, 88 S.Ct. 1889.

With our framework established, we apply the Sibron presumption to the facts of this case. Nowakowski has identified a sufficient collateral consequence to avoid mootness in this case—namely, the potential for impeachment in a future proceeding. Under New York law, the trial court in its discretion may permit cross-examination into a criminal defendant's prior bad acts or crimes upon balancing the probative value of the evidence with the risk of unfair prejudice. See generally People v. Sandoval , 34 N.Y.2d 371, 357 N.Y.S.2d 849, 314 N.E.2d 413 (1974). In these instances, the defendant bears the burden “of demonstrating that the prejudicial effect of the admission of evidence thereof for impeachment purposes would so far outweigh the probative worth of such evidence on the issue of credibility as to warrant its exclusion.” Id. at 378, 357 N.Y.S.2d 849, 314 N.E.2d 413.

Additionally, where a criminal defendant presents evidence of his good character, the prosecution may prove a conviction tending to negate such trait; similarly, where conviction of an offense “constitutes an element of the offense charged, or proof thereof is otherwise essential to the establishment of a legally sufficient case,” the previous conviction may be proved. N.Y. CRIM. PROC. LAW § 60.40[2]–[3]. In least one decision, the Appellate Division has held that a conviction for harassment in the second degree can be the subject of proper cross-examination. See People v. Hogencamp , 295 A.D.2d 643, 643–44, 743 N.Y.S.2d 608 (N.Y. 3d Dep't 2002) ; see also N.Y. CRIM. PROC. LAW § 60.40[1] (permitting independent proof of conviction where it is a proper subject of cross-examination and the defendant denies or answers equivocally). Consequently, we conclude that Nowakowski's conviction subjects him to the possibility of impeachment in a future criminal proceeding and thus presents a sufficient continuing collateral consequence to satisfy the requirements of a live case or controversy. See Evitts , 469 U.S. at 391 n.4, 105 S.Ct. 830 ; Sibron , 392 U.S. at 55–56, 88 S.Ct. 1889.

III.

We conclude by addressing our colleague's thoughtful dissenting opinion. As we read it, the dissent's principal thesis is that the Sibron presumption was designed to apply—and therefore only applies—to convictions from which significant collateral consequences are likely to result. See, e.g. , Dissenting Op., post , at 229, 236. We think the dissent's approach is incompatible with the Supreme Court's practice and statements. As an analytical matter, the dissent conflates the separate steps undertaken in Sibron and Spencer into a single inquiry about whether collateral consequences do in fact flow from the defendant's conviction. In other words, the dissent's approach would require a court to find that collateral consequences generally flow from the conviction before it applies a presumption that assumes, as a first step subject to rebuttal, that collateral consequences generally flow from the conviction. Such an approach is, of course, no presumption at all—and if it were the approach the Supreme Court intended, Spencer could have been a great deal shorter and far less concerned with the presumption's anomalous place in Article III standing analysis.

The Spencer Court described its precedents as being “willing to presume that a wrongful criminal conviction has continuing collateral consequences (or, what is effectively the same, to count collateral consequences that are remote and unlikely to occur ).” Spencer , 523 U.S. at 8, 118 S.Ct. 978 (emphasis added); see also id. at 10, 118 S.Ct. 978 (“[After Sibron ], and in summary fashion, we proceeded to accept the most generalized and hypothetical of consequences as sufficient to avoid mootness in challenges to conviction.” (emphasis added)). It also characterized its past practice as “presuming collateral consequences (or of accepting the remote possibility of collateral consequences as adequate to satisfy Article III).” Id. at 10, 118 S.Ct. 978 (emphasis added); see also id. at 12, 118 S.Ct. 978 (describing it as “our presumption of collateral consequences (or our willingness to accept hypothetical consequences)”). The dissent appeals to the single instance in Spencer that uses the terms “significant” and “likely”: “In the context of criminal conviction, the presumption of significant collateral consequences is likely to comport with reality. As we said in Sibron , it is ‘an obvious fact of life that most criminal convictions do in fact entail adverse collateral legal consequences.’ ” Id. at 12, 118 S.Ct. 978 (quoting Sibron , 392 U.S. at 55, 88 S.Ct. 1889 ). But here is the language immediately following that line in Sibron : “The mere ‘possibility’ that this will be the case is enough to preserve a criminal case from ending ‘ignominiously in the limbo of mootness.’ ” 392 U.S. at 55, 88 S.Ct. 1889 (quoting Pollard v. United States , 352 U.S. 354, 358, 77 S.Ct. 481, 1 L.Ed.2d 393 (1957), and Parker v. Ellis , 362 U.S. 574, 577, 80 S.Ct. 909, 4 L.Ed.2d 963 (1960) (Warren, C.J., dissenting)). The dissent's attempt to convert the presumption's operation into an examination of the likelihood of collateral consequences for the conviction at issue, see post , at 239, is inconsistent with the Court's actual practice, as well as Spencer 's description of that practice.

The dissent's approach is also in significant tension with Minnesota v. Dickerson , 508 U.S. 366, 113 S.Ct. 2130, 124 L.Ed.2d 334 (1993). As we note above, see supra note 15, Dickerson involved a case in which “no judgment of conviction was entered and, upon respondent's successful completion of probation, the original charges were dismissed,” and in which a Minnesota law specifically prohibited the proceeding from being “deemed a conviction for purposes of disqualifications or disabilities imposed by law upon conviction of a crime or for any other purpose.” 508 U.S. at 371 n.2, 113 S.Ct. 2130 (quoting MINN. STAT. § 152.18 ). It is hard to imagine a case—including this one—that would meet the dissent's proposed standard less than Dickerson —and yet the Spencer Court cited it as an example of the presumption without a hint of disavowal, see 523 U.S. at 10, 118 S.Ct. 978.

In short, we think the dissent's approach begs the question by demanding evidence of collateral consequences to invoke a rebuttable presumption that assumes those consequences exist. Accepting the dissent's propositions would also require us to conclude that what the Supreme Court has called a “presumption,” Spencer , 523 U.S. at 8, 118 S.Ct. 978, is not really a presumption at all; that when the Court said the presumption “is applied to criminal convictions,” id. it meant to say, but did not, that it is applied only to serious criminal convictions; and that when the Court said “a criminal case is moot only if it is shown that there is no possibility that any collateral legal consequences will be imposed on the basis of the challenged conviction,” Sibron , 392 U.S. at 57, 88 S.Ct. 1889, it did not really mean the words only , no , or any . While the Supreme Court may choose to cabin the presumption to specific offenses of a certain order of magnitude in the future, the Court has not done so yet. We therefore “leav[e] to [the Supreme Court] the prerogative of overruling its own decisions,” Rodriguez de Quijas v. Shearson/Am. Express, Inc. , 490 U.S. 477, 484, 109 S.Ct. 1917, 104 L.Ed.2d 526 (1989), and take the Court at its word. Because Nowakowski has identified the possibility of a collateral consequence stemming from a criminal conviction—a consequence significant enough to have been accepted by the Supreme Court in the past—his case is not moot.

CONCLUSION

In sum, Nowakowski was in custody for the purposes of habeas review while under a sentence of conditional discharge that obligated him to appear in specific places at specific times and subjected him to the discretion of the court to modify or revoke his discharge. In addition, Nowakowski's conviction is criminal for the purposes of the Sibron presumption, and he has identified a continuing collateral consequence under its application. His petition thus presents a live case or controversy sufficient to sustain federal jurisdiction. Accordingly, the District Court's order of May 30, 2014, is hereby VACATED, and the case is REMANDED to the District Court for further proceedings consistent with this opinion.

Livingston, Circuit Judge, dissenting:

A petitioner must, at “all stages of federal judicial proceedings,” be able to demonstrate that he has “suffered[ ] or [is] threatened with[ ] an actual injury traceable to the defendant and likely to be redressed by a favorable judicial decision.” Spencer v. Kemna , 523 U.S. 1, 7, 118 S.Ct. 978, 140 L.Ed.2d 43 (1998) (quoting Lewis v. Cont'l Bank Corp. , 494 U.S. 472, 478, 110 S.Ct. 1249, 108 L.Ed.2d 400 (1990) ). Absent a concrete, non-speculative, and redressable injury, a case ceases to present “a case or controversy under Article III ... of the Constitution.” Id. As we have long observed, “[f]ederal courts are courts of limited jurisdiction. They possess only that power authorized by Constitution and statute, which is not to be expanded by judicial decree.” Kokkonen v. Guardian Life Ins. Co. of Am. , 511 U.S. 375, 377, 114 S.Ct. 1673, 128 L.Ed.2d 391 (1994) (citations omitted).

Robert Nowakowski cited a single collateral consequence in his initial habeas petition—the potential that a subsequent suit for damages under 42 U.S.C. § 1983 might not be successful absent habeas relief—a consequence that the Supreme Court has previously determined to be insufficient to confer jurisdiction over a petition for habeas corpus on a federal court. See Spencer , 523 U.S. at 17, 118 S.Ct. 978 (discussing Heck v. Humphrey , 512 U.S. 477, 114 S.Ct. 2364, 129 L.Ed.2d 383 (1994) ). The majority does not suggest that, under traditional Article III principles, Robert Nowakowski's petition for habeas corpus would be justiciable. See Maj. Op. at 223 n.16. Instead, the majority concludes that this case is not moot for two interlocking reasons. First, in the majority's view, Sibron v. New York , 392 U.S. 40, 88 S.Ct. 1889, 20 L.Ed.2d 917 (1968), requires that we apply a presumption of collateral consequences to Nowakowski's conviction for committing a violation-level offense, notwithstanding the Supreme Court's suggestion in Spencer that such a presumption accords with Article III principles only when it is a “presumption of significant collateral consequences” that “comport[s] with reality.” 523 U.S. at 12, 118 S.Ct. 978 (emphasis added). Next, having extended this presumption further than the Supreme Court has ever extended it, the majority applies the presumption in a way that makes it all but irrebuttable, concluding that a single unlikely consequence—the remotest risk of impeachment (more remote than the risks rejected as insufficient to create jurisdiction in Spencer , 523 U.S. at 14–17, 118 S.Ct. 978 )—suffices to defeat mootness here.

I would hold that, because application of a “presumption of significant collateral consequences” to a conviction for a violation-level offense in New York “is [not] likely to comport with reality,” Supreme Court precedent neither requires nor permits us to extend that presumption to the category of judgment at issue here. Id. at 12, 118 S.Ct. 978. Assuming, arguendo , that such a presumption does apply, I would further find that the Government has clearly rebutted it. Accordingly, I would dismiss Nowakowski's petition as moot.

I.

Robert Nowakowski was convicted of harassment in the second degree, a violation-level offense in New York (specifically denominated a “petty offense,” rather than a “crime”) for which the maximum punishment is fifteen days' incarceration and/or a $250 fine. See N.Y. Penal Law § 10.00 (defining four categories of “offenses,” including “[t]raffic infraction[s],” “[v]iolation[s],” “[m]isdemeanor[s],” and “[f]elon[ies],” the latter two of which are formally labeled “[c]rime[s]”); id. § 240.26 (defining harassment in the second degree as a “violation”); N.Y. Crim Proc. L. § 1.20(39) (defining “[p]etty offense” to mean “a violation or traffic infraction”); N.Y. Penal Law § 80.05(4) (articulating the maximum fine). Nowakowski was originally sentenced to pay a $100 fine. Several years after his initial sentencing, however, and just prior to his filing suit pursuant to 42 U.S.C § 1983 seeking money damages for alleged violations of his constitutional rights in connection with his conviction, he successfully moved the New York court to resentence him to a year of conditional discharge with the sole condition of mandatory attendance at a day of community service.

Days before he was to perform his community service, Nowakowski filed a petition for habeas relief. Likely aware that he would complete his sentence prior to adjudication of his habeas motion, cf. Spencer , 523 U.S. at 7, 118 S.Ct. 978 (noting that, once a sentence expires, a petitioner must point to “some concrete and continuing injury ... if the suit is to be maintained” (citing Carafas v. LaVallee , 391 U.S. 234, 237–38, 88 S.Ct. 1556, 20 L.Ed.2d 554 (1968) ), Nowakowski asserted a single collateral consequence to defeat mootness: that, absent favorable termination in his habeas proceeding, he might be unable successfully to pursue his damages suit, cf. id. at 17, 118 S.Ct. 978 (dismissing this argument as “a great non sequitur”). At no point—either before the district court or before this Court on appeal—did Nowakowski argue that a presumption of collateral consequences did or should apply to his violation-level conviction, an offense sufficiently low-level as not to be labeled criminal by New York State. The Government, in addition to arguing that a Heck bar (whether or not it would exist here) is not a sufficient collateral consequence to keep Nowakowski's petition alive, argued that a violation-level conviction, which is not denominated a crime, does not, as a categorical matter, entail sufficient collateral consequences to support federal jurisdiction. A conviction for a violation-level offense in New York does not disqualify a defendant from voting, see N.Y. Elec. Law § 5–106, or from serving on a jury, see N.Y. Jud. Law § 510. Nor does it appear generally to disqualify a person from engaging in any type of employment. In general, the records of any arrest and prosecution that ends in conviction of a violation are sealed upon termination of the action, or destroyed or returned to the defendant. See N.Y. Crim. Proc. Law § 160.55. Such a conviction does not automatically enhance a future sentence under New York law, see N.Y. Penal Law § 70.06 (enhancing a sentence on the basis of a prior felony conviction), or under federal law, see U.S.S.G. § 4A1.2(c) (excluding prior non-felony offenses for purposes of calculating a Guidelines level if they appear on a list of petty offenses or are “similar to [those offenses appearing on the list]” unless “(A) the sentence was a term of probation of more than one year or a term of imprisonment of at least thirty days, or (B) the prior offense was similar to an instant offense”).

The district court, having no argument before it that a presumption of collateral consequences should apply, and having been presented with no reference to impeachment in any of Nowakowski's initial briefs, dismissed his petition as moot on the simple basis that Nowakowski's single alleged consequence—a potential Heck bar—had “no effect on” the mootness analysis. App'x at 3. The majority revives Nowakowski's petition, concluding that a presumption of collateral consequences (which Nowakowski himself did not argue should apply) applies, and that Nowakowski's petition is thus not moot. The majority holds that Supreme Court precedent mandates such a result. Analysis of this precedent suggests, instead, that the majority, far from applying the presumption as the Court has previously applied it, has extended it —and that this extension is unwarranted on the basis of the holdings and reasoning of Sibron and Spencer .

II.

A.

The genesis of this case can be tied to two Supreme Court cases, both decided in 1968. In Carafas v. LaVallee, the Supreme Court held that the fact that a habeas petitioner's sentence had expired would not render his petition for habeas relief moot as, though his incarceration itself could no longer be remedied, he nevertheless faced collateral consequences as a result of his convictions for two felonies—burglary and grand larceny. 391 U.S. at 236–37, 88 S.Ct. 1556. As support for its conclusion, the Court cited numerous consequences that flowed from these convictions, observing that “[the petitioner could not] engage in certain businesses[,] ... serve as an official of a labor union for a specified period of time[,] ... vote in any election held in New York State[,] ... [or] serve as a juror.” Id. at 237, 88 S.Ct. 1556. “Because of these [consequences, the petitioner had] ‘a substantial stake in the judgment of conviction which survive[d] the satisfaction of the sentence imposed on him.’ ” Id. (quoting Fiswick v. United States , 329 U.S. 211, 67 S.Ct. 224, 91 L.Ed. 196 (1946) (in which a conviction rendered a petitioner “liable to deportation and denial of naturalization, and ineligible to serve on a jury, vote, or hold office,” Spencer , 523 U.S. at 9, 118 S.Ct. 978 )).

One month later, the Court decided Sibron , which, though in the context of a direct appeal, extended the reasoning in Carafas and Fiswick and articulated the presumption that the majority applies here. In Sibron , the Court faced the question whether an appellant's completion of his sentence—six months in jail for conviction of unlawful possession of heroin—rendered his appeal moot. See 392 U.S. at 40, 88 S.Ct. 1889. Relying primarily on three cases, Fiswick , United States v. Morgan , 346 U.S. 502, 74 S.Ct. 247, 98 L.Ed. 248 (1954), and Pollard v. United States , 352 U.S. 354, 77 S.Ct. 481, 1 L.Ed.2d 393 (1957), each of which specifically addressed the collateral consequences of felony convictions, see, e.g. , Fiswick , 329 U.S. at 222, 67 S.Ct. 224 (“[U]nless pardoned, [the defendant will] carry through life the disability of a felon[,] and by reason of that fact he might lose certain civil rights.”); Pollard , 352 U.S. at 358, 77 S.Ct. 481 (relying on Morgan and Fiswick 's discussion of the civil disabilities associated with felony convictions to observe that “[t]he possibility of consequences collateral to the imposition of sentence is sufficiently substantial to justify our dealing with the merits”), Sibron determined that it did not.

The Sibron Court began by suggesting that Pollard “acknowledged the obvious fact of life that most criminal convictions do in fact entail adverse collateral legal consequences.” Sibron , 392 U.S. at 55, 88 S.Ct. 1889. It then, “without pausing to canvass the possibilities in detail,” proceeded to describe specific potential consequences that could flow from the conviction at issue in Sibron itself. Id. These included the two consequences that the majority highlights—that the conviction could “be used to impeach [Sibron's] character should he choose to put it in issue at any future criminal trial [and] ... that [the conviction] must be submitted to a trial judge for his consideration in sentencing should Sibron again be convicted of a crime.” Id. at 55–56, 88 S.Ct. 1889. The Court noted, however, that “[t]here are doubtless other collateral consequences”—a proposition likely true in the context of a serious misdemeanor conviction and certainly true in the context of the felonies at issue in the precedent on which Sibron built its holding. Id. ; see also Perez v. Greiner , 296 F.3d 123, 125 (2d Cir. 2002) (observing that, “in Sibron [ ], the Court, citing various collateral consequences such as deportation, inability to become a citizen, impeachment evidence in future criminal trials, and increased future sentences, asserted a presumption that collateral consequences attach to criminal convictions post-release”).

In the decades that followed, the Court applied the presumption articulated in Sibron in the context of felony convictions, at no point applying it to an offense as categorically low-level as Nowakowski's, nor to any offense even arguably not designated a “crime.” Even as the Court continued to apply the presumption to felony convictions, moreover, it issued two decisions limiting the presumption's scope and even questioning its application in any circumstances.

First, in Lane v. Williams , 455 U.S. 624, 102 S.Ct. 1322, 71 L.Ed.2d 508 (1982), the Court assessed whether respondents' attack on their sentences to parole terms was rendered moot when these terms expired. The Lane Court did not explicitly discuss the applicability or non-applicability of a “presumption of collateral consequences.” It did, however, find the case moot. The Court began by observing that, in Carafas , numerous collateral consequences associated with the felony conviction there (including, in that case, disenfranchisement) “were sufficient to ensure that the litigant had” a constitutionally sufficient personalized stake. Id. at 632, 102 S.Ct. 1322 (quoting Carafas , 391 U.S. at 237, 88 S.Ct. 1556 ). The Court then concluded that the “doctrine of Carafas and Sibron [was] not applicable in this case” because “[n]o civil disabilities such as those present in Carafas result from a finding that an individual has violated parole.” Id. In contrast to the “statutory consequences” Carafas discussed, “a finding that an individual has violated parole,” the Court observed, at most leads to “non-statutory consequences”—including adverse employment prospects or the potential that “the sentence imposed in a future criminal proceeding, could be affected.” Id. Such consequences would depend on “discretionary decisions” of an employer or sentencing judge. Id. at 632–33, 102 S.Ct. 1322. The Court further noted that the potential that parole violations might be “considered in a subsequent parole determination” was also insufficient to defeat mootness: Carafas “concerned existing civil disabilities,” but the finding that the respondents violated their parole could not “affect a subsequent parole determination unless [they] again violate[d] state law, [were] returned to prison, and bec[a]me eligible for parole,” a sequence of events “[r]espondents themselves [would be] able—and indeed required by law—to prevent.” Id. at 632 n.13, 102 S.Ct. 1322 (emphasis added). Further, such a finding did not in and of itself render someone ineligible for future parole, but was “simply one factor, among many” that might be considered. Id. Finally, the Court observed: “Collateral review of a final judgment is not an endeavor to be taken lightly. It is not warranted absent a showing that the complainant suffers actual harm from the judgment that he seeks to avoid.” Id.

In Spencer , the Supreme Court addressed whether a petitioner's challenge to an order revoking his parole—an order based, inter alia , on a finding that he had committed forcible rape—was rendered moot when the petitioner's sentence expired. Unlike Lane (which found the petition moot but did not expressly analyze whether a presumption applied), the Court began by explicitly concluding that a presumption of collateral consequences is not appropriately applied to a parole revocation. See 523 U.S. at 8, 118 S.Ct. 978. In so concluding, the Supreme Court criticized the presumption as applied even to criminal convictions, but it did not overturn it. It did, however, explain why the Sibron Court had applied a presumption in that case. First, Spencer noted that the Sibron Court's understanding of standing and mootness was based on a then-dominant view of such Article III doctrines as simply ensuring the sharpening of issues, a “parsimonious view” that “ha[d] since yielded to the acknowledgment that” such doctrines constitute a vital “means of defin[ing] the role assigned to the judiciary in a tripartite allocation of power.” Id. (quoting Valley Forge Christian Coll. v. Am. United for Separation of Church & State, Inc. , 454 U.S. 464, 474, 102 S.Ct. 752, 70 L.Ed.2d 700 (1982) ). Second, the Court explained that “it is an ‘obvious fact of life that most criminal convictions do in fact entail adverse collateral legal consequences,’ ” id. at 12, 118 S.Ct. 978 (quoting Sibron , 392 U.S. at 55, 88 S.Ct. 1889 ), a proposition which meant that, in the context of such convictions, “the presumption of significant collateral consequences is likely to comport with reality.” Id. The Court then explained why such a presumption should not apply to Spencer's parole revocation: unlike “most criminal convictions,” such adjudications were not categorically likely to entail significant collateral consequences. Id.

Having determined that a presumption of collateral consequences would not apply, the Court proceeded to examine, inter alia , five alleged consequences claimed by the petitioner, finding each inadequate to create a case or controversy. First, the Court held that the possibility that the revocation could be used to the petitioner's detriment in some future parole proceeding was a mere “possibility rather than a certainty or even a probability.” Id. at 14, 118 S.Ct. 978. Second, while agreeing that “the Order of Revocation could be used to increase [the petitioner's] sentence in a future sentencing proceeding,” the Court observed that such a possibility was too remote as “it was contingent upon respondents' violating the law, getting caught, and being convicted.” Id. at 15, 118 S.Ct. 978. Third and fourth, though the Court did not disagree that the order might “be used to impeach [the petitioner] should he appear as a witness or litigant in a future criminal or civil proceeding,” id. (emphasis added); or “used against him directly, pursuant to [the federal rules of evidence],” it held that, because such impeachment would turn on various “ ‘discretionary decision[s]’ ” of the prosecutor or adverse counsel, id. at 16, 118 S.Ct. 978 (quoting Lane , 455 U.S. at 632, 102 S.Ct. 1322 ); as well as the similarly discretionary decision of the “presiding judge” to admit such evidence, the possibility was overly remote, id. Finally, the Court addressed the petitioner's argument that, absent habeas relief, he would be unable to pursue a claim for damages under 42 U.S.C. § 1983. The Court concluded that such an argument was “a great non sequitur, unless one believes (as we do not) that a § 1983 action for damages must always and everywhere be available.” Id. at 17, 118 S.Ct. 978. Having determined that the consequences on which Spencer relied (consequences of substantially greater severity than the consequences at issue here) were insufficient to create a case or controversy, the Court dismissed the petition as moot.

B.

As framed by the majority, this case requires us to determine whether the presumption of collateral consequences applied in cases like Sibron should be extended as well to a conviction for a violation-level offense in New York. I would hold that because such an offense, unlike the judgments to which the Court has previously applied such a presumption, is not likely to entail significant collateral consequences, such extension is not warranted. I reach this conclusion for two interrelated reasons.

First, as is evident from the above description of the Sibron line, the Supreme Court has never before applied a presumption of collateral consequences to a judgment as low-level as Nowakowski's (whether or not denominated “criminal”)—nor, for that matter, to a conviction properly labeled a “petty offense” as that term is understood in constitutional parlance. The lowest-level offense to which the Court has previously applied the presumption was the offense in Sibron itself—unlawful possession of heroin, a misdemeanor for which Sibron received a six months' sentence. 392 U.S. at 45 n.1, 88 S.Ct. 1889. In the years since, the Court has consistently applied the presumption only to felony convictions—a consistency that made it possible for Spencer convincingly to observe that the “presumption of significant collateral consequences [was] likely to comport with reality.” Spencer , 523 U.S. at 12, 118 S.Ct. 978. This factual consistency is relevant to whether such a presumption should—and certainly need—apply here. Cf. Powers v. Hamilton Cty. Pub. Def. Comm'n , 501 F.3d 592, 602 (6th Cir. 2007) (distinguishing “the ordinary rule refinement that appellate courts necessarily engage in [from] an improper departure from binding Supreme Court precedent,” in the context of distinguishing a prior Supreme Court case that did not address the same “factual scenario”).

Second, the reasoning of Sibron and Spencer confirms that this is a factual distinction with a difference—that while such a presumption might be appropriate in the context of a more serious conviction, it is not appropriate as applied to the category of judgment here. In Sibron , the Court “acknowledged the obvious fact of life that most criminal convictions do in fact entail adverse collateral legal consequences.” 392 U.S. at 55, 88 S.Ct. 1889. In justifying its approach to mootness, the Court then briefly analyzed the consequences that might flow from the defendant's conviction. In the Court's view, it was not just that the conviction could be used to impeach the defendant and for purposes of sentencing enhancement—the Court plausibly observed that “[t]here [were] doubtless other collateral consequences.” Id. at 56, 88 S.Ct. 1889. In contrast, Spencer , in declining to apply a presumption of collateral consequences to a parole revocation, concluded that application of a presumption to such revocations, unlike application of a presumption to the convictions at issue in Sibron and its progeny, would not be “likely to comport with reality.” Spencer , 523 U.S. at 12, 118 S.Ct. 978 ; see also Lane , 455 U.S. at 632–33, 102 S.Ct. 1322 (distinguishing the parole revocation from the convictions in Sibron and Carafas not because the former was not “criminal” or a “conviction,” but because “[n]o civil disabilities such as those present in Carafas result from a finding that an individual has violated parole”). The reasoning in Spencer —which in turn characterized the cases that came before it—was thus as follows: first, the Sibron presumption was justified because collateral consequences were categorically likely to flow from convictions; and second, such consequences were less likely to flow from parole revocations, rendering a parallel presumption fundamentally inconsistent with our obligations under Article III. This Court, moreover, in deciding whether to extend the presumption to a new category of judgment, has described the reasoning of Sibron , Spencer , and Lane in precisely this way. See United States v. Mercurris , 192 F.3d 290, 293 (2d Cir. 1999). In light of the Court's prior holdings, and the reasoning of Sibron and Spencer , it is perfectly evident that a presumption of collateral consequences cannot logically be extended to a violation-level offense in New York. This is because the general consequences of such an offense are not only categorically less severe than the consequences logically associated with the convictions to which a presumption was applied in Sibron and its progeny, but also than those associated with parole revocations. Parole revocations, after all, can result in a mandated federal sentencing enhancement. See U.S.S.G. § 4A1.2(k). In contrast, even assuming that the mere risk of a sentencing enhancement is sufficient to justify application of a presumption after Lane and Spencer , cf. United States v. Dominguez–Carmona , 166 F.3d 1052, 1059–60 (10th Cir. 1999) (Lucero, J. , concurring) (“[T]he Supreme Court in Spencer [ ] .... cast into doubt the practice of using highly speculative collateral consequences to stave off mootness ....”); Mercurris , 192 F.3d at 294 (declining to apply a presumption to a determination that a conviction is an aggravated felony, as, post-Spencer , “a finding of collateral consequences cannot be based on the speculation that an individual will receive an enhanced sentence in a future sentencing proceeding in connection with a crime he has not yet committed”), there is minimal risk of such enhancement with prior violations, see N.Y. Penal Law § 70.06 ; U.S.S.G. § 4A1.2(c). As the majority observes, there may be some risk of impeachment associated with conviction of a violation-level offense—though only, it seems, in state criminal court. See Maj. Op. at 225–26. Even if evidence rules permit introduction of a violation-level conviction just as they would permit introduction of a more serious offense, however, the risk of impeachment actually occurring is not logically commensurate. And Spencer makes clear that some small risk of impeachment is simply not enough to render a presumption appropriate.

In short, it is not simply that a violation-level offense is factually distinct in severity from the offenses to which the Court has previously applied a presumption. It is that the way that such an offense, not denominated a crime in New York State, is distinct distinguishes it from those convictions and makes evident that no presumption should apply. Indeed, the basic logic of applying a presumption of collateral consequences to a serious conviction clearly forecloses application of such a presumption here. To the degree that the Court has “presum[e]d” significant consequences, it has done so because they are likely; and to the degree that the Court has “count [ed] collateral consequences that are remote and unlikely to occur,” it has done so both because such consequences invariably sit beside other unstated, and unacknowledged likely consequences, and because the seriousness of the convictions at issue suggests that even consequences that may appear remote are more likely to occur than might be apparent. Spencer , 523 U.S. at 8, 118 S.Ct. 978. This is not to say that application of a presumption of collateral consequences even to relatively serious adjudications does not sit “uncomfortably” beside traditional principles of Article III, as the Spencer majority understood. Id. at 10–11, 118 S.Ct. 978. But there is only limited discomfort, as long as the presumption is confined to such a context. In contrast, to apply a presumption to an adjudication like the one here is to require courts to presume the existence of an injury when such injury is not likely to exist . Applied in such a manner, the presumption becomes an arbitrary mechanism for the manufacture of cases and controversies that will inevitably require federal courts to hear cases on the merits in contravention of their obligations under Article III. That is the reasoning and sense of the cases from Sibron to Spencer —reasoning and sense from which the majority today departs. C.

The majority does not necessarily dispute that, were we to assess the question on a blank slate, we might plausibly hold that extension of a presumption of collateral consequences to violation-level convictions contravenes our obligations under Article III and is inconsistent with the reasoning and outcome in Spencer . Instead, the majority's primary argument is that precedent requires application of a presumption here, because the Supreme Court used the term “criminal conviction” in cases like Sibron and Spencer without qualification, and because Nowakowski's conviction is arguably penal in character for purposes of ascertaining whether certain constitutional protections attach. See Maj. Op. at 218–22, 228. At the start, to the degree that there is even doubt that the judgment in this case was the sort contemplated by Sibron and Spencer , Spencer 's admonition against extending the presumption should resolve the inquiry here. Nevertheless, the majority's reading of the Court's precedent—as well its approach to explicating those holdings—is flawed on its own terms.

First, the majority's invocation of the words “criminal conviction” in precedent like Sibron and Spencer divorces these words from their factual context and thus exaggerates the scope of the Court's prior holdings. As already noted, the Court has never applied the presumption to a judgment as categorically unlikely to entail collateral consequences as Nowakowski's petty offense, nor indeed even to any misdemeanor conviction apart from the one in Sibron itself. Even if this factual consistency did not strongly indicate that a presumption should not be applied in this case, it certainly belies any suggestion that the Court, in using the term “criminal conviction” without qualification, has already effectively resolved the inquiry before us. See Kokkonen , 511 U.S. at 379, 114 S.Ct. 1673 (noting that the “holding [of a prior case] was not remotely as permissive as its language,” and that “[i]t is to the holdings of our cases, rather than their dicta, that we must attend”); Powers, 501 F.3d at 602.

Second, even if we were to treat the words “criminal conviction” as talismanic, notwithstanding the actual holdings in which such words appear, there is still no reason to believe that the scope of this phrase includes the judgment here. The meaning of these words must be ascertained in context. Cf. United States v. Balsys , 524 U.S. 666, 673, 118 S.Ct. 2218, 141 L.Ed.2d 575 (1998) (holding that the text “any criminal case” under the Fifth Amendment's Self–Incrimination Clause does not generally include criminal cases in foreign jurisdictions, and noting that the textual argument to the contrary “overlooks the cardinal rule to construe provisions in context”). The Supreme Court has often used the short-hand “criminal conviction,” moreover, when context makes perfectly clear that it can only have meant felony conviction or, in some cases, a serious misdemeanor. For instance, in North Carolina v. Rice , the Court observed, without citing Sibron , that “[a] number of disabilities may attach to a convicted defendant even after he has left prison, and the Court has recognized the standing of such persons to challenge the legality of their convictions even when their sentences have been served.” 404 U.S. 244, 247, 92 S.Ct. 402, 30 L.Ed.2d 413 (1971). In a footnote, the Court then supported this statement by citing numerous consequences that would generally apply only to felonies, including disenfranchisement. See id. at 247 n.1, 92 S.Ct. 402 (citing, inter alia , Comment, Civil Disabilities of Felons, 53 Va.L.Rev. 403 (1967)). The Court nowhere qualified the term conviction with the words “felony” or “serious,” but no matter—the meaning was clear enough from context.

Further, in interpreting the ambit of constitutional rights, the words “criminal conviction,” “crime,” or “criminal prosecution” have frequently been used without qualification to announce the scope of a given right; notwithstanding such pronouncements, the Court has subsequently held that these rights do not apply to all criminal convictions as that term might be used in other contexts—in particular distinguishing more serious convictions from petty offenses. See Lewis v. United States , 518 U.S. 322, 325, 116 S.Ct. 2163, 135 L.Ed.2d 590 (1996) (observing that, though the Sixth Amendment is clear that “[i]n all criminal prosecutions, the accused shall enjoy the right to ... trial, by an impartial jury,” “there is [nevertheless] a category of petty crimes or offenses which is not subject to the Sixth Amendment jury trial provision” (first quoting U.S. Const. Am. VI ; then quoting Duncan v. Louisana , 391 U.S. 145, 159, 88 S.Ct. 1444, 20 L.Ed.2d 491 (1968) ); see also Duncan , 391 U.S. at 160, 88 S.Ct. 1444 (“So-called petty offenses ... have always been held to be exempt from the otherwise comprehensive language of the Sixth Amendment's jury trial provisions[, in part because] .... [t]he possible consequences to defendants from convictions for petty offenses have been thought insufficient to outweigh the benefits to efficient law enforcement and simplified judicial administration resulting from the availability of speedy and inexpensive nonjury adjudications”). The majority's determination that “criminal conviction” includes a low-level judgment like Nowakowski's (denominated a petty offense, and not criminal, by the State of New York) thus does not obviously flow from the Supreme Court's references to “criminal conviction” in cases like Sibron and Spencer , absent some ability to square this outcome with the context, logic, and reasoning of these decisions.

Yet the majority's approach to this inquiry fails this test. The majority concludes that the Court's references to “criminal conviction” in Sibron and Spencer necessarily require that a presumption apply whenever a conviction is functionally criminal “for purposes of determining the proper applicability of [certain] federal constitutional protections,” Hicks on Behalf of Feiock v. Feiock , 485 U.S. 624, 630, 108 S.Ct. 1423, 99 L.Ed.2d 721 (1988), regardless whether collateral consequences are likely to flow from the judgment in question. But beyond the word “criminal,” this approach derives no obvious support from either case. Moreover, it is an illogical—and thus extremely unlikely—approach to explicating the term “criminal conviction” as used in the context of this Article III precedent for at least two reasons.

First, the determination that a given matter should be treated as criminal for the purpose of a particular constitutional protection does not suddenly transform that adjudication into a criminal conviction for purposes of a state's collateral consequences regime: it does not, in other words, change the status of the proceeding for purposes of enhancement provisions, impeachment laws, felony impairments, or the other collateral legal consequences animating cases like Sibron and Carafas. Indeed, it is not clear how the majority's analysis in any way illuminates the actual collateral consequences flowing from an adjudication. It is perhaps for this reason that the Fifth Circuit recognized in Port v. Heard , 764 F.2d 423 (5th Cir. 1985), in the context of discussing civil and criminal contempt, that the “mootness of contempt judgments turns, not on the formalistic enterprise of labeling the contempt judgment, but on the presence vel non of a live controversy.” Compare id. at 427, with Feiock , 485 U.S. at 627, 108 S.Ct. 1423 (assessing whether a contempt proceeding was functionally civil or criminal).

Second, the doctrine on which the majority relies is not only irrelevant to the assessment of mootness: it is, in fact, counterproductive. The analyses in cases like Feiock address principally the necessity of ensuring, through careful doctrinal elaboration, that states cannot evade specific constitutional criminal protections afforded to the accused through mere nomenclature—that they cannot, by denominating a proceeding or penalty as civil, for instance, evade specific criminal procedural protections that would otherwise attach. In contrast, however, a state's designation of an offense as noncriminal may often be relevant to mootness (and indeed is relevant here), and mootness serves not as a protection afforded to individuals in the criminal process, but as a limitation on courts , confining the judiciary to its constitutionally defined role. See Spencer , 523 U.S. at 11, 118 S.Ct. 978. For these reasons, I am hard-pressed to agree with the majority that the Supreme Court intended, in using the words “criminal conviction” in the context of serious convictions, to require courts faced with difficult questions at the edge of mootness to resolve such questions not through appeal to the existence of a controversy, but through application of precedent created in an unrelated context for unrelated reasons, and that is not grounded in the animating limitations of Article III.

In short, I believe that the question whether a presumption of collateral consequences applies to a violation-level offense is an unresolved one in our jurisprudence that does not logically turn on whether or not certain constitutional protections unrelated to mootness apply. As the question before us is thus whether to extend the presumption to a new category of adjudication, I would conclude that the presumption may not be applied to a violation for the simple reason that it does not “comport with reality” to presume that “significant collateral consequences” will necessarily flow from such an offense across the range of cases. Spencer , 523 U.S. at 12, 118 S.Ct. 978.

III.

The preceding conclusion is sufficient to resolve this matter since, absent a presumption, the risk of impeachment in a future proceeding to which the majority alludes is clearly inadequate, under Spencer , to deem this a live case. See id. at 16, 118 S.Ct. 978 (deeming analogous risk of impeachment “purely a matter of speculation”). Even if I agreed with the majority that a presumption of collateral consequences applies, however, I believe it is also clear that the Government has rebutted it here.

As an initial matter, I disagree with the majority's understanding of the standard of proof that the Government must meet to rebut the presumption. The majority observes that a petitioner may present any consequence that is “merely hypothetical and speculative,” Maj. Op. at 225, and then cites language from Sibron that a petition is moot only if the government can show that “there is no possibility that any collateral legal consequences will be imposed on the basis of the challenged conviction.” Sibron , 392 U.S. at 57, 88 S.Ct. 1889. But if one were to take this statement—and thus this “standard of proof,” Maj. Op. at 224— literally , the necessary result would almost certainly be that the presumption is functionally irrebuttable.

The majority expressly disclaims that its presumption is irrebuttable, a conclusion it reaches, sensibly, on its reading of Spencer . See Maj. Op. at 224–25; see also Meister , 2007 U.S. Dist. LEXIS 98605, at *12 (observing that the “presumption [of collateral consequences] ... should be viewed in light of the Spencer Court's plainly stated reluctance to find collateral consequences absent a showing of some concrete statutory disability flowing from the conviction being attacked”). Yet once one concedes, as the majority does, that this standard cannot be taken literally and that Spencer necessarily had an effect on the strength of the presumption, one is left wondering why the majority invokes the Sibron language at all. I believe a better view is that the Government, to rebut the presumption, need show only that there is no reasonable possibility of any significant collateral consequences. Such a standard comports both with Spencer 's express construction of the presumption, see 523 U.S. at 12, 118 S.Ct. 978 (referring to Sibron as creating a “presumption of significant collateral consequences” (emphasis added)), as well as with the reasoning in Spencer itself. And such a standard would surely require us to find this case moot.

Yet even if one adopts the majority's articulation of the standard, it follows that, in affirming that this presumption is rebuttable , the majority too is not taking the text literally. In other words, even as the majority cites this standard as support for its holding, it necessarily has implicitly drawn—and other courts will have to explicitly draw—some line . To the degree then that the majority agrees that a line must be drawn—that it is not actually the case that the Government must show “no possibility [of] any collateral legal consequences”—I see no justification for the majority's conclusion that the risk of impeachment in this case should fall above that line for two reasons.

First, that risk should not be viewed in a vacuum, but in the context of the Government's broader analysis of the consequences of a violation-level offense. The majority observes that the “effect of the presumption is to accept a broader category of consequences as sufficient for purposes of avoiding mootness.” Maj. Op. at 223. Yet to the degree that the Court has accepted speculative consequences before, it has done so in the context of more serious convictions, and in the context of a presumption whose central premise is that many consequences flow from a given conviction, even if only one or two are specifically named by the Court. See Sibron , 392 U.S. at 56, 88 S.Ct. 1889. Even if the Government's categorical analysis of a violation-level offense is not enough to avoid application of the presumption, it cannot be irrelevant as well at the rebuttal stage, and it has not been irrelevant to other courts. See, e.g. , Meister , 2007 U.S. Dist. LEXIS 98605, at *12.

Second, even assuming a risk of impeachment in a vacuum is sufficient to create a justiciable controversy, the risk of such impeachment—though not nonexistent—is substantially lower in this case than in Spencer or any other Supreme Court case in which the Court has cited impeachment as the material consequence. Indeed, the very fact that New York does not label violations as crimes, while not dispositive of whether such offenses are “criminal” for purposes of certain constitutional protections, is surely relevant to the likelihood that the series of discretionary decisions necessary to result in use of that conviction to impeach Nowakowski would occur. See Spencer , 523 U.S. at 13, 118 S.Ct. 978. Even if the Court has accepted speculative or hypothetical consequences when assessing the consequences of convictions, that does not mean that any consequence, no matter how speculative or hypothetical, is sufficient to make a case justiciable. If that is so, the presumption is not rebuttable.

In short, even assuming a presumption to apply, it would seem the Government has rebutted it here. In holding otherwise, I fear that the majority—even as it claims the mantle of reasonableness in suggesting that this presumption is rebuttable —erects a wall so high that it is only in the rarest case that the Government will ever be able to leap it. Taken by itself, such an approach to rebuttal is problematic. Viewed in concert with the majority's initial determination that a presumption may apply to an offense even when that offense is unlikely to entail significant consequences, this approach guarantees that the presumption will be applied, as it has been in this very case, in a way that invents, out of thin air, federal jurisdiction. See Kokkonen , 511 U.S. at 377, 114 S.Ct. 1673 ; Larche v. Simons , 53 F.3d 1068, 1069–70 (9th Cir. 1995). IV.

In light of my analysis of mootness, I would not reach the question whether Nowakowski was in custody at the time he filed his habeas petition. I note, however, that to the degree that the majority is correct that a penalty of a day of community service—and presumably commensurate penalties—may confer upon a federal district court jurisdiction to overturn a state conviction, it follows that most petitioners in Nowakowski's position will be seeking the writ primarily, or even exclusively, to remedy collateral consequences. It is not obvious to me that this result is consistent with the Supreme Court's recognition that “[t]he custody requirement of the habeas corpus statute is designed to preserve the writ of habeas corpus as a remedy for severe restraints on individual liberty,”Hensley v. Mun. Court, San Jose Milpitas Judicial Dist., 411 U.S. 345, 351, 93 S.Ct. 1571, 36 L.Ed.2d 294 (1973), and neither the Supreme Court, nor our Court, has previously found a restraint of commensurate severity to constitute custody. But be that as it may, if habeas, an “extraordinary remedy whose operation is to a large extent uninhibited by traditional rules of finality and federalism,” id. has transformed, for some petitioners, from a remedy of severe deprivations of liberty to a mere tool through which a petitioner may challenge the collateral consequences of a conviction or adjudication, then it is all the more important that we ensure that the petition presents a live case or controversy when analyzing those collateral consequences.

I respectfully dissent.