Table of CONTENTS
I. BACKGROUND FACTS.1243
II. PROCEDURAL HISTORY .1245
III. MOOTNESS.1245
IV. RIPENESS.....'.1246
A. Introduction.1246
B. The Ex Post Facto, Bill of Attainder, and Double Jeopardy Challenges.. 1247
1. Hardship of Denying Review.1247
2. Fitness of Issues for Judicial Review.1249
C. Due Process Claims.1251
1. Burden of Persuasion.1251
2. Notice.1252
D. Summary of Unripe Claims.1252
V. REGISTRATION. 1253
A. “Punishment” Under the Ex Post Facto, Bill of Attainder, and Double
Jeopardy Clauses.1253
1. De Veau v. Braisted: Subjective Purpose.1254
2. United States v. Halper: Objective Purpose through Proportionality.. 1254
3. Austin v. United States: Objective Purpose through History.1256
4. Department of Revenue v. Kurth Ranch: Objective Purpose and
Deterrence.1258
5. California Department of Corrections v. Morales: Effect.1260
6. Kennedy v. Mendoza-Martinez: The Inquiry for the Nature of Pro-
ceedings .1261
B. Synthesizing the Jurisprudence: The Test(s).1263
C. The Registration Provisions of Megan’s Law Evaluated.1264
1. Actual Purpose.1264
2. Objective Purpose.1264
3. Effects.1266
D. Summary of Registration Claims.1267
VI. EQUAL PROTECTION.1267
VII. DUE PROCESS.1268
*1242VIII. UNCONSTITUTIONAL VAGUENESS .1269
IX. PULLMAN ABSTENTION.1270
X. CONCLUSION.1271
OPINION OF THE COURT
BECKER, Circuit Judge.Alexander Artway thought that he had paid his debt to society by serving seventeen years in jail for a sex offense. After he was released, Artway settled in a community, secured employment, and married. Then, on October 31, 1994, New Jersey enacted Megan’s Law. The Law requires certain sex offenders — including those like Artway found at sentencing to be “repetitive and compulsive” — to register with local law enforcement. It also requires community notification for registrants deemed a future risk. Artway sought an injunction against the enforcement of Megan’s Law pursuant to 28 U.S.C. § 2201 and 42 U.S.C. § 1983, arguing that it punishes him, unconstitutionally, a second time. He also alleged that the Law provides insufficient procedural protections.
After summary proceedings in which no evidence was heard and virtually no factual record developed, the District Court for the District of New Jersey held that the notification aspects of Megan’s Law violated the Ex Post Facto Clause of the United States Constitution and enjoined their enforcement against Artway. The court upheld the constitutionality of the Law’s registration component. Both sides appealed.
These cross appeals present numerous questions (some of which are quite difficult): (1) Do the registration and notification provisions of Megan’s Law constitute “punishment” within the meaning of the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses of the U.S. Constitution? (2) Is Megan’s Law unconstitutionally vague? (3) Does Megan’s Law violate equal protection or due process? (4) Are any or all of Art-way’s claims unripe or moot? and (5) Was the district court’s decision not to abstain under Railroad Commission v. Pullman Co., 312 U.S. 496, 61 S.Ct. 643, 85 L.Ed. 971 (1941), proper?
Timing is important not only to punishment, but also to proper judicial decision-making. Although we reject the State’s contention that Artway’s claims are moot because he has moved from New Jersey, ripeness problems preclude us from reaching the lion’s share of Artway’s claims. First, Artway’s claims that Megan’s Law’s notification provisions violate the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses are unripe. Sex offenders are subject to notification only if the prosecutor finds a significant risk of recidivism-a determination that, with respect to Artway, has not yet been made and cannot be easily forecasted. It is far from clear, therefore, that Artway will ever be subject to notification. Moreover, we cannot make the novel, difficult, and fact-sensitive determination whether the notification provisions constitute “punishment”-the central question under all three clauses-without a record of how notification will be implemented and what concrete effects it will have on Artway (or those similarly situated). Although Art-way’s contention that notification constitutes punishment is prima facie quite persuasive, the claim will be fit for judicial review only when Artway (or some other sex offender) submits to the notification process and the impact is chronicled in the record. Similarly, since Artway has not yet been classified under Megan’s Law, his claim that he is due more process for receiving notice of and challenging a hypothetical determination regarding his dangerousness is unripe.
With regard to Artway’s claims that are currently justiciable, we hold first that Megan’s Law’s registration component does not violate the Ex Post Facto, Double Jeopardy, or Bill of Attainder Clauses as impermissible “punishment.” As the following discussion will show, the law on “punishment” is complicated and in some disarray. We devote a significant portion of this opinion, therefore, to explaining and synthesizing caselaw on the “punishment” issue in order to formulate the correct legal test.
We also hold that (1) the “repetitive and compulsive” classification of Megan’s Law does not. offend equal protection; (2) the alleged unreliability and unfairness of Art-*1243way’s “repetitive and compulsive” determination does not violate due process; (3) Megan’s Law is not unconstitutionally vague as applied to him; and (4) the district court did not err in refusing to abstain under Pullman.
We therefore vacate the judgment of the district court insofar as it enjoins the enforcement of Tier 2 and Tier 3 notification under Megan’s Law, and affirm that judgment insofar as it holds the registration provisions (including Tier 1 notification) of the Law constitutional.
I. BACKGROUND FACTS
In 1971, a New Jersey jury convicted Art-way of sodomy. The statutory elements of Artway’s crime did not require force, but the judge found that he had used violence and, as a result, sentenced him to an indefinite term in prison. See Artway v. Pallone, 672 F.2d 1168, 1170-71 & n. 3 (3d Cir.1982).1 In addition, based in part on a prior statutory rape conviction, the judge made a finding for sentencing purposes that Artway’s conduct was “characterized by a pattern of repetitive, compulsive behavior.” See id. After serving seventeen years of the sentence, Artway was released in 1992 (he had been a fugitive from 1971 to 1975).
In 1994, the New Jersey legislature enacted Megan’s Law-formally the New Jersey’s Sexual Offender Registration Act, Pub. L.1994, Chs. 128, 133 (codified at N.J.SA. 2C:7-1 to 7-11)-in response to public outcry following the brutal rape and murder of a seven-year-old girl, Megan Kanka. Megan, her parents, and the community did not know that the murderer, who lived across the street from the Kankas, was a twice-convicted sex offender. The legislation was rushed to the Assembly floor as an emergency measure, skipping the committee process, and was debated only on the floor; no member voted against it.
Megan’s Law enacts a registration requirement and three tiers of notification. The registration provision requires all persons who complete a sentence for certain designated crimes involving sexual assault after Megan’s Law was enacted to register with local law enforcement. N.J.SA. 2C:7-2b(1). Those committing these offenses and completing all incarceration, probation, and parole before the Law’s enactment must register only if, at the time of sentencing, their conduct was found to be “characterized by a pattern of repetitive and compulsive behavior.” Id.
The registrant must provide the following information to the chief law enforcement officer of the municipality in which he resides: name, social security number, age, race, sex, date of birth, height, weight, hair and eye color, address of legal residence, address of any current temporary legal residence, and date and place of employment. N.J.SA. 2C:7-4b(1). He must confirm his address every ninety days, notify the municipal law enforcement agency if he moves, and re-register with the law enforcement agency of any new municipality. N.J.S.A. 2C:7-2d to e.
The registration agency then forwards the registrant’s information, as well as any additional information it may have, to the prosecutor of the county that prosecuted the registrant. N.J.SA 2C:7-4e to d. The prosecutor, in turn, forwards the information to the Division of State Police, which incorporates it into a central registry and notifies the prosecutor of the county in which the registrant plans to reside. Id. This information is available to law enforcement agencies of New Jersey, other states, and the United States. N.J.SA 2C:7-5. The registration information is not open to public inspection. Law enforcement agencies are authorized to release “relevant and necessary information concerning registrants when ... necessary for public protection,” but only in accordance with the notification procedures we describe below. Failure of the sex offender to comply with registration is a fourth-degree crime. Id.
At this stage, the notification provisions are triggered. The prosecutor of the county in which the registrant plans to live must *1244consider the information provided through registration and, in consultation with the prosecutor of the convicting county, determine whether the registrant poses a low, moderate, or high risk of re-offense. N.J.S.A. 2C:7-8d(1). In making that determination, the prosecutor must consider guidelines the Attorney General has promulgated pursuant to the Act. N.J.S.A. 2C:7-8a to b.
The determination of risk as low, moderate, or high places the registrant in corresponding notification categories: Tier 1, Tier 2, or Ter 3. Under Ter 1 (low risk), the prosecutor must notify law enforcement agencies likely to encounter the registrant. N.J.S.A. 2C:7-8c(1). Under Ter 2 (moderate risk), the prosecutor, working with local law enforcement agencies, must notify schools, licensed day care centers, summer camps, and designated community organizations involved in the care of children or the support of battered women or rape victims. N.J.S.A. 2C:7-8c(2). Under Ter 3 (high risk), law enforcement agencies are required to notify members of the public likely to encounter the registrant. N.J.S.A. 2C:7-8c(3).
The prosecutor makes this future risk determination using the “Registrant Risk Assessment Scale,” promulgated by the Attorney General. See Registration and Community Notification Bench Manual 26. The Scale is a matrix of thirteen categories organized into four larger headings: (1) Seriousness of Offense; (2) Offense History; (3) Characteristics of the Offender; and (4) Community Support. Id.2 The prosecutor scores each of these categories for different levels of risk-low, moderate, or high. Id. In doing so, he or she is guided by commentary that includes factual examples. Id. at 17-25. This initial risk score is multiplied by coefficients that differ by category, and the data is tabulated for a final risk assessment score. Id. at 26. Finally, the prosecutor must consider whether two exceptions apply. “If an offender has indicated that he will reoffend if released into the community and the available record reveals credible evidence to support this finding, then the offender will be deemed a high risk_” Id. at 16. Conversely, “if the offender demonstrates a physical condition that minimizes the risk of reoffense, then the offender will be deemed to be a low risk.” Id.
The form of notification under Ters 2 and 3 includes the registrant’s name, a recent photograph, his physical description, offense, address, place of employment or schooling, and a description and license plate number of the registrant’s vehicle. Id. at 39. Those notified under Ter 2 are informed that the information is not to be shared with the general public, and every notification must contain a warning about the criminal consequences of vandalism, threats and assaults against the registrant or any of his associates. Id. at 40.
The New Jersey Supreme Court, in upholding the constitutionality of Megan’s Law in Doe v. Poritz, 142 N.J. 1, 662 A.2d 367 (1995), read the following additional procedural protections into the statute. First, Ter 2 notice must be confined to those likely to encounter the registrant. Id. at 29, 662 A.2d 367. Second, the prosecutor must give the registrant notice, unless “impossible as a practical matter,” before any Ter 2 or 3 notification. Id. at 30-31, 662 A.2d 367. Third, a court must provide an opportunity for a judicial hearing, in camera, in which the registrant bears the burden of persuasion. Id. at 31-32, 662 A.2d 367.
Because every registrant is classified at a minimum under Ter 1, this lowest level of notification accompanies every registration. Ter 1 requires notice only to law enforcement, whereas Ter 2 and Ter 3 both result in notice to the community. Consequently, for purposes of the subsequent discussion, “registration” will include registration and Ter 1 notification, while “notification” will refer to Ter 2 and Ter 3 notification.
*1245II. PROCEDURAL HISTORY
Artway sought declaratory relief, pursuant to 28 U.S.C. § 2201 and 42 U.S.C. § 1988, alleging that enforcement of Megan’s Law against him would violate his federal constitutional rights, including equal protection, due process, and the right not to be punished in violation of the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses. The district court decided the ease in the most summary fashion. After the State moved to dismiss Artway’s motion for injunctive relief, Artway urged the district court to construe his original motion as one for summary judgment. The court obliged. It allowed no discovery, heard no testimony, and made no findings of fact. Instead, it ruled as a matter of law on all the complex issues pending before it.
The court opened its opinion by brushing aside a ripeness challenge to Artway’s claims. The court then held that the registration component of Megan’s Law was constitutional, but that Tier 2 and Tier 3 notification violated the Ex Post Facto Clause. In doing so, it treated this case as an abstract issue of law. The court recited caselaw on the Ex Post Facto, Cruel and Unusual Punishment, Bill of Attainder, and Double Jeopardy Clauses. It also invoked state court cases, and, as might be expected, it discussed the Scarlet Letter:3 The resulting record contains only one piece of information describing the indirect effects of Megan’s Law on Art-way: a copy of a Guardian Angel flier distributed in Artway’s community warning people to “BEWARE.” 4
But even that evidence is not discussed in the district court’s opinion. Instead, the court asserted that the registration component of Megan’s Law is constitutional “for the reasons expressed in Arizona v. Noble, [171 Ariz. 171, 829 P.2d 1217 (1992) ].” Artway v. Attorney General, 876 F.Supp. 666, 688 (D.N.J.1995). It then invalidated the notification provisions of Megan’s Law using the seven-factor test for punishment of Kennedy v. Mendoza-Martinez, 372 U.S. 144, 168-69, 83 S.Ct. 564, 567-68, 9 L.Ed.2d 644 (1963). The court enjoined New Jersey, at first preliminarily and then permanently, from enforcing the notification provisions of Megan’s Law. It did not reach Artway’s other arguments, such as the Due Process and Equal Protection challenges he presses before this Court.
Artway appeals the district court’s ruling that registration and Tier 1 notification are constitutional, and presses his Due Process and Equal Protection arguments should this Court find Tier 2 and Tier 3 constitutional. The State appeals the district court’s holding that Tier 2 and Tier 3 are unconstitutional. At this juncture, these issues all present legal questions, subject to plenary review.5 See American Medical Imaging Corp. v. St. Paul Fire & Marine Ins. Co., 949 F.2d 690, 692 (3d Cir.1991).
III. MOOTNESS
As a threshold matter, we reject the State’s assertions that Artway’s appeal is moot because he has moved out of New Jersey. Artway no longer has a live claim, the State argues, because his move from New Jersey voided his duty to register. The State points us to Lujan v. Defenders of Wildlife, 504 U.S. 555, 112 S.Ct. 2130, 119 L.Ed.2d 351 (1992), in which the Supreme Court held that environmentalists did not have injury in fact because they could not show any concrete evidence, such as a plane ticket, of their intent to return to the foreign country where the challenged environmental action would take place. Like the environmentalists in Lujan, New Jersey argues, Artway’s “bald assertion that he intends to *1246return to New Jersey ... rests on conjecture and is entirely hypothetical.”
But if the record is clear on nothing else, it shows that Artway’s obligation to register is keeping him from returning to New Jersey, and that situation presents a real controversy. The litigants in Lujan merely opined that they planned to visit the site, in a foreign country, “some day” in the future. 504 U.S. at 564 & n. 2, 112 S.Ct. at 2138 & n. 2. Artway, in contrast, lived in New Jersey-where he established a home, a family, and a job-until March 3, 1995. He left shortly after Megan’s Law was passed and has sworn that Megan’s Law is keeping him from moving back. Indeed, he brought this litigation, originally pro se, in order to remain there. Artway cannot live in New Jersey without either complying with Megan’s Law, which undoubtedly burdens him, or facing prosecution. Especially given the constitutional right to move interstate, see Shapiro v. Thompson, 394 U.S. 618, 89 S.Ct. 1322, 22 L.Ed.2d 600 (1969), this Hobson’s choice constitutes sufficient injury in fact even under Lujan’s standing analysis.
In addition to being factually different from Lujan, the State’s mootness claim is legally different from that case. Lujan addressed standing, which inquires whether someone is the proper party to bring a law suit at the beginning of the case. Doctrinally, to satisfy core Article III requirements, standing requires (1) that the plaintiff suffer injury in fact, (2) that the injury be fairly traceable to the challenged conduct, and (3) that a favorable ruling would redress the injury. See Lujan, 504 U.S. at 560-61, 112 S.Ct. at 2136-37. Mootness, on the other hand, asks whether a party who has established standing has now lost it because the facts of her case have changed over time. Thus, the threshold for satisfying the prohibition against mootness is somewhat lower than that for standing. “[T]he central question in mootness inquiries is whether changes in circumstances that prevailed at the beginning of the litigation have forestalled any occasion for meaningful relief.” Huber v. Casablanca Indus., Inc., 916 F.2d 85, 107 (3d Cir.1990) (internal quotations omitted), overruled on other grounds by Milwaukee Brewery Workers’ Pension Plan v. Jos. Schlitz Brewing Co., — U.S. -, 115 S.Ct. 981, 130 L.Ed.2d 932 (1995); accord Zellous v. Broadhead Associates, 906 F.2d 94, 100 (3d Cir.1990) (“An action becomes moot when ‘(1) there is no reasonable expectation that the alleged events will recur ... and (2) interim relief or events have completely eradicated the effects of the violation.’ ”) (quoting County of Los Angeles v. Davis, 440 U.S. 625, 631, 99 S.Ct. 1379, 1383, 59 L.Ed.2d 642 (1979)).6
The opportunity for meaningful relief is still present here. Artway ceased the activity which unquestionably granted him standing — living in New Jersey — only upon threats of enforcement. And he has sworn to his desire to return if Megan’s Law is invalidated. Cf. Begins v. Philbrook, 513 F.2d 19, 24 (1975) (holding case not moot even though plaintiffs sold second automobile on threats of benefit termination when they demonstrated continuing desire to own two ears).
IV. RIPENESS
A. Introduction
We next examine the State’s assertions that Artway’s ex post facto, double jeopardy, bill of attainder, and due process challenges are not ripe. Article III, as part of its “case or controversy” mandate, requires parties to suffer injury or come into immediate danger of suffering an injury before challenging a statute. See O’Shea v. Littleton, 414 U.S. 488, 494, 94 S.Ct. 669, 675, 38 L.Ed.2d 674 (1974). The basic rationale *1247of the ripeness requirement is “to prevent the courts, through the avoidance of premature adjudication, from entangling themselves in abstract disagreements.” Abbott Labs. v. Gardner, 387 U.S. 136, 148, 87 S.Ct. 1507, 1515, 18 L.Ed.2d 681 (1967). Ripeness prevents courts from interfering with legislative enactments until it is necessary to do so, and enhances the quality of judicial decision-making by ensuring that cases present courts an adequate record to permit effective review and decisionmaking. See id. Ripeness involves weighing two factors: (1) the hardship to the parties of withholding court consideration; and (2) the fitness of the issues for judicial review. See 387 U.S. at 149, 87 S.Ct. at 1515-16.7
B. The Ex Post Facto, Bill of Attainder, and Double Jeopardy Challenges
Artway contends that Megan’s Law imposes unconstitutional punishment under the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses. Under the Ex Post Facto Clause, the government may not apply a law retroactively that “inflicts a greater punishment, than the law annexed to the crime, when committed.” Colder v. Bull, 3 U.S. (3 Dall.) 386, 390, 1 L.Ed. 648 (1798). Under the Bill of Attainder Clause, legislatures are forbidden to engage in “[[legislative acts, no matter what their form, that apply either to named individuals or to easily ascertainable members of a group in such a way as to inflict punishment on them without a judicial trial.” United States v. Brown, 381 U.S. 437, 448-49, 85 S.Ct. 1707, 1715, 14 L.Ed.2d 484 (1965). Finally, the Double Jeopardy Clause prohibits, inter alia, “a second prosecution for the same offense after conviction ... and multiple punishments for the same offense.” United States v. Halper, 490 U.S. 435, 440, 109 S.Ct. 1892, 1897, 104 L.Ed.2d 487 (1989).
The crux of Artway’s argument is that Megan’s Law imposes unconstitutional “punishment.” In analyzing the ripeness of these challenges, we must carefully distinguish between the registration and notification provisions of Megan’s Law. We shall not, however, distinguish among the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses; their differences with respect to the requisites of “punishment,” if any, are not relevant here.
1. Hardship of Denying Review
The first factor for determining ripeness is the hardship of denying review. Abbott Labs., 387 U.S. at 149, 87 S.Ct. at 1515-16. The district court considered this factor, but failed to distinguish between the registration and notification aspects of Megan’s Law. The hardship factor inquires whether the threat of prosecution is “credible,” and not merely “speculative,” so as to be concrete for purposes of Article III. See Babbitt v. United Farm, Workers Nat’l Union, 442 U.S. 289, 298, 99 S.Ct. 2301, 2308-09, 60 L.Ed.2d 895 (1979). Although preenforcement review is the exception rather than the rule, “[w]hen the plaintiff has alleged an intention to engage in a course of conduct arguably affected with a constitutional interest, but proscribed by a statute, and there exists a credible threat of prosecution thereunder, he should not be required to await and undergo a criminal prosecution as the sole means of seeking relief.” Id. (internal quotations omitted); accord Steffel v. Thompson, 415 U.S. 452, 459, 94 S.Ct. 1209, 1216, 39 L.Ed.2d 505 (1974) (“[I]t is not necessary that petitioner first expose himself to actual arrest or prosecution to be entitled to challenge a statute he claims deters the exercise of his constitutional rights.”); Abbott Labs., 387 U.S. at 154, 87 S.Ct. at 1518-19 (holding a business’s challenge to a labeling statute ripe even though the company had not been threatened specifically with prosecution).
This Court has afforded review even when the state has taken no active measures toward prosecution. For example, in Presby*1248tery of Orthodox Presbyterian Church v. Florio, 40 F.3d 1464 (3d Cir.1994), we held that a church pastor’s preenforcement challenge to New Jersey’s anti-discrimination law was ripe for adjudication when the pastor had announced his intention to speak against homosexuality even though the government had not actually threatened to prosecute. That the state would not disavow the possibility of prosecution for activities outside the church was enough to make the threat “real and substantial.” Id. at 1468.
On the other hand, “[m]any cases deny ripeness on the straight-forward ground that the anticipated events and injury are simply too remote and uncertain to justify present adjudication.” 13A Charles A. Wright et al., Federal Practice and Procedure § 3532.2, at 138 (1984). A substantial contingency is the classic impediment to a preenforcement challenge. For example, in New Hanover Township v. United States Army Corps of Engineers, 992 F.2d 470, 473 (3d Cir.1998), we held that a challenge to construction of a municipal waste landfill was unripe because the state had not yet granted a necessary water quality certificate. Although the Army Corps of Engineers had granted another permit that the plaintiffs sought to challenge, we explained, construction of the landfill still could not commence: “[T]he effects of the Craps’ deciding that [the project] may proceed ... will not be felt in a concrete way unless and until the [state] grants [the project] a water quality certificate.” Id.; see also Acierno v. Mitchell, 6 F.3d 970, 975-77 (3d Cir.1993) (holding challenge to zoning decision unripe when review board had not yet made final decision); Wilmington Firefighters Local 1590 v. City of Wilmington, Fire Dept., 824 F.2d 262, 266 (3d Cir.1987) (holding challenge to yet uncreated promotion lists unripe because they were “purely a matter of conjecture”).
Artway urges that both the registration and notification components of Megan’s Law constitute unconstitutional “punishment” under the Ex Post Facto, Double Jeopardy, and Bill of Attainder Clauses. Artway’s challenge to the registration provisions of Megan’s Law satisfies the hardship prong. Like the petitioners in Babbitt, Stef-fel, Abbott Labs. and Florio, he faces the decision of complying with a putatively invalid law or suffering prosecution. Registration presents no contingency for Artway. If he resides in New Jersey, he must provide certain information to local law enforcement. And the high profile of Megan’s Law, and Artway’s ease in particular, virtually assures that Artway will be prosecuted if he engages in his allegedly protected conduct: returning to New Jersey without registering. In fact, the Attorney General assured the district court at oral argument that she would prosecute Artway if he failed to register. See Artway v. Attorney General, 876 F.Supp. 666, 670 n. 4 (D.N.J.1995). Under these circumstances, the threat of prosecution Art-way faces satisfies any test of the Supreme Court and of this Court: these threats are credible, real, and substantial.
In sharp contrast, Artway’s challenge to the notification provisions of Megan’s Law fails this prong. Unlike registration, notification involves a crucial contingency: only if, after registering, Artway is classified as a moderate or high risk of re-offense will he face notification. This classification hinges on a New Jersey prosecutor’s future decision to be reached after applying the Attorney General’s “Registrant Risk Assessment Seale.” See supra page 1244. The State prosecutor, possessing the pertinent information not present in this record, scores these thirteen categories for different levels of risk, employing the corresponding eleven pages of guidelines. The prosecutor then multiplies by differing coefficients, tabulates the data for a risk assessment score, and considers whether exceptions apply.
As in New Hanover Township, Ademo, and Wilmington Firefighters, whether this contingency will ever come to pass is a matter of speculation. We may not pass upon hypothetical matters. And Artway faces no hardship from denying review of his notification challenges at this point. If he registers, and if the State decides that his situation warrants community notification, he may seek to enjoin that action at that time. Thus, *1249the “hardship” factor alone precludes review of Artway’s notification claims.8
2. Fitness of Issues for Judicial Review
The second factor for evaluating ripeness, this one never mentioned by the district court, is whether the issues are fit for judicial review. Abbott Labs., 887 U.S. at 149, 87 S.Ct. at 1515-16. In making this determination, we must once again distinguish between the registration component of Megan’s Law on the one hand, and the notification provisions on the other. The principal consideration is whether the record is factually adequate to enable the court to make the necessary legal determinations. The more that the question presented is purely one of law, and the less that additional facts will aid the court in its inquiry, the more likely the issue is to be ripe, and vice-versa. Compare Duke Power Co. v. Carolina Envtl. Study Group, Inc., 438 U.S. 59, 81-82, 98 S.Ct. 2620, 2634-35, 57 L.Ed.2d 595 (1978) (“Although it is true that no nuclear accident has occurred and that such an occurrence would eliminate much of the existing scientific uncertainty surrounding this subject, it would not, in our view, significantly advance our ability to deal with the legal issues presented nor aid us in their resolution.”) with Zemel v. Rusk, 381 U.S. 1, 20, 85 S.Ct. 1271, 1282, 14 L.Ed.2d 179 (1965) (“[I]f we are to avoid rendering a series of advisory opinions, adjudication of the reach and constitutionality of [a statute under which the President prohibited travel to Cuba] must await a concrete fact situation.”).
Courts are particularly vigilant to ensure that cases are ripe when constitutional questions are at issue. See Communist Party of the United States v. Subversive Activities Control Bd., 367 U.S. 1, 81, 81 S.Ct. 1357, 1402, 6 L.Ed.2d 625 (1961) (holding unripe an ex post facto challenge to the Corrupt Practices Act especially in light of the rule to avoid unnecessary constitutional decisions). Indeed, the Supreme Court has held a constitutional challenge unripe because of the need for more detailed factual information in the record “[e]ven though the challenged statute is sure to work the injury alleged.” Babbitt v. United Farm Workers Nat’l Union, 442 U.S. 289, 300, 99 S.Ct. 2301, 2310, 60 L.Ed.2d 895 (1979).
Two Supreme Court cases illustrate the need for factual information particularly well. In Socialist Labor Party v. Gilligan, 406 U.S. 583, 92 S.Ct. 1716, 32 L.Ed.2d 317 (1972), the Court dismissed as unripe a challenge on First Amendment grounds to a state law that required candidates to swear not to attempt to overthrow the government by violence or force. The Court concluded that “the record ... is extraordinarily skimpy in the sort of proved or admitted facts that would enable us to adjudicate this claim.” Id. at 587, 92 S.Ct. at 1719. Even assuming the plaintiffs had standing to challenge the law, the Court continued, “their case has not given any particularity to the effect on them of Ohio’s affidavit requirement.” Id. at 588, 92 S.Ct. at 1719. In California Bankers Association v. Shultz, 416 U.S. 21, 94 S.Ct. 1494, 39 L.Ed.2d 812 (1974), the Court similarly declared unripe a First Amendment challenge to bank record-keeping and reporting requirements because of an insufficient factual record. Id. at 56, 94 S.Ct. at 1515. “This Court, in the absence of a concrete fact situation in which competing assoeiational and governmental interests can be weighed, is simply not in a position to determine whether an effort to compel disclosure of such records would or would not be barred .... ” Id.
Megan’s Law’s registration provisions require simply that Artway register and provide information to the local prosecutor, who in turn may provide the information only to local law enforcement agents. No private individuals or other organizations may receive this information. Registration, therefore, involves few variables in its opera*1250tion. As in Duke Power, the issue is primarily one of law and further factual information will provide little assistance. Under these circumstances, we are confident that Art-way’s registration challenge is fit for judicial review.
The notification procedures, on the other hand, involve dissemination of potentially devastating information to undetermined numbers of private citizens. Because these private citizens are not part of the trained state law enforcement mechanism, we are less certain how they will react. For instance, the one study in the record chronicles a number of incidents of harassment at the hands of private citizens as a result of the State of Washington’s notification law, but records no incidents on the part of law enforcement. We also lack concrete record evidence about what Airway’s future dangerousness classification will be, on what facts this classification will be determined, and who will be notified.9
Because Artway has not submitted to these procedures, and because the district court decided this ease without admitting any appreciable evidence, we have almost no factual grounding on which to make an assessment about notification as applied to Artway. The record contains two pieces of data: a flier distributed by the Guardian Angels warning Woodbridge residents to “BEWARE” and the brief State of Washington report describing the effects of a different law in that jurisdiction.10 While the tenor of the flier and the results of the study are worrisome indeed, they are but snippets compared to a developed record. Consistent with the basic principles of Gilligan, Shultz et alia, we cannot make complex and important determinations in a factual vacuum.
Moreover, the constitutionality of the notification provisions of Megan’s Law may well turn on the most careful parsing of the Supreme Court’s rulings on “punishment.” Not only must we decide whether a multifaceted and novel11 regulatory scheme violates con*1251stitutional safeguards, we must also discern the parameters of these safeguards themselves. As the discussion in Part V infra reveals, the law in this area needs clarification. We may not undertake this task without factual tools.
Thus, Artway’s challenge to the notification provisions of Megan’s Law fails both prongs of the ripeness test. The district court erred because, in analyzing the hardship of denying review, it did not distinguish between registration and notification; it also omitted the fitness for judicial review prong entirely. Whether Artway will ever be subject to Megan’s Law’s notification requirements remains a matter of speculation, and the record lacks the factual information necessary for this Court to decide Artway’s notification claims consistent .with its Article III obligations.
C. Due Process Claims
Two of Artway’s due process claims are also unripe. Artway argues that Megan’s Law denies him due process because; to avoid notification, he bears the burden of persuasion to demonstrate that he is not a risk of future danger. He also claims that Megan’s Law does not provide adequate notice of the State’s intention to initiate notification. The district court did not reach these issues because it held the notification provisions of Megan’s Law unconstitutional under the Ex Post Facto Clause. Since we have already discussed ripeness extensively, we analyze these claims more briefly.
1. Burden of Persuasion
The Fourteenth Amendment forbids states from denying “life, liberty, or property, without due process of law.” U.S. Const. amend. XIV. For purposes of this analysis, we will assume that notification under Megan’s Law implicates a liberty interest under state law sufficient to invoke federal due process protections. Doe found such an interest. See 142 N.J. at 104, 662 A.2d 367; accord Hewitt v. Helms, 459 U.S. 460, 466, 103 S.Ct. 864, 868-69, 74 L.Ed.2d 675 (1983) (holding that the Due Process Clause protects state created liberty interests as well as federal liberty interests). But cf. Sandin v. Conner, — U.S. -, -, 115 S.Cf. 2293, 2297-2300, 132 L.Ed.2d 418 (1995) (rejecting Hewitt’ s methodology of examining, state regulations rather than the nature of the deprivation in determining the existence of liberty interests and suggesting limits on the scope of state-created liberty interests that trigger federal due process safeguards).
Due process is a flexible concept determined by application of a three-part balancing test: (1) the private interests affected by the proceeding; (2) the risk of error imposed by the procedure created by the State; and (3) the countervailing interest in using the procedures it adopted. See Mathews v. Eldridge, 424 U.S. 319, 335, 96 S.Ct. 893, 903, 47 L.Ed.2d 18 (1976). This test applies to burdens of proof. See Heller v. Doe, 509 U.S. 312, 322-24 n. 1, 113 S.Ct. 2637, 2644 n. 1, 125 L.Ed.2d 257 (1993).
Artway argues that all three factors of the Mathews test counsel rejection of the State’s procedure, which places the burden of persuasion on the sex offender to prove that he is not dangerous in order to avoid notification. Rather, Artway contends, the State should bear the burden of persuasion and that burden should be by clear and convincing evidence. Artway submits that (1) his private interest in not being branded a dangerous sex offender is very great; (2) the fact that the State possesses greater resources counsels that it should bear a greater share of the burden (especially when Artway is called on to “prove the negative,” i.e., that he is not dangerous); and (3) the State’s interest is in getting the determination right, not in notifying in all cases. Cf. Santosky v. Kramer, 455 U.S. 745, 102 S.Ct. 1388, 71 L.Ed.2d 599 (1982) (state bears burden of persuasion by clear and convincing evidence for parental-rights termination); Addington v. Texas, 441 U.S. 418, 99 S.Ct. 1804, 60 L.Ed.2d 323 (1979) (same for civil commitment proceedings).
Artway also asserts that judicial deference to the prosecutor’s findings violates due process by establishing a constitutionally excessive presumption against him. Cf. Virgin Islands v. Parrilla, 7 F.3d 1097 (3d Cir.1993) (striking down statute creating rebuttable mandatory presumption). Under Megan’s *1252Law, the judge “shall affirm the prosecutor’s determination unless ... persuaded by a preponderance of the evidence that it does not conform to the laws and the Guidelines.” Doe, 142 N.J. at 82, 662 A2d 367.
Although Artway’s challenges on these issues are forceful, his claims are not ripe. That he will ever confront the process he challenges is entirely speculative at this point. This process is available to contest notification decisions, and Artway would be the subject of notification (as opposed to merely registration) only if he is classified as a Tier 2 (moderate risk) or Tier 3 (high risk) offender. While we know that Artway will be prosecuted if he does not register, we do not know whether, even if he does register, he will ever need to utilize the process he challenges.
2. Notice
Due process requires “notice reasonably calculated, under all circumstances, to apprise interested parties of the pendency of the action and afford them an opportunity to present their objections.” Mullane v. Central Hanover Bank & Trust, 339 U.S. 306, 314, 70 S.Ct. 652, 657, 94 L.Ed. 865 (1950). Artway argues that Megan’s Law does not provide for adequate notice of the commencement of notification proceedings. The Law requires notice to registered sex offenders classified as Tier 2 or Tier 3 before the corresponding notification occurs. However, the Act, as interpreted by Doe, dispenses with notice when “impossible as a practical matter.” 142 N.J. at 30-31, 662 A.2d 367. An erroneous notification would inflict an irreparable deprivation of his liberty interest, Artway argues, so that the State can never dispense with notice (and his corresponding right to a hearing). See United States v. Raffoul, 826 F.2d 218, 224 (3d Cir.1987) (“[A] likelihood of irreparable harm resulting from the lack of a pre-deprivation hearing is a private interest which countervails any public interest in streamlined administration.”).
But Artway’s notice claim is unripe for the same two reasons as his “punishment” and burden of persuasion challenges. First, his need for notice about proposed notification is speculative. Artway will need notice only if he is classified as a Tier 2 or Tier 3 risk. Second, the record in this case is insufficient to make this determination. The question is whether the notice requirement of Megan’s Law satisfies the strictures of due process. Mullane makes clear that the right to notice is not absolute; rather, Artway has a right to “reasonably calculated” notice. 339 U.S. at 314, 70 S.Ct. at 657. And Raffoul demonstrates that the State cannot dispense with notice when that notice is possible and irreparable harm could result. 826 F.2d at 224. Against this legal backdrop, we must evaluate Megan’s Law’s “impossible as a practical matter” standard, but we have no factual matrix against which to evaluate this standard because Artway has not submitted to Megan’s Law.12
D. Summary of Unripe Claims
In summary, we conclude that Artway’s ex post facto, double jeopardy, bill of attainder, and due process challenges to Megan’s Law’s notification provisions are not ripe.13 We therefore vacate the judgment of *1253the district court insofar as- it holds Tier 2 and Tier 3 notification unconstitutional, and direct it to dismiss Artway’s due process claims to the extent they concern notification.
Y. REGISTRATION
A. “Punishment” Under the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses
We turn now to the merits of Artway’s ripe challenge: that the registration provisions of Megan’s Law violate the Ex Post Facto, Bill of Attainder, and Double Jeopardy Clauses. We begin by recapping the nature of those protections. The Constitution provides that “[n]o state shall ... pass any ... ex post facto Law.” U.S. Const. art. I, § 10. Under the Ex Post Facto Clause, the government may not apply a law retroactively that “inflicts a greater punishment, than the law annexed to the crime, when committed.” Colder v. Bull, 3 U.S. (3 Dall.) 386, 390, 1 L.Ed. 648 (1798).
The Constitution also forbids states to “pass any Bill of Attainder.” U.S. Const. art. I, § 10.14 Under the Bill of Attainder Clause, legislatures are forbidden to enact “[l]egislative acts, no matter what their form, that apply either to named individuals or to easily ascertainable members of a group in such a way as to inflict punishment on them without a judicial trial.” United States v. Brown, 381 U.S. 437, 448-49, 85 S.Ct. 1707, 1715, 14 L.Ed.2d 484 (1965).
Finally, the Constitution provides: “[N]or shall any person be subject for the same offense to be twice put in jeopardy of life or limb.” U.S. Const. amend. V. [T]he Double Jeopardy Clause prohibits, inter alia, “a second prosecution for the same offense after conviction ... and multiple punishments for the same offense.” United States v. Halper, 490 U.S. 435, 440, 109 S.Ct. 1892, 1897, 104 L.Ed.2d 487 (1989). The threshold question under each clause, therefore, is whether the registration provisions of Megan’s Law impose “punishment.” If registration does not impose punishment, our inquiry with respect to the registration issue is at an end.15
*1254We must sort through several key cases involving these various provisions to derive (or, perhaps more appropriately given the confused state of the law, “divine”) the test for punishment. In the end, we develop a multi-part test that looks to the legislature’s subjective purpose in enacting the challenged measure, its “objective” purpose in terms of proportionality and history, and the measure’s effects.
1. De Veau v. Braisted: Subjective Purpose
We start with De Veau v. Braisted, 363 U.S. 144, 80 S.Ct. 1146, 4 L.Ed.2d 1109 (1960), in which the Supreme Court announced a subjective (or actual) legislative purpose test. In that case, the Court upheld, against bill of attainder and ex post facto challenges, a law forbidding certain unions employing former felons from collecting dues. In effect, the law barred convicted felons from working on the New York and New Jersey waterfront. The Court explained that “[t]he question in each case where unpleasant consequences are brought to bear upon an individual for prior conduct, is whether the legislative aim was to punish that individual for past activity, or whether the restriction of the individual comes about as a relevant incident to a regulation of a present situation, such as the qualifications of a profession.” Id. at 160, 80 S.Ct. at 1155 (emphasis added).
“The proof is overwhelming,” the Court continued, “that New York sought not to punish ex-felons, but to devise what was felt to be a much-needed scheme of regulation of the waterfront, and for the effectuation of that scheme it became important whether individuals had previously been convicted of a felony.” Id. This early case, emphasized by New Jersey, suggests that actual legislative purpose is the only inquiry. But subsequent cases make clear that this is no longer true.
2. United States v. Halper: Objective Purpose through Proportionality
Almost thirty years later, in United States v. Halper, 490 U.S. 435, 109 S.Ct. 1892, 104 L.Ed.2d 487 (1989), the Court articulated an “objective” legislative intent test-the test central to the arguments of both Artway and the State. Halper held that a sizeable fine, imposed in a civil proceeding after the defendant’s conviction for Medicare fraud, violated the Double Jeopardy Clause. The Court analyzed the issue by determining whether the fine served the purposes of punishment, including retribution and deterrence, or instead satisfied a remedial purpose. “Simply put, a civil as well as a criminal sanction constitutes punishment,” the Court said, “when the sanction as applied in the individual ease serves the goals of punishment.” Id.
We have recognized in other contexts that punishment serves the twin aims of retribution and deterrence. Furthermore, retribution and deterrence are not legitimate nonpunitive governmental objectives. From these premises, it follows that a civil sanction that cannot be fairly said solely to serve a remedial purpose, but rather can only be explained as also serving either retributive or deterrent purposes, is punishment, as we have come to understand that term.
Id. at 448, 109 S.Ct. at 1902 (citations and internal quotations omitted) (emphasis added).
The Court found that the fine in that case-$130,000-bore “no rational relation” to the legitimate remedial purpose-compensating the government for its $16,000 in costs. Id. at 449, 109 S.Ct. at 1902. Therefore, the Court held that the Double Jeopardy Clause barred the additional civil sanction after criminal punishment “to the extent that the second sanction may not fairly be characterized as remedial, but only as a deterrent or retribution.” Id. at 448-49, 109 S.Ct. at 1902.16
*1255Because Habper occupies such a central role in the punishment inquiry, a number of explanatory observations are in order. The first is a matter of semantics: a clear understanding of the terms “retributive,” “deterrent,” and “remedial” is critical to applying the Halper test. We therefore explain how we think the Supreme Court is using the terms; at least the reader will know how we are using them. Retribution is vengeance for its own sake. It does not seek to affect future conduct or solve any problem except realizing “justice.” Deterrent measures serve as a threat of negative repercussions to discourage people from engaging in certain behavior. Remedial measures, on the other hand, seek to solve a problem, for instance by removing the likely perpetrators of future corruption instead of threatening them (De Veau), or compensating the government for costs incurred (Halper).
Of course, as the cases point out, a measure could serve all three functions. For instance, putting someone in jail for a sex offense serves the retributive function of hurting that person, the deterrent purposes of convincing him and others not to engage in that behavior to avoid the adverse consequences, and the remedial purpose of keeping him away from others (at least outside the prison). Another complication is that measures can have one or more of these effects without having that purpose.
With this lexicon in mind, we turn to an explication of the Halper calculus, which evaluates the proportionality of ends to means. To recapitulate, the Halper test is whether “a civil sanction that cannot be fairly said solely to serve a remedial purpose, but rather can only be explained as also serving either retributive or deterrent purposes, is punishment.” Id. at 448, 109 S.Ct. at 1902. (emphasis added). The threshold question is thus whether a remedial purpose can explain the sanction. Only if the remedial purpose is insufficient to justify the measure, and one must resort also to retributive or deterrent justifications, does the measure become punitive. Only then can the measure “only be explained as also serving either retributive or deterrent purposes.”
To illustrate with a venerable statutory interpretation hypothetical, assume that someone is sent to the store in the snow for soupmeat. The trip can be explained solely by the remedial purpose of obtaining food, even though the trip through the cold could also serve retributive purposes. See id. at 447 n. 7, 109 S.Ct. at 1901 n. 7 (“[O]ur cases have acknowledged that for the defendant even remedial sanctions carry the sting of *1256punishment.”)- It therefore qualifies as non-punishment under Halper. On the other hand, assume now that, without additional justification, the agent is sent without clothes. This additional aspect of the trip cannot be explained by the remedial purpose of obtaining food; this excursion can only be explained as partly serving retributive purposes. It therefore constitutes “punishment” under the Halper test.17
Halper thus contributes an important element to our analysis: it adds an objective inquiry to supplement the actual legislative purpose test of De Veau. “This constitutional protection is intrinsically personal. Its violation can be identified only by assessing the character of the actual sanctions imposed on the individual by the machinery of the state.” Id. at 447, 109 S.Ct. at 1901; see also id. at 453, 109 S.Ct. at 1904 (Kennedy, J., concurring) (“Today’s holding, I would stress, constitutes an objective rule that is grounded in the nature of the sanction and the facts of the particular case.”).18
By acknowledging that “civil” penalties may constitute punishment, Halper departs from the practice of placing talismanic significance on the legislative labels affixed to the disputed provision and searching for the frequently unknowable and nondispositive subjective intent of the legislative body: “[T]he labels ‘criminal’ and ‘civil’ are not of paramount importance_ The notion of punishment ... cuts across the division between the civil and the criminal law.” Id. at 447-48, 109 S.Ct. at 1901.19
The Halper objective ends-means test is a step down the road to limiting especially harsh effects, but still any “sting” could be permissible with a sufficient post hoc remedial “purpose.” For example, the need for supper could explain the trip through the snow even if the temperature were below zero.
3. Austin v. United States: Objective Purpose through History
Four years after Halper, in Austin v. United States, 509 U.S. 602, 113 S.Ct. 2801, 125 L.Ed.2d 488 (1993), the Court added yet another dimension to the punishment question: a focus on history. The Court held that civil forfeiture is “punishment” subject to the Excessive Fines Clause of the Eighth Amendment. The government had argued that forfeiture of a mobile home and body shop after the owner was convicted of a drug offense served the remedial purpose of compensating the government for its costs in investigating and prosecuting these offenses. In setting out the appropriate analysis, the *1257Austin Court rescribed the key passage in Halper.
We said in Halper that “a civil sanction that cannot fairly be said solely to serve a remedial purpose, but rather can only be explained as also serving either retributive or deterrent purposes, is punishment, as we have come to understand the term.”
Id. at 610, 113 S.Ct. at 2806 (quoting Halper, 490 U.S. at 448, 109 S.Ct. at 1901-02).
The Austin Court then took a different tack than the Halper Court: it applied the Halper test primarily by examining history, rather than proportionality. “We turn, then, to consider whether, at the time the Eighth Amendment was ratified, forfeiture was understood at least in part as punishment and whether forfeiture under [the statute in question] should be so understood today.” Id. Examining history, it concluded that forfeiture has traditionally been regarded as punishment. Looking to the language and legislative history of the statute as a whole, the Court determined that these factors confirmed that the forfeiture statute served a punitive purpose, regardless of the proportionality of the particular forfeiture to the government’s costs.20 Id. at 617-23, 113 S.Ct. at 2810-12. It therefore remanded for a determination whether the forfeiture, by being “excessive,” violated the Eighth Amendment. Id.
According to Austin, a measure that has historically served punitive purposes is punishment unless the text or legislative history shows a contrary purpose. Id. at 619, 113 S.Ct. at 2810 (“We find nothing in these provisions or their legislative history to contradict the historical understanding of forfeiture as punishment.”). Thus, even if a remedial purpose could fully explain a measure, thereby satisfying Halper, it will not pass Austin muster if it has historically been considered punishment and neither the text nor the legislative history contradicts this purpose. To draw again on our soupmeat hypothetical, sending someone out into the snow would be punishment if doing so was traditionally regarded as punitive and the sender did not make his plausible remedial purposes clear. This would be the case even though a remedial purpose-fetching soupmeat-could fully explain the action. Without a convincing counter-rationale, something understood as punishment for so long simply “cannot fairly be said solely to serve a remedial purpose, but rather can only be explained as also serving retributive or deterrent purposes.” Id. at 610, 113 S.Ct. at 2806.
The Austin objective purpose analysis also represents a move toward analyzing the effect of a provision in ascertaining whether it inflicts “punishment.”21 Though it speaks of legislative “purpose,” the more likely and appropriate concern in a historical inquiry is the nature of the measure itself. Even the text and legislative history inquiry of Austin can be understood as going more to the nature of the provision itself rather than the subjective intent of the legislators.
*1258In concluding our discussion of Austin, we must question whether, as some courts have assumed, that case establishes that “punishment” for purposes of one constitutional protection is necessarily “punishment” for another. See United States v. $4.05,089.23 U.S. Currency, 33 F.3d 1210, 1219 (9th Cir.1994) (“We believe that the only fair reading of Austin is that it resolves the ‘punishment’ issue with respect to forfeiture cases for purposes of the Double Jeopardy Clause as well as the Excessive Fines Clause.”), amended on denial of rehearing, 56 F.3d 41 (1995), cert. granted, — U.S.-, 116 S.Ct. 763, 133 L.Ed.2d 709 (1996). This Court, noting the tension between Halper and Austin, has rejected the Ninth Circuit’s reading of Austin as resolving all forfeitures under § 881 as presumptively punishment for purposes of the Double Jeopardy Clause. See United States v. $184,505.01 in U.S. Currency, 72 F.3d 1160 (3d Cir.1995) (rejecting holding and reasoning of United States v. $405,089.23 U.S. Currency, 33 F.3d 1210 (9th Cir.1994)).22
Nevertheless, we believe that the historical methodology of Austin, as opposed to its broad language and holding, must be applicable to other punishment determinations: historical analysis is a staple of constitutional interpretation, including those guarantees dealing with “punishment.” Cf. Nixon v. Administrator of General Services, 433 U.S. 425, 475, 97 S.Ct. 2777, 2806, 53 L.Ed.2d 867 (1977) (examining history to determine whether access restrictions on presidential papers constituted “punishment” for Bill of Attainder Clause); Bell v. Wolfish, 441 U.S. 520, 590 n. 23, 99 S.Ct. 1861, 1901 n. 23, 60 L.Ed.2d 447 (1979) (Stevens, J., dissenting) (The Supreme Court “has probably relied upon historical analysis more often than on any of the other objective factors ... [to] determin[e] whether some government sanction is punitive.”) (citing cases).
4. Department of Revenue v. Kurth Ranch: Objective Purpose and Deterrence
One year after deciding Austin, the Court added another wrinkle in Department of Revenue v. Kurth Ranch, — U.S. -, 114 S.Ct. 1937, 128 L.Ed.2d 767 (1994), announcing that the “no deterrent purpose” rule of Halper and Austin does not apply in all situations. Kurth Ranch held that Montana’s Dangerous Drug Tax violated the Double Jeopardy Clause. The Montana law, which taxed illegal drugs and equipment at rates up to 400 percent, constituted “punishment” because it was “a concoction of anomalies, too far removed in crucial respects from a standard tax assessment to escape characterization as punishment for the purpose of Double Jeopardy analysis.” Id. at -, 114 S.Ct. at 1948. Because Montana levied this tax in a separate proceeding, after the defendants were tried and sentenced, this punishment violated the Double Jeopardy Clause. Id.
Kurth Ranch further expanded on the historical inquiry begun in Austin. It distinguished the rule of Halper-that any deterrent purpose makes a law punishment-on the ground that fines and forfeitures “are readily characterized as sanctions” whereas *1259taxes have typically served the salutary23 purpose of raising revenue. Id. at -, 114 S.Ct. at 1946. Thus, the Court explained, a high tax rate and even a deterrent purpose would not automatically render a tax punitive. Id. at -, 114 S.Ct. at 1947.
The Court then examined whether the particular tax at issue operated in the usual manner of most taxes. It differentiated among taxes with a pure revenue raising purpose, mixed-motive taxes imposed both to deter a disfavored activity and to raise revenue, and taxes imposed upon illegal activities. Pure revenue raising taxes are not “punishment,” the Court said, because they are imposed despite their negative effect on the taxed activity. Id. Even mixed-motive taxes, such as those imposed on cigarette sales, are not “punishment” because the government wishes the activity to continue to the extent that its benefits-including tax revenues-outweigh its harms. However, the Court found that these salutary justifications “vanish when the taxed activity is completely forbidden, for the legitimate revenue-raising purpose that might support such a tax could be equally well served by increasing the fine imposed upon conviction.” Id. The Court held that because a tax on illegal drugs did not operate in the usual manner,' the historically non-punitive purposes of taxes could not insulate this tax from being considered “punishment.” Id. at -, 114 S.Ct. at 1948.
The main significance of the Kurth Ranch limitation is that, at least for measures that have historically served salutary functions, even some deterrent purpose will not render a measure “punishment”: “We begin by noting that neither a high rate of taxation nor an obvious deterrent purpose automatically marks this tax a form of punishment.” Id. at -, 114 S.Ct. at 1946 (emphasis added). In these cases, courts must examine whether the particular measure at issue operates in a “usual” manner consistent with its historically salutary or mixed purposes.24
*1260Kurth Ranch also reemphasizes that at least some negative effect on the defendant does not convert a measure into “punishment.” “We note[ ], however, that whether a sanction constitutes punishment is not determined from the defendant’s perspective, as even remedial sanctions carry the ‘sting of punishment.’ ” Id. at - n. 14, 114 S.Ct. at 1945 n. 14 (citations omitted).
5. California Department of Corrections v. Morales: Effects
Most recently, California Department of Corrections v. Morales, — U.S. -, 115 S.Ct. 1597, 131 L.Ed.2d 588 (1995), contributed two additional elements to the “punishment” analysis: it further shifts the focus from a law’s purpose to its effect, and it establishes that the appropriate “punishment” analysis is flexible and context-dependent. In Morales, the Court rejected an ex post facto challenge to a California statute that decreased a prisoner’s entitlement to parole eligibility hearings. Under the law in effect at the time of the defendant’s crime, he was entitled to parole suitability hearings every year after his initial parole determination. Id. at -, 115 S.Ct. at 1600. The California legislature subsequently amended the law to allow the review board to defer subsequent suitability hearings if (1) the prisoner has been convicted of “more than one offense which involves the taking of a life,” and (2) the board “finds that it is not reasonable to expect that parole would be granted.” Id. (citing Cal.Penal Code Ann. § 3041.5(b)(2) (West 1982)). After finding the defendant unsuitable for parole, the review board invoked this new provision to delay his next suitability hearing for three years. Id.
As with the other cases discussed so far, the Court framed the question as whether the measure “increased the ‘punishment’ attached to respondent’s crime.” Id. at -, 115 S.Ct. at 1601. Rejecting the defendant’s claim that this change constituted “punishment,” the Court distinguished cases holding that legislative changes effectively increasing jail terms violated the Ex Post Facto Clause. Id. Unlike the measures in those cases, the Court said, the statute at issue “creates only the most speculative and attenuated risk of increasing the measure of punishment attached to the covered crimes.” Id. at -, 115 S.Ct. at 1605. The likelihood of parole for those covered-double murderers-is “quite remote.” Id. at -, 115 S.Ct. at 1603. Moreover, the “carefully tailored” authority of the board directs it to delay hearings only when it concludes that the hearings would be of no avail to the prisoner. Id. at -, 115 S.Ct. at 1604.
Morales makes clear that a law can constitute unconstitutional “punishment” because of its effects. The Court leads off its discussion with the declaration that “[t]he legislation at issue here effects no change in the definition of respondent’s crime.” Id. at -, 115 S.Ct. at 1601. The opinion then spends the bulk of its analysis examining the effect of the legislative change on Morales. See id. at -, 115 S.Ct. at 1601-04. In doing so, it concedes that a measure effectively extending a sentence of imprisonment *1261constitutes punishment, presumably regardless of the legislature’s motivation. See id. at -, 115 S.Ct. at 1601 (citing and distinguishing Lindsey v. Washington, 301 U.S. 397, 57 S.Ct. 797, 81 L.Ed. 1182 (1937); Miller v. Florida, 482 U.S. 423, 107 S.Ct. 2446, 96 L.Ed.2d 351 (1987); Weaver v. Graham,, 450 U.S. 24, 101 S.Ct. 960, 67 L.Ed.2d 17 (1981)). Morales concludes that the impact on the prisoner was not great enough to warrant finding an ex post facto violation. “We have long held,” the Court said, “that the question of what legislative adjustments will be held to be of sufficient moment to transgress the constitutional prohibition must be a matter of degree.” Id. at -, 115 S.Ct. at 1603 (internal quotations omitted) (emphasis added).25
Morales also highlights the flexibility of the punishment inquiry. It makes no reference or citation to De Veau, Halper, Austin, or Kurth Ranch at all. This could be read as a rejection of those standards in the ex post facto context, but we think that the better reading of this mere omission in Morales is that the appropriate “punishment” analysis depends on the context. The Court said as much: “[W]e have previously declined to articulate a single ‘formula’ for identifying those legislative changes that have a sufficient effect on substantive crimes or punishments to fall within the constitutional prohibition, and we have no occasion to do so here.” Id. (citation omitted). Morales did not need to discuss Austin and its progeny because the facts in Morales involved imprisonment; the Court needed only to discuss and distinguish the most on-point cases of Lindsey, Weaver, and Miller, supra. And in doing so it looked at negative effects on Morales as “a matter of degree.” Id.
This examination of effects, like the Austin inquiry into history, is necessary to limit what would otherwise be the untenable results of the De Veau subjective purpose inquiry and the Halper means-end calculus. While even a substantial “sting” will not render a measure “punishment,” see Halper, 490 U.S. at 447 n. 7, 109 S.Ct. at 1901 n. 7, Kurth Ranch, — U.S. at - n. 14, 114 S.Ct. at 1945 n. 14, at some level the “sting” will be so sharp that it can only be considered punishment regardless of the legislators’ subjective thoughts. For example, the legislature, with the purest heart(s), could extend the prison sentences of all previously convicted sex offenders for the sole reason of protecting potential future victims. It was simply not understood how dangerous they would be when released, the legislators could truthfully explain, and society would be safe only if sex offenders were kept behind bars. This remedial purpose would thus fully explain the continued incarceration; in the other terms of Halper, the continued imprisonment would be “rationally related” to the goal of protecting vulnerable citizens. But no Justice has ever voted to uphold a statute that retroactively increased the term of imprisonment for a past offense. See Miller v. Florida, 482 U.S. 423, 107 S.Ct. 2446, 96 L.Ed.2d 351 (1987); Weaver v. Graham, 450 U.S. 24, 101 S.Ct. 960, 67 L.Ed.2d 17 (1981).
6. Kennedy v. Mendoza-Martinez: The Inquiry for the Nature of Proceedings
Finally, before attempting a synthesis, we must briefly discuss the test employed by the district court-which was based on Kennedy v. Mendoza-Martinez, 372 U.S. 144, 83 S.Ct. 554, 9 L.Ed.2d 644 (1963)-and explain why we find its approach inappropriate. In that case, the Court held that divesting American citizenship for draft evasion or military desertion was “punishment” requiring the procedural protections of the Fifth and Sixth Amendments: “[T]he Fifth and Sixth Amendments mandate that this punishment cannot be imposed without a prior criminal trial and all its incidents, including indictment, notice, confrontation, jury trial, assistance of counsel, and compulsory process for obtaining witnesses.” Id. at 167, 83 S.Ct. at 567 (emphasis added).
Mendoza-Martinez set forth a multi-factor analysis to determine whether a measure constitutes “punishment” triggering criminal process guarantees:
whether the sanction involves an affirmative disability or restraint, [2] whether it has historically been regarded as puni*1262tive, [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 burden to which it applies is already a crime, [6] whether an alternative purpose to which it may rationally be connected is assignable for it, [7] whether it appears excessive in relation to the alternative purpose.
Id. at 168-69, 83 S.Ct. at 567-68. The district court applied this test in holding that notification under Megan’s Law was unconstitutional.
However, the Supreme Court has made clear that the Mendoza-Martinez test is not controlling for the issues in this case. See Austin, 509 U.S. at 610 n. 6, 113 S.Ct. at 2806 n. 6. Although Mendoza-Martinez used the word “punishment,” Austin explains that the seven factors are properly used to determine whether a proceeding is “so punitive that the proceeding must reasonably be considered criminal” for purposes of Sixth Amendment trial protections. Id. “In addressing the separate question whether punishment is being imposed, the Court has not employed the tests articulated in Mendoza-Martinez and Ward.” Id.
Amicus American Civil Liberties Union (ACLU) makes a clever argument on this point. The Supreme Court has said that Mendoza-Martinez does not control for determinations of whether a civil measure is “punishment.” The ACLU contends that this is because the Mendoza-Martinez “test”-which analyzes whether something is “so punitive” as to invoke criminal trial protections-is harder to prove than the test for mere “punishment.” Logically, if a measure is “so punitive” to satisfy the higher Mendoza-Martinez threshold, amicus argues, it should also be “punishment” for purposes of the challenges Artway brings, even if the reverse is not true.
Nevertheless, like the New Jersey Supreme Court in Doe, 142 N.J. at 63-73, 662 A2d 367, we think it wise to heed the Supreme Court’s advice: Mendoza-Martinez is inapplicable outside the context of determining whether a proceeding is sufficiently criminal in nature to warrant criminal procedural protections of the Fifth and Sixth Amendments.26 See Austin, 509 U.S. at 610 n. 6, 113 S.Ct. at 2806 n. 6. Even when the Court has recited the Mendoza-Martinez factors, including in Mendoza-Martinez itself, it has played them down. See Mendoza-Martinez, 372 U.S. at 167, 83 S.Ct. at 566-67 (declining to apply its own factors). It has consistently *1263insisted that these factors, really a grab-bag of many individual tests, are neither controlling nor dispositive. See United States v. Ward, 448 U.S. 242, 249, 100 S.Ct. 2636, 2641-42, 65 L.Ed.2d 742 (1980) (“[T]his list of considerations, while certainly neither exhaustive nor dispositive, has proved helpful in our own consideration of similar questions and provides some guidance.”) (emphasis added). Finally, we think that a seven factor balancing test — with factors of unknown weight that “may often point in differing directions,” Mendoza-Martinez, 372 U.S. at 169, 83 S.Ct. at 568 — is too indeterminate and unwieldy to provide much assistance to us here.27
B. Synthesizing the Jurisprudence: The Test(s)
Synthesizing these eases, we derive the following analytical framework for this case. A measure must pass a three-prong analysis-(1) actual purpose, (2) objective purpose, and (3) effect-to constitute non-punishment. We must look at actual purpose to see “whether the legislative aim was to punish.” See De Veau, 363 U.S. at 160, 80 S.Ct. at 1155. If the legislature intended Megan’s Law to be “punishment,” i.e., retribution was one of its actual purposes, then it must fail constitutional scrutiny. If, on the other hand, “the restriction of the individual comes about as a relevant incident to a regulation,” the measure will pass this first prong. Id.
If the legislature’s actual purpose does not appear to be to punish, we look next to its “objective” purpose. This prong, in turn, has three subparts. First, can the law be explained solely by a remedial purpose? See Halper, 490 U.S. at 448, 109 S.Ct. at 1901-02. If not, it is “punishment.” Second, even if some remedial purpose can fully explain the measure, does a historical analysis show that the measure has traditionally been regarded as punishment? See Austin, 509 U.S. at 610-11, 113 S.Ct. at 2806. If so, and if the text or legislative history does not demonstrate that this measure is not punitive, it must be considered “punishment.” Third, if the legislature did not intend a law to be retributive but did intend it to serve some mixture of deterrent and salutary purposes, we must determine (1) whether historically the deterrent purpose of such a law is a necessary complement to its salutary operation and (2) whether the measure under consideration operates in its “usual” manner, consistent with its historically mixed purposes. See Kurth Ranch, — U.S. at - - -, 114 S.Ct. at 1946-48. Unless the partially deterrent measure meets both of these criteria, it is “punishment.” If the measure meets both of these criteria and the deterrent purpose does not overwhelm the salutary purpose, it is permissible under Kurth Ranch.
Finally, if the purpose tests are satisfied, we must then turn to the effects of the measure.. If the negative repercussions-regardless of how they are justified-are great enough, the measure must be considered punishment. See Morales, — U.S. at -, 115 S.Ct. at 1603. This inquiry, guided by the facts of decided cases, is necessarily one “of degree.” See id.
We have thus attempted to harmonize a body of doctrine that has caused much disagreement in the federal and state courts. We realize, however, that our synthesis is by no means perfect. Only the Supreme Court knows where all the pieces belong. The Court will, we hope, provide more guidance with its decision in United States v. $405,089.23 U.S. Currency, 33 F.3d 1210 (9th Cir.1994), amended on denial of rehearing, 56 F.3d 41 (1995), cert. granted, — U.S.-, 116 S.Ct. 763, 133 L.Ed.2d 709 (1996), or some other case in the near future. With this qualification in mind, we turn to the application of this test to Megan’s Law.
*1264C. The Registration Provisions of Megan’s Law Evaluated
The registration provisions of Megan’s Law are relatively simple. They require “repetitive and compulsive” sex offenders who have completed a sentence for designated crimes to register with local law enforcement. Because Artway meets these requirements, he must register if he returns to New Jersey. In registering, Art-way must provide information including descriptions of his appearance, his genetic markers, and his residence and work place to the chief law enforcement officer of the municipality in which he chooses to reside. He must periodically confirm his residence and notify law enforcement if he moves. Unlike the notification provisions of Megan’s Law-which would require notice of Artway’s crime, his description, his whereabouts, and, critically, the State’s assessment of his future dangerousness to members of Artway’s community-registration provides this information only to law enforcement agencies. The information is not open to public inspection.
1. Actual Purpose
The first prong of our test asks whether the legislature’s actual purpose was to punish. See De Veau, 363 U.S. at 160, 80 S.Ct. at 1154-55. The only indication of actual legislative intent regarding the enacted version of Megan’s Law is the following statement of purpose in the legislation itself:
1. The Legislature finds and declares:
a. The danger of recidivism posed by sex offenders and offenders who commit other predatory acts against children, and the dangers posed by persons who prey on others as a result of mental illness, require a system of registration that will permit law enforcement officials to identify and alert the public when necessary for the public safety.
b. A system of registration of sex offenders and offenders who commit other predatory acts against children will provide law enforcement with additional information critical to preventing and promptly resolving incidents involving sexual abuse and missing persons.
N.J.S.A 2C:7-1. This passage suggests that the legislature’s actual purpose was not punishment. It speaks of “identify[ing] and alert[ing] the public” to enhance safety and “preventing and promptly resolving incidents.” Protecting the public and preventing crimes are the types of purposes De Veau found “regulatory” and not punitive. 363 U.S. at 160, 80 S.Ct. at 1154-55.
The only other legislative history, a statement in the bill as introduced in the New Jersey Senate, buttresses the conclusion that the legislature’s intent was not to punish. “The danger posed by the presence of a sex offender who has committed violent acts against children requires a system of notification to protect the public safety and welfare of the community.” Senate Bill No. 14 (introduced September 12, 1994). The section literally speaks of “notification,” but if the legislature’s actual purpose in notification was remedial, it is hard to imagine that its purpose in the predicate and less harsh step of registration was punitive.
The circumstances of this enactment, which generated such sparse legislative history, give us pause. Megan’s Law was rushed to the Assembly floor as an extraordinary measure, skipping committee consideration and debate entirely. It is just these “sudden and strong passions to which men are exposed” that the Framers designed the Ex Post Facto and Bill of Attainder Clauses to protect against. Fletcher v. Peck, 10 U.S. (6 Cranch) 87, 137-38, 3 L.Ed. 162 (1810). Nevertheless, the evidence we do have of actual legislative intent points to a non-punitive purpose.
2. Objective Purpose
The objective purpose prong asks three related questions. First, we must discern whether the law can be explained solely by a remedial purpose. See Halper, 490 U.S. at 448, 109 S.Ct. at 1901-02. Registration is a common and long-standing regulatory technique with a remedial purpose. See, e.g., New York v. Zimmerman, 278 U.S. 63, 49 S.Ct. 61, 73 L.Ed. 184 (1928) (registration of membership corporations and associations permissible); United States v. Kahriger, 345 U.S. 22, 73 S.Ct. 510, 97 L.Ed. 754 (1953) *1265(registration of professional gamblers permissible), overruled on other grounds by Marchetti v. United States, 390 U.S. 39, 88 S.Ct. 697, 19 L.Ed.2d 889 (1968); United States v. Harriss, 347 U.S. 612, 74 S.Ct. 808, 98 L.Ed. 989 (1954) (registration of lobbyists permissible).28 One need look no further than the Selective Service to find a nonpuni-tive registration system for individuals. See Gillette v. United States, 401 U.S. 437, 91 S.Ct. 828, 28 L.Ed.2d 168 (1971) (sustaining selective service system against claim that it violated free exercise).
Here, the solely remedial purpose of helping law enforcement agencies keep tabs on these offenders fully explains requiring certain sex offenders to register. Registration may allow officers to prevent future crimes by intervening in dangerous situations. Like the agent who must endure the snow to fetch the soupmeat, the registrant may face some unpleasantness from having to register and update his registration. But the remedial purpose of knowing the whereabouts of sex offenders fully explains the registration provision just as the need for dinner fully explains the trip out into the night. And the means chosen — registration and law enforcement notification only — is not excessive in any way. Registration, therefore, is certainly “reasonably related” to a legitimate goal: allowing law enforcement to stay vigilant against possible re-abuse.
Second, we must consider history, and registration does not resemble punishment through a historical analysis. Artway spends much of his brief chronicling the historical understanding of public shame as punishment. “Early forms of punishment contained strong elements of gross public humiliation _ Physical punishments ... were carried out publicly in ceremonial fashion [because it was] intended that the victim should be humiliated, for degradation figured largely in all contemporary theories of punishment.” Jon A Brilliant, Note, The Modem Day Scarlet Letter: A Critical Analysis of Modern Probation Conditions, 1989 Duke L.J. 1357, 1360-61 (internal quotations omitted); see also Ex Parte Wilson, 114 U.S. 417, 428, 5 S.Ct. 935, 940, 29 L.Ed. 89 (1885) (cataloguing “punishments that consist principally in their ignominy” as set forth in Blackstone’s Commentaries); Crime and Punishment in American History 40 (explaining that humiliating punishments were historically intended to serve as deterrents).
In particular, Artway argues that Megan’s Law is analogous to that most famous badge of punishment: the Scarlet Letter. “There can be no outrage ... against our common nature,-whatever be the delinquencies of the individual,-no outrage more flagrant than to forbid the culprit to hide his face for shame; as it was the essence of this punishment to do.” Nathaniel Hawthorne, The Scarlet Letter 63-64 (Random House 1950) (1850). Like the Scarlet Letter, Artway contends, Megan’s Law results in public ostracism and opprobrium: it would subject him to potential vigilantism, impair his opportunities to work, and damage his abilities to develop and maintain stable relationships. In his submission, its “remedial” purpose-to protect the public from him-seeks to brand him as an outcast. Such a shunning by one’s community is the essence of historical punishment, Artway contends.
Artway’s argument has considerable force, but the notification issue is not before us. We evaluate only registration, and that provision bears little resemblance to the Scarlet Letter. Registration simply requires Artway to provide a package of information to local law enforcement; registration does not involve public notification. Without this public element, Artway’s analogy fails. The Scarlet Letter and other punishments of “shame” and “ignominy” rely on the disgrace of an individual before his community. The act of registering with a discrete government entity, which is not authorized to release that information to the community at large (except in emergencies), cannot be compared to public humiliation. The officers who constitute local law enforcement, even if they are from Artway’s area, would constitute only a de minimis portion of that community. And *1266their ready access to criminal history information is an integral part of their jobs, rather than an extraordinary event likely to trigger opprobrium.
Artway relies on Weems v. United States, 217 U.S. 349, 30 S.Ct. 544, 54 L.Ed. 793 (1910), to establish that even registration is “punishment.” It does not aid his case. Weems struck down as cruel and unusual punishment a Philippine law that imposed horrible punishments for falsification of public documents. Id. at 363, 30 S.Ct. at 547-48. Any false entry, even if unintentional and with no ill effect, triggered the “cadena temporal.” Id. This punishment imposed hard and painful labor for a period from twelve years and a day to twenty years, shaekled at the wrist and the ankle, with no access to family or loved ones, the extinguishment of civil rights while serving the sentence, perpetual disqualification from political rights, such as holding office, and “surveillance.” Id. at 363-64, 30 S.Ct. at 548.
The Weems Court confronted a different issue from the one in this case. The Court held that this harsh punishment as a whole was cruel and unusual for the relatively minor offense involved. Id. at 382, 30 S.Ct. at 555. And the “surveillance” statute that made up a minor part of the total punishment differed from Megan’s Law in at least one significant respect: the unfortunate offender in Weems was required to obtain written permission before he could move. See id. at 363, 30 S.Ct. at 547-48. Given this larger context, the Court’s dictum about the harshness of “surveillance” hardly establishes that registration is “punishment.”
Finally, because registration historically is a regulatory technique with a salutary purpose, any incidental purpose to deter future offenses by past sex offenders will not invalidate it under Kurth Ranch.
3. Effects
The final prong examines whether the effects-or “sting”-of a measure is so harsh “as a matter of degree” that it constitutes “punishment.” See Morales, — U.S. at -, 115 S.Ct. at 1603. The caselaw does not tell us where the line falls that divides permissible from impermissible effects, but we know the “matter of degree” is somewhere between imprisonment and revocation of citizenship on the one hand, and loss of a profession or benefits on the other. Compare Miller v. Florida, 482 U.S. 423, 107 S.Ct. 2446, 96 L.Ed.2d 351 (1987) (increased incarceration is “punishment”) and Trop v. Dulles, 356 U.S. 86, 78 S.Ct. 590, 2 L.Ed.2d 630 (1958) (revoking citizenship is “punishment”) with De Veau v. Braisted, 363 U.S. 144, 80 S.Ct. 1146, 4 L.Ed.2d 1109 (1960) (forbidding work as union official not “punishment”); Hawker v. New York, 170 U.S. 189, 18 S.Ct. 573, 42 L.Ed. 1002 (1898) (revoking medical license is not “punishment”) and Flemming v. Nestor, 363 U.S. 603, 80 S.Ct. 1367, 4 L.Ed.2d 1435 (1960) (terminating social security benefits not “punishment”).29
Artway marshals strong reasons that notification would have devastating effects. In addition to the ostracism that is part of its very design, notification subjects him to possible vigilante reprisals and loss of employment. And unlike the mere fact of his past conviction, which might be learned from an employment questionnaire or public records, notification under Megan’s Law features the State’s determination-based overwhelmingly on past conduct-that the prior offender is a future danger to the community.30 We reemphasize, however, that as forceful as Artway’s arguments seem to be, the issue of notification is not ripe at this time.
On the other hand, registration, the only phase of Megan’s Law upon which we may pass judgment, has little impact. Most of *1267the information is already available in the public record. It is disclosed only to law enforcement, which has ready access to this criminal history. And, unlike notification, the information contains no assessment by the State that Artway is a future danger. Therefore, this impact, even coupled with the registrant’s inevitable kowtow to law enforcement officials, cannot be said to have an effect so draconian that it constitutes “punishment” in any way approaching incarceration. It is less harsh than losing a profession or benefits.
While there doubtless are some unpleasant consequences of registration-it is possible that police will leak information or engage in official harassment-we must presume that law enforcement will obey the law. Moreover, Artway, who of course bears the burden of proof to invalidate a statute on constitutional grounds, presents no evidence in this record of dire consequences flowing from registration.
D. Summary of Registration Claims
Analyzing the registration provisions of Megan’s Law under the (1) actual purpose, (2) objective purpose, and (3) effects prongs of our “punishment” test, we conclude that registration under Megan’s Law does not constitute “punishment” under any measure of the term. Hence, it does not offend the Ex Post Facto, Double Jeopardy, or Bill of Attainder Clauses. Therefore, although our analysis differs from that employed by the district court and the Supreme Court of New Jersey, we agree with their conclusion regarding registration.
YI. EQUAL PROTECTION
Turning to the remainder of Artway’s claims, we begin by rejecting his argument that Megan’s Law violates equal protection. Artway contends that Megan’s Law’s distinction between “compulsive and repetitive” sex offenders and other sex offenders is “arbitrary and discriminatory.” However, the Equal Protection Clause does not forbid all discrimination, and the distinctions made by Megan’s Law are not arbitrary.
The Equal Protection Clause provides that no state shall “deny to any person within its jurisdiction the equal protection of the laws.” U.S. Const. amend. XIV § 1. This is not a command that all persons be treated alike but, rather, “a direction that all persons similarly situated should be treated alike.” City of Cleburne v. Cleburne Living Center, 473 U.S. 432, 439, 105 S.Ct. 3249, 3254, 87 L.Ed.2d 313 (1985) (emphasis added). The level of scrutiny applied to ensure that classifications comply with this guarantee differs depending on the nature of the classification. Classifications involving suspect or quasi-suspect class, or impacting certain fundamental constitutional rights, are subject to heightened scrutiny. Id. Other classifications, however, need only be rationally related to a legitimate government goal. See Chapman v. United States, 500 U.S. 453, 465, 111 S.Ct. 1919, 1927-28, 114 L.Ed.2d 524 (1991) (applying rational basis test to classification based on nature of offense).
Megan’s Law requires persons who have committed their offense and completed all incarceration, probation and parole by the date the Law was enacted to register only if they were found to be “repetitive and compulsive” at sentencing. The challenged category-“repetitive and compulsive sex offenders”-is not a suspect or quasi-suspect class. See Cleburne, 473 U.S. at 439, 105 S.Ct. at 3253-54 (listing classes receiving heightened scrutiny as race, alienage, national origin, and sex). It also does not implicate a fundamental constitutional right.for which the Supreme Court has granted heightened equal protection scrutiny. See Chapman v. United States, 500 U.S. at 465, 111 S.Ct. at 1927-28 (applying rational basis test to classification based on nature of offense). This classification, therefore, “is subject to the general rule that legislation is presumed to be valid and will be sustained if the classification drawn by the statute is rationally related to a legitimate state interest.” Cleburne, 473 U.S. at 440, 105 S.Ct. at 3254 (citing cases).
Registration easily satisfies this requirement. Protecting vulnerable individuals from sexual offenses is certainly a legitimate state interest. Requiring registration of con*1268victed sex offenders found to be “repetitive and compulsive,” as opposed to other sex offenders or the rest of the population, is rationally related to that goal. See, e.g., State v. Wingler, 25 N.J. 161, 176, 135 A.2d 468 (1957) (holding that classification of repetitive and compulsive sex offenders “has a rational basis”); Mahfouz v. Lockhart, 826 F.2d 791, 794 (8th Cir.1987) (applying rational basis test to hold that Arkansas statute excluding sex offenders from work/study release program for inmates did not violate equal protection). The legislature could have rationally concluded that sex offenders who had completed all incarceration, probation and parole had a good chance of reintegrating into their communities and therefore posed a lower risk. Also, realizing that people who had rejoined society had the most to lose, the legislature could have rationally decided to require only “repetitive and compulsive” offenders in this category to register. Thus, this classification does not offend equal protection.
Artway’s reliance on Foucha v. Louisiana, 504 U.S. 71, 112 S.Ct. 1780, 118 L.Ed.2d 437 (1992), as establishing heightened scrutiny in this case is misplaced. Foucha held that a state statute allowing continued confinement of an individual acquitted by reason of insanity, even when that person had ceased to be mentally ill, violated due process. Id. at 78-83, 112 S.Ct. at 1784-88. A plurality indicated that doing so was also an equal protection violation. Id. at 84-85, 112 S.Ct. at 1788. But, unlike Megan’s Law, the statute in Fou-cha denied those subject to it of their physical liberty, which the Court has recognized as a fundamental constitutional right triggering heightened scrutiny. See United States v. Salerno, 481 U.S. 739, 750, 107 S.Ct. 2095, 2103, 95 L.Ed.2d 697 (1987).31
VII. DUE PROCESS
We also reject Artway’s contention that Megan’s Law denies due process by classifying former offenders on the basis of “repetitive and compulsive behavior.” This argument has two subparts. First, Artway argues that requiring him to register on the basis of the “repetitive and compulsive” finding violates due process because the finding was unreliable when made. The supposed unreliability stems from an alleged lack of training of the State employees making these determinations. Second, he contends, holding him accountable for this determination violates due process because he did not have notice at the time of sentencing of the negative implications of this finding. Artway admits that he was advised of his right to contest the “repetitive and compulsive” finding, but contends that such a finding was in his interest because it was his only hope for obtaining treatment and being placed in a treatment center, safe from the general prison population.
Although he does not spell out why using the “repetitive and compulsive” finding against him would amount to a due process deprivation, we will assume he means that such actions would be “fundamentally unfair.” Cf. Daniels v. Williams, 474 U.S. 327, 341, 106 S.Ct. 677, 680, 88 L.Ed.2d 662 (1986) (Stevens, J., concurring) (“Petitioners must show that [the state procedures] contain a defect so serious that we can characterize the procedures as fundamentally unfair, a defect so basic that we are forced to conclude that the deprivation occurred without due process.”).
But even this argument has no merit. We need not reach the issue of the fairness-whether because of unreliability or lack of incentive to oppose-of the “repetitive and compulsive” finding. One must have an interest in life, liberty, or property before due process protections are triggered. U.S. Const. amend. XIV, § 1; see also Goldberg v. Kelly, 397 U.S. 254, 90 S.Ct. 1011, 25 L.Ed.2d 287 (1970). Artway has no such interest in the reputational damage, if any, that accompanies registration. See Doe, 142 N.J. at 103-06, 662 A.2d 367; Paul v. Davis, 424 U.S. 693, 701, 96 S.Ct. 1155, 1160-61, 47 L.Ed.2d 405 (holding mere damage to *1269reputation insufficient to trigger due process); Sturm v. Clark, 835 F.2d 1009, 1012 (3d Cir.1987) (holding harm to reputation and financial interests insufficient to confer liberty interest). And at this stage, the “repetitive and compulsive” finding subjects him to no more than registration.32
Artway may have a liberty interest in notification under state law triggering federal due process protections. See Doe, 142 N.J. at 103-106, 662 A2d 367. But, as explained supra Part IV, his challenges to notification are not yet ripe.
VIII. UNCONSTITUTIONAL VAGUENESS
Artway next argues that Megan’s Law is unconstitutionally vague because it forces him to “guess” whether he is covered by the Act. We disagree. Due process requires only that a penal statute give persons of “common intelligence” fair notice about “what the State commands or forbids.” Lanzetta v. New Jersey, 306 U.S. 451, 453, 59 S.Ct. 618, 619, 83 L.Ed. 888 (1939). While Artway appears to style his complaint as a facial challenge, he has standing only to raise the vagueness of the Act as it applies to him unless he can prove that the Act is vague in substantially all its applications. See Village of Hoffman Estates v. Flipside, Hoffman Estates, Inc., 455 U.S. 489, 494-95, 102 S.Ct. 1186, 1191-92, 71 L.Ed.2d 362 (1982); United States v. Powell, 423 U.S. 87, 92, 96 S.Ct. 316, 319-20, 46 L.Ed.2d 228 (1975). Because Artway does not, and cannot, establish that Megan’s Law is vague in substantially all its applications, we deal only with the provisions of Megan’s Law as they apply to him.33
Under the relevant section of Megan’s Law, a person who has been convicted of a “sex offense” must register. Paragraph (1) of that section defines “sex offense” to include “aggravated sexual assault, sexual assault ... if the court found that the offender’s conduct was characterized by a pattern of repetitive and compulsive behavior.” Paragraph (3) of that section further defines “sex offense” to include “a sentence on the basis of criteria similar to the criteria set forth in paragraph (1) ... entered or imposed under the laws of the United States, this state or another state.” N.J. Stat. Ann. 2c:7-1b. (emphasis added).
The crux of Artway’s argument is that the “sentenced on the basis of criteria similar to” language violates due process by not specifying the predicate crimes more clearly. But Artway’s duty to register is patent under the Act. Megan’s Law requires registration for those sentenced under “similar criteria”, to “aggravated sexual assault, sexual assault ... if the court found that the offender’s conduct was characterized by a pattern of repetitive and compulsive behavior.” N.J. Stat. Ann. 2c:7-1b. Thus, Artway must register if he was sentenced for engaging in (1) “sexual assault” and (2) “repetitive and compulsive” behavior.
Artway argues (in so many words) that, because the crime of sodomy did not require an element of violence at the time he was convicted, it is unclear whether he falls under the “sexual assault” requirement. The statutory elements of the crime, however, are a *1270red herring. Artway was sentenced under New Jersey’s prior sex offender law, which required a finding of both “violence” and “repetitive and compulsive behavior.” In particular, the sentencing judge made a finding at sentencing that Artway used violence to perpetrate a sexual act. See Artway v. Pallone, 672 F.2d 1168, 1170-71 & n. 3 (3d Cir.1982) (describing crime). This is plainly sexual assault. The sentencing judge also found that Artway had engaged in “repetitive and compulsive” behavior. See id. Thus, Artway received “a sentence on the basis of criteria” similar to both elements of paragraph (1). Because the statute facially applies to Artway, he could reasonably know of his duty to register.
Artway’s citation to Hluchan v. Fauver, 482 F.Supp. 1155 (D.N.J.1980), is of no avail. Even if we found the reasoning of Fauver persuasive, it is inapposite. That case involved an equal protection challenge to regulations that classified prisoners for purpose of minimum custody eligibility. Id. at 1156. The rationality of the open-ended regulations was at issue, not the fair notice question of this vagueness claim (a due process challenge). Moreover, the regulation in Fauver was objectionable for three reasons not present in this case. First, unlike Megan’s Law, the Fauver regulations initially left “sex offense” completely undefined (thus undermining its rationality). Id. at 1157. Second, the regulations contained a provision not present in Megan’s Law for which the court was unable to “determine the meaning.” Id. Finally, the incorporation of “sex offenses” from other states presented an equal protection problem, the court thought, because “the danger exists that individuals convicted of the same criminal conduct in different jurisdictions will be treated differently.” Id. We doubt the soundness of finding equal protection violations on the basis of “dangers” that have not come to pass, but, in any event, Artway was convicted under the laws of the State of New Jersey and faces no problem with the applicability of laws of other states.
IX. PULLMAN ABSTENTION
Finally, we conclude that the district court did not err in refusing to abstain under Railroad Commission v. Pullman, 312 U.S. 496, 61 S.Ct. 643, 85 L.Ed. 971 (1941). Pullman abstention allows federal courts, in rare cases, to abstain from deciding a case if a state court’s resolution of a state law issue would obviate the need for the federal court to reach a federal constitutional issue. The doctrine attempts to avoid constitutional questions and promote principles of federalism. However, Pullman abstention “is an extraordinary and narrow exception to the duty of a District Court to adjudicate a controversy properly before it [which] can be justified ... only in exceptional circumstances.” Colorado River Water Conservation Dist. v. United States, 424 U.S. 800, 813, 96 S.Ct. 1236, 1244, 47 L.Ed.2d 483 (1976) (quoting Allegheny County v. Frank Mashuda Co., 360 U.S. 185, 188-89, 79 S.Ct. 1060, 1062-63, 3 L.Ed.2d 1163 (1959)).
Under our jurisprudence, a district court must make three findings in order to justify the Pullman exception to the general rule that federal courts must hear cases properly brought within their jurisdiction. The Court must find (1) that uncertain issues of state law underlie the federal constitutional claims brought in the district court; (2) that the state law issues are amenable to a state court interpretation that would obviate the need for, or substantially narrow, adjudication of the federal claim; and (3) that important state policies would be disrupted through a federal court’s erroneous construction of state law. See Chez Sez III Corp. v. Township of Union, 945 F.2d 628, 631 (3d Cir.1991). If all three factors are present, the federal court must then consider whether abstention is appropriate by weighing such factors as the availability of an adequate state remedy, the length of time the litigation has been pending, and the impact of delay on the litigants. Id. at 633.
Abstention is not warranted here. First, although a state law issue-whether Megan’s Law applies to him-underlies the federal constitutional claim, this issue is not “uncertain” because Megan’s Law clearly applies to him. See Part VIII supra. Second, because the applicability of Megan’s Law to Artway is patent, this issue is not “amenable” to a state law determination that *1271would obviate the need for a federal constitutional determination. The Supreme Court has used various formulations to describe “amenability,”34 but no matter which we adopt, the lack of uncertainty about the state law issue precludes satisfaction of this prong: a certain issue is not “amenable” to a contrary interpretation. The third factor-whether an improper interpretation of state law would disrupt important state policies-favors the state because the scope of Megan’s Law is an important state issue.35 Nevertheless, two of the three essential factors for abstention are lacking even before we come to the weighing factors; hence, Pullman abstention is inappropriate.
X. CONCLUSION
For the foregoing reasons, we hold that the lion’s share of Artway’s claims are unripe. In particular, we will dismiss as unripe Artway’s claims (1) that the notification provisions of Megan’s Law violate the Ex Post Facto, Bill of Attainder, or Double Jeopardy Clauses of the U.S. Constitution; and (2) that the State must provide Artway more process for receiving notice of and challenging the notification determination. We also hold unripe the claim of the Chief of Police of Wood-bridge Township that state immunity bars his “potential liability” for a hypothetical § 1983 action seeking damages.
With regard to Artway’s claims that are currently justiciable, we hold that (1) the registration component of Megan’s Law does not violate the Ex Post Facto, Double Jeopardy or Bill of Attainder Clauses as impermissible “punishment”; (2) the “repetitive and compulsive” classification of Megan’s Law does not offend equal protection; (3) the alleged unreliability and unfairness of Art-way’s “repetitive and compulsive” determination does not violate due process; and (4) Megan’s Law is not unconstitutionally vague as applied to him. Finally, we hold that the district court did not err in refusing to abstain under Pullman.
The judgment of the district court will be vacated insofar as it enjoins the enforcement of Tier 2 and Tier 3 notification under Megan’s Law, and affirmed insofar as it holds the registration provisions (including Tier 1) of the Law constitutional. The parties shall bear their own costs.
. The victim testified that Artway and two friends took her to a wooded area, stripped her, tied her to a tree, urinated on her, forced her to pose nude for photographs, and sodomized her for over an hour. See id.
. The complete list of categories is as follows: (1) Degree of Force; (2) Degree of Contact; (3) Age of Victim; (4) Victim Selection; (5) Number of Offenses/Victims; (6) Duration of Offensive Behavior; (7) Length of Time Since Last Offense; (8) History of Anti-Social Acts; (9) Response to Treatment; (10) Substance Abuse; (11) Therapeutic Support; (12) Residential Support; and (13) Employment/Educational Stability. Id.
. It also included To Kill a Mockingbird and Plato’s Dialogues in its discussion of what constitutes "punishment."
. Because Artway has never submitted to even the registration provisions of Megan’s Law, the flier is not the result of notification. Rather, Artway’s notoriety seems to have flowed from this litigation.
.One aspect of the propriety of Pullman abstention-i.e., whether important state policies would be disrupted by a federal court ruling-is reviewed for abuse of discretion. See Biegenwald v. Fauver, 882 F.2d 748, 750-51 (3d Cir.1989). However, this distinction plays no part in *1246our analysis. See infra note 34 and accompanying text.
. Mootness also contains four major exceptions: (1) wrongs that have collateral consequences, see Sibron v. New York, 392 U.S. 40, 53, 88 S.Ct. 1889, 1897-98, 20 L.Ed.2d 917 (1968); (2) wrongs that are capable of repetition yet evading review, see Roe v. Wade, 410 U.S. 113, 93 S.Ct. 705, 35 L.Ed.2d 147 (1973); (3) wrongs that are voluntarily ceased but could resume, see United States v. W.T. Grant Co., 345 U.S. 629, 73 S.Ct. 894, 97 L.Ed. 1303 (1953); and (4) wrongs to a class that continue though those to the named plaintiffs do not, see Sosna v. Iowa, 419 U.S. 393, 95 S.Ct. 553, 42 L.Ed.2d 532 (1975). These exceptions are not directly applicable here, but they further demonstrate how mootness doctrine has diverged from standing doctrine to allow courts to decide real controversies in the face of changing circumstances.
. We have sometimes employed a three-part test for ripeness in the declaratory judgment context: (1) adversity of interest; (2) conclusivity; (3) utility. See Step-Saver Data Systems, Inc. v. Wyse Technology, 912 F.2d 643, 647 (3d Cir.1990). However, the Supreme Court’s two-part test is of course still good law, and we continue to use that formulation as well. See, e.g., New Hanover Tp. v. United States Army Corps of Engineers, 992 F.2d 470 (3d Cir.1993). We deem the two-part analysis more apt for this case.
. Similarly, we cannot rule on the claim of the Chief of Police of Woodbridge Township that state immunity bars his “potential liability” for a hypothetical § 1983 action seeking damages. Artway has filed no such suit. To the extent the Police Chief's defense relates to the attorney's fees Artway is seeking, the Eleventh Amendment has no application to the award of attorneys fees under 42 U.S.C. § 1988. See Missouri v. Jenkins, 491 U.S. 274, 109 S.Ct. 2463, 105 L.Ed.2d 229 (1989).
. We recognize that some of the critical factual information, especially the effects of the proposed notification on the registrant, will be difficult to chronicle. In most cases, we assume that registrants slated for notification will seek to enjoin the notification before it happens. The actual consequences of notification on that person, of course, cannot be known at that point. Therefore, we wish to emphasize that our holding that this case is not ripe does not mean that all pre-notification challenges will be unripe. Where the fact of notification is not speculative (because the state has expressed its intent to notify), the district court enjoys flexibility to collect appropriate evidence so that the issue may be fit for judicial review. District courts may see fit to admit a broad array of evidence, including but not limited to (1) threats or actions against the registrant triggered by notice from channels other than Megan’s Law, (2) threats or actions against similarly situated registrants, especially those undergoing notification, and (3) studies of the. effects of Megan’s Law or similar notification laws. We do not suggest, however, that evidence of community reaction is mandatory before a notification challenge will be fit for judicial review.
. In addition to telling Woodbridge residents to "BEWARE,” the flier suggests that anti-registrant leafletting will be a regular result of notification:
ATTENTION: Two time convicted rapist Art-way, a 49 year old resident of the Avenel section of Woodbridge has successfully challenged Megan’s Law. After serving an 18 year sentence for sodomy, [sic] He cannot bé made the subject of communily notification. Mr. Artway said “yahoo, I jump up in the air and click my heels. I can now move to another area-in other words I can retreat-and no fliers wifi foflow me.”
(A247). The flier concludes by urging Wood-bridge residents to “keep an eye on Alexander Artway (track his movements)” and requests anonymous "information about his whereabouts.” Id.
The State of Washington study, reports that, of the 176 sex offenders who were subject to notification in that state between 1990 and 1993, 14 have suffered acts of harassment. These incidents include the following: rock and egg throwing, threats of arson, picketing, posting warning fliers throughout the community, and spray painting slogans like "Die, baby raper” and "Move or die” on the notification subject’s home and personal property. (A178). In half of the 14 cases, the harassment also extended to members of the offender's family, or to people living with the offender.
.Although forty states have sex offender registration statutes, twenty-nine of these laws have been passed since 1990. See Simeon Schopf, "Megan's Law": Community Notification and the Constitution, 29 Colum. J.L. & Soc. Probs. 117, 120 (1995). Moreover, of the minority of states whose laws permit community notification, New Jersey’s is the most far-reaching. See id.; Doe, 142 N.J. at 41 n. 9, 662 A.2d 367.
. Furthermore, the state court has not yet interpreted this standard. To the extent state court interpretation would make the standard comport with due process, abstention would probably be appropriate even if the issue were ripe. See Railroad Commission v. Pullman, 312 U.S. 496, 61 S.Ct. 643, 85 L.Ed. 971 (1941). We assume that Artway will be entitled to notice, since his whereabouts seem to be known, so long as he does not pose an immediate danger.
. Artway’s contention at oral argument that his challenge is both facial and as-applied does nothing to overcome his ripeness problem. In the limited context of the First Amendment, a facial challenge allows a litigant to argue that a law is unconstitutional-in a set of circumstances not necessarily present in his own case-on the basis of its “overbreadth." See United States v. Salerno, 481 U.S. 739, 745, 107 S.Ct. 2095, 2100, 95 L.Ed.2d 697 (1987). Artway’s challenge obviously does not rely on the First Amendment. To make a successful facial challenge in a non-First Amendment context, a litigant "must establish that no set of circumstances exists under which the Act would be valid.” Id. Artway has made no contention, let alone proved it, that notification under Megan's Law would be unconstitutional under all circumstances. For example, his "punishment” claims, which all rely on some notion of retroactivity, would fail if the sex offender committed his crime after Megan's Law was enacted. In any event, a facial challenge *1253does not-and cannot-excuse basic Article III case or controversy requirements, such as that the plaintiff actually be aggrieved by the challenged statute.
. Underlining the importance of these clauses in the eyes of the Framers, the Bill of Attainder and Ex Post Facto Clauses apply to both the federal government and the states by the original terms of the Constitution. See U.S. Const. art. I, § 9, cl. 3 (“No Bill of Attainder or ex post facto law shall be passed."); U.S. Const. art. I, § 10 (“No State shall ... pass any Bill of Attainder [or] ex post facto Law_").
. While even Artway’s ex post facto claim fails, we think that it is probably his best challenge. Bills of attainder inflict punishment “without a judicial trial.” Brown, 381 U.S. at 448-49, 85 S.Ct. at 1714-15. Artway, of course, has had a trial, at which he was convicted of the crime triggering registration. The real complaint is not that the legislature has circumvented the judicial process, but that it has changed the results of that process. This is the essence of an ex post facto challenge.
Double jeopardy is probably a stronger challenge than the bill of attainder claim, but it too has its drawbacks. Ex post facto laws are particularly objectionable because they deprive their object of all notice. See, e.g., Miller v. Florida, 482 U.S. 423, 429, 107 S.Ct. 2446, 2450-51, 96 L.Ed.2d 351 (1987); Weaver v. Graham, 450 U.S. 24, 30, 101 S.Ct. 960, 965, 67 L.Ed.2d 17 (1981). In contrast, the Double Jeopardy Clause bars the imposition of a second "punishment” in a separate proceeding even though the punishment was authorized at the time of the crime but just not sought at the same time as the first punishment. In addition, seven judges of the Ninth Circuit have recently pointed out the practical problem with broadly interpreting forfeitures as constituting “punishment” for double jeopardy purposes: those who forfeit illegal proceeds at the time of their arrest cannot be criminally prosecuted. See United States v. $405,089.23 U.S. Currency, 56 F.3d 41, 42 (9th Cir.1995) (Rymer, J., with whom Hall, Wiggins, Kozinski, O'Scannlain, Trott, and Nelson, J.J., join, dissenting from denial of rehearing). Given these equitable and practical factors, courts may be more reluctant to deem measures "punishment” in a double jeopardy challenge, especially to the extent they must make difficult judgment calls under the test described infra.
Indeed, at least one Justice has noted the equitable factor in arguing that double jeopardy does not bar punishing twice. See United States v. Hess, 317 U.S. 537, 555, 63 S.Ct. 379, 389-90, 87 L.Ed. 443 (1943) (Frankfurter; J., concurring) (“The short of it is that where two such proceedings merely carry out the remedies which Congress has prescribed in advance for a wrong, they do not twice put a man in jeopardy for the same offense.”). Of course, Justice Frankfurter’s position has not carried the day. But two current Justices have recently expressed their view *1254that the Double Jeopardy Clause does not apply to multiple punishments. See Department of Revenue v. Kurth Ranch, - U.S. -, ---, 114 S.Ct. 1937, 1955-59, 128 L.Ed.2d 767 (1994) (Scalia, J., with whom Thomas, J., joins, dissenting) (" ‘To be put in jeopardy' does not remotely mean 'to be punished,’ so by its terms this provision prohibits, not multiple punishments, but only multiple prosecutions.").
. Seemingly inconsistent language in Halper has perplexed some courts. Therefore, we explain in the margin how we think all the parts fit together. Halper declared:
*1255From these premises, it follows that a civil sanction that cannot fairly be said solely to serve a remedial purpose, but rather can only be explained as also serving either retributive or deterrent purposes, is punishment, as we have come to understand the term. We therefore hold that under the Double Jeopardy Clause a defendant who already has been punished in a criminal prosecution may not be subjected to an additional civil sanction to the extent that the second sanction may not fairly be characterized as remedial, but only as a deterrent or retribution.
490 U.S. at 448-49, 109 S.Ct. at 1901-02 (citations and internal quotations omitted) (emphasis added).
On an initial reading, the first “solely” clause of the first sentence and the second "only” sentence seem to point in a different direction than the “only be explained as also serving either retributive or deterrent purposes” language on which we base our analysis.
But the various parts of this excerpt can be reconciled; indeed they must since the majority in Halper certainly thought its declarations in this passage were consistent. As we illustrate with our subsequent soupmeat hypothetical, a measure is “punishment” if it can "only be explained as also serving either retributive or deterrent purposes.” In other words, if the measure is excessive in relation to its proffered remedial purpose, it will be “punishment.” The second sentence says the same thing if one focuses on the “fairly be characterized" language. A measure may not "fairly be characterized as remedial,” but rather may fairly be characterized "only as a deterrent or retribution" if it can "only be explained as also serving either retributive or deterrent purposes.” And the first "solely” part of the first sentence, like the second sentence, can be reconciled with the rest of the paragraph by focusing on the words "fairly be said" (as opposed to just “be said”) and “serve that purpose” (as opposed to have that effect).
This reading of the paragraph is consistent with the other language in the opinion (such as its “rational relation" discussion), the analysis of the case, and its holding: that the fine in question was punishment to the extent it vastly exceeded the government's remedial purpose-recouping its costs of prosecution-because such an excessive fine can only be explained as also serving either deterrent or retributive purposes.
. In his concurrence, Judge Shadur intimates that we may have overresolved Halper. See [Op. at 1272]. We disagree. Our task, we believe, is to derive a general rule from the Supreme Court's precedents and apply it to the facts of this case, not tailor a specific rule to the facts. We also disagree with Judge Shadur’s view that the “rule for a rare case” language of Halper limits the general ends-means test of that case. Id. As the words Judge Shadur quotes make clear, the “rule” is not the general Halper calculus, but the holding of that case: that only under the extreme factual circumstances of Halper does a fixed-penalty provision constitute “punishment” under the general means-ends test. In any event, we agree with Judge Shadur that our differences in this complex case are small indeed.
. Even though Halper was a double jeopardy case, its move away from subjective purpose should apply to ex post facto and bill of attainder claims as well. The Court explained that the subjective approach was appropriate in "identifying the inherent nature of a proceeding, or in determining the constitutional safeguards that must accompany those proceedings.” 490 U.S. at 447, 109 S.Ct. at 1901; see also infra pages 1261-62 and accompanying notes (discussing cases inteipreting this different protection under the Fifth and Sixth Amendments). However, the Court continued, “the approach is not well suited to the ‘humane interests’ safeguarded by the Double Jeopardy Clause's proscription of multiple punishments." Id. The Ex Post Facto and Bill of Attainder Clauses, of course, implicate the same "humane interests” as double jeopardy protections. The move to a more objective analysis, therefore, is better understood as a change of approach than as resting on any fundamental difference in the nature of double jeopardy, ex post facto, and bill of attainder protections.
.In moving past exclusive reliance on subjective legislative intent, the Court partly heeded the admonition of Justice Frankfurter, expressed almost half a century earlier, that such "dialectical subtleties” were an unworkable approach to "punishment” jurisprudence. See United States v. Hess, 317 U.S. 537, 554, 63 S.Ct. 379, 389-90, 87 L.Ed. 443 (1943).
. Thus, Halper and Austin are somewhat in tension. Halper, examining the proportionality of the fine in question to the government’s costs, held that a fine was "punishment” only to the extent it was disproportionate to the government’s costs. 490 U.S. at 448-49, 109 S.Ct. at 1901-02. Austin, relying primarily on history and looking at the statute as a whole (rather than the particular forfeiture in question), holds that forfeiture is “punishment” regardless of its proportionality to the government’s costs. 509 U.S. at 619-23 & n. 14, 113 S.Ct. at 2811-12 & n. 14.
After a cursoiy attempt to distinguish Halper in a footnote, Austin explains that it makes' “little practical difference whether the Excessive Fines Clause applies to all forfeitures under [the relevant statute] or only to those that cannot be characterized as purely remedial." Id. "The Clause prohibits' only the imposition of 'excessive' fines,” the Court explained, "and a fine that serves purely remedial purposes cannot be considered ‘excessive’ in any event.” Id. This may be so, but it collapses Austin's Eighth Amendment analysis into Halper's double jeopardy inquiry: “punishment” is not excessive if it is not “punishment."
. This transition was presaged in United States v. One Assortment of 89 Firearms, 465 U.S. 354, 104 S.Ct. 1099, 79 L.Ed.2d 361 (1984). In that case, the Court held that double jeopardy did not bar a civil proceeding seeking forfeiture of firearms after the owner was acquitted in a separate criminal proceeding. Though the Court still placed decisive weight on actual legislative purpose, it also inquired “whether the statutory scheme was so punitive in purpose or effect as to negate that intention.” Id. at 362-63, 104 S.Ct. at 1105 (emphasis added).
. . While this Court has rejected $405,089.23, we do not agree entirely with the reasoning of the Ninth Circuit dissenters from the denial of rehearing. The dissent criticizes the opinion as merging "the inquiry for excessive fines cases-whether the amount forfeited is partly punishment-into double jeopardy cases, where the issue is whether the amount forfeited is entirely punishment." 56 F.3d at 43.
This is incorrect. Austin adds a historical analysis (and examines the statute as a whole rather than the specific measure in question), but it does not change the underlying nature of the Halper calculus. In fact, Austin follows its statement that it must determine whether this forfeiture serves "in part to punish” by quoting the standard from Halper and citing that case. If this were not clear enough, the Court explains the "relevant question" again later in the opinion as being the Halper analysis: “Under United States v. Halper, 490 U.S. 435, 448, 109 S.Ct. 1892, 1901, 104 L.Ed.2d 487 (1989), the question is whether the forfeiture serves in part to punish, and one need not exclude the possibility that forfeiture serves other purposes to reach that conclusion.” 509 U.S. at 618 n. 12, 113 S.Ct. at 2810 n. 12. The question is not whether the measure is "partly punishment” or "entirely punishment"; the question is whether it is "punishment." And a measure that serves in part to punish (as opposed to merely having some negative effect) is "punishment.” The Halper calculus is admittedly somewhat confusing, but we have done our best to explicate it above. See supra pages 1254-56 and accompanying notes.
. We use the term “salutary” to include both remedial and otherwise beneficial goals.
. Thus, we disagree with the First Circuit's understanding of Kurth Ranch and Halper in situations involving neither fines nor taxes. See United States v. Stoller, 78 F.3d 710 (1st Cir.1996). Stoller argues that Kurth Ranch supplies the general rule-which Stoller dubs the "totality of circumstances" test-while Halper is an "exception” for "monetary" penalties.
We are unpersuaded by Stoller's limitation of Halper. Cabining Halper to “monetary" penalties is not supported by the broad language of that case. Reading nothing in Halper so strictly limiting it, we are loath to read it so narrowly without instruction from the Supreme Court. Stoller claims that the Supreme Court gave such an instruction in Kurth Ranch. But we read nothing in Kurth Ranch indicating that it supplies the general rule and Halper provides the exception. The majority opinion in Kurth Ranch, quoting Chief Justice Rehnquist’s dissent therein, explains that because "tax statutes serve a purpose quite different from civil penalties, ... Hal-per’s method.of determining whether the exaction was remedial or punitive 'simply does not work in the case of a tax statute.'" -U.S. at -, 114 S.Ct. at 1948. If so, then why not read Kurth Ranch as an "exception” for tax cases? What makes Kurth Ranch the general rule and Halper the exception in cases involving neither fines nor taxes? We believe that the better course when evaluating a measure that is neither a “civil penalty” nor a "tax” is to synthesize both Halper and Kurth Ranch and generalize them to the extent their language will support.
We think that Stoller’s limited reading of Hal-per may stem from a misunderstanding of the Halper calculus. Stoller states that, unlike monetary sanctions, many non-monetary sanctions "cannot fairly be characterized as serving only punitive purposes.” 78 F.3d.at 718, (emphasis added). It thus suggests, incorrectly, that this is what Halper requires. As footnote 16 of our opinion describes in detail, a measure constitutes "punishment” under Halper if it may "fairly be characterized only as a deterrent or.retribution.” 490 U.S. at 449, 109 S.Ct. at 1902. The accurate placement of the "only”-modifying “characterized” instead of “punitive”-changes the meaning of that phrase entirely, making the test much less strict than the First Circuit reads it.
We also disagree with Stoller's rationale that Halper is limited to "monetary" penalties because only “fines, forfeitures, and other monetary penalties ... are quantifiable in actual or approximate monetary terms.” 78 F.3d at 717. While judging the proportionality of ends to means may be slightly more difficult in a non-monetary setting, courts compare qualitative means to qualitative ends all the time. Courts regularly use ends-means analysis in equal protection and due process cases to evaluate difficult-to-quantify rights (liberty, free speech, free exercise) like the one at issue here. The feasibility of applying Halper generally is demonstrated by our soupmeat hypothetical, as well as the many cases that have used the calculus to determine the constitutionality of revoking drivers' licenses for drunk driving. See, e.g., Maryland v. Jones, 340 Md. 235, 666 A.2d 128 (1995), cert. denied, - U.S. -, 116 S.Ct. 1265, 134 L.Ed.2d 213 (1996).
*1260Furthermore, we think that Stoller's strict limitation of Halper is inconsistent with Stoller's own approach. Indeed, after arguing for Halper's inapplicability to its own case, Stoller itself proceeds to apply Halper. See 78 F.3d at 723 ("Halper expressly recognizes that civil sanctions need not be precisely calibrated in order to survive scrutiny under the Double Jeopardy Clause as long as they work 'rough remedial justice.’ We think this principle is fully transferable to the debarment context.”) (citation omitted).
Stoller’s attack on Halper is also unnecessary to its result. Even under our approach, which uses Halper to analyze proportionality as part of a larger test, the limited debarment order challenged in Stoller would not constitute "punishment.” We would be hard pressed to conclude otherwise in view óf the Supreme Court's decision in the factually similar cases of De Veau and Hawker v. New York, 170 U.S. 189, 18 S.Ct. 573, 42 L.Ed. 1002 (1898).
Nevertheless, our approach differs from the First Circuit's in our application of Halper to this situation. Given our broader test incorporating DeVeau, Austin, Kurth Ranch, and Morales, we do not think that exclusive reliance on Halper is proper. But Halper is not inapplicable, "dysfunctional," or particularly strict in non-monetary settings such as this. And Stoller's "totality of the circumstances” test, which it purports to extract from Kurth Ranch, is neither described as such by that opinion nor sufficiently determinate to be helpful (like the similar Kennedy v. Mendoza-Martinez, 372 U.S. 144, 83 S.Ct. 554, 9 L.Ed.2d 644 (1963), test rejected by the Supreme Court for Double Jeopardy analysis).
. We discuss the benchmarks for evaluating the ‘matter of degree” infra page 1266.
. Although the New Jersey Supreme Court recognized in Doe that Mendoza-Martinez does not apply to this analysis, we disagree with that court's approach insofar as it failed to take this recognition to its logical conclusion (in addition to its neglect of history under Austin and its total disregard of effects). The Doe Court notes that Mendoza-Martinez does not apply to the relevant "punishment" analysis, but continues to rely on other authorities that, like Mendoza-Martinez, pertain to the question of whether a proceeding is sufficiently criminal in nature to warrant protection under the Fifth and Sixth Amendments.
For example, although the Doe Court nominally applies the Halper and Austin tests, it loads its analysis with the assertion that "[w]here the stated legislative intent is remedial, the burden on those claiming there is a hidden punitive intent is the ‘clearest proof' of that intent.” Doe, 142 N.J. at 62, 662 A.2d 367 (citing United States v. Ward, 448 U.S. 242, 248-49, 100 S.Ct. 2636, 2641-42, 65 L.Ed.2d 742 (1980); Flemming v. Nestor, 363 U.S. 603, 617, 80 S.Ct. 1367, 1376, 4 L.Ed.2d 1435 (1960)). Ward, like Mendoza-Martinez, involves the different question whether a proceeding is effectively criminal so that the procedural protections of the Fifth and Sixth Amendments must apply. Ward, therefore, is as inapplicable to this analysis as Mendoza-Martinez itself. And Flemming was decided in the "actual purpose” era of De Veau v. Braisted, 363 U.S. 144, 80 S.Ct. 1146, 4 L.Ed.2d 1109 (1960) (decided the same year). Halper has since made clear that "the labels 'criminal' and 'civil' are not of paramount importance.” 490 U.S. at 447, 109 S.Ct. at 1901. Austin, Kurth Ranch, and Morales have further changed the analysis, sensibly we think, to include an increasing focus on objective, effect-oriented aspects of the measure in question.
The inapplicability of Mendoza-Martinez also refutes New Jersey’s argument concerning United States v. Salerno, 481 U.S. 739, 107 S.Ct. 2095, 95 L.Ed.2d 697 (1987). New Jersey argues that Salerno establishes that even preventive detention does not offend the Ex Post Facto Clause. Salerno held that preventive detention, before a trial, was not pre-trial "punishment” in violation of the Due Process Clause. Id. at 755, 107 S.Ct. at 2105-06. The Court reached this conclusion through application of the Mendoza-Martinez test. Id. at 747, 107 S.Ct. at 2101-02. Salerno, therefore, sheds little light on the test that we must apply in the context of an ex post facto inquiry.
. Even if we were to apply the Kennedy v. Mendoza-Martinez factors, they do not support a determination that registration constitutes punishment. Only one factor points toward punishment: whether the burden applies to conduct that is already criminal. The other six point toward non-punishment. Even factor (3)-whether the burden is imposed only after proof of scienter (criminal intent)-militates against a finding of “punishment” for registration because Megan’s Law also applies to those judged not guilty by reason of insanity. See N.J.S.A. 2C:7-2a.
. Lambert v. California, 355 U.S. 225, 78 S.Ct. 240, 2 L.Ed.2d 228 (1958), invalidated a registra-, tion statute, but for the different reason that it gave no notice. The registrant in that case did not bring the punishment-oriented claims that Artway makes, apparently because her facts would not support the other predicates of those challenges (e.g., she committed her offense after the enactment of that act).
. Of course, insofar as De Veau, Hawker, and Flemming undertook only an actual purpose test, they are methodologically incomplete compared with current law on “punishment.” Nevertheless, because these cases have not been overruled, we must try to read them consistently with current law. To do so, the measures challenged in these cases must survive the subsequent objective purpose and effect tests. We presume, therefore, that these cases must provide benchmarks for permissible effect.
. Past criminal conduct is the basis of 90 of the possible 111 points in the Registrant Risk Assessment Scale.
. Moreover, the heightened scrutiny the plurality hints at-"the State must have a particularly convincing reason," id. at 85, 112 S.Ct. at 1788-was probably unnecessary to decide the equal protection issue: The classification of insanity acquittees was so underinclusive that it was not even rational.
. We express no opinion regarding whether Art-way may be able to challenge his "repetitive and compulsive” finding at a notification hearing if, in fact, he is ever slated for notification. We also express no opinion about whether Artway may be able to avoid registration, on a basis other than due process, if he can prove that the "repetitive and compulsive" finding was never valid. Finally, we do not opine on the related question-which is not now posed by Artway in these terms-whether he may on some theory be able to avoid (or to terminate the need for) registration if he can prove that the original "repetitive and compulsive” finding no longer has any current validity.
. We also decline to address Artway's argument, made in his brief to this Court, that Megan’s Law does not apply to him as a matter of New Jersey law. We almost certainly cannot grant Artway’s requested relief-an injunction against state officials from enforcing this law-on this basis. See Pennhurst State Sch. and Hosp. v. Halderman, 465 U.S. 89, 104 S.Ct. 900, 79 L.Ed.2d 67 (1984) (holding that the Eleventh Amendment bars federal courts from enjoining state officials from violating state law). Moreover, confronted with the Pennhurst problem, Artway has disclaimed this claim for relief, going so far as to insist at oral argument and in a subsequent letter memorandum that it was never his intention to seek such relief from this Court.
. See Biegenwald v. Fauver, 882 F.2d 748, 752 n. 3 .(quoting 17A Charles A. Wright et at, Federal Practice and Procedure § 4242, at 42-44 (1988)). A leading commentator has interpreted the Supreme Court’s typical formulation of amenability-that the law be “fairly subject” to a state court interpretation eliminating the constitutional issues-as establishing a fairly high threshold requiring a "substantial possibility” that a state interpretation would obviate the need for a federal constitutional decision. Erwin Chemerinsky, Federal Jurisdiction 692-93 (1994).
. This Court reviews district court decisions on this factor under an abuse of discretion standard if they are "adequately explained." See Hughes v. Lipscher, 906 F.2d 961, 965 (3d Cir.1990). Here, the district court provided no explanation about why significant state policies would not be interfered with by an erroneous decision about the scope of Megan’s Law. See Artway v. Attorney General, 876 F.Supp. 666, 670 n. 4 (D.N.J.1995). It concluded simply that because Artway "is facing a criminal penalty if he does not register today ... [,] any argument for abstention obviously fails.” Id. Thus, the district court appears to have skipped straight to the discretionary balancing of hardships. We agree that the equities favor Artway, but this weighing is necessary-and appropriate-only if the three requirements for abstention are met.