Concurring and Dissenting. — The jury in this case acquitted defendant Steven Lloyd Mosley of committing a lewd act on a child under the age of 14 but convicted him of simple assault. Thereafter the trial judge found that Mosley “committed the offense as a result of sexual compulsion or for purposes of sexual gratification” and ordered him to register as a sex offender. (Pen. Code, § 290.006.) The registration order made Mosley subject to the residency restriction of Penal Code section 3003.5, subdivision (b), which prohibits any registered sex offender from living within 2,000 feet of a school or park.
*1071Mosley argues that under Apprendi v. New Jersey (2000) 530 U.S. 466 [147 L.Ed.2d 435, 120 S.Ct. 2348] (Apprendi), he was entitled to have a jury, not a judge, determine the facts supporting the registration order and residency restriction. In rejecting this claim, today’s opinion avoids deciding whether the residency restriction applies to Mosley — a statutory question — in order to decide not one but two constitutional questions concerning Apprendi’s applicability. Because the court’s methodology is as flawed as its holdings, I respectfully dissent. Section 3003.5, subdivision (b) applies to all registered sex offenders, including Mosley, and its imposition here required a determination of facts that historically lay within the jury’s domain. Mosley has stated a valid Apprendi claim because the residency restriction is a penalty that exceeds what the jury’s verdict in this case permits.
I.
In this appeal, the Attorney General argues that Penal Code section 3003.5, subdivision (b) (hereafter section 3003.5(b)) applies only to registered sex offenders who are on parole and not to misdemeanor probationers like Mosley. (All undesignated statutory references are to the Penal Code.) If the Attorney General is correct, we need not go further to decide whether the residency restriction was imposed in violation of Apprendi. So the first question we must decide is whether section 3003.5(b) applies to sex offenders who are not on parole. Today’s opinion avoids this threshold statutory question. Instead, the court assumes that section 3003.5(b) applies to registered sex offenders not on parole and proceeds to reject Mosley’s Apprendi claim. (Maj. opn., ante, at pp. 1054-1055.) This approach is highly unorthodox.
“If there is one doctrine more deeply rooted than any other in the process of constitutional adjudication, it is that we ought not to pass on questions of constitutionality” unless those questions are “unavoidable.” (Spector Motor Co. v. McLaughlin (1944) 323 U.S. 101, 105 [89 L.Ed. 101, 65 S.Ct. 152]; see Bond v. United States (2014) 572 U.S. _, _ [189 L.Ed.2d 1, 134 S.Ct. 2077, 2087]; Lyng v. Northwest Indian Cemetery Prot. Assn. (1988) 485 U.S. 439, 445-446 [99 L.Ed.2d 534, 108 S.Ct. 1319]; Escambia County v. McMillan (1984) 466 U.S. 48, 51 [80 L.Ed.2d 36, 104 S.Ct. 1577].) This doctrine promotes judicial restraint and minimizes the potential for friction between the judiciary and the political branches. (See National Federation of Independent Business v. Sebelius (2012) 567 U.S. _, _ [183 L.Ed.2d 450, 132 S.Ct. 2566, 2593].) Like the high court, this court has regularly said that we will “not reach constitutional questions unless absolutely required to do so.” (People v. Williams (1976) 16 Cal.3d 663, 667 [128 Cal.Rptr. 888, 547 P.2d 1000]; see Verdin v. Superior Court (2008) 43 Cal.4th 1096, 1102 [77 Cal.Rptr.3d 287, 183 P.3d 1250]; People v. Brown (2003) 31 Cal.4th 518, 534 *1072[3 Cal.Rptr.3d 145, 73 P.3d 1137]; Santa Clara County Local Transportation Authority v. Guardino (1995) 11 Cal.4th 220, 230 [45 Cal.Rptr.2d 207, 902 P.2d 225]; In re Michael G. (1988) 44 Cal.3d 283, 295 [243 Cal.Rptr. 224, 747 P.2d 1152]; People v. Leonard (1983) 34 Cal.3d 183, 187 [193 Cal.Rptr. 171, 666 P.2d 28].) When there is a “statutory basis” for resolving a case, we will not “render[] a decision on constitutional grounds.” (NBC Subsidiary (KNBC-TV), Inc. v. Superior Court (1999) 20 Cal.4th 1178, 1190 [86 Cal.Rptr.2d 778, 980 P.2d 337]; see College Hospital Inc. v. Superior Court (1994) 8 Cal.4th 704, 720-721 [34 Cal.Rptr.2d 898, 882 P.2d 894]; De Lancie v. Superior Court (1982) 31 Cal.3d 865, 877 & fn. 13 [183 Cal.Rptr. 866, 647 P.2d 142]; People v. Gilbert (1969) 1 Cal.3d 475, 484-485 [82 Cal.Rptr. 724, 462 P.2d 580]; Palermo v. Stockton Theatres, Inc. (1948) 32 Cal.2d 53, 66 [195 P.2d 1].)
Against this mountain of precedent, the court invents a doctrine of statutory avoidance: It avoids an issue of statutory interpretation in order to resolve the case on constitutional grounds. But, as the court does not dispute, no such doctrine appears in the case law or in any treatise on statutory interpretation. (See, e.g., Singer & Singer, 2A-3 Sutherland Statutes and Statutory Construction (7th ed. 2014) §§ 45:1-65:5; Scalia & Garner, Reading Law: The Interpretation of Legal Texts (2012).) The recognized canons of construction are so wide-ranging that it has been said “for every canon . . . there is an equal and opposite canon.” (Posner, Statutory Interpretation — in the Classroom and in the Courtroom (1983) 50 U.Chi. L.Rev. 800, 806, citing Llewellyn, The Common Law Tradition: Deciding Appeals (1960) pp. 521-535.) And yet, the canon that a court should avoid a constitutional issue if the case can be decided on statutory grounds has no complement that says a court should avoid a statutory issue if the case can be decided on constitutional grounds. Such an approach has long been rejected. (See Steamship Co. v. Emigration Commissioners (1885) 113 U.S. 33, 39 [28 L.Ed. 899, 5 S.Ct. 352].)
Today’s decision says we need not decide the statutory question “in advance of any concrete evidence of prosecutors’ intent to press charges against nonparolee sex offender registrants for noncompliance with the residency restrictions.” (Maj. opn., ante, at p. 1054.) In essence, the court says the statutory issue is not ripe for decision, even as it insists that the constitutional issue arising from a hypothetical reading of the statute is ripe. (Maj. opn., ante, at pp. 1054-1055, fn. 7.) But the Attorney General acknowledged at oral argument that she knew of at least one case in which a probationer had been charged with violating the statute. And in any event, the statutory question is squarely presented by Mosley’s claim that section 3003.5(b) on its face imposes punishment on all registered sex offenders. That claim — unlike the “constitutional challenges to section 3003.5(b) as a parole condition” in In re E.J. (2010) 47 Cal.4th 1258 [104 Cal.Rptr.3d 165, *1073223 P.3d 31] (maj. opn., ante, at p. 1054, fn. 7, italics added) — requires us to decide whether the statute applies to registrants who are not on parole. The answer to that question does not turn on whether prosecutors intend to apply the statute to any particular person or group.
Had the court properly engaged the statutory question, it would have found a clear answer in the text of section 3003.5(b): “Notwithstanding any other provision of law, it is unlawful for any person for whom, registration is required pursuant to Section 290 to reside within 2000 feet of any public or private school, or park where children regularly gather.” (Italics added.) The term “any person for whom registration is required” is unqualified and thus includes registered sex offenders who are not on parole.
The Attorney General contends that a literal reading of section 3003.5(b) would undermine the statute’s purpose of protecting the public and would clash with other provisions of section 3003.5. According to her briefing, reading section 3003.5(b) to apply to any registered sex offender, not just parolees, would increase the rate of transience among sex offenders. This would make monitoring and rehabilitation more difficult, thus inhibiting efforts to solve crimes and reduce recidivism. Moreover, the Attorney General notes that section 3003.5 appears in a chapter of the Penal Code dealing with parole conditions and that the only provision of section 3003.5 that existed before Proposition 83 took effect was subdivision (a), which prohibits a sex offender registrant “released on parole” from living with another registrant in a “single family dwelling.” If the voters truly intended section 3003.5(b) to apply to “any person for whom registration is required,” she argues, it is not clear why they put it in section 3003.5 rather than section 290, which sets forth sex offender registration requirements.
But there is no inconsistency in section 3003.5 between subdivision (a), which sets forth a residency restriction as a parole condition, and subdivision (b), which sets forth a different residency restriction that applies to parolees and nonparolees alike. Moreover, whether or not applying a residency restriction to all registered sex offenders is an effective way to promote public safety, the voters evidently believed it would be.
To the extent there is any ambiguity in the statute, we may look “to extrinsic sources such as ballot summaries and arguments for insight into the voters’ intent.” (Kwikset Corp. v. Superior Court (2011) 51 Cal.4th 310, 321 [120 Cal.Rptr.3d 741, 246 P.3d 877].) Nothing in the ballot materials on Jessica’s Law suggests that the voters intended to limit section 3003.5(b) to parolees. The official summary of the measure says Proposition 83 “[prohibits registered sex offenders from residing within 2,000 feet of any school or park.” (Voter Information Guide, Gen. Elec. (Nov. 7, 2006) official title and *1074summary of Prop. 83, p. 42, italics added.) Similarly, the Legislative Analyst explained that the measure “bars any person required to register as a sex offender from living within 2,000 feet (about two-fifths of a mile) of any school or park” and that “[a] violation of this provision would be a misdemeanor offense, as well as a parole violation for parolees.” (Id., analysis of Prop. 83 by Legis. Analyst, p. 44, italics added.) These statements suggest that section 3003.5(b) applies to any registered sex offender, including those not on parole — -just as the statute says.
II.
Having determined that section 3003.5(b) applies to Mosley, I now turn to his Apprendi claim. Today’s decision holds that “the effect of Apprendi on the residency restrictions of Jessica’s Law is obviated by a post-Apprendi decision, Oregon v. Ice (2009) 555 U.S. 160 [172 L.Ed.2d 517, 129 S.Ct. 711] (Ice).” (Maj. opn., ante, at p. 1049.) But the court misreads the Apprendi line of cases, including Ice.
Apprendi held that “[o]ther than the fact of a prior conviction, any fact that increases the penalty for a crime beyond the prescribed statutory maximum must be submitted to a jury, and proved beyond a reasonable doubt.” (Apprendi, supra, 530 U.S. at p. 490.) The high court reached this holding in the context of a New Jersey hate crime statute that prescribed longer sentences if a judge found by a “preponderance of the evidence” that the defendant committed an offense “ ‘with a purpose to intimidate an individual or group of individuals because of race, color, gender, handicap, religion, sexual orientation or ethnicity.’ ” (Id. at pp. 468-469.) Allowing the imposition of this enhanced penalty based on judicial factfinding violated the defendant’s right to a jury trial under the Sixth Amendment to the United States Constitution. (530 U.S. at pp. 490-492.)
Subsequently, the high court in Ice held that, while a jury must determine the “facts that warrant punishment for a specific statutory offense,” a judge may determine the facts that warrant consecutive rather than concurrent sentences for multiple offenses. (Oregon v. Ice, supra, 555 U.S. at p. 170 (Ice).) Ice explained that Apprendfs “animating principle is the preservation of the jury’s historic role as a bulwark between the State and the accused at the trial for an alleged offense. [Citation.] Guided by that principle, our opinions make clear that the Sixth Amendment does not countenance legislative encroachment on the jury’s traditional domain. [Citation.] We accordingly considered whether the finding of a particular fact was understood as within ‘the domain of the jury ... by those who framed the Bill of Rights.’ [Citation.]” (Ice, supra, 555 U.S. at p. 168.)
*1075Applying this inquiry in Ice, the high court said “[t]he historical record demonstrates that the jury played no role in the decision to impose sentences consecutively or concurrently. Rather, the choice rested exclusively with the judge. [Citations.]” (Ice, supra, 555 U.S. at p. 168.) “In light of this history, legislative reforms regarding the imposition of multiple sentences do not implicate the core concerns that prompted our decision in Apprendi. There is no encroachment here by the judge upon facts historically found by the jury, nor any threat to the jury’s domain as a bulwark at trial between the State and the accused.” (Id. at p. 169.)
Today’s opinion says that “[h]ere, as in Ice, there is no jury tradition connected to the sentencing decision at issue.” (Maj. opn., ante, at p. 1060.) “Instead, both residency restrictions and underlying sex offender registration requirements are modem regulatory sentencing imperatives unknown at common law. . . . They are additional examples of ‘sentencing choices or accoutrements’ ([Ice, supra, 555 U.S.] at p. 172) in which juries have played no historical role, and which do not implicate the Sixth Amendment jury trial guarantee within the meaning of Apprendi.” (Id. at p. 1060.) This is unpersuasive for several reasons.
As an. initial matter, the high court in Ice observed that all of its prior decisions applying Apprendi’s rule “involved sentencing for a discrete crime, not — as [in Ice] — for multiple offenses different in character or committed at different times.” (Ice, supra, 555 U.S. at p. 167.) Whereas “the determination of facts that warrant punishment for a specific statutory offense” fell within the jury’s traditional role (id. at p. 170), “administering multiple sentences” on multiple offenses did not (id. at p. 168). The case before us involves sentencing for a specific statutory offense, not multiple offenses.
More fundamentally, today’s decision misreads Ice when it reasons that, like the administration of multiple sentences, “sex offender registration and residency requirements are not sentencing matters in which, historically, the jury has played any traditional role at common law.” (Maj. opn., ante, at pp. 1059-1060.) In saying that “the scope of the constitutional jury right must be informed by the historical role of the jury at common law” (Ice, supra, 555 U.S. at p. 170), Ice did not suggest that the scope of the Sixth Amendment turns on whether common law juries imposed particular sentences. Rather, Ice made clear that the historical inquiry focuses on “whether the finding of a particular fact was understood as within ‘the domain of the jury.’ ” (Ice, at p. 168, italics added; see id. at p. 170 [“Apprendi’s core concern” is “a legislative attempt to ‘remove from the [province of the] jury’ the determination of facts that warrant punishment for a specific statutory offense.” (italics added)].) The fact that juries historically did not make a particular sentencing decision was evidence that a defendant had no entitlement to have a jury *1076make the factual determinations underlying the sentencing decision. In Ice, the fact that “the jury played no role in the decision to impose sentences consecutively or concurrently” meant that a judge could find the facts required to impose consecutive sentences without “encroaching] . . . upon facts historically found by the jury.” (Id. at pp. 168, 169, italics added.)
In Southern Union Co. v. United States (2012) 567 U.S. _ [183 L.Ed.2d 318, 132 S.Ct. 2344] (Southern Union), the high court reaffirmed that the salient historical inquiry is whether the determination of facts authorizing a particular sentence lay within the jury’s domain. Southern Union involved a federal statute pegging criminal fines to the duration of the violation. The question was “whether [Apprendi's] rule applies to sentences of criminal fines.” (Id. at pp. __-__ [132 S.Ct. at pp. 2348-2349].) The high court held that although some fines are not substantial enough to trigger the protections of the Sixth Amendment, “[w]here a fine is substantial enough . . . , Apprendi applies in full.” (Southern Union, at p. _ [132 S.Ct. at p. 2352]].) In so holding, the high court reiterated that “Apprendi’s ‘core concern’ is to reserve to the jury ‘the determination of facts that warrant punishment for a specific statutory offense.’ Ice, 555 U.S., at 170.” (Id. at p. _ [132 S.Ct. at p. 2350].) Southern Union further explained: “In stating Apprendi’s rule, we have never distinguished one form of punishment from another. Instead, our decisions broadly prohibit judicial factfinding that increases maximum criminal ‘sentence^],’ ‘penalties,’ or ‘punishments]’ — terms that each undeniably embrace fines. [Citations.]” (Id. at p. _ [132 S.Ct. at p. 2351].)
The high court acknowledged that “judges in the colonies and during the founding era ‘possessed a great deal of discretion’ in determining whether to impose a fine and in what amount.” (Southern Union, supra, 567 U.S. at p. _ [132 S.Ct. at p. 2353].) But the fact that historically “[t]he judge, not the jury, would normally determine fine-related sentencing facts” (id. at p. _ [132 S.Ct. at p. 2369] (dis. opn. of Breyer, J.)) was not sufficient to make Apprendi inapplicable to criminal fines. Instead, the high court said “the salient question here is what role the jury played in prosecutions for offenses that did peg the amount of a fine to the determination of specified facts— often, the value of damaged or stolen property.” (Id. at p. _ [132 S.Ct. at p. 2353].) On this question, the court concluded that “juries routinely found facts that set the maximum amounts of fines” (id. at p. _ [132 S.Ct. at p. 2356]), citing several historical examples (id. at pp. __-__ [132 S.Ct. at pp. 2353-2354]). Although none of the examples concerned the jury’s role in determining the duration of a criminal violation, the court did not hesitate to conclude that the Sixth Amendment requires such a factual determination to be made by a jury.
Thus, the applicability of Apprendi’s rule does not depend on “distinguishing] one form of punishment from another” (Southern Union, supra, 567 U.S. *1077at p. _ [132 S.Ct. at p. 2351]), nor does it require historical proof that juries as opposed to judges generally imposed a particular punishment or that juries ever made the exact same factual finding as the one at issue. Moreover, it does not matter that a particular sentencing decision was “unknown at common law” (maj. opn., ante, at p. 1060), for the same could be said of the hate crime sentencing enhancement in Apprendi itself. (See Apprendi, supra, 530 U.S. at p. 478 [“Any possible distinction between an ‘element’ of a felony offense and a ‘sentencing factor’ was unknown to the practice of criminal indictment, trial by jury, and judgment by court as it existed during the years surrounding our Nation’s founding.” (fn. omitted)].) The Sixth Amendment right to a jury trial cannot be rendered inapplicable simply by legislative development of “modern” punishments (maj. opn., ante, at p. 1060) or by “the novelty of a legislative [sentencing] scheme” (Apprendi, at p. 482). While acknowledging that “trial practices [may] change in the course of centuries,” Apprendi made clear that “practice must at least adhere to the basic principles undergirding the requirements of trying to a jury all facts necessary to constitute a statutory offense, and proving those facts beyond reasonable doubt.” (Id. at pp. 483-484.)
Nor does it help to call sex offender registration and residency requirements “ ‘sentencing . . . accoutrements’ ” akin to “ ‘supervised release following service of a prison sentence’ ” or “ ‘required attendance at drug rehabilitation programs or terms of community service.’ ” (Maj. opn., ante, at p. 1060, quoting Ice, supra, 555 U.S. at p. 171.) The high court in Southern Union suggested that this language from Ice was dicta and “more likely refers to the routine practice of judges’ imposing [sentencing options] from within a range authorized by jury-found facts” — a practice that “poses no problem under Apprendi because the penalty does not exceed what the jury’s verdict permits.” (Southern Union, supra, 567 U.S. at p. _, fn. 5 [132 S.Ct. at p. 2352, fn. 5].) The essential inquiry in Apprendi, as in Southern Union, was whether the “legislative scheme . . . removes the jury from the determination of a fact that, if found, exposes the criminal defendant to a penalty exceeding the maximum he would receive if punished according to the facts reflected in the jury verdict alone.” (Apprendi, supra, 530 U.S. at pp. 482-483; see Southern Union, at pp. __-__ [132 S.Ct. at pp. 2350-2351].)
The case before us is squarely analogous to Apprendi in a crucial respect: Imposition of the sex offender registration requirement and residency restriction, like the hate crime sentence enhancement in Apprendi, turns on a finding of the defendant’s motive or intent. (See Apprendi, supra, 530 U.S. at p. 494 [“it does not matter whether the required finding is characterized as one of intent or of motive”].) Section 290.006 authorizes a court to require registration “if the court finds at the time of conviction or sentencing that the person committed the offense as a result of sexual compulsion or for purposes of sexual gratification.” (Italics added.) “By its very terms, this *1078statute mandates an examination of the defendant’s state of mind — a concept known well to the criminal law as the defendant’s mens rea.” (Apprendi, at p. 492.) The key historical question — “whether the finding of a particular fact was understood as within ‘the domain of the jury ... by those who framed the Bill of Rights’ ” (Ice, supra, 555 U.S. at p. 168)—has an easy answer in this case. (Whether “the defendant’s likelihood of reoffense” must also be found by a jury (maj. opn., ante, at p. 1061) is not before us today, and we have never decided whether such a finding is necessary to expose a defendant to discretionary registration under section 290.006.)
In sum, the circumstances here directly implicate Apprendi’s rule. The jury acquitted Mosley of committing a lewd act on a child under the age of 14 — an offense requiring proof of sexual intent (§ 288, subd. (a)) — and instead convicted him of simple assault (§ 240). A simple assault conviction, by itself, does not result in any registration requirement or residency restriction. The “effect” of the judge’s finding under section 290.006 was to “expose the defendant to a greater punishment than that authorized by the jury’s guilty verdict.” (Apprendi, supra, 530 U.S. at p. 494.) At least that is so if, as Mosley contends, section 3003.5(b)’s residency restriction constitutes punishment under the Sixth Amendment.
III.
Turning to this last question, I agree with the Court of Appeal that the residency restriction is punitive based on the factors set forth in Kennedy v. Mendoza-Martinez (1963) 372 U.S. 144, 168-169 [9 L.Ed.2d 644, 83 S.Ct. 554] (Mendoza-Martinez). Applying those factors, the high court in Smith v. Doe (2003) 538 U.S. 84, 97-106 [155 L.Ed.2d 164, 123 S.Ct. 1140] (Smith) held that Alaska’s sex offender registration scheme was not punitive for purposes of the ex post facto clause, and this court has reached the same conclusion about California’s sex offender registration requirements for purposes of the Sixth Amendment (People v. Picklesimer (2010) 48 Cal.4th 330, 343-344 [106 Cal.Rptr.3d 239, 226 P.3d 348] (Picklesimer)). But section 3003.5(b)’s residency restriction goes beyond registration requirements and imposes more severe burdens that drive the analysis toward a different result.
In first determining the voters’ intent behind section 3003.5(b), today’s opinion acknowledges that Proposition 83 “did not specifically state why the voters adopted the residency restrictions in particular” and that Proposition 83’s official title and findings “suggest that, overall, the electorate had a mix of punitive, deterrent, and protective motives.” (Maj. opn., ante, at pp. 1063-1064.) Nevertheless, the court puts decisive weight on the ballot arguments of Proposition 83’s supporters, while dismissing the fact that Proposition 83 placed the residency restriction in the Penal Code and made *1079its violation a criminal offense. (Maj. opn., ante, at pp. 1064-1065; cf. Smith, supra, 538 U.S. at p. 94 [“formal attributes of a legislative enactment, such as the manner of its codification or the enforcement procedures it establishes, are probative of the legislature’s intent” though “not dispositive”].)
On balance, I believe the voters primarily intended section 3003.5(b) as a regulatory measure. But this nonpunitive intent is not stated as clearly or as conclusively as when a legislature “expressed] the objective of the law in the statutory text itself.” (Smith, supra, 538 U.S. at p. 93.) As a result, the Court of Appeal was right to question whether the “ ‘ “clearest proof’ ’ ” of punitive effect is necessary to outweigh the voters’ apparent intent. (Id. at p. 92; see id. at p. 110 (conc. opn. of Souter, J.); id. at pp. 114-115 (dis. opn. of Ginsburg, J.).) Ultimately, however, we need not resolve the appropriate standard because even if the clearest proof is required, the Mendoza-Martinez factors decisively show that section 3003.5(b) is punitive in effect.
Here, as in Smith, “[t]he factors most relevant to our analysis are whether, in its necessary operation, the regulatory scheme; has been regarded in our history and traditions as a punishment; imposes an affirmative disability or restraint; promotes the traditional aims of punishment; has a rational connection to a nonpunitive purpose; or is excessive with respect to this purpose.” (Smith, supra, 538 U.S. at p. 97.) These factors are “ ‘neither exhaustive nor dispositive,’ ” but are “ ‘useful guideposts.’ ” (Ibid.) In applying these factors, we must evaluate the effects of section 3003.5(b) in light of its applicability to all registered sex offenders, not just offenders like Mosley whose victims were children. (See Mendoza-Martinez, supra, 372 U.S. at p. 169 [“these factors must be considered in relation to the statute on its face”].)
In assessing whether a residency restriction “has been regarded in our history and traditions as a punishment” (Smith, supra, 538 U.S. at p. 97), the Court of Appeal said it “is sufficiently close to banishment, property deprivation, and a probation condition to be deemed traditional punishment.” These semblances are discernible but inexact. Unlike banishment, which directly forbids offenders from “returning] to their original community” (Smith, at p. 98), section 3003.5(b) declares certain areas off-limits for establishing a home but not for any other purpose. Unlike “the punitive confiscation of property,” which traditionally involves a fine or seizure of land specifically intended to burden the defendant’s property interests (Nixon v. Administrator of General Services (1977) 433 U.S. 425, 474 & fn. 38 [53 L.Ed.2d 867, 97 S.Ct. 2777]), the deprivation of property interests worked by section 3003.5(b) seems incidental to the statute’s main purpose of restricting liberty. And unlike probation or parole, which involves conditional release and ongoing *1080supervision (see Smith, at p. 101), section 3003.5(b) establishes an independent legal obligation without individualized supervision or constructive custody. Comparing the residency restriction to traditional forms of punishment does not strongly suggest it is punitive.
By contrast, section 3003.5(b) clearly “imposes an affirmative disability or restraint” that suggests its punitive character. (Smith, supra, 538 U.S. at p. 97.) “Here, we inquire how the effects of the [statute] are felt by those subject to it.” (Id. at pp. 99-100.) Unlike a registration requirement, which “imposes no physical restraint” and “leaves [sex offenders] free to change . . . residences” (id. at p. 100), the residency restriction directly limits where sex offenders may live and establish a home. In addition, whereas the registration scheme in Smith disclosed and disseminated information that was “already a matter of public record” (id. at p. 101), section 3003.5(b) imposes on sex offenders a burden that is separate from and additional to all other requirements with which they must comply.
Moreover, section 3003.5(b) predictably results in severe geographic limitations on compliant housing in densely populated areas, as our decision today in In re Taylor (2015) 60 Cal.4th 1019 [184 Cal.Rptr.3d 682, 343 P.3d 867] (Taylor) confirms. (See id. at pp. 1039-1040 [§ 3003.5(b) “effectively barred petitioners access to approximately 97 percent of the multifamily rental housing units in San Diego County that would otherwise be available to them”]; see also Cal. Sex Offender Management Bd., Homelessness Among California’s Registered Sex Offenders: An Update (Sept. 2011) p. 7 (hereafter 2011 Update) [“A number of metropolitan areas have developed maps showing the areas where, according to the language of Prop 83, affected sex offenders may not live. Observers agree that the vast majority of potential housing locations in urban areas are now included in the off-limits territory. San Francisco, for example, has virtually no realistic places where a paroled sex offender may legally live.”].)
The breadth of the restriction has “led to dramatically escalating levels of homelessness among sex offenders.” (2011 Update, supra, at p. 1; see id. at p. 7 [sex offenders registered as transient nearly tripled from 2,050 in 2007 (just after Jessica’s Law was passed) to 6,012 in 2011]; Taylor, supra, 60 Cal.4th at pp. 1040-1041.) For many sex offenders, the residency restriction also adversely affects access to transportation, employment opportunities, health care, drag and alcohol rehabilitation programs, and other social services. (See Taylor, at p. 1040.) And even among those who are able to obtain compliant housing and access to needed services, a restriction on where one may choose to establish a home can itself work a significant deprivation. (See Kelo v. New London (2005) 545 U.S. 469, 494-495 [162 L.Ed.2d 439, 125 S.Ct. 2655] (dis. opn. of O’Connor, J.); Cleburne v. Cleburne Living Center, Inc. (1985) *1081473 U.S. 432, 473 [87 L.Ed.2d 313, 105 S.Ct. 3249] (conc. & dis. opn. of Marshall, J.) (Cleburne).) Unlike similar laws in other states, section 3003.5(b) contains no provision allowing a sex offender to stay in a home that he or she owned before the law’s enactment, before the sex offense conviction, or before a newly located school or park is established in the area. (Cf., e.g., Iowa Code § 692A.114, subd. 3.c., d.); 720 Ill. Comp. Stat. 5/11-9.3(b-5); Okla. Stat. tit. 57, § 590A.)
Today’s opinion acknowledges the realities described in Taylor, which addressed an as-applied challenge, but disclaims any awareness of how section 3003.5(b) generally affects sex offenders throughout the state. (Maj. opn., ante, at p. 1067, fn. 15.) The court says the terms of the residency restriction are “potentially burdensome” but “limited” because they “do not regulate a registered sex offender’s daily activities” and “do not dictate where he or she may travel, visit, shop, eat, work, or play.” (Id. at p. 1066.) This blinks reality and common sense.
If the only real effect of the residency restriction were to control where sex offenders may sleep at night, leaving them free to spend their waking hours near parks and schools, then Jessica’s Law would have been an idle enactment. Clearly, the desired and anticipated effect of the law was to keep sex offenders away from neighborhoods where children play or attend school— not just at night, but at all times. (See Voter Information Guide, supra, argument in favor of Prop. 83, p. 46 [Prop. 83 will “keep[] [child molesters] away from schools and parks” and will “[c]reate predator free zones around schools and parks” (capitalization & italics omitted)].) By prohibiting sex offenders from living near schools or parks, section 3003.5(b) has the effect of eroding or severing their ties to the community and ordinary civic life. (See Pleasant Grove City v. Summum (2009) 555 U.S. 460, 469 [172 L.Ed.2d 853, 129 S.Ct. 1125] [parks are traditional public forums “ ‘ “used for purposes of assembly, communicating thoughts between citizens, and discussing public questions” ’ ”]; Cleburne, supra, 473 U.S. at p. 461, fn. 5 (conc. & dis. opn. of Marshall, J.) [group home for mentally retarded persons “was specifically located near a park, a school, and a shopping center so that its residents would have full access to the community at large”]; Evans v. Newton (1966) 382 U.S. 296, 301-302 [15 L.Ed.2d 373, 86 S.Ct. 486] [a park is a “public facility” that “serves the community” and can play “an integral part” of a city’s activities]; Ed. Code, § 38131, subd. (a) [each school is “a civic center”]; Presidential Com. on Election Admin., The American Voting Experience: Report and Recommendations of the Presidential Commission on Election Administration (Jan. 2014) p. 33 [“schools [are] the preferred venue for polling places”].) To say that section 3003.5(b) “do[es] not regulate a registered sex offender’s daily activities” (maj. opn., ante, at p. 1066) provides no realistic appraisal of “how the effects of the [statute] are felt by those subject to it.” (Smith, supra, 538 U.S. at pp. 99-100.) The effect, even *1082if not the intent, of the residency restriction is to ostracize sex offenders as “pariahs who do not belong in the community.” (Cleburne, at p. 473 (conc. & dis. opn. of Marshall, J.).)
The nature and severity of these burdens inform whether section 3003.5(b) “promote[s] the traditional aims of punishment — retribution and deterrence.” (Mendoza-Martinez, supra, 372 U.S. at p. 168.) Although “[a]ny number of governmental programs might deter crime without imposing punishment” (Smith, supra, 538 U.S. at p. 102), the residency restriction hardly qualifies as ordinary government regulation. As noted, whereas the registration requirement in Smith publicized “information about the individual’s conviction [that] was already in the public domain” (id. at p. 100), the residency restriction imposes greater and qualitatively different burdens than a registration requirement. To the extent that potential offenders weigh the costs and benefits of criminal conduct, the lifetime disabilities resulting from the residency restriction substantially increase the costs of sexual offenses and thereby promote deterrence. Moreover, it is significant that section 3003.5(b) “makes no individualized determination of the dangerousness of a particular registrant. . . . When a restriction is imposed equally upon all offenders, with no consideration given to how dangerous any particular registrant may be to public safety, that restriction begins to look far more like retribution for past offenses than a regulation intended to prevent future ones.” (Commonwealth v. Baker (Ky. 2009) 295 S.W.3d 437, 444 (Baker) [finding similar residency restriction (Ky. Rev. Stat. § 17.545) to be punitive for purposes of ex post facto analysis].)
The remaining factors are whether section 3003.5(b) “has a rational connection to a nonpunitive purpose” and whether it “is excessive with respect to this purpose.” (Smith, supra, 538 U.S. at p. 97.) Today’s opinion effectively collapses these two factors into one: It finds the residency restriction rationally connected to protecting children around schools and parks, and it then finds no excessiveness by analogizing the inquiry to ordinary rational basis review. (Maj. opn., ante, at p. 1069 [“The electorate could reasonably conclude . . . .”]; id. at p. 1069, fn. 16, citing FCC v. Beach Communications, Inc. (1993) 508 U.S. 307, 314 [124 L.Ed.2d 211, 113 S.Ct. 2096].) But the inquiry here focuses on the effects of the statute, not whether there is some “ ‘conceivable basis which might support it.’ ” (Beach Communications, at p. 315.) In stating that “[t]he question is whether the . . . means chosen are reasonable in light of the nonpunitive objective” (Smith, at p. 105), Smith drew no analogy between the excessiveness inquiry and rational basis review. Instead, the high court evaluated reasonableness in this context by reference to the statute’s actual workings and empirical underpinnings (id. at pp. 103-105).
*1083Although section 3003.5(b)’s nonpunitive purpose of protecting children is not a sham or pretext for imposing punishment, the residency restriction is plainly excessive with respect to that purpose. The most conspicuous feature of section 3003.5(b) in this regard is its categorical applicability to all registered sex offenders — regardless of whether the offender ever targeted children, regardless of whether the offense was violent or nonviolent, regardless of risk profile. Section 3003.5(b) differs from analogous residency restrictions that limit their applicability to sex offenders whose crimes involved children. (See, e.g., Del. Code tit. 11, § 1112(a), (b)(4); Fla. Stat. § 775.215(2)(a), (3)(a); Ind. Code § 35-42-4-ll(a); Iowa Code § 692A.114, subd. l.c.; S.C. Code § 23-3-535(B).) And it differs from analogous restrictions that limit their applicability to sex offenders who have committed a certain level of aggravated offense or who have been assessed as meeting a certain threshold of dangerousness. (See, e.g., Ark. Code § 5-14-128(a); Neb. Rev. Stat. §§ 29-4016(4), (5), 29-4017(1).)
The indiscriminate character of section 3003.5(b) also stands in contrast to the differentiated approach set forth in the community notification provisions of California’s sex offender registration statute. Under section 290.46, the state Department of Justice must maintain a public Web site that provides information about registered sex offenders. This Web site contains a wealth of information about individuals convicted of the most serious sex crimes, including their names, photographs, and addresses. (§ 290.46, subd. (b).) However, the Web site displays less information about individuals who pose a lower risk to the community. For example, it provides the ZIP codes but not the addresses of some offenders. (§ 290.46, subds. (c), (d).) And the lowest risk offenders may file an application to have all of their information removed from the Web site. (§ 290.46, subd. (e).) These provisions are designed to protect communities from dangerous individuals without intruding more than necessary on the privacy rights of low risk offenders. Section 3003.5(b), by contrast, imposes the same onerous restraint on all sex offenders, even those who pose such a low risk of recidivism that their information is excluded from online community notification.
In Smith, the high court reasoned that because sex offender registration imposes a relatively “minor” burden, “the State can dispense with individual predictions of future dangerousness” without casting into doubt the nonpunitive character of the regulatory scheme. (Smith, supra, 538 U.S. at p. 104.) By contrast, when a statute imposes a severe restraint such as involuntary commitment, “[t]he magnitude of the restraint ma[kes] individual assessment appropriate.” (Ibid.) Though less onerous than involuntary commitment, section 3003.5(b)’s residency restriction is more onerous than registration. It is sufficiently onerous that “a lack of individual assessment,” which attenuates the nexus to a regulatory purpose, “render[s] the statute punitive.” (Baker, supra, 295 S.W.3d at p. 446.)
*1084The excessiveness of the residency restriction in relation to its nonpunitive purpose is reinforced by findings of the California Sex Offender Management Board (Board). Created in 2005 by bipartisan legislation (Stats. 2006, ch. 338, p. 2668), the Board is the state entity within the Department of Corrections and Rehabilitation charged with “address [ing] any issues, concerns, and problems related to the community management of adult sex offenders.” (§ 9002, subd. (a); see ibid. [“The main objective of the board ... is to achieve safer communities by reducing victimization.”].) In a 2011 report, the Board concluded: “Based on all that is known about sex offender recidivism and about the nature of most sex offenses involving children, there is no evidence that residence restrictions are related to preventing or deterring sex crimes against children. To the contrary, the evidence strongly suggests that residence restrictions are likely to have the unintended effect of increasing the likelihood of sexual re-offense.” (2011 Update, supra, at p. 1.) According to the Board, residency restrictions like section 3003.5(b) are ineffective because most sex crimes against children are perpetrated not by strangers lurking near schools or parks, but by family members or acquaintances who victimize children inside their homes or other private settings. (2011 Update, at pp. 9-13.) At the same time, by limiting sex offenders’ access to housing and, in turn, employment, health care, transportation, and rehabilitative services, residency restrictions often result in “an unstable lifestyle” that “represents a major risk factor for re-offending.” (Id. at p. 14.) These general findings, like the recidivism data cited in Smith and in today’s opinion, properly inform whether the statute is punitive in effect. (See maj. opn., ante, at p. 1068, quoting Smith, supra, 538 U.S. at p. 103 [relying on secondary sources]; Smith, at p. 104 [same]; but cf. maj. opn., ante, at p. 1063, fn. 12 [criticizing my reliance on secondary sources].)
To be sure, “[t]he excessiveness inquiry ... is not an exercise in determining whether the legislature has made the best choice possible to address the problem it seeks to remedy.” (Smith, supra, 538 U.S. at p. 105.) Even so, section 3003.5(b) is sweeping and indiscriminate “in its necessary operation” (Smith, at p. 97) and imposes significant disabilities and restraints that scarcely further the statute’s protective purpose, if at all. Section 3003.5(b)’s residency restriction, though regulatory in intent, is punitive in effect.
This conclusion, in the context of an Apprendi claim, does not mean the residency restriction cannot validly be imposed on persons who are subject to registration under section 290.006. But it does mean the facts authorizing imposition of the restriction must be proven to a jury beyond a reasonable doubt. In this case, the residency restriction is “a penalty exceeding the maximum [Mosley] would receive if punished according to the facts reflected in the jury verdict alone.” (Apprendi, supra, 530 U.S. at p. 483.) Accordingly, *1085I respectfully dissent from today’s decision upholding the imposition of the residency restriction on Mosley.
Because the residency restriction is separable from section 290’s registration requirements, and because our precedent holds that the registration requirements are not punitive (Picklesimer, supra, 48 Cal.4th at pp. 343-344), I agree that “the judge-imposed registration order remains separately valid and extant.” (Maj. opn., ante, at p. 1070.) I thus join the court in reversing the Court of Appeal’s judgment to the extent it relieved Mosley of the requirement that he register as a sex offender.
Werdegar, J., concurred.
Appellant’s petition for a rehearing was denied April 29, 2015. Werdegar, J., and Liu, J., were of the opinion that the petition should be granted.