People v. Valencia

Full title: THE PEOPLE, Plaintiff and Respondent, v. DAVID J. VALENCIA, Defendant and…

Court: SUPREME COURT OF CALIFORNIA

Date published: Jul 3, 2017

3 Cal.5th 347 (Cal. 2017)

220 Cal. Rptr. 3d 230

397 P.3d 936

Opinion

S223825 S223676

07-03-2017

The PEOPLE, Plaintiff and Respondent, v. David J. VALENCIA, Defendant and Appellant. The People, Plaintiff and Respondent, v. Clifford Paul Chaney, Defendant and Appellant.

Michael Satris, Bolinas, under appointment by the Supreme Court, for Defendant and Appellant Clifford Paul Chaney. Stephanie L. Gunther, Bakersfield, under appointment by the Supreme Court, for Defendant and Appellant David John Valencia. Three Strikes Project and Michael S. Romano, San Francisco, for George Gascon, Bill Landsdowne and David Mills as Amici Curiae on behalf of Defendants and Appellants. Richard Such and John T. Philipsborn, San Francisco, for California Attorneys for Criminal Justice as Amici Curiae on behalf of Defendants and Appellants. Kamala D. Harris and Xavier Becerra, Attorneys General, Dane R. Gillette and Gerald A. Engler, Chief Assistant Attorneys General, Michael P. Farrell, Assistant Attorney General, Stephen G. Herndon, Rachelle A. Newcomb, Peter W. Thompson and Darren K. Indermill, Deputy Attorneys General, for Plaintiff and Respondent.

Cantil-Sakauye, C.J.

Michael Satris, Bolinas, under appointment by the Supreme Court, for Defendant and Appellant Clifford Paul Chaney.

Stephanie L. Gunther, Bakersfield, under appointment by the Supreme Court, for Defendant and Appellant David John Valencia.

Three Strikes Project and Michael S. Romano, San Francisco, for George Gascon, Bill Landsdowne and David Mills as Amici Curiae on behalf of Defendants and Appellants.

Richard Such and John T. Philipsborn, San Francisco, for California Attorneys for Criminal Justice as Amici Curiae on behalf of Defendants and Appellants.

Kamala D. Harris and Xavier Becerra, Attorneys General, Dane R. Gillette and Gerald A. Engler, Chief Assistant Attorneys General, Michael P. Farrell, Assistant Attorney General, Stephen G. Herndon, Rachelle A. Newcomb, Peter W. Thompson and Darren K. Indermill, Deputy Attorneys General, for Plaintiff and Respondent.

Cantil-Sakauye, C.J.In November 2012, California voters enacted Proposition 36, the Three Strikes Reform Act of 2012 (Proposition 36 or Three Strikes Reform Act). With some exceptions, Proposition 36 modified California's "Three Strikes" law to reduce the punishment imposed when a defendant's third felony conviction is not serious or violent. ( Pen. Code,¹ § 667, subd. (e)(2)(C), as amended by Prop. 36, § 2, approved by the voters at Gen. Elec. (Nov. 6, 2012).) It also enacted a procedure governing inmates sentenced under the former Three Strikes law whose third strike was neither serious nor violent, permitting them to petition for resentencing in accordance with Proposition 36's new sentencing provisions. (§ 1170.126, subd. (e), as added by Prop. 36, § 2, approved by the voters at Gen. Elec. (Nov. 6, 2012).) The resentencing provisions provide, however, that an inmate will be denied resentencing if "the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§ 1170.126, subd. (f), as added by Prop. 36, § 6, approved by the voters at Gen. Elec. (Nov. 6, 2012).) Proposition 36 did not define the phrase "unreasonable risk of danger to public safety."Two years later, in November 2014, California voters approved Proposition 47, the Safe Neighborhoods and Schools Act (Proposition 47). Proposition 47 reduced certain drug-related and theft-related offenses that previously were felonies or "wobblers"² to misdemeanors. (§ 1170.18, added by Prop. 47, § 14, approved by the voters at the Gen. Elec. (Nov. 4, 2014).) It also enacted a procedure permitting inmates who are serving felony sentences for offenses that Proposition 47 reduced to misdemeanors to petition to have their felony convictions reclassified as misdemeanors and to be resentenced based on the reclassification. Like Proposition 36, Proposition 47 gave resentencing courts discretion to decline to impose a lesser sentence if resentencing "would result in an unreasonable risk of danger to public safety." (§ 1170.18, subd. (b)(3).) In contrast to Proposition 36, however, Proposition 47 restricted that discretion by defining the phrase "unreasonable risk of danger to public safety." (§ 1170.18, subd. (c).) It stated: "As used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit a new violent felony within the meaning of" section 667, subdivision (e)(2)(C)(iv). (§ 1170.18, subd. (c).) The cited subdivision of section 667 identifies eight types of particularly serious or violent felonies, known colloquially as "super strikes."³

2 Wobblers are "a special class of crimes involving conduct that varies widely in its level of seriousness," and may therefore be "chargeable or ... punishable as either a felony or a misdemeanor." (People v. Park (2013) 56 Cal.4th 782, 789, 156 Cal.Rptr.3d 307, 299 P.3d 1263 ; see also People v. Kunkel (1985) 176 Cal.App.3d 46, 51, fn. 3, 221 Cal.Rptr. 359.)

3 The felonies commonly referred to as super strikes are (1) a " 'sexually violent offense,' " as defined in Welfare & Institutions Code, section 6600, subdivision (b) ; (2) oral copulation or sodomy, or sexual penetration of a child under 14 years of age and more than 10 years younger than the defendant, as defined in sections 286, 288a and 289; (3) a lewd or lascivious act involving a child under 14 years of age, in violation of section 288; (4) any homicide offense, including attempted homicide, as defined in sections 187 to 191.5; (5) solicitation to commit murder, as defined in section 653f; (6) assault with a machine gun on a peace officer or firefighter, as defined in section 245, subdivision (d)(3); (7) possession of a weapon of mass destruction, as defined in section 11418, subdivision (a)(1); and (8) any serious and/or violent felony offense punishable in California by life imprisonment or death. (See § 667, subd. (e)(2)(C)(iv) (I-VIII).)

We granted review in these two cases to resolve two related issues concerning Proposition 47's effect on resentencing proceedings under Proposition 36. In People v. Valencia (S223825), we address whether Proposition 47's definition of "unreasonable risk of danger to public safety" (§ 1170.18, subd. (c)) applies to resentencing proceedings under Proposition 36. People v. Chaney (S223676) presents the question whether, if Proposition 47's definition of unreasonable risk of danger to public safety applies to resentencing proceedings under the Three Strikes Reform Act, does the definition apply retroactively to Proposition 36 resentencing petitions that a court has already denied but are not yet final on appeal.For the reasons set forth below, we hold that Proposition 47 did not amend the Three Strikes Reform Act. Accordingly, we need not address whether, in People v. Chaney (S223676), the measure applies retroactively to Proposition 36 resentencing petitions that have been denied.

I. FACTS

Defendant David J. Valencia has a lengthy criminal record. In 1995, he was convicted of kidnapping. (§ 207, subd. (a).) In 1996, he was convicted of making criminal threats (§ 422), resisting arrest by threat or violence (§ 69), and driving under the influence of alcohol ( Veh. Code, § 23152, subd. (a) ). In 2000, he was convicted of corporal injury to a spouse or cohabitant. (§ 273.5.) In addition, between 1987 and 2007, he was convicted of 12 misdemeanors. Finally, in 2009, Valencia struck his wife during an argument over whether she should drive while drunk, causing a laceration on her head that was closed with staples. He was convicted a second time of corporal injury to a spouse or cohabitant (§ 273.5), which qualified as a third strike offense, and he was sentenced to an indeterminate term of 25 years to life.

In 2013, following the enactment of the Three Strikes Reform Act, Valencia petitioned for resentencing. He was eligible for resentencing because (1) his third strike was neither serious nor violent, as defined by statute, (2) his third strike was not among the other crimes excluded from the sentencing reforms,⁴ and (3) he had not suffered a prior conviction for a super strike. (See People v. Johnson (2015) 61 Cal.4th 674, 681-682, 189 Cal.Rptr.3d 794, 352 P.3d 366 ( Johnson ).)

4 A prisoner is not eligible for resentencing if his or her third strike offense involved large quantities of controlled substances (§ 1170.12, subd. (c)(2)(C)(i)) or is one of various enumerated sex offenses (id., subd. (c)(2)(C)(ii)). A prisoner is also ineligible if he or she used a firearm, was armed with a firearm or deadly weapon, or intended to cause great bodily injury. (Id., subd. (c)(2)(C)(iii).)

The People opposed resentencing, arguing that releasing Valencia would pose an unreasonable risk of danger to public safety based on his criminal history of alcohol abuse and domestic violence. In August 2013, after considering evidence from both parties and hearing testimony from Valencia and his mother, the sentencing court denied Valencia's petition: "I cannot grant this. I just feel that he is a danger. He is an unreasonable risk to public safety. And, in particular, to the women he's around. I just can't do this."

On appeal, Valencia argued that Proposition 47 had amended the Three Strikes Law by narrowing the sentencing court's discretion to deny resentencing on the ground that the petitioner posed an unreasonable risk to public safety. The Court of Appeal rejected this contention and affirmed the judgment. All three justices took the view that, read literally, Proposition 47's plain language—defining "unreasonable risk of danger to public safety" "[a]s used throughout this [Penal] Code"—applied to resentencing proceedings under Proposition 36. Two justices concluded, however, that "the literal meaning [of Proposition 47's amendment] does not comport with the purpose of the [Three Strikes Reform] Act, and applying it to resentencing proceedings under the [Three Strikes Reform] Act would frustrate, rather than promote, that purpose and the intent of the electorate in enacting both initiative measures." The third justice found the text's plain language insurmountable. But because he concluded that the electorate did not clearly manifest an intent to apply Proposition 47 retroactively to sentencing petitions like Valencia's that had already been denied, he concurred in the result.

In 2005, Clifford Paul Chaney was arrested for felony drunk driving after having been convicted of three separate incidents of driving under the influence in the prior 10 years. ( Veh. Code, § 23550.) In exchange for having other counts dismissed, Chaney admitted he had six prior 1984 convictions for armed robbery and for conspiracy to commit robbery, which were all serious or violent felonies, and was sentenced as a third strike offender for the 2005 felony drunk-driving conviction, receiving an indeterminate sentence of 25 years to life.

In 2013, Chaney unsuccessfully sought resentencing under the Three Strikes Reform Act. The resentencing court stated that it was unconvinced "that [Chaney] would not re-engage in alcohol use and place the public at risk." Accordingly, the court concluded that Chaney posed an "unreasonable risk of danger to public safety" (§ 1170.126, subd. (f)) and denied his petition.

On appeal, Chaney argued that Proposition 47's narrower definition of "unreasonable risk of danger to public safety" (§ 1170.18, subd. (c)) applied retroactively to resentencing proceedings under the Three Strikes Reform Act. (§ 1170.126.) The Court of Appeal rejected Chaney's argument and concluded that Proposition 47's definition of "unreasonable risk of danger to public safety" did not apply retroactively to him because his resentencing petition was decided before the effective date of Proposition 47.

II. DISCUSSION

A. The Three Strikes Reform Act (Proposition 36) and the Safe Neighborhoods and Schools Act (Proposition 47)

1. Proposition 36: The Three Strikes Reform Act

We recently summarized the Three Strikes Reform Act in Johnson, supra , 61 Cal.4th 674, 189 Cal.Rptr.3d 794, 352 P.3d 366. As we noted, "[p]rior to its amendment by the [Three Strikes Reform] Act, the Three Strikes law required that a defendant who had two or more prior convictions of violent or serious felonies receive a third strike sentence of a minimum of 25 years to life for any current felony conviction, even if the current offense was neither serious nor violent. (Former §§ 667, subds. (d), (e)(2)(A), 1170.12, subds. (b), (c)(2)(A).)" ( Id . at p. 680-681, 189 Cal.Rptr.3d 794, 352 P.3d 366.) We explained that Proposition 36 "amended the Three Strikes law with respect to defendants whose current conviction is for a felony that is neither serious nor violent. In that circumstance, unless an exception applies, the defendant is to receive a second strike sentence of twice the term otherwise provided for the current felony, pursuant to the provisions that apply when a defendant has one prior conviction for a serious or violent felony." ( Id . at p. 681, 189 Cal.Rptr.3d 794, 352 P.3d 366.) As noted above, Proposition 36 excluded from its reforms third strikes involving large quantities of controlled substances, specified sex offenses, and offenses committed when the defendant was armed or intended to cause great bodily harm. It also excluded defendants with prior convictions for super strikes. (See ante , fns. 3 & 4.)

Most pertinent here is Proposition 36's resentencing provision, which "provides a procedure by which some prisoners already serving third strike sentences may seek resentencing in accordance with the new sentencing rules. (§ 1170.126.)" ( Johnson , supra , 61 Cal.4th at p. 682, 189 Cal.Rptr.3d 794, 352 P.3d 366.) An inmate who is serving a third strike sentence that would have yielded a second strike sentence under Proposition 36's new sentencing rules "shall be resentenced" as second strike offender "unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§ 1170.126, subd. (f).)

In exercising its discretion to deny resentencing, the court has broad discretion to consider: (1) the inmate's "criminal conviction history, including the type of crimes committed, the extent of injury to victims, the length of prior prison commitments, and the remoteness of the crimes"; (2) his or her "disciplinary record and record of rehabilitation while incarcerated"; and (3) "[a]ny other evidence the court, within its discretion, determines to be relevant in deciding whether a new sentence would result in an unreasonable risk of danger to public safety." (§ 1170.126, subd. (g)(1)-(3).) Thus, as the Legislative Analyst explained in the Voter Information Guide, "[i]n determining whether an offender poses [an unreasonable risk of danger to public safety], the court could consider anyevidence it determines is relevant, such as the offender's criminal history, behavior in prison, and participation in rehabilitation programs." (Voter Information Guide, Gen. Elec. (Nov. 6, 2012) analysis of Prop. 36, p. 50, italics added.)

Following the enactment of Proposition 36, Courts of Appeal have rejected arguments that the phrase "unreasonable risk of danger to public safety," as used in section 1170.126, subdivision (f), is unconstitutionally vague. (See, e.g., People v. Garcia (2014) 230 Cal.App.4th 763, 769-770, 178 Cal.Rptr.3d 883 ; People v. Flores (2014) 227 Cal.App.4th 1070, 1075, 174 Cal.Rptr.3d 390 ["Surely a superior court judge is capable of exercising discretion, justly applying the public safety exception, and determining whether a lesser sentence would pose an unreasonable risk of harm to the public safety"].)

2. Proposition 47: The Safe Neighborhoods and Schools Act

Proposition 47 reclassified as misdemeanors certain drug- and theft-related offenses that previously were felonies or wobblers. (§ 1170.18, subds. (a), (b); see Health & Saf. Code §§ 11350 [possession of a controlled substance], 11357 [same], 11377 [possession of concentrated cannabis]; Pen. Code §§ 459.5 [shoplifting], 473 [forgery], 476a [writing bad checks], 490.2 [grand theft], 496 [receiving stolen property], and 666 [petty theft with a prior conviction].) Proposition 47 also added a provision allowing felony offenders "serving a sentence for a conviction" for offenses now reclassified as misdemeanors to petition to have their sentences recalled and to be resentenced. (§ 1170.18, subd. (a); see also People v. Morales (2016) 63 Cal.4th 399, 404, 203 Cal.Rptr.3d 130, 371 P.3d 592.)

The resentencing procedure provides that if an inmate "would have been guilty of a misdemeanor" had Proposition 47 been in effect at the time of the offense, and he or she has no prior convictions for super strikes or any offense that requires registration as a sex offender, the inmate may petition for a recall of his or her sentence and resentencing in accordance with Proposition 47's reclassification of certain offenses as misdemeanors. (§ 1170.18, subds. (a)-(i).) If the court determines that the petitioner meets these criteria, it must recall the felony sentence and resentence the petitioner based on the new classification of the offense as a misdemeanor, "unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§ 1170.18, subd. (b).)

In exercising this discretion, the resentencing court may consider: (1) the petitioner's "criminal conviction history, including the type of crimes committed, the extent of injury to victims, the length of prior prison commitments, and the remoteness of the crimes"; (2) his or her "disciplinary record and record of rehabilitation while incarcerated"; and (3) "[a]ny other evidence" the court deems relevant. (§ 1170.18, subd. (b)(1)-(3).)

Significantly, however, in contrast to Proposition 36, Proposition 47 limits the trial court's discretion to deny resentencing by defining the phrase "unreasonable risk of danger to public safety" narrowly. In connection with resentencing under Proposition 47, " 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit a [super strike]." (§ 1170.18, subd. (c); see ante , fn. 3.) Thus, under Proposition 47 a resentencing court may not deny a petition for reclassification and resentencing for certain theft and drug possession felonies to misdemeanors for an otherwise eligible petitioner unless it finds that the resentencing would pose an unreasonable risk that the petitioner will commit a super strike.

Of most significance to the present matter, Proposition 47 also provides that its definition of "unreasonable risk of danger to public safety" is effective "[a]s used throughout this Code." (§ 1170.18, subd. (c).) Because section 1170.126, subdivision (g)(3), added by Proposition 36, the Three Strikes Reform Act, also requires the court to assess whether the petitioner's resentencing would pose "an unreasonable risk of danger to public safety," this raises the question of whether Proposition 47 amended the resentencing criteria for eligible third strike offenders under the previously enacted measure, Proposition 36. The parties do not contest that such an amendment would be more favorable to three strike inmates and result in the release of more recidivist serious and/or violent offenders than had been originally contemplated under Proposition 36.

B. Proposition 47's Definition of "Unreasonable Risk of Danger to Public Safety" Does Not Apply to the Three Strikes Reform Act

Viewed in isolation, Proposition 47's statement that its definition of " 'unreasonable risk of danger to public safety' " is to be "used throughout this Code" (§ 1170.18, subd. (c)) would be understood to result in the amendment of the entire Penal Code, which of course includes the resentencing criteria for the Three Strikes Reform Act. But our courts have recognized that the meaning of isolated statutory language can be informed by and indeed must be consistent with the provisions of the relevant statute as whole. In the present matter, there are several reasons why the meaning of the phrase "[a]s used throughout this Code" in section 1170.18, subdivision (c) should be subject to careful scrutiny.

First, the phrase "[a]s used throughout this Code" in section 1170.18, subdivision (c), when considered in the context of both section 1170.18 and other provisions of Proposition 47 as a whole, renders the entirety of section 1170.18, subdivision (c) ambiguous as to its potential application to the resentencing criteria for the Three Strikes Reform Act. Second, the ballot materials for Proposition 47 supplied no notice to voters that the measure intended to amend the resentencing criteria of the Three Strikes Reform Act and that three strike inmates previously convicted of serious or violent felonies could be released as a consequence. Third, despite creating detailed procedures for the resentencing of low-level felons to misdemeanants, Proposition 47 provided no similar procedural guidance for the resentencing of three strike inmates under its definition of "unreasonable risk of danger to public safety," which further detached any perceived connection with the Three Strikes Reform Act.

Finally, under the circumstances of this matter, it is unreasonable to assign dispositive significance to the legal presumptions we normally apply to voters who approve an initiative in order to interpret the phrase "[a]s used throughout this Code." Those presumptions, even if applicable, would not alter our conclusion that the statutory language is ambiguous and that this ambiguity should be resolved by construing section 1170.18, subdivision (c) not to apply to petitioners under other provisions of the Penal Code such as the Three Strikes Reform Act. Furthermore, as we will explain, neither the initiative's text nor its supporting materials describe any intention to amend the criteria for the resentencing of recidivist serious or violent felons, and both the Attorney General, who is required by law to summarize ballot measures, and the Legislative Analyst, who is required by law to provide and explain to voters a measure's potential impacts, did not interpret the phrase "[a]s used throughout this Code" as referring to the sentencing criteria for the Three Strikes Reform Act. Given these circumstances, it is unreasonable to assume or presume that voters, with greater acumen than the legal professionals of the offices of the Attorney General and Legislative Analyst, somehow discerned a connection with the Three Strikes Reform Act. Thus, those legal presumptions provide no assistance in determining whether such a connection existed. 1. The Rules of Statutory Interpretation

We have long recognized that the language used in a statute or constitutional provision should be given its ordinary meaning, and "[i]f the language is clear and unambiguous there is no need for construction, nor is it necessary to resort to indicia of the intent of the Legislature (in the case of a statute) or of the voters (in the case of a provision adopted by the voters)." ( Lungren v. Deukmejian (1988) 45 Cal.3d 727, 735, 248 Cal.Rptr. 115, 755 P.2d 299.) To that end, we generally must "accord [ ] significance, if possible, to every word, phrase and sentence in pursuance of the legislative purpose," and have warned that "[a] construction making some words surplusage is to be avoided." ( Dyna-Med, Inc. v. Fair Employment & Housing Com. (1987) 43 Cal.3d 1379, 1387, 241 Cal.Rptr. 67, 743 P.2d 1323.)

But "[t]he words of the statute must be construed in context, keeping in mind the statutory purpose, and statutes or statutory sections relating to the same subject must be harmonized, both internally and with each other, to the extent possible."( Dyna-Med, Inc. v. Fair Employment & Housing Com. , supra , 43 Cal.3d at p. 1387, 241 Cal.Rptr. 67, 743 P.2d 1323.) "Where uncertainty exists consideration should be given to the consequences that will flow from a particular interpretation." ( Ibid . )

2. Prior Cases Involving Ambiguous Statutory Language

A reason to further explore the meaning of statutory language and to consider extrinsic evidence of legislative intent is where statutory language is ambiguous when considered "in the context of the statute and initiative as a whole." ( People v. Superior Court (Pearson) (2010) 48 Cal.4th 564, 571, 107 Cal.Rptr.3d 265, 227 P.3d 858.) This is because an initiative's " 'language must also be construed in the context of the statute as a whole' " and its " 'overall ... scheme.' " ( Professional Engineers in California Government v. Kempton (2007) 40 Cal.4th 1016, 1037, 56 Cal.Rptr.3d 814, 155 P.3d 226, quoting People v. Rizo (2000) 22 Cal.4th 681, 685, 94 Cal.Rptr.2d 375, 996 P.2d 27.)

In Building Industry Assn. v. City of Camarillo (1986) 41 Cal.3d 810, 226 Cal.Rptr. 81, 718 P.2d 68, this court interpreted the meaning of the phrase "governing body," as used in Evidence Code former section 669.5, subdivision (a), a statute that sought to limit growth control legislation "enacted by the governing body of a city, county, or city and county." Opponents of a local growth control ordinance, adopted by city voters as a local initiative, claimed the ordinance was invalid because it did not comply with the requirements of Evidence Code section 669.5. This court acknowledged that, under its plain meaning, "[t]he term 'governing body' excludes the electorate." ( Building Industry Assn. , supra , 41 Cal.3d at p. 818, 226 Cal.Rptr. 81, 718 P.2d 68.) But the court further noted that "the words 'governing body' cannot be considered by themselves, but must be analyzed within the context of the entire statute," which also included an exception for certain voter-approved initiatives, and when so read, "[a]mbiguity does exist within [Evidence Code] section 669.5." ( Id . at pp. 818-819, 226 Cal.Rptr. 81, 718 P.2d 68, italics omitted.) This court then reviewed the legislative history of Evidence Code section 669.5 in order to reach the conclusion that it does apply to growth control legislation enacted by local initiatives. ( Id . at pp. 819-820, 226 Cal.Rptr. 81, 718 P.2d 68.)

In Kennedy Wholesale, Inc. v. State Bd. of Equalization (1991) 53 Cal.3d 245, 279 Cal.Rptr. 325, 806 P.2d 1360, this court interpreted a provision of article XIII A, section 3 of the California Constitution (hereafter, article XIII A, section 3 ) that stated that any changes in state taxes collected for increasing revenue " 'must be imposed by an Act passed by no less than two-thirds of all members elected to each of the two houses of the Legislature....' " ( Id . at p. 248, 279 Cal.Rptr. 325, 806 P.2d 1360, quoting Cal. Const., art. XIII A, § 3.) Opponents of Proposition 99, a measure passed by the voters in 1988 that increased the tax on tobacco products, contended that article XIII A, section 3 rendered the measure void because only a supermajority of the Legislature may impose such taxes. This court acknowledged that, considered in isolation, the plain meaning of article XIII A, section 3 would have such an effect. ( Kennedy Wholesale , supra , 53 Cal.3d at p. 249, 279 Cal.Rptr. 325, 806 P.2d 1360.) Instead, however, the court found section 3"ambiguous when read in the context of the whole Constitution." ( Ibid . ) We construed section 3 in light of article IV, section 1 of the state Constitution, through which "the people expressly 'reserve to themselves the powers of initiative and referendum.' " ( Kennedy Wholesale , at p. 249, 279 Cal.Rptr. 325, 806 P.2d 1360, quoting Cal. Const., art. IV, § 1.) Accordingly, we concluded that "[t]o interpret section 3 as giving the Legislature exclusive power to raise taxes would implicitly repeal article IV, section 1, pro tanto" even though article XIII A, section 3"does not even mention the initiative power, let alone purport to restrict it." ( Kennedy Wholesale , at p. 249, 279 Cal.Rptr. 325, 806 P.2d 1360.) This court then turned to the ballot materials of the measure that enacted article XIII A, section 3 and found "[n]othing in the official ballot pamphlet supports the inference that the voters intended to limit their own power to raise taxes in the future by statutory initiative." ( Kennedy Wholesale , at p. 250, 279 Cal.Rptr. 325, 806 P.2d 1360.)

In People v. Hazelton (1996) 14 Cal.4th 101, 58 Cal.Rptr.2d 443, 926 P.2d 423, we interpreted a provision enacted in 1994 by Proposition 184, the initiative version of the Three Strikes law that mandated life sentences as the third strike penalty for those who have " 'two or more prior felony convictions, as defined in paragraph (1) of subdivision (b) ' " of former section 1170.12. ( Hazelton , supra , 14 Cal.4th at p. 105, 58 Cal.Rptr.2d 443, 926 P.2d 423, quoting former § 1170.12, subd. (c)(2)(A).) At issue was whether former section 1170.12 permitted third strike penalties for similar out-of-state prior felony convictions, which were defined not in paragraph (1) but in paragraph (2) of subdivision (b) in section 1170.12. Instead, paragraph (1) defined an eligible prior felony conviction as "[a]ny offense defined in subdivision (c) of Section 667.5 as a violent felony or any offense defined in subdivision (c) of Section 1192.7 as a serious felony in this state." (Former § 1170.12, subd. (b)(1), as added by Prop. 184, Gen. Elec. (Nov. 8, 1994).) The defendant in Hazelton contended that the life sentence required by former section 1170.12, subdivision (c)(2)(A), should not apply to him because his prior out-of-state conviction was not included in "paragraph (1) of subdivision (b)," and the life sentence provision did not expressly refer to out-of-state felony convictions described in paragraph (2) of subdivision (b) in former section 1170.12.

We rejected this argument by concluding that "[r]ead in the context of section 1170.12 as a whole, the language of subdivision (c)(2)(A) is ambiguous regarding the inclusion of out-of-state convictions...." ( Hazelton , supra , 14 Cal.4th at p. 105, 58 Cal.Rptr.2d 443, 926 P.2d 423.) This court observed that introductory language in section 1170.12, subdivision (b) stated: " 'Notwithstanding any other provision of law and for the purposes of this section , a prior conviction of a felony shall be defined as:' " followed by the list of subparts that included both paragraph (1) and paragraph (2) of subdivision (b). ( Hazelton , at p. 106, 58 Cal.Rptr.2d 443, 926 P.2d 423, quoting former § 1170.12, subd. (b).) When read in light of this introductory language, we concluded that "[s]ubdivision (c)(2)(A), therefore, arguably proffers a different and inconsistent definition to the extent it limits the definition of a prior felony for purposes of the three strikes provision to only adult California convictions." ( Hazelton , at p. 106, 58 Cal.Rptr.2d 443, 926 P.2d 423.) As a result, we turned to the ballot materials of the initiative that enacted section 1170.12 and found "no evidence in the legislative history that the voters intended to exclude out-of-state convictions from the purview of the initiative's third strike penalty." ( Hazelton, at p. 108, 58 Cal.Rptr.2d 443, 926 P.2d 423.)

Consequently, Building Industry Assn. , Kennedy Wholesale , and Hazelton all established that statutory language, even if it appears to have a clear and plain meaning when considered in isolation, may nonetheless be rendered ambiguous when the language is read in light of the statute as a whole or in light of the overall legislative scheme.

3. Ambiguity of Section 1170.18's Application to the Three Strikes Reform Act

Section 1170.18, subdivision (c) states: "As used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit a new violent felony within the meaning of" section 667, subdivision (e)(2)(C)(iv). The dissents argue this language should have made it plain to voters that only offenders at risk of committing a super strike could be denied resentencing under either Proposition 47 or the Three Strikes Reform Act. But the initiative's language is ambiguous when viewed in the context of the subject matter of Proposition 47 as a whole and, in particular, in the context of section 1170.18 in which subdivision (c) is located. The ambiguity is compounded because the alleged effect on the Three Strikes Reform Act is not reflected in the uncodified provisions of Proposition 47 that set forth the purposes of the measure.

First, although the words "as used throughout this code" in section 1170.18, subdivision (c) may appear clear and unambiguous when viewed in isolation, the ambiguity regarding the intended meaning and scope of subdivision (c) becomes apparent when this language is viewed in context. To begin with, the subdivision was enacted as part of an initiative measure—Proposition 47—whose primary focus was reducing the punishment for a specifically designated category of low-level felonies from felony to misdemeanor sentences. The measure did not purport to alter the sentences for felonies other than those that the measure reduced to misdemeanors.Moreover, section 1170.18, subdivision (c) itself contains language that reasonably supports the conclusion that it was intended to apply only to persons who are eligible for resentencing to misdemeanors under the provisions of Proposition 47. To reiterate, section 1170.18, subdivision (c) states: "As used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit a new violent felony...." (Italics added.) As the Attorney General contends, the use of the phrase "the petitioner" in section 1170.18, subdivision (c) is significant because section 1170.18, subdivision (a), provides that an eligible person may file a petition for a recall of sentence in order to reduce the conviction to a misdemeanor, and, thereafter, the statute repeatedly refers to such a person as "the petitioner." (§§ 1170.18, subds. (b), (c), (l), (m).) For the reasons described in Justice Kruger's concurring opinion, it is unclear whether the petitioners referred to in section 1170.18, subdivision (c) include "any petitioner in any proceeding governed by any provision of the Penal Code, including Proposition 36" or whether it refers "only to the lower-level offenders who have petitioned for relief under Penal Code section 1170.18, subdivision (a)." (Conc. opn. of Kruger, J., post , 220 Cal.Rptr.3d at pp. 260–261, 397 P.3d at p. 962.) Moreover, the use of the definite article "the," instead of "a" petitioner, is an indication that Proposition 47's definition of "unreasonable risk of danger to public safety" applies only to those individuals applying for relief under Proposition 47.

If Proposition 47's definition of "unreasonable risk of danger to public safety" was intended to apply to petitioners both under Proposition 36 and Proposition 47, then section 1170.18, subdivision (c), should have referred to "a petitioner" or otherwise made clear that it applied both to individuals seeking relief under section 1170.18, subdivision (a), and to those seeking relief under section 1170.126, subdivision (b), part of the Three Strikes Reform Act. To read the statute as applying to the latter individuals would transform the meaning of the phrase "the petitioner."

Additionally, the location of the provision in question within section 1170.18 is informative. As noted, section 1170.18 creates the procedure through which those persons who, having been previously convicted and sentenced for those low-level felonies now reducible to misdemeanors, may petition to have their convictions resentenced as misdemeanors.

Section 1170.18 begins by listing the previously designated low-level felony convictions that are eligible for relief under Proposition 47 in subdivision (a). This is followed by subdivision (b), which describes the criteria a resentencing court may apply in granting or denying relief to Proposition 47 eligible offenders. ( § 1170.18, subd. (b).) The next subdivision of section 1170.18 is subdivision (c), and it defines the phrase "unreasonable risk of danger to public safety," which is one of the resentencing criteria listed in subdivision (b) for the resentencing of the low-level felonies described in subdivision (a). Thus, the placement of subdivision (c) tends to reinforce its application to crimes that had previously been low-level felonies. Given that nothing in the subject matter of section 1170.18 as a whole suggests a connection to the Three Strikes Reform Act or serious and violent felonies, the placement of subdivision (c) in section 1170.18 obscures any indication that the subdivision was intended to amend the resentencing criteria for the Three Strikes Reform Act. (See, e.g., Lonicki v. Sutter Health Central (2008) 43 Cal.4th 201, 210-211, 74 Cal.Rptr.3d 570, 180 P.3d 321 [relying upon the location of statutory provision in ascertaining its meaning and scope].)

It is true that the use of the phrase "[a]s used throughout this Code" in section 1170.18, subdivision (c) could support an alternative interpretation of its intended scope, under which the limiting definition it sets forth would apply to persons seeking resentencing under other statutory provisions in addition to section 1170.18. For this reason, these two opposing interpretations render section 1170.18, subdivision (c) ambiguous. But, as further explained below, the interpretation that takes into account both the meaning of the phrase "the petitioner" and the location of subdivision (c) itself within section 1170.18 appears more consistent with the express purpose and intended scope of Proposition 47, as stated in its uncodified sections, which we describe below.

Second, another indication that section 1170.18, subdivision (c) is ambiguous rests on the fact that the subdivision, as construed by defendant and the dissenting opinions, is difficult to reconcile with Proposition 47's uncodified preamble, which was presented to the electorate. The preamble, which describes Proposition 47 to the voters, contains uncodified sections expressing the measure's "Findings and Declarations" and its "Purpose and Intent." (Voter Information Guide, Gen. Elec. (Nov. 4, 2014) (Voter Information Guide) text of Prop. 47, §§ 2, 3, p. 70.) "In considering the purpose of legislation, statements of the intent of the enacting body contained in a preamble, while not conclusive, are entitled to consideration." ( People v. Canty (2004) 32 Cal.4th 1266, 1280, 14 Cal.Rptr.3d 1, 90 P.3d 1168.) As we explain, these uncodified provisions are inconsistent with and beyond the scope of any intention to amend the resentencing criteria for the Three Strikes Reform Act or to release additional three strike inmates.

The initiative's Findings and Declarations state that the "act ensures that sentences for people convicted of dangerous crimes like rape, murder, and child molestation are not changed." (Voter Information Guide, supra , text of Prop. 47, § 2, p. 70.) Similarly, Proposition 47 reiterates its Purpose and Intent as being to "[e]nsure that people convicted of murder, rape, and child molestation will not benefit from this act," and to "[r]equire misdemeanors instead of felonies for nonserious, nonviolent crimes like petty theft and drug possession, unless the defendant has prior convictions for specified violent or serious crimes." (Id ., § 3, p. 70.) The Purpose and Intent of the initiative is also to "[a]uthorize consideration of resentencing for anyone who is currently serving a sentence for any of the offenses listed herein that are now misdemeanors." (Ibid .)

Thus, the expressly stated central objectives of Proposition 47 were to redesignate specified minor felony offenses as misdemeanors and to permit those persons previously convicted of these same low-level felonies to seek resentencing as misdemeanors. Proposition 47 further assured voters that persons convicted of murder, rape, and child molestation would not benefit from Proposition 47 and specifically declared that the "act ensures that sentences for people convicted of dangerous crimes ... are not changed." (Voter Information Guide, supra , text of Prop. 47, § 2, p. 70, italics added.)

It is not unreasonable to view the term "dangerous crimes" as encompassing the serious and violent felonies addressed by both the Three Strikes law and the Three Strikes Reform Act. The uncodified introductory provisions of Proposition 47 further assured voters that the measure would reduce low-level felony convictions to misdemeanors, "unless the defendant has prior convictions for specified violent or serious crimes." (Voter Information Guide, supra , text of Prop. 47, § 3, p. 70.) By design, those convicted under either the former Three Strikes law or the amended version enacted by Proposition 36 are persons convicted of dangerous crimes and have prior convictions for various violent or serious crimes. These uncodified introductory provisions, therefore, are inconsistent with any intention to make the resentencing provisions of the Three Strikes Reform Act more favorable to the resentencing and release of three strike inmates, who are felons with recidivist convictions for serious or violent felonies.

Also, an extension of Proposition 47's definition of "an unreasonable risk of danger to public safety" to the previously enacted Three Strikes Reform Act would conflict with the measure's stated purpose to "[e]nsure that people convicted of murder, rape, and child molestation will not benefit from this act." (Voter Information Guide, supra , text of Prop. 47, § 3, p. 70.)

As previously discussed, the list of super strike offenses is limited to a set of eight categories of serious or violent felonies. (See, ante , 220 Cal.Rptr.3d at p. 235, fn. 3, 397 P.3d at p. 940, fn. 3.) The only sexual offenses included in that set are specific sexually violent offenses involving threats, force, violence, duress, menace, or fear of immediate and unlawful bodily injury and specific sexual offenses involving minors under the age of 14. (See § 667, subd. (e)(2)(C)(iv) (I-VIII).) Thus, contrary to Proposition 47's Purpose and Intent, three strike inmates previously convicted of nonforcible rape or molestation of a child over the age of 14 would stand to benefit under Proposition 47's definition of "an unreasonable risk of danger to public safety." Proposition 47's uncodified Purpose and Intent, therefore, appears inconsistent with the application of the phrase "[a]s used throughout this Code," under its section 1170.18, subdivision (c), which would have the effect of extending possible sentencing relief to convicted rapists and child molesters.

Construing the phrase "[a]s used throughout this Code" as a more expansive standard to permit the resentencing of three strikes offenders would thus be inconsistent with Proposition 47's uncodified findings, declarations, purpose, and intent. This stark inconsistency reinforces the conclusion that the meaning of section 1170.18, subdivision (c) is ambiguous.

4. The Voter Information Materials Concerning Proposition 47

Accordingly, we turn to evidence, outside the measure's express provisions, to ascertain the voter's intent in approving the initiative. Specifically, we examine the materials that were before the voters.⁵ ( Robert L. v. Superior Court (2003) 30 Cal.4th 894, 905, 135 Cal.Rptr.2d 30, 69 P.3d 951 ( Robert L. ) [in order resolve questions of purpose and ambiguity, "we look to the materials that were before the voters"].)

5 Even assuming a few advocacy groups may have predicted that Proposition 47 could affect three strikes resentencing, there is no evidence that the electorate as a whole, or indeed any significant part of it, was aware of such statements. (See Farmers Ins. Exchange v. Superior Court (2006) 137 Cal.App.4th 842, 857, 40 Cal.Rptr.3d 653 ["We cannot presume that the electorate as a whole was aware of statements made in an article published in legal periodicals before the election, or in a magazine article published after the election...."].)

Nothing in the materials accompanying the text of Proposition 47 suggested that the initiative would alter the resentencing criteria under the previously enacted Three Strikes Reform Act, resulting in the potential release of additional recidivist serious or violent felony offenders. "We cannot presume that ... the voters intended the initiative to effect a change in law that was not expressed or strongly implied in either the text of the initiative or the analyses and arguments in the official ballot pamphlet." ( Farmers Ins. Exchange v. Superior Court , supra , 137 Cal.App.4th at pp. 857-858, 40 Cal.Rptr.3d 653.)⁶

6 We recognize that the materials in the ballot pamphlet may not touch on every aspect of an initiative, no matter how minor. "A statute, of course, must prevail over any summary. Were it not so, no statute could ever be enacted whole and entire. For every summary, by definition, is incomplete." (In re Cervera (2001) 24 Cal.4th 1073, 1079-1080, 103 Cal.Rptr.2d 762, 16 P.3d 176.) When, for example, an initiative contains a clear and unambiguous provision that, because of its relatively limited significance, is not mentioned in ballot summary or arguments, the absence of such a reference will not nullify its effectiveness. Here, however, the language of Proposition 47 is not free from ambiguity. And the application of its definition of dangerousness to inmates seeking resentencing under the Three Strikes Reform Act is a matter of such substantial import that the voters could reasonably expect that, if Proposition 47 applied to such inmates, the ballot materials would mention it.

Proposition 47 voter information materials focused on describing the reclassification of certain theft- and drug-related felonies to misdemeanors and did not make any reference to three strike inmates. For instance, the "Official Title and Summary," which was prepared by the Attorney General, states that the measure "[r]equires [a] misdemeanor sentence instead of [a] felony for certain drug possession offenses," as well as "petty theft, receiving stolen property, and forging/writing bad checks." (Voter Information Guide, supra , Summary of Prop. 47, p. 34.) The Official Title and Summary further clarifies that a "felony sentence for these offenses" would remain intact if the person "has [a] previous conviction for crimes such as rape, murder, or child molestation or is [a] registered sex offender" or unless the court "finds [an] unreasonable public safety risk." (Ibid ., italics added.) The Official Title and Summary makes no mention of the Three Strikes Law, the Three Strikes Reform Act, three strike inmates, or life sentences for recidivist serious or violent felons.

Nor does the Legislative Analyst's Analysis of Proposition 47 included in the Voter Information Guide make any reference to the Three Strikes Law, the Three Strikes Reform Act, three strike inmates, or life sentences for recidivist serious or violent felons. Under the Elections Code, the Legislative Analyst must "prepare an impartial analysis of the measure describing the measure and including a fiscal analysis of the measure showing the amount of any increase or decrease in revenue or cost to state or local government." ( Elec. Code, § 9087, subd. (a).) Accordingly, if the Legislative Analyst had recognized that Proposition 47 would have also negated life sentences and facilitated the resentencing and release of third strike inmates under the more restrictive definition of an "unreasonable risk of danger to public safety," the Analyst would have been required to provide the voters with an estimate of the fiscal scope of such an application. Yet the analysis provides no estimate of any effect on third strike offenders and their crimes nor any discussion of the fiscal effects on the state prison system from the invalidation of life sentences for recidivist serious or violent felons or on the parole system in handling the release of inmates caused by a resentencing of third strike inmates under a standard that would be more favorable to them.

In addition, the Legislative Analyst must provide an analysis that is "easily understood by the average voter" and it "may contain background information, including the effect of the measure on existing law and the effect of enacted legislation which will become effective if the measure is adopted, and shall generally set forth in an impartial manner the information the average voter needs to adequately understand the measure." ( Elec. Code, § 9087, subd. (b).)

However, the analysis mentions no "effect" on the "existing" Three Strikes Reform Act. Instead, it focuses on Proposition 47's proposal to reduce punishment for nonserious and nonviolent felony crimes. In explaining what Proposition 47 proposed, the Legislative Analyst declared: "This measure reduces penalties for certain offenders convicted of nonserious and nonviolent property and drug crimes," and would allow "certain offenders who have been previously convicted of such crimes to apply for reduced sentences." (Voter Information Guide, supra , analysis of Prop. 47, p. 35, italics added.) The analysis then lists the offenses for which the measure would reduce punishment—grand theft, shoplifting, receiving stolen property, writing bad checks, check forgery, and drug possession. (Id . at pp. 35-36.) Thus, the Legislative Analyst discerned that Proposition 47 would provide reduced punishment for only nonserious and nonviolent property and drug offenses.⁷

7 In his dissenting opinion, Justice Liu correctly notes that the analysis provided by the Legislative Analyst "may contain background information, including the effect of the measure on existing law." (Elec. Code, § 9087, subd. (b), italics added.) Quoting Justice Rushing's majority opinion in People v. Cordova (2016) 248 Cal.App.4th 543, 203 Cal.Rptr.3d 700 (Cordova ), however, his dissent further contends that the Analyst, under these circumstances, is only required to make " 'a rational judgment about what effects are most likely to matter to voters.' " (Dis. opn. of Liu, J., post, 220 Cal.Rptr.3d at pp. 275–276, 397 P.3d at pp. 974–975, quoting Cordova, supra, 248 Cal.App.4th at p. 559, 203 Cal.Rptr.3d 700.) But the majority in Cordova cited no authority for its described scope of the Analyst's duties. More important, the majority in Cordova held that Proposition 47's definition of "unreasonable risk of danger to public safety" applies to determinations of dangerousness under the Three Strikes Reform Act of 2012 (Cordova, at p. 578, 203 Cal.Rptr.3d 700 ), and we granted review of it (People v. Cordova, rev. granted Aug. 31, 2016, S236179). Our decision today, therefore, effectively overrules the majority decision in Cordova. (Cal. Rules of Court, rule 8.1115, subd. (e)(2) ["any published opinion of a Court of Appeal in a matter in which the Supreme Court has ordered review and deferred action pending the decision, is citable and has binding or precedential effect, except to the extent it is inconsistent with the decision of the Supreme Court or is disapproved by that court"].) Moreover, assuming, without deciding, that Cordova accurately describes the scope of the Analyst's duty under the Elections Code, it would seem logical to conclude that it would matter to voters whether a measure under their consideration might result in the release of additional recidivist serious or violent felony offenders.

Turning to the assessments of Proposition 47's impacts, the Legislative Analyst again focused solely on the effect of the reduction of certain crimes from felonies to misdemeanors: "We estimate that about 40,000 offenders annually are convicted of the above crimes [grand theft, shoplifting, receiving stolen property, writing bad checks, check forgery, and drug possession] and would be affected by the measure." (Voter Information Guide, supra , analysis of Prop. 47, at p. 36.) Furthermore, the Legislative Analyst expressed uncertainty regarding the measure's fiscal impact, noting that "[i]n particular, it would depend on the way individuals are currently being sentenced for the felony crimes changed by this measure." (Ibid .) There was no discussion of the probable fiscal effects related to the resentencing of third strike inmates, especially notable considering the significant costs associated with life sentences for recidivist serious or violent felons.

Finally, the arguments in favor of and opposed to Proposition 47 discussed only the reductions of certain felonies to misdemeanors and did not mention the previously enacted Three Strikes Reform Act.⁸ (Voter Information Guide, supra , arguments regarding Prop. 47, at pp. 38-39.)

8 In his dissenting opinion, Justice Cuéllar contends that the ballot arguments made by opponents of Proposition 47 warned voters that persons with violent criminal histories might benefit from the measure. (Dis. opn. of Cuéllar, J., post, 220 Cal.Rptr.3d at p. 290, 397 P.3d at p. 986.) This ballot argument, however, is made in the context of discussing only the 10,000 felons whose theft- and drug- related convictions would be reduced under Proposition 47. As previously discussed, the Legislative Analyst made no estimated count of any three strike inmates that would benefit from the measure nor did the drafters or proponents of the measure provide such a number.

Accordingly, nothing in the voter information pamphlet for Proposition 47 informed the electorate that the definition of "unreasonable risk of danger to public safety," contained in section 14 of the measure, proposed section 1170.18, subdivision (c), was intended to amend and narrow the resentencing criteria under Proposition 36, the Three Strikes Reform Act. The materials for Proposition 47 signaled no relationship at all with that prior initiative. In fact, based on the analysis and summary they prepared, there is no indication that the Legislative Analyst or the Attorney General were even aware that the measure might amend the resentencing criteria governing the Three Strikes Reform Act.

5. The Absence of Procedures for Resentencing Third Strike Offenders under the New Definition

Further weighing against the defendants' proposed interpretation of the phrase "[a]s used throughout this Code" is the fact that Proposition 47 provides no guidance whatsoever concerning how to implement the resentencing of three strike inmates under the new definition of "unreasonable risk of danger to public safety." (§ 1170.18, subd. (c).) Given the effective date of Proposition 47 in 2014, any amendatory effect on the previously enacted 2012 Three Strikes Reform Act raises serious procedural questions that the measure does not address. This circumstance further widens the gap concerning the voters' understanding of whether Proposition 47 would amend the resentencing criteria for three strike inmates.

As previously described, Proposition 36, the Three Strikes Reform Act, took effect on November 7, 2012, and it gave inmates "two years after the effective date of the act" or "at a later date upon a showing of good cause" to file any petition for resentencing. (§ 1170.126, subd. (b).) Proposition 47 took effect on November 5, 2014, two days short of the two-year window enacted by the Three Strikes Reform Act to file a resentencing petition. Assuming Proposition 47 were construed to amend the Three Strikes Reform Act, its provisions are silent with respect to whether its definition of "an unreasonable risk of danger to public safety" in section 1170.18, subdivision (c), would apply only to inmates who had not yet sought resentencing, or whether it would also apply to inmates whose resentencing petitions had already been filed but not acted upon, and to those whose petitions had already been denied.

An application only to inmates who had not yet sought resentencing would mean that such inmates would have only two days to seek resentencing under the new definition, unless the passage of Proposition 47 could constitute "good cause" for an untimely resentencing petition under section 1170.126, subdivision (b). In any event, it is reasonable to infer that, by November 5, 2014, most of the three strike inmates eligible for resentencing under the Three Strikes Reform Act would have already petitioned for resentencing.Even if we were to assume without deciding that the new definition may apply to any inmate who had filed a resentencing petition, but the proceedings on that petition were not yet final (see In re Estrada (1965) 63 Cal.2d 740, 744, 48 Cal.Rptr. 172, 408 P.2d 948 [without legislative guidance to the contrary, courts ordinarily presume that newly enacted legislation ameliorating criminal punishment is intended to apply to all cases not yet reduced to final judgment on the statute's effective date] ), such an application would serve, only a subset of third strike inmates—those who had unsuccessfully petitioned for resentencing and whose judgments on those petitions were not final as of November 5, 2014. Although Proposition 47 created new procedures for the resentencing of felons as misdemeanants, it created no similar procedural mechanism enabling previously unsuccessful third strike petitioners to seek a second opportunity for resentencing using the new definition of "an unreasonable risk of danger to public safety." (§ 1170.18, subd. (c).)

If one of the purposes of Proposition 47 was to narrow the criteria applicable to resentencing third strike offenders, one would expect the drafters to have addressed at least some of the above described problems. Instead, the measure was silent on these matters.⁹ This further exacerbated the unlikelihood that the Attorney General, the Legislative Analyst, and the voters understood that the measure would potentially release not just low-level, nonviolent offenders, but also three strike inmates with repeat convictions for serious and violent felonies. In fact, had Proposition 47 tried to answer the questions surrounding its procedural applicability to the Three Strikes Reform Act, it would have alerted the public to its proposed direct applicability to that prior measure.¹⁰

9 As defendants note, following the November 2014 election, one of Proposition 47's drafters, a law professor at Stanford School of Law, stated that three strike inmates who had been previously denied relief under the resentencing provisions of the Three Strikes Reform Act "could return to court and cite Proposition 47's new definition of an 'unreasonable risk of danger.' " (St. John & Gerber, Prop. 47 jolts landscape of California justice system, L.A. Times (Nov. 5, 2014) online at < http://www.latimes.com/local/politics/la-me-ff-pol-proposition47-20141106-story.html> [as of July 3, 2017].) But " '[t]he opinion of drafters or legislators who sponsor an initiative is not relevant since such opinion does not represent the intent of the electorate and we cannot say with assurance that the voters were aware of the drafters' intent.' " (Robert L., supra, 30 Cal.4th at p. 904, 135 Cal.Rptr.2d 30, 69 P.3d 951.) This caution applies with even greater force here where a single drafter expressed an intent and interpretation only after the passage of the measure.

10 Of course, as the dissenters point out, the Legislature can always enact necessary procedures to implement directives in a voter-enacted measure. But the utter absence of procedures for revised three strikes resentencing is significant because the measure included detailed procedures for resentencing Proposition 47 petitioners. Without similar attention to three strikes resentencing, or indeed any mention of the Three Strikes Reform Act, the measure gave voters no notice of a potential change in that law.

6. The Presumptions Concerning an Initiative Adopted by Voters

In favor of applying the defendants' proposed interpretation, defendants and the dissents rely heavily on two presumptions we generally apply to measures approved by the initiative process. The first is the assumption that voters who approve an initiative are presumed to " 'have voted intelligently upon an amendment to their organic law, the whole text of which was supplied [to] each of them prior to the election and which they must be assumed to have duly considered....' " ( Amador Valley Joint Union High Sch. Dist. v. State Bd. of Equalization (1978) 22 Cal.3d 208, 243-244, 149 Cal.Rptr. 239, 583 P.2d 1281, quoting Wright v. Jordan (1923) 192 Cal. 704, 713, 221 P. 915.) The second is a presumption, which we also apply to the Legislature, that the voters, in adopting an initiative, did so being "aware of existing laws at the time the initiative was enacted." ( Professional Engineers in California Government v. Kempton , supra , 40 Cal.4th at p. 1048, 56 Cal.Rptr.3d 814, 155 P.3d 226 ; see also In re Lance W. (1985) 37 Cal.3d 873, 890, fn. 11, 210 Cal.Rptr. 631, 694 P.2d 744.)

Even if we assume for the sake of argument that these presumptions apply here, they do not alter our conclusion that the electorate did not intend to apply Proposition 47's definition of "unreasonable risk of danger to public safety" to inmates seeking resentencing under the Three Strikes Reform Act, because the presumptions have no bearing on our determination that the statutory language is ambiguous and that this ambiguity should be resolved by construing the definition as inapplicable to those inmates. In any event, we have found compelling reasons not to apply such presumptions in prior cases, and, to an even greater degree, they are also inapt in the present matter.

a) The Presumption that Voters Have Considered the Text of the Laws Proposed by an Initiative

In Taxpayers to Limit Campaign Spending v. Fair Pol. Practices Comm. (1990) 51 Cal.3d 744, 274 Cal.Rptr. 787, 799 P.2d 1220 ( Taxpayers ), we declined to credit the presumption that voters had thoroughly studied the probable impact of proposed initiatives. ( Id. at pp. 769-770, 274 Cal.Rptr. 787, 799 P.2d 1220.) There, we considered the voters' simultaneous approval of two overlapping ballot measures regulating campaign contributions and spending. We examined the reach of California Constitution article II, section 10, subdivision (b), which provides: "If provisions of 2 or more measures approved at the same election conflict, those of the measure receiving the highest affirmative vote shall prevail." In determining the voters' intent, the court had to decide whether this constitutional provision required the measure with the highest votes to prevail entirely or whether the rules of statutory construction required the individual provisions of each measure to be grafted into a hybrid form and put into effect, despite any conflicting provisions. ( Taxpayers , supra , 51 Cal.3d at p. 747, 274 Cal.Rptr. 787, 799 P.2d 1220.) We observed that both a state commission and the Court of Appeal had attempted to reconcile the measures using the latter approach, but noted that each had reached "divergent conclusions" concerning whether various provisions from each measure were irreconcilable and whether the voters would have adopted various parts of the measure that had received fewer votes if they had "not been part of that measure." ( Id . at p. 760, 274 Cal.Rptr. 787, 799 P.2d 1220.)

In rejecting the contention that the voters intended a hybrid version of the measures if they were both approved, this court acknowledged the "presumption that the voters thoroughly study and understand the content of complex initiative measures." ( Taxpayers , supra , 51 Cal.3d at p. 768, 274 Cal.Rptr. 787, 799 P.2d 1220.) But we refused to assume further "that voters not only recognized that they were approving initiatives with fundamentally conflicting provisions intended to regulate the same subject, but also analyzed the remaining provisions in order to predict which would be implemented if either measure received a lesser affirmative vote." ( Ibid . ) This court remarked "that even the most conscientious voters may lack the time to study ballot measures with that degree of thoroughness." ( Id . at p. 770, 274 Cal.Rptr. 787, 799 P.2d 1220.) " ' "We think the assertion may safely be ventured that it is only the few persons who earnestly favor or zealously oppose the passage of a proposed law initiated by petition who have attentively studied its contents and know how it will probably affect their private interests." ' " ( Ibid ., quoting Wallace v. Zinman (1927) 200 Cal. 585, 592-593, 254 P. 946, quoting in turn State v. Richardson (1906) 48 Or. 309, 319, 85 P. 225.)The present case is similar. In defining the phrase "unreasonable risk of danger to public safety," it is peculiar that section 1170.18, subdivision (c), extended that definition to the Three Strikes Reform Act by use of the phrase "[a]s used throughout this Code." When Proposition 47 was presented to the voters, the phrase "unreasonable risk of danger to public safety" appeared in only one other section of the Penal Code. (§ 1170.126, subd. (f).) Given this unique circumstance, it is highly unusual that the drafters of Proposition 47 did not simply make a direct reference to the single relevant provision intended to be amended— section 1170.126, the resentencing statute previously enacted by the Three Strikes Reform Act. Such a direct reference would have expressly disclosed to the voters a connection between Proposition 47 and the Three Strikes Reform Act. Instead, Proposition 47 applied its definition of "unreasonable risk of danger to public safety" to the entire Penal Code, which has thousands of sections encompassing hundreds of thousands of provisions, only one of which was relevant to the proposed new definition.

Consequently, without an express reference to section 1170.126, part of the Three Strikes Reform Act, the average voter would not have known the impact or import of the phrase "[a]s used throughout this Code" in section 1170.18, subdivision (c), unless they had exhaustively sifted through the voluminous Penal Code in order to find the single other reference to the phrase "unreasonable risk of danger to public safety." We similarly recognized in Taxpayers that it is unreasonable to presume that the voters had such a "degree of thoroughness" that they " ' "attentively studied" ' " the measure and analyzed various provisions using the acumen of a legal professional. ( Taxpayers , supra , 51 Cal.3d at p. 770, 274 Cal.Rptr. 787, 799 P.2d 1220.)

Indeed, in the present matter, even professional bodies charged with the duty of enforcing the law and assessing the effects of proposed legislation did not identify the need to refer to the Three Strikes Reform Act's section 1170.126. First, the Attorney General's Official Title and Summary fails to note or describe any effect Proposition 47 might have in facilitating the release of serious or violent recidivist felons. In general, the Official Title and Summary prepared by the Attorney General plays an important role in preempting voter confusion and manipulation in the initiative process. If the Attorney General had omitted a key provision from Proposition 47's title and summary, the initiative's drafters should have brought this to her attention during the measure's public review period. (See Elec. Code, §§ 9002, subd. (a), 9004, subd. (a).) Their failure to do so suggests no such change was contemplated.¹¹ Second, the Legislative Analyst's Analysis of Proposition 47, included alongside the Attorney General's Official Title and Summary in the Voter Information Guide, also make no mention of the Three Strikes Law, the Three Strikes Reform Act, three strike inmates, or the life sentences for serious or violent recidivist felons.

11 After the review period, the Attorney General issued the following Official Title and Summary for Proposition 47, which ultimately appeared on the November 2014 ballot:

"Criminal Sentences. Misdemeanor Penalties. Initiative Statute.

• Requires misdemeanor sentence instead of felony for certain drug possession offenses.

• Requires misdemeanor sentence instead of felony for the following crimes when amount involved is $950 or less: petty theft, receiving stolen property, and forging/writing bad checks.

• Allows felony sentence for these offenses if person has previous conviction for crimes such as rape, murder, or child molestation or is registered sex offender.

• Requires resentencing for persons serving felony sentences for these offenses unless court finds unreasonable public safety risk.

• Applies savings to mental health and drug treatment programs, K-12 schools, and crime victims.

Summary of Legislative Analyst's Estimate of Net State and Local Government Fiscal Impact:

• Net state criminal justice system savings that could reach the low hundreds of millions of dollars annually. These savings would be spent on school truancy and dropout prevention, mental health and substance abuse treatment, and victim services.

• Net county criminal justice system savings that could reach several hundred million dollars annually."

Defendants and our dissenting colleagues find these circumstances unremarkable and engage in the fiction that we should still apply the presumption that the voters thoroughly study and understand the content of complex initiative measures, even though the implications for three strikes resentencing were apparently opaque to the Attorney General and the Legislative Analyst. They were almost certainly opaque to the average voter as well. It is not reasonable to apply a presumption of voter awareness when the text of the initiative and the voter information guide supporting it make no reference whatsoever to any effect on a different law.¹² b) The Presumption that Voters Are Aware of Existing Laws

12 Concurring with the result in Taxpayers, Justice Mosk warned: "It is a sad fact of democratic life that many voters have neither the time nor the will to scrutinize a measure's contents in the detail needed to discern a conflict not squarely presented in the ballot arguments or title" and that the drafter of an initiative "could take advantage of this reality to offer the voters a competing initiative that artfully conceals its true aim by failing to declare its opposition to the target initiative in materials the voter can readily comprehend." (Taxpayers, supra, 51 Cal.3d at p. 772, 274 Cal.Rptr. 787, 799 P.2d 1220 (conc. & dis. opn. of Mosk, J.).)

In Robert L. , supra , 30 Cal.4th 894, 135 Cal.Rptr.2d 30, 69 P.3d 951, we refused to presume that voters were aware of the legal meaning of the term "wobbler." (See ante , 220 Cal.Rptr.3d at p. 234, fn. 2, 397 P.3d at p. 940, fn. 2.) There, we interpreted the meaning of a gang enhancement adopted by the 2000 electorate in Proposition 21, finding no evidence presented to the voters of the interpretation put forth by the defendant. The enhancement provides that "[a]ny person who is convicted of a public offense punishable as a felony or a misdemeanor," committed for the benefit of a criminal street gang, is to be punished by imprisonment in the county jail, or by imprisonment in the state prison for one, two, or three years. (§ 186.22, subd. (d).) We rejected the defendant's contention that the phrase " 'punishable as a felony or a misdemeanor' " in Proposition 21 meant that the enhancement applied only to so-called wobbler offenses. ( Robert L. , supra , 30 Cal.4th at p. 902, 135 Cal.Rptr.2d 30, 69 P.3d 951.)

In reaching this conclusion, we first analyzed the statute's plain meaning. We acknowledged that wobblers are the only "offenses punishable as a misdemeanor or felony," but we refused to apply the standard presumption that voters are aware of existing law when approving an initiative. ( Robert L. , supra , 30 Cal.4th at p. 901, 135 Cal.Rptr.2d 30, 69 P.3d 951, italics omitted.) Although the term "wobbler" is commonly used by "attorneys, judges, and law enforcement personnel who are familiar with criminal law," we observed that the word "does not have a meaning defined by statute or commonly understood by the electorate." ( Id . at p. 902, 135 Cal.Rptr.2d 30, 69 P.3d 951.) We refused to attribute this level of knowledge to the electorate such that voters would have recognized the phrase "punishable as a felony or a misdemeanor" as restricting the enhancement to "wobbler" offenses. "We are confident that the average voter, unschooled in the patois of criminal law, would have [instead] understood the plain language of section 186.22(d) to encompass all misdemeanors and all felonies."¹³ ( Robert L. , at p. 902, 135 Cal.Rptr.2d 30, 69 P.3d 951.)

13 In his dissenting opinion, Justice Cuéllar contends that our holding in Robert L. did not disregard the presumption that voters are presumed to be aware of existing laws. (Dis. opn. of Cuéllar, J., post, 220 Cal.Rptr.3d at p. 291, 397 P.3d at p. 987.) But in Robert L., we rejected the petitioner's argument that that "section 186.22(d) 'perfectly describes wobblers,' because '[w]obblers are the only public offenses punishable as a misdemeanor or felony,' and the voters, who are presumed to know this fact, therefore intended section 186.22(d) to be limited to wobblers." (Robert L., supra, 30 Cal.4th at p. 901, 135 Cal.Rptr.2d 30, 69 P.3d 951, original italics omitted, new italics added.) In rejecting this contention, we refused to apply such a presumption, recognizing that "the average voter, unschooled in the patois of criminal law, would have understood" a different meaning. (Id. at p. 902, 135 Cal.Rptr.2d 30, 69 P.3d 951.) Thus, in Robert L., we found it unreasonable to presume that a lay voter would understand or give credit to a term of legal art.

Similarly, in this matter, without further guidance from Proposition 47's drafters or its Voter Information Guide materials, it is unreasonable to conclude that the measure's reference to an "unreasonable risk of danger to public safety" would have triggered awareness on the part of voters that this was precisely the same language applied to govern the resentencing of three strike inmates.

7. Conclusion

The parties agree that the application of Proposition 47's definition of an "unreasonable risk of danger to public safety" to the resentencing proceedings of three strike inmates would ease the burden for recidivist serious or violent offenders to have their life sentences vacated, and render them more likely to be released. If Proposition 47 had truly intended to amend the Three Strikes Reform Act in order to allow additional three strike inmates, who by definition, have records for multiple serious or violent felonies to be resentenced, one would expect its drafters to have mentioned or referred to such a purpose and intention in the measure's preamble. They did not.

Moreover, such an amendment necessarily would suggest disfavor with the broad discretion that Proposition 36, two years earlier, had given resentencing courts to determine which offenders are too dangerous to the public to be eligible for resentencing. If Proposition 47 actually intended to curtail this discretion, just two years after Proposition 36 went into effect, and confine trial courts to consider only whether the resentencing of a recidivist serious or violent offender posed a danger of committing one of eight categories of super strike offenses, one would expect to see some mention of the Three Strikes Law, the Three Strikes Reform Act, three strike inmates, life sentences, or why the resentencing discretion required reform in the text of the initiative. But there is none.

Instead, Proposition 47 explicitly assured voters that the sentences of persons convicted of dangerous crimes and various sex crimes would not change.

In describing to voters Proposition 47's title and summary, the Attorney General failed to note or identify any effect the measure might have in facilitating the release of serious or violent recidivist felons. In describing to voters Proposition 47's effect on public safety, the criminal justice system, and fiscal policy, the Legislative Analyst also failed to note or identify any effect the measure might have in facilitating the release of serious or violent recidivist felons.

Our dissenting colleagues argue that our decision here subverts the will of the voters and impermissibly interferes with the sanctity of our state's initiative process. On the contrary, adopting the construction advocated by defendants and the dissents as to the scope of a phrase in a measure without notice to the voters, not mentioned by the Attorney General or Legislative Analyst, and contrary to the stated purposes and assurances described in the measure's own preamble, would not protect the voters' right to directly enact laws but could very likely encourage the subversion and manipulation of that democratic right.

"While we are of the opinion that statutes dealing with the initiative should be liberally construed to permit the exercise by the electors of this most important privilege, we are also of the opinion that statutes passed for the purpose of protecting electors from confusing or misleading situations should be enforced." ( Clark v. Jordan (1936) 7 Cal.2d 248, 252, 60 P.2d 457.) The statutory provisions imposed on the Attorney General and Legislative Analyst in order to educate voters about the effect of proposed initiatives and to protect them from being misled or confused, would have no meaning and would not serve their intended purpose of protecting the democratic process if we adopted the defendants' and the dissents' interpretation of Proposition 47. In that scenario we would be imputing to voters an intent that could not reasonably have existed, where even the Attorney General and Legislative Analyst, in advising voters, apparently were unaware of that professed intent.

We cannot infer the realization of a voter intent where there was nothing to enlighten it in the first instance. As we have aptly recognized, and reiterated in Robert L. , " '[i]n the case of a voters' initiative statute ... we may not properly interpret the measure in a way that the electorate did not contemplate: the voters should get what they enacted, not more and not less.' "¹⁴ ( Robert L. , supra , 30 Cal.4th at p. 909, 135 Cal.Rptr.2d 30, 69 P.3d 951, quoting Hodges v. Superior Court (1999) 21 Cal.4th 109, 114, 86 Cal.Rptr.2d 884, 980 P.2d 433.)For the reasons described above, we decline to credit defendants' proposed interpretation of section 1170.18, subdivision (c) as enacted by Proposition 47, the Safe Neighborhoods and Schools Act. Given the circumstances, it is reasonable to interpret section 1170.18, subdivision (c)'s definition of an "unreasonable risk of danger to public safety" as applicable only to the resentencing proceedings that are authorized under Proposition 47.

14 In addition to the complaint that we are not making a finding of drafting error, to varying degrees of forcefulness, the dissents also fault us for purportedly failing to contend that we are not adopting their interpretation on the basis that it would generate absurd results or that their interpretation would frustrate the manifest purposes of Proposition 47, regardless of any ambiguity in its language. But there is no value in raising and then rejecting arguments that the majority does not rely upon to reach its ultimate conclusion. (Branch v. Smith (2003) 538 U.S. 254, 280-281, 123 S.Ct. 1429, 155 L.Ed.2d 407 [addressing the dissent's various irrelevant arguments in favor of a different interpretation of redistricting statutes].) The arguments distract from the question of whether Proposition 47, as a whole, is ambiguous as to any intent to facilitate the release of recidivist serious and violent felons in light of the measure's express promise that dangerous criminals would not benefit from the act. Nor do the dissents adequately address the omission by the entities entrusted to ensure that voters are fully informed and not misled by the provisions of an initiative, the Attorney General and the Legislative Analyst.

C. Equal Protection Does Not Compel the Application of Proposition 47's Definition of "Unreasonable Risk of Danger to Public Safety" to the Three Strikes Reform Act

We further conclude that refusing to construe Proposition 47's definition as applicable to the previously enacted Three Strikes Reform Act does not violate equal protection or due process; nor does our determination raise any question concerning cruel and unusual punishment under the federal Constitution.

"The concept of equal protection recognizes that persons who are similarly situated with respect to a law's legitimate purposes must be treated equally. [Citation.] Accordingly, ' "[t]he first prerequisite to a meritorious claim under the equal protection clause is a showing that the state has adopted a classification that affects two or more similarly situated groups in an unequal manner." ' [Citation.] 'This initial inquiry is not whether persons are similarly situated for all purposes, but "whether they are similarly situated for purposes of the law challenged." ' [Citation.]" ( People v. Brown (2012) 54 Cal.4th 314, 328, 142 Cal.Rptr.3d 824, 278 P.3d 1182, italics omitted.)

It is not difficult to conclude that those resentenced under Proposition 36 are not similarly situated to those resentenced under Proposition 47. These are two very different populations of offenders. As the text of Proposition 47 indicates, that measure focused on offenders convicted of a set of low-level, nonserious, nonviolent felonies and reduced them to misdemeanors. In contrast, Proposition 36 concerned the resentencing of recidivist offenders who had two prior violent or serious felony convictions and a third nonserious, nonviolent felony conviction, and who are serving terms of 25 years to life. Obviously, those recidivists with criminal records involving serious or violent felonies, who may have already been incarcerated for a lengthy period, raise concerns for public safety different from those who committed what were previously low-level felonies. Thus, the two groups are not similarly situated for purposes of resentencing because one group consists of recidivist serious or violent offenders, who may have the propensity to commit serious or violent felony crimes and the other generally consists of low-level offenders.¹⁵ For these same reasons, the defendants' Eighth Amendment challenge fails as well because there is a valid reason to apply less lenient resentencing criteria to third strike offenders, as opposed to low-level offenders whose crimes are now reclassified as misdemeanors.

15 Accordingly, we also reject application of the rule of statutory construction that when statutes are in pari materia similar phrases appearing in each should be given like meanings because Proposition 47 and the Three Strikes Reform Act do not address the same subject. (California Fed. Savings & Loan Assn. v. City of Los Angeles (1995) 11 Cal.4th 342, 352, 45 Cal.Rptr.2d 279, 902 P.2d 297.) In addition, we observe that this doctrine might carry greater force concerning provisions enacted by the same initiative or initiatives on the same ballot. But here, the electorate that enacted Proposition 47 in 2014 was obviously different from the one that had endorsed Proposition 36 in 2012. More important, because the voters in 2014 were not advised of any overlap between the two measures when they considered Proposition 47, reliance on the appearance of the phrase "unreasonable risk of public safety" as proof that voters intended the same meaning is not justified, especially where the later measure, at best, tacitly attempted to amend the same language used in the former measure.

Because we hold that Proposition 47's definition of "unreasonable risk of danger to public safety" does not apply to resentencing proceedings under the Three Strikes Reform Act, we need not address whether, in People v. Chaney (S223676), that definition applies retroactively to petitions that a court has already denied but are not yet final on appeal.

III. DISPOSITION

For the foregoing reasons, we affirm the judgments of the Courts of Appeal in both People v. Valencia (S223825) and People v. Chaney (S223676).

We Concur:

Chin, J.

Corrigan, J.

Kruger, J.

CONCURRING OPINION BY KRUGER, J.

Kruger, J.

A provision of the Safe Neighborhoods and Schools Act of 2014 (better known as Proposition 47) permits inmates serving felony sentences for certain drug- and theft-related offenses to petition to have their sentences recalled and to be resentenced as misdemeanants. ( Pen. Code, § 1170.18, subd. (a).) If the inmate otherwise satisfies the statutory criteria for resentencing, the court must grant the petition unless it "determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (Id. , subd. (b).) The provision goes on to specify that, "[a]s used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit" one of an enumerated list of violent felonies, colloquially known as super strikes. (Id. , subd. (c) ( section 1170.18(c) ); see Pen. Code, § 667, subd. (e)(2)(C)(iv).)

The question before us is whether this definition of "unreasonable risk of danger to public safety" applies only to petitioners who seek resentencing under Proposition 47, or whether it also applies to individuals who have petitioned for relief under a different provision of the Penal Code—specifically, life prisoners seeking resentencing under the earlier enacted Three Strikes Reform Act of 2012 (better known as Proposition 36), which contains a parallel recall procedure. I agree with the majority that the text of this definitional provision contains conflicting indications about its intended scope. I also agree that standard interpretive aids resolve the ambiguity in favor of the narrower interpretation. Every other provision of Proposition 47 reflects a clear and exclusive focus on affording relief to individuals who have committed specified drug- and theft-related offenses, and neither the stated purposes of the proposition nor the ballot materials alerted voters to any possibility that a favorable vote might also result in a significant change to the separate statutory scheme governing the resentencing of life prisoners under the "Three Strikes" law. Although this is certainly a choice the voters could make, I do not think we can say it is a choice the voters have already made.

I.

At issue in this case are two related statutory schemes: Proposition 36, which amended the Three Strikes law, and Proposition 47. As first enacted in 1994, the Three Strikes law provided that the prison sentence of a defendant convicted of a felony who had a single prior conviction for a violent or serious felony (a strike) was doubled ( Pen. Code, § 1170.12, former subd. (c)(1)), and a defendant convicted of a felony (no matter how minor) who had two or more prior strikes would be sentenced to a minimum of 25 years to life in prison (id. , former subd. (c)(2)).¹ Nearly 20 years later, partly in response to the swelling prison population and rising incarceration costs, voters passed Proposition 36. It provides that a defendant with two prior strikes who commits a felony that is neither violent nor serious is no longer subject to a 25-years-to-life sentence; instead, the sentence is merely doubled, as is the case for a defendant with only one prior strike. ( Pen. Code, §§ 667, subd. (e), 1170.12, subd. (c).)

1 All statutory references are to the Penal Code unless otherwise noted.

Proposition 36 also permits inmates who, before its enactment, had received a life sentence under the Three Strikes law as the result of a conviction of a nonserious and nonviolent felony with two or more prior strikes to file a petition asking the trial court to instead impose a second-strike sentence. ( Pen. Code, § 1170.126.) But not all such inmates may apply. Those whose third strike was a sex offense, involved large amounts of drugs, involved a firearm or deadly weapon, or was committed with the intent to inflict great bodily injury may not seek resentencing. ( Pen. Code, § 1170.12, subd. (c)(2)(C).) Also barred from applying are inmates with prior convictions for certain extremely serious crimes that have come to be referred to as super strikes: violent sex crimes, child molestation, homicide, solicitation of murder, assaulting a police officer or firefighter with a machine gun, possession of a weapon of mass destruction, or any crime punishable by life imprisonment or death. (Id ., subd. (c)(2)(C)(iv).) Finally, the trial court may deny a petitioner's request for resentencing when "the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety " ( Pen. Code, § 1170.126, subd. (f), italics added), based on the petitioner's criminal record, disciplinary record in prison, and other relevant evidence. Proposition 36 did not define the phrase "unreasonable risk of danger to public safety."

In November 2014, two years after the enactment of Proposition 36, the voters approved Proposition 47. Proposition 47 reclassifies certain drug offenses and theft-related crimes, which previously were either felonies or wobblers, as misdemeanors. Much like Proposition 36, Proposition 47 creates a recall procedure that permits previously sentenced individuals to seek the benefit of its ameliorative changes to the law. Such a petition may be brought by a person whose felony conviction was based on conduct that the initiative reclassified as a misdemeanor ( Pen. Code, § 1170.18, subd. (a) ), but resentencing may be denied when "the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety" (id ., subd. (b)), based on the petitioner's criminal record, disciplinary record in prison, and other relevant evidence. But unlike Proposition 36, Proposition 47 defines "unreasonable risk of danger to public safety": "As used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit a new violent felony within the meaning of clause (iv) of subparagraph (C) of paragraph (2) of subdivision (e) of section 667." (Id ., subd. (c).) The specified clause lists the super strikes. In other words, section 1170.18(c) states that a court may find a petitioner to be dangerous to public safety only if it finds that there is an unreasonable risk that the petitioner will commit a super strike offense.

II.

The question before us is one of statutory interpretation: whether Proposition 47's definition of "unreasonable risk of danger to public safety," set forth in section 1170.18(c), applies to petitioners who have received life sentences under the Three Strikes law and are seeking resentencing under Proposition 36.

California cases have established a set of standard rules for the construction of voter initiatives. "We interpret voter initiatives using the same principles that govern construction of legislative enactments. [Citation.] Thus, we begin with the text as the first and best indicator of intent. [Citations.] If the text is ambiguous and supports multiple interpretations, we may then turn to extrinsic sources such as ballot summaries and arguments for insight into the voters' intent." ( People v. Mentch (2008) 45 Cal.4th 274, 282, 85 Cal.Rptr.3d 480, 195 P.3d 1061.) On the other hand, " ' "[i]f there is no ambiguity, then we presume the lawmakers meant what they said," ' " and the plain meaning of the language ordinarily will govern. ( Mays v. City of Los Angeles (2008) 43 Cal.4th 313, 321, 74 Cal.Rptr.3d 891, 180 P.3d 935.) In interpreting a voter initiative, we "give effect to the voters' formally expressed intent, without speculating about how they might have felt concerning subjects on which they were not asked to vote." ( Ross v. RagingWire Telecom., Inc. (2008) 42 Cal.4th 920, 930, 70 Cal.Rptr.3d 382, 174 P.3d 200 ( Ross ).)

Proposition 47's resentencing provision states, as relevant here: "As used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit a new violent felony within the meaning of clause (iv) of subparagraph (C) of paragraph (2) of subdivision (e) of section 667." ( § 1170.18(c).) The central interpretive task before us is to identify the "petitioner[s]" to whom this provision refers. Does it refer to petitioners seeking resentencing under the provisions of Proposition 47? Or does it refer to any petitioner in any proceeding provided for in the Penal Code—in which case the definition also applies to life prisoners seeking resentencing under Proposition 36?

The text of section 1170.18(c) alone does not provide a clear answer. On the one hand, the breadth of the introductory clause of section 1170.18(c) ( "As used throughout this Code") suggests that the "unreasonable risk" definition is designed to apply broadly, for purposes of evaluating the dangerousness of any "petitioner" in any petition proceeding under any provision of the Penal Code. And not coincidentally, the phrase "unreasonable risk of danger to public safety" appears only in one other place in the Penal Code: the earlier-enacted, parallel resentencing provision of Proposition 36 ( § 1170.126, subd. (f) ), which contains an essentially identical petition procedure.

But on the other hand, reading section 1170.18(c) in the context of surrounding provisions, the focus of the provision appears much narrower. Subdivision (a) of Penal Code section 1170.18 explains that "[a] person ... serving a sentence for a conviction ... of a felony or felonies who would have been guilty of a misdemeanor under the act that added this section ... may petition for a recall of sentence...." The next subdivision explains that "[u]pon receiving a petition under subdivision (a), the court shall determine whether the petitioner satisfies the criteria in subdivision (a). If the petitioner satisfies the criteria in subdivision (a), the petitioner's felony sentence shall be recalled and the petitioner resentenced to a misdemeanor ... unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety. In exercising its discretion, the court may consider all of the following: [¶] (1) The petitioner's criminal conviction history.... [¶] (2) The petitioner's disciplinary record and record of rehabilitation.... [¶] (3) Any other evidence the court ... determines to be relevant...." ( Pen. Code, § 1170.18, subd. (b), italics added.)In total, the phrase "the petitioner" appears seven times in subdivision (b) of Penal Code section 1170.18. Each time, the phrase is unmistakably used to refer to a person who has filed "a petition under subdivision (a)"—that is, a person who "would have been guilty of a misdemeanor" under Proposition 47 and has asked to be resentenced accordingly, barring a finding that his or her resentencing "would pose an unreasonable risk of danger to public safety." The natural inference is that when the very next subdivision of the statute supplies a definition of "unreasonable risk" that refers to the "risk that the petitioner will commit a new [super strike]" ( § 1170.18(c), italics added), the definition uses the phrase in the same way—that is, to refer to a person who has filed a petition under subdivision (a) to reduce a conviction for a minor felony or wobbler to a misdemeanor, and not to other persons who have filed other petitions in other proceedings governed by other provisions of the Penal Code. (See, e.g., People v. Blackburn (2015) 61 Cal.4th 1113, 1125, 191 Cal.Rptr.3d 458, 354 P.3d 268 [" 'it is generally presumed that when a word is used in a particular sense in one part of a statute, it is intended to have the same meaning if it appears in another part of the same statute' "]; Wilcox v. Birtwhistle (1999) 21 Cal.4th 973, 979, 90 Cal.Rptr.2d 260, 987 P.2d 727 ["words or phrases given a particular meaning in one part of a statute must be given the same meaning in other parts of the statute"]; see also Department of Revenue of Ore. v. ACF Industries, Inc. (1994) 510 U.S. 332, 342, 114 S.Ct. 843, 127 L.Ed.2d 165 [" ' " 'identical words used in different parts of the same act are intended to have the same meaning' " ' "].)

The remainder of the statute reinforces the inference. The phrase "the petitioner" appears in just two other places in the statute, and each time, it again clearly refers to persons who have filed a petition authorized by subdivision (a). (See Pen. Code, § 1170.18, subd. (l ) ["If the court that originally sentenced the petitioner is not available, the presiding judge shall designate another judge to rule on the petition or application."]; id. , subd. (m) ["Nothing in this section is intended to diminish or abrogate any rights or remedies otherwise available to the petitioner or applicant."].)

Defendants argue that if the "unreasonable risk" definition in section 1170.18(c) does not apply to Proposition 36 petitioners, then the introductory language providing that the definition is to apply "throughout this Code" is superfluous. Because no other provision of the Penal Code applies to Proposition 47 petitioners, section 1170.18(c) would operate no differently had the drafters of Proposition 47 omitted the introductory clause entirely. But while we do generally strive to construe enactments to avoid rendering any word or provision surplusage ( City of Alhambra v. County of Los Angeles (2012) 55 Cal.4th 707, 724, 149 Cal.Rptr.3d 247, 288 P.3d 431 ), we have also made clear that, like all such interpretive canons, the canon against surplusage is a guide to statutory interpretation and is not invariably controlling. ( People v. Cruz (1996) 13 Cal.4th 764, 782, 55 Cal.Rptr.2d 117, 919 P.2d 731 ; see also In re J.W. (2002) 29 Cal.4th 200, 209, 126 Cal.Rptr.2d 897, 57 P.3d 363 ; accord, e.g., Arlington Central School Dist. Bd. of Ed. v. Murphy (2006) 548 U.S. 291, 299, fn. 1, 126 S.Ct. 2455, 165 L.Ed.2d 526 ["While it is generally presumed that statutes do not contain surplusage, instances of surplusage are not unknown."]; Lamie v. United States Trustee (2004) 540 U.S. 526, 536, 124 S.Ct. 1023, 157 L.Ed.2d 1024 ["[O]ur preference for avoiding surplusage constructions is not absolute."].)^{2 2}

2 Justice Liu's dissenting opinion suggests that a reviewing court may construe statutory language as surplusage only when the resulting interpretation renders that language redundant. (Dis. opn., of Liu, J., 220 Cal.Rptr.3d at p. 266, 397 P.3d at p. 966.) There is, however, no such rigid rule; other instances of surplusage, too, are not unknown. (See, e.g., King v. Burwell (2015) ––– U.S. ––––, 135 S.Ct. 2480, 2489-2492, 192 L.Ed.2d 483 ; Chickasaw Nation v. United States (2001) 534 U.S. 84, 89-90, 122 S.Ct. 528, 151 L.Ed.2d 474 (Chickasaw Nation ); People v. Watts (2016) 2 Cal.App.5th 223, 236-237, 206 Cal.Rptr.3d 202.)
In Chickasaw Nation, for example, the United States Supreme Court held that a statutory cross-reference should be treated as surplusage, even though doing so gave no effect to the language at issue. (Chickasaw Nation, supra, 534 U.S. at p. 94, 122 S.Ct. 528 ; see also pp. 97-98, 122 S.Ct. 528 (dis. opn. of O'Connor, J.).) The court explained that it was impossible to give effect to the cross-reference without either "seriously rewriting the language of the rest of the statute" (id. at p. 89, 122 S.Ct. 528 ) or adopting a construction that, though "comprehensible," would expand the statute well beyond its apparently intended scope (id. at p. 90, 122 S.Ct. 528 ). The court further explained that while courts ordinarily endeavor to "give effect to each word 'if possible,' " that canon of construction "is sometimes offset by the canon that permits a court to reject words 'as surplusage' if 'inadvertently inserted or if repugnant to the rest of the statute....' [Citation.] And the latter canon has particular force here where the surplus words consist simply of a numerical cross-reference in a parenthetical." (Id. at p. 94, 122 S.Ct. 528 ; see also pp. 90-91, 92, 122 S.Ct. 528 [concluding that Congress had simply "fail[ed] to delete an inappropriate cross reference" that "might easily escape notice" in the legislative process.] ) This case raises similar considerations, though it does not concern the meaning of a specific cross-reference to another statute; it instead concerns the meaning of a general reference to the Penal Code, which, it is argued, should be understood as an oblique cross-reference to Proposition 36.

2 For a definition of "super strike," see maj. opn., ante, 220 Cal.Rptr.3d at p. 235, fn. 3, 397 P.3d at p. 940, fn. 3.

In truth, either interpretation of section 1170.18(c) raises oddities. If we accept the proposition that the phrase "throughout this Code" must be given independent significance, then we also must accept that Proposition 47 contains a definitional provision that uses the phrase "the petitioner" differently from every other provision of the Act in which that phrase appears—including its immediately preceding subdivision, in which the same phrase appears no fewer than seven times, and which contains the operative "unreasonable risk" language to which the definition attaches. This, too, would be a departure from ordinary interpretive rules.

All this leads to the unavoidable conclusion that the text of section 1170.18(c)'s "unreasonable risk" definition is indeed ambiguous. It is possible that it was meant to apply to any petitioner in any proceeding governed by any provision of the Penal Code, including Proposition 36. It is also possible that it applies only to the lower-level offenders who have petitioned for relief under Penal Code section 1170.18, subdivision (a). To determine section 1170.18(c)'s intended scope, we must look beyond the provision itself.

This wider look ultimately settles the matter, because, with the arguable exception of section 1170.18(c)'s introductory phrase, every other provision of the initiative focuses exclusively on the category of offenders covered by Proposition 47's operative provisions—that is, the lower-level offenders who have committed certain drug- and theft-related crimes the law now treats as misdemeanors. No provision of Proposition 47 manifests an intent to change the law applicable to other categories of offenders, nor were the voters otherwise informed that their vote on the initiative might result in any such change.

Uncodified section 3 of Proposition 47, entitled "Purpose and Intent," identifies six statutory purposes: (1) to "[e]nsure that people convicted of murder, rape, and child molestation will not benefit from this act," (2) to create a "Safe Neighborhoods and Schools Fund," (3) to "[r]equire misdemeanors instead of felonies for nonserious, nonviolent crimes like petty theft and drug possession, unless the defendant has prior convictions for specified violent or serious crimes," (4) to "[a]uthorize consideration of resentencing for anyone who is currently serving a sentence for any of the offenses listed herein that are now misdemeanors," (5) to "[r]equire a thorough review of criminal history and risk assessment of any individuals before resentencing to ensure that they do not pose a risk to public safety," and (6) to "save significant state corrections dollars" and to "increase investments in programs that reduce crime and improve public safety." (Voter Information Guide, Gen. Elec. (Nov. 4, 2014) text of Prop. 47, at p. 70.)

The fourth and fifth items on this list describe the purposes of Proposition 47's resentencing provision: namely, to permit persons who have been convicted of specified minor offenses that the initiative designates as misdemeanors to obtain reduction of their sentences, while ensuring that none of those individuals "pose[s] a risk to public safety." (Voter Information Guide, supra , at p. 70, 221 Cal.Rptr. 359.) Neither provision contains any indication that the voters also intended the resentencing provision to make a significant alteration to the existing law governing resentencing for life prisoners under Proposition 36. On the contrary, uncodified section 3 suggests that, as far as resentencing is concerned, the only persons to whom section 1170.18(c) applies are those "currently serving a sentence for any of the offenses listed herein that are now misdemeanors." (Voter Information Guide, supra , at p. 70, 221 Cal.Rptr. 359.)

The remaining text of Proposition 47 is to similar effect. (See Horwich v. Superior Court (1999) 21 Cal.4th 272, 276, 87 Cal.Rptr.2d 222, 980 P.2d 927 [" 'we do not construe statutes in isolation, but rather read every statute "with reference to the entire scheme of law of which it is part so that the whole may be harmonized and retain effectiveness" ' "].) It describes the reclassification of certain felonies and wobblers as misdemeanors, it explains the relief that is available to persons previously convicted of those offenses as felonies, and it discusses what to do with the money saved by the reclassification. The initiative makes no reference to any other provision of law, including Proposition 36 or the Three Strikes law, or to any other kind of petition besides the recall petition authorized by Penal Code section 1170.18, subdivision (a).

As the majority opinion explains (maj. opn., ante , 220 Cal.Rptr.3d at pp. 245–248, 397 P.3d at pp. 949–951), the ballot materials accompanying Proposition 47 also focus exclusively on the lower-level offenders who are the subject of Proposition 47's operative provisions. (See Robert L. v. Superior Court (2003) 30 Cal.4th 894, 905, 135 Cal.Rptr.2d 30, 69 P.3d 951 [to determine the voters' intent in adopting ambiguous language, "we look to the materials that were before the voters"].) On the subject of resentencing, the analysis of the initiative prepared by the Legislative Analyst states that Proposition 47 "reduces penalties for certain offenders convicted of nonserious and nonviolent property and drug crimes," and "allows certain offenders who have been previously convicted of such crimes to apply for reduced sentences" (Voter Information Guide, supra , at p. 35). After describing the crimes that are reduced to misdemeanors by the initiative, the analysis states that an offender who is serving a felony sentence for one of those crimes may "apply to the court to have their felony conviction changed to a misdemeanor." (Id. at p. 36.) The analysis contains no indication that the initiative's definition of dangerousness would also apply to inmates sentenced under the Three Strikes law who are seeking reduction of their sentence under Proposition 36.

This omission is not, of course, dispositive. But neither is it irrelevant. Our cases recognize that, as a practical matter, voters often rely on the experts employed by the Attorney General and the Legislative Analyst to summarize proposed initiatives and to discuss their significant effects. Here, the use of the phrase "[a]s used throughout this Code" in section 1170.8(c) was, it appears, sufficiently oblique that neither the Attorney General nor the Legislative Analyst appeared to recognize the possibility that its purpose might be to make a significant amendment to the resentencing provisions of Proposition 36. When these trained experts were apparently unaware of the initiative's possible effect on Proposition 36, and thus did not alert the voters to the possibility of such an effect, it becomes more difficult to conclude that the voters understood that a "yes" vote would have that consequence.

The ballot arguments both for and against Proposition 47 likewise made no mention of petitioners seeking relief under Proposition 36. As pertinent to the resentencing provision, the proponents argued that the initiative "[s]tops wasting prison space on petty crimes and focuses law enforcement resources on violent and serious crime" (Voter Information Guide, supra , at p. 38), and that it "[s]tops wasting money on warehousing people in prisons for nonviolent petty crimes" (ibid .), whereas the opponents stressed that the initiative would make thousands of felons eligible for early release and that judges could block their release only in very rare cases. Neither side suggested that Proposition 47 might have any effect on other categories of offenders.

Finally, the circumstances surrounding the enactment of Proposition 47 reinforce this conclusion. Proposition 36 prescribed a two-year window, beginning on November 7, 2012, during which life prisoners could file a petition for resentencing. Proposition 47 took effect on November 5, 2014, two days before that window closed. Had the voters intended to make significant changes to the procedures governing Proposition 36 resentencing so shortly before the window closed, one might have expected to see some acknowledgment of this state of affairs, if not some effort to address how the changes might apply to petitions that had already been filed or adjudicated. Proposition 47 contains no such acknowledgment.

In sum, aside from the unelaborated reference to "this Code" in section 1170.18(c)'s definitional provision, every provision of the statute, as well as the ballot materials accompanying the initiative, focuses exclusively on the mechanisms for providing relief for the lower-level offenders covered by Proposition 47 itself. All available interpretive aids thus point in the same direction: The phrase "the petitioner" in section 1170.18(c) refers to an offender seeking to reduce convictions for specified minor felonies to misdemeanors under Proposition 47.^{3 3}

3 To read the statute otherwise would, I fear, lay the groundwork for abuse of the initiative system. I see no reason to question the intentions or good faith of the drafters of Proposition 47—neither of which is particularly relevant to the analysis in any event. But if the drafters had intended to submit to voters a provision that modified the existing resentencing scheme under Proposition 36, they chose a curious way to accomplish that goal. To give dispositive effect to an oblique reference to an entire statutory code, in the face of considerable evidence suggesting the intended scope of the statute is materially narrower, would undoubtedly encourage drafters in future cases to deploy similarly oblique references to hide the true scope of proposed legislation from the electorate.
The alternative is to recognize the practical reality that voters are sometimes asked to vote on statutory language that is not drawn with the precision one might hope for. In the initiative process, there are no hearings, no committee reports, no opportunities to propose amendments from the floor. Our job, ultimately, is to give effect to the voters' intent, "without speculating about how they might have felt concerning subjects on which they were not asked to vote." (Ross, supra, 42 Cal.4th at p. 930, 70 Cal.Rptr.3d 382, 174 P.3d 200.)

3 The court questions why I emphasize the absence of a drafting error, the inapplicability of the absurdity doctrine, and the lack of any finding that the manifest purposes of the law are frustrated by a particular interpretation of the law. (See maj. opn., ante, 220 Cal.Rptr.3d at p. 254, fn. 14, 397 P.3d at pp. 956–957, fn. 14). Yet the fact that these matters are not addressed in the court's opinion is precisely why I mention them. Although California law underscores the importance of effectuating a statute's purpose, it also limits the circumstances under which courts may simply refuse to give effect to a statute as it was enacted. It is quite surprising that today's decision avoids any meaningful discussion of cases where our court has declined to apply the literal meaning of a provision because it conflicts with statutory purposes.
--------

III.

It is entirely possible that, if the choice had been put to them, the voters might have decided that Proposition 47's definition of "unreasonable risk of danger to public safety" should apply equally to inmates serving life sentences under the Three Strikes law who are seeking resentencing under Proposition 36. Proposition 36 itself supplied no definition of the term, instead conferring more open-ended discretion on judges to evaluate a life prisoner's dangerousness. (Cf. People v. Conley (2016) 63 Cal.4th 646, 659, 203 Cal.Rptr.3d 622, 373 P.3d 435 [advising that the public safety requirement "must be applied realistically, with careful consideration of the [Three Strikes] Reform Act's purposes of mitigating excessive punishment and reducing prison overcrowding"].) There is nothing inherently improbable or absurd about a conclusion that life prisoners should be entitled to resentencing under Proposition 36 unless a court determines that there is an unreasonable risk that they will commit future super strikes. That is not the only possible policy choice the voters could make, but it is a plausible one.

But voters can make that choice only if the question is presented in the initiative on which they have been asked to vote. The question was not presented in Proposition 47, and so it is not a choice we can say the voters have already made.

Voter initiatives play a unique role in California's system of government. They are designed to "enable the people to amend the state Constitution or to enact statutes when current government officials have declined to adopt (and often have publicly opposed) the measure in question." ( Perry v. Brown (2011) 52 Cal.4th 1116, 1125, 134 Cal.Rptr.3d 499, 265 P.3d 1002.) In interpreting a voter initiative, we are bound to respect both the choices the voters have made and the limits of those choices. "For a court to construe an initiative statute to have substantial unintended consequences strengthens neither the initiative power nor the democratic process...." ( Ross , supra , 42 Cal.4th at p. 930, 70 Cal.Rptr.3d 382, 174 P.3d 200.) For these reasons, I concur.

We Concur: Chin, J.

Corrigan, J.

DISSENTING OPINION BY LIU, J.

Liu, J.

In 1994, California voters enacted the "Three Strikes" law, one of the toughest sentencing laws in the country. This measure required any person convicted of two prior serious or violent felonies to receive a sentence of 25 years to life imprisonment for a third felony conviction, even if the third felony was not serious or violent. (Pen. Code, former § 1170.12, subds. (b), (c)(2)(A) ; all undesignated statutory references are to the Penal Code.) For over two decades, we have applied the Three Strikes law in accordance with its plain meaning, and we have done so regardless of whether the text, history, or ballot materials addressed the particular application of the statute at issue. (See In re Cervera (2001) 24 Cal.4th 1073, 1078–1080, 103 Cal.Rptr.2d 762, 16 P.3d 176 ( Cervera ); People v. Deloza (1998) 18 Cal.4th 585, 594, 76 Cal.Rptr.2d 255, 957 P.2d 945 ; People v. Benson (1998) 18 Cal.4th 24, 30–35, 74 Cal.Rptr.2d 294, 954 P.2d 557 ; People v. Fuhrman (1997) 16 Cal.4th 930, 939, 67 Cal.Rptr.2d 1, 941 P.2d 1189.) Today the court departs from plain meaning in deciding whether the terms of a recent initiative broaden a statutory entitlement to resentencing for certain inmates whose third strike was neither serious nor violent. This turnabout is as unorthodox in its methodology as it is unsettling in its implications for the initiative process and the limited role of courts in interpreting statutes.

In 2012, the California electorate passed Proposition 36. This initiative reduced the punishment for nonviolent, nonserious third-strike felonies and allows eligible persons who had received third-strike sentences for such felonies to petition for resentencing. ( § 1170.126.) An eligible petitioner "shall be resentenced ... unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§ 1170.126, subd. (f).)

In 2014, California voters passed Proposition 47. This measure reduced certain low-level crimes from felonies to misdemeanors and allows eligible persons previously convicted of such crimes to seek recall of the felony sentence and resentencing to a misdemeanor. ( § 1170.18, subd. (a).) An eligible person "shall" be resentenced "unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§ 1170.18, subd. (b).) Proposition 47 then says: "As used throughout this Code, 'unreasonable risk of danger to public safety' means an unreasonable risk that the petitioner will commit" one of several particularly violent felonies known as "super strikes." ( § 1170.18, subd. (c) (hereafter section 1170.18(c) ).) Everyone agrees that "this Code" means the Penal Code and that the resentencing provision of Proposition 36 is the only other place in the Penal Code where the term "unreasonable risk of danger to public safety" is used.

The question here is whether Proposition 47's definition of "unreasonable risk of danger to public safety" applies to that term as used in Proposition 36. The plain meaning of section 1170.18(c) supplies the answer: Proposition 47's definition applies to that term "[a]s used throughout this Code" ( § 1170.18(c) ), and Proposition 36's resentencing provision is part of the Penal Code.

Instead of following this straightforward analysis, the court today nullifies the phrase "As used throughout this Code" in section 1170.18(c). This judicial override of statutory text defies basic rules of interpretation and the democratic values those rules are designed to serve. We interpret voter initiatives as we interpret all legislative enactments: " 'we begin with the text as the first and best indicator of intent.' " ( Kwikset Corp. v. Superior Court (2011) 51 Cal.4th 310, 321, 120 Cal.Rptr.3d 741, 246 P.3d 877.) " 'Absent ambiguity, we presume that the voters intend the meaning apparent on the face of an initiative measure [citation] and the court may not add to the statute or rewrite it to conform to an assumed intent that is not apparent in its language.' [Citation.] Of course, in construing the statute, '[t]he words ... must be read in context, considering the nature and purpose of the statutory enactment.' " ( People ex rel. Lungren v. Superior Court (1996) 14 Cal.4th 294, 301, 58 Cal.Rptr.2d 855, 926 P.2d 1042 ( Lungren ).) " 'A court must, where reasonably possible, harmonize statutes, reconcile seeming inconsistencies in them, and construe them to give force and effect to all of their provisions.' " ( Pacific Palisades Bowl Mobile Estates, LLC v. City of Los Angeles (2012) 55 Cal.4th 783, 805, 149 Cal.Rptr.3d 383, 288 P.3d 717.)

A statute's plain meaning need not be given effect if the text clearly reveals a drafting error, but the court does not say there was a drafting error here. A statute's plain meaning need not be given effect if doing so would produce absurd results, but the court does not claim any absurdity here. The plain meaning rule does not apply if the text is in fact ambiguous, but the court identifies no ambiguity in the phrase "As used throughout this Code." Indeed, what two ways are there to read those words? Finally, we have said the plain meaning rule does not apply if it would contravene the manifest purpose of the statute, but the evidence of a purpose to restrict the applicability of Proposition 47's definition of "unreasonable risk of danger to public safety" is underwhelming. The court says nothing indicates the voters intended or knew that Proposition 47's definition of that term would apply to Proposition 36. Nothing, that is, except the unambiguous words they enacted , which we must presume the voters duly considered against the backdrop of existing laws.

Today's decision crosses the line from statutory interpretation to judicial legislation. When the voters enacted Proposition 47, they spoke clearly on how widely its resentencing criteria would apply. We cannot now tell inmates like David Valencia and Clifford Chaney, whose third strike was neither serious nor violent, that what Proposition 47 plainly says is not what the voters really meant. I respectfully dissent.

I.

At the outset, it is essential to note that the phrase "As used throughout this Code" in section 1170.18(c) has a plain meaning, and no one suggests otherwise. No member of the court contends that this phrase is itself ambiguous . Yet the court contends that the entirety of section 1170.18(c) is ambiguous when considered in context. This seriously distorts the concept of ambiguity.

The court relies on cases "where statutory language is ambiguous when considered 'in the context of the statute and initiative as a whole.' " (Maj. opn., ante , 220 Cal.Rptr.3d at p. 240, 397 P.3d at p. 944.) But in those cases, the consideration of context did not create ambiguity in statutory text that was otherwise unambiguous. Instead, the consideration of context showed that the text itself was susceptible to multiple meanings and therefore ambiguous. (See Kennedy Wholesale, Inc. v. State Bd. of Equalization (1991) 53 Cal.3d 245, 249, 279 Cal.Rptr. 325, 806 P.2d 1360 [statutory text is "susceptible of the interpretation that plaintiff has proposed" (italics added) ]; People v. Hazelton (1996) 14 Cal.4th 101, 105–106, 58 Cal.Rptr.2d 443, 926 P.2d 423 [the statutory term " 'prior felony convictions ...' could be interpreted ... to refer to the forum in which the prior conviction was obtained" or "the nature of the prior conviction"]; Building Industry Assn. v. City of Camarillo (1986) 41 Cal.3d 810, 819–820, 226 Cal.Rptr. 81, 718 P.2d 68 ["the Legislature must have assumed that measures enacted by a 'governing body' included initiatives adopted by the electorate"].) No one could disagree with the principle that statutory language must be construed in the context of the statute as a whole, and there are many cases where context sheds light on alternative ways of interpreting language. But those cases provide no support for what the court has done here.

The court contends that section 1170.18(c) is ambiguous because its use of the term "the petitioner" may be understood to mean only a Proposition 47 petitioner, which is what the term means in subdivisions (b), (l), and (m) of section 1170.18. This interpretation of "the petitioner" is not unreasonable. But it does not establish that there are two ways of reading the entirety of section 1170.18(c). Instead, reading "the petitioner" to mean what the court says it means requires us to wholly disregard the introductory phrase "As used throughout this Code."

Although the concurring opinion suggests that the phrase may be read as tolerable "surplusage" (conc. opn., ante , 220Cal.Rptr.3d at pp. 259–260, 397 P.3d at p. 961), all the instances of tolerable surplusage cited in the main text of the concurring opinion involve language that was construed as redundant. (See People v. Cruz (1996) 13 Cal.4th 764, 782–783, 55 Cal.Rptr.2d 117, 919 P.2d 731 ; In re J.W. (2002) 29 Cal.4th 200, 209–210, 126 Cal.Rptr.2d 897, 57 P.3d 363 ; Arlington Central School Dist. Bd. of Ed. v. Murphy (2006) 548 U.S. 291, 299, fn. 1, 126 S.Ct. 2455, 165 L.Ed.2d 526 ; Lamie v. United States Trustee (2004) 540 U.S. 526, 536, 124 S.Ct. 1023, 157 L.Ed.2d 1024.) It is one thing to tolerate redundancy in interpreting statutory text; it is quite another to give no independent significance to operative language that cannot be construed as redundant of anything else in the statute. In a case of redundancy, the text is still given a meaning; here the court gives the phrase "As used throughout this Code" no meaning at all.

In a footnote, the concurring opinion says Chickasaw Nation v. United States (2001) 534 U.S. 84, 94, 122 S.Ct. 528, 151 L.Ed.2d 474, held that "a statutory cross-reference should be treated as surplusage, even though doing so gave no effect to the language at issue." (Conc. opn., ante , 220 Cal.Rptr.3d at p. 260, fn. 2, 397 P.3d at p. 961, fn. 2.) But Chickasaw Nation said the statutory cross-reference was language that "Congress included inadvertently"—in other words, "a drafting mistake." ( Chickasaw Nation , at pp. 90, 91, 122 S.Ct. 528.) No member of the court contends that the phrase "As used throughout this Code" in section 1170.18(c) was a drafting error. The concurring opinion's reliance on King v. Burwell (2015) 576 U.S. ––––, –––– – ––––, 135 S.Ct. 2480, 2489–2492, 192 L.Ed.2d 483, is also unavailing. The high court there said it was deciding between two plausible interpretations of a statutory provision, not between one interpretation that gives effect to all the words of the provision and another that does not. ( Id. at p. ––––, 135 S.Ct. at p. 2491 ].) And in People v. Watts (2016) 2 Cal.App.5th 223, 206 Cal.Rptr.3d 202, the court declined to apply the rule against surplusage because doing so would not " 'result[ ] in a reasonable reading of the legislation.' " ( Id. at p. 237, 206 Cal.Rptr.3d 202.) No member of the court contends that giving effect to the plain meaning of "As used throughout this Code" in section 1170.18(c) would be unreasonable. In short, the cases cited by the concurring opinion serve only to underscore how novel the court's analysis is here.

There is indeed nothing unreasonable about reading "the petitioner" in section 1170.18(c) to include any petitioner who is subject to evaluation for "unreasonable risk of danger to public safety" "throughout this Code," including a petitioner who seeks resentencing under Proposition 36. Both Proposition 47 and Proposition 36 refer to "the petitioner" in setting forth similar petitioning schemes. (Compare § 1170.18, subds. (a), (b), (l), (m) with § 1170.126, subds. (b), (f), (g).) Reading "the petitioner" in section 1170.18(c) to include "the petitioner" under either Proposition 47 or Proposition 36 readily harmonizes the term with section 1170.18(c)'s unambiguous specification of how widely its definition of "unreasonable risk of danger to public safety" applies—i.e., "As used throughout this Code."

Thus, the issue before us presents a choice between (1) giving "the petitioner" a reasonable construction that effectuates the plain meaning of "As used throughout this Code" and (2) giving "the petitioner" a different, reasonable construction that ignores the plain meaning of "As used throughout this Code." These two options do not show that section 1170.18(c) as a whole is ambiguous. Deciding the meaning of "the petitioner" involves a choice between accepting and disregarding the plain meaning of "As used throughout this Code." The analytical framework for making that choice consists of the rules governing when to follow and when to depart from plain meaning, not the rules for resolving statutory ambiguity.

Nor is it correct to say that section 1170.18(c) is ambiguous because giving effect to the plain meaning of the phrase "As used throughout this Code" would be "inconsistent with Proposition 47's uncodified findings, declarations, purpose, and intent." (Maj. opn., ante , 220 Cal.Rptr.3d at p. 245, 397 P.3d at p. 949.) The "stark inconsistency" the court alleges (ibid. ) is illusory, as explained further below. But setting that issue to one side, the contention that the plain meaning of an operative provision "conflict[s] with" the statute's purposes (id. at p. 245, 397 P.3d at p. 948) is an argument for disregarding plain meaning; it is not an argument for ambiguity. Were it otherwise, then every case involving whether to effectuate or disregard plain meaning in light of the statute's purposes could be recast as a case about ambiguity. As today's decision shows, distorting the concept of ambiguity in this way significantly, and unjustifiably, expands the power of courts to rewrite statutes.

A finding of ambiguity opens the door to consideration of extrinsic sources that may tip the balance between two reasonable readings of the text. But in order to apply a statute in a manner that disregards its plain meaning, the burden of justification is much steeper. Plain meaning may be disregarded if the text contains a clear drafting error or if the consequences would be unreasonable or absurd. (See, e.g., People v. Broussard (1993) 5 Cal.4th 1067, 1071, 22 Cal.Rptr.2d 278, 856 P.2d 1134 ; People v. Skinner (1985) 39 Cal.3d 765, 775, 217 Cal.Rptr. 685, 704 P.2d 752.) As noted, the court does not allege any drafting error or absurd consequences here. We have also said courts may disregard plain meaning when it "contravenes clear evidence of a contrary legislative intent." ( Ornelas v. Randolph (1993) 4 Cal.4th 1095, 1105, 17 Cal.Rptr.2d 594, 847 P.2d 560 ( Ornelas ).) In truth, this is the standard of justification that the court's decision to disregard the plain meaning of the phrase "As used throughout this Code" must meet. The court is not unaware of this, for why else does today's opinion argue at such length that giving effect to the phrase's plain meaning would violate the voters' intent? This labored effort to prove that the voters who enacted the phrase did not really mean it is the real ground of today's decision, notwithstanding the court's attempt to conjure ambiguity out of unambiguous text.

II.

To show that the voters did not mean what the statute says, the court points to (1) uncodified language in Proposition 47 illuminating the initiative's purpose, (2) the absence of procedural mechanisms for applying Proposition 47's definition of "unreasonable risk of danger to public safety" to Proposition 36 resentencing, and (3) extrinsic evidence that voters were unaware of, and therefore could not have intended, the plain meaning of the phrase "As used throughout this Code." None of these, singly or together, is convincing.

A.

According to the court, uncodified provisions of Proposition 47 make clear that the initiative was not intended to affect the resentencing criteria for Proposition 36 petitioners. (Maj. opn., ante , 220 Cal.Rptr.3d at pp. 244–245, 397 P.3d at pp. 947–949.) The court quotes the initiative's "Findings and Declarations" section, which says "[t]his act ensures that sentences for people convicted of dangerous crimes like rape, murder, and child molestation are not changed." (Voter Information Guide, Gen. Elec. (Nov. 4, 2014) text of Prop. 47, § 2, p. 70 (Voter Information Guide).) The court also quotes the "Purpose and Intent" section, which says "it is the purpose and intent of the people of California to: [¶] (1) Ensure that people convicted of murder, rape, and child molestation will not benefit from this act. [¶] .... [¶] (3) Require misdemeanors instead of felonies for nonserious, nonviolent crimes like petty theft and drug possession, unless the defendant has prior convictions for specified violent or serious crimes. [¶] (4) Authorize consideration of resentencing for anyone who is currently serving a sentence for any of the offenses listed herein that are now misdemeanors." (Id. , § 3, at p. 70.)

The court reasons that because three-strikes offenders who are eligible for resentencing under Proposition 36 necessarily have been convicted of violent or serious crimes, the application of Proposition 47's definition of "unreasonable risk of danger to public safety" to resentencing under Proposition 36 would contravene Proposition 47's purpose to "ensure[ ] that sentences for people convicted of dangerous crimes ... are not changed." (Voter Information Guide, supra , text of Prop. 47, § 2, at p. 70.) It would also conflict with Proposition 47's purpose to "[e]nsure that people convicted of murder, rape, and child molestation will not benefit from this act" (id. , § 3, subd. (1), at p. 70), the court says, because three-strikes offenders convicted of nonforcible rape or child molestation involving minor victims over the age of 14 are eligible for resentencing under Proposition 36.

This line of argument is problematic for several reasons. First, although we have looked to statements of purpose contained in a preamble to inform our interpretation of an operative statutory provision that is ambiguous (see Carter v. California Dept. of Veterans Affairs (2006) 38 Cal.4th 914, 922–927, 44 Cal.Rptr.3d 223, 135 P.3d 637 ; People v. Canty (2004) 32 Cal.4th 1266, 1280–1281, 14 Cal.Rptr.3d 1, 90 P.3d 1168 ), I am not aware of any case—the court cites none—in which we have relied on statements of purpose in a preamble to nullify an operative provision that is unambiguous . Indeed, we have repeatedly declined to construe such statements of purpose in a manner that restricts or alters a statute's operative provisions. (See, e.g., Scher v. Burke (2017) 3 Cal.5th 136, 148–149, 218 Cal.Rptr.3d 643, 395 P.3d 680 ; Equilon Enterprises v. Consumer Cause, Inc. (2002) 29 Cal.4th 53, 58–61, 124 Cal.Rptr.2d 507, 52 P.3d 685 ; Briggs v. Eden Council for Hope & Opportunity (1999) 19 Cal.4th 1106, 1117–1119, 81 Cal.Rptr.2d 471, 969 P.2d 564.)

Second, one reason courts do not assign much weight to preamble language relative to operative provisions is that such language is often phrased at a high level of generality and typically includes multiple purposes that point in different directions. (Cf. Sonic-Calabasas A, Inc. v. Moreno (2013) 57 Cal.4th 1109, 1167, 163 Cal.Rptr.3d 269, 311 P.3d 184 [" '[N]o statute ... pursues its "broad purpose" at all costs.' "]; ibid. [" ' "[I]t frustrates rather than effectuates legislative intent simplistically to assume that whatever furthers the statute's primary objective must be the law." ' "].) That is certainly true here. The court, while citing the purposes above, neglects to mention other purposes stated in Proposition 47: to "save significant state corrections dollars on an annual basis" and to "increase investments in programs that reduce crime and improve public safety, such as prevention programs in K-12 schools, victim services, and mental health and drug treatment, which will reduce future expenditures for corrections." (Voter Information Guide, supra , text of Prop. 47, § 3, subd. (6), at p. 70; see id. , § 2, at p. 70 ["The people enact [Proposition 47] to ensure that prison spending is focused on violent and serious offenses, to maximize alternatives for nonserious, nonviolent crime, and to invest the savings generated from this act into prevention and support programs in K–12 schools, victim services, and mental health and drug treatment."].) Applying Proposition 47's definition of "unreasonable risk of danger to public safety" to Proposition 36 petitioners whose third strike is not serious or violent is fully consistent with Proposition 47's purpose of maximizing alternatives for nonserious, nonviolent crime, saving significant state correction dollars each year, and increasing investment in crime prevention, victim services, and mental health and drug treatment to reduce recidivism.

Third, apart from the court's one-sided presentation of Proposition 47's purposes, there is no inconsistency between the purposes the court focuses on and the application of Proposition 47's definition of "unreasonable risk of danger to public safety" to Proposition 36. It is true that inmates eligible for resentencing under Proposition 36 have been convicted of violent or serious felonies. But that is also true of many inmates eligible for resentencing under Proposition 47, which excludes only persons who have been convicted of a "super strike" or a registrable sex offense. ( § 1170.18, subd. (i) ; see, e.g., Harris v. Superior Court (2016) 1 Cal.5th 984, 987–988, 209 Cal.Rptr.3d 584, 383 P.3d 648 [Prop. 47 petitioner had a prior strike for robbery]; People v. Hernandez (2017) 10 Cal.App.5th 192, 195–196, 215 Cal.Rptr.3d 880 [Prop. 47 petitioner had two prior strikes]; People v. Cortez (2016) 3 Cal.App.5th 308, 310, 207 Cal.Rptr.3d 510 [Prop. 47 petitioner admitted a prior strike].) Indeed, the Legislative Analyst made clear in the ballot materials for Proposition 47 that some petitioners eligible for Proposition 47 resentencing had been sent to state prison instead of county jail for now-reclassified nonserious, nonviolent felonies "because they had a prior serious or violent conviction." (Voter Information Guide, supra , analysis by the Legis. Analyst, at p. 36.)

Further, the fact that a smaller share of eligible Proposition 47 petitioners, compared to eligible Proposition 36 petitioners, have prior serious or violent felony convictions does not compel the inference that Proposition 47's definition of "unreasonable risk of danger to public safety" must apply only to the former and not the latter. The definition establishes the level of risk that the electorate found acceptable to authorize a reduced sentence in each individual case. (See Voter Information Guide, supra , text of Prop. 47, § 3, subd. (5), at p. 70 [Prop. 47 "[r]equire[s] a thorough review of criminal history and risk assessment of any individuals before resentencing to ensure that they do not pose a risk to public safety"].) It is perfectly rational for the voters to have established the same risk threshold for evaluating individuals from both populations, even if the populations differ on average in terms of criminal history. The Court of Appeal in Valencia's case said it would be "logical" and not "absurd" to treat differently "someone with multiple prior serious and/or violent felony convictions whose current offense is (or would be, if committed today) a misdemeanor" versus "someone whose current offense is a felony." But even so, there is also nothing illogical or absurd about using the same risk criteria in evaluating both types of recidivist offenders. That is simply a policy judgment— and one that the voters made clearly in the text of Proposition 47.

Nor does it help the court's analysis that Proposition 47, but not Proposition 36, excludes from resentencing eligibility persons who have committed certain sex offenses involving minor victims over the age of 14. It is true that Proposition 47 declares a general purpose of "ensur[ing] that people convicted of murder, rape, and child molestation will not benefit from this act." (Voter Information Guide, supra , text of Prop. 47, § 3, subd. (1), at p. 70.) But Proposition 36 likewise states in its uncodified preamble that its purpose is "to restore the original intent of California's Three Strikes Law—imposing life sentences for dangerous criminals like rapists, murderers, and child molesters" and to "[r]equire that murderers, rapists, and child molesters serve their full sentences—they will receive life sentences, even if they are convicted of a new minor third strike crime." (Voter Information Guide, Gen. Elec. (Nov. 6, 2012), text of Prop. 36, § 1, subd. (1), p. 105.) It is entirely reasonable that the electorate understood Proposition 47's intent not to benefit "people convicted of murder, rape, and child molestation" in the context of the initiative's other purposes in the same way it understood Proposition 36's intent not to benefit "murderers, rapists, and child molesters" in the context of that initiative's other purposes just two years earlier.

It is important to keep in mind the burden that the court's analysis must satisfy here. In order to disregard plain meaning, there must be "clear evidence of a contrary legislative intent" ( Ornelas , supra , 4 Cal.4th at p. 1105, 17 Cal.Rptr.2d 594, 847 P.2d 560, italics added); the plain meaning must be inconsistent with the "manifest purpose" of the statute ( Arias v. Superior Court (2009) 46 Cal.4th 969, 979, 95 Cal.Rptr.3d 588, 209 P.3d 923, italics added). It is hardly enough to cite generally worded preamble language that declares several competing purposes. Such language falls well short of the clarity and precision required to nullify the plain meaning of an operative provision.

B.

As a further reason to reject section 1170.18(c)'s plain meaning, the court identifies several questions that would arise from applying Proposition 47's definition of "unreasonable risk of danger to public safety" to Proposition 36 petitioners: Would the definition apply prospectively or retroactively? Would Proposition 47's new definition constitute "good cause" for filing a Proposition 36 resentencing petition beyond the statutory two-year window? ( § 1170.126, subd. (b).) And what about Proposition 36 petitioners whose resentencing petitions are already final? The fact that Proposition 47 does not address these "serious procedural questions," the court says, is an indication that the initiative was not intended to amend the resentencing criteria for Proposition 36 petitioners. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 248, 397 P.3d at p. 951.)

Conspicuously absent from the court's discussion of this point is any citation to precedent suggesting that the absence of details addressing the procedure for implementing a statutory provision is a reason to disregard the provision's plain meaning. As to whether Proposition 47's narrowing of the definition of unreasonable risk constitutes "good cause" for filing a new or renewed Proposition 36 resentencing petition beyond the two-year window, isn't this the kind of question that lawmakers regularly leave for the courts to answer? Similarly, when we encounter a statute that does not say whether it applies retroactively or prospectively, we do not simply throw up our hands in confusion and infer that the lawmakers, having not addressed this issue, did not mean to enact what they enacted. Instead, we regularly give effect to such statutes in a manner guided by interpretive presumptions in our case law. (See, e.g., People v. Conley (2016) 63 Cal.4th 646, 656–657, 203 Cal.Rptr.3d 622, 373 P.3d 435 ( Conley ); In re Pedro T. (1994) 8 Cal.4th 1041, 1049, 36 Cal.Rptr.2d 74, 884 P.2d 1022 ; People v. Hajek (2014) 58 Cal.4th 1144, 1195–1196, 171 Cal.Rptr.3d 234, 324 P.3d 88 ; People v. Brown (2012) 54 Cal.4th 314, 142 Cal.Rptr.3d 824, 278 P.3d 1182.)

Here, I would apply the presumption that new statutes reducing criminal punishment apply to cases that have not yet proceeded to final judgment, absent evidence of contrary legislative intent. (See In re Estrada (1965) 63 Cal.2d 740, 745, 48 Cal.Rptr. 172, 408 P.2d 948 ["[w]hen the Legislature amends a statute so as to lessen the punishment" for a crime, "[i]t is an inevitable inference that the Legislature must have intended that the new statute imposing the new lighter penalty now deemed to be sufficient should apply to every case to which it constitutionally could apply"].) The Attorney General says the Estrada presumption does not apply because section 1170.18(c) merely narrows resentencing discretion and does not reduce the punishment for a particular crime. But by narrowing the definition of "unreasonable risk of danger to public safety," Proposition 47 recalibrates the universe of third-strike offenders entitled to lesser punishment for a defined category of offenses, i.e., nonserious, nonviolent felonies. ( § 1170.126, subd. (f) [a third-strike offender eligible for resentencing "shall" be resentenced unless the court determines that resentencing the offender "would pose an unreasonable risk of danger to public safety"].) Proposition 47's definition of unreasonable risk plainly " 'represents a legislative judgment that the lesser penalty or the different treatment is sufficient to meet the legitimate ends of the criminal law' " for a subset of defendants whose third strike was a nonviolent, nonserious felony. ( Estrada , at p. 745, 48 Cal.Rptr. 172, 408 P.2d 948.) This would not be the first time we have applied the Estrada presumption to a statute revising trial courts' sentencing discretion to allow for lesser punishment. (See People v. Francis (1969) 71 Cal.2d 66, 76, 75 Cal.Rptr. 199, 450 P.2d 591.) Because nothing in the text or history of Proposition 47 rebuts the Estrada presumption, I would hold that Proposition 47's definition of "unreasonable risk of danger to public safety" applies to all Proposition 36 resentencing petitions that have not yet proceeded to final judgment, including the two defendants' petitions here.

The main point is that courts do not infer that a statute does not mean what it says simply because the statute does not say more about how it is to be implemented. Lawmakers know that courts have plenty of tools to answer questions that statutes leave unanswered. The court's approach, which effectively " 'dictate[s] to legislative drafters the forms in which laws must be written' " ( Conley , supra , 63 Cal.4th at p. 656, 203 Cal.Rptr.3d 622, 373 P.3d 435 ), is unprecedented, as the court's failure to cite any precedent confirms. C.

In addition to citing the lack of procedures to guide Proposition 47's applicability to resentencing under Proposition 36, the court says the ballot materials for Proposition 47 "signaled no relationship at all" with Proposition 36. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 248, 397 P.3d at p. 951.) The court treats this silence as a reason to jettison the usual presumptions that the voters have duly considered the initiative's text and are aware of existing laws. (Id. at pp. 249–253, 397 P.3d at pp. 952–955.) According to the court, we should not give effect to the plain meaning of the phrase "As used throughout this Code" in section 1170.18(c) because apart from the statutory text, nothing else informed the voters that Proposition 47's definition of "unreasonable risk of danger to public safety" would apply to Proposition 36 petitioners.

This line of argument is as startling as it is threatening to the orderly development of the law. (See Eskridge et al., Legislation and Statutory Interpretation (2d ed. 2006) p. 231 ["plain meaning is the most obvious and perhaps the most objective focal point for all of us to know what the rule of law requires of us and our neighbors"].) Is the judiciary free to nullify an unambiguous statutory provision whenever it can be shown that the ballot materials did not inform the electorate of its consequences? Our precedent, including two recent decisions construing Proposition 47, makes clear this is not the law. We do not impute such ignorance to the voters out of respect for the democratic process and out of concern for overstepping the judicial function. If we can rewrite statutes on the ground that the voters were not aware of what they were enacting, there will be no end to the mischief that courts and litigants can inflict on the initiative process.

We have long followed the rule that voters "must be assumed to have voted intelligently upon an [initiative measure], the whole text of which was supplied each of them prior to the election, and which they must be assumed to have duly considered, regardless of any insufficient recitals in the instructions to voters or the arguments pro and con of its advocates or opponents accompanying the text of the proposed measure." ( Wright v. Jordan (1923) 192 Cal. 704, 713, 221 P. 915, italics omitted ( Wright ).) This rule has led us, repeatedly, to enforce the plain meaning of an initiative's text even when its consequences were not apparent from the ballot materials.

In People v. Romanowski (2017) 2 Cal.5th 903, 215 Cal.Rptr.3d 758, 391 P.3d 633, we held that theft of access card (e.g., credit or debit card) information is a crime eligible for reduced punishment under Proposition 47. We explained that Proposition 47 reduced various theft crimes from felonies to misdemeanors where the value of the property taken does not exceed $950. ( Romanowski , at p. 908, 215 Cal.Rptr.3d 758, 391 P.3d 633.) Although neither the initiative's text nor the ballot materials mentioned theft of access card information in describing reclassified theft offenses, we observed that the text of Proposition 47 does refer to " '[s]ection 487 or any other provision of law defining grand theft .' " ( Romanowski , at p. 908, 215 Cal.Rptr.3d 758, 391 P.3d 633, quoting § 490.2, italics added by Romanowski .) We said: "We deny a phrase like 'any other provision of law' its proper impact if we expect a penal statute—whether enacted by the Legislature or the electorate—to further enumerate every provision of the Penal Code to which it is relevant. And we generally presume that the electorate is aware of existing laws. [Citation.] Here this means we must presume that voters were at least aware that the Penal Code sets out 'grand theft' crimes that included theft of access card account information." ( Romanowski , at pp. 908–909, 215 Cal.Rptr.3d 758, 391 P.3d 633.)

In People v. Gonzales (2017) 2 Cal.5th 858, 216 Cal.Rptr.3d 285, 392 P.3d 437, we held that the new crime of shoplifting created by Proposition 47 includes any entry into a commercial establishment during regular business hours with intent to take property worth no more than $950, including entering a bank to cash a forged check. We acknowledged that this holding diverged from the "colloquial understanding" of shoplifting ( Gonzales , at p. 871, 216 Cal.Rptr.3d 285, 392 P.3d 437 ), and our examination of the ballot materials revealed nothing that mentioned its application to nonlarcenous thefts ( id. at pp. 869–870, 216 Cal.Rptr.3d 285, 392 P.3d 437 ). But we observed that Proposition 47's definition of shoplifting "expressly mentions the burglary statute" ( Gonzales , at p. 869, 216 Cal.Rptr.3d 285, 392 P.3d 437 ), and we said the burglary statute had long been construed to cover theft by false pretenses ( id. at pp. 867–868, 216 Cal.Rptr.3d 285, 392 P.3d 437 ). Reasoning that "[t]he electorate 'is presumed to be aware of existing laws and judicial construction thereof,' " we concluded that the voters intended the definition of shoplifting to have the same scope as the burglary statute. ( Id. at p. 869, 216 Cal.Rptr.3d 285, 392 P.3d 437.)

In Day v. City of Fontana (2001) 25 Cal.4th 268, 105 Cal.Rptr.2d 457, 19 P.3d 1196, we held that a provision of Proposition 213 "precludes uninsured drivers from recovering noneconomic damages in actions against local public entities for nuisance and dangerous condition of property." ( Day , at p. 282, 105 Cal.Rptr.2d 457, 19 P.3d 1196.) Finding that "[s]uch actions fall squarely within the terms" of the statute barring liability ( id. at p. 282, 105 Cal.Rptr.2d 457, 19 P.3d 1196 ; see id. at pp. 273, 277, 105 Cal.Rptr.2d 457, 19 P.3d 1196 ), we said "it is of no consequence here that the ballot materials did not specifically refer to the act's application in actions against local public entities for nuisance and dangerous condition of property" ( id. at p. 282, 105 Cal.Rptr.2d 457, 19 P.3d 1196 ). Ballot arguments, we explained, " 'are not legal briefs and are not expected to cite every case the proposition may affect.' [Citations.] Here we may reasonably infer from the ballot arguments that a primary aim of Proposition 213 was to protect insured motorists and to reduce automobile insurance rates. [Citation.] Such arguments, however, did not imply that protection of insured motorists was the initiative's sole aim; nor did they suggest that reductions in automobile insurance premiums would be the initiative's only effect." ( Id. at pp. 278–279, 105 Cal.Rptr.2d 457, 19 P.3d 1196.)

In Lungren , supra , 14 Cal.4th 294, 58 Cal.Rptr.2d 855, 926 P.2d 1042, we held that Proposition 65's prohibition on releasing toxic chemicals " 'into any source of drinking water' " covered the discharge of toxic chemicals into faucet water. The defendants argued that the analysis of the Legislative Analyst, which spoke of the initiative's effect on government-issued permits, suggested to the electorate that Proposition 65 had to do with surface and ground waters, not plumbing faucets. ( Lungren , at p. 308, 58 Cal.Rptr.2d 855, 926 P.2d 1042.) We said: "The Legislative Analyst did not suggest that all the effects and ramifications of the Act were being set forth in his brief summation.... In light of the explicit language and purpose of the statute, and the generality and brevity of the Legislative Analyst's commentary, the latter cannot plausibly be viewed as implicitly limiting the scope of the statute in the manner advocated by defendants." ( Ibid. )

In Amwest Surety Ins. Co. v. Wilson (1995) 11 Cal.4th 1243, 48 Cal.Rptr.2d 12, 906 P.2d 1112, we concluded that Proposition 103 applied to surety insurance, finding the statutory text "clear" and not "ambiguous." ( Amwest , at p. 1260, 48 Cal.Rptr.2d 12, 906 P.2d 1112.) The plaintiff insurance company observed that the ballot materials "did not mention surety insurance" and thus "it is unlikely the voters understood that Proposition 103 applied to surety insurance." ( Ibid. ) But we declined the plaintiff's suggestion " 'to make [the proposition] say what the voters thought it meant.' " ( Amwest , at pp. 1260–1261, 48 Cal.Rptr.2d 12, 906 P.2d 1112.)

In Delaney v. Superior Court (1990) 50 Cal.3d 785, 268 Cal.Rptr. 753, 789 P.2d 934, we considered whether an initiative amending the newsperson's shield law in our state Constitution applied only to confidential as opposed to all unpublished information collected by reporters. We found it "clear and unambiguous" that the enacted text—which protected any newsperson for " 'refusing to disclose any unpublished information' "—applied to all such information. ( Id. at p. 798, 268 Cal.Rptr. 753, 789 P.2d 934, italics added by Delaney .) The defendant, who sought to compel testimony from two journalists, suggested we rely on the Legislative Analyst's statement and ballot arguments to find that the voters intended the shield law to apply only to confidential information. ( Id. at pp. 801–802, 268 Cal.Rptr. 753, 789 P.2d 934.) We declined to do so: "We cannot conclude that, by emphasizing one purpose, perhaps the primary purpose of the measure, the argument misled voters into thinking confidentiality was the only purpose, especially when the measure itself made clear that all unpublished information would be protected. Moreover, a possible inference based on the ballot argument is an insufficient basis on which to ignore the unrestricted and unambiguous language of the measure itself." ( Id. at pp. 802–803, 268 Cal.Rptr. 753, 789 P.2d 934.)

And in Cervera , supra , 24 Cal.4th 1073, 103 Cal.Rptr.2d 762, 16 P.3d 176, we held that the Three Strikes law does not authorize prison conduct credits for use against indeterminate terms despite legislative history and ballot materials suggesting otherwise. We observed that the text of the Three Strikes laws authorize the award of prison conduct credits pursuant to a separate statute addressing only the application of such credits to determinate sentences. ( Cervera , at p. 1078, 103 Cal.Rptr.2d 762, 16 P.3d 176.) We acknowledged that the Legislative Analyst's analysis, the ballot arguments, and various legislative committee analyses discussed the availability of prison conduct credits without distinguishing between determinate and indeterminate terms. ( Id. at p. 1079, 103 Cal.Rptr.2d 762, 16 P.3d 176.) "But," we concluded, "it was the Three Strikes law that was enacted, not any of the documents within its legislative or initiative history. A statute, of course, must prevail over any summary. Were it not so, no statute could ever be enacted whole and entire. For every summary, by definition, is incomplete. A summary is a model of the body of the statute, well executed or not as the case may be. It is not a procrustean bed for the stretching or lopping off of its limbs. The summary must yield to the statute, not the statute to the summary." ( Id. at pp. 1079–1080, 103 Cal.Rptr.2d 762, 16 P.3d 176.)The lessons of this body of precedent are clear. The unambiguous language of an initiative is the best evidence of the voters' intent. Ballot materials may help illuminate the purpose of an initiative statute. But the fact that ballot materials do not speak to the application of the statute in a particular case is not a reason to disregard the plain meaning of the enacted text. Ballot materials are necessarily incomplete and serve as summaries. "The summary must yield to the statute, not the statute to the summary." ( Cervera , supra , 24 Cal.4th at p. 1080, 103 Cal.Rptr.2d 762, 16 P.3d 176.)

Against the backdrop of this case law, it is mystifying how the court can place such heavy reliance on the fact that the ballot materials prepared by the Attorney General and Legislative Analyst did not mention Proposition 47's effects on resentencing three-strikes inmates. (Maj. opn., ante , 220 Cal.Rptr.3d at pp. 245–248, 251–252, 253–254, 397 P.3d at pp. 949–951, 954–955, 955–956.) The court makes reference to "[t]he statutory provisions imposed on the Attorney General and Legislative Analyst in order to educate voters about the effect of proposed initiatives and to protect them from being misled or confused." (Id. at p. 254, 397 P.3d at p. 956.) But as the Court of Appeal explained in People v. Cordova (2016) 248 Cal.App.4th 543, 558–559, 203 Cal.Rptr.3d 700 ( Cordova ), the contents of the ballot pamphlet "are constrained by considerations of space, time, and subjective determinations of materiality. The official summary of any ballot measure is authored by the office of the Legislative Analyst. ( Elec. Code, §§ 9087, 9086, subd. (b).) The summary is required only to 'generally set forth in an impartial manner the information the average voter needs to adequately understand the measure.' ( Elec. Code, § 9087, subd. (b), italics added.) Note the absence of any directive that the Legislative Analyst attempt to provide voters with a complete understanding of the measure, which would be a practical impossibility in any event; few judges or lawyers would be so arrogant as to profess that they completely understand any provision of law, at least in the sense of being able to forecast all of its effects. (See In reGabriel G. [ (2008) ] 134 Cal.App.4th [1428,] 1437 ["Legislation often has unintended consequences."].) In manifest recognition of this fact, the governing statute states only that the ballot summary 'may contain' such 'background information' as 'the effect of the measure on existing law.' ( Elec. Code, § 9087, subd. (b), italics added.) The Legislative Analyst is thus called upon only to make a rational judgment about what effects are most likely to matter to voters, and to describe them in a fair and intelligible way. Inherent in this undertaking is the necessity of informational triage—of determining what details are necessary to form an 'adequate[ ] understanding [ing],' and what details may be omitted. The preparer of such a summary necessarily exercises a discretionary function requiring courts to allow considerable latitude when the result is challenged as incomplete or inaccurate. (See Brennan v. Board of Supervisors (1981) 125 Cal.App.3d 87, 96, 177 Cal.Rptr. 677 ["Faced with the difficult task of simplifying a complex proposal, the Committee drafted a summary which, if not all-encompassing, at least briefly described its major subjects."].)

"When the Legislative Analyst fails to mention some effect of a ballot measure, it remains open to the measure's official proponents and opponents to use their space in the ballot pamphlet to supply any perceived lack. But they too must practice triage; their arguments are restricted to 500 words to open and 250 words in rebuttal. (See Elec. Code, §§ 9062, 9069 ; cf. id. , § 9041.) This means the advocates must select a limited number of points to include in their voter information guide arguments, relying on other media to pursue issues deemed of lesser moment. Here, both sides evidently concluded that the effect of section 1170.18(c) on Proposition 36 petitions was not a powerful enough ground of argument to warrant mention in the guide. That decision can hardly furnish an occasion for judicial nullification." (Fn. omitted.) Today's opinion does not contend that Cordova 's description of the Legislative Analyst's duties under the Elections Code is incorrect. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 247, fn. 7, 397 P.3d at p. 950, fn. 7; see Cal. Rules of Court, rule 8.1115, subd. (e)(2) [a published Court of Appeal opinion on a matter in which we have granted review "is citable and has binding or precedential effect, except to the extent it is inconsistent with the decision of the Supreme Court or is disapproved by that court"].)

Instead of grappling with the precedent above, the court relies on Robert L. v. Superior Court (2003) 30 Cal.4th 894, 135 Cal.Rptr.2d 30, 69 P.3d 951 and Taxpayers to Limit Campaign Spending v. Fair Political Practices Comm. (1990) 51 Cal.3d 744, 274 Cal.Rptr. 787, 799 P.2d 1220 ( Taxpayers ). But in Robert L. , we rejected a technical reading of the phrase "punishable as a felony or a misdemeanor" in section 186.22, subdivision (d) in favor of how "the average voter ... would have understood the plain language." ( Robert L. , at p. 902, 135 Cal.Rptr.2d 30, 69 P.3d 951.) In this case, the court is rejecting how the average voter would have understood the plain language of section 1170.18(c) in favor of nullifying the disputed language altogether.It is also far-fetched to say this case is "similar" to Taxpayers in demanding an unreasonable degree of thoroughness on the part of the electorate. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 251, 397 P.3d at p. 953.) In Taxpayers , the voters approved two overlapping ballot initiatives. We held that the initiative receiving more votes should be given effect, rejecting the view that we should attempt to reconcile the competing measures. ( Taxpayers , supra , 51 Cal.3d at p. 770, 274 Cal.Rptr. 787, 799 P.2d 1220.) No case had ever assumed that "voters not only recognized that they were approving initiatives with fundamentally conflicting provisions intended to regulate the same subject, but also analyzed the remaining provisions in order to predict which would be implemented if either measure received a lesser affirmative vote." ( Id. at p. 768, 274 Cal.Rptr. 787, 799 P.2d 1220.) Hybridizing the two initiatives, we said, would "implement a fictitious electoral intent." ( Ibid. ) Here, the only issue is the voters' understanding of a single unambiguous and easily intelligible phrase: "As used throughout this Code."

The court's theory that the electorate was likely ignorant of the plain meaning of section 1170.18(c) when it enacted Proposition 47 is misguided at a more fundamental level. The presumption that voters have attentively studied the text of measures they enact and are aware of existing laws is not primarily based on the notion that most voters do in fact read the text or study the background law. Instead, it is based on the importance of separating the judicial function from the legislative function in our system of government. Without a strong presumption that the legislative authority means what it says in its official enactments, the lawmaking process would regularly be subject to "collateral attack" ( Wright , supra , 192 Cal. at p. 713, 221 P. 915 ), with courts rewriting plain language to fit their conception of what the lawmakers really meant. We should not countenance such lack of judicial restraint or such erosion of legislative accountability.

In this case, the court claims neither a drafting error nor absurd consequences. The court instead contends that the voters, having duly enacted section 1170.18(c), were caught unawares and did not actually intend to enact what they did. This holding is highly irregular and goes beyond our legitimate authority to interpret statutes. We should enforce the statute as written and leave any corrections up to the usual lawmaking process. ( Cal. Const., art. I, §§ 9, 10, subd. (c).)

III.

I conclude with a bit of context that may further illuminate why many will look askance at today's decision. In the early 1980s, as the "law and order" movement accelerated in this state and throughout the nation, California voters enacted a major criminal justice reform initiative known as "The Victims' Bill of Rights" (Proposition 8). (See Brosnahan v. Brown (1982) 32 Cal.3d 236, 242–245, 186 Cal.Rptr. 30, 651 P.2d 274 ( Brosnahan ).) This initiative affected numerous areas of our criminal law: restitution awards, evidentiary standards, bail, sentencing practices, diminished capacity defenses, plea bargaining, and more. The petitioners in Brosnahan , claiming that the initiative violated the single-subject rule, argued that "the complexity of Proposition 8 may have led to confusion or deception among voters, who were assertedly uninformed regarding the contents of the measure." ( Id. at p. 251, 186 Cal.Rptr. 30, 651 P.2d 274.) We rejected this argument as follows:

"... Proposition 8 received widespread publicity. Newspaper, radio and television editorials focused on its provisions, and extensive public debate involving candidates, letters to the editor, etc., described the pros and cons of the measure. In addition, before the election each voter received a pamphlet containing (1) the title and summary prepared by the Attorney General, (2) a detailed analysis of the measure by the Legislative Analyst, and (3) a complete 'Text of the Proposed Law.' This text contained the entirety of the 10 sections of the Victims' Bill of Rights and included in 'strikeout type' the text of former article I, section 12, of the Constitution. Each voter also was given written arguments in favor of Proposition 8 and rebuttal thereto, and written arguments against Proposition 8 and rebuttal thereto. [Citations.]

"Moreover, as we stated in [ Fair Political Practices Com. v. Superior Court (1979) 25 Cal.3d 33, 42, 157 Cal.Rptr. 855, 599 P.2d 46 ] in disposing of an identical contention that the measure was too complicated, 'Our society being complex, the rules governing it whether adopted by legislation or initiative will necessarily be complex. Unless we are to repudiate or cripple use of the initiative, risk of confusion must be borne.' [Citation.]

"Petitioners' entire argument that, in approving Proposition 8, the voters must have been misled or confused is based upon the improbable assumption that the people did not know what they were doing. It is equally arguable that, faced with startling crime statistics and frustrated by the perceived inability of the criminal justice system to protect them, the people knew exactly what they were doing. In any event, we should not lightly presume that the voters did not know what they were about in approving Proposition 8. Rather, in accordance with our tradition, 'we ordinarily should assume that the voters who approved a constitutional amendment "... have voted intelligently upon an amendment to their organic law, the whole text of which was supplied each of them prior to the election and which they must be assumed to have duly considered ." ' [Citations.]" ( Brosnahan , supra , 32 Cal.3d at p. 252, 186 Cal.Rptr. 30, 651 P.2d 274, italics added by Brosnahan .)

This court has applied Proposition 8 in accordance with the full sweep of its plain language, even if the text or ballot materials did not alert the electorate to particular applications. (See People v. Alvarez (2002) 27 Cal.4th 1161, 1174–1175, 119 Cal.Rptr.2d 903, 46 P.3d 372 ; In re Lance W. (1985) 37 Cal.3d 873, 885–888, 210 Cal.Rptr. 631, 694 P.2d 744.) And, as noted (ante , 220 Cal.Rptr.3d at p. 264, 397 P.3d at p. 964), we have taken the same approach in applying the Three Strikes law.

Today California voters, like many citizens throughout America, have had second thoughts about our past criminal justice policies. Fiscal, moral, religious, and public safety considerations have motivated an unusually wide spectrum of leaders to reexamine penal laws and sentencing practices. (See, e.g., Chettiar & Waldman, Solutions: American Leaders Speak Out on Criminal Justice (2015); Ball, Do the Koch Brothers Really Care About Criminal-Justice Reform? (Mar. 3, 2015) The Atlantic.) It is in this context that the electorate, by large majorities, passed Proposition 36 and Proposition 47. Like the "tough on crime" initiatives from previous decades, Proposition 47 was vigorously debated, and statewide advocacy groups highlighted its various implications, including its effect on resentencing petitions under Proposition 36. The court notes that one of the authors of Proposition 47, as reported the day after the 2014 general election, said three-strikes inmates previously denied resentencing " 'could return to court and cite Proposition 47's new definition of an "unreasonable risk of danger." ' " (Maj. opn., ante , 220 Cal.Rptr.3d at p. 249, fn. 9, 397 P.3d at p. 952, fn. 9, quoting St. John & Gerber, Prop. 47 Jolts Landscape of California Justice System , L.A. Times (Nov. 5, 2014).)

The court today concludes that the drafters of Proposition 47 pulled a fast one on an uninformed public. But it is at least "equally arguable that, faced with startling [prison spending] statistics and frustrated by the perceived inability of the criminal justice system to [reduce costs], the people knew exactly what they were doing" when they passed Proposition 47. ( Brosnahan , supra , 32 Cal.3d at p. 252, 186 Cal.Rptr. 30, 651 P.2d 274.) Many voters may have reasonably believed that inmates whose third strike was neither serious nor violent should not have received 25-years-to-life sentences or that some three-strikes inmates serving lengthy sentences have rehabilitated themselves or have aged out of crime. "In any event, we should not lightly presume that the voters did not know what they were about in approving" the initiative ( ibid. ) when the application of its definition of "unreasonable risk of danger to public safety" to inmates serving a third-strike sentence for a nonviolent, nonserious offense is fully consistent with Proposition 47's stated purpose "to ensure that prison spending is focused on violent and serious offenses," "to maximize alternatives for nonserious, nonviolent crime," and to "save significant state corrections dollars." (Voter Information Guide, supra , text of Prop. 47, §§ 2, 3, at p. 70.)

We should apply the same interpretive approach to this ameliorative statute that we applied to the punitive statutes of an earlier era. The aberrant approach to statutory interpretation in today's decision disserves the initiative process, the inmates who are now its beneficiaries, and the judicial role itself.

I would reverse the judgments of the Courts of Appeal in these cases and remand each case to the resentencing court to consider whether the petitioner poses an "unreasonable risk of danger to public safety" as defined in section 1170.18(c).

We Concur:

Werdegar, J.

Cuéllar, J.

DISSENTING OPINION BY CUÉLLAR, J.

Cuéllar, J.

In 2014, Californians enacted Proposition 47, the Safe Neighborhoods and Schools Act. This initiative allowed certain offenders convicted of less serious nonviolent felonies and wobblers to petition to have their felony convictions reclassified as misdemeanors and their felony sentences recalled for resentencing. ( Pen. Code, § 1170.18, added by Prop. 47, § 1, approved by voters at Gen. Elec. (Nov. 4, 2014).)^{1 1} Proposition 47, however, carved out an exception precluding relief when reclassification would "result in an unreasonable risk of danger to public safety." The measure defined the nature of this particular risk throughout the Penal Code. The phrase "[a]s used throughout this Code" in section 1170.18, subdivision (c) means what it says. Thus, the definition applies in the context of Proposition 36, the Three Strikes Reform Act of 2012.

1 What is commonly referred to as the Three Strikes law "consists of two, nearly identical statutory schemes": the first enacted by the Legislature in March 1994 (Pen. Code, former § 667, subds. (b) -(i) ), and the second enacted by ballot initiative in November 1994 (Pen. Code, former § 1170.12, added by Prop. 184, as approved by voters, Gen. Elec. (Nov. 8, 1994)). (People v. Superior Court (Romero) (1996) 13 Cal.4th 497, 504-505, 53 Cal.Rptr.2d 789, 917 P.2d 628.)

1 All further statutory citations are to the Penal Code unless otherwise noted.

The parties accept that Proposition 47 leaves intact the eligibility criteria to petition for resentencing under the Three Strikes Reform Act. Under Proposition 47's more specific definition of "unreasonable risk of danger to public safety," courts retain discretion to determine whether sentence recall is warranted, though subject to more precise constraints than would apply under the previously undefined Proposition 36 standard. Permitting Proposition 47's definition to apply wherever the phrase "unreasonable risk of danger to public safety" appears throughout the Penal Code tends to advance the goal of concentrating state corrections spending on the most dangerous offenders. And giving effect to the terms of section 1170.18, subdivision (c) would not result in automatic, nor immediate, release of Proposition 36 petitioners. Thus, the ordinary understanding of this provision's language is consistent with the statute's overall structure and context, and by imbuing the language with its ordinary meaning we promote the broad remedial purposes of both the Safe Neighborhoods and Schools Act and the Three Strikes Reform Act.

So why does the court conclude that "throughout this Code" ( § 1170.18, subd. (c) ) cannot possibly mean "throughout this Code" and instead must mean "as applicable only to the resentencing proceedings that are authorized under Proposition 47"? (Maj. opn., ante , 220 Cal.Rptr.3d at p. 255, 397 P.3d at p. 957.) Not because the ordinary meaning of "[a]s used throughout this Code" in any way suggests that these words narrow rather than expand the applicability of Proposition 47's definition. Indeed, one needs more than even the Theory of Relativity to make "throughout this Code" mean, as a matter of ordinary usage, "only ... under Proposition 47." Not because it's clear from the text, structure, or context of the initiative that the term "the petitioner" excludes all petitioners except those encompassed by Proposition 47. Not because the court here undertakes a thorough analysis of past cases where we have, in fact, failed to apply the literal meaning of a statutory provision. Not because the initiative contains a drafting error. And not because it can be said that giving effect to the phrase as written results in absurd consequences. Instead, the court finds that "it is reasonable" to confine Proposition 47's definition of the exception precluding relief to the initiative in which it was introduced. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 255, 397 P.3d at p. 957.)

The most striking problem with the court's conclusion, for me, is that it's not clear what justifies it. Even if one were to accept the implicit idea that we are simply weighing competing "reasonable" definitions of an opaque statute in this case, there is nothing "reasonable" about failing to apply a statutory definition throughout the code when it is supposed to apply "throughout this Code." For the majority, sentencing judges should retain somewhat greater discretion under the Three Strikes Reform Act to deny relief to individuals who'd have a marginally stronger basis to obtain such relief if the phrase "[a]s used throughout this Code" were given proper effect. To reach this conclusion, the majority appears to rely on an implausibly cramped reading of Proposition 47's purposes. Nothing in the court's analysis persuasively supports a reading of the statute that essentially ignores the words "throughout this Code" and their place in the statutory scheme. Nor do the majority opinion and concurrence persuade in insisting that "the petitioner" in section 1170.18, subdivision (c) must mean only the Proposition 47 petitioner based on the provision's use of the definite article "the." By eschewing the statutory provision's ordinary meaning in this context, the court redrafts Proposition 47 without a credible rationale. In the process, the court fails to give effect not only to the text and explicit purposes of the initiative, but also to a variety of tenets we generally apply when we interpret statutes. Which is why I respectfully dissent.

I.

The Legislature enacted the "Three Strikes" law in 1994 " 'to ensure longer prison sentences and greater punishment for those who commit a felony and have been previously convicted of one or more serious and/or violent felony offenses.' " ( § 667, subd. (b).) That same year, California voters enacted a nearly identical measure through ballot initiative. ( § 1170.12, as added by Prop. 184, § 1, Gen. Elec. (Nov. 8, 1994); see also People v. Sasser (2015) 61 Cal.4th 1, 11, 186 Cal.Rptr.3d 540, 347 P.3d 522 ; In reCervera (2001) 24 Cal.4th 1073, 1075, 103 Cal.Rptr.2d 762, 16 P.3d 176 ( Cervera ).) Under the Three Strikes law, any defendant convicted of a new felony who had previously been convicted of one "strike"—that is, a qualifying serious or violent felony—faced a sentence equal to "twice the term otherwise provided as punishment for the current felony conviction." ( § 1170.12, former subd. (c)(1).) For those defendants convicted of any new felony with two or more prior strikes, the Three Strikes law required courts to impose "an indeterminate term of life imprisonment with a minimum term of" at least 25 years. ( § 1170.12, former subd. (c)(2)(A); see Teal v. Superior Court (2014) 60 Cal.4th 595, 596, 179 Cal.Rptr.3d 365, 336 P.3d 686 ( Teal ).) One particularly controversial aspect of the law was that the third strike felony triggering a life sentence could be nonviolent and relatively minor. (See In re Coley (2012) 55 Cal.4th 524, 528–529, 146 Cal.Rptr.3d 382, 283 P.3d 1252.)Over the next 20 years, California's prison population surged. By 2008, the state's correctional facilities housed roughly 156,000 persons, almost double the number of inmates the state's prisons were designed to hold. ( Brown v. Plata (2011) 563 U.S. 493, 501, 131 S.Ct. 1910, 179 L.Ed.2d 969 ( Brown v. Plata ).) In 2009, a three-judge federal court concluded that overcrowding was the "primary cause of the state's unconstitutional failure to provide adequate medical and mental care" to its prisoners and ordered the state to reduce its prison population to 137.5 percent of design capacity within two years. ( Coleman v. Schwarzenegger (E.D.Cal. 2009) 922 F.Supp.2d 882, 949 ; see also id. at pp. 969–970.) In May 2011, the United States Supreme Court affirmed the three-judge court's remedial order, holding that "the court-mandated population limit [was] necessary to remedy the violation of prisoners' constitutional rights." ( Brown v. Plata , at p. 502, 131 S.Ct. 1910.)

The Legislature and the Governor responded by enacting the Criminal Justice Realignment Act of 2011. (Stats. 2011, ch. 15, § 1; Stats. 2011, 1st Ex. Sess. 2011–2012, ch. 12, § 1 (Realignment Act); see also People v. Scott (2014) 58 Cal.4th 1415, 1418, 171 Cal.Rptr.3d 638, 324 P.3d 827.) The Realignment Act revised sentencing for nonserious, nonviolent, and non-sex registrant offenders, directing them from the state to local jurisdictions, and shifting fiscal responsibility for those offenders from the state to local governments. (See Scott , at pp. 1418–1419, 171 Cal.Rptr.3d 638, 324 P.3d 827 ; § 1170, subd. (h)(2), (3), (5).)

California voters then began using the initiative process to reform some of the most punitive features of the criminal justice system. In 2012, voters approved Proposition 36, the Three Strikes Reform Act. (As approved by voters, Gen. Elec. (Nov. 6, 2012).) In the wake of Brown v. Plata and the Realignment Act's passage, Proposition 36 did two things to help advance the goal of reallocating scarce prison spending and space from low-risk offenders to more dangerous ones. First, the measure amended the Three Strikes law to reduce the punishment for second strike defendants who commit third strike felony offenses that are neither serious nor violent. (See §§ 667, subd. (e), 1170.12, subd. (c).) Second, as relevant in this case, Proposition 36 added a resentencing provision permitting third strike offenders presently serving an indeterminate term of life imprisonment for a nonserious or nonviolent felony to petition for a reduction of their term to a second strike sentence. (See § 1170.126, subd. (b) ; Teal , supra , 60 Cal.4th at p. 597, 179 Cal.Rptr.3d 365, 336 P.3d 686.)

A third strike inmate who satisfies certain specified criteria "shall be resentenced" as a second strike offender to "twice the term otherwise provided for the current felony conviction," "unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§§ 1170.126, subd. (f), 1170.12, subd. (c)(1).) Despite the inclusion of mandatory language, Proposition 36 supplied a judicial safeguard permitting discretionary denial of resentencing petitions. When exercising that discretion, the court may consider: "(1) The petitioner's criminal conviction history, including the type of crimes committed, the extent of injury to victims, the length of prior prison commitments, and the remoteness of the crimes"; "(2) The petitioner's disciplinary record and record of rehabilitation while incarcerated"; and "(3) Any other evidence the court, within its discretion, determines to be relevant in deciding whether a new sentence would result in an unreasonable risk of danger to public safety." (§ 1170.126, subd. (g)(1)–(3).)

Two years after the passage of Proposition 36, in November 2014, California voters approved additional changes in sentencing law by enacting Proposition 47, the Safe Neighborhoods and Schools Act. (As approved by voters, Gen. Elec. (Nov. 4, 2014).) Proposition 47 reclassified certain drug- and theft-related felony and wobbler offenses as misdemeanors. (See § 1170.18, subds. (a), (b).) According to its provisions, an inmate who "would have been guilty of a misdemeanor" had the measure been effective at the time of the offense "may petition for a recall of sentence before the trial court that entered the judgment of conviction in his or her case to request resentencing in accordance with" relevant drug- and theft-related Penal Code provisions amended or added by Proposition 47. ( § 1170.18, subd. (a).) If, under these provisions, the court determines that the inmate petitioner would have been guilty of a misdemeanor rather than a felony, the felony sentence "shall be recalled" and the petitioner resentenced to a misdemeanor. ( § 1170.18, subd. (b).)

Through a procedure modeled on Proposition 36's, Proposition 47 permits the sentencing court to deny a resentencing petition from an inmate otherwise entitled to relief if "the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§ 1170.18, subd. (b).) In determining whether the petitioner poses an unreasonable risk of danger to public safety under Proposition 47, the court considers factors exactly identical to those it considers for risk assessment purposes under Proposition 36: "(1) The petitioner's criminal conviction history, including the type of crimes committed, the extent of injury to victims, the length of prior prison commitments, and the remoteness of the crimes"; "(2) The petitioner's disciplinary record and record of rehabilitation while incarcerated"; and "(3) Any other evidence the court, within its discretion, determines to be relevant in deciding whether a new sentence would result in an unreasonable risk of danger to public safety." (§ 1170.18, subd. (b)(1)–(3).) "As used throughout this Code," the statute states, the term " 'unreasonable risk of danger to public safety' " means an "unreasonable risk that the petitioner will commit a new [super strike]."² ( § 1170.18, subd. (c).)

II.

So what does Proposition 47's phrase "[a]s used throughout this Code" mean here? When we interpret statutes, giving effect to legislative purpose is the touchstone of our mission. (See Goodman v. Lozano (2010) 47 Cal.4th 1327, 1332, 104 Cal.Rptr.3d 219, 223 P.3d 77 ["In interpreting a statute, our primary goal is to determine and give effect to the underlying purpose of the law"].) "[A]long with the text of the statutory provision directly at issue, structure and context can be critical in determining whether ambiguity exists and in discerning the Legislature's intended purpose." ( Poole v. Orange County Fire Authority (2015) 61 Cal.4th 1378, 1393, 191 Cal.Rptr.3d 551, 354 P.3d 346 (conc. opn. of Cuéllar, J.).)

The text of the statute is integral to our understanding of the statute's purpose. (See Robert L. v. Superior Court (2003) 30 Cal.4th 894, 901, 135 Cal.Rptr.2d 30, 69 P.3d 951 ( Robert L. ), quoting People v. Birkett (1999) 21 Cal.4th 226, 231, 87 Cal.Rptr.2d 205, 980 P.2d 912 [in interpreting a voter initiative, as with ordinary statutory construction, " 'we turn first to the language of the statute, giving the words their ordinary meaning' "].) We treat the text as an especially important indication of legislative purpose, typically even as the most reliable indicator of purpose. ( Murphy v. Kenneth Cole Productions, Inc. (2007) 40 Cal.4th 1094, 1103, 56 Cal.Rptr.3d 880, 155 P.3d 284 ; Hsu v. Abbara (1995) 9 Cal.4th 863, 871, 39 Cal.Rptr.2d 824, 891 P.2d 804.) We must take "the language ... as it was passed into law, and [we] must, if possible without doing violence to the language and spirit of the law, interpret it so as to harmonize and give effect to all its provisions." ( People v. Garcia (1999) 21 Cal.4th 1, 14, 87 Cal.Rptr.2d 114, 980 P.2d 829, citation omitted ( Garcia ).)

As the court suggests, we are indeed bound to give effect to the statute's underlying purposes. In doing so, we must consider the function of the specific statutory provisions as written. But giving effect to a provision's intended purpose requires us to consider not only the statutory design in its wider legal context, but also the purpose of a specific provision within the framework of that design. Everything about that inquiry in this case points in the same direction: "throughout this Code" means throughout this code.

Consider first the particular terms of the provision most directly at issue. By defining a phrase as it is "used throughout this Code" ( § 1170.18, subd. (c) ), the statute directs that the definition apply throughout the entire Penal Code, not just in resentencing proceedings under Proposition 47. (See People v. Bucchierre (1943) 57 Cal.App.2d 153, 166, 134 P.2d 505 ["The words 'as in this code provided' [citation] refer to the Penal Code"].) The court does not point to any part of the statute or initiative indicating that "throughout this Code" is a phrase carrying special meaning in Proposition 47.

In addition, we strive to give effect to all the words in a statute, avoiding surplusage whenever possible. (See City and County of San Francisco v. Farrell (1982) 32 Cal.3d 47, 54, 184 Cal.Rptr. 713, 648 P.2d 935 ["In construing the words of a statute or constitutional provision to discern its purpose, the provisions should be read together; an interpretation which would render terms surplusage should be avoided, and every word should be given some significance, leaving no part useless or devoid of meaning"].) The term "unreasonable risk of danger to public safety" appears in only one other provision in the entire Penal Code— section 1170.126, Proposition 36's analogous resentencing section. Hence, a refusal to apply section 1170.18's definition "throughout this Code" reduces that phrase to an inoperative appendage.

We may decline to give full effect to the ordinary meaning of a statute's language in certain circumstances. None of those circumstances is present in this case. For example, we have found judicial correction of language appropriate in instances of drafting error, where it has "appear[ed] clear that a word has been erroneously used." ( People v. Skinner (1985) 39 Cal.3d 765, 775, 217 Cal.Rptr. 685, 704 P.2d 752 ; see id. at pp. 775–779, 217 Cal.Rptr. 685, 704 P.2d 752 [construing statutory provision's use of the word "and" instead of "or" to be inadvertent].) But the court carefully refrains from asserting that a drafting error occurred in Proposition 47. Indeed, it acknowledges that at least one of Proposition 47's authors has publicly taken the position that the measure's definition of "unreasonable risk of danger" applies to resentencing proceedings under the Three Strikes Reform Act. (See maj. opn., ante , 220 Cal.Rptr.3d at p. 249, fn. 9, 397 P.3d at p. 952, fn. 9.)

Nor is there any good reason to believe the consequences of construing the phrase "throughout this Code" to mean throughout the Penal Code will prove absurd. On occasion, we have disregarded the literal meaning of a statutory provision when an "obvious absurdity" would otherwise result. ( In re Thierry S. (1977) 19 Cal.3d 727, 741, fn. 13, 139 Cal.Rptr. 708, 566 P.2d 610.) And it is certainly true that anticipating the consequences of a particular interpretation can help inform our understanding of whether an interpretation is consistent with the purpose of the statutory design. (See Horwich v. Superior Court (1999) 21 Cal.4th 272, 280, 87 Cal.Rptr.2d 222, 980 P.2d 927 ; id. at pp. 287–288, 87 Cal.Rptr.2d 222, 980 P.2d 927.) But the absurdity doctrine is to be invoked sparingly—only when we determine that "as a matter of law, the [enacting body] did not intend the statute to have its literal effect." ( Gorham Co., Inc. v. First Financial Ins. Co. (2006) 139 Cal.App.4th 1532, 1544, 44 Cal.Rptr.3d 197 ; see also California School Employees Assn. v. Governing Bd. of South Orange County Community College Dist. (2004) 124 Cal.App.4th 574, 588, 21 Cal.Rptr.3d 451 ["We must exercise caution using the 'absurd result' rule; otherwise, the judiciary risks acting as a ' "super-Legislature" ' by rewriting statutes to find an unexpressed legislative intent"].) In light of the striking similarities between the resentencing procedures of the two initiatives, the results of applying the definition consistently between Proposition 36 and Proposition 47 prove entirely reasonable. Once a petitioner's eligibility is established under either Proposition 36 or Proposition 47, the petitioner must be resentenced "unless the court, in its discretion, determines that resentencing the petitioner would pose an unreasonable risk of danger to public safety." (§§ 1170.126, subd. (f), 1170.18, subd. (b), italics added.) The sentencing court considers the exact same listed factors in making the required determination of dangerousness under either resentencing scheme. Moreover, the groups of petitioners eligible to apply for relief under Propositions 36 and 47 are not wholly distinct. If an inmate is currently serving a sentence for a third strike offense that also happens to be an enumerated offense in Proposition 47, that individual may petition for resentencing under either or both acts. Given these comparable and overlapping resentencing structures, there is nothing absurdabout finding that the same method for assessing the risk of dangerousness applies to petitioners under Proposition 36 and Proposition 47 alike.

What the court posits is that an amendment circumscribing the basis for denying relief under the Three Strikes Reform Act "necessarily would suggest disfavor" of the broader discretion afforded courts under Proposition 36's originally undefined phrase "unreasonable risk of danger to public safety." (Maj. opn., ante , 220 Cal.Rptr.3d at p. 253, 397 P.3d at p. 956.) Not so. The electorate was entitled to apply its later-approved, more focused standard of dangerousness uniformly throughout the Penal Code without first making any determination that Proposition 36's undefined standard had turned out to be problematic or unfavorable. Moreover, Proposition 47's focused standard preserves the sentencing court's authority, in proceedings under Proposition 36, to consider the petitioner's criminal history, the extent of injury to victims, the length of prior terms of incarceration, and any other relevant evidence when making that determination. (See § 1170.126, subd. (g).) The considerations that inform a petitioner's risk of committing a super strike in the future are of course not limited to the single question of whether a petitioner has previously been convicted of a super strike. And if the sentencing court, after considering all relevant evidence, were to find that the Proposition 36 petitioner does not pose an unreasonable risk of committing a super strike, the petitioner would not be immediately released into the general public. Instead, relief would mean a reduction of a life term for a nonviolent, nonserious felony to a doubled term of imprisonment. (See §§ 1170.126, subd. (b), 667, subd. (e)(1).) Thus, the court—correctly—does not argue that allowing Proposition 47's definition of "unreasonable risk of danger" to apply throughout the Penal Code would compel absurd results.

Finally, we have held that a literal construction of an enactment's language does not control "when such a construction would frustrate the manifest purpose of the enactment as a whole." ( Arias v. Superior Court (2009) 46 Cal.4th 969, 979, 95 Cal.Rptr.3d 588, 209 P.3d 923 ; accord, California School Employees Assn. v. Governing Board (1994) 8 Cal.4th 333, 340, 33 Cal.Rptr.2d 109, 878 P.2d 1321.) This last exception to general statutory interpretation principles appears to animate the court's view, insofar as the court concludes that that certain aspects of the initiative appear "inconsistent with" giving effect to the statutory text. (Maj. opn., ante , 220 Cal.Rptr.3d at pp. 244–245, 397 P.3d at pp. 947–949.) But, according to some of its stated purposes, Proposition 47 sought to assess the public safety risk of inmates and reduce corrections spending. Proposition 36's petition mechanism is also markedly similar to Proposition 47's in terms of its procedures and its identification of factors bearing on an inmate's risk of danger to public safety. The court has not gone so far as to find that Proposition 47's manifest purposes would be frustrated by applying "unreasonable risk" definition to the Three Strikes Reform Act, as our precedent has up until now required. Today's departure from our ordinary mode of statutory interpretation unjustifiably augments the judicial power to rewrite duly enacted statutes in accordance with courts' own preferences.³ III.

The court appears to concede that the phrase "[a]s used throughout this Code"—on its own—has a clear ordinary meaning. (See maj. opn., ante , 220 Cal.Rptr.3d at p. 242, 397 P.3d at p. 946; see also Marshall v. Pasadena Unified School Dist. (2004) 119 Cal.App.4th 1241, 1255, 15 Cal.Rptr.3d 344 ["There is nothing ambiguous about the phrase 'as used in this code.' In enacting [Public Contract Code] section 1102, the Legislature did not merely define the term 'emergency' for a particular chapter, article or division...."].) But the court would find the statutory language ambiguous when considered in the context of the entire statute and initiative.

The provisions of the initiative contain no reference to the Three Strikes law or the Three Strikes Reform Act, or to persons sentenced to life imprisonment. Such silence cannot be taken to resolve the issue, or else we would effectively impose a requirement that initiative provisions specify every possible consequence that would result from voter approval. (See People v. Romanowski (2017) 2 Cal.5th 903, 908–909, 215 Cal.Rptr.3d 758, 391 P.3d 633 ["We deny a phrase like 'any other provision of law' its proper impact if we expect a penal statute—whether enacted by the Legislature or the electorate—to further enumerate every provision of the Penal Code to which it is relevant"].) In harmonizing provisions of a ballot measure consistent with the spirit of the law, "it matters not whether the drafters, voters or legislators consciously considered all the effects and interrelationships of the provisions they wrote and enacted." ( Garcia, supra , 21 Cal.4th at p. 14, 87 Cal.Rptr.2d 114, 980 P.2d 829.)

The court does not show that applying Proposition 47's definition to the resentencing standard under the Three Strikes Reform Act would frustrate—or even, under its diluted standard, prove inconsistent with—the manifest purposes of Proposition 47. Portions of Proposition 47's uncodified sections demonstrate the initiative's focus on reducing certain nonserious and nonviolent felonies to misdemeanors, while "ensur[ing] that sentences for people convicted of dangerous crimes like rape, murder, and child molestation are not changed." (Voter Information Guide, Gen. Elec. (Nov. 4, 2014) (Voter Information Guide), text of Prop. 47, § 2, p. 70.) But application of the definition in accordance with the statute's literal text appears consistent with other stated purposes of the initiative. One of Proposition 47's uncodified purposes is to "[r]equire a thorough review of criminal history and risk assessment of any individuals before resentencing to ensure that they do not pose a risk to public safety." (Voter Information Guide, supra , text of Prop. 47, § 3, p. 70.) Given that Proposition 47 supplied a definition by which resentencing courts are to assess whether a petitioner poses an "unreasonable risk of danger to public safety," this purpose would be promoted if more inmates, such as Valencia and Chaney, were evaluated under the same forward-looking risk measurement standard as a precursor to any sentence reduction. Moreover, another one of Proposition 47's stated purposes is to "save significant state corrections dollars" and "increase investments in programs that reduce crime and improve public safety ... which will reduce future expenditures for corrections." (Voter Information Guide, supra , text of Prop. 47, § 3, p. 70.) This purpose would be advanced by allowing the "unreasonable risk of danger to public safety" definition to apply throughout the code, because by reducing state corrections spending on some greater increment of inmates who are no longer deemed unreasonably dangerous, limited state resources could be more efficiently spent to assure continued confinement for unreasonably dangerous criminals.

The majority and concurrence also find ambiguity in the language of the initiative because section 1170.18, subdivision (c) refers to an unreasonable risk that "the petitioner"—rather than "a petitioner" or "any petitioner"—will commit a new enumerated felony. (Italics added; see maj. opn., ante , 220 Cal.Rptr.3d at pp. 242–243, 397 P.3d at pp. 946–947; conc. opn., ante , at pp. 258–260, 397 P.3d at pp. 960–961.) Both those opinions contend that the use of the definite article "the" offers support for the position that Proposition 47's definition applies only to individuals applying for relief under Proposition 47.

Even if we assumed that the term "the petitioner" in the preceding provision of section 1170.18, subdivision (b) denotes a person who filed his or her petition under subdivision (a)—that is, under Proposition 47—that phrase could very well adopt a different meaning within the immediately surrounding context of subdivision (c). If the phrase, "[a]s used throughout this Code," is to have any significance, it must encompass the only other instance in which the phrase is used in the Penal Code, which is in the Three Strikes Reform Act. Nothing about the use of the phrase "the petitioner" deprives the provision of its literal meaning. Section 1170.18, subdivision (c) relates to the petitioner whose opportunity for relief is affected by an assessment of his or her "unreasonable risk of danger to public safety." Thus, the phrase "the petitioner" plainly has relevance elsewhere in the Penal Code, specifically in the Three Strikes Reform Act. Those words, "[a]s used throughout this Code," would serve an unintelligible function if section 1170.18, subdivision (c)'s definition were to pertain only to the Proposition 47 petitioner. There is no indication that twisting the language "[a]s used throughout this Code" to somehow mean "as applicable only to the resentencing proceedings that are authorized under Proposition 47" advances the goals of the statute. (Compare § 1170.18, subd. (c) with maj. opn., ante , 220 Cal.Rptr.3d at p. 255, 397 P.3d at p. 957.)

The court appears to have found ambiguity not because it has found statements in the rest of Proposition 47 directly undermining the statute's clear language, but instead because it has not found redundant statements bolstering this language. The court posits that "one would expect [the Proposition 47] drafters to have mentioned or referred to a purpose and intention" to amend the Three Strikes Reform Act in the initiative's preamble. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 253, 397 P.3d at pp. 955–956.) But why would this be so? The amendment was sufficiently conveyed through the text of the law distributed to and approved by voters.

Nor should a lack of procedural detail in the statute bear on how we construe an unambiguous term. A prospective-only application of Proposition 47's definition to the Three Strikes Reform Act would mean the new definition would pertain to any petitioner who has not yet had a court decide whether he or she presents "an unreasonable risk of danger to public safety" as of Proposition 47's effective date. The court believes that Proposition 47's failure to include procedures for resentencing Three Strikes Reform Act petitioners under the new definition "widens the gap concerning the voters' understanding" of Proposition 47's effects. (See maj. opn., ante , 220 Cal.Rptr.3d at p. 248, 397 P.3d at p. 951.) But Proposition 36's two-year deadline for filing a petition may be extended "upon a showing of good cause." ( § 1170.126, subd. (b).) Nothing in the statute expressly bars Chaney, who is currently serving a third strike sentence, from filing a second petition. Thus, the question arises: If Proposition 47's definition of "unreasonable risk" were to govern proceedings under section 1170.126, would the new definition give Chaney "good cause" to file a second petition? Unfortunately for Chaney and other similarly situated prisoners, the court's decision forecloses relief at an earlier point in the analysis. While it is true that Proposition 47 itself does not specify whether a new definition constitutes good cause for a belated or second petition, such a question would require only judicial resolution. The court does not identify any instance where a lack of procedural detail stopped us from giving effect to the plain meaning of the substantive provisions of a statute.

Under today's holding, the clarity and meaning of text would be called into question whenever any statute refers to an application of a provision "throughout this Code." The phrase itself is straightforward, and remains so even when considering the definite article preceding the word "petitioner." ( § 1170.18, subd. (b).) But the court's opinion implicitly imposes an ersatz clear statement rule whenever any statute seeks to modify a previous one, and then elevates it to become a clearer statement requirement that voters and legislators confirm—through additional, redundant provisions and declarations of purpose—that they do indeed mean what they say, in order for their enactments to be given effect.

IV.

Having detected what it perceives to be ambiguity in the language of the statute, the court proceeds to examine evidence of voter "intent" outside the measure's provisions. Ballot materials extrinsic to the initiative may be relevant in the interpretation of statutes forged through the initiative process. But this does not mean that the ballot materials' failure to expressly mention one particular consequence can contravene relatively simple language. Official ballot materials are not detailed legal memoranda, nor can we command they discuss every nuance or legal issue an initiative may touch. ( Santa Clara County Local TransportationAuthority v. Guardino (1995) 11 Cal.4th 220, 237, 45 Cal.Rptr.2d 207, 902 P.2d 225 ["Ballot arguments are not legal briefs and are not expected to cite every case the proposition may affect"]; see also Day v. City of Fontana (2001) 25 Cal.4th 268, 282, 105 Cal.Rptr.2d 457, 19 P.3d 1196 ["it is of no consequence here that the ballot materials did not specifically refer to the act's application in actions against local public entities for nuisance and dangerous condition of property" because "[s]uch actions fall squarely within the terms of [Civil Code] section 3333.4, and the statute's operation in such cases promotes rather than defeats the declared purpose of Proposition 213 to restore balance to the justice system with respect to violators of the financial responsibility law"].)

Here, a total of seven brief bullet points constitute Proposition 47's Official Title and Summary. This abbreviated statement prepared by the Attorney General may serve as a useful guide to voters, but it by no means replaces the statute itself. In Cervera , we considered whether the original Three Strikes law authorized a defendant with three strikes to receive prison conduct credits under article 2.5 of the Penal Code—which by its terms only applied to determinate sentences—for use against his mandatory indeterminate term of life imprisonment. ( Cervera , supra , 24 Cal.4th at p. 1076, 103 Cal.Rptr.2d 762, 16 P.3d 176.) Although some of the "documents within the history of the bill and the initiative measure that would each become the Three Strikes law" assumed that prison conduct credits could apply to third-strikers ( id. at p. 1079, 103 Cal.Rptr.2d 762, 16 P.3d 176 ), we concluded that the statute's language did not authorize such reductions of indeterminate life sentences. "A statute, of course, must prevail over any summary. Were it not so, no statute could ever be enacted whole and entire. For every summary, by definition, is incomplete.... The summary must yield to the statute, not the statute to the summary." ( Id. at pp. 1079–1080, 103 Cal.Rptr.2d 762, 16 P.3d 176.)Ballot materials are not and cannot be dispositive in discerning an initiative's intended purpose, in particular where—as here—the statute's text, read in context, adequately communicates a provision's meaning and intended purpose. In Delaney , we acknowledged that "repeated references in the [ballot] argument to confidentiality and the like permit the inference the proponents of the measure intended to protect only confidential information." ( Delaney v. Superior Court (1990) 50 Cal.3d 785, 802, 268 Cal.Rptr. 753, 789 P.2d 934.) But we declined to conclude that "by emphasizing one purpose, perhaps the primary purpose of the measure, the argument misled voters into thinking confidentiality was the only purpose, especially when the measure itself made clear that all unpublished information would be protected." ( Id. at pp. 802–803, 268 Cal.Rptr. 753, 789 P.2d 934 ; see also id. at p. 803, 268 Cal.Rptr. 753, 789 P.2d 934 ["It would be a strained approach to constitutional analysis if we were to give more weight to a possible inference in an extrinsic source (a ballot argument) than to a clear statement in the Constitution itself"].)

Moreover, just as with the language of the uncodified initiative, there are several statements in the ballot materials consistent with an application of Proposition 47's definition to the Three Strikes Reform Act. The "Argument in Favor of Proposition 47" disclosed that the initiative would "[i]mprove public safety," "[r]educe prison spending and government waste," and "focus [ ] law enforcement dollars on violent and serious crime." (Voter Information Guide, supra , argument in favor of Prop. 47, p. 38.) These declared purposes of reallocating prison spending to the most dangerous criminals and allowing others to pursue resentencing would be advanced by permitting Three Strikes Reform Act petitioners to demonstrate that—under the same forward-looking risk assessment employed under Proposition 47—they do not pose an unreasonable risk of danger to public safety.

Opponents of the initiative also argued that "[t]he proponents of this dangerous measure have already admitted that Proposition 47 will make 10,000 felons eligible for early release. According to independent analysis, many of those 10,000 felons have violent criminal histories " and "Felons with prior convictions for armed robbery, kidnapping, carjacking, child abuse, residential burglary, arson, assault with a deadly weapon, and many other serious crimes will be eligible for early release under Prop. 47." (Voter Information Guide, supra , argument against Prop. 47, p. 39.) The rebuttal to the opponents' argument clarified that Proposition 47 would not require automatic release of anyone. (Voter Information Guide, at p. 39.) Yet the voters enacted Proposition 47 despite cautionary statements that persons with violent criminal histories could benefit from the passage of the initiative. (Cf. Robert L. , supra , 30 Cal.4th at pp. 906–907, 135 Cal.Rptr.2d 30, 69 P.3d 951 [voters' approval of initiative despite warnings contained in opponents' argument was indicative of voter intent].)We afford the Legislative Analyst's review of an initiative the same interpretive weight as other ballot materials not enacted into law. Comments of the Legislative Analyst, like other ballot statements presented to the electorate, " 'may be helpful but are not conclusive in determining the probable meaning of initiative language.' " ( San Francisco Taxpayers Assn. v. Board of Supervisors (1992) 2 Cal.4th 571, 580, 7 Cal.Rptr.2d 245, 828 P.2d 147 ; see also Carman v. Alvord (1982) 31 Cal.3d 318, 331, 182 Cal.Rptr. 506, 644 P.2d 192 ["We should not assume that [the Legislative Analyst's] brief comments accurately reflected the full intent of the drafters or the understanding of the electorate"].) For Proposition 47, the Legislative Analyst estimated the fiscal effects of reducing felonies to misdemeanors and did not expressly mention the Three Strikes Reform Act. It is possible the Legislative Analyst did not apprehend the significance of the statutory phrase "[a]s used throughout this Code." Or perhaps the analyst omitted reference to Proposition 36 in deciding to relay, within a limited space, what he ascertained to be the most prominent aspects of Proposition 47. Incompleteness of the analysis does not necessarily render an absent consequence incongruous with the statutory text. After all, the analyst never wrote that he understood Proposition 47's definition of "unreasonable risk" to not apply to resentencing Three Strikes Reform Act. Even assuming arguendo there existed a contradiction between the Legislative Analyst's statements and an application of Proposition 47's dangerousness risk assessment definition throughout the Penal Code, we have previously rejected the notion that extrinsic materials—even the Legislative Analyst's analysis—would trump plain language of the initiative. ( San Francisco Taxpayers , at p. 580, 7 Cal.Rptr.2d 245, 828 P.2d 147 ["The case for rejecting the Legislative Analyst's views is even more compelling here, where the contradiction is in the language of the initiative"].)

Both the majority opinion and the concurrence appear to use the ballot statements made by the Attorney General and the Legislative Analyst as a reference point for voter awareness and understanding. (See, e.g., maj. opn., ante , 220 Cal.Rptr.3d at p. 239, 397 P.3d at p. 944 [expressing skepticism at notion that voters may have "greater acumen than the legal professionals"]; conc. opn., ante , at p. 262, 397 P.3d at p. 963 ["When these trained experts were apparently unaware of the initiative's possible effect ... it becomes more difficult to conclude that voters understood" the consequences of their affirmative vote].) Such an approach downplays the fact that the ballot materials authored by the Attorney General and Legislative Analyst are indisputably incomplete summaries of the ballot measure, constrained by word limits. An even more troubling feature of the court's argument is that it suggests that a few "legal professionals" or "trained experts" hold ultimate authority to decide (advertently or not) which portions of the people's enactments are ratified. Under today's holding, pronouncements of individual officers can defeat the actual language of laws approved by the electorate. V.

To the extent that the court's decision is calibrated to apply only in the initiative context,it departs from our longstanding commitment to analyze voter-approved statutes using the same approach we employ for Legislature-enacted statutes. Lawmaking by ballot initiative is so fundamental to California's democracy that our state Constitution speaks of the power of initiative "not as a right granted the people, but as a power reserved by them." ( Associated Home Builders etc., Inc. v. City of Livermore (1976) 18 Cal.3d 582, 591, 135 Cal.Rptr. 41, 557 P.2d 473 ; Cal. Const., art. IV, § 1.) When voters opt to wield the legislative power, they step into the shoes of legislators. Thus, traditionally, judicial interpretation of ballot initiatives has been "governed by the same rules that apply in construing a statute enacted by the Legislature." ( People v. Park (2013) 56 Cal.4th 782, 796, 156 Cal.Rptr.3d 307, 299 P.3d 1263 ; accord, People v. Johnson (2015) 61 Cal.4th 674, 682, 189 Cal.Rptr.3d 794, 352 P.3d 366.) Voter enactments are to be afforded legitimacy commensurate to that which is furnished to enactments by our Legislature, and our respect for the doctrine of separation of powers should not wane when we encounter a voter-enacted statute.We generally presume voters " 'have voted intelligently upon an amendment to their organic law, the whole text of which was supplied [to] each of them prior to the election and which they must be assumed to have duly considered.' " ( Amador Valley Joint Union High Sch. Dist. v. State Bd. of Equalization (1978) 22 Cal.3d 208, 243–244, 149 Cal.Rptr. 239, 583 P.2d 1281, quoting Wright v. Jordan (1923) 192 Cal. 704, 713, 221 P. 915.) This presumption, we have acknowledged, serves "to further the fundamental right of the electorate to enact legislation through the initiative process." ( Taxpayers to Limit Campaign Spending v. Fair Pol. Practices Com. (1990) 51 Cal.3d 744, 768, 274 Cal.Rptr. 787, 799 P.2d 1220.) Although the court identifies Taxpayers as one case in which we recognized some limit to the degree of thoroughness with which we presume voters review ballot measures, Taxpayers is a far cry from the case before us. There, the voters simultaneously approved two initiatives with "fundamentally conflicting provisions intended to regulate the same subject." ( Ibid . ) Provision-by-provision reconciliation of nonconflicting aspects of those initiatives was riddled with uncertainty, as "illustrated by the divergent conclusions" reached by lower courts and administrative agencies. ( Id. at p. 760, 274 Cal.Rptr. 787, 799 P.2d 1220.) In contrast, in the instant case, we are tasked with assessing whether reasonable voters plausibly meant to give effect to a ballot measure's clear language, thus affecting a law previously enacted by ballot measure. These unremarkable circumstances do not warrant a departure from our standard presumption that the electorate votes intelligently after considering the provided text of the law.As another baseline presumption to ground our interpretations, we presuppose that the voters, in adopting an initiative, did so being "aware of existing laws at the time the initiative was enacted." ( Professional Engineers in California Government v. Kempton (2007) 40 Cal.4th 1016, 1048, 56 Cal.Rptr.3d 814, 155 P.3d 226 ; see also People v. Shabazz (2006) 38 Cal.4th 55, 65, fn. 8, 40 Cal.Rptr.3d 750, 130 P.3d 519 [rejecting argument that "voters were unlikely to have considered the doctrine of transferred intent" because "it is well-settled that voters ' "are presumed to know the law" ' "]; In re Lance W. (1985) 37 Cal.3d 873, 890, fn. 11, 210 Cal.Rptr. 631, 694 P.2d 744 ["the electorate would be deemed to know of the superseding impact of federal constitutional provisions on state laws or constitutional provisions which conflict with and restrict rights guaranteed by the United States Constitution" because "[t]he adopting body is presumed to be aware of existing laws and judicial construction thereof"]; Penziner v. West American Finance Co. (1937) 10 Cal.2d 160, 174, 74 P.2d 252 ["It must be presumed that the Legislature in proposing and the electorate in adopting the constitutional amendment acted with full knowledge of the existence of the prior statute relating to the same general subject"].) It is not untenable at all—and indeed, more consistent with our previous case law on statutory construction—to presume that reasonable voters were aware of the Three Strikes Reform Act when they voted to approve Proposition 47 two years later.

Contrary to the court's assertion, we did not "refuse[ ] to apply" this presumption in Robert L. (Maj. opn., ante , 220 Cal.Rptr.3d at p. 252, 397 P.3d at p. 955.) What we did was interpret the term at issue in the context of the entire statute. We observed that the ordinary meaning of the phrase "a public offense punishable as a felony or a misdemeanor" would not have been restricted to wobbler offenses, and identified discrepancies between that phrase and the statutory description of so- called wobbler offenses. ( Robert L. , supra , 30 Cal.4th at p. 901, 135 Cal.Rptr.2d 30, 69 P.3d 951.) Our conclusion was also supported by the fact that the term wobbler was, at that time, "a legal term of art of recent vintage" with no "meaning defined by statute or commonly understood by the electorate." ( Id. at p. 902, 135 Cal.Rptr.2d 30, 69 P.3d 951.) Our opinion in Robert L. does not signify that we disregard the presumption that voters are aware of existing law; it means that we cannot assume that every term is unambiguous in its meaning. We must acknowledge when statutory language is reasonably susceptible to multiple interpretations. Yet, in attempting to justify their interpretation in this case, the court does not contend that the word "code" is commonly understood by the electorate to refer to only the initiative containing the provision at issue. Its decision to strike the phrase "throughout this Code" from section 1170.18, subdivision (c) and limit the effectiveness of the provision's definition to resentencing proceedings authorized only under Proposition 47 is all the more indefensible.At times, both the majority opinion and the concurrence appear to imply that courts should ascertain the subjective, exact thoughts of the voters casting ballots for the initiative—presumably the electoral majority who prevailed at the ballot box. It is undisputed that the statutory language of the Safe Neighborhoods and Schools Act was provided to every individual who cast a vote for Proposition 47. Yet the majority opinion finds the phrase "[a]s used throughout this Code" was insufficient to provide "notice" to the voters as to the scope of the definition at issue. (See maj. opn., ante , 220 Cal.Rptr.3d at p. 254, 397 P.3d at p. 956.) The concurrence similarly concludes that the question of whether to apply Proposition 47's definition throughout the Penal Code was "not presented" to voters who were asked to approve Proposition 47. (Conc. opn., ante , at p. 263, 397 P.3d at p. 964.) In relying on these rationales, these opinions essentially demand an extra degree of assurance—beyond what we can glean from any reasonable understanding of the statutory text, the statute's structure, or its context—that the electorate possessed actual awareness of the meaning of language contained in section 1170.18, subdivision (c). Recall how firmly settled is our responsibility to interpret legislation consistently, whether enacted by voters or by their representatives in the Legislature. (See, e.g., Lesher Communications, Inc. v. City of Walnut Creek (1990) 52 Cal.3d 531, 540, 277 Cal.Rptr. 1, 802 P.2d 317, quoting Legislature v. Deukmejian (1983) 34 Cal.3d 658, 675, 194 Cal.Rptr. 781, 669 P.2d 17 [statutes enacted by initiative " 'are subject to the same constitutional limitations and rules of construction as are other statutes' "].) Democracy involves both lofty aspirations and messier realities. In the absence of an unambiguous reason to require an extraneous statement of purpose in either this context or in lawmaking generally, we give effect to statutory provisions in question. Nothing about Proposition 47 or its purposes calls for a plainer statement than "[a]s used throughout this Code" for one of its terms to apply "throughout this Code."

Although we analyze text, structure, and context to give effect to a statutory provision's intended purpose or purposes as reasonable voters or legislators might have understood them, we are hardly in a position to ascertain the subjective intentions of each member of an enacting majority. When we analyze a statute, a particularly compelling mix of humility and analytical clarity should prevent us from implying that we are articulating legislators' or voters' subjective states of mind. ( Garcia , supra , 21 Cal.4th at p. 14, fn. 8, 87 Cal.Rptr.2d 114, 980 P.2d 829. ["We do not say anything regarding the 'probability' any or all legislators or voters subjectively contemplated a given meaning; we simply do not know."] ) To some extent, we must attribute constructive notice on the part of the enacting body in order to properly give effect to the purposes of any enacted law. We do so because, in construing statutes, our task is not to ask whether the enacting body "consciously considered all the effects and interrelationships of the provisions" but is instead to take the language of the law and "if possible without doing violence to the language and spirit of the law, interpret it so as to harmonize and give effect to all its provisions." ( Id. at p. 14, 87 Cal.Rptr.2d 114, 980 P.2d 829.)

It is one of democracy's recurring challenges that the rules and standards governing society—whether enacted by legislation or initiative—are often enormously complicated in both their content and effect. ( Fair Political Practices Com. v. Superior Court (1979) 25 Cal.3d 33, 42, 157 Cal.Rptr. 855, 599 P.2d 46 ["Our society being complex, the rules governing it whether adopted by legislation or initiative will necessarily be complex"].) Whether or not the complexity of any individual provision is fully understood by a voter casting a ballot on an initiative, we "should not lightly presume" that voters are unaware of what they are doing in approving an initiative—even when the contents of a ballot measure are intricate and potentially confusing. ( Brosnahan v. Brown (1982) 32 Cal.3d 236, 252, 186 Cal.Rptr. 30, 651 P.2d 274.) The tenets we typically apply to voter-enacted statutes are rooted in our respect for the initiative process and, more generally, the legislative function. Thus, they start from the premise that the electorate must take a considerable measure of responsibility in enacting complicated laws. The court has not offered a convincing reason for us to here dispense with these fundamental interpretive canons.

It is a well-settled principle of statutory interpretation that, where both text and purpose are clear, courts shall not endeavor to rewrite language of a ballot measure. (See People v. Skinner (1985) 39 Cal.3d 765, 775, 217 Cal.Rptr. 685, 704 P.2d 752 ; see also Ross v. RagingWire Telecommunications, Inc. (2008) 42 Cal.4th 920, 930, 70 Cal.Rptr.3d 382, 174 P.3d 200 ["the initiative power is strongest when courts give effect to the voters' formally expressed intent"].) Here, the court contravenes this principle by attempting to discern the purpose of the voters' enactment without consideration of a relevant provision's clear text as a crucial indicator of purpose.

The court sees in section 1170.18, subdivision (c)'s terms—"[a]s used throughout this Code"—not a clear, expansive indication that Proposition 47's definition of "unreasonable risk of danger to public safety" applies across the Penal Code as the provision's literal terms indicate, but the opposite: a quirky limitation which restricts the "unreasonable risk" definition from applying "throughout this Code." None of the court's arguments to justify this remarkable conclusion strike me as persuasive. This would be a more difficult case if there were any compelling reason to doubt that Proposition 36 and Proposition 47 share fundamental, common purposes. But the purposes of both voter initiatives are advanced by giving effect to Proposition 47's explicit terms, which provide that the definition of "unreasonable risk of danger to public safety" applies "throughout this Code"— rather than only throughout proceedings authorized under Proposition 47 or, effectively, "throughout this Code, except the Three Strikes Reform Act." With respect, I must dissent.We Concur:

Werdegar, J.

Liu, J.