People v. Burgos, 16 Cal.5th 1

Opinion of the Court by Guerrero, C. J.

In 2021, the Legislature passed Assembly Bill No. 333 (2021–2022 Reg. Sess.) (Assembly Bill 333), known as the STEP Forward Act of 2021. (Stats. 2021, ch. 699, § 1.) Assembly Bill 333 amended Penal Code section 186.22 by imposing new substantive requirements relating to gang enhancements and the criminal offense of gang participation.¹ (Stats. 2021, ch. 699, § 4.) Assembly Bill 333 also added section 1109, which provides that, if requested by the defense, a trial court must try a gang enhancement charge separately from the underlying offense. (§ 1109, subd. (a); Stats. 2021, ch. 699, § 5.) The statute likewise provides that gang-participation offenses must be tried separately from all other counts that do not require gang evidence as an element of the crime. (§ 1109, subd. (b); Stats. 2021, ch. 699, § 5.)

We granted review to determine whether section 1109’s provisions governing bifurcation apply retroactively to cases in which the judgment is not yet final. This question has divided the Courts of Appeal. (Compare People v. Burgos (2022) 77 Cal.App.5th 550, 565–568 (Burgos) [concluding that § 1109 applies retroactively], People v. Montano (2022) 80 Cal.App.5th 82, 105–108 (Montano) [same], People v. Ramos (2022) 77 Cal.App.5th 1116, 1128–1131 (Ramos) [same] with People v. Boukes (2022) 83 Cal.App.5th 937, 946–949 [concluding § 1109 does not apply retroactively], review granted Dec. 14, 2022, S277103 (Boukes), People v. Perez (2022) 78 Cal.App.5th 192, 207 [same], review granted Aug. 17, 2022, S275090 (Perez), People v. Ramirez (2022) 79 Cal.App.5th 48, 65 [same], review granted Oct. 12, 2022, S275341 (Ramirez).)

Ordinarily, statutes are presumed to apply only prospectively, unless the Legislature expressly declares otherwise. This well-settled principle is codified at section 3 of the Penal Code and appears in other codes as well. (See, e.g., Code Civ. Proc., § 3; Civ. Code, § 3.) In In re Estrada (1965) 63 Cal.2d 740 (Estrada), however, we held that an amendment to a statute that lessened punishment for a crime gave rise to an inference of contrary legislative intent; that is, that the Legislature must have intended that the amendment mitigating punishment would apply retroactively to every case to which it constitutionally could apply. (Id. at p. 745.)

We conclude that the Estrada inference of retroactivity does not extend to section 1109. Estrada’s principle of statutory interpretation infers retroactivity from statutory reductions in punishment, not from the type of prophylactic rules of criminal procedure embodied in section 1109’s bifurcation provisions. Accordingly, the general presumption of prospective-only application applies to section 1109. The equal protection clauses of the federal and state Constitutions (U.S. Const., 14th Amend.; Cal. Const., art. I, § 7, subd. (a)) do not compel a different result. Because the Court of Appeal concluded that section 1109 applies retroactively, we reverse its judgment.

I. FACTUAL AND PROCEDURAL BACKGROUND

Around midnight on August 29, 2015, Francisco Burgos, Damon Stevenson, Jr., James Richardson, Derrik Lozano, and Gregory Byrd approached Gabriel Cortez and Danny Rodriguez near a convenience store in San Jose. A member of the group asked Cortez and Rodriguez where they were from, whether they “banged,” and if they were from “Meadowfair,” a criminal street gang. After Cortez and Rodriguez responded that they were from “right here,” a member of the group responded, “Well, we’re Crip[s].” The group proceeded to rob Cortez and Rodriguez at gunpoint and threatened to shoot them unless they left immediately. Cortez and Rodriguez ran to Cortez’s home.

The People filed an information charging Burgos, Stevenson, Richardson, Lozano, and Byrd with two counts of second degree robbery. (§§ 211, 212.5, subd. (c).) Each count also included gang enhancement and firearm enhancement allegations. (§§ 186.22, subd. (b)(1)(C), 12022.53, subds. (b) & (e)(1).) Burgos, Stevenson, and Richardson (collectively, defendants) moved to bifurcate trial on the gang enhancements.² The trial court denied the requests. At trial, the prosecution presented evidence that the robberies were committed in association with a criminal street gang.

A jury found defendants guilty of two counts of second degree robbery and also found true the gang enhancement allegations. The trial court found true prior conviction allegations as to Burgos and Stevenson, and Richardson admitted his prior conviction. The court sentenced each defendant to 21 years of imprisonment.

While defendants’ appeals were pending, the Legislature passed Assembly Bill 333, which became effective January 1, 2022. (See People v. Tran (2022) 13 Cal.5th 1169, 1206 (Tran).) Assembly Bill 333 amended section 186.22 by narrowing the definitions of “ ‘criminal street gang,’ ” “common benefit,” and “ ‘pattern of criminal gang activity,’ ” and adding the requirement that gang members “collectively engage in” the “pattern of criminal gang activity.” (§ 186.22, subds. (e)(1)–(2), (f) & (g).) These changes “ha[d] the effect of ‘increas[ing] the threshold for conviction of the section 186.22 offense and the imposition of the enhancement.’ ” (Tran, at p. 1207.) Assembly Bill 333 also “added section 1109, which requires, if requested by the defendant, a gang enhancement charge to be tried separately from all other counts that do not otherwise require gang evidence as an element of the crime.” (Tran, at p. 1206.)

The Attorney General conceded that the amendments to section 186.22 apply retroactively, and that the true findings on defendants’ gang enhancement allegations must be vacated. (Burgos, supra, 77 Cal.App.5th at p. 563; cf. Tran, supra, 13 Cal.5th at p. 1207 [Assembly Bill 333’s substantive changes apply retroactively to all nonfinal cases under Estrada].) However, the Attorney General maintained that section 1109’s bifurcation provisions apply only prospectively. (Burgos, at p. 564.)

A divided Court of Appeal held that section 1109 also applies retroactively. (Burgos, supra, 77 Cal.App.5th at pp. 564–569.) The majority concluded that the statute meets the criteria for retroactive application under the reasoning of Estrada, supra, 63 Cal.2d 740, because it “reduces the possible punishment for a class of persons [and] provides a possible ameliorating benefit to a class of criminal defendants.” (Burgos, at p. 565.)

We granted review, limited to the issue of whether section 1109’s bifurcation provisions apply retroactively to cases not yet final.

II. DISCUSSION

Defendants contend that section 1109 applies retroactively, not just prospectively. Burgos also claims that applying section 1109 only prospectively would violate equal protection principles. We find both arguments unconvincing.

A. Section 1109’s Bifurcation Provisions Do Not Apply Retroactively

1. Language and Legislative History of Section 1109

Prior to the enactment of section 1109, trial courts were authorized, in their discretion, to bifurcate trials so that a gang enhancement allegation would be tried separately from a charged offense, when appropriate to avoid undue prejudice to the defense. (People v. Hernandez (2004) 33 Cal.4th 1040, 1049 (Hernandez).) This authority came from “ ‘section 1044, which vests the trial court with broad discretion to control the conduct of a criminal trial.’ ” (Id. at p. 1048; see § 1044 [“It shall be the duty of the judge to control all proceedings during the trial”].) Section 1109 now requires the trial court to bifurcate such trials if the defense so requests.

Section 1109 provides in relevant part: “(a) If requested by the defense, a case in which a gang enhancement is charged . . . shall be tried in separate phases as follows: [¶] (1) The question of the defendant’s guilt of the underlying offense shall be first determined. [¶] (2) If the defendant is found guilty of the underlying offense and there is an allegation of a[] [gang] enhancement . . . , there shall be further proceedings to the trier of fact on the question of the truth of the enhancement. . . . [¶] (b) If a defendant is charged with a violation of subdivision (a) of Section 186.22, this count shall be tried separately from all other counts that do not otherwise require gang evidence as an element of the crime. This charge may be tried in the same proceeding with an allegation of an enhancement under subdivision (b) or (d) of Section 186.22.”

In enacting Assembly Bill 333, the Legislature made several findings and declarations related to section 1109’s bifurcation provisions. The Legislature declared that “[g]ang enhancement evidence can be unreliable and prejudicial to a jury because it is lumped into evidence of the underlying charges which further perpetuates unfair prejudice in juries and convictions of innocent people.” (Stats. 2021, ch. 699, § 2, subd. (d)(6).) The Legislature further declared that “[s]tudies suggest that allowing a jury to hear the kind of evidence that supports a gang enhancement before it has decided whether the defendant is guilty or not may lead to wrongful convictions” and “[t]he mere specter of gang enhancements pressures defendants to accept unfavorable plea deals rather than risk a trial filled with prejudicial evidence and a substantially longer sentence.” (Id., subd. (e).) The Legislature resolved that “[b]ifurcation of trials where gang evidence is alleged can help reduce its harmful and prejudicial impact.” (Id., subd. (f).)

2. Section 3 and the Presumption of Prospective Application

The question of whether a statute applies retroactively begins with section 3 of the Penal Code. Enacted in 1872, this provision specifies that no part of the Penal Code “is retroactive, unless expressly so declared.” (§ 3.)

“ ‘[T]he language of section 3 erects a strong presumption of prospective operation, codifying the principle that, “in the absence of an express retroactivity provision, a statute will not be applied retroactively unless it is very clear from extrinsic sources that the [lawmakers] . . . must have intended a retroactive application.” [Citations.] Accordingly, “ ‘a statute that is ambiguous with respect to retroactive application is construed . . . to be unambiguously prospective.’ ” ’ ” (People v. Buycks (2018) 5 Cal.5th 857, 880 (Buycks); see People v. Brown (2012) 54 Cal.4th 314, 324 (Brown) [same]; People v. Floyd (2003) 31 Cal.4th 179, 184 (Floyd) [§ 3 “embodies the general rule of statutory construction that ‘when there is nothing to indicate a contrary intent in a statute it will be presumed that the Legislature intended the statute to operate prospectively and not retroactively’ ”]; Tapia v. Superior Court (1991) 53 Cal.3d 282, 287 (Tapia) [“It is well settled that a new statute is presumed to operate prospectively absent an express declaration of retrospectivity or a clear indication that the electorate, or the Legislature, intended otherwise”]; People v. Hayes (1989) 49 Cal.3d 1260, 1274 (Hayes) [same]; cf. Evangelatos v. Superior Court (1988) 44 Cal.3d 1188, 1207–1208 (Evangelatos) [“Like similar provisions found in many other codes,” including Pen. Code, § 3, Civ. Code, § 3 “reflects the common understanding that legislative provisions are presumed to operate prospectively, and that they should be so interpreted ‘unless express language or clear and unavoidable implication negatives the presumption’ ”]; cf. Aetna Cas. & Surety Co. v. Ind. Acc. Com. (1947) 30 Cal.2d 388, 393 (Aetna) [“It is an established canon of interpretation that statutes are not to be given a retrospective operation unless it is clearly made to appear that such was the legislative intent”].)

However, the presumption that a statute will apply only prospectively “is a canon of statutory interpretation rather than a constitutional mandate. [Citation.] Accordingly, ‘the Legislature can ordinarily enact laws that apply retroactively, either explicitly or by implication.’ ” (People v. Frahs (2020) 9 Cal.5th 618, 627 (Frahs).) “In applying this principle, we have been cautious not to infer retroactive intent from vague phrases and broad, general language in statutes.” (Brown, supra, 54 Cal.4th at p. 319; see also Aetna, supra, 30 Cal.2d at p. 396 [“it must be assumed that the Legislature was acquainted with the settled rules of statutory interpretation, and that it would have expressly provided for retrospective operation of the amendment if it had so intended”].)

3. The Estrada Exception to Prospective Application of Statutes

Although section 3’s presumption of prospectivity ordinarily applies if a statute does not contain an express statement of retroactivity or otherwise clearly and unavoidably indicate that the Legislature intended the statute to operate retroactively, this court has recognized an additional, related exception to this rule.

In Estrada, we held that an amendment to a statute that reduces the punishment for a particular criminal offense gives rise to an inference that the statute applies retroactively to all judgments not yet final on appeal. (Estrada, supra, 63 Cal.2d at p. 745.) We explained our reasoning as follows: “There is one consideration of paramount importance. It leads inevitably to the conclusion that the Legislature must have intended, and by necessary implication provided, that the amendatory statute should prevail. When the Legislature amends a statute so as to lessen the punishment it has obviously expressly determined that its former penalty was too severe and that a lighter punishment is proper as punishment for the commission of the prohibited act. It 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 amendatory act imposing the lighter punishment can be applied constitutionally to acts committed before its passage provided the judgment convicting the defendant of the act is not final. This intent seems obvious, because to hold otherwise would be to conclude that the Legislature was motivated by a desire for vengeance, a conclusion not permitted in view of modern theories of penology.” (Id. at pp. 744–745.)

At the same time, we reiterated in Estrada that section 3’s general rule of prospectivity still applied to an amended statute unless there were “other factors that indicate the Legislature must have intended” that it would operate retroactively. (Estrada, supra, 63 Cal.2d at p. 746.) Accordingly, Estrada stands for the principle that, “ ‘where the amendatory statute mitigates punishment and there is no saving clause, . . . the amendment will operate retroactively so that the lighter punishment is imposed.’ ” (People v. Nasalga (1996) 12 Cal.4th 784, 792 (plur. opn. of Werdegar, J.) (Nasalga), quoting Estrada, at p. 748; cf. People v. Gentile (2020) 10 Cal.5th 830, 852 [under Estrada, “Newly enacted legislation lessening criminal punishment or reducing criminal liability presumptively applies to all cases not yet final on appeal at the time of the legislation’s effective date” unless it “sets out a specific mechanism as the exclusive avenue for retroactive relief”], superseded by statute on another ground as stated in People v. Wilson (2023) 14 Cal.5th 839, 869; see also People v. DeHoyos (2018) 4 Cal.5th 594, 601–603 [holding that a statute reducing punishment did not apply retroactively under Estrada because its resentencing provisions sufficiently indicated a contrary intent]; People v. Conley (2016) 63 Cal.4th 646, 657–658 (Conley) [same].)

We have applied Estrada’s inference of retroactivity to legislation that created an affirmative defense, contracted a criminal offense, or otherwise lessened punishment in some meaningful manner. These laws have included statutes addressing penalty enhancements as well as statutes concerned with substantive offenses. (E.g., People v. Prudholme (2023) 14 Cal.5th 961, 968–969 [statute reducing maximum probation term for nonviolent offenses applies retroactively]; People v. Wright (2006) 40 Cal.4th 81, 95 [newly enacted affirmative defense to transporting marijuana applies retroactively]; Nasalga, supra, 12 Cal.4th at p. 798 (plur. opn. of Werdegar, J.) [amendments that increase monetary amount of property loss to trigger sentencing enhancements apply retroactively]; Tapia, supra, 53 Cal.3d at pp. 300–301 [statute specifying that certain death-penalty qualifying special circumstances must be intentional applies retroactively]; People v. Rossi (1976) 18 Cal.3d 295, 302 [amendatory statute decriminalizing the commission of certain sexual acts applies retroactively].) We have also applied the Estrada inference to statutes that give trial courts discretion to impose lesser punishment. (People v. Stamps (2020) 9 Cal.5th 685, 699 [statute eliminating restriction on trial court’s ability to strike serious felony enhancement applies retroactively]; People v. Francis (1969) 71 Cal.2d 66, 76 (Francis) [modified treatment of marijuana possession from straight felony to either felony or misdemeanor applies retroactively].)

More recently, we have applied the Estrada inference to statutes that, while not limited to reducing punishment for a particular crime, created a concrete avenue for certain individuals charged with a criminal offense to be treated more leniently or avoid punishment altogether. (Frahs, supra, 9 Cal.5th at pp. 624, 629 [statute creating pretrial diversion program in lieu of criminal prosecution for individuals suffering from qualifying mental health disorders applies retroactively]; People v. Superior Court (Lara) (2018) 4 Cal.5th 299, 303 (Lara) [statute prohibiting prosecutors from directly filing criminal charges against minors in “adult” criminal court and giving juvenile courts sole discretion to decide whether to prosecute minors as adults applies retroactively].) We explained in Lara that although Estrada was not directly on point, “its rationale does apply. The possibility of being treated as a juvenile in juvenile court — where rehabilitation is the goal — rather than being tried and sentenced as an adult can result in dramatically different and more lenient treatment. Therefore, Proposition 57 reduces the possible punishment for a class of persons, namely juveniles.” (Lara, at p. 303.) “For this reason,” we concluded, “Estrada’s inference of retroactivity applies.” (Ibid.)

Notably, however, we have recognized “the limited role Estrada properly plays in our jurisprudence of prospective versus retrospective operation” and reiterated that Estrada does not “weaken[] or modify[]” section 3’s default rule that amendments to a statute apply prospectively. (Brown, supra, 54 Cal.4th at p. 324; see Evangelatos, supra, 44 Cal.3d at pp. 1208–1209 [“California continues to adhere to the time-honored principle . . . that in the absence of an express retroactivity provision, a statute will not be applied retroactively unless it is very clear from extrinsic sources that the Legislature or the voters must have intended a retroactive application. The language in Estrada . . . should not be interpreted as modifying this well-established, legislatively-mandated principle”].) Rather, Estrada informs the application of the default rule “in a specific context by articulating the reasonable presumption that a legislative act mitigating the punishment for a particular criminal offense is intended to apply to all nonfinal judgments.” (Brown, at p. 324.)

We have found Estrada inapplicable to statutes that, although arguably lessening punishment in some sense, did not implicate the central rationale behind the Estrada inference. In Brown, we held that a statute that temporarily increased the rate at which local prisoners could earn conduct credits against their sentences for good behavior did not apply retroactively under Estrada (Brown, supra, 54 Cal.4th at pp. 318, 320), even though it resulted in less time served. We emphasized that Estrada “supports an important, contextually specific qualification to the ordinary presumption that statutes operate prospectively: When the Legislature has amended a statute to reduce the punishment for a particular criminal offense, we will assume, absent evidence to the contrary, that the Legislature intended the amended statute to apply to all defendants whose judgments are not yet final on the statute’s operative date. [Citation.] We based this conclusion on the premise that ‘ “[a] legislative mitigation of the penalty for a particular crime represents a legislative judgment that the lesser penalty or the different treatment is sufficient to meet the legitimate ends of the criminal law.” ’ ” (Brown, at p. 323, fn. omitted.)

We concluded in Brown that although the statute’s increased rate for earning conduct credits functioned in practice to reduce punishment, it did not fall within the scope of Estrada. We explained that “the rule and logic of Estrada is specifically directed to a statute that represents ‘ “a legislative mitigation of the penalty for a particular crime” ’ [citation] because such a law supports the inference that the Legislature would prefer to impose the new, shorter penalty rather than to ‘ “satisfy a desire for vengeance” ’ [citation]. The same logic does not inform our understanding of a law that rewards good behavior in prison.” (Brown, supra, 54 Cal.4th at p. 325.) We rejected the argument that Estrada should be understood to apply more broadly to any statute that reduces punishment in any manner, explaining that this contention “would expand the Estrada rule’s scope of operation in precisely the manner we forbade” in our prior decisions, and “does not in any event represent a logical extension of Estrada’s reasoning.” (Brown, at p. 325.)

Consistent with our reasoning in Brown, both our case law and decisions by the Courts of Appeal have declined to apply the Estrada presumption to new legislation that modified aspects of how criminal cases are investigated or tried.³ (Accord, Tapia, supra, 53 Cal.3d at p. 299 [we apply laws that “address the conduct of trials” that are in effect at the time of trial]; People v. Mortimer (1873) 46 Cal. 114, 118 [Pen. Code procedures govern trial of crimes defined and committed before the code’s enactment].) In People v. Robertson (1989) 48 Cal.3d 18, 51 (Robertson), we held that a newly enacted death penalty law prohibiting the consideration of nonstatutory factors in aggravation, which the former death penalty law allowed, did not apply retroactively under Estrada. We reasoned that the law “had no bearing on the criminality of defendant’s conduct or the severity of punishment therefor; hence the statute in effect at the time of the offenses governs” the penalty phase retrial. (Ibid.)

In People v. Cervantes (2020) 55 Cal.App.5th 927 (Cervantes), the Court of Appeal held that legislative amendments to section 859.5, which expanded the requirement of recorded custodial interrogations to all persons suspected of committing murder, did not come within Estrada’s limited exception to section 3’s general rule of prospective application because the amendments did not alter or reduce criminal punishment or treatment for past criminal conduct. (Cervantes, at p. 940.) The Cervantes court further held that the logic of Estrada did not apply because the legislative findings accompanying section 859.5’s amendments indicated the changes “were not designed to provide a clear and significant benefit to defendants; they were designed to reduce biased interpretation of, and ensure the accuracy of the evidence of, the communication that occurs in an interrogation.” (Cervantes, at p. 941.)

Similarly, in People v. Sandee (2017) 15 Cal.App.5th 294 (Sandee), the Court of Appeal rejected the claim that the Electronic Communications Privacy Act (§ 1546 et seq.), which limited a government entity’s access to electronic device information, applied retroactively under Estrada. (Sandee, at p. 305, fn. 7.) The court determined that “[t]he retroactivity principle discussed in Estrada is not applicable here because the [statute] does not have the effect of lessening the punishment for a crime.” (Ibid.)

In sum, we have adhered to section 3’s default rule of prospective operation in a variety of contexts and have applied Estrada’s limited inference of retroactivity only to statutes that “are analogous to the Estrada situation” and by their nature implicate “Estrada’s logic” (Lara, supra, 4 Cal.5th at p. 312, some italics omitted); that is, statutes that either reduce the punishment for a criminal offense or create discretion to reduce such punishment, or narrow the scope of criminal liability, because such enactments give rise to an “inevitable inference that the Legislature [or electorate] 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” (Estrada, supra, 63 Cal.2d at p. 745).

4. Conflicting Court of Appeal Decisions

As noted, the question of section 1109’s retroactivity has divided the Courts of Appeal.

In this case, the Court of Appeal majority concluded that section 1109’s bifurcation provisions apply retroactively under Estrada and its progeny. (Burgos, supra, 77 Cal.App.5th at pp. 564–569.) The majority reasoned that section 1109 applies “to a distinct class of defendants — those charged with gang enhancements” (Burgos, at p. 565), and “the legislative findings in Assembly Bill 333 also show the Legislature intended to reduce punishment specifically for people of color — who overwhelmingly comprise the class of defendants charged with gang enhancements” (id. at p. 566). Because the legislative findings “repeatedly cite the disparate levels of punishment suffered by people of color under the old law,” the majority reasoned, section 1109 cannot be considered only “as a ‘prophylactic rule of criminal procedure’ designed to enhance the fairness of proceedings.” (Burgos, at p. 566.) The majority highlighted two ameliorative effects of the bifurcation provisions: the “increased possibility of acquittal — which necessarily reduces possible punishment,” and the reduction of “pressure to accept longer sentences” through unfavorable plea deals, which “will necessarily reduce the degree of punishment for many defendants.” (Id. at p. 567.)

The majority then rejected the argument that “different parts of Assembly Bill 333 should be treated differently under Estrada.” (Burgos, supra, 77 Cal.App.5th at p. 567.) It reasoned that the “legislative findings setting forth the ameliorative purposes of the bill apply to the entire bill, and they specifically address the reasons for the new bifurcation rules.” (Ibid.) Considering Assembly Bill 333’s substantive changes to section 186.22, which apply retroactively under Estrada, “[i]t would be especially incongruous for the Legislature to make one isolated section of a bill prospective-only without stating so expressly, expecting instead that a court would somehow discern this anomaly.” (Burgos, at pp. 567–568.)

The dissenting justice below disagreed. In his view, “section 1109 is not an ameliorative statute within the meaning of the Estrada rule, and therefore it is subject to the general rule that Penal Code provisions are presumed to be prospective-only.” (Burgos, supra, 77 Cal.App.5th at p. 569 (dis. opn. of Elia, J.).) The dissent reasoned that our prior applications of Estrada involved statutes that “were facially ‘ameliorative’ because they either reduced the punishment or created discretion to reduce the punishment for a criminal offense, or narrowed the scope of criminal liability.” (Burgos, at p. 571 (dis. opn. of Elia, J.).) Section 1109, by contrast, is “a prophylactic rule of criminal procedure expressly intended to employ new procedures aimed at enhancing the fairness of future criminal proceedings. It makes no change to any crime or defense and makes no change to any punishment provision, and it does not create the possibility of lesser punishment or any other ‘ameliorative’ benefit from which it could be inferred that failing to extend that benefit retroactively must have been motivated by a ‘desire for vengeance.’ ” (Burgos, at p. 572 (dis. opn. of Elia, J.).) Because section 1109 “makes a ‘purely procedural’ change to a trial procedure that will not have any impact ‘directly’ or indirectly on punishment,” the dissent reasoned, the general presumption of prospectivity applies. (Burgos, at p. 572 (dis. opn. of Elia, J.).)

In Montano, supra, 80 Cal.App.5th 82 and Ramos, supra, 77 Cal.App.5th 1116, the Courts of Appeal agreed with the Burgos majority. In Montano, the appellate court relied on the uncodified preamble in Assembly Bill 333 in concluding that section 1109 is ameliorative within the meaning of Estrada. (Montano, at pp. 105–106.) The Montano court also determined that section 1109 was sufficiently similar to the statutes in Lara and Frahs to bring it within the scope of Estrada’s inference of retroactivity. (Montano, at pp. 106–108.) The Ramos court likewise determined that Assembly Bill 333’s findings and declarations that the Legislature intended “to benefit a certain class of criminal defendants by reducing the potentially harmful and prejudicial impact of gang evidence through bifurcation of gang allegations” invoked “the logic of Estrada.” (Ramos, at p. 1131.)

Other Courts of Appeal, however, have found the dissent in Burgos persuasive. (Boukes, supra, 83 Cal.App.5th 937, review granted; Perez, supra, 78 Cal.App.5th 192, review granted; Ramirez, supra, 79 Cal.App.5th 48, review granted.) In Ramirez, the appellate court concluded that “the Estrada presumption does not apply to section 1109 because section 1109 ‘is not an ameliorative statute within the meaning of the Estrada rule.’ ” (Ramirez, at p. 65, quoting Burgos, supra, 77 Cal.App.5th at p. 569 (dis. opn. of Elia, J.).) Likewise, in Perez, the Court of Appeal held that because section 1109 is a procedural statute that ensures a jury will not be prejudiced by gang-related evidence, and not one that reduces the punishment imposed or reduces the scope of application of the gang statute, it does not give rise to an inference of retroactivity under Estrada. (Perez, at p. 207.) The Perez court contrasted “the new law in Lara, which was a new procedural law that had the effect of potentially reducing the punishment for a class of defendants,” with bifurcation under section 1109, which “ensures a jury will not be prejudiced by the introduction of evidence to support gang enhancement allegations.” (Perez, at p. 207.) The Boukes court agreed, concluding that section 1109 does not apply retroactively because it does not reduce punishment imposed on gang enhancements. (Boukes, at p. 948.)

With this background in mind, we turn to the question of whether section 1109’s bifurcation provisions apply retroactively.

5. Discussion

Before we address whether section 1109’s provisions provide the sort of relief contemplated in Estrada such that an inference of retroactivity arises, we resolve the threshold question of whether the statute expressly or impliedly addresses retroactive application. As noted, if the statute expressly states that it is retroactive or otherwise clearly and unavoidably indicates that the Legislature intended retroactive operation, that determination governs, regardless of whether section 1109 lessens punishment within the meaning of Estrada.

The parties agree that section 1109 does not contain an express statement as to its retroactivity. Defendants maintain, however, that the legislative findings and declaration of purpose accompanying Assembly Bill 333 clearly indicate the Legislature intended its provisions would apply retroactively. They contend that the Legislature‘s stated objective to reduce the prejudicial impact of gang evidence, and the concerns it expressed regarding racial and socioeconomic disparities relating to prosecution under the gang enhancement statute, demonstrate that the Legislature intended the statute‘s bifurcation provisions would apply retroactively.

There is no question that the legislative findings accompanying Assembly Bill 333 reflect significant concerns about gang enhancements in general, including about their usefulness in stemming crime and their disproportionate impact on people of color in particular.⁴ However, we do not discern from these findings a ” ‘clear and unavoidable implication’ ” that the Legislature intended Assembly Bill 333‘s bifurcation provisions to apply retroactively. (Brown, supra, 54 Cal.4th at p. 319.) Moreover, there is nothing in the legislative history to suggest that the Legislature considered the issue of retroactivity at all. (Cf. Evangelatos, supra, 44 Cal.3d at p. 1211 [enactment‘s stated objective to remedy system of inequity and injustice in tort actions does not suggest that the retroactivity question was considered].)

Indeed, our opinion in Tran, which concerned Assembly Bill 333‘s amendments to the gang enhancement statute, also suggests that the bill‘s legislative history does not shed much light on whether the Legislature intended for its provisions to have retroactive operation. If it did, it would not have been necessary in Tran to resort to the secondary question of whether the amendments to section 186.22 apply retroactively under Estrada. (Tran, supra, 13 Cal.5th at pp. 1206–1207.)

Most of the legislative findings appear to relate most directly to the provisions of Assembly Bill 333 that substantively narrow the gang enhancement statute, not section 1109. The legislative findings describe the function of section 1109 in more equivocal terms. (Stats. 2021, ch. 699, § 2, subd. (d)(6) [“Gang enhancement evidence can be unreliable and prejudicial to a jury” (italics added)]; id., subd. (e) [“Studies suggest that allowing a jury to hear the kind of evidence that supports a gang enhancement before it has decided whether the defendant is guilty or not may lead to wrongful convictions” (italics added)]; id., subd. (f) [“Bifurcation of trials where gang evidence is alleged can help reduce its harmful and prejudicial impact” (italics added)].)⁵ These limited assertions align with the measured nature of change in subdivision (a) of section 1109, which gives defendants the election of a bifurcated or a unitary trial, rather than making bifurcation mandatory whenever gang evidence may be introduced. In short, the legislative findings do not clearly indicate the Legislature intended section 1109 to have retroactive operation. (See Buycks, supra, 5 Cal.5th at p. 880 [statute that is ambiguous regarding retroactive application applies prospectively].)

With no “express declaration of retroactivity or a clear and compelling implication that the Legislature intended” to apply the statute retroactively (Hayes, supra, 49 Cal.3d at p. 1274), we apply section 3‘s presumption that section 1109 operates prospectively unless the statute “lessen[s] . . . punishment” within the meaning of Estrada.⁶ (Estrada, supra, 63 Cal.2d at p. 745.) We conclude that it does not.

As noted, section 1109 requires the trial court to try an underlying offense separately from a gang enhancement if requested by the defense, and it mandates that a substantive gang offense be tried separately from nongang-related offenses. By its terms, section 1109 does not directly or potentially reduce the punishment for an offense. Nor does it change the elements of a substantive offense, defense, or penalty enhancement. Likewise, it does not create an alternative avenue for certain individuals to receive lesser or no punishment. Instead, section 1109 reflects a prophylactic procedural rule that modifies the sequence of trial proceedings. In an effort to minimize the potentially prejudicial impact of gang evidence, it gives defendants the option to request bifurcation at trial of the gang enhancement allegations charged under section 186.22, subdivision (b) and the other charged offenses.

Because the procedures established by section 1109 do not alter the criminality of defendant‘s conduct or the severity of punishment, the logic of Estrada does not apply. (See Robertson, supra, 48 Cal.3d at p. 51.) The further any particular amendatory act is from a clear reduction of punishment, or removal of a class of defendants from the criminal system altogether, the less “inevitable” and “obvious” (Estrada, supra, 63 Cal.2d at p. 745) and more speculative the inference of retroactivity becomes. (Brown, supra, 54 Cal.4th at p. 325 [“a statute increasing the rate at which prisoners may earn credits for good behavior does not represent a judgment about the needs of the criminal law with respect to a particular criminal offense, and thus does not support an analogous inference of retroactive intent“].) We decline to extend the reach of our holding in Estrada to a statute so untethered from its logic.

The reasoning we applied in Estrada leads us to conclude that there are also practical reasons why it makes sense to distinguish between the retroactive effect of new legislation reducing the punishment prescribed by law and the retroactive effect of new legislation that makes the types of changes we see in section 1109. The Legislature, as the lawmaking body of the state (Cal. Const., art. IV, § 1), is accorded considerable deference in deciding how to set the penalties for particular offenses. (E.g., People v. Wilkinson (2004) 33 Cal.4th 821, 840–841.) When the Legislature determines that one of its enactments operates too harshly, we presume that it wants its reevaluation, as embodied in a new law, to apply as broadly as possible within the limits of finality. (Estrada, supra, 63 Cal.2d at p. 745.)

But it is more difficult to draw such an inference when a statute makes the types of changes reflected in section 1109. While the Legislature may determine that certain additional procedures are warranted to enhance the fairness of criminal proceedings, as it did in enacting section 1109, the Legislature might decide to adopt a rule of prospective application for such changes for reasons other than “a desire for vengeance.” (Estrada, supra, 63 Cal.2d at p. 745.) Some changes adopted by the Legislature might enhance the fundamental fairness already guaranteed by the Constitution and by existing statutory protections. But applying these changes retroactively to already-concluded proceedings will inevitably come with systemic costs (which may affect the resources available to ensure the timely and effective administration of justice in other cases). Thus, even if the Legislature decides it makes good policy sense to provide a defendant with a right to bifurcation as specified in section 1109 on a going forward basis, it could also reasonably intend, for reasons other than “vengeance” (Estrada, at p. 745), that the change in the law should not apply to defendants who have already been tried. (Cf. Jenkins v. Delaware (1969) 395 U.S. 213, 218 [potential incongruities “inherent in prospective decision making, i.e., some defendants benefit from the new rule while others do not,” “must be balanced against the impetus the technique provides for the implementation of long overdue reforms, which otherwise could not be practicably effected” (italics omitted)].)

It makes practical sense for courts to look to the Legislature to weigh the costs and benefits of retroactive versus only prospective application, and not to infer retroactive intent from legislative silence in these circumstances. That is because “we can[not] . . . say with confidence, as we did in Estrada, that the enacting body lacked any discernible reason to limit application of the law with respect to cases pending on direct review.” (Conley, supra, 63 Cal.4th at pp. 658–659; see id. at p. 658 [declining to apply Estrada where “the enacting body creates a special mechanism for application of the new lesser punishment to persons who have previously been sentenced, and . . . the body expressly makes retroactive application of the lesser punishment contingent on a court‘s evaluation of the defendant‘s dangerousness“].) Here, we cannot presume the Legislature‘s silence on section 1109‘s retroactivity equates with an intent to undo or jeopardize completed trials conducted under valid and constitutional laws (and where evidence supports guilt beyond a reasonable doubt).

Defendants contend that section 1109 is sufficiently similar to the statutes at issue in Lara and Frahs to bring it within the scope of Estrada. We disagree. It is true that the statutes we considered in Lara and Frahs had substantial procedural components. But the core of the changes made by those statutes was to allow for the possibility of channeling cases involving certain classes of offenders the Legislature identified as having diminished culpability — juvenile offenders (in Lara) and offenders with mental health disorders (in Frahs) — away from ordinary criminal proceedings, thereby reducing those offenders’ exposure to the punishment otherwise prescribed under applicable criminal law. (Lara, supra, 4 Cal.5th at p. 303; Frahs, supra, 9 Cal.5th at p. 624.) The crux of the statutes at issue in Lara and Frahs, and what mattered for purposes of our analysis, was the path to reduced punishment based on a legislative judgment that the punishments prescribed by law for certain offenses are frequently inappropriate for certain classes of persons who bear reduced culpability for their crimes. (Frahs, at p. 624 [“The statute here is similar to the scheme we considered in Lara, in that section 1001.36 by design and function provides a possible ameliorating benefit for a class of persons . . . by offering an opportunity for diversion and ultimately the dismissal of charges“].)

Section 1109, in contrast, does not reflect an intent to benefit defendants in the same way. It is by design and function a procedural change affecting the sequence of trial proceedings in situations where the defendant elects to invoke the statute. Unlike with the statutory changes at issue in Lara and Frahs, in enacting section 1109, the Legislature did not identify any class of defendants having a characteristic potentially demonstrating diminished culpability deserving of less punishment. Instead, section 1109 is designed to provide fairer proceedings for defendants subject to the same potential punishment. No ” ‘obvious’ and ‘inevitable’ inference” of retroactivity arises based on the statute‘s operation. (See dis. opn. of Evans, J., post, at p. 4.)

Defendants who elect a bifurcated trial will still be tried in a criminal court. They will face the same charges and the same potential sentences for those charges if convicted. And in some instances, the same gang evidence introduced to establish the elements of a gang enhancement might be admissible at a bifurcated trial on the underlying charge. (See Hernandez, supra, 33 Cal.4th at p. 1048 [“the criminal street gang enhancement is attached to the charged offense and is, by definition, inextricably intertwined with that offense“]; id. at pp. 1049–1050 [“To the extent the evidence supporting the gang enhancement would be admissible at a trial of guilt, any inference of prejudice would be dispelled“]; see also People v. Chhoun (2021) 11 Cal.5th 1, 31 [“The People are generally entitled to introduce evidence of a defendant‘s gang affiliation and activity if it is relevant to the charged offense“]; Boukes, supra, 83 Cal.App.5th at p. 948, review granted [holding any error in lack of bifurcation was harmless beyond a reasonable doubt because evidence of defendant‘s gang affiliation “would have been admissible for reasons aside from any effort to prove the special circumstance and gang enhancement findings, for instance, defendant‘s motivation for the killing“].)

Under these circumstances, the rationale of Estrada does not apply; we cannot say it is “inevitable” that the Legislature would intend retroactive application of the bifurcation provisions without expressly so declaring. (Estrada, supra, 63 Cal.2d at p. 745.) Neither Frahs nor Lara goes so far as defendants here propose. New bifurcation procedures that may, in some instances, be beneficial to a criminal defendant in that they conceivably could result in the exclusion of gang-related evidence during the trial of charged offenses are not the equivalent of a change in the legislated punishment that must be applied to all nonfinal cases on appeal.⁷ (Estrada, at p. 745; Brown, supra, 54 Cal.4th at p. 325, fn. 15 [legislative intent to equalize custody conduct credits, which served to reduce punishment by decreasing days in custody, “does not, by itself, provide a logical basis for inferring the Legislature wished to do so retroactively“].) We decline to apply Estrada‘s inference of retroactivity when, as here, there is no alteration to criminality or the legislated punishment.

Defendants also contend that even if the Legislature‘s stated objectives in passing Assembly Bill 333 do not provide a clear indication that the Legislature intended that the measure‘s bifurcation provisions would apply retroactively, they nonetheless demonstrate that section 1109 is an ameliorative statute within the meaning of Estrada. The Court of Appeal majority likewise relied on Assembly Bill 333‘s legislative findings in concluding that Estrada‘s inference of retroactivity applies. (Burgos, supra, 77 Cal.App.5th at pp. 566–567.)

This logic is unavailing. Again, we decline to extend the reach of Estrada‘s inference to statutes like section 1109 without a clearer indication of legislative intent to mitigate punishment. As noted, section 1109‘s provisions do not directly or potentially reduce the punishment for an offense, change the elements of a substantive offense, defense, or penalty, or create an alternative avenue for certain individuals to receive lesser or no punishment. The uncodified legislative findings concerned with bifurcation establish an intent to promote fairness and reduce the potential for prejudice in trial proceedings where a gang enhancement is alleged. But they do not reflect an intent to lessen punishment within the meaning of Estrada and its progeny. Thus, they are insufficient to rebut section 3‘s “strong presumption of prospective operation.” (Brown, supra, 54 Cal.4th at p. 324.) “[T]he mere fact that the statute is remedial and subject to the rule of liberal construction” cannot, without more, give rise to the implication that the Legislature intended it to have retroactive application. (Aetna, supra, 30 Cal.2d at p. 395.)⁸

In this respect, we also find it notable that defendants and the Court of Appeal majority focus primarily on the legislative findings accompanying Assembly Bill 333 rather than the text of section 1109 itself.⁹ Likewise, the dissent makes a single passing reference to section 1109‘s actual provisions, noting only that it grants a criminal defendant charged with one or more gang enhancements “the right to have the question of guilt be tried in ‘separate phases.’ ” (Dis. opn. of Evans, J., at p. 1.) But this is the inverse of our usual Estrada inquiry. Estrada and its progeny focused on the substantive statutory provisions in determining whether they gave rise to an inference of retroactivity. Consistent with standard principles of statutory interpretation, our precedent instructs that we should focus on the statute itself before examining legislative findings or legislative history. (See, e.g., Frahs, supra, 9 Cal.5th at p. 635; Lara, supra, 4 Cal.5th at pp. 303, 309.) Indeed, in determining whether Assembly Bill 333‘s amendments to section 186.22 applied retroactively under Estrada, we relied on section 186.22‘s substantive changes to the elements of the gang enhancement rather than the accompanying legislative findings and statements of purpose, concluding that section 186.22 applied retroactively because it increased the threshold for conviction of the gang offense and imposition of the gang enhancement. (Tran, supra, 13 Cal.5th at pp. 1206–1207.) Although legislative findings may be a useful aid in confirming whether we should presume retroactive intent based on the ameliorative nature of a statute, we are disinclined to bypass section 1109‘s substantive provisions, which do not actually operate to mitigate punishment, and rely on legislative findings relating to potential prejudice in determining whether Estrada‘s inference of retroactivity applies.

We do not dispute that the changes implemented through section 1109 are designed to serve the salutatory goals of enhancing the overall fairness of criminal proceedings and reducing the potential for prejudice. (Stats. 2021, ch. 699, § 2, subds. (d)(6), (e) & (f).) Yet we have declined to infer retroactive intent under the principles of Estrada based solely on the presence of similar remedial purposes. In Brown, we rejected the argument that the Estrada inference extends “broadly to any statute that reduces punishment in any manner” (Brown, supra, 54 Cal.4th at p. 325), explaining that such a position “would expand the Estrada rule‘s scope of operation in precisely the manner we forbade in Evangelatos” (ibid.). In the earlier Evangelatos decision, we rejected the argument that a tort reform proposition‘s purpose, to rectify the “system of inequity and injustice” (Civ. Code, § 1431.1, subd. (a)), demonstrated that the electorate intended for it to apply retroactively. (Evangelatos, supra, 44 Cal.3d at p. 1213.) We explained: “Although the ‘findings and declaration of purpose’ included in the proposition clearly indicate that the measure was proposed to remedy the perceived inequities resulting under the preexisting joint and several liability doctrine and to create what the proponents considered a fairer system under which ‘defendants in tort actions shall be held financially liable in closer proportion to their degree of fault’ [citation], such a remedial purpose does not necessarily indicate an intent to apply the statute retroactively. Most statutory changes are, of course, intended to improve a preexisting situation and to bring about a fairer state of affairs, and if such an objective were itself sufficient to demonstrate a clear legislative intent to apply a statute retroactively, almost all statutory provisions and initiative measures would apply retroactively rather than prospectively. In light of the general principles of statutory interpretation set out above, and particularly the provisions of Civil Code section 3 [articulating the general rule of prospective application], the contention is clearly flawed.” (Ibid.) These same principles apply here and support the prospective application of section 1109‘s bifurcation procedures.¹⁰

Our Courts of Appeal have expressed a similar understanding of Estrada‘s limited applicability in a variety of contexts. In Cervantes, for example, the Court of Appeal held that statutory amendments that expanded the circumstances in which custodial interrogations must be recorded applied only prospectively (Cervantes, supra, 55 Cal.App.5th at p. 941), notwithstanding the fact that the Legislature had declared in the enactment that recording interrogations during police questioning of suspects “decreases wrongful convictions based on false confessions and enhances public confidence in the criminal justice process” (Stats. 2016, ch. 791, § 1, subd. (b)). (See id., subd. (a)(2) [citing national study finding that false confessions extracted during police questioning of suspects have been identified as a leading cause of wrongful convictions].) And in Sandee, the Court of Appeal held that the Electronic Communications Privacy Act did not apply retroactively under Estrada notwithstanding the likelihood that the act‘s increased restrictions on law enforcement access to electronic devices would benefit defendants, because the law did not have the effect of lessening the punishment for a crime. (Sandee, supra, 15 Cal.App.5th at p. 305, fn. 7.) The objective of the statutes at issue in these decisions, like section 1109, is to enhance the fairness of a criminal proceeding. But they do not implicate the logic of Estrada.

The dissent posits that “section 3‘s presumption of prospective effect does not apply to ‘new provisions’ — like the one here — ‘which clearly benefit only defendants.’ ” (Dis. opn. of Evans, J., post, at p. 5, citing Tapia, supra, 53 Cal.3d at p. 300.) Neither the text of section 3 nor our cases support this proposition. Our cases have instead made clear that the presumption applies except to changes that mitigate punishment. The statutory provisions at issue in Tapia applied retroactively because they benefitted defendants in the specific manner contemplated by Estrada and its progeny. (See Tran, supra, 13 Cal.5th at p. 1207, quoting Tapia, at p. 301 [”Estrada applies to statutory amendments ‘which redefine, to the benefit of defendants, conduct subject to criminal sanctions’ ” (italics added)].) Specifically, the statutory provisions found to operate retroactively in Tapia “provide[] that the killing of a prosecutor or judge, to constitute a special circumstance, must be intentional” and “also provide[] that an accomplice, for a non-felony-murder special circumstance to be found true, must have had the intent to kill.” (Tapia, at p. 301.) Section 1109, by contrast, does not create an affirmative defense, contract a criminal offense, or otherwise lessen punishment by redefining the law‘s treatment of a defendant‘s conduct. There is no basis in our case law for inferring that the Legislature intended a rule about the conduct of trials to apply retroactively to cases that have already been tried, absent a clear indication to that effect.

It is, of course, true that procedures designed to enhance the fairness and accuracy of particular proceedings might indirectly result in reduced punishment for individual offenders, insofar as the procedures might increase the chances of acquittal. But the question before us is whether section 1109 constitutes a change within the meaning and logic of Estrada, and we have not previously applied the Estrada inference so broadly. As the dissenting Court of Appeal justice aptly put it, “Although the Legislature expressly intended section 1109 to enhance the fairness of future proceedings, there is a manifest distinction between the Legislature‘s creation of new criminal procedures designed to enhance fairness and its enactment of provisions that reduce the possibility of punishment. The underlying rationale for the Estrada rule is to avoid the inference that the Legislature was bent on vengeance. No such inference arises from the nonretroactivity of section 1109‘s provisions. Instead, the Legislature has clearly decided that these new procedures offer a better way to avoid the risk of undue prejudice than the previous procedures of discretionary bifurcation coupled with restrictive jury instructions.” (Burgos, supra, 77 Cal.App.5th at pp. 573–574 (dis. opn. of Elia, J.).) To hold that any new procedures which could inure to a defendant‘s benefit come within the Estrada exception to section 3‘s default rule on that basis alone is not a “logical extension of Estrada‘s reasoning” and would represent a significant expansion of Estrada‘s scope of operation “in precisely the manner we [previously] forbade.” (Brown, supra, 54 Cal.4th at p. 325.)

Defendants also assert that the retroactivity of Assembly Bill 333 should be considered as a whole. Because the legislative findings pertain to the entirety of the bill and it is undisputed that the amendments to section 186.22 have retroactive application, defendants argue, section 1109‘s provisions should also apply retroactively. But most of the findings relate to the former gang enhancement provisions, not discretionary bifurcation. Moreover, we do not examine retroactivity at the level of the entire enactment. We focus on the nature of each provision rather than its company. (E.g., Tapia, supra, 53 Cal.3d at pp. 286, 289, 301; Francis, supra, 71 Cal.2d at p. 78 [rejecting argument that legislative intent applicable to chaptered laws “is ‘presumptively the same’ “; to the contrary, “the Legislature manifestly could have different intents with respect to different sections contained in one chapter“]; Burgos, supra, 77 Cal.App.5th at p. 573 (dis. opn. of Elia, J.) [“Many legislative bills amend numerous (sometimes hundreds of) statutes, and whether a specific amendatory statute is subject to the Estrada rule depends on the nature of the amendment, not the mere fact that the amendment was enacted in the company of other amendments in a single legislative bill“].) The mere placement of multiple provisions in a single bill does not provide a “clear and compelling implication” that the Legislature intended all would have retroactive application. (Hayes, supra, 49 Cal.3d at p. 1274.) Accordingly, in the face of the Legislature‘s silence regarding Assembly Bill 333‘s application, we conclude that Estrada‘s inference of retroactivity applies to the punishment-mitigating changes to section 186.22, but not provisions governing the conduct of trial in section 1109. (See Brown, supra, 54 Cal.4th at p. 322 [when statutory language does not clearly support interpretation of retroactive application, courts must “resolve such ambiguities in favor of prospective operation,” which “is precisely the function of section 3 and the default rule it embodies“].)

In sum, we see no justification in Estrada and its progeny for applying the Estrada inference of retroactivity to section 1109. We therefore turn to Burgos‘s alternative claim that equal protection principles require the statute to be applied retroactively.

B. Prospective Application of Section 1109 Does Not Violate Equal Protection Principles

In the alternative, Burgos argues that prospective-only application of section 1109 would violate the equal protection clauses of the federal and state Constitutions (U.S. Const., 14th Amend.; Cal. Const., art. I, § 7, subd. (a)). He contends there is no rational basis to distinguish between two classes of individuals charged with gang enhancements, those convicted before the effective date of section 1109, whose judgments are not yet final, and those convicted after section 1109‘s enactment. We disagree.

“The equal protection clause of the Fourteenth Amendment to the United States Constitution provides that no state may ‘deny to any person within its jurisdiction the equal protection of the laws.’ [Citation.] This provision is ‘essentially a direction that all persons similarly situated should be treated alike.’ [Citation.] ‘At core, the requirement of equal protection ensures that the government does not treat a group of people unequally without some justification.’ ” (People v. Hardin (2024) 15 Cal.5th 834, 847, fn. omitted (Hardin).)

“The degree of justification required to satisfy equal protection depends on the type of unequal treatment at issue.” (Hardin, supra, 15 Cal.5th at p. 847.) “[W]hen a statute involves neither a suspect classification nor a fundamental right, the ‘general rule is that legislation is presumed to be valid and will be sustained if the classification drawn by the statute is rationally related to a legitimate state interest.’ ” (Ibid.) Burgos concedes that rational basis review applies.

We recently clarified that “when plaintiffs challenge laws drawing distinctions between identifiable groups or classes of persons, on the basis that the distinctions drawn are inconsistent with equal protection, courts no longer need to ask at the threshold whether the two groups are similarly situated for purposes of the law in question. The only pertinent inquiry is whether the challenged difference in treatment is adequately justified under the applicable standard of review. The burden is on the party challenging the law to show that it is not.” (Hardin, supra, 15 Cal.5th at pp. 850–851.)

We have repeatedly rejected an equal protection argument based on the timing of a statute lessening punishment for a particular offense. (Floyd, supra, 31 Cal.4th at pp. 188–190; Baker v. Superior Court (1984) 35 Cal.3d 663, 668–669; accord, People v. Willis (1978) 84 Cal.App.3d 952, 956; People v. Superior Court (Gonzales) (1978) 78 Cal.App.3d 134, 142–143.) We have explained the basis for this holding as follows: ”Estrada itself recognized that when the Legislature has amended a statute to lessen the punishment, its determination as to which statute should apply to all convictions not yet final, ’either way, would have been legal and constitutional.’ [Citations.] That the Legislature‘s choice, either way, would be constitutional is the foundation for our oft-repeated statement that, in this type of circumstance, the problem ‘is one of trying to ascertain the legislative intent — did the Legislature intend the old or new statute to apply?’ [Citations.] Defendant‘s equal protection argument presumes that the Estrada rule is constitutionally compelled. As we have stated repeatedly, it is not.” (Floyd, supra, 31 Cal.4th at pp. 188–189; see People v. Aranda (1965) 63 Cal.2d 518, 532 [“A refusal to apply a statute retroactively does not violate the Fourteenth Amendment“], superseded by statute as stated in People v. Capistrano (2014) 59 Cal.4th 830, 868.) Burgos offers no persuasive reason to refrain from applying our prior precedent to an amended statute that is not ameliorative under Estrada.¹¹

Accordingly, we hold that equal protection principles do not require section 1109 to be applied retroactively.

C. Burgos‘s Additional Arguments Exceed the Scope of Review

Burgos also contends that even if section 1109 applies prospectively only, the admission of gang evidence at his trial was unduly prejudicial and violated his due process rights. The Court of Appeal did not reach this and other claims brought by the defense in light of its holding that Assembly Bill applies retroactively. (Burgos, supra, 77 Cal.App.5th at p. 564, fn. 4.)

We decline to address Burgos‘s claims of prejudice because they exceed the scope of our review. Our briefing order was limited to the question of whether “the provision of Penal Code section 1109 governing the bifurcation at trial of gang enhancements from the substantive offense or offenses appl[ies] retroactively to cases that are not yet final.” Having answered this question in the negative, we remand the matter to the Court of Appeal to resolve any remaining claims in the first instance.

III. DISPOSITION

We conclude that section 1109 does not apply retroactively. Because the Court of Appeal majority concluded otherwise, we reverse its judgment and remand the matter for further proceedings consistent with our opinion.

We also disapprove People v. Montano, supra, 80 Cal.App.5th 82 and People v. Ramos, supra, 77 Cal.App.5th 1116, to the extent they are inconsistent with our opinion.

GUERRERO, C. J.

We Concur:

CORRIGAN, J.

KRUGER, J.

GROBAN, J.

JENKINS, J.

I agree with the majority opinion that the inference of retroactivity from In re Estrada (1965) 63 Cal.2d 740 does not extend to Penal Code section 1109. (Maj. opn., ante, at p. 2.) However, I write separately to emphasize a point we have made before: It would be preferable for “the Legislature, and the electorate with respect to ballot measures, to consider the retroactive application of new laws and to regularly express their intent regarding if and how they should be applied retroactively.” (People v. Prudholme (2023) 14 Cal.5th 961, 980.) Instead, without an express statement from the Legislature, we are left to attempt to glean what we think the Legislature intended. In the last 10 years alone, we have been compelled to make a retroactivity determination at least nine times, including the present case. None of the new laws at issue in these cases contained an express statement regarding retroactivity.¹

Our Courts of Appeal have been compelled to make this determination with even greater regularity. There have been well over a dozen published Court of Appeal cases in just the last two years that have addressed the retroactivity of new laws where no express statement regarding retroactivity was provided by the Legislature.²

Without a clear statement from the Legislature, we scour uncodified legislative findings, declarations of purpose, statements on the floor, and committee reports in an effort to decipher the Legislature‘s intent. This process is admittedly less precise than simply looking to a clear statement in the statute itself. Indeed, two of my distinguished colleagues here, analyzing exactly the same legislative materials relied on by the majority, have come to the opposite conclusion regarding retroactivity. (Dis. opn., post, at p. 1.) Going forward, the Legislature can remain silent regarding retroactivity and we will continue to use these less than perfect analytical tools to determine what we believe it meant. Or, instead, the Legislature can take a more preferable course and plainly state whether or not the statute in question is retroactive.

GROBAN, J.

I Concur:

CORRIGAN, J.

Whether ameliorative legislation, such as Penal Code section 1109,¹ applies retroactively is a question of legislative intent. (In re Estrada (1965) 63 Cal.2d 740, 745 (Estrada).) One may readily discern the Legislature‘s intent with respect to section 1109 by examining its text and context, the uncodified legislative findings accompanying its enactment, and its legislative history. (See In re Marriage of Bouquet (1976) 16 Cal.3d 583, 587 (Bouquet) [“Consistent with Estrada‘s mandate, we must address ‘all pertinent factors’ when attempting to divine the legislative purpose“].) That examination convinces me that the Legislature intended this statute, which was explicitly intended to prevent the conviction of innocent defendants, to have retroactive effect. I therefore respectfully dissent.

As relevant here, section 1109 grants a criminal defendant who is charged with one or more gang enhancements (§ 186.22, subds. (b), (d)) the right to have the question of guilt be tried in “separate phases” (§ 1109, subd. (a)): trial on the underlying offense and then, if necessary, trial on the truth of the gang enhancement. (Id., subd. (a)(1) & (2).)

Placing the choice whether to bifurcate proceedings involving a gang enhancement in the defendant‘s hands represented a sea change in how gang cases are tried. Previously, defendants had to rely entirely on the trial court‘s discretion to bifurcate trial of a gang enhancement from nongang offenses. (See People v. Hernandez (2004) 33 Cal.4th 1040, 1049–1050.) In practice, that discretion was exercised only “rarely.” (Sen. Com. on Public Safety, Analysis of Assem. Bill No. 333 (2021–2022 Reg. Sess.) as amended May 28, 2021, p. 8.) Because of what was then regarded as the efficiencies of joinder, courts denied bifurcation even when admission of the gang evidence would have been excluded as “unduly prejudicial” in the trial on the nongang offenses. (Hernandez, at p. 1050; see Sen. Com. on Public Safety, Analysis of Assem. Bill No. 333, supra, as amended May 28, 2021, at p. 9.)

The legislative history of section 1109 explained the magnitude of the potential prejudice: evidence supporting a gang enhancement ” ‘may be so extraordinarily prejudicial, and of so little relevance to guilt, that it threatens to sway the jury to convict regardless of the defendant‘s actual guilt.’ ” (Sen. Com. on Public Safety, Analysis of Assem. Bill No. 333, supra, as amended May 28, 2021, p. 8, italics added.) The legislative history also noted the testimony of a San Joaquin County deputy district attorney and expert on gang prosecutions before the Committee on the Revision of the Penal Code, who ” ‘agreed that bifurcating evidence of gang involvement from evidence related to the underlying charges would reduce the risk of unfairly prejudicing juries and convicting innocent people.’ ” (Id. at p. 10, italics added.)

This unfairness was of profound concern to the Legislature. In clear and forceful language, the Legislature found and declared that “[g]ang enhancement evidence can be unreliable and prejudicial to a jury because it is lumped into evidence of the underlying charges which further perpetuates unfair prejudice in juries and convictions of innocent people.” (Stats. 2021, ch. 699, § 2, subd. (d)(6), italics added.) It relied on studies showing that “allowing a jury to hear the kind of evidence that supports a gang enhancement before it has decided whether the defendant is guilty or not may lead to wrongful convictions” and noted, as a result, that “[t]he mere specter of gang enhancements pressures defendants to accept unfavorable plea deals rather than risk a trial filled with prejudicial evidence and a substantially longer sentence.” (Id., § 2, subd. (e).) Citing the 2020 Annual Report of the Committee on Revision of the Penal Code, the Legislature found that this unfairness was borne disproportionately by “people of color, creating a racial disparity.” (Stats. 2021, ch. 699, § 2, subd. (d)(1).)

The disparity identified by the Legislature is substantial. Nearly 90 percent of individuals in the state‘s gang database are Black or Latino (Stats. 2021, ch. 699, § 2, subd. (d)(10)), and over 98 percent of people sentenced to prison for a gang enhancement in our state‘s largest county are people of color. (Id., subd. (d)(4).) The Legislature further found that “gang enhancement statutes criminalize entire neighborhoods historically impacted by poverty, racial inequality, and mass incarceration as they punish people based on their cultural identity, who they know, and where they live.” (Id., subd. (a).)

Like the Court of Appeal below, I believe the Legislature‘s concern about the conviction of innocent defendants indicates it wanted to extend the benefit of this ameliorative statute “to every case to which it constitutionally could apply.” (Estrada, supra, 63 Cal.2d at p. 745.) Legislatures as well as courts are aware that avoiding “the injustice that results from the conviction of an innocent person has long been at the core of our criminal justice system.” (Schlup v. Delo (1995) 513 U.S. 298, 325.) A legal maxim, well known to lawyers and nonlawyers alike, states ” ‘that it is better that ninety-nine . . . offenders should escape, than that one innocent man should be condemned.’ ” (Ibid.) Against this shared background, the Legislature‘s explicit concern about preventing “wrongful convictions” of “innocent people” (Stats. 2021, ch. 699, § 2, subds. (e), (d)(6)), which it supported with empirical studies (id., subd. (e)), cannot reasonably be read as “equivocal.” (Maj. opn., ante, at p. 21.)

In short, section 1109 is an ameliorative statute that by design and function is aimed at preventing the conviction of innocent defendants. It is an “obvious” and “inevitable” inference the Legislature would have been as concerned about wrongful convictions of innocent people that were then occurring or already had occurred as it was about wrongful convictions that could occur three months later, once the statute took effect. (Estrada, supra, 63 Cal.2d at p. 745; cf. In re Johnson (1970) 3 Cal.3d 404, 413 [“the more directly the new rule in question serves to preclude the conviction of innocent persons, the more likely it is that the rule will be afforded retrospective application“].) “[T]o hold otherwise would be to conclude that the Legislature was motivated by a desire for vengeance,” as there can be no penological justification for denying the benefit of section 1109 to innocent defendants whose convictions are not yet final. (Estrada, at p. 745.)

Consequently, this is yet another instance where the ” ‘logic’ ” and “rationale” (maj. opn., ante, at pp. 12–13) of Estrada require a statute to be applied retroactively. As we have previously explained, section 3‘s presumption of prospective effect does not apply to “new provisions” — like the one here — “which clearly benefit only defendants.” (Tapia v. Superior Court (1991) 53 Cal.3d 282, 300; see People v. Conley (2016) 63 Cal.4th 646, 657 (Conley) [“The Estrada rule rests on an inference that, in the absence of contrary indications, a legislative body ordinarily intends for ameliorative changes to the criminal law to extend as broadly as possible“].) Section 1109 should therefore be given retroactive effect.

I.

The majority opinion evades this conclusion by disregarding the development of the Estrada doctrine over the past several decades and focusing instead only on the narrow question whether the statute here “mitigate[s] punishment.” (Maj. opn., ante, at p. 34.) While it is true that the particular statute in Estrada did act “to lessen the punishment” (Estrada, supra, 63 Cal.2d at p. 745), we have long since recognized that Estrada can apply to many other types of statutes, including those that “created an affirmative defense” (maj. opn., ante, at p. 10), “contracted a criminal offense” (ibid.), or granted courts ”discretion to impose lesser punishment” (ibid., italics added). Our precedent likewise holds that a statute may be retroactive even when it is not restricted to “a particular crime” (id. at p. 11) and even though ”Estrada was not directly on point,” so long as ” ‘its rationale’ ” or ” ‘logic’ ” applies (id., at pp. 11, 16).

The foregoing amply demonstrates the folly of an unduly narrow focus on whether an ameliorative statute has a particular kind of effect on punishment. Consider People v. Brown (2012) 54 Cal.4th 314 (Brown), which similarly attempted to cabin Estrada‘s scope only to the specific circumstances which, to that point, our case law happened to have addressed. In rejecting an attempt to make retroactive a temporary increase in the rate at which county jail inmates could earn conduct credits, Brown purported to limit Estrada to ” ’ “[a] legislative mitigation of the penalty for a particular crime” ’ ” and deemed it significant that the increase in credits for good behavior “does not represent a judgment about the needs of the criminal law with respect to a particular criminal offense.” (Brown, at p. 325, second italics added.)

A few years later, however, we reversed field and repudiated Brown‘s narrow reading of Estrada. We deemed it irrelevant that a new procedural provision prohibiting prosecutors from charging juveniles with crimes directly in adult court, and instead conditioning transfer to adult court on a fitness hearing in juvenile court, “did not ameliorate the punishment, or possible punishment, for a particular crime.” (People v. Superior Court (Lara) (2018) 4 Cal.5th 299, 308, italics added (Lara).) Although the “effect” of the new rule was “different from the statutory changes in Estrada,” we reasoned that “the same inference of retroactivity should apply” because “[t]he possibility of being treated as a juvenile in juvenile court . . . can result in dramatically different and more lenient treatment.” (Lara, at pp. 308, 303.) Shortly after that, we applied the Estrada retroactivity exception to a new legislative scheme that, for most charged felonies, offered defendants the possibility of mental health diversion. (People v. Frahs (2020) 9 Cal.5th 618 (Frahs).) And in People v. Prudholme (2023) 14 Cal.5th 961 (Prudholme), we applied the Estrada retroactivity exception to a statute reducing the maximum allowable probation term for a wide range of offenses, despite the fact that the new statute was not limited to “any particular offense” (id. at p. 976) and did not even involve “punishment” but instead modified the terms of “an act of leniency” (id. at p. 968) — and despite the fact that Estrada was “not directly on point” (ibid.).

What mattered in Prudholme, as well as in these other cases, was not whether the particular statute made a “punishment-mitigating change[]” (maj. opn., ante, at p. 36) nor whether the statute happened to mirror statutes we had previously deemed to be retroactive under Estrada, but whether the “rationale of Estrada” applied. (Prudholme, supra, 14 Cal.5th at p. 968.) That rationale inquires whether the statute “by design and function provides a possible ameliorating benefit” (Frahs, supra, 9 Cal.5th at p. 624) as to punishment (see Estrada, supra, 63 Cal.2d at p. 745) or as to the issue of guilt or innocence (see People v. Wright (2006) 40 Cal.4th 81, 95 [newly enacted affirmative defense]) for an identifiable “class of persons” (Lara, supra, 4 Cal.5th at p. 308; Frahs, at p. 624). Where, as here, the Legislature enacts an ameliorative statute aimed at benefiting defendants, it “represents a legislative judgment that the lesser penalty or the different treatment is sufficient to meet the legitimate ends of the criminal law.” (Estrada, at p. 745, italics added.)

Even if the test for retroactivity were limited to whether the statute reflected an intent to lessen punishment, section 1109 certainly satisfies the test. The Legislature explicitly contemplated that bifurcating trial on a gang enhancement from trial on the underlying felony — like a statute narrowing the definition of a crime or recognizing a new affirmative defense — would significantly increase the possibility of acquittal and thereby ” ‘reduce the possibility of punishment.’ ” (Maj. opn., ante, at p. 35.) It is true that section 1109 is procedural, but (as the majority concedes (see maj. opn., ante, at p. 26)) so too were the statutes we considered in Lara and Frahs. (See also People v. Stamps (2020) 9 Cal.5th 685, 699 [statute “[e]liminating the prior restriction on the court‘s ability to strike a serious felony enhancement“]; People v. Francis (1969) 71 Cal.2d 66, 76 [statute “vest[ing] in the trial court discretion to impose either the same penalty as under the former law or a lesser penalty“].) Section 1109, like these other procedural statutes, is retroactive because it reflects an intent to “ma[k]e a reduced punishment possible.” (Frahs, supra, 9 Cal.5th at p. 629.)

The majority opinion‘s occasional effort to broaden its characterization of Estrada‘s rationale — i.e., as applicable to “statutes that either reduce the punishment for a criminal offense or create discretion to reduce such punishment, or narrow the scope of criminal liability” (maj. opn., ante, at p. 16) — falls short of the mark. This characterization does not articulate an actual rationale; it merely describes the individual instances in which this court has previously found Estrada to apply.² What‘s worse, this mechanical approach replicates Brown‘s improperly narrow reading of Estrada, which we shortly thereafter effectively repudiated in Lara, Frahs, and Prudholme. I fear this zigzag approach to applying Estrada will sow confusion in the law and exacerbate the uncertainty that has already led the lower courts time and again to arrive at conflicting decisions as to whether a given statute is retroactive. (Compare, e.g., People v. Burgos (2022) 77 Cal.App.5th 550, 565–567 with People v. Perez (2022) 78 Cal.App.5th 192, 207.) Yet we are not without the tools to solve the problem. The question for courts, properly understood, is whether a statute, by design and function, provides a clear and significant benefit to defendants as to punishment or on the question of guilt and innocence. Unlike the majority, I find it unimaginable the Legislature would have restricted its concern about convicting the innocent to only a subset of those it could constitutionally reach.

II.

The majority opinion seeks to downplay the Legislature‘s focus on preventing wrongful convictions of innocent people in two ways. First, it tries to divert attention away from the express legislative findings accompanying the enactment of section 1109. Second, it insists the Legislature could not have meant what it said about the risk of convicting innocent people when irrelevant gang evidence is included in a trial on the underlying substantive offense. Neither effort is persuasive.

A.

The majority opinion intimates there is something untoward about considering the explicit legislative findings included in the uncodified section of the bill that enacted section 1109. (See maj. opn., ante, at pp. 30–31.) Not so. As the majority opinion elsewhere concedes, the question of whether new legislation applies retroactively ” ’ “is one of trying to ascertain the legislative intent.” ’ ” (Id. at p. 38.) From the beginning we have instructed courts making that inquiry to consider “all pertinent factors” (Estrada, supra, 63 Cal.2d at p. 746), and we have reaffirmed that broad review continuously (see, e.g., Bouquet, supra, 16 Cal.3d at p. 587 [“Consistent with Estrada‘s mandate, we must address ‘all pertinent factors’ “]), including in our most recent retroactivity decisions (see, e.g., Prudholme, supra, 14 Cal.5th 961, 967 (Prudholme) [quoting Estrada]). The majority opinion‘s contention that we have “disapproved” this practice (maj. opn., ante, at p. 14, fn. 3) is manifestly mistaken.

It is also beyond dispute that the Legislature‘s explicit findings are a valuable resource in ascertaining legislative intent. We regularly consult uncodified provisions, which are “part of the statutory law” (Carter v. California Dept. of Veterans Affairs (2006) 38 Cal.4th 914, 925), to ascertain “[t]he general intent” of legislation (Munson v. Del Taco, Inc. (2009) 46 Cal.4th 661, 669) as well as its “express purpose and intended scope” (People v. Valencia (2017) 3 Cal.5th 347, 362). (See generally People v. Allen (1999) 21 Cal.4th 846, 858, fn. 13 [uncodified sections “express the Legislature‘s view on some aspect of the operation or effect of the bill“].) Those findings, which are set forth above, evidence the Legislature‘s special concern about the risk of convicting innocent people.

B.

In fact, it‘s no exaggeration to say that convicting the innocent is what the law abhors most. Such a scenario is the subject of some of our culture‘s most enduring literature and film, from To Kill a Mockingbird to The Shawshank Redemption. Instances of convicting the innocent haunt our nation‘s history. (See, e.g., Stevenson, Just Mercy: A Story of Justice and Redemption (2014); The Central Park Five (Sundance Selects 2012).) It is therefore perplexing that the majority opinion describes the Legislature‘s findings on this fundamental point to be merely ” ’ “vague phrases” ’ ” and ” ’ “broad general language.” ’ ” (Maj. opn., ante, at p. 14, fn. 3.)

I view the findings differently. The Legislature stated clearly and specifically that lumping gang enhancement evidence with “evidence of the underlying charges . . . perpetuates unfair prejudice in juries and convictions of innocent people.” (Stats. 2021, ch. 699, § 2, subd. (d)(6).) The Legislature supported this finding with empirical studies showing that the jury‘s consideration of “the kind of evidence that supports a gang enhancement . . . may lead to wrongful convictions.” (Id., § 2, subd. (e).) Finally, the legislative history noted that the proposed remedy — section 1109 — would be effective. As noted above, a prosecution gang expert opined that ” ‘bifurcating evidence of gang involvement from evidence related to the underlying charges would reduce the risk of unfairly prejudicing juries and convicting innocent people.’ ” (Sen. Com. on Public Safety, Analysis of Assem. Bill No. 333, supra, as amended May 28, 2021, p. 10.)

To characterize the Legislature‘s conscious effort to remedy this type of injustice as merely a general desire “to promote fairness and reduce the potential for prejudice in trial proceedings” (maj. opn., ante, at p. 29) is not only a profound understatement but a category error. Section 1109 is not merely a “new procedure[] which could inure to a defendant‘s benefit.” (Maj. opn., ante, at p. 35.) Rather, it enables juries, at the defendant‘s option, to carefully and soberly assess the evidence regarding the charged offense without being distracted by the irrelevant and inflammatory suggestion the defendant is allegedly associated with a criminal street gang. Never before have we denied retroactive effect to a statute that “by design and function provides a possible ameliorating benefit for a class of persons” (Frahs, supra, 9 Cal.5th at p. 624) — and certainly not when the statute is aimed at preventing the conviction of an innocent defendant. Not even once.

C.

The remaining cases on which the majority opinion relies — none of which was “designed to provide a clear and significant benefit to defendants” (People v. Cervantes (2020) 55 Cal.App.5th 927, 941) — are soundly distinguishable. Cervantes considered the retroactive effect of an amended statute requiring that custodial interrogations of murder suspects be recorded. (§ 859.5, subds. (a), (b), (d); Stats. 2016, ch. 791, § 2.) In enacting the amendments, the Legislature had made findings that recorded interrogations would generally promote fairness neutrally to both the prosecution and the defense and reduce the likelihood of prejudice arising from inaccurate reports of what transpired. (Stats. 2016, ch. 791, § 1, subd. (a)(2), (3).) In particular, the Legislature believed that recorded interrogations would not only “decrease[] wrongful convictions based on false confessions” (id., § 1, subd. (b)), but also ” ‘prevent disputes about how an officer conducted himself or herself or treated a suspect during the course of an interrogation, prevent a defendant from lying about the account of events he or she originally provided to law enforcement, and spare judges and jurors the time necessary and the need to assess which account of an interrogation to believe.’ ” (Cervantes, at p. 941, quoting Stats. 2016, ch. 791, § 1, subd. (b).)

Because the amendments were aimed at improving the justice system generally, they were not retroactive. (Cervantes, at p. 941; see also People v. Sandee (2017) 15 Cal.App.5th 294, 305 & fn. 7 [statute limiting governmental searches of cell phones, designed to protect Californians’ privacy, was not retroactive].)³ These examples of “new procedures which could inure to a defendant‘s benefit” (maj. opn., ante, at p. 35) or general improvements in “how criminal cases are investigated or tried” (id. at p. 13) shed little light on section 1109 and the stakes highlighted by the Legislature.

The majority opinion‘s hypothesized legislative justification for refusing to apply section 1109 retroactively likewise does not withstand scrutiny. The majority opinion suggests that the “systemic costs” of retroactive application could justify a decision not to apply the new bifurcation policy to defendants who have already been tried, noting these “completed trials” were “conducted under valid and constitutional laws (and where evidence supports guilt beyond a reasonable doubt).” (Maj. opn., ante, at pp. 25–26.) But this stated concern fails to distinguish the case before us from any of the other instances in which we have found Estrada to apply. Anytime the Legislature reduces punishment, expands the opportunities for diversion, limits the ability to try a minor as an adult, or increases the possibility of acquittal by, say, adding a new element to (or otherwise narrowing) an offense, retroactive application of the new statute could equally “undo or jeopardize completed trials conducted under valid and constitutional laws (and where evidence supports guilt beyond a reasonable doubt).” (Id. at p. 26.) In Frahs, supra, 9 Cal.5th 618, for example, a defendant who had been convicted of second degree robbery, found to have suffered a prior serious felony conviction, and sentenced to nine years in prison was allowed to seek eligibility for mental health diversion under section 1001.36 — despite the fact the defendant had been found guilty beyond a reasonable doubt under valid and constitutional laws as they then existed.

In this case, the Court of Appeal found it “likely” the jury improperly relied on evidence of Francisco Burgos‘s and his codefendants’ gang membership to establish their guilt of two counts of robbery, given that neither victim identified any of the defendants at trial. (People v. Burgos, supra, 77 Cal.App.5th at p. 569.) Denying Burgos, whose conviction is not yet final, the opportunity to demonstrate he would not have been convicted had the gang enhancement been bifurcated “can, by hypothesis, serve no purpose other than to satisfy a desire for vengeance.” (Frahs, supra, 9 Cal.5th at p. 628.) While the Legislature may be free, as a constitutional matter, to apply such a statute either retroactively or prospectively (see maj. opn., ante, at pp. 38–39) — or even to provide for an alternate avenue of relief, and thereby rebut the inference of retroactive effect (see, e.g., Conley, supra, 63 Cal.4th at pp. 658–659) — the majority opinion cites no authority suggesting that “systemic costs” (maj. opn., ante, at p. 25) have ever led the Legislature to intend prospective-only application of an ameliorative statute that “by design and function” (Frahs, at p. 624) was aimed at preventing the wrongful conviction of innocent people. Given its express findings, there is no indication the Legislature had “a fear of too much justice,” which in this circumstance “would be repugnant to deeply rooted conceptions” against punishing the innocent. (McCleskey v. Kemp (1987) 481 U.S. 279, 339 (dis. opn. of Brennan, J.).)

III.

For all these reasons, I am unwilling to conclude the Legislature would have been so nonchalant about denying a remedy for innocent defendants who were convicted simply because the jury was allowed to consider evidence of their alleged association with gangs that was unrelated to the charged crimes a few days, weeks, or months before the effective date of section 1109. By finding otherwise, the court today seems to have lost sight of the fact that the general statutory presumption in section 3 that statutes operate prospectively “is not a straitjacket.” (Estrada, supra, 63 Cal.2d at p. 746.) As we recently observed, this presumption “should not be followed blindly in complete disregard of factors that may give a clue to the legislative intent.” (Prudholme, supra, 14 Cal.5th at p. 967.) Section 1109‘s explicit focus on the risk of wrongful conviction of innocent people offers much more than a clue.

Unlike the majority opinion, I believe that a statute which, by design and function, provides a clear benefit to defendants on the question of guilt or innocence should apply in every case to which it constitutionally could apply — regardless of whether the statute might be characterized as more procedural than substantive. It seems rather unlikely the Legislature would have been more concerned about the elusive line between statutes that are procedural and those that are substantive than whether the previous statutory regime may have led to the conviction of innocent people whose judgments could still be corrected.

I therefore respectfully dissent.

EVANS, J.

I Concur:

LIU, J.

See next page for addresses and telephone numbers for counsel who argued in Supreme Court.

Name of Opinion People v. Burgos

Procedural Posture (see XX below)

Original Appeal

Original Proceeding

Review Granted (published) XX 77 Cal.App.5th 550

Review Granted (unpublished)

Rehearing Granted

Opinion No. S274743

Date Filed: June 3, 2024

Court: Superior

County: Santa Clara

Judges: Cynthia A. Sevely and Helen E. Williams

Counsel:

Laurie Wilmore, under appointment by the Supreme Court; and Conrad Petermann, under appointment by the Court of Appeal, for Defendant and Appellant Francisco Burgos.

Jean M. Marinovich, under appointment by the Supreme Court, for Defendant and Appellant Damon Stevenson, Jr.

Solomon R. Wollack, under appointment by the Supreme Court, for Defendant and Appellant James Richardson.

Rob Bonta, Attorney General, Lance E. Winters, Chief Assistant Attorney General, Jeffrey M. Laurence, Assistant Attorney General, Alice B. Lustre, Seth K. Schalit and J. Michael Chamberlain, Deputy Attorneys General, for Plaintiff and Respondent.

Counsel who argued in Supreme Court (not intended for publication with opinion):

Solomon R. Wollack

Attorney at Law

P.O. Box 23933

Pleasant Hill, CA 94523

(925) 671-2501

Jeffrey M. Laurence

Assistant Attorney General

455 Golden Gate Avenue, Suite 11000

San Francisco, CA 94102

(415) 510-3812

Notes

1

All further statutory references are to the Penal Code unless otherwise indicated. See People v. Tran (2022) 13 Cal.5th 1169, 1206–1207 (amendments to Pen. Code, § 186.22 apply retroactively to nonfinal cases); Prudholme, supra, 14 Cal.5th at page 963 (Assembly Bill No. 1950 (2019–2020 Reg. Sess.) (Assembly Bill 1950) applies retroactively to nonfinal cases); People v. Stamps (2020) 9 Cal.5th 685, 699 (Senate Bill No. 1393 (2017–2018 Reg. Sess.) applies retroactively to nonfinal cases); People v. Frahs (2020) 9 Cal.5th 618, 624, 628–637 (Pen. Code, § 1001.36 applies retroactively to nonfinal cases); People v. Superior Court (Lara) (2018) 4 Cal.5th 299, 304 (Proposition 57 (approved Nov. 8, 2016) applies retroactively to nonfinal cases); People v. Gentile (2020) 10 Cal.5th 830, 852 (Senate Bill No. 1437 (2017–2018 Reg. Sess.) does not apply retroactively to nonfinal cases on direct appeal); People v. DeHoyos (2018) 4 Cal.5th 594, 602–603 (Proposition 47 (approved Nov. 4, 2014) does not apply retroactively to nonfinal cases on direct appeal); People v. Conley (2016) 63 Cal.4th 646, 657 (Proposition 36 (approved Nov. 6, 2012) does not apply retroactively to nonfinal cases on direct appeal); see also People v. Esquivel (2021) 11 Cal.5th 671, 675–676 (collecting cases). All subsequent references are to the Penal Code unless otherwise specified.

2

Lozano pleaded guilty prior to trial and Byrd was acquitted at trial. See, e.g., People v. Slaton (2023) 95 Cal.App.5th 363, 376, review granted November 15, 2023, S282047 (Evid. Code, § 352.2 does not apply retroactively to nonfinal cases on direct appeal); People v. Ramos (2023) 90 Cal.App.5th 578, 596 , review granted July 12, 2023, S280073 (same); People v. Venable (2023) 88 Cal.App.5th 445, 456, review granted May 17, 2023, S279081 (Evid. Code, § 352.2 applies retroactively to nonfinal cases); People v. Jackson (2023) 93 Cal.App.5th 207, 212, review granted September 13, 2023, S281267 (Assembly Bill 1950 modified the defendant‘s probationary period); People v. Shelly (2022) 81 Cal.App.5th 181, 186 (Assembly Bill 1950 applies retroactively to nonfinal cases); People v. Govan (2023) 91 Cal.App.5th 1015, 1020 (Assembly Bill No. 518‘s (2021–2022 Reg. Sess.) amendments to Pen. Code, § 654 apply retroactively to nonfinal cases); In re S.S. (2023) 89 Cal.App.5th 1277, 1288 (Assembly Bill No. 2361‘s (2021–2022 Reg. Sess.) amendments to Welf. & Inst. Code, § 707 apply retroactively to nonfinal cases); People v. E.M. (2022) 85 Cal.App.5th 1075, 1088–1090 [trial court erred in denying recall on the ground that Senate Bill No. 1393 did not apply retroactively to defendant‘s case because it was final on appeal]; People v. Superior Court (Ortiz) (2022) 81 Cal.App.5th 851, 860, footnote 2 (Pen. Code, § 1001.95 applies retroactively to nonfinal cases); In re N.L. (2022) 81 Cal.App.5th 463, 469 (Senate Bill No. 383 (2021–2022 Reg. Sess.) applies retroactively to nonfinal cases); People v. Zabelle (2022) 80 Cal.App.5th 1098, 1109 (Pen. Code, § 1170‘s amended statutory language applies retroactively to nonfinal cases); People v. Singh (2022) 81 Cal.App.5th 147, 153 (Assembly Bill No. 1259 (2021–2022 Reg. Sess.) applies retroactively to nonfinal cases). In fact, it‘s not even an accurate description. If the test under Estrada were solely whether ” ‘the Legislature has amended a statute to lessen the punishment’ ” (maj. opn., ante, at p. 38), then the amended statute at issue in Brown, supra, 54 Cal.4th 314, which temporarily increased the rate at which county inmates could earn conduct credits (and thus “earlier release from custody” (id. at p. 322)), would have had retroactive effect. But Brown correctly did not restrict its analysis to the statute‘s effect on punishment. We reasoned instead that the amended credits statute, enacted in the face of “a state fiscal emergency” (id. at p. 318), was not designed to confer a benefit on defendants; rather, it was designed to “offer[] incentives for future good behavior.” (Id. at p. 324, italics added.)

3

The dissenting opinion attempts to downplay the significance of our decision in Brown, asserting that we “reversed field and repudiated Brown’s narrow reading of Estrada” a few years later. (Dis. opn. of Evans, J., post, at p. 6; see id. at pp. 8–9.) Rather than renouncing it, this court has continued to cite Brown with approval in discussing principles of retroactivity. (See, e.g., People v. Esquivel (2021) 11 Cal.5th 671, 676 [including Brown in summary of this court’s relevant precedent evaluating whether specific legislation ameliorates punishment]; Buycks, supra, 5 Cal.5th at p. 880 [reaffirming Brown’s holding]; People v. Hajek and Vo (2014) 58 Cal.4th 1144, 1195–1196 [same]; Lara, supra, 4 Cal.5th at p. 312 [agreeing that conduct credits at issue in Brown were “not ‘analogous’ to the Estrada situation, and Estrada’s logic d[id] not apply” to them].)
In contrast, the dissent relies on cases that have either been disapproved by our court as improperly expanding Estrada’s reach or do not involve the interpretation of statutes at all. (See dis. opn. of Evans, J., post, at p. 1, citing In re Marriage of Bouquet (1976) 16 Cal.3d 583, 587; dis. opn. of Evans, J., post, at p. 4, citing In re Johnson (1970) 3 Cal.3d 404, 413.) In Californians for Disability Rights v. Mervyn’s, LLC (2006) 39 Cal.4th 223, 229–230, we explained that “at least in modern times, we have been cautious not to infer the voters’ or the Legislature’s intent on the subject of prospective versus retrospective operation from ‘vague phrases’ [citation] and ‘broad, general language’ [citation] in statutes, initiative measures and ballot pamphlets. We have also disapproved statements to the contrary in certain older cases,” including Marriage of Bouquet. And Johnson addresses the standards for determining “[t]he retrospective effect of a law-making opinion” (Johnson, at p. 410, italics added), which are fundamentally different from determining the retroactive operation of a statute. (See, e.g., Frlekin v. Apple Inc. (2020) 8 Cal.5th 1038, 1057 [“ ‘ “The general rule that judicial decisions are given retroactive effect is basic in our legal tradition” ’ ”].) The statute at issue in People v. Robertson (1989) 48 Cal.3d 18, 51, did not implicate a defendant‘s guilt or innocence, and the statutory scheme in Evangelatos v. Superior Court (1988) 44 Cal.3d 1188 did not even implicate the criminal law at all. (Id. at p. 1210, fn. 15 [“the Estrada decision provides no guidance for the resolution of this case“].)

4

For example, the legislative findings accompanying Assembly Bill 333 state that “[c]urrent gang enhancement statutes criminalize entire neighborhoods historically impacted by poverty, racial inequality, and mass incarceration as they punish people based on their cultural identity, who they know, and where they live” (Stats. 2021, ch. 699, § 2, subd. (a)), “[n]o empirical studies have produced evidence showing that gang enhancements meaningfully reduce crime or violence” (id., subd. (c)), and “[t]he gang enhancement statute is applied inconsistently against people of color, creating a racial disparity” (id., subd. (d)(1)).

5

We disagree with the dissent‘s characterization of the legislative findings pertaining to section 1109 as “clear and forceful” (dis. opn. of Evans, J., post, at p. 2) on the relevant issue before us — i.e., whether the limited bifurcation provisions apply retroactively to nonfinal judgments. We of course are “aware that avoiding ‘the injustice that results from the conviction of an innocent person has long been at the core of our criminal justice system’ ” (dis. opn. of Evans, J., post, at p. 4, citing Schlup v. Delo (1995) 513 U.S. 298, 325), and we are cognizant that the Legislature identified racial disparities in the use of gang enhancements (dis. opn. of Evans, J., post, at p. 3). But we also note that strongly worded legislative findings relating to racial bias and unfairness in the criminal justice system do not necessarily convey that the Legislature intended for section 1109 — which concerns the order of presentation of evidence at trial, not the substantive scope of the gang enhancements — to apply retroactively to cases that have already been tried. (Compare Stats. 2020, ch. 317, § 2, subds. (a)–(j) with Pen. Code, former § 745, subd. (j) [“This section applies only prospectively in cases in which judgment has not been entered prior to January 1, 2021“].)

6

Stevenson incorrectly suggests that the Legislature‘s failure to include an express provision regarding the prospective application of Assembly Bill 333 supports a finding that the Legislature intended the statute to apply retroactively. As stated above, if a statute is silent as to retroactivity, it is generally presumed to apply prospectively. (§ 3; Brown, supra, 54 Cal.4th at p. 320; Evangelatos, supra, 44 Cal.3d at p. 1209 [under “the general principle of prospectivity, the absence of any express provision directing retroactive application strongly supports prospective operation of the measure“].)

7

In Assembly Bill 333‘s findings and declarations, the Legislature cited a decision by this court in which we recognized that gang evidence may in some instances be prejudicial (Stats. 2021, ch. 699, § 2, subd. (e), citing People v. Williams (1997) 16 Cal.4th 153, 193). But in that case we actually upheld the trial court‘s denial of the defendant‘s motion to exclude gang evidence because the evidence was relevant to motive and identity and its probative value outweighed its prejudicial effect (Williams, at p. 193).

8

The dissent emphasizes that the Legislature intended to prevent the conviction of “innocent” defendants serving “wrongful convictions.” (Dis. opn. of Evans, J., post, at pp. 1–4, 9–12, 14–16.) We acknowledge the important remedial purposes of the changes enacted by the Legislature, but we decline to infer retroactivity based on such “broad, general language in statutes.” (Brown, supra, 54 Cal.4th at p. 319.) The Legislature chose to adopt optional bifurcation procedures that operate in a limited manner; as noted, gang evidence may still be admissible where relevant to the underlying charges. A voluntary procedure which may (or may not) alter when gang evidence is admitted at trial is not the type of ameliorative reduction in punishment that gives rise to Estrada‘s inference of retroactivity.

9

In evaluating whether section 1109 applies retroactively under Estrada, the Court of Appeal addressed the statutory language only in passing, observing simply that it applies to a distinct class of defendants (those charged with gang enhancements or active participation in a street gang). (Burgos, supra, 77 Cal.App.5th at p. 565.) That section 1109 applies to a class of persons is of minimal significance in determining whether it applies retroactively under Estrada, since it is axiomatic that an amendatory act will affect a class of persons. Although we took note of the fact that the stated purposes of the enactments in Lara and Frahs were to address specific classes of individuals, we did so in the context of identifying the Legislature‘s desire to reduce their reentry into the criminal system. (See Lara, supra, 4 Cal.5th at p. 309; Frahs, supra, 9 Cal.5th at p. 626.)

10

The dissent points out that “the statutory scheme in Evangelatos v. Superior Court (1988) 44 Cal.3d 1188 did not even implicate the criminal law.” (Dis. opn. of Evans, J., post, at p. 13, fn. 5.) But both Civil Code section 3 and Penal Code section 3 contain a default presumption in favor of prospective application.

11

In Hardin, we explained that our rejection of the two-step framework for determining equal protection challenges did not “call into question any of this court‘s precedent that purported to dispose of an equal protection challenge upon deciding that the challenged disparate treatment did not involve groups that were similarly situated for purposes of the law in question.” (Hardin, supra, 15 Cal.5th at p. 851.) Thus, Hardin does not call into question any of our prior precedent rejecting an equal protection argument based on the timing of a statute lessening punishment for a particular offense, even assuming such precedent might be said to have been premised on an application of the two-step framework.

Case Details

Case Name: People v. Burgos

Court Name: California Supreme Court

Date Published: Jun 3, 2024

Citations: 16 Cal.5th 1; 292 Cal. Rptr. 3d 587; S274743

Docket Number: S274743

Court Abbreviation: Cal.