Opinion of the Court by Jenkins, J.

In 1987, petitioner William Milton was convicted of two robberies in Illinois. In 1999, he was convicted of robbery in California, and the prosecution sought an enhanced sentence on the ground the two Illinois robbery convictions were “serious felony” convictions that were also “strikes” under the “Three Strikes” law (Pen. Code,1 §§ 667, subds. (b)–(j), 667.5, subd. (c), 1170.12, subd. (a)). Because an out-of-state robbery qualifies as a serious felony only if it “includes all of the elements of [a California robbery]” (§§ 667, subd. (d)(2); see 1192.7, subd. (19)), which an Illinois robbery does not, the prosecution asserted the Illinois robberies were serious felonies under section 1192.7, subdivision (c)(8) and (23), which provide that any felony in which the defendant personally uses a firearm or a dangerous or deadly weapon is a serious felony. After reviewing the record from the Illinois robbery cases, including the charging document and sentencing hearing transcript, the trial court found petitioner used a firearm in committing both Illinois robberies and imposed a third strike sentence. The Court of Appeal affirmed the judgment, and we denied review.

1 All undesignated statutory references are to the Penal Code.

1 In re MILTON Opinion of the Court by Jenkins, J.

In 2017, petitioner filed a petition for writ of habeas corpus, seeking resentencing under People v. Gallardo (2017) 4 Cal.5th 120, 124–125 (Gallardo), where we held that a trial court violates a defendant’s Sixth Amendment right to a jury trial when it makes factual findings about the nature of a defendant’s prior conviction in imposing an enhanced sentence based on that prior conviction. The Court of Appeal denied the petition on the ground that Gallardo was not retroactive to petitioner’s judgment, which had been final since 2000. (People v. Milton (2019) 42 Cal.App.5th 977, 982, 987 (Milton).) The Courts of Appeal that have addressed the question of Gallardo’s retroactivity have reached conflicting results. We conclude the Gallardo rule does not apply retroactively to final judgments. Accordingly, we affirm the Court of Appeal’s denial of the petition for writ of habeas corpus.

FACTUAL AND PROCEDURAL BACKGROUND

In 1987, an information was filed in Illinois charging petitioner with armed robbery (Ill. Rev. Stat., former ch. 38, § 18-2(a)) and simple robbery (Ill. Rev. Stat., ch. 38, former § 18-1). The information alleged petitioner committed armed robbery by taking money from the victim “while ar[med] with a dangerous weapon, a gun . . . by threatening the imminent use of force” and that he committed simple robbery by taking money from his victim “by threatening the imminent use of force.” Petitioner pleaded guilty to simple robbery, and an Illinois jury found him guilty of armed robbery. The Illinois court held a combined sentencing hearing for the two convictions. At the hearing, the Illinois prosecutor recounted the testimony of the armed robbery victim as follows: “Mr. Milton got out of the car, pointed a gun at [the

2 In re MILTON Opinion of the Court by Jenkins, J.

victim], and threatened him, forced him into the car where [the victim] was robbed of his goods.” The court stated, “[Y]ou used a gun. You stopped the victim. . . . You forced this individual into the automobile.” For the simple robbery, the prosecutor stated the victim was “accosted by” petitioner, who “approache[d] [the victim] with a weapon, threaten[ed] him, and . . . [the victim] lost his entire paycheck . . . to Mr. Milton.” The court added, “You stopped [the victim], and again at the point of a gun you took . . . cash from this individual.” The court stated that the “stipulated facts” for the simple robbery were “that the victim . . . left [the market] after cashing his check. He was stopped. Money was demanded from the victim by . . . Milton, who possessed a handgun. And [money] was taken from the victim . . . In addition . . . , [petitioner] made a statement to the Waukegan Police Department that he participated in and did take the money as is described in this stipulation.” Before pronouncing its sentence, the court stated, “In each of the two respective offenses you deliberately held a gun — a loaded gun — upon an individual. . . . I’m going to tell you that he who participates in an offense of violence against another with a gun is going to be punished.” (See People v. Milton (1989) 182 Ill.App.3d 1082, 1095 [538 N.E.2d 1227, 1236] [Illinois sentencing court imposed an aggravated sentence as to Milton’s armed robbery conviction based on several factors including gun use].) In 1998, shortly after he was released from prison in Illinois for the two Illinois robberies and other subsequent offenses, petitioner was charged with another robbery, this time in California, and a jury found him guilty of second degree robbery (§ 211). Petitioner waived his right to a jury trial as to the truth of his prior convictions. In a bifurcated proceeding,

3 In re MILTON Opinion of the Court by Jenkins, J.

he admitted he had two prior felony convictions from Illinois and that one of them was an armed robbery conviction that qualified as a strike. He argued his Illinois simple robbery conviction was not a strike. The California prosecutor acknowledged that robbery under Illinois law, unlike under California law, does not require the specific intent to permanently deprive the victim of the property. The prosecutor argued, however, that the trial court could rely on certified documents from Illinois in determining that petitioner used a firearm during both robberies and that the robberies were therefore serious felonies under section 1192.7, subdivision (c)(8) and (23), which provide that a felony in which the defendant personally uses a firearm or a dangerous or deadly weapon is a serious felony. Petitioner’s counsel argued the court was not authorized to look beyond the facts of the Illinois convictions in determining firearm use and that, in any event, “the stipulated facts” from “the transcript . . . of the plea” showed only that petitioner possessed a handgun, not that he used one, in committing the simple robbery. The court stated it saw “nothing wrong” with relying on the certified documents and found petitioner used a gun during both robberies and that both prior convictions were strikes. The court imposed a third strike sentence of 25 years to life, plus five years for the prior serious felony enhancement (§ 667, subd. (a)(1)). Petitioner appealed and contended, among other things, that his Illinois simple robbery conviction did not qualify as a strike. The Court of Appeal affirmed, concluding the trial court was “entitled to look at the entire record of conviction to determine the substance of the foreign convictions” and that “the abstract of the judgment, the stipulated facts of the

4 In re MILTON Opinion of the Court by Jenkins, J.

offense in question and the Illinois court’s sentencing comments,” which “were admissible as part of the ‘entire record’ of [the] conviction,” provided substantial evidence to support the trial court’s finding that petitioner “obtained the proceeds of both robberies by pointing and threatening the victims with a handgun.” Petitioner filed a petition for review, which we denied. He subsequently filed five petitions for writ of habeas corpus in this court, each of which we denied. In 2016, petitioner filed a sixth petition for writ of habeas corpus in our court, asserting the trial court erred in finding his Illinois simple and armed robberies were serious felonies for purposes of the Three Strikes law.2 We denied the petition “without prejudice to any relief to which petitioner might be entitled after this court decides [Gallardo],” which was then pending in our court. Following our decision in Gallardo, petitioner filed the instant petition, arguing his Illinois robberies were not serious felonies under Gallardo. We issued an order directing the Department of Corrections and Rehabilitation to show cause, returnable to the Court of Appeal, “why petitioner is not entitled to relief pursuant to [Gallardo] . . . and why Gallardo should not apply retroactively

2 As noted, petitioner previously admitted his Illinois armed robbery was a serious felony. He later changed his position to assert, as he does here, that both of his Illinois robberies were not serious felonies. Because we affirm the denial of petitioner’s instant writ petition on retroactivity grounds, we need not, and will not, address the Attorney General’s argument that, if Gallardo applies retroactively, petitioner’s prior admission regarding the Illinois armed robbery precludes him from asserting it is not a serious felony.

5 In re MILTON Opinion of the Court by Jenkins, J.

on habeas corpus to final judgments of conviction.” Following briefing, the Court of Appeal denied the petition in a published opinion. (Milton, supra, 42 Cal.App.5th at pp. 993–994.) The court concluded the trial court’s factfinding regarding firearm use would have been improper under Gallardo and that if Gallardo applied retroactively, remand would have been necessary. (Id. at p. 999.) The court held, however, that petitioner was not entitled to relief because Gallardo was not retroactive to his final judgment under federal and state tests for retroactivity. (Id. at pp. 993–994.) As we noted above, the Courts of Appeal that have considered Gallardo’s retroactivity are split on the issue. (E.g., Milton, supra, 42 Cal.App.5th 977 [Gallardo is not retroactive]; In re Brown (2020) 45 Cal.App.5th 699, review granted June 10, 2020, S261454 [Gallardo is retroactive]; In re Scott (2020) 49 Cal.App.5th 1003, review granted Aug. 12, 2020, S262716 [not retroactive]; In re Haden (2020) 49 Cal.App.5th 1091, review granted Aug. 12, 2020, S263261 [not retroactive].) We granted review to resolve the conflict in the Courts of Appeal.

DISCUSSION

I. Sentence Enhancements Based on Prior Felony

Convictions Under the Three Strikes law, a prior conviction for a “serious felony” (§ 1192.7, subd. (c)) or “violent felony” (§ 667.5, subd. (c)) is a strike and subjects a defendant to increased punishment. (§§ 1170.12, subd. (a), 667, subds. (b)–(j).) For a prior out-of-state conviction to qualify as a strike under section 667, subdivision (d)(2), the out-of-state offense must include all of the elements of a serious or violent felony in California. (People v. Warner (2006) 39 Cal.4th 548, 552–553.)

6 In re MILTON Opinion of the Court by Jenkins, J.

The California offense of robbery (§ 211), which is a serious felony (§ 1192.7, subd. (c)(19)), is a specific intent crime that requires “ ‘the intent to permanently deprive the person of the property.’ ” (People v. Jackson (2016) 1 Cal.5th 269, 343.) Illinois robberies are general intent crimes, and the definitions of robbery and armed robbery in Illinois do not include this specific intent element. (People v. Jamison (2001) 197 Ill.2d 135, 161 [756 N.E.2d 788, 801]; People v. Lee (1998) 294 Ill.App.3d 738, 743 [691 N.E.2d 117, 120].) Because the Illinois robbery statutes do not contain all the elements of California’s robbery statute, petitioner’s Illinois robbery convictions do not qualify as strikes under section 667, subdivision (d)(2). An out-of-state felony, however, also qualifies as a serious felony under California law if the defendant personally used a firearm or a dangerous or deadly weapon in committing the offense. (§ 1192.7, subd. (c)(8), (23); People v. Le (2015) 61 Cal.4th 416, 425.) Thus, if petitioner personally used a firearm in the commission of the Illinois felonies, those prior convictions would be serious felony convictions and strikes under California’s Three Strikes law. At the time the trial court sentenced petitioner in 1999, California law permitted trial courts to examine “the entire record” of a prior conviction “to determine the substance of” that conviction — i.e., the conduct underlying the conviction — for sentence enhancement purposes. (People v. Guerrero (1988) 44 Cal.3d 343, 355 (Guerrero).) A year later, the United States Supreme Court held in Apprendi v. New Jersey (2000) 530 U.S. 466 (Apprendi) that “any fact that increases the penalty for a crime beyond the prescribed statutory maximum” must be found by a jury. (Id. at p. 490.) The high court, however, preserved the so-called Almendarez-Torres exception

7 In re MILTON Opinion of the Court by Jenkins, J.

(Almendarez-Torres v. United States (1998) 523 U.S. 224), under which “the fact of a prior conviction” used to impose an enhanced punishment for a later offense may be found by the court. (Apprendi, at p. 490.) Applying these principles, the Apprendi court concluded the defendant was improperly denied a jury trial on the factual predicate of an enhancement allegation that was attached to his pending charges. (Id. at pp. 474, 490.) In People v. McGee (2006) 38 Cal.4th 682 (McGee), we determined that Apprendi’s extension of the right to have a jury make factual findings did not apply to enhancements based on prior convictions. (McGee, at pp. 709–710.) We held the Sixth Amendment permits courts to determine whether the offense qualifies as a strike and that while the inquiry is a “limited one” that “focus[es] on the elements of the offense of which the defendant was convicted,” a court may also review the record of the prior conviction to determine whether “the conviction realistically may have been based on conduct that would not constitute a serious felony under California law.” (McGee, at p. 706.) We concluded the defendant therefore had no right to a jury finding on whether his prior conviction qualified as a serious felony and that the trial court did not err in examining the victim’s preliminary hearing testimony from the prior case to determine the nature of the prior conviction. (McGee, at pp. 689, 709.) Less than a decade later, the United States Supreme Court extended the right to have a jury make factual determinations about the nature of a prior conviction that is used to increase punishment. (See Descamps v. United States (2013) 570 U.S. 254 (Descamps); Mathis v. United States (2016) 579 U.S. 500 (Mathis).) In Descamps, the district court

8 In re MILTON Opinion of the Court by Jenkins, J.

enhanced the defendant’s punishment under the Armed Career Criminal Act (ACCA, 18 U.S.C. § 924(e)), an elements-based federal recidivist sentencing scheme that requires imposition of an increased sentence where the elements of the prior conviction match those of an enumerated offense. (Descamps, at p. 254; see Taylor v. United States (1990) 495 U.S. 575, 600, 601 [under the ACCA, the sentencing court compares only the elements of the offenses and does not look at the “facts underlying the prior convictions” or “the facts of each defendant’s conduct”].) The prior conviction in Descamps was a California burglary, which did not include an unlawful entry element as required by the ACCA. (Descamps, at pp. 254, 255– 259.) In determining that the California burglary conviction qualified as an ACCA predicate offense, the district court reviewed a transcript of the plea colloquy, which showed the defendant did not object to the prosecutor’s statement that the burglary “ ‘involve[d] the breaking and entering of a grocery store.’ ” (Descamps, supra, 570 U.S. at p. 259.) The high court concluded this was error and held that the district court’s review of a prior conviction to determine whether it qualifies under the ACCA must be limited to the elements of those offenses (the “categorical approach”), except to the extent a limited inquiry into the record of the prior conviction is necessary to determine which part of a divisible statute was violated (the “modified categorical approach”). (Descamps, at pp. 267–270.) The Descamps court concluded that because California’s burglary statute is not a divisible statute divided into lawful and unlawful entry alternatives, the prosecutor’s references to the defendant’s breaking and entering was extraneous and could play no role in the district court’s efforts

9 In re MILTON Opinion of the Court by Jenkins, J.

to determine whether the conviction qualified as an ACCA predicate offense. (Descamps, at pp. 277–278.) Subsequently, in Mathis, supra, 579 U.S. 500, the United States Supreme Court considered whether an Iowa burglary conviction qualified as a prior conviction for ACCA purposes. Emphasizing that the ACCA involves an “elements-only inquiry” where “a sentencing judge may look only to ‘the elements of the [offense], not to the facts of [the] defendant’s conduct,’ ” the high court reaffirmed Descamps’s holding “that the prior crime qualifies as an ACCA predicate if, but only if, its elements are the same as, or narrower than, those of the generic offense.” (Mathis, supra, at pp. 510, 503.) In both Descamps and Mathis, the United States Supreme Court rested its decision on cases interpreting the ACCA but also drew upon Sixth Amendment principles for its holding. The Descamps court explained that the district court’s decision failed to consider “the categorical approach’s Sixth Amendment underpinnings. . . . The Sixth Amendment contemplates that a jury — not a sentencing court — will [make findings about the defendant’s underlying conduct], unanimously and beyond a reasonable doubt.” (Descamps, supra, 570 U.S. at p. 269.) Similarly, the Mathis court stated that allowing courts to make factual determinations regarding prior convictions would raise Sixth Amendment concerns. (Mathis, supra, 579 U.S. at pp. 508, 511.)

II. Our Gallardo Decision

Shortly after the high court’s decisions in Descamps and Mathis, we revisited our earlier decision in McGee in Gallardo. (Gallardo, supra, 4 Cal.5th at p. 124.) In Gallardo, the prosecution alleged the defendant’s prior conviction for

10 In re MILTON Opinion of the Court by Jenkins, J.

aggravated assault under former section 245, subdivision (a)(1), a divisible statute, qualified as a strike. (Gallardo, supra, 4 Cal.5th at p. 123.) The defendant had pleaded guilty to that offense, but her plea did not specify whether she had used a deadly weapon (a serious felony) or force likely to produce great bodily injury (not a serious felony). (Id. at p. 125; § 1192.7, subd. (c)(31) [“assault with a deadly weapon” is a serious felony].) To resolve this ambiguity, the trial court reviewed a transcript of the victim’s preliminary hearing testimony that the defendant used a knife and found, based on that testimony, that the prior conviction qualified as a strike. (Gallardo, at p. 125.) Although the trial court complied with the procedure we approved of in McGee, we concluded in Gallardo that the procedure was no longer viable in light of Descamps and Mathis, which informed us that a defendant’s “constitutional right to a jury trial swe[pt] more broadly than our case law previously recognized.” (Gallardo, supra, 4 Cal.5th at p. 138.) We “disapprove[d] People v. McGee” “insofar as it authorize[d] trial courts to make findings about the conduct that ‘realistically’ gave rise to a defendant’s prior conviction.” (Gallardo, at pp. 125, 134.) We held the court’s factfinding role regarding prior convictions was now “limited to identifying those facts that were established by virtue of the [prior] conviction itself — that is, facts the jury was necessarily required to find to render a guilty verdict, or that the defendant admitted as the factual basis for a guilty plea.” (Id. at p. 136; see id. at p. 124 [the 6th Amend. “ ‘contemplates that a jury,’ ” not the court, will make such factual findings], quoting Descamps, supra, 570 U.S. at p. 269.) Applying these principles, we concluded the trial court violated the defendant’s

11 In re MILTON Opinion of the Court by Jenkins, J.

right to a jury trial, and we remanded the matter for the court to determine what the defendant necessarily admitted about the nature of her crime when she entered her plea. (Gallardo, at p. 138.) Justice Chin concurred in the majority’s conclusion that a defendant has the right to a jury trial on the nature of his or her prior conviction but disagreed with the majority’s remedy of remanding the matter for the trial court to review the record of conviction. (Gallardo, supra, 4 Cal.5th at p. 140 (conc. & dis. opn. of Chin, J.).) He proposed instead that a jury should review the record of conviction and that the defendant should have the option of presenting live testimony on remand: “The proper remedy for a violation of defendant’s jury trial right is to give her that jury trial.” (Ibid.) The majority rejected Justice Chin’s suggested remedy, which neither party had requested. (Id. at p. 138.) The majority stated that having a jury empaneled for the sole purpose of reviewing the record of conviction would raise concerns about compliance with Apprendi and would not involve procedural safeguards such as cross-examination of witnesses “that normally apply in criminal proceedings.” (Id. at p. 139.) The majority also stated, “Our precedent instructs that determinations about the nature of prior convictions are to be made by the court, rather than a jury, based on the record of conviction. (See McGee, supra, 38 Cal.4th at p. 695.) We have explained that the purpose of the latter limitation is to avoid forcing the parties to relitigate long-ago events, threatening defendants with ‘harm akin to double jeopardy and denial of speedy trial.’ (Guerrero, supra, 44 Cal.3d at p. 355.) The Attorney General has not asked us to reconsider this aspect of our precedent.” (Gallardo, supra, at p. 138.)

12 In re MILTON Opinion of the Court by Jenkins, J.

III. Gallardo Does Not Apply Retroactively

A. General Principles California courts have applied two tests for retroactivity, often referred to as the federal and state tests. (In re Thomas (2018) 30 Cal.App.5th 744, 754 (Thomas).) Under both tests, a judicial decision that creates a “new rule” is generally not given retroactive effect in cases on collateral review that were final when the rule was announced. (Teague v. Lane (1989) 489 U.S. 288, 306 (Teague); Donaldson v. Superior Court (1983) 35 Cal.3d 24, 36.) Thus, the threshold question under both tests is whether a judicial decision constitutes a new rule. (In re Ruedas (2018) 23 Cal.App.5th 777, 799 (Ruedas).) A new rule, however, will nevertheless be given retroactive effect under either test if it is substantive, as opposed to procedural. (Teague, supra, 489 U.S. at p. 311; People v. Mutch (1971) 4 Cal.3d 389, 395–396.) Therefore, the second question under both tests is whether the new rule announced by the judicial decision is procedural or substantive. Under the federal test, rules that are both new and procedural do not apply retroactively to final judgments, without exception. (Edwards v. Vannoy (2021) 593 U.S. ___, [141 S.Ct. 1547, 1551–1552, 1560] (Edwards) [the United States Supreme Court’s decision striking down Louisiana’s nonunanimous jury verdict law is not retroactive to final judgments because it is a new procedural rule].) 3 In contrast,

3 Until recently, a new procedural rule could nevertheless be retroactive under the federal test if it was a “watershed” rule that altered “ ‘our understanding of the bedrock procedural elements that must be found to vitiate the fairness

13 In re MILTON Opinion of the Court by Jenkins, J.

in California, a new procedural rule may nevertheless be retroactive under People v. Johnson (1970) 3 Cal.3d 404 (Johnson), a case in which this court recited three factors the United States Supreme Court had previously instructed courts to consider in determining whether a new rule applies retroactively to cases predating the announcement of the rule: “ ‘ “(a) the purpose to be served by the new standards, (b) the extent of the reliance by law enforcement authorities on the old standards, and (c) the effect on the administration of justice of a retroactive application of the new standards.” ’ ” (Id. at p. 410; In re Lopez (2016) 246 Cal.App.4th 350, 359, fn. 2 [“[t]he California Supreme Court has articulated a different three-part test for determining retroactivity of judicial opinions involving questions of procedure”].) “The application of [the various] procedural bars and limitations on the retroactivity of changes in the criminal law serves to protect the finality of judgments on collateral review.” (In re Martinez (2017) 3 Cal.5th 1216, 1222 (Martinez); see Teague, supra, 489 U.S. at p. 306 [the government has a

of a particular conviction.’ ” (Teague, supra, 489 U.S. at p. 311.) Historically, there was such a high bar for a rule to be considered “watershed” that Gideon v. Wainwright (1963) 372 U.S. 335, which held that indigent defendants have the right to counsel, was the only case that had met this exception. (Whorton v. Bockting (2007) 549 U.S. 406, 418.) Last year, the high court in Edwards, supra, 593 U.S. ___, [141 S.Ct. at pp. 1551–1552, 1560] acknowledged how narrow the watershed exception has been and eliminated it. In light of Edwards, which was decided six months after petitioner filed his reply brief, petitioner withdrew his argument that Gallardo announced a watershed rule.

14 In re MILTON Opinion of the Court by Jenkins, J.

legitimate interest in having judgments remain final, and collateral review “ ‘is not designed as a substitute for direct review’ ”].) Some California courts have applied the federal retroactivity test in deciding the retroactivity of new procedural rules (e.g., In re Moore (2005) 133 Cal.App.4th 68, 77 [new procedural rule not retroactive under Teague]), while others have applied the state Johnson test (e.g., Thomas, supra, 30 Cal.App.5th at pp. 760–761 [new procedural rule not retroactive under Johnson]) or both tests (e.g., Ruedas, supra, 23 Cal.App.5th at p. 799). We have neither explicitly disavowed Johnson nor explicitly adopted the federal Teague test for state collateral review proceedings (In re Hansen (2014) 227 Cal.App.4th 906, 910), and we need not decide in this case what test applies to new procedural rules, because we reach the same conclusion under both tests. B. Gallardo Is a New Rule Under Both Federal and State Law Under federal law, a case will generally be considered to have announced a new rule where it “breaks new ground,” i.e., “the result was not dictated by precedent existing at the time the defendant’s conviction became final.” (Teague, supra, 489 U.S. at p. 301.) The result in a given case is not dictated by precedent if it is “susceptible to debate among reasonable minds” (Butler v. McKellar (1990) 494 U.S. 407, 415) or if “reasonable jurists may disagree” (Sawyer v. Smith (1990) 497

U.S. 227, 234).

The California cases such as Guerrero that existed at the time petitioner’s conviction became final did not dictate our decision in Gallardo because Gallardo invalidated the prior procedure that Guerrero and other cases had approved. Nor

15 In re MILTON Opinion of the Court by Jenkins, J.

did federal precedent such as Apprendi, which existed at the time petitioner’s conviction became final, dictate our Gallardo decision. In McGee, which we decided after Apprendi, we expressly distinguished Apprendi and declined to construe it in a manner that afforded a defendant the right to a jury trial on the nature of a prior conviction. (McGee, supra, 38 Cal.4th at pp. 696–697.) In other words, we did not view Apprendi and other existing United States Supreme Court precedent as dictating the result we later reached in Gallardo. Furthermore, as we stated in Gallardo, the law regarding a sentencing court’s authority to make factual findings about prior convictions was unsettled after Apprendi was decided: “In the wake of Apprendi, questions arose about the scope of the so-called Almendarez-Torres exception to the general Sixth Amendment rule forbidding judicial factfinding in criminal cases.” (Gallardo, supra, 4 Cal.5th at p. 128; see id. at pp. 127, 138 [acknowledging it would have been difficult for the parties in Gallardo to know how Apprendi and Descamps would have affected their case].) We conclude that Gallardo announced a new rule under the federal test because precedent that existed at the time petitioner’s conviction became final did not dictate our decision in Gallardo.4

4 Petitioner argues Gallardo is nevertheless retroactive under In re Gomez (2009) 45 Cal.4th 650 (Gomez), but that case is distinguishable. There, this court decided the retroactivity of the high court’s holding in Cunningham v. California (2007) 549 U.S. 270 (Cunningham) — that California’s determinate sentencing laws violated the rule articulated in Blakely v. Washington (2004) 542 U.S. 296 (Blakely) that an aggravating fact used to increase a sentence

16 In re MILTON Opinion of the Court by Jenkins, J.

Under California law, a rule is new where the decision “(1) explicitly overrules a precedent of [the California Supreme Court] [citation], or (2) disapproves a practice impliedly sanctioned by prior decisions of [the California Supreme Court] [citation], or (3) disapproves a longstanding and widespread practice expressly approved by a near-unanimous body of lower-court authorities.” (People v. Guerra (1984) 37 Cal.3d 385, 401 (Guerra); see Donaldson v. Superior Court, supra, 35 Cal.3d at p. 36 [a decision that “only elucidates and enforces prior law” does not create new law].) Petitioner concedes, and we agree, that “Gallardo satisfies this threshold inquiry, as it disapproved prior California Supreme Court law” by “overrul[ing] the prior precedent set forth in Guerrero and McGee.” In Gallardo, we explicitly overruled McGee, which had, until then, provided controlling precedent on a trial court’s authority to make

beyond the standard range had to be found by a jury. Blakely had already been decided by the time the petitioner was sentenced, but Cunningham had not. (Gomez, supra, 45 Cal.4th at p. 653.) This court held in Gomez that Cunningham was retroactive to the petitioner’s final judgment because it did not announce a “new rule,” but instead was “dictated” by Blakely. (Gomez, supra, 45 Cal.4th at p. 658.) In so concluding, this court emphasized that the high court made it clear in Cunningham that it was “simply appl[ying]” Blakely’s “ ‘bright-line-rule’ ” to California’s sentencing laws and that it “d[id] not view its application of Blakely to California law as an extension or modification of the [Blakely] rule.” (Gomez, supra, 45 Cal.4th at pp. 660, 658.) In contrast, in deciding Gallardo, we did not “simply appl[y]” Apprendi or other precedent that existed at the time petitioner’s conviction became final. (Id. at p. 660.)

17 In re MILTON Opinion of the Court by Jenkins, J.

factual findings when determining whether a prior conviction qualified as a strike. (See Saffle v. Parks (1990) 494 U.S. 484, 488 [“[t]he explicit overruling of an earlier holding no doubt creates a new rule”].) We did not merely “explain or refine the holding of a prior case, . . . apply an existing precedent to a different fact situation, . . . or . . . draw a conclusion that was clearly implied in or anticipated by previous opinions.” (Guerra, supra, 37 Cal.3d at p. 399 [listing “common examples of decisions that do not establish a new rule of law”].) C. Gallardo Is a Procedural, Not Substantive Rule, Under Both Federal and State Law Federal and state cases distinguish between substantive and procedural rules in similar ways. We therefore review federal and state authorities relevant to this substantive- procedural dichotomy together and conclude the Gallardo rule is procedural under both the federal and state tests. Both federal and state cases have held that a rule is substantive rather than procedural where it “ ‘alters the range of conduct or the class of persons that the law punishes.’ ” (Welch v. United States (2016) 578 U.S. 120, 129 (Welch); see Martinez, supra, 3 Cal.5th at p. 1222.) “ ‘This includes decisions that narrow the scope of a criminal statute by interpreting its terms, as well as constitutional determinations that place particular conduct or persons covered by the statute beyond the State’s power to punish.’ ” (Welch, at p. 129; see In re Lopez, supra, 246 Cal.App.4th at p. 357.) In Welch, the high court considered the retroactivity of its holding in Johnson v. United States (2015) 576 U.S. 591, 597, invalidating as unconstitutionally vague a provision of the ACCA that defined “violent felony” as any felony that

18 In re MILTON Opinion of the Court by Jenkins, J.

“ ‘otherwise involves conduct that presents a serious potential risk of physical injury to another.’ ” (Welch, supra, 578 U.S. at p. 123.) The high court held this new rule was substantive because all defendants whose sentences had been enhanced under the now-invalidated provision were categorically excluded from enhanced punishment as a result of the new rule. In other words, the new rule “changed the substantive reach of the [ACCA], altering ‘the range of conduct or the class of persons [the ACCA] punishes.’ ” (Welch, at p. 129.) In Martinez, we held that our holding in People v. Chiu (2014) 59 Cal.4th 155 that a natural and probable consequences theory of liability can no longer serve as a basis for a first degree murder conviction was a substantive change in the law that applied retroactively to final judgments. (Martinez, supra, 3 Cal.5th at pp. 1222–1223.) As in Welch, the new rule we announced in Chiu was substantive because all defendants who had been convicted of first degree murder under the now- invalidated natural and probable consequences theory were categorically entitled to relief from their convictions as a result of the new rule. “Procedural rules, by contrast, ‘regulate only the manner of determining the defendant’s culpability.’ [Citation.] Such rules alter ‘the range of permissible methods for determining whether a defendant’s conduct is punishable.’ [Citation.] ‘They do not produce a class of persons convicted of conduct the law does not make criminal, but merely raise the possibility that someone convicted with use of the invalidated procedure might have been acquitted otherwise.’ ” (Welch, supra, 578 U.S. at p. 129; see Schriro v. Summerlin (2004) 542 U.S. 348, 352 (Schriro) [procedural rules have a “more speculative connection to innocence”].) “If a new rule regulates only the

19 In re MILTON Opinion of the Court by Jenkins, J.

procedures for determining culpability, the Teague balance generally tips in favor of finality. The chance of a more accurate outcome under the new procedure normally does not justify the cost of vacating a conviction whose only flaw is that its procedures ‘conformed to then-existing constitutional standards.’ ” (Welch, supra, 578 U.S. at p. 131.) Some examples of procedural rules are ones that “alter[] only the procedures used to obtain the conviction,” “ ‘allocate decisionmaking authority’ between judge and jury, [citation], or regulate the evidence that the court [may] consider in making its decision.” (Id. at pp. 131, 130.) In its most recent retroactivity case, the United States Supreme Court stated that its decision striking down Louisiana’s nonunanimous jury verdict law, which permits conviction by a 10-2 vote, was a procedural rule because it “alter[ed] ‘only the manner of determining the defendant’s culpability.’ ” (Edwards, supra, 593 U.S. ___, [141 S.Ct. at p. 1562], quoting Schriro, supra, 542 U.S. at p. 353; see Jones v. Mississippi (2021) ___ U.S. ___, ___ [141 S.Ct. 1307, 1318, fn. 4].) Similarly, the rule from Crawford v. Washington (2004) 541 U.S. 36 that the confrontation clause gives defendants the right to confront and cross-examine witnesses was procedural because, while it narrowed the class of persons subject to criminal punishment, it merely changed the “procedure for determining whether the admission of hearsay statements violated the confrontation clause.” (In re Moore, supra, 133 Cal.App.4th at p. 75; see Whorton v. Bockting, supra, 549 U.S. at p. 417 [“it is clear and undisputed that the [Crawford] rule is procedural and not substantive”]; Ruedas, supra, 23 Cal.App.5th at p. 793 [the rule from People v. Sanchez (2016) 63 Cal. 4th 665 that an expert witness’s out-of-court

20 In re MILTON Opinion of the Court by Jenkins, J.

testimonial statements about case-specific facts violates the confrontation clause (Sanchez rule) is procedural, not substantive].) Petitioner argues the Gallardo rule is substantive because, “[b]y limiting imposition of an increased sentence to circumstances where the prior conviction itself, as distinct from the underlying conduct, supports the increased sentence,” the rule “ ‘alter[ed] the range of conduct or the class of persons that the law punishes’ ” and “effectively ‘modifie[d] the elements’ [citation] of the prior conviction allegation.” We disagree. Unlike Welch, Martinez, and other cases in which courts have found rules to be substantive in nature, Gallardo did not change the law in a way that excludes certain types of offenders from the reach of the law (here, the Three Strikes law) nor did it categorically provide relief to all defendants who were sentenced under the pre-Gallardo rule. 5 Nor did our decision in Gallardo “ ‘narrow the scope of a criminal statute’ ” (Welch, supra, 578 U.S. at p. 129) for example, by eliminating out-of-state prior convictions or convictions involving personal use of a firearm or a deadly weapon as qualifying serious felonies. Rather, the Gallardo rule “regulate[d] the evidence that the court could consider” in making prior conviction

5 For example, in both In re Scott, supra, 49 Cal.App.5th at p. 1019, review granted, and In re Haden, supra, 49 Cal.App.5th at p. 1107 (conc. opn. of Tucher, J.), review granted, the Courts of Appeal held that Gallardo is not retroactive but stated that even if it were, the petitioners in those cases would not be entitled to relief.

21 In re MILTON Opinion of the Court by Jenkins, J.

determinations (Welch, supra, 578 U.S. at p. 130) by precluding courts from looking at anything other than “those facts that were established by virtue of the [prior] conviction itself — that is, facts the jury was necessarily required to find to render a guilty verdict, or that the defendant admitted as the factual basis for a guilty plea” (Gallardo, supra, 4 Cal.5th at p. 136). Petitioner argues that “the class of persons who may be subject to the punishment has been limited as well” because, by limiting the evidence on which a court may rely in making prior conviction determinations, Gallardo effectively “narrow[ed] the universe of the defendants for whom a sentence can be enhanced based on a prior conviction.” However, most, if not all, new procedural rules — such as the high court’s invalidation of Louisiana’s nonunanimous jury verdict law (Edwards, supra, 593 U.S. ___, [141 S.Ct. 1547]) or the Crawford and Sanchez rules — likely “narrow[] the universe of defendants” “subject to the punishment” in the sense that fewer defendants will have been convicted if the new rules had been in effect at the time they were tried. However, as discussed above, more is required before we may conclude that a new rule is substantive in nature. A new rule is not substantive merely because it “ ‘raise[s] the possibility that someone convicted with use of the invalidated procedure might have been acquitted otherwise.’ ” (Welch, supra, 578 U.S. at p. 129.) Notably, we described the trial court’s error in Gallardo as one concerning the “form” of judicial factfinding, stating, “[T]he court engaged in a form of factfinding that strayed beyond the bounds of the Sixth Amendment” in finding the defendant used a knife. (Gallardo, supra, 4 Cal.5th at p. 136.)

22 In re MILTON Opinion of the Court by Jenkins, J.

And, consistent with our view of the court’s error in Gallardo, we remanded the matter, instructing the trial court to reconsider its factual finding of the defendant’s knife use “in accordance with this opinion,” that is, by considering only “ ‘the record of the prior proceeding’ ” to determine “ ‘what facts [the defendant] necessarily admitted in entering her plea.’ ” (Id. at pp. 140, 130, 137.) We did not remove the defendant or any group of people from the reach of applicable sentencing laws, and we did not conclude that the conduct underlying her prior conviction, i.e., whether she used a knife, was no longer relevant. Instead, we described the procedural changes that would apply to the court’s determination on remand of whether the prior conviction qualified as a prior strike.6 We therefore conclude the Gallardo rule, which “prescribe[d] the manner of finding facts to increase the defendant’s sentence” by changing “ ‘ “the range of permissible methods for determining whether a defendant’s conduct is punishable,” ’ ” was procedural in nature. (Milton, supra, 42 Cal.App.5th at p. 992; In re Brown, supra, 45 Cal.App.5th at p. 716, review granted [disagreeing with the Milton court on the ultimate question of Gallardo’s retroactivity but agreeing that the new rule that Gallardo announced was procedural, not substantive].)

6 Although the offense in Gallardo involved a divisible offense, Gallardo did not decide that the Three Strikes law could have no possible application in cases involving indivisible offenses. Accordingly, the Court of Appeal in Milton stated that if Gallardo applied retroactively, petitioner would not be automatically entitled to relief, but the matter would need to be remanded for a redetermination of the prior offenses under the procedure announced in Gallardo. Other Courts of Appeal have reached similar conclusions after Gallardo.

23 In re MILTON Opinion of the Court by Jenkins, J.

Our conclusion is buttressed by cases that have held that rules protecting a defendant’s Sixth Amendment right to a jury finding on facts needed to support increased sentences are procedural. “[T]he United States Supreme Court has made it clear that Apprendi, and cases following it, did not alter state substantive law.” (People v. Anderson (2009) 47 Cal.4th 92, 118.) In People v. Amons (2005) 125 Cal.App.4th 855, 865, the Court of Appeal held that the holding in Blakely that an aggravating fact other than a prior conviction used to increase a sentence beyond the statutory standard range must be found by a jury “is a procedural rule that affects only the manner of determining the defendant’s punishment.” (See Schardt v. Payne (9th Cir. 2005) 414 F.3d 1025, 1036 [Blakely, which “allocated some of the decision-making authority previously held by judges to juries,” is a procedural rule].) And in Schriro, supra, 542 U.S. at page 358, the United States Supreme Court explained that its holding in Ring v. Arizona (2002) 536 U.S. 584, that a defendant has a Sixth Amendment right to have a jury find facts sufficient to impose the death penalty, was a procedural rule.7

7 The cases on which petitioner relies in arguing the Gallardo rule was substantive do not support his position. In Montgomery v. Louisiana (2016) 577 U.S. 190, the United States Supreme Court held that a new rule invalidating mandatory sentencing schemes that require juveniles to be sentenced to life in prison without parole (LWOP) applied retroactively. In People v. Trujeque (2015) 61 Cal.4th 227, we held that a decision barring prosecution of a juvenile as an adult after the juvenile court has commenced adjudicatory proceedings applied retroactively. In both cases, the new rules exempted a whole category of people (juveniles) from ever

24 In re MILTON Opinion of the Court by Jenkins, J.

facing trial or receiving mandatory LWOP sentences and redefined the class of people who could be punished. Allen v. Ives (9th Cir. 2020) 950 F.3d 1184, which petitioner cites, as well as Holt v. U.S. (7th Cir. 2016) 843 F.3d 720 and Hill v. Masters (6th Cir. 2016) 836 F.3d 591, on which Justice Liu relies, are also distinguishable. In Allen v. Ives, the court stated that Descamps and Mathis announced a substantive rule because they “alter[ed] ‘the range of conduct’ ” the law punished. (Allen v. Ives, supra, 950 F.3d at p. 1192.) As we noted above, in Descamps and Mathis, the high court interpreted the ACCA, an elements-based statutory scheme, in a way that limited its substantive reach to prior convictions that matched the elements of an ACCA offense. (Descamps, supra, 570 U.S. at pp. 277–278 [California burglary is not a qualifying offense under the ACCA]; Mathis, supra, 579 U.S. at p. 509 [Iowa burglary is not a qualifying offense under the ACCA].) Thus, because of the Descamps and Mathis decisions, an entire category of defendants, i.e., those whose sentences were enhanced under the ACCA based on prior California or Iowa burglary convictions, are no longer subject to sentence enhancements based on those prior convictions. In contrast to the ACCA, the Three Strikes law does not define qualifying offenses strictly by their elements but looks to the conduct underlying the offense. (See, e.g., Guerrero, supra, 44 Cal.3d at p. 352.) For example, unlike the ACCA, section 1192.7 defines as “serious felonies” offenses that do not correspond to any established offense under California law; these offenses instead focus on the factual content of the crime, such as gun use. In other words, these offenses would never be established merely by the fact of conviction or by consideration of formal elements alone. Because it is the Three Strikes law itself that required (and still requires) a judge to consider “the nature of the conduct underlying a prior conviction” when determining whether that conviction qualifies as a strike, Gallardo was not, and could not have been, a “state law analog to Descamps and Mathis” as Justice Liu asserts it was intended to be. (Dis. opn. of Liu, J., post, at p. 5.) In Gallardo, which involved application of the Three Strikes law, we did not

25 In re MILTON Opinion of the Court by Jenkins, J.

Because Gallardo announced a new procedural rule, it is not retroactive to final judgments under the federal retroactivity test. (Edwards, supra, 593 U.S. ___, [141 S.Ct. at pp. 1551–1552, 1560] [under the federal test, new rules of criminal procedure are not retroactive on collateral review, without exception].) D. Gallardo Is Also Not Retroactive Under the State Johnson Test We likewise conclude Gallardo is not retroactive under the state test we announced in Johnson. Under Johnson, the retroactivity of a new rule is determined by “ ‘ “(a) the purpose to be served by the new standards, (b) the extent of the reliance by law enforcement authorities on the old standards, and (c) the effect on the administration of justice of a retroactive application of the new standards.” ’ ” (Johnson, supra, 3 Cal.3d at p. 410.) The first factor — the purpose of the new rule — is the critical factor in determining retroactivity. The other “factors of reliance and burden on the administration of justice are of significant relevance only when the question of retroactivity is a close one after the purpose of the new rule is considered.” (Ibid.; see Guerra, supra, 37 Cal.3d at pp. 401–402.)

make substantive changes to the Three Strikes law (pp. 20–23, ante) but instead addressed the procedural issue of how the court should go about making the determination of whether a prior conviction counts as a strike. Namely, we altered the source material from which a judge could draw in making that determination: after Gallardo, a judge may consider only those “facts that were necessarily found in the course of entering the conviction.” (Gallardo, supra, 4 Cal.5th at p. 134.)

26 In re MILTON Opinion of the Court by Jenkins, J.

Johnson is referred to as “the old federal standard” (Ruedas, supra, 23 Cal.App.5th at p. 799) because it is based on factors the United States Supreme Court articulated in a number of retroactivity cases beginning with Linkletter v. Walker (1965) 381 U.S. 618, 629 (Linkletter). In Linkletter, the high court set forth several factors relevant to the retroactivity analysis, including “the prior history of the rule in question, its purpose and effect, and whether retrospective operation will further or retard its operation.” (Linkletter, supra, 381 U.S. at p. 629.) The court noted that “in each of the three areas in which we have applied our rule retrospectively the principle that we applied went to the fairness of the trial — the very integrity of the fact-finding process.” (Id. at p. 639, fn. omitted.) And the court suggested that this basic-fairness-and- reliability test would apply differently on direct and collateral review. While “a change in law will be given effect while a case is on direct review,” the “effect of the subsequent ruling of invalidity on prior final judgments when collaterally attacked is subject to no set ‘principle of absolute retroactive invalidity’ but depends upon a consideration of” multiple factors, including “ ‘public policy in the light of the nature both of the statute and of its previous application.’ ” (Id. at p. 627, italics added.) Two years after Linkletter, in Stovall v. Denno (1967) 388 U.S. 293, 297, the high court refined its analysis into a three-part test based in part on the factors listed in Linkletter: (a) the purpose to be served by the new standards; (b) the extent of the reliance by law enforcement authorities on the old standards; and (c) the effect on the administration of justice of a retroactive application of the new standards. In 1970, we adopted the Linkletter-Stovall approach in Johnson and surveyed the high court’s retroactivity decisions

27 In re MILTON Opinion of the Court by Jenkins, J.

for guidance on how to apply the three factors. (Johnson, supra, 3 Cal.3d at p. 410.) In discussing the high court’s retroactivity cases, we — like the high court — noted that in many of those cases, concerns about basic fairness determined whether a new rule was retroactive. We observed, for example, that the high court gave retroactive effect to new rules protecting a defendant’s right to counsel at various stages of a trial “because denial of counsel ‘must almost invariably deny a fair trial.’ ” (Id. at p. 411.) We noted that in contrast, the high court did not give retroactive effect to a rule that provided defendants with the right to counsel at lineup because “the absence of counsel does not render a lineup unfair per se.” (Id. at p. 412.) We also noted that cases requiring juries, rather than judges, to decide serious criminal cases were not retroactive “because although those cases recognized that juries may serve to prevent arbitrariness and repression, they did not rest on any assumption that nonjury trials are more likely than jury trials to be unfair or unreliable.” (Ibid., citing DeStefano v. Woods (1968) 392 U.S. 631.) Based on that review of cases from the high court, we also stated that reliability was a core concern: “Fully retroactive decisions are seen as vindicating a right which is essential to a reliable determination of whether an accused should suffer a penal sanction. . . . [¶] On the other hand, decisions which have been denied retroactive effect are seen as vindicating interests which are collateral to or relatively far removed from the reliability of the fact-finding process at trial.” (Johnson, supra, 3 Cal.3d at pp. 411–412.) Meanwhile, during the same period, disagreements developed at the high court over the Linkletter-Stovall approach. The primary disagreement was sparked by the

28 In re MILTON Opinion of the Court by Jenkins, J.

court’s post-Linkletter conflation of direct and collateral review for retroactivity purposes. Justice Harlan, in two influential minority opinions, criticized this development and emphasized the need to return to Linkletter’s original distinction between cases on collateral and direct review, with much more stringent retroactivity standards to be applied to cases on collateral review. (Desist v. United States (1969) 394 U.S. 244, 258 (Desist) (dis. opn. of Harlan, J.); Mackey v. United States (1971) 401 U.S. 667, 692–693, 696 (Mackey) (conc. & dis. opn. of Harlan, J.).) 8 The high court largely adopted Justice Harlan’s proposals by eventually holding that new rules will apply retroactively on direct review (Griffith v. Kentucky (1987) 479 U.S. 314, 322), while they will only rarely be applied retroactively on collateral review (Teague, supra, 489 U.S. at pp. 308, 310, 317, 319–320 [emphasizing the need to respect the finality of judgments on collateral review]). Following the United States Supreme Court’s decision in Teague, we have not had much occasion to apply Johnson’s

8 In these opinions, Justice Harlan stated that new substantive rules should apply retroactively, while new procedural rules should be retroactive only where the procedures used were “so fundamentally devoid of the necessary elements of procedural due process as to require upsetting [the] conviction in spite of the fact that it was perfectly lawful when made final.” (Mackey, supra, 401 U.S. at pp. 699–700 (conc. & dis. opn. of Harlan, J.).) He also emphasized the importance of the finality of judgments and urged that courts should apply new rules retroactively on direct review but should not do so on collateral review except in very limited situations. (Desist, supra, 394 U.S. at p. 258 (dis. opn. of Harlan, J.).)

29 In re MILTON Opinion of the Court by Jenkins, J.

three-part test. It appears the last time we did so was in People v. Carrera (1989) 49 Cal.3d 291, 327–328, an automatic appeal we decided six months after the high court decided Teague. However, a review of our post-Johnson cases, including more recent retroactivity cases from our court, indicates that we have incorporated some of the key developments in retroactivity law that the high court made in the post-Johnson years — developments that are largely consistent with Justice Harlan’s understanding of what Linkletter, the landmark case that ultimately gave rise to Johnson, originally required. For example, in Johnson, we did not originally distinguish between cases on direct review and collateral review. But nearly all of our post-Johnson cases applying the Johnson test concerned retroactive application of a rule to nonfinal convictions. We later recognized a distinction between direct and collateral review in Guerra, where we stated that “even a . . . decision . . . that cannot serve as a basis for collateral attack on a final judgment . . . ordinarily governs all cases . . . on direct review . . . .” (Guerra, supra, 37 Cal.3d at p. 400; see Martinez, supra, 3 Cal.5th at p. 1222 [the various “procedural bars and limitations on the retroactivity of changes in the criminal law serves to protect the finality of judgments on collateral review”].) We also did not distinguish between substantive and procedural rules in Johnson, but in cases decided post-Johnson, we have made that distinction clear by adopting the view that all substantive rules are retroactive. (E.g., Martinez, supra, 3 Cal.5th at pp. 1222–1223.) And, in the years following Johnson, we have incorporated some of the high court’s statements regarding fairness of the proceedings, focusing on both fairness and reliability as integral aspects of

30 In re MILTON Opinion of the Court by Jenkins, J.

our retroactivity determinations. (E.g., People v. Meyers (1987) 43 Cal.3d 250, 267 [a new rule invalidating a jury selection procedure was not retroactive to cases in which juries were selected before the rule was announced because its purpose was to further 6th Amend. values, and the new rule “ ‘did not rest on the premise that every criminal trial’ ” conducted under the former jury selection procedure “ ‘was necessarily unfair’ ”]; People v. McDaniel (1976) 16 Cal.3d 156, 166, fn. 4 [citing Stovall for the proposition that denial of counsel not only would “adversely affect the truth-finding process” but would also “almost invariably lead to a denial of a fair trial”].) With a fuller appreciation for the pre- and post-Johnson developments that provide the framework for our analysis, including the added importance we place on the finality of judgments when evaluating the retroactivity of cases on collateral review, we return to the question of whether the Gallardo rule is retroactive under the Johnson test. As we stated in Guerra — which concerned a nonfinal case on direct review — for a new rule to apply retroactively, its “ ‘major’ ” or “primary purpose” must be “to promote reliable determinations of guilt or innocence,” i.e., “ ‘ “ ‘to overcome an aspect of the criminal trial that substantially impairs its truth- finding function and so raises serious questions about the accuracy of guilty verdicts.’ ” ’ ” (Guerra, supra, 37 Cal.3d at pp. 402, 403.) For these purposes, the threshold for applying a case retroactively on collateral review is necessarily demanding, given the important systemic interests in the stability and finality of judgments. (See Martinez, supra, 3 Cal.5th at p. 1222.) Petitioner argues Gallardo is retroactive under this standard because the “fundamental purpose of Gallardo is to promote fair and reliable determinations of the

31 In re MILTON Opinion of the Court by Jenkins, J.

defendant’s guilt or innocence on the allegation that he suffered a prior conviction qualifying as a strike under California law.” We disagree. Although the rule we announced in Gallardo modified the permissible procedures for finding facts about a defendant’s prior convictions, the factfinding procedures in place prior to Gallardo did not lack basic integrity or fairness in a manner akin to the practices the Linkletter court identified as paradigmatic examples of basic unfairness: denying an indigent defendant an attorney, foreclosing a criminal appeal because of inability to pay, or using an unfair procedure for determining whether a confession admitted in evidence is actually voluntary. (See Linkletter, supra, 381 U.S. at p. 639, fn. 20.) During the many years in which it was the sentencing court’s role to make findings about the nature of prior convictions, for example, a pre-Gallardo sentencing court “still had to apply the beyond-a-reasonable-doubt standard of proof” in determining whether a prior conviction was a serious or violent felony. (Milton, supra, 42 Cal.App.5th at pp. 995, citing People v. Frierson (2017) 4 Cal.5th 225, 233; see People v. Woodell (1998) 17 Cal.4th 448, 461 [there was evidence to support a finding, beyond a reasonable doubt, that the defendant’s out-of-state prior conviction involved the use of a deadly weapon for purposes of determining whether the prior conviction was a serious felony].) In addition, a pre-Gallardo sentencing court was restricted to reviewing the record of the prior conviction and “no further,” which ensured the court would not base its determination on potentially unreliable information outside the record of conviction. (Guerrero, supra, 44 Cal.3d at p. 355 [noting the restriction is “fair” and “reasonable”].)

32 In re MILTON Opinion of the Court by Jenkins, J.

Moreover, there were other safeguards in place to ensure the sentencing court would not base its findings on unreliable material in the record of conviction. In People v. Reed (1996) 13 Cal.4th 217, for example, we held the sentencing court erred in relying on a probation report in determining the conduct underlying the defendant’s prior conviction (i.e., that he used a deadly weapon) because a court is permitted to rely only on documents that “reliably reflect[] the facts” of the prior offense. (Id. at p. 223.) We further observed that in contrast to the unreliable probation report, a preliminary hearing transcript on which the sentencing court relied was sufficiently reliable “because the procedural protections afforded the defendant during a preliminary hearing tend to ensure the reliability of such evidence. Those protections include the right to confront and cross-examine witnesses and the requirement those witnesses testify under oath, coupled with the accuracy afforded by the court reporter’s verbatim reporting of the proceedings.” (Ibid.) Because there were various safeguards in place that rendered the pre-Gallardo procedure reasonably fair and reliable, it cannot be said that our “ ‘major’ ” or “primary purpose” in announcing the Gallardo rule was “ ‘ “ ‘to overcome an aspect of the criminal trial that substantially impairs its truth-finding function and so raises serious questions about the accuracy of guilty verdicts.’ ” ’ ” (Guerra, supra, 37 Cal.3d at pp. 402, 403.) Petitioner asserts that reliability and fairness must have been core concerns of our decision in Gallardo because, in rejecting Justice Chin’s proposed remedy of a remand for a jury trial, the Gallardo majority noted that procedural safeguards such as cross-examination of witnesses would not be in place if we were to empanel a jury for the purpose of reviewing the

33 In re MILTON Opinion of the Court by Jenkins, J.

record of conviction from the prior case. (Citing In re Haden, supra, 49 Cal.App.5th at p. 1103 (conc. opn. of Tucher, J.), review granted [the Gallardo court’s “response to [Justice Chin’s dissent] reveals the breadth of interests at stake].) 9 We do not believe the Gallardo majority’s rejection of the remedy proposed by Justice Chin can bear the weight petitioner places upon it. Notably, neither the Gallardo majority nor Justice Chin expressed any concern that the sentencing court’s factfinding about the defendant’s knife use substantially impaired the truth-finding function as compared to what a jury in the prior proceeding might have found.10 Instead, we stated in Gallardo

9 Amicus curiae Office of State Public Defender makes a similar argument: “[I]f this Court believed the only issue was the Sixth Amendment limit on judicial factfinding, it would have embraced the dissent’s proposed remedy and shifted factfinding to a sentencing jury.” 10 “[J]udicial factfinding is not inherently unreliable or less reliable than jury factfinding.” (Milton, supra, 42 Cal.App.5th at p. 998.) “[A]lthough . . . cases recognize[] that juries may serve to prevent arbitrariness and repression, they d[o] not rest on any assumption that nonjury trials are more likely than jury trials to be . . . unreliable.” (Johnson, supra, 3 Cal.3d at p. 412.) “[F]or every argument why juries are more accurate factfinders [than judges], there is another why they are less accurate.” (Schriro, supra, 542 U.S. at p. 356; see id. at p. 357 [“it is hard to see how a trial in which a judge finds only aggravating factors could be” “impermissibly inaccurate”]; In re Consiglio (2005) 128 Cal.App.4th 511, 515 [same].) Even the majority of the Court of Appeal in In re Brown, which held Gallardo is retroactive, acknowledged that a court’s factfinding regarding the conduct underlying a prior conviction would not be more reliable than that of a jury: “We recognize that the factfinding process might not be any less reliable if

34 In re MILTON Opinion of the Court by Jenkins, J.

that we were going to “reconsider McGee” because the high court’s “further explication” of Sixth Amendment principles in Descamps and Mathis informed us that a defendant has the right to have a jury determine all facts relating to the nature of a prior conviction that are used to impose an increased sentence. (Gallardo, supra, 4 Cal.5th at pp. 124, 136.) The high court’s post-McGee decisions instructed us that a sentencing court was no longer permitted to make certain factual findings, even if those findings might have been entirely reliable, because a defendant has a Sixth Amendment right to have a jury of his peers make those types of findings. (See Blakely, supra, 542 U.S. at pp. 301, 313 [a defendant’s right to a jury “ ‘of his equals and [neighbors]’ ” is based on the “Framers’ paradigm for criminal justice” — “the common-law ideal of limited state power accomplished by strict division of authority between judge and jury”].) Thus, the “ ‘major’ ” or “primary purpose” of the Gallardo rule (Guerra, supra, 37 Cal.3d at p. 402) was not to “vindicat[e] a right [that] is essential to a reliable determination of whether an accused should suffer a penal sanction” (Johnson, supra, 3 Cal.3d at p. 411) but to ensure our procedure was consistent with the high court’s “further explication” of “Sixth Amendment principles” (Gallardo, at pp. 136, 124). Petitioner asserts the prior procedure, which allowed courts to enhance sentences based on factfinding regarding the conduct underlying prior convictions, was “fundamentally unfair” and raised reliability concerns because defendants

conducted by the sentencing judge, and might even be better.” (In re Brown, supra, 45 Cal.App.5th at pp. 721–722.)

35 In re MILTON Opinion of the Court by Jenkins, J.

lacked the incentive to contest facts regarding that conduct in the prior proceedings. We disagree, both as to petitioner’s case and as a general matter. The record here shows that the Illinois sentencing court relied on petitioner’s gun use in imposing an aggravated sentence on the armed robbery conviction. Because gun use could result in a longer sentence, petitioner would have had the incentive to contest it at his original sentencing and presumably would have done so if there were any question whether he used a gun in the commission of either robbery. More to the point, California defendants prior to Gallardo, in challenging whether the prosecution had proven a fact about a prior conviction beyond a reasonable doubt, could raise their lack of incentive to challenge that fact during the original proceedings, in the course of arguing the beyond-a-reasonable-doubt standard was not satisfied. (See People v. Smith (1988) 206 Cal.App.3d 340, 346 [“Smith was not barred from presenting evidence the burglaries were not residential; his motive to do so was strong, particularly since his trial occurred pre-Alfaro”]; People v. Johnson (1989) 208 Cal.App.3d 19, 24 [“Since appellant’s 1983 convictions occurred after section 667 was enacted, the residential allegations in the information were neither irrelevant nor superfluous. Given the consequences of section 667, appellant had ample reason to contest the residential nature of the burglary charged when he pled guilty”]; People v. Skeirik (1991) 229 Cal.App.3d 444, 464 [“Both defendant and the state had an important incentive to contest the designation of his first degree conviction” as having been related to a deadly weapon, given statute that, at the time of conviction, forbade probation for defendant “ ‘who at the time of the

36 In re MILTON Opinion of the Court by Jenkins, J.

perpetration of said crime . . . was himself armed with a deadly weapon’ ”].) We are also mindful that courts prior to Gallardo were attempting to navigate a different fairness concern: the concern that courts should treat conduct underlying in-state and out-of-state offenses identically for purposes of our state’s recidivist statutes. The factfinding procedures in place prior to Gallardo — though ultimately inconsistent with the Sixth Amendment principles upon which our decision in Gallardo rested — were intended, in part, to avoid the unfairness of treating defendants with essentially identical underlying conduct very differently under California’s recidivist sentencing laws. Given minor variations in the way different jurisdictions define criminal offenses — as in this case11 — our pre-Gallardo decisions reflected a concern that a strictly elements-based approach to assessing out-of-state prior convictions could unfairly treat defendants with essentially identical underlying conduct very differently simply because of the happenstance of where they committed their crimes. Our pre-Gallardo cases approved a different approach in part to ameliorate that fairness concern. Our cases had adopted the pre-Gallardo approach, in other words, in part to help ensure

11 The definitions of California robbery and Illinois robbery are very similar. (See 720 Ill. Comp. Stat. Ann. 5/18-1 [“A person commits robbery when he or she knowingly takes property . . . from the person or presence of another by the use of force or by threatening the imminent use of force”]; § 211 [“Robbery is the felonious taking of personal property in the possession of another, from his person or immediate presence, and against his will, accomplished by means of force or fear”].)

37 In re MILTON Opinion of the Court by Jenkins, J.

that “[a] defendant whose prior conviction was suffered in another jurisdiction” would be “subject to the same punishment as a person previously convicted of an offense involving the same conduct in California.” (People v. Myers (1993) 5 Cal.4th 1193, 1201.)12 Gallardo, of course, concluded the Constitution requires a different approach. But we are not, in sum, persuaded that the “ ‘major’ ” or “primary purpose” of the Gallardo rule was “ ‘ “ ‘to overcome an aspect of the criminal trial that substantially impairs its truth-finding function and so raises serious questions about the accuracy of guilty verdicts.’ ” ’ ” (Guerra, supra, 37 Cal.3d at pp. 402, 403.) Nor are we persuaded that our decision invalidated prior procedures that were fundamentally unfair or seriously undermined the accuracy or reliability of criminal sentencing procedures. (See Linkletter, supra, 381 U.S. at p. 639 & fn. 20.) As noted, the first Johnson factor is critical in determining retroactivity, and the second factor of law enforcement’s reliance on the old rule, and the third factor of the burden on the administration of justice “are of significant relevance only when the question of

12 The point here is not, as Justice Groban suggests (dis. opn. of Groban, J., post, at p. 13), that the two defendants convicted of robbery, one in Illinois and one in California, necessarily engaged in identical conduct simply because they were both convicted of robbery. The point is that two defendants who actually did commit identical conduct would necessarily be treated differently under our state’s recidivist statutes unless courts were, pre-Gallardo, permitted to look beyond the elements of each state’s definition of robbery when determining whether each defendant’s offense qualified as a strike.

38 In re MILTON Opinion of the Court by Jenkins, J.

retroactivity is a close one after the purpose of the new rule is considered.” (Johnson, supra, 3 Cal.3d at p. 410.) We conclude the first Johnson factor’s effect on the issue of retroactivity is determinative and that we need not discuss whether the second and third Johnson factors also weigh against applying Gallardo retroactively.

DISPOSITION

In light of all the pertinent considerations, we conclude the rule we announced in Gallardo is a new procedural rule and that it is not retroactive to cases on collateral review under both state and federal tests for retroactivity. Thus, it does not apply to petitioner’s final judgment. Accordingly, we affirm the decision of the Court of Appeal denying the petition for writ of habeas corpus. 13

13 Petitioner argues that “alternatively,” his sentence is unauthorized and may be corrected at any time because it violated “the Sixth Amendment principles discussed in Gallardo.” (Citing, e.g., United States v. Johnson (1982) 457 U.S. 537, 550; People v. Scott (1994) 9 Cal.4th 331, 354–355.) The argument is circular. If we conclude the Gallardo rule is retroactive, petitioner’s sentence was unlawful at the time of sentencing and is unauthorized. If the Gallardo rule is not retroactive, his sentence was lawful and is authorized. (See In re Brown, supra, 45 Cal.App.5th at p. 731 (dis. opn. of Menetrez, J.), review granted [“We cannot apply Gallardo retroactively to render the sentence legally unauthorized and then infer from that lack of legal authorization that Gallardo must be retroactive”], italics omitted & added.) We also reject petitioner’s argument that his sentence was unauthorized by Apprendi. For the reasons we have explained, Apprendi and other United States Supreme Court cases that existed when petitioner’s conviction became final did not dictate our decision in Gallardo. (See pt. III B, ante.)

39 In re MILTON Opinion of the Court by Jenkins, J.

JENKINS, J.

We Concur:

CANTIL-SAKAUYE, C. J.

CORRIGAN, J.

KRUGER, J.

GUERRERO, J.

40 In re MILTON

S259954

Dissenting Opinion by Justice Liu

I join the dissenting opinion of Justice Groban. As he explains (dis. opn. of Groban, J., post, at pp. 2‒4), our decision in People v. Gallardo (2017) 4 Cal.5th 120 (Gallardo) operates retroactively because it substantively altered what constitutes “ha[ving] been convicted of a serious felony” and thereby modified what counts as a “strike” under the “Three Strikes” sentencing law. (Pen. Code, § 667, subd. (a)(1); all undesignated statutory references are to this Code.) Moreover, our decision in People v. Johnson (1970) 3 Cal.3d 404 furnishes an independent basis for the retroactivity of the Gallardo rule because the purpose of the rule is to promote the reliability of factual determinations concerning prior convictions. (Dis. opn. of Groban, J., post, at pp. 4‒14.) I write separately to underscore how the characterization of Gallardo in today’s opinion may reopen serious questions as to the constitutionality of the Three Strikes law. Under the Three Strikes law, “[a] person convicted of a serious felony who previously has been convicted of a serious felony in this state or of any offense committed in another jurisdiction that includes all of the elements of any serious felony” shall receive an enhanced sentence. (§ 667, subd. (a)(1).) The term “serious felony” is defined by reference to section 1192.7, subdivision (c), which lists 42 types of criminal activity that qualify as serious felonies.

1 In re MILTON Liu, J., dissenting

Our decision in Gallardo marked an important shift in how this sentencing enhancement statute is to be applied. Before Gallardo, a defendant was subject to an enhanced sentence wherever the defendant had a prior conviction and the record of that conviction revealed that the “conduct [which] likely (or ‘realistically’) supported the defendant’s [prior] conviction” fell within the statutory definition of a serious felony. (Gallardo, supra, 4 Cal.5th at p. 124.) But under the new standard set out in Gallardo, that is not enough. After Gallardo, a defendant stands “convicted of a serious felony” within the meaning of section 667, subdivision (a)(1) only if the “facts that were necessarily found [by the trier of fact] in the course of entering the conviction” establish that the defendant engaged in conduct satisfying the serious felony definition. (Gallardo, at p. 134.) Under the pre-Gallardo regime, such a showing would have been sufficient, but it would not have been necessary. Accordingly, I understand Gallardo to have narrowed who counts as having been “convicted of a serious felony” under the Three Strikes law. (§ 667, subd. (a)(1).) This case requires us to decide whether Gallardo’s holding applies retroactively. The high court has held that a new rule must be applied retroactively where the rule is “substantive” rather than procedural (see Teague v. Lane (1989) 489 U.S. 288, 311) and that “[a] rule is substantive rather than procedural if it alters the range of conduct or the class of persons that the law punishes” (Schriro v. Summerlin (2004) 542 U.S. 348, 353). “This includes decisions that narrow the scope of a criminal statute by interpreting its terms.” (Id. at p. 351.) Gallardo did not change the range of conduct made subject to additional punishment under the Three Strikes law;

2 In re MILTON Liu, J., dissenting

the definition of “serious felony” was left untouched. But Gallardo did change the class of persons who can be so punished; our decision narrowed the scope of the sentencing enhancement statute by interpreting its terms. Like Justice Groban, I would hold that Gallardo announced a substantive rule that must be given retroactive effect under Teague. Today’s opinion holds that Gallardo’s rule is merely procedural and “did not remove the defendant or any group of people from the reach of applicable sentencing laws.” (Maj. opn., ante, at p. 23.) “Rather,” this court says, “the Gallardo rule ‘regulate[d] the evidence that the court could consider’ in making prior conviction determinations [citation] by precluding courts from looking at anything other than ‘those facts that were established by virtue of the [prior] conviction itself — that is, facts the jury was necessarily required to find to render a guilty verdict, or that the defendant admitted as the factual basis for a guilty plea.’ ” (Id. at pp. 21–22.) But such a reading of Gallardo is at odds with the concerns that motivated our opinion. In Apprendi v. New Jersey (2000) 530 U.S. 466 (Apprendi), the high court held that under the Sixth Amendment, “any fact that increases the penalty for a crime beyond the prescribed statutory maximum, other than the fact of a prior conviction, must be submitted to a jury, and proved beyond a reasonable doubt.” (Apprendi, at p. 490.) Apprendi preserved an exception for sentencing enhancements based solely on the fact of a prior conviction — the Almendarez- Torres exception — on the ground that “there is a vast difference between accepting the validity of a prior judgment of conviction entered in a proceeding in which the defendant had the right to a jury trial and the right to require the prosecutor

3 In re MILTON Liu, J., dissenting

to prove guilt beyond a reasonable doubt, and allowing the judge to find the required fact under a lesser standard of proof.” (Id. at p. 496; see Almendarez-Torres v. United States (1998) 523 U.S. 224, 230, 244 (Almendarez-Torres).) Both before and after Gallardo, the determination of whether a defendant’s prior conviction counts as a “convict[ion] of a serious felony” (§ 667, subd. (a)(1)) has been a determination to be made by a judge, not by a jury. For this arrangement to comply with Apprendi, it must fall within the limited Almendarez-Torres exception. Gallardo was animated by our recognition that there are serious questions as to whether our previous Three Strikes jurisprudence can be reconciled with Apprendi in light of the high court’s discussions of that case in Descamps v. United States (2013) 570 U.S. 254 (Descamps) and Mathis v. United States (2016) 579 U.S. 500 (Mathis). Descamps and Mathis concerned the application of a federal sentencing enhancement statute, the Armed Career Criminal Act of 1984 (ACCA). (18 U.S.C. § 924, subd. (e).) Although we acknowledged that both cases “were decided on statutory, rather than constitutional grounds,” we explained that “the high court’s interpretation of the relevant federal statute was informed by an understanding of . . . Sixth Amendment principles, and the court’s explication of those principles was both considered and unequivocal: The jury trial right is violated when a court adds extra punishment based on factfinding that goes ‘beyond merely identifying a prior conviction’ by ‘tr[ying] to discern what a trial showed, or a plea proceeding revealed, about the defendant’s underlying conduct.’ ” (Gallardo, supra, 4 Cal.5th at pp. 134, 135, quoting Descamps, supra, 570 U.S. at p. 269.)

4 In re MILTON Liu, J., dissenting

Descamps and Mathis, we said, “are persuasive evidence that the Almendarez-Torres exception [to Apprendi] is narrower” than this court had previously supposed. (Gallardo, supra, 4 Cal.5th at p. 132.) Those decisions suggest that our pre-Gallardo case law could not be reconciled with the Almendarez-Torres exception because the statutory scheme as previously interpreted made a judge the trier of fact as to not merely “the fact of a prior conviction” but the nature of the conduct underlying a prior conviction. (Gallardo, at p. 130, quoting Apprendi, supra, 530 U.S. at p. 490.) We understood the high court to have interpreted the ACCA in Descamps and Mathis in a manner that avoids this potential Apprendi problem (Gallardo, at p. 133), and we construed the Three Strikes law to avoid that same problem (Gallardo, at p. 135 [“We are persuaded, and we will follow the [high] court’s guidance.”]). In other words, our decision in Gallardo was explicitly intended to be a state law analog to Descamps and Mathis. With this in mind, it is notable that the federal Courts of Appeals for the Sixth, Seventh, and Ninth Circuits have each concluded that the rules of Descamps and Mathis are substantive and must therefore be applied retroactively under Teague. (See Allen v. Ives (9th Cir. 2020) 950 F.3d 1184, 1192 [“To the extent that Mathis and Descamps may be thought to have announced a new rule, we have no trouble concluding that the rule is one of substance rather than procedure. . . . We have previously recognized that decisions that alter the substantive reach of a federal statute apply retroactively”]; Holt v. United States (7th Cir. 2016) 843 F.3d 720, 722 [“While Holt’s appeal was pending we held that the version of the Illinois burglary statute under which he had been convicted is

5 In re MILTON Liu, J., dissenting

indeed not a ‘violent felony’ because it does not satisfy the definition of ‘burglary’ used in Mathis v. United States [citation] for indivisible statutes. . . . [S]ubstantive decisions such as Mathis presumptively apply retroactively on collateral review.”]; Hill v. Masters (6th Cir. 2016) 836 F.3d 591, 595–596 [“The Government concedes that, after Descamps . . . Maryland’s second-degree assault statute no longer constitutes a crime of violence for the purpose of the career-offender enhancement. [Citation.] Thus, were Hill to be sentenced today, he would not qualify as a career offender. The Government further concedes that Descamps and Royal apply retroactively”].) Today’s decision makes this court an outlier. The court suggests these cases are inapposite because, unlike the federal statute at issue in Descamps and Mathis, the Three Strikes law allows courts to look beyond the elements of an offense and consider “the conduct underlying the offense.” (Maj. opn., ante, at p. 25, fn. 7.) It is true that, under Gallardo, a judge charged with determining whether a defendant is a person “previously convicted of a serious felony” for the purposes of section 667, subdivision (a)(1) makes this determination not by reference to the elements of the offense for which the conviction was entered, but rather by reference to the facts necessarily found by the trier of fact in entering the conviction. But that does not mean the Three Strikes law allows courts to find facts concerning the conduct underlying the offense. Under Gallardo, judges who administer the Three Strikes law are not charged with deciding questions about a defendant’s actual conduct, but only with determining what a jury necessarily found in entering a verdict of conviction. That is not a question about what the defendant did; it is a question about what a jury did (or, in the case of a bench trial or guilty

6 In re MILTON Liu, J., dissenting

plea, what a court did). We made clear in Gallardo that, were we to construe our statutory scheme to allow judges to “go[] ‘beyond merely identifying a prior conviction’ by ‘tr[ying] to discern what a trial showed, or a plea proceeding revealed, about the defendant’s underlying conduct,’ ” this would likely violate the requirements of the Sixth Amendment. (Gallardo, supra, 4 Cal.5th at p. 135, quoting Descamps, supra, 570 U.S. at p. 269.) I fear that the court’s reading of Gallardo today may resurrect the very Apprendi problem that Gallardo sought to avoid in construing the Three Strikes law. Before Gallardo, we had interpreted the scheme to allow judges to resolve factual questions concerning the nature of the conduct underlying prior convictions. We might have addressed this problem by having trial courts empanel a new jury to review the record of conviction and make its own judgment about the nature of the conduct on which a prior conviction was based. That was the approach suggested by Justice Chin in his dissent from Gallardo. (Gallardo, supra, 4 Cal.5th at pp. 140–144 (dis. opn. of Chin, J.).) Had we taken that approach, the question of what counts as having been “convicted of a serious felony” (§ 667, subd. (a)(1)) would have been unchanged; it still would turn on the nature of the conduct underlying the conviction. The change that would have been worked by Justice Chin’s proposed approach is that a jury, not a judge, would decide whether the conduct underlying the conviction constituted a serious felony. But that was not the path we chose. Instead, we clarified what counts as having been “convicted of a serious felony,” such that a judge may sit as a trier of fact only as to whether the facts necessarily found in the course of entering the prior conviction satisfy the statutory definition of “serious

7 In re MILTON Liu, J., dissenting

felony,” and not as to any details of the underlying conduct. We did not reallocate any decisionmaking authority from the judge to a jury; we clarified the question that the sentencing judge is tasked with deciding. In sum, I cannot agree that the only change worked by Gallardo was the promulgation of a new rule of evidence for determining the nature of the conduct underlying a prior conviction. That characterization of Gallardo misapprehends its significance in reconciling the Three Strikes law with Apprendi in light of Descamps and Mathis. For this reason, and for the reasons set forth by Justice Groban, I respectfully dissent.

LIU, J.

I Concur:

GROBAN, J.

8 In re MILTON

S259954

Dissenting Opinion by Justice Groban

William Milton is currently serving an indeterminate 25- years-to-life sentence because he purportedly used a gun in committing two robberies in Illinois over 10 years before he was convicted of a California robbery. Without a finding that he previously used a firearm in committing the Illinois robberies, Milton would have received a maximum prison term of five years. The allegation that he personally used a gun in the Illinois robberies was never pleaded, presented to a jury, nor admitted by Milton as part of a plea. Nonetheless, the California sentencing court relied on handwritten notes and statements from the Illinois sentencing judge “to determine what really happened” during Milton’s Illinois crimes and impose his “Three Strikes” sentence. The Attorney General, Milton, the majority, and I agree that this was error under our subsequent decision in People v. Gallardo (2017) 4 Cal.5th 120 (Gallardo). (See In re Milton (2019) 42 Cal.App.5th 977, 999 (Milton).) Nonetheless, the majority concludes that Milton has no recourse — he will continue to serve a 25-years-to-life sentence based upon a factual allegation that was never pleaded, found true by a jury, nor admitted as part of a plea — because the rule effectuated by Gallardo is “procedural,” and “reliability and fairness [were not] core concerns of our decision . . . .” (Maj. opn., ante, at pp. 23, 32.) I do not agree that the rule set forth in Gallardo, which could mean the difference between a life in prison or a short determinate term

1 In re MILTON Groban, J., dissenting

for some petitioners, constitutes a mere procedural change in the law. As explained below, in my view, Gallardo should be given retroactive effect because it substantively changed the class of persons punishable under the Three Strikes law. (See Welch v. United States (2016) 578 U.S. 120, 129 (Welch).) Moreover, even if Gallardo’s rule is deemed procedural, I would still apply it retroactively because its core purpose is to enhance the reliability of the factfinding process for prior conviction determinations. (See People v. Johnson (1970) 3 Cal.3d 404, 411 (Johnson).) 1. GALLARDO ANNOUNCED A NEW SUBSTANTIVE RULE As the majority explains, “a rule is substantive rather than procedural where it ‘ “alters the range of conduct or the class of persons that the law punishes.” ’ ” (Maj. opn., ante, at p. 18.) Whereas a procedural rule “ ‘ “regulate[s] only the manner of determining the defendant’s culpability.” ’ ” (Id. at p. 19.) The majority concludes the rule announced in Gallardo falls into the procedural category because it merely “ ‘regulate[d] the evidence that the court could consider’ in making prior conviction determinations.” (Id. at p. 21.) From the majority’s perspective, Gallardo “did not remove the defendant or any group of people from the reach of applicable sentencing laws, and we did not conclude that the conduct underlying [Sulma Marilyn Gallardo’s] prior conviction, i.e., whether she used a knife, was no longer relevant.” (Id. at p. 22.) I disagree. Sure, the range of conduct that meets the relevant statutory definition of “serious felony” was unchanged by Gallardo. Before and after Gallardo, an out-of-state felony has qualified as a “serious felony” under California law if the

2 In re MILTON Groban, J., dissenting

defendant personally used a firearm or a dangerous or deadly weapon in committing the offense. (Pen. Code, § 1192.7, subd. (c)(8), (23).) Thus, the majority rightly observes that Gallardo “did not [hold] that the conduct underlying [the defendant’s] prior conviction, i.e., whether [Gallardo] used a knife, was no longer relevant.” (Maj. opn., ante, at p. 22.) But the prosecutor’s task in the Three Strikes context is not simply to prove that the defendant merely committed a felony that the statutory scheme defines as serious or violent; the prosecutor must prove that the defendant was convicted of such a felony (or felonies). (See Pen. Code, §§ 1170.12, subd. (a), 667, subds. (b)–(j); see also maj. opn., ante, at p. 6.) Because Gallardo altered what can serve as the basis for a prior conviction finding, it substantively redefined the class of persons eligible for punishment under the Three Strikes law. (See In re Martinez (2017) 3 Cal.5th 1216, 1222.) More specifically, under the old regime, a defendant could be found to have suffered a prior serious felony conviction wherever the record revealed that, “realistically,” the conduct that supported the conviction satisfied the definition. (People v. McGee (2006) 38 Cal.4th 682, 706.) But now, because of Gallardo, a defendant can stand convicted of a serious felony within the meaning of the statute only if it can be shown that the “facts that were necessarily found in the course of entering the conviction” satisfy the serious felony definition. (Gallardo, supra, 4 Cal.5th at p. 134; see also id. at p. 136 [“a court considering whether to impose an increased sentence based on a prior qualifying conviction may not determine the ‘nature or basis’ of the prior conviction based on its independent conclusions about what facts or conduct ‘realistically’ supported the conviction”]; id. at p. 138 [“While a

3 In re MILTON Groban, J., dissenting

trial court can determine the fact of a prior conviction without infringing on the defendant’s Sixth Amendment rights, it cannot determine disputed facts about what conduct likely gave rise to the conviction”].) Where a defendant’s conviction does not establish the facts necessary to render a prior offense a serious felony, a judge can no longer impose a Three Strikes sentence. In other words, Gallardo narrowed the universe of people eligible for Three Strikes treatment to those whose juries necessarily found true (or the defendant necessarily admitted as part of a guilty plea) the strike qualifying facts. So understood, Gallardo did not simply alter the procedures used, or evidence properly considered, when determining whether an individual suffered a prior strike conviction, it substantively changed the punishable class. (See Welch, supra, 578 U.S. at p. 129; id. at p. 130 [“Johnson [v. United States (2015) 576 U.S. 591] affected the reach of the underlying statute rather than the judicial procedures by which the statute is applied. Johnson is thus a substantive decision and so has retroactive effect . . . in cases on collateral review” (italics added)]; see also People v. Trujeque (2015) 61 Cal.4th 227, 251 [“Using the high court’s rationale, it seems fair to characterize Breed’s [Breed v. Jones (1975) 421 U.S. 519, 531] double jeopardy rule as more substantive than procedural because without the rule’s retroactive application, a defendant would otherwise ‘face[] a punishment that the law cannot impose upon him’ ”].) I would therefore apply Gallardo retroactively to cases final on appeal.

4 In re MILTON Groban, J., dissenting

2. GALLARDO’S PRIMARY PURPOSE IS TO PROMOTE

RELIABLE DETERMINATIONS OF GUILT OR INNOCENCE

Even if I were to accept the majority’s view that Gallardo merely set forth a new procedural rule (maj. opn., ante, pp. 20– 23), I would still apply it retroactively to final convictions under the state balancing test for the retroactive application of new rules. (See Johnson, supra, 3 Cal.3d 404.) As the majority explains, under our state test, if the judicial decision establishes a new rule, courts determine whether to apply the new rule based on three considerations — the purpose of the new rule, the reliance placed on the old rule, and the effect retroactive application would have on the administration of justice. (Id. at p. 410; maj. opn., ante, at pp. 25–26.) “The first factor — the purpose of the new rule — is the critical factor in determining retroactivity.” (Maj. opn., ante, at p. 26.) State courts are “free to give greater retroactive impact to a decision than the federal courts choose to give.” (Johnson, at p. 415.) Pursuant to Johnson, “[f]ully retroactive decisions are seen as vindicating a right which is essential to a reliable determination of whether an accused should suffer a penal sanction.” (Id. at p. 411; see maj. opn., ante, at p. 27 [quoting Johnson on this point].) I agree with petitioner that “[t]he fundamental purpose of Gallardo is to promote fair and reliable determinations of the petitioner’s guilt or innocence on the allegation that he suffered a prior conviction qualifying as a strike under California law.” In rejecting petitioner’s argument, the majority reasons that a pre-Gallardo sentencing court’s factfinding, beyond a reasonable doubt, based on the prior record of conviction was “reasonably fair and reliable.” (Maj. opn., ante, at p. 32; see id. at pp. 31–35.) I disagree. The majority overlooks the fact that,

5 In re MILTON Groban, J., dissenting

by limiting a sentencing court to simply “identifying those facts that were established by virtue of the conviction itself,” Gallardo cabined a trial court’s authority in a way that was primarily meant to enhance the reliability of prior serious felony determinations. (Gallardo, supra, 4 Cal.5th at p. 136.) The sentencing judge is no longer permitted to fact find — judges may not “ ‘[try] to discern what a trial showed, or a plea proceeding revealed, about the defendant’s underlying conduct’ ” (Id. at p. 135, quoting Descamps v. United States (2013) 570 U.S. 254, 269 (Descamps).) Instead, sentencing courts may now only identify those facts that were already found in “the deliberate and considered way the Constitution guarantees.” (Descamps, at p. 273.) In fact, the manner in which Milton was sentenced here highlights the inherent unreliability of a trial court’s pre- Gallardo prior conviction determinations — and how Gallardo remedied this unreliability. As the majority explains, in 1987, Milton was convicted of two robberies in Illinois: one armed robbery (convicted by jury) and one simple robbery (convicted by plea). The certified record from the Illinois case contained some handwritten notes on what appears to be a charging document. According to the notes, in the simple robbery, petitioner accosted the victim, demanded money, and took $338. The notes also indicated that petitioner “had a gun.” At sentencing, the Illinois court stated that, with respect to the plea/simple robbery case, the “stipulated facts” indicated petitioner “possessed a handgun.” Later in imposing sentence, the Illinois court observed in aggravation that, in both cases, petitioner “deliberately held a gun — a loaded gun — upon an individual.”

6 In re MILTON Groban, J., dissenting

In the California case, the prosecutor conceded that petitioner did not admit the arming allegation as part of his plea to the Illinois simple robbery, but argued that the Illinois judge’s handwritten notes (which the prosecutor believed the court made at sentencing) and his comments at sentencing indicated that petitioner used a gun. Defense counsel argued that the original arming allegation was dismissed and that it was unclear “who or what stipulated to anything” and where the “facts” in the notes came from. In any event, the stipulation only stated that petitioner “possessed a gun.” It did not show he used a firearm during the commission of the robbery, which is necessary to qualify the offense as a serious felony.1 The sentencing judge concluded that the Illinois robbery convictions qualified as California serious felonies noting that, “I see nothing wrong with going beyond the court record to determine what really happened. And in doing so, I am satisfied that the defendant used a gun in both robberies.” In other words, Milton was sentenced to a Three Strikes term because the sentencing judge was personally “satisf[ied],” based on details presented in notes and statements that were not necessary to Milton’s prior underlying guilt

1 Cf. People v. Chambers (1972) 7 Cal. 3d 666, 672 (discussing the firearm use enhancement under Pen. Code, § 12022.5 and explaining that “[b]y employing the term ‘uses’ instead of ‘while armed’ the Legislature requires something more than merely being armed”).

7 In re MILTON Groban, J., dissenting

determinations, that he used a gun.2 This was not a reliable basis upon which to sentence Milton to an indeterminate 25- years-to-life term. Now look at the impact Gallardo would have if Milton were sentenced today; the sentencing court would be limited to simply identifying those “facts that were necessarily found in the course of entering [Milton’s prior] conviction[s].” (Gallardo, supra, 4 Cal.5th at p. 134.) Thus, if Milton were sentenced today, the court would not be able to consider the purported facts in the Illinois judge’s handwritten notes or the judge’s

2 Here, the California sentencing court essentially attached an uncharged personal firearm use enhancement to petitioner’s prior Illinois robberies to render them strikes. Such action arguably undermined additional case law from our court and the high court separate and apart from Gallardo. (See, e.g., Apprendi v. New Jersey (2000) 530 U.S. 466, 476 [“ ‘under the Due Process Clause of the Fifth Amendment and the notice and jury trial guarantees of the Sixth Amendment, any fact . . . that increases the maximum penalty for a crime must be charged in an indictment, submitted to a jury, and proven beyond a reasonable doubt’ ”]; People v. Anderson (2020) 9 Cal.5th 946, 953 [while citing our case law strictly interpreting statutory pleading and proof requirements, we noted that those requirements derive from “a bedrock principle of due process,” one that also applies to sentence enhancements since a defendant has the “ ‘right to fair notice of the specific sentence enhancement allegations that will be invoked to increase punishment for his crimes’ ”].) Furthermore, Milton makes the persuasive argument that such an allowance unfairly deprived him of the benefit of his plea bargain. In his simple robbery case in Illinois, an arming allegation was dismissed in exchange for his guilty plea, but the California court punished him as though he had pleaded guilty to a greater offense.

8 In re MILTON Groban, J., dissenting

statements at sentencing, which were of “questionable reliability” (Milton, supra, 42 Cal.App.5th at p. 995) and not “necessarily found in the course of entering [his] conviction[s]” (Gallardo, at p. 134). Gallardo’s limitation thereby eliminates the risk that an individual like Milton will be sentenced to a Three Strikes term on an unreliable basis. The pre-Gallardo risks to reliability are elucidated by the high court’s rationale for limiting a sentencing court to comparing elements between a prior crime of conviction and the relevant definition of a predicate offense for purposes of an enhanced sentence under the Armed Career Criminal Act. (See Descamps, supra, 570 U.S. at p. 257; Mathis v. United States (2016) 579 U.S. 500 (Mathis).) “The Armed Career Criminal Act (ACCA or Act), 18 U.S.C. § 924(e), increases the sentences of certain federal defendants who have three prior convictions ‘for a violent felony,’ including ‘burglary, arson, or extortion.’ To determine whether a past conviction is for one of those crimes, courts use what has become known as the ‘categorical approach’: They compare the elements of the statute forming the basis of the defendant’s conviction with the elements of the ‘generic’ crime — i.e., the offense as commonly understood. The prior conviction qualifies as an ACCA predicate only if the statute’s elements are the same as, or narrower than, those of the generic offense.” (Descamps, at p. 257.) In Descamps, in explaining the rationale for its elements- based approach, the high court observed that the meaning of records from prior convictions “will often be uncertain” as to nonelements of an offense because “[a] defendant, after all, often has little incentive to contest facts that are not elements of the charged offense — and may have good reason not to.”

9 In re MILTON Groban, J., dissenting

(Descamps, supra, 570 U.S. at p. 270.) The high court reiterated this same logic in Mathis, supra, 579 U.S. at page 512, explaining that “[s]tatements of ‘non-elemental fact’ in the records of prior convictions are prone to error precisely because their proof is unnecessary. [Citation.] At trial, and still more at plea hearings, a defendant may have no incentive to contest what does not matter under the law; to the contrary, he ‘may have good reason not to’ — or even be precluded from doing so by the court. [Citation.] When that is true, a prosecutor’s or judge’s mistake as to means, reflected in the record, is likely to go uncorrected. [Citation.] Such inaccuracies should not come back to haunt the defendant many years down the road by triggering a lengthy mandatory sentence.” (Fn. omitted; see also Sessions v. Dimaya (2018) __ U.S. __ [138 S.Ct. 1204, 1218] (plur. opn.) [“This Court has often described the daunting difficulties of accurately ‘reconstruct[ing],’ often many years later, ‘the conduct underlying [a] conviction’ ”].) We expressed a desire to remedy similar concerns in adopting Gallardo’s rule. In concluding that the sentencing court improperly relied on the preliminary hearing transcript from Gallardo’s prior plea proceedings to conclude that she used a knife during her prior assault offense, we observed, “A sentencing court reviewing that preliminary transcript has no way of knowing whether a jury would have credited the victim’s testimony had the case gone to trial. And at least in the absence of any pertinent admissions, the sentencing court can only guess at whether, by pleading guilty to a violation of Penal Code section 245, subdivision (a)(1), defendant was also acknowledging the truth of the testimony indicating that she had committed the assault with a knife.” (Gallardo, supra, 4 Cal.5th at p. 137, italics added.) Gallardo thus remanded

10 In re MILTON Groban, J., dissenting

the matter “to permit the People to demonstrate to the trial court, based on the record of the prior plea proceedings, that defendant’s guilty plea encompassed a relevant admission about the nature of her crime.” (Id. at p. 139.) By limiting the sentencing judge to “identifying those facts that were established by virtue of the conviction itself” (Gallardo, supra, 4 Cal.5th at p. 136), Gallardo prohibits guesswork and reliance on inaccurate or incomplete records, such as that which occurred here. Gallardo thus necessarily eliminated the risk that an individual will be sentenced to a Three Strikes term on an unreliable basis. Stated differently, Gallardo’s “primary purpose” is “to promote reliable determinations of guilt or innocence.” (People v. Guerra (1984) 37 Cal.3d 385, 402.) Gallardo should therefore be applied retroactively on collateral review under Johnson. In rejecting the notion that Gallardo’s “ ‘primary purpose’ ” was to promote reliability, the majority underscores, inter alia, that a pre-Gallardo sentencing court was required to apply a beyond a reasonable doubt standard of proof to prior serious felony determinations and was limited to reviewing the record of the prior conviction. (Maj., opn., ante, at pp. 32–37.) But the majority fails to account for the fact that, before Gallardo, a sentencing judge could look to facts that were not “established by virtue of the conviction itself.” (Gallardo, supra, 4 Cal.5th at p. 136.) These unestablished facts, which were not necessary to the conviction, were “prone to error precisely because their proof [was] unnecessary” (Mathis, supra, 579 U.S. at p. 512). As such, any determination based on these extraneous facts would have been unreliable irrespective of the standard of proof they were used to satisfy. In this way, it simply does not matter that pre-Gallardo courts

11 In re MILTON Groban, J., dissenting

were utilizing a beyond a reasonable doubt standard. What matters is that, in order to reach this conclusion, they were relying on information of “questionable reliability” (Milton, supra, 42 Cal.App.5th at p. 995) that they are not permitted to rely on after Gallardo. Furthermore, the majority rejects Milton’s claim that the pre-Gallardo procedure “raised reliability concerns because defendants lacked the incentive to contest facts regarding [conduct not critical to their convictions] in the prior proceedings.” (Maj. opn., ante, at pp. 34–35.) The majority observes that Milton had incentive to contest his gun use because the Illinois sentencing judge relied on it to impose an aggravated sentence. (Id. at p. 35.) But even if Milton’s potential incentive to challenge his gun use in Illinois lends reliability to the sentencing judge’s factfinding in Milton’s case (a premise I disagree with),3 this circumstance has no bearing

3 Even if the circumstances of this individual petitioner were relevant to our determination, the majority asks too much of him. When Milton was convicted in Illinois, he had no apparent notice that a California court might, years later, look at the judge’s notes and statements in the record from his Illinois sentencing hearing to dramatically enhance his sentence for an unrelated offense. I disagree with the premise that though Milton was not even charged with committing a crime involving personal use of a deadly weapon, he was nonetheless fully incentivized to contest personal use of a deadly weapon. Under these circumstances, and in light of the important reliability concerns at stake, I do not believe the onus should be on a petitioner to foresee such circumstances and create a fuller record or, years later, require him to “raise [his] lack of incentive to challenge that fact during the original

12 In re MILTON Groban, J., dissenting

on the question of Gallardo’s global impact on the reliability of guilt determinations. The issue we must decide here is whether Gallardo applies “retroactively to final judgments.” (Maj. opn., ante, at p. 2.) As the majority acknowledges, in order to answer this question, Johnson principally compels us to consider “the purpose of the new rule” in Gallardo. (Id. at p. 26, italics added.) The fact that one defendant might have had reason to challenge an unproven factual allegation used to enhance his sentence tells us nothing about Gallardo’s purpose. Moreover, I read the majority’s decision to apply far more broadly than to just this petitioner (id. at p. 38 [“In light of all the pertinent considerations, we conclude the rule we announced in Gallardo is . . . not retroactive to cases on collateral review under both state and federal tests for retroactivity”]), and therefore unique circumstances that may or may not apply to this petitioner should play no part in that

proceedings, in the course of arguing the beyond-a-reasonable- doubt standard [for prior conviction determinations] was not satisfied.” (Maj. opn., ante, at p. 35.) The Court of Appeal decisions cited by the majority in support of a contrary conclusion do not persuade me otherwise. (See id. at p. 35.) Notably, in two of the cases, People v. Johnson (1989) 208 Cal.App.3d 19, 24 and People v. Skeirik (1991) 229 Cal.App.3d 444, 464, the sentence enhancing statutes at issue existed at the time the defendant was convicted of the prior offense, thereby making the defendant’s incentive to contest certain underlying conduct an arguably closer case; Milton, by contrast, was convicted of the Illinois robberies before the 1994 passage of the Three Strikes law. (See Ewing v. California (2003) 538 U.S. 11, 15 [detailing California’s passage of the Three Strikes law].)

13 In re MILTON Groban, J., dissenting

determination. Even if the majority were correct and Milton did have incentive to contest his gun use, there is no basis for concluding that all petitioners “as a general matter” (maj. opn, ante, at p. 35) had similar incentives in their prior proceedings. Finally, the majority asserts that pre-Gallardo courts were trying to “avoid the unfairness of treating [out-of-state] defendants with essentially identical underlying conduct very differently under California’s recidivist sentencing laws” based on “minor variations in the way different jurisdictions define criminal offenses.” (Maj. opn., ante, at p. 36.) But these differences were not “minor,” and the defendants did not engage in “identical underlying conduct.” (Ibid.) For instance, as relevant to Milton’s crimes, “[a]n essential element of the California crime of robbery is ‘the [specific] intent to permanently deprive the person of the property,’ ” whereas “robbery and armed robbery are general intent crimes in Illinois.” (Milton, supra, 42 Cal.App.5th at p. 985.) The difference between general and specific intent can greatly distinguish a defendant’s culpability. (See People v. Moore (2018) 19 Cal.App.5th 889, 893 [“General intent crimes require only a general criminal intent to commit the proscribed act, while specific intent crimes require an additional intent to do some further act or achieve some further consequence”].) Such differences are neither “minor” (maj. opn., ante, at p. 36) nor the result of mere “happenstance” (ibid.) and thus illustrate the importance of Gallardo’s new rule. Now, under Gallardo, “the conviction itself” (Gallardo, supra, 4 Cal.5th at p. 136) must qualify a petitioner for Three Strikes treatment, which ensures that an out-of-state offender “actually did commit identical conduct.” (Maj. opn., ante, at p. 37, fn. 12.) For all these reasons, we should apply Gallardo retroactively and

14 In re MILTON Groban, J., dissenting

remand for an accurate assessment of Milton’s prior convictions (just as we did in Gallardo). (See id. at pp. 139– 140.) 3. CONCLUSION The upshot of the majority’s holding is that even though Milton’s prior use of a gun was not pleaded or proven at trial, or admitted to by plea, and even though the parties and the majority all agree that the sentencing court engaged in improper factfinding to conclude he personally used a gun, his indeterminate 25-years-to-life sentence must stand. I would not let Milton continue to serve a Three Strikes term without certainty that this severe punishment is supported by facts “necessarily found in the course of entering [his] prior conviction[s].” (Gallardo, supra, 4 Cal.5th at p. 134.) Milton should get the chance to be resentenced based upon the more reliable rule we laid out in Gallardo, the very same rule that would apply if he or anyone else were sentenced today. I dissent.

GROBAN, J.

I Concur:

LIU, J.

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

Name of Opinion In re Milton __________________________________________________________

Procedural Posture (see XX below) Original Appeal Original Proceeding Review Granted (published) XX 42 Cal.App.5th 977 Review Granted (unpublished) Rehearing Granted __________________________________________________________

Opinion No. S259954 Date Filed: August 22, 2022 __________________________________________________________

Court: Superior County: Los Angeles Judge: Ronald J. Slick __________________________________________________________

Counsel:

Brad Kaiserman, under appointment by the Supreme Court, for Petitioner William Milton.

Mary K. McComb, State Public Defender, Elias Batchelder, Erik Levin and Jessie Hawk, Deputy State Public Defenders, as Amicus Curiae on behalf of Petitioner William Milton.

Xavier Becerra and Rob Bonta, Attorneys General, Gerald A. Engler and Lance E. Winters, Chief Assistant Attorneys General, Susan Sullivan Pithey, Assistant Attorney General, Michael C. Keller, Eric J. Kohm, Louis W. Karlin and Nicholas J. Webster, Deputy Attorneys General, for Respondent the People. Counsel who argued in Supreme Court (not intended for publication with opinion):

Brad Kaiserman 5870 Melrose Avenue # 3396 Los Angeles, CA 90038 (310) 367-7632

Nicholas J. Webster Deputy Attorney General 300 South Spring Street, Suite 1702 Los Angeles, CA 90013 (213) 269-6103