Opinion
PANELLI,J.
—Proposition 115, the “Crime Victims Justice Reform Act,” changed criminal law in several respects on June 6, 1990. We granted review to determine whether the measure’s provisions should be applied to prosecutions of crimes committed before its effective date. We conclude that certain provisions addressing the conduct of trials, and certain other provisions changing the law to the benefit of defendants, may be applied to such prosecutions. The remainder of the measure’s provisions may not.
Facts
The People have accused petitioner Robert Alan Tapia of committing first degree murder with special circumstances on February 12, 1989. The prosecution is pending in the Superior Court of Tulare County. Voir dire has not yet commenced.
Proposition 115 took effect on June 6, 1990, the day after the voters approved the measure. Shortly thereafter, the superior court ruled that it would apply the measure’s procedural provisions to the case before it and, accordingly, conduct voir dire under the new statute. The new voir dire statute provides that the court rather than the attorneys “shall conduct the examination of prospective jurors” and that the examination “shall be con[287]*287ducted only in aid of the exercise of challenges for cause.” (Prop. 115, § 7, codified as Code Civ. Proc., § 223.) Seeking to have the superior court’s order vacated, Tapia petitioned the Court of Appeal for a writ of mandate. The Court of Appeal summarily denied relief. We granted review and directed issuance of an alternative writ. We also stayed proceedings in the superior court pending our decision.
Discussion
As stated, we granted review to determine whether the provisions of Proposition 115 should be applied to prosecutions of crimes committed before its effective date. To answer this question we must address two issues. The first is whether the presumption of prospectivity applies to this initiative. (See Evangelatos v. Superior Court (1988) 44 Cal.3d 1188, 1206-1209 [246 Cal.Rptr. 629, 753 P.2d 585] (Evangelatos).) The second concerns the meaning of the terms “prospective” and “retrospective.” We do not address any other issue concerning the applicability or validity of the measure’s provisions.1
I.
We may quickly dispose of the first issue. It is well settled that a new statute is presumed to operate prospectively absent an express declaration of retrospectivity or a clear indication that the electorate, or the Legislature, intended otherwise. (See People v. Hayes (1989) 49 Cal.3d 1260, 1274 [265 Cal.Rptr. 132, 783 P.2d 719]; Evangelatos, supra, 44 Cal.3d at pp. 1206-1209; Aetna Cas. & Surety Co. v. Ind. Acc. Com. (1947) 30 Cal.2d 388, 393 [182 P.2d 159] (Aetna Casualty); Jones v. Union Oil Co. (1933) 218 Cal. 775, [25 P.2d 5]; cf. In re Estrada (1965) 63 Cal.2d 740, 746 [48 Cal.Rptr. 172, 408 P.2d 948].)2 Both the text of Proposition 115 and the related ballot arguments are entirely silent on the question of retrospectivity. Thus, as to most of Proposition 115’s provisions we see no reason to depart from the ordinary rule of construction that new statutes are intended to operate prospectively.3
[288]*288II.
There remains the question of what the terms “prospective” and “retrospective” mean. Tapia argues that a law is being applied retrospectively if it is applied to the prosecution of a crime committed before the law’s effective date. For some types of laws, the test which Tapia proposes is clearly appropriate. Certainly a law is retrospective if it defines past conduct as a crime, increases the punishment for such conduct, or eliminates a defense to a criminal charge based on such conduct. Such a law, as applied to a past crime, “change[s] the legal consequences of an act completed before [the law’s] effective date,” namely the defendant’s criminal behavior. (Weaver v. Graham (1981) 450 U.S. 24, 31 [67 L.Ed.2d 17, 24, 101 S.Ct. 960]; see also People v. Weidert (1985) 39 Cal.3d 836, 851 [218 Cal.Rptr. 57, 705 P.2d 380], quoting Weaver v. Graham, supra; cf. Aetna Casualty, supra, 30 Cal.2d at p. 394 [a law is retrospective if it changes “the legal effects of past events”].) Application of such a law to past crimes would also violate the constitutional rule against ex post facto legislation. (U.S. Const., art. I, § 10, cl. 1; Cal. Const., art. I, § 9.)
Tapia’s proposed test is not appropriate, however, for laws which address the conduct of trials which have yet to take place, rather than criminal behavior which has already taken place. Even though applied to the prosecution of a crime committed before the law’s effective date, a law addressing the conduct of trials still addresses conduct in the future. This is a principle that courts in this state have consistently recognized. Such a statute “ ‘is not made retroactive merely because it draws upon facts existing prior to its enactment .... [Instead,] [t]he effect of such statutes is actually prospective in nature since they relate to the procedure to be followed in the future.” (Strauch v. Superior Court (1980) 107 Cal.App.3d 45, 49 [165 Cal.Rptr. 552], quoting Olivas v. Weiner (1954) 127 Cal.App.2d 597, 600-601 [274 P.2d 476].) For this reason, we have said that “it is a misnomer to designate [such statutes] as having retrospective effect.” (Morris v. Pacific Electric Ry. Co. (1935) 2 Cal.2d 764, 768 [43 P.2d 276].)
We previously addressed this issue in Estate of Patterson (1909) 155 Cal. 626 [102 P. 941]. The case involved the proof of a will that was destroyed in the great San Francisco fire of 1906. The testatrix died later that year, unaware that her will was gone. In 1907, the Legislature amended the Civil Code to permit proof of a will “ ‘shown to have been ... by public calamity destroyed in the lifetime of the testator, without his knowledge . . . .’ ” (Estate of Patterson, supra, 155 Cal. at p. 634, quoting former Code Civ. Proc., § 1339, added by Stats. 1907, ch. 100, § 1, p. 122.) Holding the new statute applicable, we stated that “[i]t is a mistake to characterize the amendment of section 1339 as a retrospective law. It relates wholly to what [289]*289shall be done upon the trial of the application for probate, the proof that must be furnished and the facts which must be established. It applies only to trials which take place after its enactment. It can have no effect whatever on previous trials or enactments. It is prospective only in its nature.” (Estate of Patterson, supra, 155 Cal. at p. 638.)
Courts came to the same conclusion in subsequent decisions. In Strauch v. Superior Court, supra, 107 Cal.App.3d 45, the court held to be prospective a statute which imposed on plaintiffs in malpractice suits the requirement of filing a certificate of merit, even as applied to causes of action that accrued before the statute’s effective date. The new statute operated prospectively because it did not “create a new cause of action or deprive a malpractice defendant of any defense on the merits or affect vested rights.” (Id., at p. 49.) Similarly, in Andrus v. Municipal Court (1983) 143 Cal.App.3d 1041 [192 Cal.Rptr. 341] (disapproved on other grounds in Evangelatos, supra, 44 Cal. 3d at p. 1207, fn. 11), the court immediately applied a new statute eliminating the right to appeal from certain orders of the superior court denying extraordinary relief. Noting that “[Retroactive is not an apt word, of course,” the court found it “absurd” to “subscribe to the notion that the Legislature desired to postpone the demise of a procedural loophole . . . .” (Andrus v. Municipal Court, supra, at p. 1047.) Once again, the court in Republic Corp. v. Superior Court (1984) 160 Cal.App.3d 1253 [207 Cal.Rptr. 241] immediately applied a new, mandatory statute requiring dismissal for prolonged failure to prosecute and making the plaintiff’s diligence irrelevant. The court observed that application of the new statute to a case filed before its enactment was not “retroactive” because the statute’s effect was “ ‘actually prospective in nature . . . .’ ” (Id., at p. 1257, quoting Strauch v. Superior Court, supra, 107 Cal.App.3d at pp. 48-49.)
From these cases, it is evident that a law governing the conduct of trials is being applied “prospectively” when it is applied to a trial occurring after the law’s effective date, regardless of when the underlying crime was committed or the underlying cause of action arose. Tapia challenges this conclusion, arguing that we previously rejected it in Aetna Casualty, supra, 30 Cal.2d 388, in Evangelatos, supra, 44 Cal.3d 1188, and in People v. Hayes, supra, 49 Cal. 3d 1260. Those opinions, however, do not support his argument. In each opinion, we refused to apply a statute so as to change the legal consequences of the parties’ past conduct. In determining whether such statutes changed “the legal effects of past events” (Aetna Casualty, supra, 30 Cal.2d at p. 394) we sometimes used the terms “substantive” and “procedural.” (Id., at p. 395; Evangelatos, supra, 44 Cal.3d at p. 1226, fn. 26.) However, we also made it clear that it is the law’s effect, not its form or label, which is important. (Aetna Casualty, supra, 30 Cal.2d at p. 394; Evangelatos, supra, 44 Cal.3d at pp. 1225-1226, fn. 26.)
[290]*290In Aetna Casualty, supra, 30 Cal.2d 388, we annulled an award of the Industrial Accident Commission. The commission had erroneously applied a statute which increased the amount of compensation for disabilities in a case involving injuries which had occurred prior to the statute’s effective date. Defending its award before this court, the commission argued that the statute in question “relate[d] solely to matters of procedure or remedy.” (30 Cal.2d at p. 393.) We acknowledged both the rule articulated in Estate of Patterson, supra, 155 Cal. 626, and our previous observation that “ ‘it is a misnomer to designate [procedural changes] as having retrospective effect.’ ” (Aetna Casualty, supra, 30 Cal.2d at p. 394, quoting Morris v. Pacific Electric Ry. Co., supra, 2 Cal.2d at p. 768.) We offered the further explanation that “[i]n such cases the statutory changes are said to apply not because they constitute an exception to the general rule of statutory construction [i.e., the presumption of prospectivity], but because they are not in fact retroactive. There is then no problem as to whether the Legislature intended the changes to operate retroactively.” (Aetna, supra, 30 Cal.2d at p. 394.) However, we did not apply this reasoning to the statute at issue because the statute was “substantive in its effect, and its operation would be retroactive, since it imposes a new or additional liability and substantially affects existing rights and obligations.” (Id., at p. 395.) The law in question increased employers’ liability for industrial disabilities and, thus, changed the legal consequences of their past conduct.
We confronted a similar situation in Evangelatos, supra, 44 Cal.3d 1188. The initiative at issue in that case, Proposition 51, limited an individual tortfeasor’s liability for noneconomic damages to his proportionate share of fault. Misconstruing the rule derived from Estate of Patterson, supra, 155 Cal. 626, several amici curiae contended that the measure should be applied retrospectively on the ground that it was “procedural” rather than “substantive.” We rejected the argument because “retroactive application of Proposition 51 to preexisting causes of action would have a very definite substantive effect” on both plaintiffs and defendants. (Evangelatos, supra, 44 Cal.3d at pp. 1225-1226, fn. 26.) Like the law at issue in Aetna Casualty, supra, Proposition 51 changed the legal consequences of the parties’ past conduct. Specifically, Proposition 51 reduced the amount of damages which a plaintiff could recover from a particular defendant and made it impossible for defendants to recover contribution for noneconomic damages.4
Thus, contrary to Tapia’s argument, it is clear that neither Aetna Casualty, supra, nor Evangelatos, supra, repudiated the general rule that statutes [291]*291addressing the conduct of trials are prospective. Instead, in each case we held the rule inapplicable to statutes which changed the legal consequences of past conduct by imposing new or different liabilities based upon such conduct. (Aetna Casualty, supra, 30 Cal.2d at p. 395; Evangelatos, supra, 44 Cal.3d at p. 1226, fn. 26.)
Tapia also interprets our opinion in People v. Hayes, supra, 49 Cal.3d 1260, as supporting his position. It does not. In Hayes we considered the effect of Evidence Code section 795, which requires the exclusion of prehypnotic testimony unless certain statutory procedures were followed at the time of hypnosis. As in our previous cases, we began by reaffirming the presumption that new statutes operate prospectively and proceeded to determine what “prospective” operation meant in the case before us. We did not hold that the statute would apply, or not, based upon the date the alleged crime was committed. Instead, we looked to the date of the conduct regulated by the statute. Because the prehypnotic evidence in question predated the statute, we held that “[t]o invoke section 795 to exclude such evidence . . . would be tantamount to giving the statute retroactive effect.” (People v. Hayes, supra, 49 Cal. 3d at p. 1274.) The past conduct to which the statute attached legal consequences was the use of hypnosis; the date of the offense was irrelevant.
III.
Tapia next argues that a definition of “retrospective law” originally formulated in American States W. S. Co. v. Johnson (1939) 31 Cal.App.2d 606, 613 [88 P.2d 770] (American States), encompasses—and thus prohibits—the application of a new law to the prosecution of crimes committed before the law’s effective date. In American States, the Court of Appeal defined a retrospective law “[as] one which affects rights, obligations, acts, transactions and conditions which are performed or exist prior to the adoption of the statute.” We repeated th& American States formulation in Evangelatos, supra, 44 Cal.3d at page 1206, and in Aetna Casualty, supra, 30 Cal.2d at page 391. Based upon this formulation, Tapia argues that a law addressing the conduct of trials, such as the new voir dire provision, “affects [a] right[] . . . which . . . exist[ed] prior to the adoption of the statute,” the right in question being the supposed right to have voir dire conducted under the law existing at the time of the offense.
As the foregoing examination of our opinions demonstrates, however, we have not invoked the American States formulation to justify such a result.5 [292]*292Tapia’s argument appears to be an effort to incorporate into the definition of “retrospectivity” the now obsolete concept of “substantial protections.” It was formerly held that a law violated the constitutional prohibition against ex post facto legislation if it eliminated a “substantial protection,” whether substantive or procedural, existing at the time of the offense. However, the United States Supreme Court has repudiated that analysis. (Collins v. Youngblood (1990) _ U.S. _ [111 L.Ed.3d 30, 39, 110 S.Ct. 2715, 2719] (Collins).) Although we have employed the federal concept of “substantial protections” in deciding questions under the ex post facto clause (People v. Smith (1983) 34 Cal.3d 251, 260 [193 Cal.Rptr. 692, 667 P.2d 149]; People v. Ward (1958) 50 Cal.2d 702, 707 [328 P.2d 777, 76 A.L.R.2d 911]; see pt. IV, post), the concept has never formed a part of the definition of “retrospectivity” in the context of statutory construction.
IV.
Finally, Tapia argues that our opinion in People v. Smith, supra, 34 Cal. 3d 251 (Smith), dictates a contrary result. In Smith we held that the provisions of Proposition 8, the “Victims’ Bill of Rights,” would apply only to prosecutions of offenses committed after the measure’s effective date. We did not distinguish between provisions which addressed the defendant’s criminal behavior and provisions which addressed the conduct of trials.
The defendant in Smith was found guilty of robbery after the trial court denied his motion to suppress a confession. On appeal, defendant argued that his confession had been obtained in violation of the California Constitution. Proposition 8 took effect after we granted review and long after the defendant’s crime and trial. One provision of that measure amended the California Constitution to provide that “relevant evidence shall not be excluded in any criminal proceeding.” (Cal. Const., art. I, § 28, subd. (d).) We held that Proposition 8 could not be applied to validate the trial court’s ruling, erroneous at the time it was made, denying the motion to suppress. Thus, the narrow question immediately before us in Smith was truly one of retrospectivity, that is, whether a new rule should be applied to change “the legal effects of past events.” (Aetna Casualty, supra, 30 Cal.2d at p. 394; see also Weaver v. Graham, supra, 450 U.S. at pp. 30-31 [67 L.Ed.2d at pp. 24-25].)
In Smith, going beyond the facts of the case, we also held that Proposition 8 in its entirety would apply only to crimes committed after the measure’s effective date. Because Proposition 115, like Proposition 8, includes [293]*293various types of provisions, Tapia argues that we should follow the same approach in this case. However, the reasons supporting our decision in Smith do not dictate the same result in the case now before us.
The first reason we gave for our decision was that the “primary stated purpose of Proposition 8 [was] to deter the commission of crimes.” (Smith, supra, 34 Cal.3d at p. 258, italics added.) Based on that purpose, we concluded that the electorate “must have intended the measure to apply only to offenses that could be deterred, i.e., that had not already been committed by the time Proposition 8 was adopted.”6 (Smith, supra, 34 Cal.3d at pp. 258-259, italics in original.) In contrast, the voters in adopting Proposition 115 expressly declared that their purposes were to reduce the unnecessary “costs of criminal cases” and to “create a system in which justice is swift and fair . . . .” (Prop. 115, § 1, subds. (b), (c).)7 We can best effectuate this purpose by giving the earliest possible application to reforms designed to accelerate the adjudication of criminal cases.
The second reason we gave for our decision in Smith was the desire to avoid resolving doubts about the initiative’s constitutionality arising from the rule against ex post facto legislation. (Smith, supra, 34 Cal. 3d at p. 259.) In 1983, when we decided Smith, some authorities suggested that even procedural reforms might violate the rule against ex post facto legislation unless they left “untouched all the substantial protections with which existing law surround[ed] the person accused of crime” at the time the crime was committed. (Thompson v. Utah (1898) 170 U.S. 343, 352 [42 L.Ed. 1061, 1061, 18 S.Ct. 620], italics added; see Smith, supra, 34 Cal.3d at p. 260.) The vagueness inherent in the term “substantial protections,” which the high court did not define, led us to observe that “[a]s to all but the most obvious examples of ex post facto legislation,. . . the general rule is that there is no general rule.” (Smith, supra, 34 Cal.3d at p. 260.) For this reason, we concluded that case-by-case analysis of Proposition 8’s provisions under the ex post facto clause was not practical.
The United States Supreme Court has greatly simplified ex post facto law since Smith. In Collins, supra,_ U.S._ [111 L.Ed.2d 30,110 S.Ct. 2715], the high court endorsed its earlier formulation of the law in Calder v. Bull (1798) 3 U.S. (3 Dall.) 386, 390 [1 L.Ed. 648, 650] (Calder), and Beazell v. Ohio (1925) 269 U.S. 167, 169-170 [70 L.Ed. 216, 217-218, 46 S.Ct. 68] [294]*294(Beazell), as “faithful to our best knowledge of the original understanding” of the ex post facto clause. (Collins, supra,_U.S. at p._ [111 L.Ed. 2d at p. 39, 110 S.Ct. at p. 2719].) Under that exclusive formulation, “ ‘any statute [1] which punishes as a crime an act previously committed, which was innocent when done; [2] which makes more burdensome the punishment for a crime, after its commission, or [3] which deprives one charged with crime of any defense available according to law at the time when the act was committed, is prohibited as ex post facto.' ”8 (Collins, supra,_ U.S. at p._ [111 L.Ed.2d at p. 39, 110 S.Ct. at p. 2719], quoting Beazell, supra, 269 U.S. at pp. 169-170 [70 L.Ed. at p. 217].)
In Collins the Supreme Court also rejected the proposition that a law violates the ex post facto clause simply because it eliminates a “substantial protection” existing at the time an oifense was committed.9 According to the high court, references in its earlier cases “to ‘substantial protections’ and ‘personal rights’ should not be read to adopt without explanation an undefined enlargement of the [ex post facto clause].” (Collins, supra, _ U.S. at p._ [111 L.Ed.2d at p. 41, 110 S.Ct. at p. 2721].) While the Legislature, or the electorate, cannot immunize a law from scrutiny under the ex post facto clause simply by labelling it “procedural,” “the prohibition which may not be evaded is the one defined by the Calder categories.” (Ibid. [Ill L.Ed.2d at p. 41, 110 S.Ct. at p. 2721].)
Accordingly, the United States Supreme Court has resolved the analytical difficulty which led us to conclude in Smith that it would be impractical to consider ex post facto challenges to Proposition 8’s provisions on a case-by-case basis. After Collins, we need not determine “how substantial is the [295]*295right that the statute impairs and how significant is that impairment.” (Smith, supra, 34 Cal.3d at p. 260.) Instead, we can resolve such challenges by applying the exclusive Colder categories.
Lastly on this point, Tapia argues that we should perpetuate “substantial protection” analysis by adopting it as a matter of state law. We decline to do so. While we unquestionably have the power to interpret a provision of the state Constitution differently than its federal counterpart (Cal. Const., art. I, § 24), neither the language nor the history of the state ex post facto clause supports a different interpretation.
The language of the state and federal clauses is identical in relevant part: each declares that no “ex post facto law” shall be passed. (Cal. Const., art. I, § 9; U.S. Const., art. I, § 10, cl. 1.) The state clause was first proposed in the constitutional convention of 1849 as an amendment to a committee draft of article I, the “Declaration of Rights,” apparently as an afterthought. (Rep. of the Debates in the Convention of California on the Formation of the State Constitution, in Sept, and Oct. 1849 (1850) at p. 43.) The record of the debates does not include any discussion of the amendment.10 In 1879, the delegates to the second constitutional convention voted to preserve the language adopted in 1849, again without any recorded debate on its meaning.11 In 1971, as part of an ongoing project of constitutional reform, the Constitution Revision Commission proposed to change the wording of the clause from the negative (“No ... ex post facto law . . . shall ever be passed”) to the positive (“A[n]... ex post facto law . . . may not be passed”). However, the commission did not propose any substantive change.12 The voters approved the provision, as reworded, in the November 1974 General Election.
[296]*296Nor does an independent meaning for the state ex post facto clause emerge from our own opinions. Although this court has often cited the state clause together with the federal, we have never given the state clause an independent interpretation or held, in any case, that the state clause required a different result than the federal clause. Moreover, nothing in our opinions supports the notion that “substantial protection” analysis has an independent footing in the state Constitution. The United States Supreme Court introduced “substantial protection” analysis in 1883 with its opinion in Kring v. Missouri, supra, 107 U.S. 221. California, however, was not quick to follow. It was not until 1930 that we cited Kring (People v. Dawson (1930) 210 Cal. 366, 368 [292 P. 267]), and not until 1958 that we clearly followed “substantial protection” analysis (People v. Ward, supra, 50 Cal.2d 702, 707) as the supremacy clause required us to do. (U.S. Const., art VI, cl. 2.)
Before Kring, this court’s understanding of the rule against ex post facto laws was inconsistent with “substantial protection” analysis. Before us in People v. Mortimer (1873) 46 Cal. 114 was a new statute giving the prosecution, for the first time, the right to open and to close the argument to the jury in criminal cases. The issue was whether the trial court had erred by applying the statute to the trial of a crime committed before the statute’s effective date. (Id., at pp. 114, 116.) We considered the claim of error under Penal Code section 6,13 which we said was “intended ... to steer clear of any difficulty arising from the enactment of ex post facto laws, which are prohibited by the Constitution of the United States. But laws changing the mere forms of procedure in a criminal action” do not fall within the ex post facto clause’s prohibition. (46 Cal.3d at p. 118.) We also stated that “[i]t is clear . . . that no constitutional difficulty would be encountered in requiring past offenses to be tried under new forms of procedure . . . .” (Ibid.) We did not even cite the state Constitution, let alone suggest that it dictated a different result or required us to consider whether the particular procedural law in question eliminated a “substantial protection.” As already mentioned, we would not clearly follow “substantial protection” analysis for another 75 years. (See People v. Ward, supra, 50 Cal.2d at p. 707.)
As the high court’s recent opinion in Collins, supra, _ U.S. _ [111 L.Ed.2d 30, 110 S.Ct. 2715], demonstrates, our apparent reluctance to [297]*297employ “substantial protection” analysis was justified. Federal law had compelled us to follow that analysis, since we were never free to adopt an ex post facto rule less favorable to the defense than the federal rule.14 After Collins, there is no reason not to return to the original understanding of the scope of the rule against ex post facto laws expressed in both federal and state Constitutions.
The third reason we gave for our decision in Smith was the desire to “avoid a number of practical consequences adverse to the administration of justice and the right of fair trial.” (Smith, supra, 34 Cal.3d at p. 262.) As we have already discussed, the intervening simplification of ex post facto law eliminates the major administrative difficulty that we foresaw in Smith. Although we cannot altogether avoid further litigation concerning Proposition 115’s application in particular cases, we must weigh that administrative burden against the substantial benefits that immediate application of the measure offers by reducing time spent in voir dire and preliminary hearings, among other things. On balance, considerations of administrative efficiency, as well as the electorate’s stated goals of reducing delay and unnecessary cost, favor giving Proposition 115 the earliest possible application.
V.
It remains to be determined which of Proposition 115’s provisions may and may not be applied to the prosecution of crimes committed before the measure’s effective date. The provisions fall into four categories: (A) provisions which change the legal consequences of criminal behavior to the detriment of defendants; (B) provisions which address the conduct of trials; (C) provisions which clearly benefit defendants; and (D) a single provision which codifies existing law.
A.
The first category of provisions, those which change the legal consequences of criminal behavior to the detriment of defendants, cannot be applied to crimes committed before the measure’s effective date. These [298]*298provisions include section 915 (amending Pen. Code, § 189), which adds crimes to the list of felonies supporting a conviction of first degree murder; those portions of section 10 (amending Pen. Code, § 190.2) which add new special circumstances; that portion of section 10 (codified at Pen. Code, § 190.2, subd. (d)), which provides that an accomplice, for a felony-murder special circumstance to be found true, must have been a major participant and have acted with reckless indifference to human life;16 section 11 (adding Pen. Code, § 190.41), which provides that the corpus delicti of a felony-based special circumstance need not be proved independently of the defendant’s extrajudicial statement; section 12 (amending Pen. Code, § 190.5), which subjects persons between the ages of 16 and 18 to the penalty of life without possibility of parole for first degree murder with special circumstances; sections 13 and 14 (adding Pen. Code, §§ 206 and 206.1), which define the new crime of torture; and section 26 (adding Pen. Code, § 1385.1), which precludes a judge from striking a special circumstance that has been admitted or found to be true.
Application of these provisions to crimes committed before the measure’s effective date would be “retrospective” because each would change the legal consequences of the defendant’s past conduct. (Weaver v. Graham, supra, 450 U.S. at p. 31 [67 L.Ed.2d at p. 24]; see also People v. Weidert, supra, 39 Cal.3d at p. 851, quoting Weaver v. Graham; cf. Aetna Casualty, supra, 30 Cal.2d at p. 394.) Such application would also likely violate the rule against ex post facto legislation, since each of these provisions appears to define conduct as a crime, to increase punishment for a crime, or to eliminate a defense.17(Collins, supra,___ U.S. at p._ [111 L.Ed.2d at p. 39, 110 S.Ct. [299]*299at p. 2719].) Accordingly, these provisions may only be applied to prosecutions of crimes committed on or after June 6, 1990.
B.
Other provisions of Proposition 115 address the conduct of trials rather than the definition of, punishment for, or defenses to crimes. These provisions include section 2 (adding Cal. Const., art. I, § 14.1), which eliminates postindictment preliminary hearings; section 4 (adding Cal. Const., art. I, § 29), which gives the People the right to due process and a speedy trial; section 5 (adding Cal. Const., art. I, § 30), which provides that the Constitution shall not be construed to prohibit joinder, makes hearsay evidence admissible at preliminary hearings, and makes discovery reciprocal; sections 6, 7, and 7.5 (repealing Code Civ. Proc., §§ 223 and 223.5, and adding a new Code Civ. Proc., § 223), which reform voir dire; sections 8, 15, 16, 17, and 18 (adding Evid. Code, § 1203.1, amending Pen. Code, §§ 859, 866, and 872, and adding Pen. Code, § 871.6), which reform preliminary hearing procedures; section 19 (adding Pen. Code, §954.1), which provides that the absence of cross-admissibility is not a ground for severance; section 20 (adding Pen. Code, § 987.05), which requires appointment of counsel who is ready to proceed; section 21 (adding Pen. Code, § 1049.5), which provides that felony trials shall take place within 60 days of arraignment; section 22 (adding Pen. Code, § 1050.1), which authorizes continuances to maintain joinder; sections 23, 24, 25, and 27 (adding Pen. Code, §§ 1054 to 1054.7, and repealing Pen. Code, §§ 1102.5, 1102.7, and 1430), which reform discovery procedures and provide for reciprocal discovery; and section 28 (adding Pen. Code, § 1511), which provides for appellate review of trial dates and continuances.
Tapia has advanced several arguments in an effort to show that application of such provisions in his case, even though addressed on their face to the conduct of future trials and not to past criminal behavior, will nevertheless be “retrospective” as applied to him. First, he argues that the new voir dire provisions will limit his counsel’s ability to explore potential jurors’ exposure to pretrial publicity. If counsel had known of the new rule, to complete the argument, he might have sought an earlier ruling on a motion to change venue rather than waiting for the results of voir dire. The flaw in this argument is the premise. There is no reason to believe that the new voir dire rules will be applied to deprive Tapia’s counsel of any information to which voir dire is legitimately directed. Under the new provisions, “the court may permit the parties, upon a showing of good cause, to supplement the examination by such further inquiry as it deems proper, or shall itself [300]*300submit to the prospective jurors upon such a showing, such additional questions by the parties as it deems proper.” (§ 7, codified as Code Civ. Proc., § 223.)
Next, Tapia argues that the new provisions for reciprocal discovery would be retrospective as applied to him because his counsel has conducted his investigation based upon the assumption that discovery would not be reciprocal. In this case the objection is purely hypothetical because the prosecutor has informed the trial court that he does not intend to utilize the reciprocal discovery provisions. All parties and amici curiae, however, have stressed the importance of an early resolution of issues concerning Proposition 115’s application and have asked us to address such issues in this case.
Application of the discovery provisions to compel production of evidence obtained by defense counsel before Proposition 115’s effective date would be retroactive under the principles we have already discussed. This is because counsel can only be guided, while conducting an investigation, by the discovery rules then in force. In contrast, the date on which the crime was committed has no bearing on the discoverability of the results of counsel’s investigation. Discovery rules address trial preparation, not criminal behavior. Trial courts are able to determine, in camera if necessary, whether particular evidence claimed to be subject to discovery was obtained before or after Proposition 115’s effective date.
Finally, Tapia argues that counsel might not have permitted the trial to be continued if counsel had known that to do so would have required him to defend the case under the new provisions which Proposition 115 introduced. In this way, to complete the argument, the new provisions change the legal consequences of counsel’s past conduct. In our view, this argument does not demonstrate that the new provisions are retrospective. In view of the many reasons that can affect the decision to seek a continuance, to say that any such decision was motivated by reliance on the state of the law unaffected by Proposition 115 is speculative in the extreme. Moreover, it has always been understood in this state that the rules governing the conduct of trials are subject to change by the Legislature, or the electorate, insofar as the Constitution permits. (People v. Mortimer, supra, 46 Cal. at pp. 118-119.)
Accordingly, the provisions enumerated above may be applied to pending cases regardless of when the charged offense is alleged to have occurred.
C.
The third category of new provisions consists of those which clearly benefit only defendants. Each of these provisions is contained in section 10. [301]*301Specifically, the section provides that the killing of a prosecutor or judge, to constitute a special circumstance, must be intentional. (§ 10, codified at Pen. Code, § 190.2, subd. (a)(ll), (12).) The section also provides that an accomplice, for a non-felony-murder special circumstance to be found true, must have had the intent to kill. (§ 10, codified at Pen. Code, § 190.2, subd. (c).)
Application of these provisions to trials of crimes committed before Proposition 115’s operative date may change the legal consequences of a defendant’s criminal conduct. Such application is permissible, however, because the provisions favor defendants. Although we usually presume that new statutes are intended to operate prospectively, that presumption “is not a straitjacket.” (In re Estrada, supra, 63 Cal.2d at p. 746.) In past cases we have not applied the presumption to statutes changing the law to the benefit of defendants. Instead, we have assumed that “[w]hen the Legislature amends a statute so as to lessen the punishment it has obviously expressly determined that its former penalty was too severe and that a lighter punishment is proper as punishment for the commission of the prohibited act” and “ ‘sufficient to meet the legitimate ends of the criminal law.’ ” (Id., at p. 745, quoting People v. Oliver (1956) 1 N.Y.2d 152 [151 N.Y.S.2d 367, 134 N.E.2d 197, 201-202].) We have applied the same reasoning to statutes which redefine, to the benefit of defendants, conduct subject to criminal sanctions. (People v. Rossi (1976) 18 Cal.3d 295 [134 Cal.Rptr. 64, 555 P.2d 1313].)18 These authorities compel the conclusion that the provisions listed above may be applied to pending cases.
D.
Finally, section 10 also codifies the rule of People v. Anderson, supra, 43 Cal.3d 1104, to the effect that an actual killer, for a special circumstance to be found true, need not have had the intent to kill unless the applicable special circumstance specifically so requires. (§ 10, codified at Pen. Code, § 190.2, subd. (b); see People v. Anderson, supra, at p. 1147, overruling Carlos v. Superior Court (1983) 35 Cal.3d 131 [197 Cal.Rptr. 79, 672 P.2d 862]; see also People v. Hendricks (1987) 43 Cal.3d 584, 596, fn. 3 [238 Cal.Rptr. 66, 737 P.2d 1350].) This provision, which does not change [302]*302existing law, may be applied to crimes committed before Proposition 115’s effective date.
Disposition
The alternative writ is discharged. The judgment of the Court of Appeal is affirmed.
Lucas, C. J., Kennard, J., Arabian, J., and Baxter, J., concurred.