OPINION OF THE COURT

BRICKLEY, J.

These cases require us to decide whether the prohibition in either the United States or Michigan Constitution against placing a person twice in jeopardy prohibits, in a single trial, convictions of both first-degree criminal sexual conduct under MCL 750.520b(1)(c); MSA 28.788(2)(1)(c) (penetration under circumstances involving any "other felony") and the underlying "other felony" of either armed robbery or kidnapping used to prove the charge of first-degree criminal sexual conduct. We hold that convictions of both first-degree criminal sexual conduct and the underlying felony of armed robbery or kidnapping in a single trial are not barred by either constitution.

Herbert Robideau was convicted of first-degree criminal sexual conduct, two counts of armed robbery, MCL 750.529; MSA 28.797, and possession of a firearm during the commission of a felony, MCL 750.227b; MSA 28.424(2). The Court of Appeals affirmed his convictions, finding no violation of the prohibitions against double jeopardy, notwithstanding that one of the armed robbery counts was used as the "other felony" to establish the crime of first-degree criminal sexual conduct. 94 Mich. App. 663; 289 N.W.2d 846 (1980). We granted leave to appeal. 412 Mich. 871 (1981).

Chester Bouknight was convicted of first-degree criminal sexual conduct, two counts of armed robbery, and possession of a firearm during the commission of a felony. One of the armed robbery counts was used as the "other felony" to establish the crime of first-degree criminal sexual conduct. The Court of Appeals vacated that armed robbery conviction, finding that, although the multiple convictions did not violate the federal constitution, the convictions of first-degree criminal sexual conduct and the underlying "other felony" did violate the Michigan Constitution. 106 Mich. App. 798; 308 N.W.2d 703 (1981). We granted leave to appeal. 412 Mich. 871 (1981).

Freddy Lee Brown was convicted of two counts of first-degree criminal sexual conduct (penetration under circumstances involving any other felony and another penetration while armed, MCL 750.520b[e]; MSA 28.788[2][1][e]). Brown was also convicted of kidnapping, MCL 750.349; MSA 28.581, the "other felony" used to establish one count of first-degree criminal sexual conduct. The Court of Appeals reversed Brown's convictions for penetration under circumstances involving any other felony and for kidnapping because of an instructional error on the authority of People v Barker, 411 Mich. 291; 307 N.W.2d 61 (1981). The Court affirmed his conviction under MCL 750.520b(1)(e); MSA 28.788(2)(1)(e) and remanded the case for retrial. The Court instructed that on remand Brown, consistent with the provisions of the federal constitution, could not again be convicted of both penetration under circumstances involving any other felony and the underlying "other felony" of kidnapping. 105 Mich. App. 58; 306 N.W.2d 392 (1981). We granted leave to appeal. 412 Mich. 864 (1981).

Randall Dale Griffin was convicted of first-degree criminal sexual conduct and kidnapping, with the kidnapping being used to establish the "other felony" for first-degree criminal sexual conduct. The Court of Appeals affirmed his conviction in an unpublished opinion per curiam, released August 14, 1980. We granted leave to appeal. 412 Mich. 871 (1981).

The basic contour of the Double Jeopardy Clause was succinctly set forth in North Carolina v Pearce, 395 U.S. 711, 717; 89 S Ct 2072; 23 L Ed 2d 656 (1969):

"It protects against a second prosecution for the same offense after acquittal. It protects against a second prosecution for the same offense after conviction. And it protects against multiple punishments for the same offense."

In Benton v Maryland, 395 U.S. 784; 89 S Ct 2056; 23 L Ed 2d 707 (1969), the United States Supreme Court held that the federal Double Jeopardy Clause was applicable to actions by the states. The present cases concern the third protection, the protection against multiple punishment, and, in order to decide them, as well as to give some guidance in an area that has been described as a state of "confusion upon confusion" and a "veritable Sargasso Sea", we must begin with a review of the multiple-punishment aspects of the Double Jeopardy Clause as declared by the United States Supreme Court, as best those decisions can be interpreted.

We begin by stating what is, at least at the time of this decision, settled law. Where multiple punishment is involved, the Double Jeopardy Clause acts as a restraint on the prosecutor and the courts, not the Legislature. Brown v Ohio, 432 U.S. 161; 97 S Ct 2221; 53 L Ed 2d 187 (1977).

The Double Jeopardy Clause prohibits a court from imposing more punishment than that intended by the Legislature. "[T]he question whether punishments imposed by a court after a defendant's conviction upon criminal charges are unconstitutionally multiple cannot be resolved without determining what punishments the Legislative Branch has authorized". Whalen v United States, 445 U.S. 684, 688; 100 S Ct 1432; 63 L Ed 2d 715 (1980). The most recent expression of this principle is found in Ohio v Johnson, ___ US ___; 104 S Ct 2536, 2541; 81 L Ed 2d 425, 433 (1984):

"In contrast to the double jeopardy protection against multiple trials, the final component of double jeopardy — protection against cumulative punishments — is designed to ensure that the sentencing discretion of courts is confined to the limits established by the legislature. Because the substantive power to prescribe crimes and determine punishments is vested with the legislature, United States v Wiltberger, 18 US (5 Wheat) 76, 93; 5 L Ed 37 (1820), the question under the Double Jeopardy Clause whether punishments are `multiple' is essentially one of legislative intent, see Missouri v Hunter, 459 U.S. 359; 103 S Ct 673; 74 L Ed 2d 535 (1983)."

As an aid to determining the intent of a legislature, the Supreme Court has used an approach which, although first established in Gavieres v United States, 220 U.S. 338, 342; 31 S Ct 421; 55 L Ed 489 (1911), is now known as the Blockburger test because of its use in Blockburger v United States, 284 U.S. 299, 304; 52 S Ct 180; 76 L Ed 306 (1932):

"The applicable rule is that where the same act or transaction constitutes a violation of two distinct statutory provisions, the test to be applied to determine whether there are two offenses or only one, is whether each provision requires proof of a fact which the other does not."

If two statutes constitute the "same offense" under the Blockburger test, it is presumed that the Legislature did not intend to allow the defendant to be punished under both statutes.

"The assumption underlying the rule is that [the Legislature] ordinarily does not intend to punish the same offense under two different statutes." Whalen v United States, supra, pp 691-692.

The presumption raised by the test can be rebutted by a clear indication of legislative intent, and when such an intent is found, it must be respected, regardless of the outcome of the Blockburger test. Albernaz v United States, 450 U.S. 333, 340;

101 S Ct 1137; 67 L Ed 2d 275 (1981); Missouri v Hunter, 459 U.S. 359; 103 S Ct 673; 74 L Ed 2d 535 (1983).

Although the Blockburger test is simply stated, it is subject to more than one interpretation, depending on the statutes involved. When one of the two statutes involved is a necessarily lesser included offense of the other, application of the Blockburger test will always raise the presumption that the two statutes involve the "same offense". The necessarily included lesser offense will never have an element not required by the greater offense. Problems arise, however, when one statute is not a lesser included offense of the other. Such is the case with the so-called compound and predicate crimes that are involved in the cases before us. The compound crime of first-degree criminal sexual conduct requires proof of a penetration during the commission of any other felony. The "any other felony" is the predicate crime, in these cases, either armed robbery or kidnapping.

If one looks only to the statutory language, the compound and predicate crimes are separate under the Blockburger test. First-degree criminal sexual conduct requires proof of a penetration, an element not required by either kidnapping or robbery. Armed robbery requires proof of an armed taking from the person or in the presence of another, and kidnapping requires the movement or confinement of another. None of these elements are, in the strict sense, required for the crime of first-degree criminal sexual conduct. Any felony will do.

If, however, one looks to how the crimes with which the defendants were charged were actually proved — the evidence used — the compound crime of first-degree criminal sexual conduct and the predicate crime of either robbery or kidnapping are not separate offenses under Blockburger. In the cases of Robideau and Bouknight, the prosecution convicted the defendants of the predicate crime of armed robbery and also used the predicate crime of armed robbery to establish the "other felony" to convict the defendants of the compound crime of first-degree criminal sexual conduct. Similarly, in the cases of Griffin and Brown, the defendants were convicted of the predicate crime of kidnapping, and the kidnapping was used to establish the "other felony" to convict them of first-degree criminal sexual conduct. Looking to the facts of the case instead of at the statutes in the abstract in all four cases before us, the prosecutor was required to prove the predicate crime to establish the compound crime. If Blockburger is applied to the facts of the cases, the compound and predicate crimes are presumptively the same offense.

This Court and the United States Supreme Court have applied both approaches to the Blockburger test for reasons not always explicable. The resulting confusion has been compounded by a failure to adequately distinguish between the different interests involved in successive-prosecution versus single-prosecution cases. Nevertheless, we must begin with a review of some of the double jeopardy cases decided by the United States Supreme Court to determine whether the Blockburger test is to be applied in the abstract or to the facts of the cases before us, or whether it should be applied at all.

In Blockburger, the defendant was convicted in a single trial of selling narcotics not in or from their original stamped package and of selling narcotics without an official written order. The two convictions resulted from the same sale. The Court looked to the statutory definitions of the crimes, as evidenced by the Court's statement that "[e]ach of the offenses created requires proof of a different element". Id., p 304. The Court concluded that because the first statute required proof that the narcotics were not sold in or from their original container, while the second did not, and the second statute required proof that the sale was made without an official order, while the first did not, the two crimes were separate. Therefore, the defendant could be doubly punished, notwithstanding the legislative "double-description" of crimes that had made one act into two crimes.

Although Blockburger did not specifically state that the test used was for determining legislative intent by looking to the statutory elements of a crime, later decisions made that point clear. In Gore v United States, 357 U.S. 386; 78 S Ct 1280; 2 L Ed 2d 1405 (1958), the Court stressed that Blockburger had been decided as a matter of legislative intent. In Iannelli v United States, 420 U.S. 770, 785, fn 17; 95 S Ct 1284; 43 L Ed 2d 616 (1975), the Court, in concluding that under the Blockburger test conspiracy to violate gambling laws and violation of gambling laws were separate offenses, stated that "the test focuses on the statutory elements of the offense * * * notwithstanding a substantial overlap in the proof offered to establish the crimes".

The Blockburger test was applied in a case involving a successive prosecution in Brown v Ohio, 432 U.S. 161; 97 S Ct 2221; 53 L Ed 2d 187 (1977). The case involved a defendant who was convicted of joy-riding. After his release from jail, he was charged with automobile theft, a crime of which joy-riding, under Ohio law, was a necessarily included lesser offense, i.e., all the elements of joy-riding were also elements of automobile theft. As is always the case when true necessarily included lesser and greater offenses are involved, the Court's application of the Blockburger test resulted in a finding that the two crimes constituted the "same offense" and, therefore, were not to be separately punished.

Justice Brennan, in his concurring opinion, did not apply the Blockburger test. He instead treated the case as if issues of multiple punishment were not involved and proposed deciding the case on the ground that the interest of the defendant in avoiding a second trial barred the second prosecution. In previous cases involving successive prosecutions, the Court had applied the factual approach of In re Nielsen, 131 U.S. 176; 9 S Ct 672; 33 L Ed 118 (1889). If the same evidence used in the first trial was to be substantially used in the second trial, the defendant could not constitutionally be tried a second time. As stated in Nielsen, supra, p 188:

"[A] person [who] has been tried and convicted for a crime which has various incidents included in it * * * cannot be a second time tried for one of those incidents without being twice put in jeopardy for the same offense."

But Justice Brennan would have gone beyond the approach of Nielsen. He viewed all successive prosecutions as barred by the Double Jeopardy Clause if the second trial involved a part of the same transaction litigated in the first trial. So long as the same transaction of criminal events was involved, Justice Brennan would require that all charges possible be brought against the defendant in the first trial.

The majority's decision in Brown was criticized as reaching the right decision for the wrong reason. See Westen Drubel, Toward a General Theory of Double Jeopardy, 1978 Supreme Court Review 81, 157. Perhaps, however, the most valid criticism is that the Court failed to fully explain its reasoning. The opinion never addressed why the traditional successive prosecution analysis of Nielsen, supra, was not applied.

In our opinion, the Brown Court's failure to apply the reasoning of Nielsen appears to be explained by its statement that "[w]here the judge is forbidden to impose cumulative punishment for two crimes at the end of a single proceeding, the prosecutor is forbidden to strive for the same result in successive proceedings". Brown, supra, p 166. It thus appears that Brown was decided as if the only issue were one of multiple punishment. Multiple punishment could not be imposed under the statutes in question, regardless of the number of trials. It was therefore unnecessary for the Court to decide whether, if allowed, multiple punishments could be imposed in separate trials.

In Harris v Oklahoma, 433 U.S. 682; 97 S Ct 2912; 53 L Ed 2d 1054 (1977), another case involving successive prosecution which also presented to the Court its first compound and predicate crimes, the Court inexplicably reverted to the analysis of Nielsen and ignored Blockburger. The Oklahoma Legislature had linked two crimes, allowing the predicate crime of robbery to act as one of the aggravating crimes which raised second-degree murder to the compound crime of first-degree felony murder. The defendant had been tried and convicted of first-degree felony murder in the course of a robbery. The defendant was subsequently tried and convicted of the predicate robbery.

The Supreme Court's opinion per curiam reversing the subsequent robbery conviction, without mentioning Blockburger, simply stated that the conviction of the greater crime could not be had without a conviction of the lesser crime, and for authority relied upon Nielsen. Unlike Brown, the case was decided on the ground that it involved a successive prosecution. Doing so suggests that the crimes in question had satisfied the Blockburger test and that multiple punishments in a single trial would be allowed. That result can be reached only if one applies the Blockburger test to the elements of the crimes in question in the abstract, i.e., first-degree murder requires proof of a killing, which robbery, of course, does not, and robbery requires proof of a robbery, which is not required to establish first-degree murder (under Oklahoma law, any specified felony would do). Applying the Blockburger test to the facts of the case before the Court would show that robbery was, in fact, required to prove first-degree murder, and it would, therefore, have been unnecessary to consider the case under the analysis of Nielsen.

If the Court in Harris had indeed silently applied the Blockburger test in the abstract to the statutes before it, its method of analysis could be considered consistent with prior law. After all, numerous decisions had stated that Blockburger looked to the statutory elements of a crime. Yet, when faced with the next case involving a compound and predicate crime, the Court applied Blockburger to the facts of the case.

In Whalen v United States, 445 U.S. 684; 100 S Ct 1432; 63 L Ed 2d 715 (1980), the defendant was charged in a single prosecution with the compound crime of first-degree felony murder and also with rape, a predicate crime which elevated second-degree murder to first. In an apparent effort to achieve the same result as that reached in Harris, notwithstanding that Harris was decided in the context of successive prosecutions, the Court, despite a footnote disclaimer to the contrary, refocused Blockburger away from the elements of the charged crimes. The result was that the Court transformed the Blockburger test into one that looked to the conduct which led to the charges against the defendant or to the actual facts proved at trial, all the while suggesting that they could be indicative of legislative intent:

"In this case, resort to the Blockburger rule leads to the conclusion that Congress did not authorize consecutive sentences for rape and for a killing committed in the course of the rape, since it is plainly not the case that `each provision requires proof of a fact which the other does not.' A conviction for killing in the course of rape cannot be had without proving all the elements of the offense of rape. See United States v Greene, 160 US App DC 21, 34; 489 F.2d 1145, 1158 (1973). Cf. Harris v Oklahoma, 433 U.S. 682, 682-683; 97 S Ct 2912; 53 L Ed 2d 1054 (1977). The Government contends that felony murder and rape are not the `same' offense under Blockburger, since the former offense does not in all cases require proof of a rape; that is, DC Code § 22-2401 (1973) proscribes the killing of another person in the course of committing rape or robbery or kidnaping or arson, etc. Where the offense to be proved does not include proof of rape — for example, where the offense is a killing in the perpetration of a robbery — the offense is of course different from the offense of rape, and the Government is correct in believing that cumulative punishments for the felony murder and for a rape would be permitted under Blockburger. In the present case, however, proof of rape is a necessary element of proof of the felony murder, and we are unpersuaded that this case should be treated differently from other cases in which one criminal offense requires proof of every element of another offense." Whalen, supra, pp 693-694.

The dissenting justices recognized the majority's transformation of Blockburger into a factual test, noting that doing so was without precedent and contrary to the avowed purpose of the test to determine legislative intent. Nevertheless, the dissenters did not propose to decide the case using Blockburger. They instead proposed looking to all sources of legislative history rather than relying on the "arbitrary assumptions" and "syntactical subtleties" implicit in any application of Blockburger to compound and predicate crimes. Whalen, supra, p 712.

The application of Blockburger to the facts of a given case when multiple punishment was involved proved to be a short-lived practice. In Albernaz v United States, 450 U.S. 333, 338; 101 S Ct 1137; 67 L Ed 2d 275 (1981), the Court repaired to its description of the Blockburger test as one which "`focuses on the statutory elements of the offense'". The Court found that the offense of conspiracy to import marijuana and conspiracy to distribute marijuana could be separately punished even though on the facts of the case there was but one conspiracy. The Court's conclusion that the two statutes before it were separate under Blockburger, however, did not end its inquiry. The Court noted that the Blockburger test was merely a means to determine legislative intent. The presumptions created by the test could be rebutted by a clear indication of legislative intent to the contrary.

The Supreme Court found such a clear indication of legislative intent in Missouri v Hunter, supra. Hunter involved a single prosecution where the defendant was convicted of the compound crime of "armed criminal action" (possession of a dangerous weapon during the commission of any felony) and the underlying predicate crime on the facts of the case, armed robbery. The Missouri Court of Appeals reversed the defendant's conviction of armed criminal activity, relying on decisions of the Missouri Supreme Court, which had applied the Blockburger test to the facts of the cases before it and concluded that the Double Jeopardy Clause was offended by convictions of both armed criminal action and the predicate felony. See State v Haggard, 619 S.W.2d 44 (Mo, 1981). The Missouri Supreme Court reached that conclusion notwithstanding the Missouri Legislature's express provision for double punishment in the statute.

The United States Supreme Court did not disapprove the Missouri Supreme Court's interpretation of Blockburger, but found that the Missouri Legislature had expressly rebutted the presumption raised by the statute's failure to pass the Blockburger test. Even though armed criminal action and the underlying predicate, armed robbery, were the same offense under the Whalen factual approach to Blockburger, multiple punishment was permissible because the Legislature had so intended.

The United States Supreme Court's treatment of issues of multiple punishment suggests a struggle, and an unsuccessful one, to prescribe a single standard applicable to the statutes considered. It especially evidences recent difficulty in multiple punishment cases where compound, or legislatively linked, crimes are involved. In multiple punishment cases, the Court, with one exception, has applied the Blockburger test to the statutory elements of crimes. This one exception, however, which looked to the facts of a case involving a compound and predicate crime, is analogous to the present case. The question left unanswered by the Supreme Court is whether the analysis of Whalen is the proper analysis when compound and predicate crimes are involved. The inconsistency of federal case law forces us to turn to state law for guidance.

In comparison with the United States Supreme Court's treatment of issues of multiple punishment, our Court's experience has been similarly rocky, although we have followed a somewhat different path. As was the experience of the United States Supreme Court, the first double jeopardy cases to come before this Court involved successive prosecutions. In People v Cox, 107 Mich. 435; 65 N.W. 283 (1895), the defendant was tried and convicted of keeping a bawdy house. He was later tried and convicted of keeping the same house as a house of ill fame under a separate statute. In an opinion by Justice HOOKER, this Court reversed the defendant's conviction, finding that on the facts of the case one statute was included within the other. The opinion of the United States Supreme Court in Nielsen, supra, was found to be "decisive".

By the time this Court first dealt with an issue of multiple punishment in a single trial, a factual, same-evidence approach was well established in successive prosecution cases. See People v Cook, 236 Mich. 333; 210 N.W. 296 (1926); People v Beverly, 247 Mich. 353; 225 N.W.2d 481 (1929). In People v White, 390 Mich. 245; 212 N.W.2d 222 (1973), this Court even extended the double jeopardy protection from successive prosecutions by adopting, under the Michigan Constitution, Justice Brennan's transactional approach. See Ashe v Swenson, 397 U.S. 436, 448; 90 S Ct 1189; 25 L Ed 2d 469 (1970) (Brennan, J., concurring). This approach expands the United States Supreme Court's same-evidence test by forbidding a second trial when part of the criminal transaction has been subject to a prior trial.

In People v Martin, 398 Mich. 303; 247 N.W.2d 303 (1976), this Court faced its first single-prosecution, multiple-punishment case. The defendant had been convicted both of delivery of heroin and possession of heroin. Without reference to Blockburger or to legislative intent, we found that the prohibition against double jeopardy had been violated because "[p]ossession of the heroin present in this case was that necessary to its delivery". Id., p 307. The rationale for looking to the facts of the case to see if any facts were doing double duty with respect to the two statutes was People v Cook, supra, a successive prosecution case. See, also, People v Stewart (On Rehearing), 400 Mich. 540; 256 N.W.2d 31 (1977), which applied the same analysis and reached the same result under the statutes replaced by those considered in Martin.

This Court first faced a compound and predicate crime situation in Wayne County Prosecutor v Recorder's Court Judge, 406 Mich. 374; 280 N.W.2d 793 (1979), app dis sub nom Brintley v Michigan, 444 U.S. 948; 100 S Ct 418; 63 L Ed 2d 317 (1979). There, the two defendants had been convicted of the compound crime of possession of a firearm during the commission of a felony and also of an underlying predicate felony, either murder or armed robbery. We applied the Blockburger test to the statutes in the abstract and found that the two crimes were separate: murder does not require use of a firearm and felony-firearm does not require the predicate felony to be murder; armed robbery does not require the use of a firearm and felony-firearm does not require the predicate felony to be armed robbery. More importantly, we recognized for the first time that issues of multiple punishment were to be decided by determining what the Legislature intended. Because the Legislature had expressly provided that felony-firearm was to be punished separately from and consecutive to the predicate felony, all double jeopardy concerns were satisfied. Since the decision in Wayne County Prosecutor, however, we have failed to look to legislative intent.

In People v Jankowski, 408 Mich. 79; 289 N.W.2d 674 (1980), the defendant, in a transaction arising out of a robbery of a gas station, was charged with and convicted in a single trial of larceny over $100, larceny in a building, and armed robbery. We applied the Blockburger test, but to the facts of the case rather than to the statutes in the abstract. As a matter of Michigan constitutional law, we found that multiple punishment under the statutes was impermissible because on the facts of the case before the Court the same facts proved all three crimes.

In People v Wilder, 411 Mich. 328; 308 N.W.2d 112 (1981), we confronted, as did the United States Supreme Court in Whalen, a single prosecution where the defendant was convicted of both the compound crime of first-degree felony murder and of the underlying predicate felony. Again, we applied a method of analysis taken from successive-prosecution cases such as Cook and Harris, and looked to the facts of the case. Because, in the case before the Court, proof of the charged predicate felony also served as proof of the compound felony of first-degree murder, we found that convictions of both crimes on the facts of the case violated the prohibition against double jeopardy.

We next decided People v Carter, 415 Mich. 558; 330 N.W.2d 314 (1982). There, the defendant was convicted in a single trial of both aiding and abetting the crime of extortion and conspiracy to commit extortion. We found no federal double jeopardy violation after applying the Blockburger test to the statutes in the abstract. Each crime required proof of an element not required by the other. We also found no violation of the prohibition against double jeopardy under a factual approach even though the facts used to establish the aiding and abetting could also have established the conspiracy. Other facts also had been introduced into evidence which could have proved the conspiracy. In essence, for the reason that the prosecutor over-proved his case, we found no violation of the Michigan prohibition against double jeopardy. Still, we recognized that it is the intent of the Legislature which should be determinative. That realization is implicit in our statement that the crimes involved were not of the "class or category". Id., p 586.

Our latest experience with issues of multiple punishment came in People v Wakeford, 418 Mich. 95; 341 N.W.2d 68 (1983). There, we dealt not with the issue of multiple punishment under two statutes, but, instead, with multiple punishment under a single statute. The issue before the Court was whether the robbery of two grocery store cashiers constituted one or two robberies. Our focus was entirely on the robbery statute and legislative intent. Because only one statute was involved, an analysis under Blockburger was unnecessary. Our task was only to determine what act the Legislature had made criminal.

We concluded in Wakeford that the Legislature intended separate robbery convictions for each person robbed, the focus of robbery being on the assault against a person and not on the owner of the property taken. In doing so, we rejected the claim that the defendant's convictions were invalid because some of the same evidence was used in proving both robberies. Wakeford, however, does not answer the question now before us. Multiple punishment analysis when only one statute is involved has never been burdened by a shifting interpretation of Blockburger.

As can be seen from the above review, when more than one statute is involved, Michigan's double jeopardy analysis has been no more consistent than federal double jeopardy analysis. We do come to the conclusion that much of the inconsistency in double jeopardy analysis results from the failure to clearly distinguish between single-prosecution and successive-prosecution cases. We will, therefore, undertake that task.

Successive-prosecution cases involve the core values of the Double Jeopardy Clause, the common-law concepts of autrefois acquit and convict. See Bartkus v Illinois (On Rehearing), 359 U.S. 121, 151; 79 S Ct 676; 3 L Ed 2d 684 (1959) (Black, J., dissenting). Where successive prosecutions are involved, the Double Jeopardy Clause protects the individual's interest in not having to twice "run the gauntlet", in not being subjected to "embarrassment, expense and ordeal", and in not being compelled "to live in a continuing state of anxiety and insecurity", with enhancement of the "possibility that even though innocent he may be found guilty". Green v United States, 355 U.S. 184, 187, 190; 78 S Ct 221; 2 L Ed 2d 199 (1957).

To protect this interest, the United States Supreme Court has adopted a factual-evidentiary approach, Nielsen, supra, and, where appropriate, a collateral estoppel approach, Ashe v Swenson, supra. Both these approaches look to the evidence produced at the first trial to determine if the same facts or issues will be proved at the second. But these approaches are limited. They allow for the possibility that a transaction could be divided into discrete offenses, with successive trials for each offense. To preclude that possibility, this Court, in People v White, adopted Justice Brennan's transactional approach as the standard by which the propriety of a second prosecution is judged under the Michigan Constitution.

Different interests are involved when the issue is purely one of multiple punishments, without the complications of a successive prosecution. The right to be free from vexatious proceedings simply is not present. The only interest of the defendant is in not having more punishment imposed than that intended by the Legislature. The intent of the Legislature, therefore, is determinative. Under neither the federal nor the Michigan double jeopardy provisions does this Court sit as a superlegislature, instructing the Legislature on what it can make separate crimes. As previously described, prior decisions of this Court have applied a factual test in single-trial multiple-punishment cases, creating areas in which arguably the Legislature cannot now act. To the extent that those decisions interpret the prohibition against double jeopardy as a substantive limitation on the Legislature, we now disavow them.

We are therefore left only with the question of what the Legislature intended in cases such as those at bar. As a means of determining that end, we find the Blockburger test to have questionable status in the Supreme Court of the United States and find the propriety of its use in any case to be questionable. When applied in the abstract to the statutory elements of an offense, it merely serves to identify true lesser included offenses. While it may be true that the Legislature ordinarily does not intend multiple punishments when one crime is completely subsumed in another, Blockburger itself is of no aid in making the ultimate determination. Although its creation of a presumption may make a court's task easier, it may also induce a court to avoid difficult questions of legislative intent in favor of the wooden application of a simplistic test.

Because Blockburger was developed to deal with situations where an identifiable single act falls under the coverage of two statutes, it is even less helpful when applied to a compound crime. In these crimes, the Legislature has intentionally converted what would normally be two discrete acts into one legislatively created "act".

The difficulties with the Blockburger test lead us to the conclusion that it should be abandoned. The United States Supreme Court has declared that it is but a test of statutory construction and not a principle of constitutional law. Missouri v Hunter, supra. Indeed, it would be quite contrary to established principles of federalism for the United States Supreme Court to impose on the states the method by which they must interpret the actions of their own legislatures. We, therefore, find it within our power to reject the Blockburger test, preferring instead to use traditional means to determine the intent of the Legislature: the subject, language, and history of the statutes.

At times, this will be a difficult task. The Legislature rarely reveals its intentions with a specific statement. Yet the difficulty in determining the intent of the Legislature cannot be cause for furthering the use of a test which reveals little, if anything, about legislative intent.

Some general principles can be stated. Statutes prohibiting conduct that is violative of distinct social norms can generally be viewed as separate and amenable to permitting multiple punishments. A court must identify the type of harm the Legislature intended to prevent. Where two statutes prohibit violations of the same social norm, albeit in a somewhat different manner, as a general principle it can be concluded that the Legislature did not intend multiple punishments. For example, the crimes of larceny over $100, MCL 750.356; MSA 28.588, and larceny in a building, MCL 750.360; MSA 28.592, although having separate elements, are aimed at conduct too similar to conclude that multiple punishment was intended.

A further source of legislative intent can be found in the amount of punishment expressly authorized by the Legislature. Our criminal statutes often build upon one another. Where one statute incorporates most of the elements of a base statute and then increases the penalty as compared to the base statute, it is evidence that the Legislature did not intend punishment under both statutes. The Legislature has taken conduct from the base statute, decided that aggravating conduct deserves additional punishment, and imposed it accordingly, instead of imposing dual convictions.

We do not intend these principles to be an exclusive list. Whatever sources of legislative intent exist should be considered. If no conclusive evidence of legislative intent can be discerned, the rule of lenity requires the conclusion that separate punishments were not intended.

As evidenced by the divisions in the United States Supreme Court in Whalen and in this Court in Wayne County Prosecutor, legislatively linked compound and predicate crimes are not subject to an easy analysis. Without a doubt, sexual penetration, regardless of the circumstances, involves social norms distinct from those norms violated by the predicate felony, in these cases, robbery or kidnapping. Yet, because the compound crime of first-degree criminal sexual conduct incorporates all the elements of the predicate felony, it cannot be concluded that the compound and predicate felonies are entirely distinct. The focus of first-degree criminal sexual conduct, however, is clearly on penetration. Cf. People v Willie Johnson, 406 Mich. 320; 279 N.W.2d 534 (1979). The common element of the many ways to establish first-degree criminal sexual conduct is penetration. Although not conclusive, that suggests a legislative intent that first-degree criminal sexual conduct and the underlying predicate felony were to be separately punished.

First-degree criminal sexual conduct carries a maximum penalty of life imprisonment. The predicate crimes in the instant cases also carry maximum penalties of life imprisonment. This very strongly suggests that the Legislature intended that the compound crime of first-degree criminal sexual conduct and the predicate crimes of robbery and kidnapping be separately punished.

1974 PA 266, which created the crime of criminal sexual conduct, was a radical restructuring of the rape laws. Convictions were made more easily obtainable and punishments increased. See Boyle, The Criminal Sexual Conduct Act, 43 Det Lawyer 6 (Oct, 1975). Unlike robbery and armed robbery, where robbery is subsumed in armed robbery and a greater punishment is imposed, first-degree criminal sexual conduct and those predicate crimes which normally occur along with first-degree criminal sexual conduct, including armed robbery and kidnapping, have equal punishments. If the predicate crime is considered subsumed, there is no greater punishment. To suggest that the Legislature intended first-degree criminal sexual conduct to subsume a co-equal felony would be to attribute to the Legislature a total lack of purpose. It is only when a conviction is allowed of both the predicate felony and of first-degree criminal sexual conduct that there is any purpose in having the crime of penetration under circumstances involving any other felony.

If we were to conclude that only one conviction could result from fact situations such as the cases at bar, the result would be that the defendants, having completed the predicate felonies of kidnapping and robbery, could then embark on one of the most heinous crimes possible, with no risk either of a second conviction or a statutorily increased maximum sentence. We would not lightly attribute such an unthinking and callous attitude to the Legislature in its adoption of the compound felony of first-degree criminal sexual conduct. The contrary conclusion unmistakably comes forth. The Legislature intended that the crimes of first-degree criminal sexual conduct and either robbery or kidnapping be separately punished.

In Robideau, the decision of the Court of Appeals is affirmed.

In Bouknight, the decision of the Court of Appeals is reversed.

In Griffin, defendant's convictions must be reversed because the trial court failed to instruct the jury in accordance with the dictates of People v Barker, 411 Mich. 291; 307 N.W.2d 61 (1981). The cause is remanded to the circuit court for further proceedings consistent with this opinion.

In Brown, the decision of the Court of Appeals is affirmed, the directions to the circuit court on remand are vacated, and the cause is remanded to the circuit court for further proceedings consistent with this opinion.

WILLIAMS, C.J., and RYAN and BOYLE, JJ., concurred with BRICKLEY, J.

LEVIN, J. (separate opinion).

The defendants were convicted of criminal sexual conduct in the first degree on evidence of sexual penetration in circumstances involving the commission of another felony — armed robbery in the cases of Robideau and Bouknight, and kidnapping in the cases of Griffin and Brown.

Griffin's appeal should not be decided at this time because the issue presented by his motion for a directed verdict of acquittal on the kidnapping charge is implicated in People v Wesley, argued May 4, 1983, which will be decided at an early date.

I

In each of these consolidated cases, the defendants committed an aggravated theft or kidnapping and a sexual assault, thus violating two societal norms. The Double Jeopardy Clause is not violated by their convictions of a theft or kidnapping offense and a criminal sexual conduct offense.

II

While the Legislature may define an offense as including elements of another offense, the question whether offenses are the "same" within the meaning of the Double Jeopardy Clause presents an issue of constitutional construction to be decided by the Court. Punishing the offender beyond the 15-year maximum for criminal sexual conduct in the third degree because the sexual assault was committed in circumstances involving the commission of another felony and additionally punishing him for that felony punishes him twice for the same felony.

When multiple CSC 1 — robbery/kidnapping convictions become final, a defendant is entitled, by virtue of the Double Jeopardy Clause, to have the judgment entered against him amended so that he is not punished twice for the same offense. The prosecutors in the instant cases should now be put to an election whether the sentence imposed for CSC 1 should be reduced to the maximum sentence for CSC 3, with no change in the sentence for the predicate robbery or kidnapping conviction, or whether, alternatively, to eliminate the sentence for the predicate robbery or kidnapping conviction with no change in the sentence imposed for CSC 1 and with no change in sentences for non-predicate armed robbery or kidnapping counts.

Robideau was sentenced to concurrent terms of 26-1/2 to 40 years for convictions on two counts of armed robbery (one, the predicate felony) and 20 to 35 years for CSC 1 and a consecutive 2-year term for felony-firearm.

Bouknight was sentenced to concurrent terms of 8 to 30 years for CSC 1 and for convictions on two counts of armed robbery (one, the predicate felony) and a consecutive 2-year term for felony-firearm.

Brown was sentenced to concurrent terms of 20 to 40 years for kidnapping (the predicate felony) and two counts of CSC 1, one based on the kidnapping conviction and the other on being armed with a weapon.

Griffin was sentenced to concurrent sentences of 15 to 25 years for kidnapping and CSC 1.

Thus, if the prosecutor were required to elect and were to elect in Robideau to eliminate the sentence on the predicate armed robbery conviction, Robideau would be required to serve 26-1/2 to 40 years for the non-predicate armed robbery conviction and, concurrently, 20 to 35 years for CSC 1 and a 2-year consecutive sentence for felony-firearm. A similar election in Bouknight would require Bouknight to serve 8 to 30 years for the non-predicate armed robbery conviction and, concurrently, 8 to 30 years for CSC 1 and the 2-year consecutive term for felony-firearm. In Brown, if the sentence for the CSC 1 conviction based on a predicate felony were reduced to the maximum sentence for CSC 3, Brown would still be required to serve 20 to 40 years for kidnapping and, concurrently, 20 to 40 years for the CSC 1 conviction based on being armed with a weapon. In Griffin, however, the prosecutor would be put to an election between the 15- to 25-year sentence for kidnapping and a reduction of the sentence for the sexual assault to the maximum for CSC 3 or alternatively 15 to 25 years for CSC 1 and elimination of the kidnapping sentence.

In all events, the defendants would, as a practical matter, serve exactly the same number of days in prison and have exactly the same number of felony convictions.

III

Less than two years ago, in People v Carter, 415 Mich. 558, 582; 330 N.W.2d 314 (1982), this Court, on review of its earlier multiple punishment decisions in People v Martin, 398 Mich. 303; 247 N.W.2d 303 (1976), People v Stewart (On Rehearing), 400 Mich. 540; 256 N.W.2d 31 (1977), People v Jankowski, 408 Mich. 79; 289 N.W.2d 674 (1980), and People v Wilder, 411 Mich. 328; 308 N.W.2d 112 (1981), stated that Michigan's Double Jeopardy Clause offers "broader double jeopardy protection than do the federal standards. * * * However, in Michigan the emphasis is not on the theoretical elements of the crimes involved, but rather upon the proof of facts adduced at trial".

The Court today "disavows" its earlier decisions to the extent that they depart from the United States Supreme Court's latest pronouncement in Missouri v Hunter, 459 U.S. 359; 103 S Ct 673; 74 L Ed 2d 535 (1983), which is without decisional support in that Court's earlier decisions and is based solely on dictum.

This Court's new test launches the trial courts, the Court of Appeals, and this Court on an elusive search for clear or conclusive evidence of legislative intent. In finding such evidence in the instant cases, the majority states that the circumstance that the predicate crimes of armed robbery and kidnapping carry maximum penalties of life in prison, as does CSC 1, "very strongly suggests" that the Legislature intended separate punishments for CSC 1 and those predicate crimes. This ignores that the predicate felony might be a 2-, 4-, 10- or 15-year felony and that it is unlikely the Legislature intended separate punishment in respect to some predicate felonies and not others and suggests the absence of clear or conclusive evidence of legislative intent to provide for cumulative or separate sentencing.