Defendant appeals from convictions of rape in the second degree, ORS 163.365, and sexual abuse in the first degree, ORS 163.427. Defendant was sentenced on each conviction to 75 months’ imprisonment pursuant to ORS 137.700 to be served concurrently and a post-prison supervision period of 10 years, less time served in prison. Defendant appeals, and we affirm.

In his first assignment of error, defendant argues that the trial court erred when it ordered him to pay witness fees in the amount of $25. ORS 161.665(1) now provides, in part:

“[T]he court, only in the case of a defendant for whom it enters a judgment of conviction, may include in its sentence thereunder a provision that the convicted defendant shall pay as costs expenses specially incurred by the state in prosecuting the defendant. Costs include a reasonable attorney fee for counsel appointed pursuant to ORS 135.045 or 135.050 and a reasonable amount for expenses approved under ORS 135.055. * * * Costs shall not include expenses inherent in providing a constitutionally guaranteed jury trial or expenditures in connection with the maintenance and operation of government agencies that must be made by the public irrespective of specific violations of law.”

At trial, three witnesses testified for the prosecution, only one of which was a police officer. Apparently, the trial court imposed the witness fees for the nonpolice witnesses. Defendant argues that the award of witness fees to the state for those witnesses falls within the ORS 161.665 prohibition for recovery of expenses inherent in providing a constitutionally guaranteed jury trial.

In State v. Hastings, 24 Or App 123, 544 P2d 590, rev den (1976), we held that costs incurred so that the state could prove an indictment against the defendant in the form of witness fees were not an expenditure within the exclusion of ORS 161.665. Defendant argues that Hastings was wrongly decided and that under ORS 136.602,1 the state is financially *418responsible for its witnesses and a defendant is financially responsible only for his or her witnesses. We disagree with defendant’s arguments. Witness fees for witnesses called by the prosecution are not the kind of expenses inherent in providing a constitutionally guaranteed jury trial or expenditures in connection with the maintenance and operation of government agencies that must be made by the public irrespective of specific crimes. Unlike general costs of prosecutions, such as district attorneys’ salaries, sheriffs’ salaries, jurors’ fees or police investigations, the costs of witness fees are costs specific to a defendant’s own case. As we said in Hastings, “[pjayment of witness fees in a particular case is not an expenditure made ‘irrespective of specific violations of law.’ ” 24 Or App at 125. Moreover, the language of ORS 136.602(1) does not require a contrary interpretation of ORS 161.665(1). ORS 136.602(1) provides that “[ejxcept as otherwise specifically provided by law” the state will initially pay the expenses of prosecution witnesses. When read in conjunction with ORS 161.665, it is clear that the trial court has the authority to order a defendant to repay witness fees that have been paid in the first instance by the state.

Next, defendant argues that ORS 137.700 violates Article I, section 16, of the Oregon Constitution, which provides that “all penalties shall be proportioned to the offense.” ORS 137.700 is a codification of Ballot Measure 11 adopted by the people in the 1994 general election. Or Laws 1995, ch 2 (amended by Or Laws 1995, ch 421, § 1; Or Laws 1995, ch 422, § 47). The measure adopted mandatory minimum sentences for certain felonies. Id. Defendant argues:

“Ballot Measure 11 as applied to the crime of rape in the second degree and sexual abuse in the first degree is disproportionate because ORS 137.700-mandated sentenced] for those crimes, Class B felonies, [are] greater than the *419greatest actual sentence that could be imposed for some Class A felonies.

“The legislature has classified all felonies within the criminal code into three categories of escalating seriousness. Those categories are Class C felonies (the least serious), Class B felonies and Class A felonies (the most serious barring murder, aggravated murder and treason). ORS 161.535.

“The sentences to be imposed upon these different classifications are set forth in other statues and rules. See e.g., ORS 161.605 (setting forth indeterminate sentences); Or Laws 1989, ch 790, § 87 (legislature approved the sentencing guidelines); Ballot Measure 11 (voters enacted determinate sentences for some crimes); ORS 137.700. Although the legislature and voters have changed the punishment for the different classification of felonies over the years, they have not changed the classification system itself.

í]í # if;

“Because defendant’s punishment for committing a Class B felony under ORS 137.700 exceeds the greatest punishment an offender with the same or greater criminal history could receive for committing some Class A felonies - more serious crimes according to the legislature’s classification system, the sentence mandated by ORS 137.700 for rape in the second degree and sexual [abuse] in the [first] degree is disproportionate to the offense. As such, it is unconstitutional.”2 (Internal footnote omitted.)

Defendant cites no authority for his argument. In State v. Shumway, 291 Or 153, 630 P2d 796 (1981), Merrill v. Gladden, 216 Or 460, 337 P2d 774 (1959), and Cannon v. Gladden, 203 Or 629, 281 P2d 233 (1955), Article I, section 16, of the Oregon Constitution, was held applicable to situations where sentences for lesser included crimes were more *420severe than for their completed counterparts. The comparison was made between related crimes, and such comparisons demonstrated the irrationality of imposing a greater penalty for a less severe offense. In Cannon, the defendant was sentenced to life imprisonment for the crime of assault with intent to commit rape. Under the law that existed at the time, one found guilty of either statutory or forcible rape could not be imprisoned for more than 20 years. In Merrill, the plaintiff was convicted of the crime of assault with intent to rob and sentenced to serve a term of 20 years. When the plaintiff brought a habeas corpus proceeding in circuit court, the circuit court judge entered an order directing his release. At that time, the penalty for unarmed robbery was a maximum of 15 years, and the penalty for armed robbery had a maximum penalty of life because of the aggravating circumstance. The court held that “the legislature has full authority to provide a greater penalty for a crime committed in an aggravated manner than one that is not.” Merrill, 216 Or at 465. The court concluded:

“Since it was the legislative intent to include both an armed robbery and an unarmed robbery within the assault with intent to rob statute, and since the indictment was sufficient to describe an armed assault, it follows that the trial court was not prohibited by constitution from entering the sentence of 20 years, pronounced against the plaintiff.” Id. at 470.

In Shumway, the defendant was convicted of intentional homicide and sentenced to life. Under the existing law, he was required to serve 25 years before becoming eligible for parole, as required by an initiative measure. Under other statutes governing intentional homicide and aggravating circumstances, the defendant would have been eligible for parole either 15 or 20 years after sentencing, depending on the particular aggravating circumstance. The court concluded:

“Under this statutory scheme, a defendant receives a lesser minimum sentence to be served before being eligible for parole for aggravated intentional homicide than he does for an unaggravated intentional homicide. This is in violation of Art[icle] I, [section] 16[,] of the Oregon Constitution *421and that provision in ORS 163.115(5) requiring the defendant to serve not less than 25 years before becoming eligible for parole is invalid and cannot be applied to the defendant * * Shumway, 291 Or at 164.

In State v. Turner, 296 Or 451, 455-56, 676 P2d 873 (1984), the defendant argued that a 15-year minimum imprisonment term of a 30-year dangerous offender sentence for attempted rape was disproportionate to the sentence for murder because there was no minimum term that could be imposed for murder at that time. He relied on the holdings in Cannon and Shumway. The com! noted first that in those cases, it had compared sentences for different degrees of the same crime where the sentence for the lesser crime was more severe than the sentence for the greater crime. The court said, “[t]hese cases[, Cannon, Merrill and Shumway,] are distinguishable from the present case because here we are asked to compare the sentences of unrelated crimes.” Turner, 296 Or at 456. The court went on to point out that upon a murder conviction, a sentence for the remainder of a person’s life is a greater sentence than a 30-year sentence with a 15-year minimum. The court concluded the 15-year minimum sentence imposed on the defendant for a crime where no homicide was involved was not disproportionate.

In Isom, 313 Or at 399-400, the defendant argued that the death penalty for aggravated murder after an escape was unconstitutionally disproportionate under Article I, section 16, of the Oregon Constitution, to the sentence for intentional murder during an escape. The defendant relied on the holdings in Cannon and Shumway. The court rejected his argument:

“Those cases, however, do not support defendant’s contention here. Each of those cases concerned a situation in which, due to definitional peculiarities, a lesser included offense carried a greater penalty than the principle offense. That is not the situation. * * * The legislature has chosen to subject all such persons to the maximum potential penalty. Defendant’s opinion makes sense, but so does that which we attribute to the legislature. There was a rational basis for the legislature to conclude that both classes of escapees are dangerous.” Isom, 313 Or at 400 (emphasis in original).

*422In light of these precedents, defendant’s argument is not well taken. On each occasion that the Supreme Court has had the opportunity to consider an argument under section 16, it has determined whether there is a rational basis for the severity of a particular penalty for a particular offense in comparison to the penalties imposed for other offenses. Under the current status of the law, there are some presumptive sentences that exceed the minimum sentences under Measure 11. In those cases, a court continues to sentence under the sentencing guidelines as enacted by the legislature. However, there are other crimes in which Measure 11 imposes a greater sentence than that imposed under the sentencing guidelines. In those cases, the trial court cannot impose both the Measure 11 sentence and the presumptive guidelines sentence because they conflict. Although both sentencing schemes exist concurrently, the enactment of Measure 11 means that the Measure 11 sentence trumps the guidelines sentence.3 State ex rel Huddleston v. Sawyer, 324 Or 597, 601-05, 932 P2d 1145, cert den_US_, 118 S Ct 557, 139 L Ed 2d 399 (1997).

In this case, defendant argues that Measure 11 is unconstitutional because it requires a 75-month sentence for the crimes of first-degree sexual abuse and second-degree rape. Both crimes are Class B felonies, and their Measure 11 sentences are in excess of the penalties allowed for some Class A felonies, such as first-degree burglary and first-degree arson.4 However, the guideline penalties for those *423Class A felonies are part of a different sentencing structure. First-degree sexual abuse and second-degree rape are not lesser-included offenses of first-degree burglary and first-degree arson. Rather, they are unrelated offenses for which the people through the initiative process have promulgated a different degree of severity than that imposed originally by the legislature. The people, as well as the legislature, are entitled to make determinations about what punishments should be imposed for crimes. See Or Const, Art IV, § 1(2).5 See also Vannatta v. Keisling, 324 Or 514, 528, 931 P2d 770 (1997) (holding that since the adoption of Article IV, section 1, of the Oregon Constitution, in 1902, “the people have shared with the Legislature Assembly the power to enact laws”). Thus, so long as there is a rational basis for the people of the State of Oregon to conclude that first-degree sexual abuse and second-degree rape deserve greater penalties than felonies such as first-degree burglary and first-degree arson, the requirement of section 16 is met. See Isom, 313 Or at 400.

The differences between first-degree sexual abuse and second-degree rape and first-degree burglary and first-degree arson are apparent. The former involve injury to and the defilement of human beings. They are “person” crimes. A primary principle of criminal law is “the protection and safety of the people of the state.” Tuel v. Gladden, 234 Or 1, 6, 379 P2d 553 (1963). The latter are “property’ crimes. It is rational to consider property crimes less grievous than crimes against persons. Moreover, defendant fails to take into account that the legislature also considers crimes against persons to be more aggravated than crimes against property. For instance, if the crime of burglary in the first degree is committed when the burglar is armed with a deadly weapon or causes or threatens physical injury to the victim, it is classified as a 9 on the crime seriousness scale. OAR 253-04-002, Appendix 3.6 If the burglary cannot be classified as a *4249 but the offense was committed in an occupied dwelling, it is classified as an 8 on the scale. OAR 253-04-002, Appendix 3. If the burglary is classified as a 9 and a substantial and compelling reason exists to depart from the presumptive sentence, a defendant’s maximum sentence authorized under the sentencing guidelines could be up to 144 months with a double departure. OAR 253-04-002, Appendix 3; OAR 253-08-001 to -004. Similarly, a burglar could be sentenced up to 90 months with a double departure if the offense was classified as an 8. OAR 253-04-002, Appendix 3; OAR 253-08-001 to -004. In addition, when the crime of arson in the first degree represents a threat of serious physical injury, the guidelines provide for a maximum allowable sentence of 260 months’ imprisonment with a double departure. OAR 253-04-002, Appendix 3; OAR 253-08-001 to -004. Under the circumstances, we conclude that the people of the State of Oregon rationally could believe that the just desserts for first-degree sexual abuse or second-degree rape is a 75-month sentence, and accordingly, we reject defendant’s argument.

In his final assignment of error, defendant argues on various grounds that Measure 11 violates the federal constitution. Those arguments have been rejected in Huddleston, 324 Or 597; State v. Skelton, 153 Or App 580, 589 nn 8-9, 957 P2d 585, rev den 327 Or 448 (1998); George, 146 Or App 449; and State v. Parker, 145 Or App 35, 929 P2d 327 (1996), rev den 324 Or 654 (1997). Defendant does not make an “as applied” challenge to Measure 11 on federal constitutional grounds.

Affirmed.

dissenting.

The statutes that apply to defendant’s crimes show on their face that his Measure 11 sentence is disproportionate to the seriousness of his offenses. Because the majority fails even to recognize the relevant issues, I dissent.

The constitutional violation in this case is the result of the following actions of the legislative authority of the state:1 (1) it has expressly determined the relative seriousness of almost all felonies; (2) it has established sentencing *425guidelines that produce sentences for most felonies that are roughly proportionate to the determination of the relative seriousness of the crimes; (3) it has, through Measure 11, established sentences for some felonies that are far greater than the sentences for other felonies that it has determined to be equally serious. As a result, persons who commit equally serious crimes will receive significantly disproportionate punishments when one person’s crimes are Measure 11 offenses and the other person’s are not. That situation on its face appears to violate the express requirement of Article I, section 16, of the Oregon Constitution, that all punishments be proportioned to the offense.

The majority not only finds nothing constitutionally defective in this situation; it does not even recognize that there is a constitutional question that requires consideration. Because I cannot so easily dismiss a patent violation of an express constitutional requirement, I dissent.

The relevant facts are straightforward. Defendant was convicted of rape in the second degree and sexual abuse in the first degree, both arising out of the same criminal episode. In accordance with ORS 137.700 (Measure 11), the court sentenced him to concurrent mandatory minimum 75-month terms, with no opportunity for work release or good-time reduction.2 On appeal, defendant asserts that those sentences violate Article I, section 16, of the Oregon Constitution (section 16), because they are not proportioned to the offenses *426for which he was convicted. Although the focus of his argument is on the distinction between categories of felonies, I agree with the basis of his understanding of section 16 and of its effect on his sentences. I would remand for resentencing under the sentencing guidelines.3

The foundation of defendant’s attack on the sentences is that Measure 11 has distorted the legislature’s attempt to make sanctions for crimes proportionate to the legislative determination of their relative seriousness. That distortion, he asserts, resulted in his sentence being significantly harsher than sentences for non-Measure 11 offenses that the legislature has determined to be as serious as, or more serious than, the ones that he committed. At the trial court he emphasized two aspects of Measure 11: First, a Measure 11 sentence in his case is much greater than the presumptive sentence for the same offense under the sentencing guidelines; second, a Measure 11 sentence for the Class B felonies of which he was convicted is greater than the presumptive guidelines sentences for a number of Class A felonies. On appeal, he focuses on the latter point, the disparity between his Measure 11 sentence for the lesser offense of a Class B felony and the guidelines sentence for the greater offense of a Class A felony. Defendant’s essential point has considerable force when applied to the guidelines themselves, because they now contain the primary legislative determination of the relative seriousness of felonies. I consider it in that context.

I begin with the requirement of proportionality in section 16. That requirement, in essence, is that the punishment for crimes be related to their seriousness: punishments for equally serious crimes should be approximately equal, while punishments for crimes that are more serious should be greater than punishments for crimes that are less serious. That requirement is related to, but distinct from, the separate prohibitions in section 16 of excessive fines and of punishments that are cruel and unusual. The Oregon courts have been hesitant to hold that punishments for distinct offenses *427are disproportionate to each other, in part because of the difficulty of a court making the essentially legislative determination of whether one offense is more serious than another. Thus, the only punishments that the courts have actually overturned as disproportionate under section 16 involved greater punishments for an attempt to commit a crime than for the completed offense; in that situation there is no difficulty in determining which offense is more serious.4

I next describe the sentencing guidelines and how they affect proportionality analysis under section 16. The guidelines constitute a careful legislative determination of the relative seriousness of almost all felonies and provide for sentences that the legislature has determined to be proportioned to the seriousness of the offense and the dangerousness of the offender. As a result of the guidelines, courts can now measure the proportionality of sentences effectively, because they can rely on the legislature’s own determination of the seriousness of offenses. Measure 11 superimposed new, generally more severe, sentences for some felonies onto the existing guidelines without modifying the existing determination of the relative seriousness of the offenses involved. Because of the guidelines, a court is able to evaluate the proportionality under section 16 of those new sentences without intruding into an essentially legislative arena. Such an evaluation reveals that offenses that the legislature has determined to be equally serious may have quite different punishments and that crimes that are less serious may have greater punishments than those that are more serious. That result is as much a violation of section 16 as is a greater sentence for an attempt than for the completed offense; in both situations, the unquestionably lesser offense receives the greater punishment. Because defendant’s sentence under Measure 11 is subject to those flaws, it is invalid.

Section 16 provides, in part:

“Excessive bail shall not be required, nor excessive fines imposed. Cruel and unusual punishments shall not be *428inflicted, but all penalties shall be proportioned to the offense.” (Emphasis added.)

Aside from the prohibitions of excessive bail and excessive fines, section 16 expressly contains two related but distinct concepts, each of which applies to punishments for crimes: it both prohibits cruel and unusual punishments and requires that penalties be proportioned to the offense. Sustar v. County Court for Marion Co., 101 Or 657, 665, 201 P 445 (1921). In this respect, section 16 is different from the Eighth Amendment, which expressly prohibits only cruel and unusual punishments.5 Some cases do not adequately distinguish between those two concepts, but a careful analysis shows that the requirement of proportionality raises distinct considerations that operate independently of the prohibition of cruel and unusual punishment.

The Supreme Court first applied section 16 as a limit on criminal sanctions in State v. Ross, 55 Or 450, 104 P 596, 106 P 1022 (1910), appeal dismissed 227 US 150, 33 S Ct 220, 57 L Ed 458 (1913). Part of the defendant’s sentence in Ross was a fine of $576,853.74; if he could not pay the fine, he had to work it off at the rate of one day in the county jail for every two dollars of the fine, or approximately 790 years in jail. In its original opinion, the court deleted the alternative of jail, noting that, while it was within the letter of the law, it was cruel and unusual punishment; the court left the fine intact. 55 Or at 474. On reconsideration, it also remitted the fine as excessive, explaining that it was greater than the defendant could pay with a lifetime of effort. Id. at 480.

In contrast to its extensive discussion of other issues in the case, the court in Ross limited its analysis of section 16 to conclusory statements and citations to a few cases from other jurisdictions; it apparently believed that the answer was obvious. The primary value of the case for later decisions was to make clear that a statutorily authorized sentence *429could still violate section 16 and to provide an example of evaluating the severity of a sentence. The court did not expressly rely on or discuss the requirement of proportionality.

The first case that did expressly consider the requirement of proportionality under section 16 was Sustar, which was a writ of review proceeding, not a direct criminal appeal. The petitioner argued that his sentence of six months in the county jail and a $500 fine for possession of two quarts of moonshine was both disproportionate and cruel and unusual. The court first stated that the trial court had gone to the verge by giving the petitioner the maximum penalty under the law; nevertheless, it stated, a fine within the limits of the statute was generally legal and not subject to judicial review. Sustar, 101 Or at 662. It then cited a number of cases from other states that held that similar sentences for prohibition violations were not cruel, unusual, or excessive. Id. at 662-64.

After these comments, the court turned to the relevant portion of section 16, which, it noted,

“contains three prohibitions and one command:

“ T. Excessive bail shall not be required.

“ ‘2. Nor excessive fines imposed.

“ ‘3. Cruel and unusual punishment shall not be inflicted.

“ ‘4. All penalties shall be proportioned to the offense.’ ”

Id. at 665 (quoting section 16).

The court relied on Weems to construe section 16. After quoting the statement of the United States Supreme Court that “ ‘[i]t is a precept of justice that punishment for crime should be graduated and proportioned to the offense,’ ” 101 Or at 665, (quoting 217 US at 367), the Oregon court added:

“In order to justify the court in declaring punishment cruel and unusual with reference to its duration, the punishment must be so proportioned to the offense committed as to shock the moral sense of all reasonable men as to what is right and proper under the circumstances: Weems v. United States, supra.”

In considering the petitioner’s case under these principles, the court emphasized that the prohibition law carried out a *430fundamental public policy and that it was the duty of the officers of the law and the courts to enforce that policy. It then affirmed the denial of the writ of review, explaining:

“There is nothing before this court authorizing it to declare that the penalty in the instant case is excessive. The statute is not in conflict with the constitutional provision hereinbefore noted. We assume that the penalty was proportioned to the offense as commanded by Section 16, Article I, Oregon Constitution. There may, or may not, have been aggravating circumstances. A writ of review is a special proceeding, and we cannot review the evidence.” Sustar, 101 Or at 665-66 (emphasis added).

Sustar is unclear on the precise relationship between the requirement of proportionality and the prohibition of cruel and unusual punishment, a lack of clarity that can be seen in the court’s treatment of Weems. The court first quoted Weems as stating as a general principle that justice requires proportionality; immediately thereafter it cited the same case to support a standard for cruel and unusual punishment that can be met only by a high degree of disproportionality. Within one paragraph, thus, the court shifted from a discussion of the requirement of proportionality to the prohibition of cruel and unusual punishment, concepts that, as it had recognized immediately before, are separate and distinct aspects of section 16. What made the shift possible, and what confused some later courts, was that the standard that the court adopted for cruel and unusual punishment has a proportionality component.6

It is essential, however, that what the Oregon court described was the standard for determining when a prison *431term is cruel and unusual: the sentence must be so disproportionate to the offense as to shock the moral sense of all reasonable people. That standard, which cannot be found in Weems, at least relates to the same constitutional terms that the Court considered in Weems. It does not, however, relate to the express requirement of proportionality that section 16 contains and that the Eighth Amendment lacks. Yet, by both defining cruel and unusual punishment in a way that included gross disproportionality and also recognizing and applying a separate constitutional proportionality requirement, Sustar made possible subsequent confusion between the two standards.7

Sustar contains criteria for determining the proportionality of a sentence that are separate and distinct from its express standard for cruel and unusual punishment. Those criteria suggest that the essence of the requirement of proportionality is the existence of a reasonable relationship between the punishment and the offense. The court’s direct (and correct) quotation from Weems treated such a relationship as a basic requirement of justice. The court also stated that the petitioner’s sentence went to the verge, suggesting that a greater punishment might well have been disproportionate. It examined the penalties that other states provided for the same offense, suggesting that such an examination was relevant under section 16. In its dispositive paragraph, the court noted that there may, or may not, have been aggravating circumstances. Because the evidence was not before it, the court assumed that the penalty was proportioned to the offense, suggesting that it might have reached a different conclusion if it had had the evidence. 101 Or at 665-66. In short, the court in Sustar evaluated the proportionality of the sentence independently of the shock the moral sense standard that it created for determining when punishment is cruel and unusual. Its decision is unclear primarily because *432the petitioner’s sentence was not disproportionate under any standard, and the court therefore did not need to be more explicit than it was.8

Two subsequent cases indicate that the court did not treat Sustar as adopting the shock the moral sense test for proportionality issues. In Kelley v. Meyers, 124 Or 322, 263 P 903 (1928), the defendant was convicted of assisting in the escape of a prisoner. He challenged the sentence, arguing that it was disproportionate because the length of the sentence varied depending on the crime for which the escapee was incarcerated. The court upheld the statute, noting that it is a more serious offense to help a person charged with or convicted of highway robbery to escape than to help a person charged with or convicted of a misdemeanor. The different punishment was not arbitrary and thus did violate section 16. “[T]he difference in the penalties to be imposed * * * has a proper relation to the enormity of the offense committed or attempted[.]” 124 Or at 328-30 (quotation at 330). In State v. Coffman, 171 Or 166, 173, 136 P2d 687 (1943), the court held that a sentence of two years’ imprisonment for negligent homicide was neither disproportionate nor cruel and unusual, noting the difference between that sentence and the almost 790 years involved in Ross, on which the defendant relied. In both of these cases the court examined the relationship between the offense and the sentence and, exercising its independent judgment, determined that it was proportionate.

In contrast, in Cannon v. Gladden, 203 Or 629, 281 P2d 233 (1955), the court relied on the Sustar standard for cruel and unusual punishment to decide what, as the court subsequently recognized, was a clear proportionality case. The plaintiff in Cannon sought habeas corpus relief on the ground that he was serving a life sentence for assault with intent to commit rape when, if he had been convicted of the rape charged in the indictment, his sentence could have been no more than 20 years. The court cited Ross for its authority to modify an unconstitutionally excessive sentence and *433quoted its discussion in Sustar of the requirements of section 16, including the distinction between the prohibition of cruel and unusual punishment and the requirement of proportionality. It then held that a greater sentence for a lesser offense is so disproportioned to the offense as to shock the moral sense of all reasonable men.

“It is unthinkable, and shocking to the moral sense of all reasonable men as to what is right and proper, that in this enlightened age jurisprudence would countenance a situation where an offender, either on a plea or verdict of guilty to the charge of rape, could be sentenced to the penitentiary for a period of not more than 20 years, whereas if he were found guilty of the lesser offense of assault with intent to commit rape he could spend the rest of his days in the bastille.” Cannon, 203 Or at 632-33.

One would have thought that a greater punishment for a lesser offense raised straightforward proportionality issues, with no need to discuss cruel and unusual punishment or shocking the moral sense. That, in fact, is how the first cases to discuss the holding in Cannon treated it. In State ex rel Gladden v. Kelly, 213 Or 197, 203, 324 P2d 486 (1958), the court construed Cannon as holding that “life imprisonment for attempted rape was excessive and not in proportion to the offense when the greater crime provided a lesser penalty.” In Merrill v. Gladden, 216 Or 460, 337 P2d 774 (1959), the court cited Cannon as holding that, in order for the plaintiff to show a violation of section 16, he had to show that assault with intent to rob was a lesser-included offense of robbery and that the penalty for it was disproportionate to the penalty for the completed offense. 216 Or at 462-63. It then held that,

“since our constitution provides that ‘all penalties must be proportioned to the offense,’ we conclude that the penalty assessable for the crime of assault with intent to rob cannot be greater than that provided for the accomplished robbery.” Id. at 464.

The Merrill court did not refer to the Sustar standard but rather resolved the case as one involving proportionality, not cruel and unusual punishment.9

*434In State v. Shumway, 291 Or 153, 630 P2d 796 (1981), the court also focused on the proportionality requirements of section 16.10 In Shumway, the defendant was convicted of murder and, under the applicable statute, received a sentence of life imprisonment with a 25-year minimum. On appeal, he argued that that sentence was disproportionate to the authorized sentence for aggravated murder, for which the minimum sentence would have been only 15 or 20 years. In its lead opinion, the Supreme Court agreed with that contention and deleted the minimum sentence, leaving the life term in place. It cited Cannon as holding that the life sentence for attempted rape was disproportionate to the offense and quoted the discussion of proportionality in Merrill. 291 Or at 164. It did not refer either to Sustar or to shocking the moral sense.

In State v. Turner, 296 Or 451, 676 P2d 873 (1984), the defendant received a minimum term for his burglary conviction and an additional minimum term as a dangerous offender. On appeal, he relied on Shumway in arguing that those terms were unconstitutional, asserting that, if there was no minimum term for murder, there could not be one for his less serious offenses. The Supreme Court first explained Cannon and Shumway as involving proportionality issues, quoting Merrill for the purpose. It then distinguished those cases on the ground that they involved comparisons between different degrees of the same crime. It pointed out that a life sentence, when considered without reference to any possible parole, is greater than the minimum terms that the defendant received.

“A sentence of imprisonment for the remainder of a person’s life for murder is greater than any maximum or minimum *435statutory term authorized by the Oregon legislature for any less severe crime. The minimum sentences imposed on this defendant under ORS 144.110(1) are not disproportionate.” 296 Or at 457.

In State v. Rogers, 313 Or 356, 379-80, 836 P2d 1308 (1992), the defendant argued that the death penalty was cruel and unusual in his case, which he described as involving “aggravated misdemeanor murder.” The court noted that that argument was essentially one of proportionality. It cited Sustar as holding that punishment is cruel and unusual when it is so disproportioned to the offense as to shock the moral sense of reasonable men and then held that the death penalty for the defendant’s offense did not meet that standard. The defendant relied on the Eighth Amendment as well as section 16, and both the defendant and the court treated his argument as involving the prohibition of cruel and unusual punishment, not the requirement of proportionality.

In State v. Isom, 313 Or 391, 837 P2d 491 (1992), the defendant received a death sentence after being convicted of committing a murder while on escape status from prison. The murder was not related to the escape, and the statute did not require such a relationship for receiving the death penalty. The defendant argued, relying on Cannon and Shumway, that the death penalty was therefore disproportionate to the offense. The Supreme Court noted that in both of the cases on which he relied a lesser-included offense carried a greater penalty than the principal offense. It then refused to perform its own evaluation of the relative seriousness of the offenses that the defendant described.

“Defendant’s view, that a person in the process of escaping is more dangerous than a defendant who already has escaped, is simply defendant’s opinion. The problem for defendant is that the legislature apparently had a different opinion, i.e., it believed that murders intentionally committed by escapees, whether during or after the escape, pose a considerable threat to the safety and peace of mind of the general populace. * * * Defendant’s opinion makes sense, but so does that which we attribute to the legislature. There was a rational basis for the legislature to conclude that both classes of escapees are dangerous.” 313 Or at 400 (emphasis in original).

*436The court then considered the defendant’s argument that the penalty was disproportionate because it did not apply to certain escapees. In response, the court cited Cannon as having adopted the shock the moral sense standard for disproportionality; it appears that the defendant also assumed that to be the correct standard. The court rejected the defendant’s argument, holding that one part was based on a misunderstanding of the law and that the other part involved a reasonable distinction between escapees from state and federal custody. Id. at 401-02. Thus, the court actually determined that the statute was not disproportionate in any respect.11

Our cases on section 16 reflect the Supreme Court’s. In Woolstrum v. Board of Parole, 89 Or App 600, 603-04, 750 P2d 509, rev den 305 Or 672 (1988), we cited the shock the moral sense standard in dictum in discussing a proportionality-based attack on the standards for setting a parole release date. Our holding was that the petitioner had not shown that any unconstitutionality in the standards affected him. In contrast, in State v. Spinney, 109 Or App 573, 577-78, 820 P2d 854 (1991), rev dismissed 313 Or 75 (1992), we upheld the sentencing guidelines against a proportionality challenge by emphasizing that the guidelines were designed to do precisely what section 16 requires: impose a sentence that is proportioned to the offense and the offender. In State v. Rice, 114 Or App 101, 106-07, 836 P2d 731, rev den 314 Or 574 (1992), the defendant argued that his sentence of seven consecutive 90-day jail terms for misdemeanors of negotiating bad checks was both cruel and unusual and disproportionate, because under the guidelines a conviction for the felony of first-degree theft would normally result in a probationary sentence. In affirming the sentences, we carefully distinguished between the defendant’s arguments concerning disproportionality and the arguments concerning *437cruel and unusual punishment, referring to the shock the moral sense standard only with respect to the latter.12

The holdings in our previous Measure 11 cases involving section 16 issues are consistent with Spinney and Rice. In State v. Lawler, 144 Or App 456, 469-70, 927 P2d 99 (1996), we cited Cannon and Shumway and noted that under section 16 a sentencing scheme is unconstitutionally disproportionate, among other reasons, if it authorizes a more serious sentence for a lesser-included offense. The defendant did not argue that Measure 11 was defective in that fashion as to the crime with which he was charged or of which he was convicted. In State v. George, 146 Or App 449, 456-57, 934 P2d 474 (1997), and State v. Rhodes, 149 Or App 118, 123, 941 P2d 1072 (1997), rev den 326 Or 390 (1998), we cited Cannon, Teague, and Isom for the shock the moral sense test and evaluated the sentences in isolation, not in relation to those authorized for other offenses that are equally or more serious. Neither defendant attempted to compare his sentence with those for other offenses.

Although the reasoning in the section 16 proportionality cases is not completely consistent, both the holdings of those cases and the weight of the opinions show a consistent theme: the requirement of proportionality is independent of the prohibition of cruel and unusual punishment. Isolated statements in some cases might suggest otherwise, but the cases as a whole make the point clear. The way in which later cases emphasize that Cannon is based on the requirement of proportionality may be the best example. On the other hand, aside from the obvious example of a harsher sentence for a lesser-included offense, the standards for determining when a sentence is disproportionate are unclear. It is still possible, however, to draw some conclusions about those standards.

*438The heart of the proportionality requirement is that there must be a correlation between the punishment and the relative seriousness of the offense.13 Sentences authorized for similar offenses in this and other jurisdictions are relevant to that determination. See Kelly and Sustar. The court may also consider its own evaluation of the seriousness of the offense and the harshness of the punishment, see Coffman and Ross, although, as the discussion in Isom shows, it hesitates to do so when the evaluation involves a value judgment that is essentially legislative in nature. Other than Ross, the only times that the court has expressly held a punishment to be disproportionate were when the statutes provided a harsher punishment for an attempt to commit an offense than for the completed offense itself. In such a situation, the court does not need to evaluate the relative seriousness of the offenses; it is inherent in their nature.

In some of its recent statements, the Supreme Court may have suggested that a sentence can be disproportionate only if the punishment for a lesser-included offense is harsher than the punishment for the greater offense. That also seems to be the view of the majority in this case. Such a conclusion, however, is directly at odds with a court’s duty under the constitution and is inconsistent with the Supreme Court’s first discussions of section 16. The court must evaluate the proportionality of any sentence when the defendant challenges it; the Cannon and Shumway situation is simply the most obvious kind of violation of that provision. The best interpretation of the court’s more recent statements is that they emphasize its normal hesitation to determine that one offense is more serious than another, because that is primarily a legislative determination.

*439Before the adoption of the sentencing guidelines, there was little legislative guidance in this area; as the court suggested in Isom, while a defendant might be able to make a plausible argument, there are usually also plausible arguments the other way, and when that is the case the choice between them is the legislature’s. There was no plausible argument, however, when the punishment for a lesser-included offense was greater than that for the greater offense. By definition, the lesser-included offense was less serious. There was nothing to guide the court in going any further. Although it recognized in its earlier cases that it could make such decisions independently, except for Ross it declined to do so.

Recent statutory and administrative changes have altered the context for proportionality analysis of felonies, because they now provide explicit legislative determinations of the relative seriousness of most felonies. The first step came with the adoption of the current criminal code in 1971. In that code, the legislature classified most felonies, with the maximum penalty for each offense based on its classification.14 See ORS 161.535, ORS 161.605 and ORS 161.625. However, under the code the trial court retained broad discretion in sentencing within the maximum limits, giving an appellate court little assistance in evaluating the constitutional proportionality of a specific sentence. Experience with that discretion led the legislature to be concerned that sentencing was inconsistent from case-to-case and from court-to-court, concerns that the Parole Board’s adoption of a matrix to determine when an offender should be released did not fully resolve. See Laird Kirkpatrick, Mandatory Felony Sentencing Guidelines: The Oregon Model, 25 UC Davis LR 695, 696-98 (1992). The result was the adoption of the sentencing guidelines. The relationship between those guidelines and Measure 11 is the issue that we should be deciding in this case.

*440The sentencing guidelines are administrative rules that the legislature has expressly approved, Or Laws 1989, ch 790, § 87, and that establish the framework for sentencing for almost all felony convictions; the legislature retains control over any amendments to the guidelines. ORS 137.667(4). In authorizing the development and adoption of the guidelines, the legislature instructed the appropriate bodies to take into consideration

“factors relevant to establishment of appropriate sentences, including severity of the offense, criminal history of the offender, aggravating and mitigating circumstances, performance under probationary supervision, prevention of recidivism, possibility of reformation or deterrence and the effective capacity of state and local correctional facilities and other sentencing sanctions available.” Or Laws 1987, ch 619, § 2(3); see also § 2(2)(b).

The guidelines, thus, must include a carefully considered determination of the relative seriousness of, and appropriate sanctions for, those felonies that they cover, which include those for which defendant was convicted.

The heart of the guidelines is the Sentencing Guidelines Grid, which consists of a “Crime Seriousness Scale” on one axis and a “Criminal History Scale” on the other. For each conviction, the seriousness of the offense and the nature of the defendant’s criminal history intersect at a grid block that establishes the presumptive sentence for that conviction. The crime seriousness levels range from 1 (least serious) to 11 (most serious). The criminal history columns range from A (most serious) to I (least serious). See State v. Davis, 315 Or 484, 486-87, 847 P2d 834 (1993). In determining the seriousness level of the various felonies, the Criminal Justice Council gave greater weight to attacks on personal safety, particularly crimes of violence, than to property offenses. It also created a number of subcategories for some felonies, so that the same offense might appear at two or more seriousness levels depending on the circumstances of its commission. Kirkpatrick, 25 Mandatory Felony Sentencing Guidelines at 702-05. The legislature, by approving the guidelines, has determined that the offenses at each crime seriousness level are equally serious. See OAR 213-04-0002(1). That express legislative determination of the relative seriousness *441of felonies now makes possible an analysis of the proportionality of sanctions.

The majority avoids the conflict between Measure 11 sentences and the sentencing guidelines by deciding that a Measure 11 sentence “trumps” a guidelines sentence whenever they conflict. As a matter of statutory construction, that approach is consistent with the Supreme Court’s resolution of the issue. See State ex rel Huddleston, 324 Or at 601-05. The problem is that the majority does not deal with the basic constitutional problem that the guidelines present. The crucial factor for section 16 proportionality analysis is not the ultimate sentence but the ranking of the crime on the crime seriousness scale. That is the place where the legislature determined the relative seriousness of felonies, and Measure 11 neither repeals nor modifies it. As a result, the effect of Measure 11 is to require sentences for some felonies that are far greater than the guidelines sentences for equally serious felonies. That is an effect of Measure 11 that the majority does not mention.

In deciding that a Measure 11 sentence prevails over a guidelines sentence under the applicable statutes, the Supreme Court held that a Measure 11 sentence constitutes a determinate sentence that, under ORS 137.637, will be the effective sentence when it is longer than a presumptive guidelines sentence. That conclusion says nothing about the relationship between a Measure 11 sentence and the legislature’s determination, in the crime seriousness scale, of the relative seriousness of felonies. Measure 11 simply superimposes a new set of sentences onto the guidelines sentences without regard to the effect of those new sentences on the existing sentencing structure. The majority’s suggestion that the mere enactment of Measure 11 somehow superseded the careful ranking of the relative seriousness of felonies is without any support in the statute itself or in any other source of which I am aware. Measure 11 focuses solely on those crimes that it covers, not on their relationship to other crimes that the statute does not cover.

The legislature’s actions since the adoption of Measure 11 confirm that that statute does not affect the continuing effectiveness of the crime seriousness scale. After Measure 11 became effective, the Oregon Criminal Justice *442Commission amended the crime seriousness scale by adding some crimes and by changing the position of others on the scale. Among the changes, it moved Assault in the Second Degree from category 8 to category 9 and Robbery in the Second Degree from category 7 to category 9. OAR ch 213, Appendix 2 (effective November 1,1997). Each of those felonies is a Measure 11 offense. The legislature thereafter expressly approved those changes, Or Laws 1997, ch 691, § 1, something that would have been a meaningless action if the crime seriousness scale were no longer relevant for Measure 11 crimes. I conclude that Measure 11 had no effect on the legislature’s determination of the relative seriousness of felonies, including Measure 11 felonies.

The guidelines, including the legislative determination of the relative seriousness of felonies, therefore remain in effect for Measure 11 crimes. However, Measure 11, rather than the guidelines, now generally sets the minimum sentences for those offenses, and those sentences are generally much longer than the presumptive guidelines sentences. The result is to raise a serious constitutional problem of proportionality under section 16. The guidelines contain a number of offenses that are at either the same or a higher level on the crime seriousness scale as some Measure 11 offenses. Thus, persons who commit those Measure 11 crimes now receive greater sentences than persons who commit other, non-Measure 11, crimes that the legislature has determined to be equally or more serious. The problem can be illustrated by examining defendant’s case.

Both of defendant’s convictions are for crimes that are crime seriousness level 8 under the guidelines. His criminal history score is H (no more than three adult A misdemeanors or two juvenile nonperson felonies). As a result, his presumptive guidelines sentence is 19 to 20 months on each conviction, with the option of probation if the court were to make certain findings. A guidelines sentence would also have the possibility of a good-time reduction based on his conduct in prison. In contrast, under Measure 11, the mandatory minimum sentence for each offense was 75 months, with no possibility of a good-time reduction.

*443At the time that defendant committed these crimes, the guidelines ranked a number of non-Measure 11 crimes as equally or more serious than his offenses. Those other crimes included: (1) arson in the first degree (level 10 when there was a threat of serious physical injury, level 9 if the damage was over $50,000, and level 8 if the damage was over $25,000)15; (2) solicitation of aggravated murder and attempted aggravated murder (level 1016); (3) burglary in the first degree (level 9 if the criminal was armed with a deadly weapon or if he or she caused or threatened physical injury to the victim, level 8 if the burglarized dwelling was occupied at the time of the crime); (4) delivery of drugs (level 8 in certain circumstances); (5) criminally negligent homicide (level 8); (6) inmate in possession of a firearm (level 8); (7) compelling or promoting prostitution (level 8)17; (8) a second conviction of stalking (level 8); and (9) use of a child in a display of sexually explicit conduct (level 8).

None of those non-Measure 11 offenses had a presumptive sentence that in defendant’s case would have been as severe as the Measure 11 sentence that he received, even ignoring the prohibition of work release or good-time credits. Indeed, no presumptive guidelines sentence reached the level of defendant’s Measure 11 sentence until crime seriousness 10, criminal history F (two or three nonperson felonies), where the presumptive sentence was 71 to 80 months. Thus, the punishment for defendant’s Measure 11 offense is comparable to the punishment for a person who committed a crime that was two steps more serious and who had a significantly greater criminal record than did defendant. Yet nothing in Measure 11 changes the determination in the guidelines of the relative seriousness of crimes in the crime seriousness scale.18

*444Section 16 requires that punishment be proportioned to the seriousness of the offense. In Cannon and Shumway, the Supreme Court enforced that requirement when statutes imposed a greater punishment for a lesser-included offense than for the greater offense. As a result of the sentencing guidelines, it is now possible to apply the reasoning behind those cases to a broader range of felonies. The courts are no longer faced with choosing between competing value judgments as to whether one offense is more serious than another; the legislature has made the choice. We now can — and, because we can, under section 16 we must— extend the existing proportionality analysis to defendant’s situation. His Measure 11 sentence is greater than the sentence that he would have received for other crimes that are equally or more serious than those that he committed. His punishment, thus, was not proportioned to his offense. Defendant’s sentence violates section 16, and we should remand for the court to resentence him under the guidelines.19 Because the majority does not do so, I dissent.