OPINION

COATS, Chief Judge.

Yako William "Billy" Collins was convicted of sexual assault in the first degree.1 Superior Court Judge Eric Smith sentenced Collins to twenty-five years of imprisonment with five years suspended. Collins appeals his conviction and sentence. We affirm Collins's conviction, but remand his case to the superior court to reconsider whether Collins's case should be referred to the three-judge panel for sentencing.

Factual and procedural background

The victim in this case, J.H., lived with her parents and her sister in Wasilla. Collins was the boyfriend of J.H.'s sister, and he also lived with the family. Collins and J.H.'s sister shared an apartment over the garage.

On the evening of March 12, 2008, the family was hosting a bonfire and barbeque. J.H. and her boyfriend, accompanied by J.H.'s sister and Collins, spent the evening drinking in the apartment over the garage. At some point, Collins and J.H.'s sister left the apartment, and J.H. had sex with her boyfriend. J.H. and her boyfriend then went to J.H.'s room inside the main house.

Sometime after midnight, either Collins or J.H.'s father asked the boyfriend to leave J.H.'s room, and he complied. J.H. then went to sleep. Sometime later, J.H. awoke to find Collins lying on top of her, having sexual intercourse with her. J.H. testified that she had not consented to have sex with him that night. J.H. testified that she tried to push Collins off her, and told him that what he was doing was wrong and that he needed to stop. Despite J.H.'s protests, Collins continued to have sex with J.H.

Collins stopped when he heard J.H.'s sister approaching the bedroom. He told J.H.'s sister that he and J.H. had just been talking. J.H. did not contradict Collins or immediately report the sexual assault because she wanted to "forget it ever happened." However, she told her boyfriend the next day that Collins had sexually assaulted her.

Later that day, J.H.'s father received a text message from J.H.'s sister, informing him that Collins had had sex with J.H. When J.H.'s father confronted her about this text message, she began to cry. She told her father that Collins had raped her. J.H. told her father that she had not said anything earlier because she was afraid no one would believe her.

J.H.'s father called the police. Following an investigation, a grand jury indicted Collins for first- and second-degree sexual assault.

Collins did not testify at his trial or present any other witnesses. He contended that the State's evidence supported a reasonable conclusion that J.H. had consensual sex with him, and that she later falsely claimed that she was raped because she was afraid of her father.

The jury convicted Collins of first-degree sexual assault (and acquitted him of second-degree sexual assault). Superior Court Judge Eric Smith sentenced Collins to twenty-five years with five years suspended — twenty years to serve.

Judge Smith did not commit plain error by limiting evidence related to J.H.'s sex with her boyfriend earlier in the evening of the assault

Collins argues that Judge Smith erred by limiting the evidence he could introduce on the subject of J.H.'s sexual intercourse with her boyfriend on the night of the sexual assault.

After the sexual assault was reported to the police, J.H. was examined by a sexual assault nurse examiner, Astin Mills. J.H. told Mills that she had been sexually assaulted. She also told Mills that she had consensual sex with her boyfriend earlier in the evening. Mills observed substantial injuries to J.H.'s genital area.

In a pre-trial motion, Collins sought to admit the evidence that J.H. had intercourse with her boyfriend on the evening of the assault, arguing that the evidence was relevant to provide an alternative explanation for J.H.'s injuries. Collins indicated that he might want to call the boyfriend as a witness. The State did not oppose admitting the evidence to provide an alternative explanation for J.H.'s injuries. But the State contended that the identity and age of J.H.'s boyfriend were not relevant.

Judge Smith ruled that Collins could introduce evidence that J.H. had sexual intercourse with her boyfriend through the testimony of J.H. He also ruled that if Collins established that the identity of J.H.'s boyfriend was relevant, or that he needed to introduce more on this subject than J.H.'s testimony, he could make an application to the court to introduce additional evidence.

During the trial, J.H. testified that earlier in the evening, before the sexual assault, she had sex with her boyfriend. In addition, the jury heard a recording in which J.H. told Nurse Mills she had sex that evening with her boyfriend "Nick."

On appeal, Collins argues that Judge Smith erred in not allowing him to introduce additional evidence on J.H.'s prior sexual activity with her boyfriend. He argues that his attorney told the court he wanted to call the boyfriend as a witness. But, as just explained, Collins's attorney only stated that he might want to call the boyfriend as a witness. Moreover, Judge Smith invited Collins to revisit the issue after J.H. testified if he wished to introduce more evidence on the subject, but Collins never asked for any further relief. Collins must therefore show plain error.

We conclude that Judge Smith did not abuse his discretion, much less commit plain error, by limiting the evidence Collins could introduce on the subject of J.H.'s sexual intercourse with her boyfriend. Collins never asked to call J.H.'s boyfriend, and he made no offer of proof establishing that the boyfriend's testimony would have benefitted his case. Nor does it appear that Collins preserved his objection to the court's decision to exclude evidence of the boyfriend's last name and age. In any event, this information does not appear to have been particularly probative.

Collins has not shown that Judge Smith abused his discretion by limiting evidence of J.H.'s father's alleged history of violence

Collins's defense at trial was that J.H. falsely claimed Collins raped her because she was afraid of how her father might react if she admitted she had consensual sex with Collins. Collins conceded that there was no evidence that J.H.'s father had ever been violent with J.H., but he sought to introduce evidence that J.H. was aware of her father's history of violence against others.

To resolve this issue, Judge Smith allowed the parties to question J.H. outside the presence of the jury. During this questioning, J.H. testified that her father had never been violent toward her. She testified that, although she had seen her father argue with his former girlfriends, she had never seen those arguments escalate into physical violence.

Based on this testimony, Judge Smith ruled that Collins could ask J.H. if she was afraid of her father. But he ruled that J.H.'s testimony that she had observed her father argue with former girlfriends was not relevant.

When the trial resumed, J.H. testified that she was not afraid of her father. She testified that she had not wanted her father to know about the rape because she simply did not want to deal with all the consequences of reporting a rape.

On appeal, Collins argues that Judge Smith unduly restricted his cross examination of J.H. about her father's history of violence, and that he should have been able to present "the full panoply of evidence" regarding the father's assaultive history. But Collins never offered any evidence that J.H.'s father had a history of violence, much less that J.H. was aware of this history. The only evidence on this issue was that J.H. had seen her father argue with former girlfriends. J.H. did not indicate that there was anything about this behavior that made her fearful of her father. We conclude that Judge Smith did not abuse his discretion by limiting this evidence.

Why we remand Collins's case for resentencing

In 2006, the legislature amended the sentencing statutes to sharply increase the sentencing ranges for sexual offenders. Under this revision, Collins, as a first time felony offender convicted of first-degree sexual assault, was subject to a sentencing range of twenty to thirty years. He was ultimately sentenced at the lower end of that range to twenty years to serve. If Collins had committed his offense before this change in the law, he would have faced a sentencing range of eight to twelve years.

Collins was twenty years old when he committed the offense. He had no juvenile record, and a limited adult record. At age nineteen, he was convicted of fourth-degree theft and fined fifty dollars. And while on bail release in this case, he was convicted of providing false information to the police, and sentenced to ninety days in jail. (Collins's bail condition required him to abstain from alcohol. He consumed alcohol while on bail release and, when contacted by the police, gave a false name because he knew he was in violation of his bail conditions.)

When an offense is subject to a presumptive range of imprisonment and no statutory mitigating factor has been proven, a sentencing judge has no discretion to impose a sentence below the low end of the presumptive range.2 In these circumstances, the defendant can only obtain a lesser sentence by persuading the judge that he is entitled to referral to the three-judge sentencing panel.3 Collins asked Judge Smith to send his case to the three-judge panel because he had "extraordinary potential for rehabilitation" and because manifest injustice would result from a sentence within the presumptive range. In support of this argument, Collins submitted letters vouching for his character, including strong letters of support from both the victim and her mother.

Judge Smith decided not to refer the case to the three-judge panel. He reasoned that, although "there are a lot of ... instinctual reasons for believing [Collins was] unlikely to commit this kind of crime again," he did not have sufficient information to reach this conclusion. He pointed out that Collins had not provided the court with a sex offender assessment or a substance abuse evaluation. He concluded that, in spite of Collins's youth and good reputation, and the letters of support he received from J.H. and others, his case was not sufficiently exceptional to justify referral to the three-judge panel.

Recently, in Luckart v. State,4 we considered the sentence appeal of another young first felony offender convicted of attempted sexual assault in the first degree, as well as several other crimes. In that case, the superior court had concluded that it would be manifestly unjust to sentence Luckart within the presumptive range of twenty-five to thirty-five years to serve given his youth, his lack of an adult criminal record, his cooperation with the police, and his apparent remorse.5 The three-judge panel agreed that a sentence within the presumptive range would be manifestly unjust.6 But the panel erroneously concluded, based on a misreading of the governing statute, that it had no authority to impose a sentence below the presumptive range because it had not found that Luckart had extraordinary potential for rehabilitation.7 We remanded the case to the three-judge panel, holding that once the panel ruled that a sentence within the presumptive range was manifestly unjust, it was required to impose a sentence below the presumptive term.8

Luckart, like Collins, was sentenced under the presumptive sentencing ranges enacted by the legislature in 2006. In his concurring opinion in Luckart, the present author recommended that the three-judge panel consider the legislative history of the 2006 revisions to the sentencing ranges in determining the appropriate sentence for Luckart pointing out that the legislature had based those sentencing ranges on certain assumptions about typical sex offenders that might not apply in Luckart's case.9

When we reviewed Judge Smith's decision in this case, we again had occasion to examine the legislative history of the presumptive sentencing ranges for sex offenders. That legislative history makes clear that the current sentencing ranges are based on legislative findings that the typical sex offender is a repeat offender with very poor prospects for rehabilitation.10 Because it was not obvious from the sentencing record that these legislative assumptions applied to Collins, we asked the parties and amicus curiae to submit supplemental briefing addressing whether sentencing judges should take account of these legislative assumptions in deciding whether referral to the three-judge panel is warranted in a particular sex offense prosecution. Having considered that briefing and argument, we conclude that the answer is yes.

The legislature addressed the purposes and rationale behind its revisions to the sentencing ranges for sex offenses in a detailed letter of intent.11 The legislature explained that it had increased the presumptive sentencing ranges for sex offenses because they are serious crimes that are prevalent in Alaska and have far-reaching negative impacts on victims, victims' families, and society.12 But the legislature also declared that it had increased the sentencing ranges based on its findings that sex offenders usually have committed multiple sex offenses by the time they are caught, that they often do not respond to rehabilitative treatment, and that they therefore cannot be safely released into society.13

The legislature also explicitly recognized that there would be cases in which a sentence within the presumptive range would be manifestly unjust.14 The legislature anticipated that sentencing judges would provide a necessary safety valve to prevent unfairness in particular cases by imposing a sentence below the presumptive range based on statutory mitigating factors or by referring the case to the three-judge panel.15

Given this legislative history, and given the legislative assumptions underlying the increased presumptive sentencing ranges enacted in 2006, we conclude that defendants convicted of sex offenses — particularly, young defendants with no significant criminal record — should be able to obtain referrals to the three-judge sentencing panel if they can show that these assumptions do not apply to them.

As this Court explained in Knight v. State,16 the presumptive sentencing range for any given class of criminal case represents the legislature's assessment of "the appropriate sentence for typical cases in that class," which we assumed to be "[the] relatively broad category into which most cases will fall."17 We explained that the aggravating and mitigating factors codified in AS 12.55.155(c)-(d) helped to identify the "relatively narrow circumstances that tend to make a given case atypical and place it outside the relatively broad presumptive middle ground."18

In other words, our prior decisions dealing with presumptive sentencing, and with referrals to the three-judge panel, are premised on an underlying assumption about the presumptive sentencing ranges codified in AS 12.55.125: the assumption that the legislature intended for these presumptive ranges to quantify, in the absence of evidence to the contrary, the sentences that should be imposed on defendants who had committed a typical offense within the statutory definition of the crime, and who had typical prospects for rehabilitation.

Based on the structure and content of the presumptive sentencing statutes, we can reasonably add one further assumption: that the defendant's criminal history is adequately reflected by the defendant's status as a first, second, or third felony offender.

The legislature has enacted several aggravating factors that apply to defendants whose criminal history is materially worse than one would normally suppose from the defendant's classification as a first, second, or third felony offender. There is a statutory aggravator for defendants with a prior history of assault,19 for defendants with a prior history of similar offenses,20 for defendants with a prior history of more serious offenses,21 and for defendants who have a greater number of prior offenses than is reflected in the defendant's categorization as a first, second, or third felony offender.22

These aggravating factors, if proved, allow the sentencing judge to exceed the presumptive sentencing ranges that would otherwise apply. It therefore follows that the legislature viewed these defendants as "atypical" in the sense that they should not presumptively receive a sentence within the normal sentencing range. Rather, when one or more of these aggravators are proved, a defendant's case no longer falls within "[the] relatively broad [middle] category" of defendants who are typically dangerous, and who have typical prospects for rehabilitation.23

But these assumptions about the presumptive sentencing ranges, and about which defendants should be subjected to these presumptive ranges, do not apply to the presumptive sentencing ranges established by the legislature for sex offenses.

As can be seen from the legislative history, the legislature enacted those presumptive sentencing ranges based on the assumption that defendants being sentenced for sex offenses have likely committed many other sex offenses before they were caught, and on the further assumption that the defendants being sentenced for sex offenses are particularly resistant to rehabilitative efforts.24 In other words, these presumptive ranges for sex offenders were designed for defendants who, by normal standards, are a typically dangerous, and who have atypically poor prospects for rehabilitation.

Accordingly, we conclude that a defendant's case should be referred to the three-judge sentencing panel, for consideration of sentences outside the presumptive range, if the defendant shows, by clear and convincing evidence, that the legislature's assumptions do not apply to him — either that the defendant does not have a history of unprosecuted sexual offenses, or that the defendant has prospects for rehabilitation which, in other offenders, would be considered "normal" (or "good"). If the defendant can prove either or both of these things, then the defendant's case should be referred to the three-judge panel because the legislature designed the presumptive ranges for sex offenders with significantly different characteristics. In such cases, the three-judge panel should decide whether, given all the circumstances of the case, it would be manifestly unjust to sentence the defendant to a term of imprisonment within the applicable presumptive range.

Now that we have clarified this area of the law, we must remand Collins's case to the superior court so that Judge Smith can reassess whether Collins's case should be referred to the three-judge panel.

We must address one further aspect of Collins's sentencing. At the sentencing hearing, Judge Smith rejected two statutory mitigating factors proposed by Collins — AS 12.55.155(d)(9) (conduct among the least serious within the definition of the offense) and (d)(12) (minor harm). We have examined the record, and we conclude that it supports Judge Smith's rulings on these two mitigators.

Conclusion

We AFFIRM Collins's conviction, but we VACATE the superior court's denial of his request for referral to the three-judge sentencing panel. The superior court must reevaluate its decision on this matter in conformity with this opinion.

(We do not retain jurisdiction of this case.)

BOLGER, Judge, dissenting.

By my reading, the lead opinion establishes two new mitigating factors based on the legislative history of the 2006 sentencing legislation — mitigating factors that apply only to sex offenses. In my opinion, these mitigating factors are not supported by this legislative history.

The first factor is based on the court's conclusion that the legislature believed that "the typical sexual offender is a repeat offender." This conclusion is apparently based on legislative findings regarding two studies reporting polygraph interviews with convicted sex offenders.1 Based on this conclusion, the lead opinion establishes a mitigating factor that applies to any sex offender who does not have a prior history of sex offenses.

I do not agree that these findings warrant a new mitigating factor. A sentencing judge generally does not have a polygraph interview to rely on. So the problem with requiring proof of unreported offenses is that the offenses cannot be documented: The legislature found that only sixteen percent of victims report a sexual assault2 and that only twenty-seven percent of reported sex crimes result in an arrest.3 These unreported and unprosecuted offenses will be unavailable to the sentencing judge; we should not make this routine circumstance the basis for a mitigating factor.

In addition, this mitigating factor would be inconsistent with the statutory context outlined in the lead opinion. The legislature did not change the aggravating factors that apply to these sex offenses. If the prosecution can prove that an offender is guilty of previously unreported sex offenses, then those offenses may support the aggravating factors for a prior history of aggravated assaultive behavior4 or a prior history of similar offenses.5 This court has approved the application of these aggravating factors to sexual assault convictions on numerous occasions.6

One could argue that the 2006 legislation impliedly repealed or amended these aggravating factors, but this approach would be unwarranted. We should not infer an amendment of the aggravating factor statute because the enforcement of this statute does not raise any actual conflict with the 2006 legislation.7 We should presume that the legislature was aware of our interpretation of this statute when it enacted the 2006 legislation.

So it appears to me that the presumptive sentencing ranges that were modified in the 2006 legislation remain subject to the aggravating factors that apply when the prosecution can prove a sex offender is guilty of previously unreported sex offenses. As noted in the lead opinion, these aggravating factors are intended to identify circumstances that tend to make a case atypically aggravated.8 If the prosecution cannot prove these aggravating factors, then the sentence properly belongs within the presumptive range; the absence of these factors does not make the case atypically mitigated.

The second mitigating factor in the lead opinion is based on the court's conclusion that the legislature found that most sex offenders have "very poor prospects for rehabilitation." This observation is apparently based on legislative findings regarding the ineffectiveness of institutional treatment programs.9 From these findings, the opinion reasons that it would be manifestly unjust to sentence a sex offender with normal prospects for rehabilitation to a sentence within the normal presumptive sentencing range. I disagree.

According to the legislative findings, a sex offender with moderate prospects for rehabilitation (and a correspondingly moderate risk of recidivism) may be unusual in the sense that he may have a better chance of rehabilitation than the usual sex offender. But it does not follow that it would be manifestly unjust to sentence a sex offender with a moderate risk of recidivism to a sentence within the presumptive range. The legislature recognized that sex offenses can have a serious impact on the victim and society.10 So a sex offender who poses a moderate risk of recidivism may still pose an unacceptable danger to the community. We should maintain the same standard for a sex offender that we have previously established for other defendants, requiring the defendant to show particularly favorable prospects for rehabilitation in order to establish a non-statutory mitigating factor.11

In the trial court, Collins argued that he had unusually favorable prospects for rehabilitation, but the sentencing judge found that Collins had not proven this mitigating factor. Collins did not submit a psychological evaluation or a sex offender assessment to establish how he could be supervised in the community. And the sentencing judge found that Collins's conduct on pretrial release was not good-he became intoxicated and then lied about his identity when he was confronted by the police. I conclude that the sentencing judge made a reasonable decision that Collins has not established any mitigating factors sufficient to warrant a referral to the three-judge panel.