OPINION

STEWART, Judge.

In 2004, the Alaska legislature made two significant changes to AS 28.35.030, the law prohibiting driving while under the influence. Before that time, to convict a motorist of driving while under the influence based solely on the result of a chemical test of the motorist's breath or blood, the jury had to find that the motorist's blood alcohol level was at or above the legal limit at the time of driving. The legislature in 2004 redefined the offense so that the defendant's blood alcohol level at the time of driving is no longer determinative — now, to find a motorist guilty of that offense, the jury need only find that a chemical test properly administered within four hours of driving showed a blood alcohol level at or above the legal limit at the time of the test.

Second, the legislature prohibited defendants in all prosecutions for driving while under the influence from offering evidence to show that they were less intoxicated at the time of driving than at the time of a later chemical test. (The legislature made one exception to this prohibition: the defendant may offer evidence to show that he consumed alcohol between the time of driving and the time of the chemical test.) That is, the legislature barred defendants from offering expert testimony or other evidence to show that the alcohol they had consumed before driving had not been fully absorbed into their bloodstream at the time of driving because, for instance, they ate a large meal right before drinking, or consumed the alcohol in "big gulps" shortly before they got in their car.

Douglas L. Valentine was convicted under this new law. He argues that his conviction should be overturned because the 2004 changes to the driving while under the influence law violate due process and equal protection, are vague and overbroad, create impermissible presumptions, violate the constitutional right to an independent test, and violate the supreme court's rule-making power. Having reviewed the record in Valentine's case and the legislative history of the 2004 amendments, we conclude that the statute survives all of Valentine's constitutional challenges.

Facts and proceedings

On June 18, 2005, at about 8:45 p.m., Fairbanks Police Sergeant Dan Welborn stopped Valentine for speeding. When he contacted Valentine, he noticed that he had a moderate odor of alcohol and that his eyes were watery and bloodshot. He administered three field sobriety tests, which Valentine failed, and then arrested him. At the station, Valentine submitted to a breath test, which showed a blood alcohol level of .099 percent. Valentine requested an independent test, which he obtained at 9:45 p.m. That blood test showed a blood alcohol level of .119 percent. The State charged Valentine under both theories in AS 28.35.030: it alleged that Valentine was "under the influence" at the time of driving under subsection (a)(1), and that a chemical test showed that his "blood alcohol" level was above the legal limit under subsection (a)(2).

Before trial, Valentine filed a motion to dismiss, challenging the constitutionality of the 2004 amendments to AS 28.35. District Court Judge Winston S. Burbank denied the motion to dismiss. In his decision, Judge Burbank incorporated by reference an earlier ruling by District Court Judge Raymond M. Funk rejecting the same constitutional claims in a consolidated Fairbanks case, State v. Baxley, Marshall Tyler. The effect of the district court's ruling was to prohibit Valentine from offering evidence to show that, even though his blood alcohol level was above the legal limit at the time of his two chemical tests, he was not guilty of driving while under the influence under either theory because, at the time he drove, the alcohol he had consumed had not yet been fully absorbed into his bloodstream.

Valentine's case then proceeded to trial before Judge Donald D. Hopwood. At trial, Valentine argued that he had only consumed two beers, that people absorb alcohol at different rates, that there was ample evidence that he was not impaired at the time he performed the field sobriety tests, and that the State had no evidence of his actual blood alcohol level at the time of driving.

The jury convicted Valentine in a general verdict (that is, a verdict that did not specify whether he was convicted under subsection (a)(1) or (a)(2)). Valentine now appeals the denial of his motion to dismiss. He supplemented the record on appeal with the expert testimony presented in Judge Funk's consolidated case, State v. Baxley. Background on the 2004 amendments to AS 28.35

In Conrad v. State, this court addressed the question of whether the pre-2004 version of AS 28.35.030(a)(2) required the State to prove that the defendant's blood alcohol level met or exceeded the legal limit at the time of driving, or whether it was enough for the State to show that a chemical test properly administered within four hours of driving showed a blood alcohol level that met or exceeded the legal limit.

Conrad submitted to a breath test about one hour and fifteen minutes after he was stopped by the police. That breath test showed a blood alcohol level of 154 percent. Thirty minutes later, an independent blood test showed that Conrad's blood alcohol level was .131 percent. At trial, Conrad presented what he called a "big gulp" defense: he claimed he had quickly consumed two beers right before he drove and that the alcohol from these beers had not been fully absorbed into his bloodstream at the time he was stopped. Therefore, he argued, he was below the legal limit at the time he was driving even though a breath test more than one hour later showed a blood alcohol level above the legal limit.

Although Conrad was able to argue his defense to the jury, the trial judge in effect instructed the jury that his defense was no defense — the judge told the jury that, under the version of AS 28.35.030(a)(2) in effect at that time, Conrad's guilt hinged on the result of his breath test, not on his alcohol level at the time he was driving. The jury convicted Conrad.

On appeal, Conrad argued that the court's instruction was error. We agreed, ruling that "a defendant's guilt under [former] AS 28.35.030(a)(2) "hinges on' the defendant's blood alcohol content at the time the defendant operated or controlled a motor vehicle." We observed that a number of states had defined the offense in terms of the defendant's test result — and we suggested that the Alaska legislature might have the authority to do so as well. But we concluded that the wording and legislative history of former AS 28.35.030(a)(2) "yield no such expression of legislative intent."

We elaborated on this ruling when we denied the State's petition for rehearing. The statute at issue in Conrad provided that a person committed the offense of driving while intoxicated if the person operated or controlled a motor vehicle "when, as determined by a chemical test taken within four hours after the alleged offense was committed, there [was] 0.08 percent or more by weight of alcohol in the person's blood." We found the statute's use of "when" ambiguous, and we resolved that ambiguity against the State:

The problem here is the ambiguous reference of the adverb "when". Does the statute mean that a person is guilty if they operate or control a motor vehicle "when . . . [their blood contains] 0.08 percent or more . . . alcohol [by weight]"? Or does the statute mean that a person is guilty if they operate or control a motor vehicle"[and later], as determined by a chemical test taken within four hours [of] the alleged offense . . ., [their blood contains] 0.08 percent or more . . . alcohol [by weight]"?

Several years later, in 2004, the legislature responded to the Conrad decision by amending AS 28.35.030. First, the legislature eliminated the adverb "when" from AS 28.35.030(a) to resolve the ambiguity identified in Conrad:

(a) A person commits the crime of driving while under the influence of an alcoholic beverage, inhalant, or controlled substance if the person operates or drives a motor vehicle or operates an aircraft or a water-craft

(1) while under the influence of an alcoholic beverage, intoxicating liquor, inhalant, or any controlled substance, singly or in combination; or

(2) and if [WHEN], as determined by a chemical test taken within four hours after the alleged operating or driving [OFFENSE WAS COMMITTED], there is 0.08 percent or more by weight of alcohol in the person's blood or 80 milligrams or more of alcohol per 100 milliliters of blood, or if [WHEN] there is 0.08 grams or more of alcohol per 210 liters of the person's breath[.]

Under the amended statute, a person commits the offense of driving while under the influence if (1) the person operates or drives a motor vehicle while under the influence or if (2) the person operates or drives a motor vehicle and "if, as determined by a chemical test taken within four hours after the alleged operating or driving, there is 0.08 percent or more by weight of alcohol in the person's blood or 80 milligrams or more of alcohol per 100 milliliters of blood, or if there is 0.08 grams or more of alcohol per 210 liters of the person's breath." Thus, under the plain language of the amended statute, the defendant's guilt under subsection (a)(2) now hinges on the chemical test result, not on the defendant's blood alcohol level at the time of driving. (Guilt under subsection (a)(1) still hinges on the motorist's impairment at the time of driving.) During House Judiciary Committee discussion of this amendment, representatives of the Department of Law emphasized that the amendment was designed to remove the ambiguity in the statute that we identified in Conrad.

At the same time, the legislature added a new subsection, AS 28.35.030(s), which prohibited defendants from raising a "big gulp" or other "delayed absorption" defense to a charge of driving while under the influence under subsection (a)(1) or subsection (a)(2):

(s) In a prosecution under (a) of this section, a person may introduce evidence on the amount of alcohol consumed before or after operating or driving the motor vehicle, aircraft, or watercraft to rebut or explain the results of a chemical test, but the consumption of alcohol before operating or driving may not be used as a defense that the chemical test did not measure the blood alcohol at the time of the operating or driving. Consumption of alcohol after operating or driving the motor vehicle, aircraft, or watercraft may be used to raise such a defense.

Deputy Attorney General Susan Parkes explained that this amendment was designed to eliminate the "big gulp" defense in driving while under the influence prosecutions:

[The amendment] addresses the "big gulp defense"[.] The current law says if a person is given [a breath] test within four hours of driving and [the result] is .08 [percent blood alcohol] or above, that person is legally intoxicated. However, the courts have allowed people to argue that they took a "big gulp" right before driving and although the [chemical test] was accurate at the time taken, the person's blood alcohol level was lower than that [at the time of driving]. [The Department of Law] does not believe the intent of the, legislation was to get into the middle of a battle of experts, so [the amendment] would do away with that defense and over turn a case called Conrad, issued by the Court of Appeals.

Subsection (s) was amended several times before the legislature arrived at the final language. Language was added to clarify that defendants may introduce evidence that they drank after driving. Language was also added to clarify that defendants may introduce evidence of how much alcohol they consumed before driving to rebut or explain the results of the chemical test — for instance, a defendant may offer evidence that he only had two drinks before driving, and that the chemical test therefore must have been inaccurate — but not to argue that the alcohol he consumed before driving had not been fully absorbed into his bloodstream at the time of driving.

During House Judiciary Committee discussion of the House version of the bill, Representative Max Gruenberg questioned whether judges, rather than legislators, should be deciding what evidence is relevant in a driving while under the influence prosecution. Representative Les Gara interjected that the committee was "changing the substance of the crime. . . . This isn't really a relevance issue; we're changing policy here." Representative Gruenberg then noted that the bill would make it a crime to have a blood alcohol level above the legal limit "even though a person might potentially have not been behind the wheel at that moment in time." Representatives Gara and Ralph Samuels agreed with that assessment, and Samuels added: "If it's in your stomach, it's as good as being in your blood, is what we're saying." Susan Parkes of the Department of Law echoed: "[W]hat's being prohibited is people [who have an illegal] level of alcohol in their bodies getting into a car and driving."

Finally, the legislature amended AS 28.35.033(c) to add the following underlined language:

Except as provided in AS 28.35.030(s), the provisions of (a) of this section [providing, inter alia, that a .08 percent blood alcohol level gives rise to a presumption that a person is "under the influence"] may not be construed to limit the introduction of any other competent evidence bearing upon the question of whether the person was or was not under the influence of intoxicating liquor.

The effect of this amendment to AS 28.35.033(c) was to prohibit defendants from offering "big gulp" or other "delayed absorption" evidence to rebut a chemical test result that is offered to show that the defendant was "under the influence" at the time of driving under AS 28.35.030(a)(1).

This legislative history, and the plain language of the amended statutes, indicate that the legislature intended to accomplish two things through this legislation. First, it intended to hinge a defendant's guilt in a driving while under the influence prosecution under subsection (a)(2) on the result of a chemical test properly administered within four hours of driving (if the test result is attributable to alcohol ingested before or during the operation of a motor vehicle). And, second, it intended to eliminate a "delayed absorption" defense to any prosecution under subsection (a)(1) or (a)(2) that relies in whole or in part on the result of a chemical test, so that defendants may not argue based on alcohol absorption rates that the result of their chemical test did not reflect their actual alcohol level at the time of driving.

The amended statutes are not void for vagueness

Valentine argues that the amended version of the driving while under the influence law is void for vagueness because guilt hinges on the result of a chemical test administered within four hours of driving, rather than on a motorist's blood alcohol level at the time of driving.

When no First Amendment rights are at stake, we consider two factors in determining if a statute is unconstitutionally vague. The first is whether the statute gives adequate notice to the ordinary citizen of what conduct is prohibited. A statute may not be "so vague that men of common intelligence must necessarily guess at its meaning and differ as to its application." The second consideration is whether the statute invites arbitrary enforcement because it gives too much discretion to prosecuting authorities. To establish this second factor, the defendant must show a history or strong likelihood of uneven application.

Valentine argues that the statute is impermissibly vague because it is difficult for a motorist to predict if, or when, within a four-hour period his alcohol level will reach or exceed the legal limit. He claims that the only way to prevent a driving while under the influence statute from being impermissibly vague is to rest guilt on the motorist's impairment at the time of driving.

To support this claim, Valentine cites the Pennsylvania Supreme Court's decision in Commonwealth v. Barud. The statute at issue in that case barred operating a motor vehicle "if the amount of alcohol by weight in the blood of the person is 0.10% or greater at the time of a chemical test of a sample . . . obtained within three hours after the person drove, operated, or was in actual physical control of the vehicle." The Pennsylvania court struck down the statute, concluding that it did not provide a reasonable standard to guide conduct because it was too difficult for a motorist to predict when, or if, he would reach a .10 percent blood alcohol level within a three-hour period.

At the outset, we note that Valentine's claim is not really a vagueness challenge. Valentine is not arguing that the language of the statute is unconstitutionally imprecise — nor could he persuasively, since the statute unambiguously prohibits a blood alcohol level of .08 percent or more as measured by a chemical test properly administered within four hours of driving. Valentine's real claim is that the statute violates due process because it is too difficult for a motorist who chooses to drink alcohol before driving to gauge if, or precisely when, his conduct will violate the statute.

Regardless of how Valentine's argument is characterized, other courts that have addressed this issue (with the exception of the Pennsylvania Supreme Court in Barud) have rejected this claim. As one court explained in upholding a statute that made it illegal to have a blood alcohol level of .10 percent as determined by a chemical test taken within two hours of driving:

While it certainly must be hard to predict if one's BAC has reached or will reach 0.10 percent within any time frame, this, does not mean that the two-hour rule provides no notice of the prohibited conduct. To the contrary, the two-hour rule, just as the former time-of-driving rule, gives fair notice that, although driving after drinking is not illegal per se, driving becomes illegal after a certain level of consumption; the fact that people cannot sense exactly what their BAC is at any given moment or will be in two hours does hot change this. As the Arizona State Supreme Court put it, "[w]here a statute gives fair notice of what is to be avoided or punished, it should not be declared void for vagueness simply because it may be difficult for the public to determine how far they can go before they are in actual violation."

The Georgia Supreme Court similarly held that regardless of whether a statute forbids a certain blood alcohol level at the time of driving or within three hours of driving, "a person who drinks a substantial amount of alcohol . . . is put on notice that he chooses to drive at his own peril."

Valentine argues that Alaska's statute is especially vague because the police have four hours from the time of driving — rather than the more typical two or three hours in states such as Pennsylvania and Georgia — in which to obtain a chemical test. Valentine has no standing to challenge the statute on this ground because, as he concedes, his breath and blood tests were administered within one hour of driving. But in any event, according to the expert testimony advanced by Valentine, most people reach their peak alcohol level within one hour after they stop drinking. Consequently, the fact that the police have four, instead of two or three, hours in which to obtain a sample would not make it appreciably more difficult for the typical motorist to evaluate whether they had consumed too much alcohol to legally drive.

We find the view adopted by the majority of courts persuasive and conclude that the amended driving while under the influence statutes give adequate notice of what conduct is prohibited. Valentine has not argued, much less shown, that there is a history or strong likelihood that the law will be unevenly applied. We therefore find no merit to his claim that the statute is impermissibly vague, and therefore violates due process. The amended statutes are not unconstitutionally overbroad

Valentine next argues that the amended driving while under the influence law is overbroad because it punishes people for innocent conduct. He argues that the law unconstitutionally criminalizes lawful conduct because a motorist can be convicted of driving while under the influence even if his blood alcohol level is below the legal limit at the time of driving.

This claim fails because, under the revised statutes, driving with a blood alcohol level below the statutory limit is not necessarily lawful conduct. As discussed earlier, the plain language and legislative history of the amended statutes indicate that it is now illegal to drive or operate a motor vehicle after ingesting enough alcohol to register a blood alcohol level of .08 percent or more at the time of a chemical test administered within four hours of driving. As Representative Samuels explained: "If it's in your stomach, it's as good as being in your blood." Or, as Representative Gara declared:

We know that you're not a technician and that you didn't time the amount you had to drink to get you home before you got drunk — you essentially just got lucky that you hadn't exceeded .08 at the time you got in your car, but you had been drinking irresponsibly. And I think even though you're not technically at a .08 percent [at the time of driving], we want to punish you for having that much to drink and getting in your car, because I think the truth is, even at .05 and .06, you're putting the public in danger.

To the extent that Valentine is arguing that the legislature has no authority to penalize a motorist whose blood alcohol level is lower than .08 percent at the time of driving, that claim is meritless. The legislature violates substantive due process by enacting a law that has "no reasonable relationship to a legitimate government purpose." The State has a legitimate interest in preventing driving while under the influence and it is the legislature's role, within constitutional limits, to determine the precise blood alcohol level or degree of impairment that makes operating a car unacceptably dangerous to the public. Valentine has not argued that the legislature's judgment is wrong, and that a motorist who drives or operates a motor vehicle after ingesting enough alcohol to register a blood alcohol level of .08 percent within the next four hours poses no risk to others. Indeed, the legislature previously recognized, in enacting AS 28.33.030(a)(2), the statute prohibiting operating a commercial motor vehicle with a blood alcohol level of .04 percent or more, that driving with a blood alcohol level of less than .08 percent might pose an unacceptable risk.

For these reasons, we reject Valentine's claim that the amended statutes are unconstitutionally overbroad because a defendant may be convicted of driving while under the influence even if his alcohol level at the time of driving is below the statutory limit.

The amended statutes do not impose criminal liability without mens rea

Valentine argues that the amended law unconstitutionally allows a person to be convicted without any proof of mens rea with respect to the circumstance of driving while impaired or with a blood alcohol level above the legal limit.

Valentine acknowledges that this court has repeatedly held that the offense of driving while under the influence does not require proof that the defendant was aware that he was legally impaired or that his blood alcohol level was above the legal limit — it is enough that the defendant knowingly drank and knowingly drove. As this court explained in Morgan v. Anchorage

It certainly does not make sense to allow a defendant to claim that his intentional consumption of alcohol impaired his ability to know that he was intoxicated. It does make sense to require a person who drinks and drives to be responsible for not drinking to the point where he is under the influence of alcohol. He should drive at his peril rather than only at the public's peril.

Valentine argues that this line of cases is at odds with our other decisions requiring a mens rea of at least criminal negligence with respect to the circumstance of driving with a revoked or suspended license. But as we explain below, these cases are not irreconcilable; they simply reflect the different proof requirements and policy interests underlying prosecutions for driving with a suspended or revoked license and driving while under the influence.

To convict a defendant of driving with a suspended or revoked license, the State does not have to show that the defendant was aware that his license was suspended or revoked; nor does the State have to show that the defendant understood the legal effect or scope of that revocation or Suspension. All the State must prove with respect to that circumstance is criminal negligence: that a reasonable person in the defendant's position would have been aware of a substantial and unjustifiable risk that his license was revoked or suspended, and that the defendant's failure to perceive that risk constituted a gross deviation from the standard of care that a reasonable person would exercise.

To convict a defendant of driving while under the influence, the State is not required to prove any mens rea with respect to the circumstance of having consumed too much alcohol to legally drive. But the State must prove that the defendant was legally intoxicated (either because his blood alcohol level met or exceeded the statutory limit within four hours of driving, or because he was under the influence at the time of driving) and that he knowingly drank and drove; it is implicit in' the proof of these elements that the defendant was criminally negligent as to the circumstance of being too impaired to legally drive. The conduct of consuming alcohol puts a person on notice that he may be impaired. Morgan and its progeny simply preclude a defendant from arguing that he did not have the required mens rea to commit the offense because, even though he knowingly drank and drove, he was too intoxicated to know he was impaired.

Valentine also argues that the line of cases following Morgan was implicitly overruled by the Alaska Supreme Court in State v. Hazelwood. In Hazelwood, the supreme court explained that, when a criminal statute proscribes conduct only when a relevant circumstance is present, a conviction under that statute must be predicated on a finding that the defendant was at least negligent with respect to that circumstance. For instance, a defendant cannot be convicted of transporting illegally taken game unless the defendant was at least negligent with respect to the fact that the game was illegally taken — because if the game was not illegally taken, transporting it would not be a crime. Valentine argues by analogy that a person cannot be. convicted of driving while under the influence unless the person was at least negligent with respect to the circumstance of being legally intoxicated — because drinking and driving is otherwise not sanctionable. This claim fails for the same reason discussed above: proof that the defendant was legally intoxicated and that the defendant knowingly consumed alcohol and drove is sufficient to establish that the defendant was negligent with respect to the circumstance that he was too impaired to legally drive. We therefore reject Valentine's claim that the amended law is unconstitutional because it imposes criminal liability without proof of mens rea. The amended statutes did not deny Valentine due process by creating impermissible presumptions

Valentine argues that the amended law creates an impermissible presumption that a motorist was at least as intoxicated at the time of driving as at the time of a chemical test administered within four hours of driving.

We have previously identified two possible defects in a presumption directed against the accused in a criminal case: "(1) the nexus between the fact proved and the fact to be presumed may be so insubstantial as to violate due process[;] and (2) the presumption may undermine the government's duty to prove guilt beyond a reasonable doubt."

Valentine attacks the amended driving while under the influence statutes on both these grounds. First, he argues that the presumption that a motorist was at least as intoxicated at the time of driving as at the time of a later chemical test is unconstitutional even if it is permissive and rebuttable because it has no basis in science. He argues that it is more likely, given alcohol elimination and absorption rates, that a motorist will be less intoxicated at the time of driving than at the time of a later chemical test. In advancing this claim, Valentine relies on expert testimony from State v. Baxley and the general rule that even a permissive presumption must be based on empirically valid inferences.

Second, Valentine argues that this presumption is unconstitutional because the defendant can no longer rebut it by offering "delayed absorption" evidence. He argues that, because the defendant cannot offer this rebuttal evidence, the presumption verges on a mandatory conclusive presumption — that is, a presumption that removes from the jury the element that the defendant was under the influence at the time of driving once the State proves that the defendant's chemical test result was at least .08 percent blood alcohol. He argues that this presumption violates Evidence Rule 303 by impermissibly shifting the burden of production (and perhaps the burden of persuasion) to the defendant.

As to Valentine's first claim, he has not shown that the presumption that a motorist was at least as intoxicated at the time of driving as at the time of a later chemical test lacks an empirical basis. Valentine advances expert testimony to show that it normally takes forty-five minutes to one hour for alcohol to be fully absorbed after a person stops drinking, and that it could take as long as four hours for someone who ate a large meal or had an upset stomach. That means some people will have a lower alcohol level at the time of driving than at the time of a later chemical test. But that is not enough to invalidate the presumption. We owe considerable deference to the legislative facts underlying a presumption. The inference that a person's alcohol level at the time of driving was at least as high as at the time of a later chemical test doubtless takes into account many factors beyond alcohol absorption rates — including the fact that many people who drink enough alcohol to become legally intoxicated do so over a period of time, not all at once right before driving a car.

Valentine's other claims must be addressed separately for prosecutions under subsections (a)(1) (the "under the influence" theory) and (a)(2) (the "blood alcohol" theory). With respect to prosecutions under subsection (a)(2), Valentine's claims fail for the reasons already discussed: the amended law does not create a presumption that the motorist was at least as intoxicated at the time of driving as at the time of the later chemical test. Under subsection (a)(2), it does not matter how intoxicated the motorist was at the time of driving. What matters is whether the motorist ingested enough alcohol before or while driving to have a blood alcohol level at or above .08 percent at the time of a chemical test administered within four hours of driving. The defendant may attack the accuracy of the chemical test, or show that alcohol was consumed after driving. What he cannot offer is evidence to show that the test did not measure his blood alcohol level at the time of driving because the alcohol he consumed before or while driving had not been fully absorbed — evidence that is no longer relevant under the amended statutes.

Under AS 28.35.030(a)(1), guilt still hinges on whether the motorist is "under the influence" at the time of driving. If the State charges the defendant under this subsection, and offers no evidence of a chemical test result, the defendant is still permitted to offer evidence to show that he was not under the influence at the time of driving. That is because subsection (s) only prohibits "delayed absorption" evidence that is offered to show "that the chemical test did not measure the blood alcohol at the time of operating or driving." Thus, for instance, if the State presented testimony that the defendant must have been under the influence because he drank six beers before driving, the defendant could offer testimony to show that he drank those beers immediately before driving and that the alcohol had not been sufficiently absorbed into his blood to impair his driving.

But if the defendant is charged under subsection (a)(1) and the State offers a chemical test result as evidence that the defendant was under the influence at the time of driving, the defendant may not offer "delayed absorption" evidence to argue that the chemical test result did not accurately reflect his impairment at the time of driving.

To understand the significance of this, one must look to AS 28.35.033(a), a related statute which the legislature left unchanged in 2004. That statute outlines certain inferences juries may draw from a defendant's chemical test result in deciding if the defendant was under the influence at the time of driving. Under the statute, the jury may infer from a chemical test result that shows a blood alcohol level of .04 percent or less that the defendant was not under the influence at the time of driving. If a test shows that the defendant's blood alcohol level was more than .04 percent but less than .08 percent, there is no presumption, but the jury may consider the chemical test result, together with other competent evidence, to determine if the defendant was under the influence at the time of driving. If a test shows that the defendant's alcohol level was .08 or more, the jury may infer from this fact alone that the defendant was under the influence of alcohol. We have previously explained that these presumptions apply to the defendant's impairment at the time of driving.

Consistent with this statute, the jurors in Valentine's case received the standard instructions informing them (1) that they "may" infer that he was under the influence if a chemical test showed that his blood alcohol level was at least .08 percent and (2) that they "may" infer that his chemical test result accurately reflected his alcohol level at the time of driving. We have previously upheld similar jury instructions based on the presumptions in AS 28.35.033(a) against claims that they impermissibly shifted the burden of proof to the defendant.

Valentine's jury was further instructed that it could consider "any . . . competent evidence" to determine if Valentine was under the influence at the time of driving. To underscore this point, Valentine in closing argument detailed all the evidence he claimed showed he was not impaired at the time of driving, and reminded the jury that it was free to disregard any inference based on his chemical test results that he was under the influence when he drove.

As the record in Valentine's case demonstrates, there is no merit to the claim that the amended statutes create a mandatory conclusive presumption that directs the jury to find that the defendant was under the influence at the time of driving based on a chemical test result showing a blood alcohol level of at least .08 percent. Rather, as was true before the 2004 revisions to the driving under the influence law, the jury may, but need not, infer from such a test result that the defendant was under the influence at the time of driving.

It is true that, under the 2004 amendments, a defendant may no longer offer "big gulp" or other "delayed absorption" evidence to show that his chemical test result did not reflect his impairment at the time of driving. But we conclude that the legislature had the authority to redefine the offense in this manner. The legislature wished to discourage motorists from "playing a form of Russian roulette with the driving public" by consuming large amounts of alcohol and then racing the clock to reach their destination before the alcohol was absorbed into their bloodstream and they became too impaired to drive. Prohibiting a "delayed absorption" defense was a reasonable way to deter this conduct. We cannot say there is no circumstance in which prohibiting this evidence would infringe the defendant's right to present a defense. For instance, a court would violate the defendant's due process rights if it allowed the State to offer evidence of alcohol elimination and absorption rates to show that the defendant was more intoxicated at the time of driving than at the time of a later chemical test, and then prohibited the defendant from countering the State's proof with a "delayed absorption" defense. But that is not the circumstance presented here. Legislative enactments are presumed to be constitutional, and Valentine has not met his burden to show a violation of due process in this case. The 2004, amendments do not infringe the rule-making power of the Alaska Supreme Court

Valentine also argues that the 2004 amendments unconstitutionally infringe the rule-making power of the Alaska Supreme Court. Specifically, he argues that the legislation conflicts with Alaska Evidence Rule 402 by preventing defendants charged with driving while under the influence from introducing evidence relevant to their alcohol level at the time of driving.

This claim is without merit. Evidence Rule 402 provides that "[a]ll relevant evidence is admissible, except as otherwise provided . . . by enactments of the Alaska Legislature[.]" Evidence Rule 402 thus expressly permits the legislature to define the parameters of relevant evidence, within constitutional limits. The language of AS 28.35.033(c) makes clear that the legislature intended to restrict the evidence that could be admitted in driving while under the influence prosecutions. There is thus no merit to Valentine's claim that the legislature infringed the supreme court's rule-making power.

The amended law does riot violate the constitutional right to an independent test

Valentine next argues that the 2004 amendments violate a motorist's constitutional right to an independent test because the defendant can no longer use the result of his independent test to support his claim that his alcohol level was rising at the time of the test, and was probably lower at the time of driving.

In Snyder v. State, the Alaska Supreme Court held that an accused has the right to an independent test for two reasons: (1) to give the defendant the opportunity to challenge the reliability of the State-administered breath test and, more broadly, (2) to safeguard the accused's right to obtain evanescent exculpatory evidence. Under the amended statute, these rights are still protected.

Valentine argues that the amended statute. violates due process by forcing the defendant into a "Hobson's choice" because any independent test that shows a blood alcohol level at or above .08 percent will necessarily be incriminating. But the right to an independent test has never guaranteed an exculpatelry test result.

Valentine also asserts that the new statute rewards post-driving drinking by permitting a "big gulp" defense in that circumstance, thus encouraging motorists to keep a flask of alcohol in the car in case they are pulled over. It is unclear how this argument relates to Valentine's claim that the statute violates the right to an independent test. To the extent that Valentine is arguing that this claimed discrepancy violates due process, that argument fails. Alaska Statute 28.35.030(s) permits a defendant to offer evidence that he drank after driving to establish that the chemical test did not measure his alcohol level (or impairment) at the time of driving. It does not, as Valentine claims, preserve a "delayed absorption" defense for post-driving drinkers while barring that defense for pre-driving drinkers. Under the statute, evidence that the motorist drank after driving would not be admitted to show that alcohol had not been fully absorbed into chemical test. the bloodstream at the time of driving, but rather to show that the motorist's test result was above the legal limit due to alcohol that was consumed after driving. The legislature has a strong interest in discouraging motorists from consuming too much alcohol before driving a motor vehicle; it does not have this same interest in penalizing a motorist for drinking after driving. Allowing motorists to offer evidence of post-driving drinking is thus reasonably related to the statute's purpose, The amended statutes do not violate equal protection

Lastly, Valentine argues that the amended statutes violate Alaska's guarantee of equal protection of the laws. He argues that the statutes are overinclusive because they "criminaliz[e] those with rising blood-alcohol levels" and underinclusive because they "omit those with declining blood-alcohol levels." In other words, he argues that there is an insufficiently tight fit between the goal of the amended statutes and the means chosen to accomplish that goal because some motorists who had a blood alcohol level above the legal limit at the time of driving will escape responsibility for that offense while other motorists who were below the legal limit when driving will be convicted, based on the happenstance of when the state administers its chemical test.

The amended driving while under the influence law does treat similarly situated people differently, in that guilt in some cases may depend on the arbitrary timing of the chemical test. The question is whether the means-to-ends fit is imprecise enough to violate equal protection given the nature of the interest at stake.

In evaluating whether legislation violates the equal protection clause of the Alaska Constitution, this court applies a flexible test that is dependent on the importance of the rights involved:

Initially, [the court] must look to the purpose of the statute, viewing the legislation as a whole, and the circumstances surrounding it. It must be determined that this purpose is legitimate, that it falls within the police power of the state. Examining the means used to accomplish the legislative objectives and the reasons advanced therefore, the court must then determine whether the means chosen substantially further the goals of the enactment. Finally, the state interest in the chosen means must be balanced against the nature of the constitutional right involved.

The State has a compelling interest in protecting the public from impaired drivers. And although the right to drive is an important interest, there is no constitutionally protected right to drink and drive. Therefore, the law will survive an equal protection challenge if it substantially furthers the goals of the legislature.

The general purpose of the driving while under the influence law is to deter and penalize driving while under the influence. Although arguably the most effective means of accomplishing that aim would be to convict only motorists whose alcohol level reaches a certain level while driving, such precision is impossible given the technology now in use and the inevitable delays involved in administering a DataMaster or other chemical test.

Valentine has offered no evidence or argument to refute the legislature's judgment that a person who drives after consuming enough alcohol to register a .08 blood alcohol level within four hours is a danger to the public. And there is a rational basis for that judgment. A motorist cannot necessarily predict how long it will take him to drive to his destination or, given the many variables influencing how fast alcohol is absorbed, how impaired he will become en route. The 2004 amendments substantially further the legislature's goal of deterring a motorist who has consumed enough alcohol to reach a blood alcohol level of .08 percent or more within the four hours after driving or operating a motor vehicle. Given that the only interest at stake is the interest in drinking and driving, that is enough for the amended statute to survive constitutional scrutiny. While it is true that some who commit the offense will escape detection, and others will not, that is true of all criminal offenses.

Valentine also claims that the statute violates equal protection by hinging guilt for a pre-driving drinker on his alcohol level at the time of the chemical test while hinging guilt for a post-driving drinker on his alcohol level at the time of driving. But the statute does not treat similarly situated persons differently. All people charged with driving while under the influence may offer evidence to show that their alcohol level was above the legal limit at the time of a chemical test because of alcohol they consumed after driving. As discussed earlier, the legislature has little interest in penalizing motorists for drinking after driving.

For these reasons, we reject Valentine's equal protection challenge.

Conclusion

Valentine's conviction is AFFIRMED.