concurring.
I am writing to explain in more detail what this litigation is about and why I join in the court’s decision.
In 1990, Callan was convicted of theft and sentenced to serve 3 years in prison. At the end of 2 years in prison, Callan had accumulated 1 year of good time credit under AS 33.20.010 — an amount equal to the time remaining on his sentence. He was therefore released on parole. AS 33.20.030-040.
While on parole, Callan committed burglary, and he was sentenced to 3 years’ imprisonment for this crime. Based on the burglary, the Parole Board revoked Callan’s' 1 year of good time credit from the earlier theft conviction. Thus, Callan was sent to prison to serve the 1 remaining year from his theft sentence and to serve a consecutive 3-year sentence for his burglary.
Approximately one year later, Callan filed an application for post-conviction relief. The relief Callan sought was in the nature of a declaratory judgement: an order directing the Department of Corrections to calculate his sentence and his parole date in a particular way.
The Department of Corrections calculated that Callan would spend 2 years and 8 months in prison, and then he would spend 1 year and 4 months on mandatory parole. The Department’s calculation was based on the working assumption that Callan’s two sentences should be treated as one composite term of 4 years’ imprisonment. Because a prisoner accrues good time credit at the rate of 1 day for every 2 days of good behavior in prison, Callan would be entitled to release on mandatory parole after serving two-thirds of this 4-year composite term — or approximately 2 years, 8 months — and then Callan would spend the remaining 1 year, 4 months on mandatory parole.1
Callan argued that the Department’s calculation was mistaken. He asserted that, because he had been sentenced to two separate, consecutive sentences, his time in prison should be credited first to one sentence, then to the other. Callan argued that he should be deemed to first be serving his 1 year of revoked good time credit from his theft conviction, then- (at the end of that 1-year sentence) to be serving the 3-year sentence he received for the burglary. According to Cal-lan’s calculation, his first 8 months in jail satisfied the 1-year theft sentence (that is, he served 8 months and he earned 4 months of good time credit). At the end of those 8 months, Callan argued, he was entitled to immediate release on parole from the 1-year theft sentence. Callan further asserted that the Department of Corrections had no authority to defer this theft parole until the end of his burglary sentence. Thus, Callan argued, he must be deemed to have simultaneously served his 4 months of mandatory parole from the theft sentence at the same time he was serving the first 4 months of his burglary sentence.
The practical effect of this difference between the Department of Corrections’ calculation and Callan’s calculation lies in the date of Callan’s ultimate release from state supervision. Under the Department’s calculation, Callan would be released from state supervision 4 years after he entered prison. That is, Callan would serve 2 years, 8 months in prison and then spend 1 year, 4 months on mandatory parole. Under Callan’s ealcula*859tion, he would be released from mandatory parole 4 months earlier. According to Cal-lan, his 4-month mandatory parole from the theft sentence was served during months 9 through 12 of his imprisonment (that is, simultaneously with the first 4 months of Cal-lan’s burglary sentence), so that he would have only 1 year of mandatory parole remaining (all attributable to his burglary sentence). when he was released from prison.
Superior Court Judge Mark C. Rowland decided this issue in favor of the State. Since the issue Callan raises is purely one of law, this court exercises its independent judgment when reviewing the superior court’s interpretation of the statutes. Glidden v. State, 842 P.2d 604, 606 (Alaska App. 1992).
Alaska adopted its good time and mandatory parole statutes from former federal legislation. Morton v. Hammond, 604 P.2d 1 (Alaska 1979). The federal cases interpreting these federal statutes — former 18 U.S.C. §§ 4161-66 — support the Department of Corrections’ method of computation. That is, in analogous circumstances, federal courts have ruled that a prisoner’s various consecutive sentences should be treated as one composite sentence when calculating the prisoner’s good time credit and parole release date. See McCray v. United States Board of Parole, 542 F.2d 558, 560 (10th Cir.1976); Briest v. United States Bureau of Prisons, 459 F.2d 284 (8th Cir.1972).
Under the former federal statutes, a prisoner’s rate of good time accrual depended- on the length of the prisoner’s sentence (prisoners with longer sentences earned good time credit at a faster rate). See State v. McCallion, 875 P.2d 93, 95 n. 3 (Alaska App.1994). It is therefore possible to explain these federal decisions as the product of the rule of lenity — the rule that ambiguous penal statutes should be interpreted in favor of defendants. However, in addition to being supported by federal precedent, the State’s method of calculation appears more logical and fair. Under Callan’s proposed calculation, a prisoner’s final release date would be greatly, and arbitrarily, affected by the choice of which sentence he or she was deemed to be serving first.
Suppose a defendant was simultaneously sentenced to consecutive terms of 6 years and 3 years. Assuming that the defendant received a one-third reduction for good time credit against each sentence, the 6-year sentence from crime A would entail 4 years of actual imprisonment and 2 years of mandatory parole, while the 3-year sentence for crime B would entail 2 years of actual imprisonment and 1 year of mandatory parole.
Under the State’s calculation, this defendant would serve 6 years in jail and then 3 years on parole, for a total of 9 years under state supervision. Using Callan’s method of calculation, the defendant’s total time under state supervision would vary, depending upon which sentence was considered “served” first.
If the prisoner were deemed to be serving the 3-year sentence for crime B first, then the prisoner’s year of mandatory parole for that crime would be served simultaneously with his first year of imprisonment for crime A. Thus, the prisoner would spend a total of 6 years in jail (during which time he would satisfy his year of parole on crime B), and then he would spend 2 years on parole from crime A — a total of 8 years under state supervision. But if the prisoner were deemed to be serving the 6-year sentence for crime A first, then his 2 years of mandatory parole for that crime would be served simultaneously with his 2 years of actual imprisonment for crime B. Under this scenario, the prisoner would spend only 7 years under state supervision — 6 years in jail and then only 1 year on parole from crime B.
The fact that Callan’s method of calculation leads to such arbitrary and unfair results is a strong argument against it. See Endell v. Johnson, 738 P.2d 769 (Alaska App.1987) (rejecting a proposed interpretation of the statute granting credit for time served because the interpretation led to arbitrary and unfair differences in the amount of time defendants spent in prison). Moreover, it appears strained (at best) to suggest that AS 33.20 should be construed to place a prisoner *860on “parole” at the same time that he or she is actually imprisoned and serving another sentence.
For these reasons, the superior court rightly rejected Callan’s proposed method of calculation and his assertion that he was simultaneously paroled and imprisoned,
. It appears that, following his re-incarceration, Callan lost 120 days of this good time for infractions of prison rules. However, for purposes of deciding the issue in this appeal, this additional loss of good time is irrelevant.