Shabazz v. Gabry

OPINION GILMORE, District Judge. This is a class action suit questioning whether the 1992 amendments to Michigan’s parole laws governing frequency of parole review hearings violate the Ex Post Facto Clause of the United States Constitution as retroactively applied to various plaintiff subclasses. The case comes before the court on cross-motions for summary judgment. Plaintiffs are inmates in the custody of the Michigan Department of Corrections who committed their crimes on or before September 22, 1992. Defendants are Gary Gabry, Chairman of the Michigan Parole Board, and other members of the Parole Board. For the reasons discussed below, the court GRANTS in part and DENIES in part plaintiffs’ motion for summary judgment, and GRANTS in part and DENIES in part defendants’ motion for summary judgment, holding that Mich.Comp.Laws §§ 791.234(6) and 791.244(1) (1992) are ex post facto laws in violation of the United States Constitution, Article I, Section 10, Clause 1, as applied to all plaintiffs who committed their crimes and were convicted between 1977 and 1982 receiving parolable life sentences or long, indeterminate sentences, and all plaintiffs who committed their crimes and were convicted after 1982 receiving mandatory life sentences, parolable life sentences or long, indeterminate sentences. The court declares these laws invalid as applied to these plaintiffs and orders defendant to reinstate a parole hearing schedule equivalent to that in place at the time these plaintiffs committed their crimes. 1 The court also holds that the 1992 amendments do not violate the Ex Post Facto Clause as applied to the remaining members of the plaintiff class, including all inmates who committed their crimes and were convicted before 1977 receiving mandatory life, parolable life or long, indeterminate sentences, and all inmates who committed their *121 crimes and were convicted between 1977 and 1982 receiving mandatory life sentences. I. BACKGROUND Plaintiffs are inmates in the custody of the Michigan Department of Corrections (“MDOC”) who committed their crimes on or before September 22, 1992. They filed this action under 42 U.S.C. § 1983 alleging violations of their constitutional rights. Specifically, plaintiffs challenge the constitutionality of the 1992 legislative amendments to Michigan’s parole laws, which reduced the frequency of parole hearings for all inmates serving long, indeterminate sentences (LIDs), parolable life Sentences, and mandatory life sentences. Plaintiffs assert that the new parole statutes, as applied to them, violate the Ex Post Facto Clause of the Constitution, Article I, § 10. In September of 1994, this court adopted a Report and Recommendation filed by the magistrate judge certifying a plaintiff class. For purposes of this opinion, the court has separated plaintiffs into the following three (3) subclasses: 2 1. Those inmates who committed their crimes and were convicted between 1982-1992 receiving mandatory life sentences, parolable life sentences or long, indeterminate sentences. 2. Those inmates who committed their crimes and were convicted between 1977-1982 receiving parolable life or long, indeterminate sentences. 3. Those inmates who committed their crimes and were convicted between 1977-1982 receiving mandatory life sentences, and those inmates who committed their crimes and were convicted prior to 1977 receiving mandatory life, parolable life or long, indeterminate sentences. Plaintiffs in each subclass assert that the retroactive application of 1992 amendments to Michigan’s parole statutes violated the Ex Post Facto Clause of the Constitution by depriving them of the right to a certain number of hearings before the parole board to determine their eligibility for early release. To determine whether a new law violates the Ex Post Facto Clause, the court must look at the relevant law in effect at the time an offense was committed. See Weaver v. Graham, 450 U.S. 24, 29, 101 S.Ct. 960, 964, 67 L.Ed.2d 17 (1981); Kellogg v. Shoemaker, 46 F.3d 503, 509 (6th Cir.1995). Hence, this court will begin by examining the laws governing frequency of parole hearings at the time members of the plaintiff subclasses committed their crimes. A. FREQUENCY OF REVIEW FOR SUBCLASS 1 Subclass 1 consists of those inmates who committed their crimes’ and were convicted between 1982-1992 receiving mandatory life sentences, parolable life sentences or long, indeterminate sentences (“LIDs”). Between 1982 and 1992, the frequency of parole board review for these inmates was governed by M.C.L. §§ 791.234 and 791.244. Pursuant to M.C.L. § 791.234, parolable lifers/LIDs were to be interviewed after four years, and then every two years thereafter (4 + 2 + 2, etc.), and pursuant to M.C.L. § 791.244, mandatory lifers were to be interviewed on the same 4+2 + 2, etc. schedule. The 1982 statutes had no effect on the eligibility requirements for early release of mandatory lifers, and had no effect on M.C.L. § 791.234, which provided that lifers/LIDs were not eligible for parole until they had served 10 years of their sentence. Nor did the 1982 statutes change M.C.L. § 791.233(l)(b), which provided that certain LIDs were eligible for “special parole,” allowing for release prior to their tenth year of incarceration. The 1982 statutes affected only the frequency of scheduled hearings before the parole board. B. FREQUENCY OF REVIEW FOR SUBCLASS 2 Subclass 2 consists of those inmates who committed their crimes and were con *122 victed between 1977-1982 receiving parolable life sentences or LIDs. Between 1977 and 1982, parole board jurisdiction and parole eligibility were governed by statute, but there was no statute governing the frequency of parole hearings. Instead, during this period parole board review was governed by administrative regulation, Michigan Administrative Code (MAC) Rule 791.7710 (“Administrative Rule 710”), which was enacted and codified in 1977 pursuant to the Administrative Procedures Act (“APA”). This Rule provides: If release is denied, the board shall furnish the resident written notice: (a) Stating with particularity the grounds on which its determination is based. (b) Including, whenever possible, a statement of what the resident should do to better qualify for parole. (e) Except for those residents who are within 18 months of sentence expiration, setting a new hearing date, to be no more than 12 months from the minimum eligibility date or previous pass-over date. MAC r. 791.7710 (effective 11/30/77). By its own terms, this rule does not apply to inmates who are not eligible for parole. 3 However, defendants have argued that this rule was also not intended to apply to the parola-ble lifers/LIDs in subclass 2, claiming that the rule applied only to LIDs for whom parole eligibility was based on service of a minimum sentence. The court rejects this argument and holds that the plain language of Rule 710 does not limit its application to inmates for whom parole eligibility is based on service of a minimum sentence. Moreover, the court notes that in codifying certain other MAC Rules, the MDOC chose to expressly exclude parolable lifers and LIDs. See e.g. MAC r. 791.7701. Hence, if the MDOC had intended to limit the application of Rule 710, it could have done so expressly. Thus, the court concludes that Rule 710 applied to members of subclass 2 and required reinterview within 12 months of the last “pass-over” date if parole was denied. 4 Accordingly, between 1977 and 1982, the members of subclass 2 should have been interviewed after seven years, again at three years and then every year thereafter (7 + 3 + 1 + 1, etc.). 5 C. FREQUENCY OF REVIEW FOR SUBCLASS 3 Subclass 3 consists of those inmates who committed their crimes and were convicted between 1977-1982 receiving mandatory life sentences, and those inmates who committed their crimes and were convicted prior to 1977 receiving mandatory life, parolable life or LIDs. When the members of this subclass were convicted and sentenced, there was no applicable statute or codified administrative regulation governing the frequency of parole board review. Instead, the frequency of parole review for these inmates was governed by the internal policies and procedures of the parole board itself and by policy directives established by the MDOC. Pursuant to these procedures and directives, the practice of the parole board was to interview parola-ble lifers/LIDs after seven years of incarceration, then every three years thereafter (7 + 3 + 3, etc.), and mandatory lifers after ten years of incarceration, and every three years thereafter (10 +3 + 3., etc.). D. THE 1992 LEGISLATIVE CHANGES In 1992, the Michigan legislature amended M.C.L. §§ 791.234 and 791.244 to provide that all mandatory lifers, parolable lifers and LIDs were to receive their first parole interview after ten years, and then only every five years thereafter (10 + 5 + 5, etc.). The *123 amendments further provided that the new schedule would apply to all inmates regardless of whether they committed their crimes before or after the effective date of the act on September 22, 1992. 6 The record in this case shows that under the 1992 amendments, the new dates for initial interviews and reinterviews were determined by computer-generated lists. The new schedule was implemented without regard to the nature of an inmate’s crime or his likelihood of parole. Thus, by way of example, a parolable life/LID who had been interviewed in years 4, 6, and 8 would have his next interview automatically “flopped” from year 10 to year 13 regardless of whether he received favorable recommendations in years 4, 6 and 8 and was considered a likely candidate for release on parole in the near future. II. This court must determine whether the 1992 amendments to Michigan’s parole laws, which postponed regularly scheduled parole reviews for all mandatory lifers, parolable lifers and LIDs from the fourth year and every two years thereafter to the tenth year and every five years thereafter, violate the Ex Post Facto Clause of the United States Constitution as applied to any of the three plaintiff subclasses. The court will begin by discussing some of the relevant law governing this complex issue. A. Article I, § 10 of the Constitution forbids the states from passing any “ex post facto law.” In Collins v. Youngblood, 497 U.S. 37, 110 S.Ct. 2715, 111 L.Ed.2d 30 (1990), the Supreme Court identified three primary legislative acts which implicate the core concerns of the Ex Post Facto Clause: 1. Those that punish as a crime an act previously committed that was innocent when done; 2. Those that aggravate a crime or make it greater than it was when committed; and 3.Those that make more burdensome the punishment for a crime after is commission. Id. at 52, 110 S.Ct. at 2724. It is the third category that we are concerned with in this case. In accordance with this original understanding, Collins noted with approval Justice Chase’s opinion in Colder v. Bull, 3 U.S. (3 Dall.) 386, 1 L.Ed. 648 (1798), which explained that a law falls within the prohibition on ex post facto laws if it “changes the punishment, and inflicts a greater punishment, than the law annexed to the crime when committed.” Id., 497 U.S. at 42, 110 S.Ct. at 2719. In a case construing federal parole law, the Sixth Circuit has held that “parole consideration is part of the law annexed to the crime.” Ruip v. United States, 555 F.2d 1331, 1335 (6th Cir.1977). Hence, the issue before the court is whether the statutory change in scheduled parole hearings inflicts a greater punishment on any members of the plaintiff class than the parole laws annexed to plaintiffs’ crimes when committed. B. THE MORALES DECISION In the recent case of California Dept. of Corrections v. Morales, — U.S. -, 115 S.Ct. 1597, 131 L.Ed.2d 588 (1995), the Supreme Court considered whether amendments to a California statute governing frequency of parole board review violated the Ex Post Facto Clause. Petitioner Morales had received a sentence of 15 years to life for the 1980 murder of his wife. As required by California law, Morales received a parole hearing in 1989, at which time he was found unsuitable for parole for numerous reasons including the particularly heinous nature of his crime, and the fact that he had committed the murder while on parole for an earlier murder. According to the law in effect at the time Morales committed the second murder, he would have been entitled to subsequent parole suitability hearings on an annual basis. However, in 1981, the law was amended to allow the parole board to defer subsequent *124 hearings for up to three years for a prisoner convicted of more than one offense involving the taking-of a life if the board (1) determined that it was not reasonable to expect that parole would be granted at a hearing during the intervening years and (2) articulated the basis for its finding. In Mr. Morales’ case, the Board made the requisite findings and scheduled his rehearing for 1992 rather than 1990. Morales filed a habeas corpus petition in which he asserted that the new law, as applied to him, violated the Ex Post Facto Clause. The district court denied the petition, but the Ninth Circuit reversed, holding that the amendment to California’s parole statute constituted an ex post facto law. In reaching this conclusion, the Ninth Circuit relied heavily on the fact that “a prisoner cannot be paroled without first having a parole hearing.” Morales v. Calif. Department of Corrections, 16 F.3d 1001, 1004 (9th Cir.1994). In other words, the court regarded an inmate as ineligible for parole in the period between interviews. The Supreme Court reversed, finding that the challenged statute did not violate the Ex Post Facto Clause. However, the Court declined to announce a bright-line test for determining when legislative changes affecting an inmate’s opportunities for parole constitute ex post facto laws, commenting: [W]e have long held that the question of what legislative adjustments “will be held to be of sufficient moment to transgress the constitutional prohibition must be a matter of degree.” (citations omitted) ... We have previously declined to articulate a single “formula” for identifying those legislative changes that have a sufficient effect on substantive crimes or punishments to fall within the constitutional prohibition ... and we have no occasion to do so here. Id., 115 S.Ct. at 1603. The closest that the Court came to providing a rule governing the ex post facto challenge before this court is as follows: In evaluating the constitutionality of the 1981 amendment [to California’s parole law], we must determine whether it produces a sufficient risk of increasing the measure of punishment attached to the covered crimes. Id. (emphasis added). Applying this “sufficient risk” test to the facts of the Morales case, the Court determined that the amendment “create[d] only the most speculative and attenuated possibility of producing the prohibited effect of increasing the measure of punishment for covered crimes, and such conjectural effects are insufficient under any threshold we might establish under the Ex Post Facto Clause.” Id. Critical to the Court’s decision in Morales were the following considerations: 1. The California amendment applied only to double murders — a class of prisoners for whom the likelihood of release on parole is quite remote. Id., 115 S.Ct. at 1599. 2. The Parole Board’s authority under the amendment was carefully tailored to its goal of eliminating the “costly and time-consuming responsibility of scheduling parole hearings for prisoners who have no chance of being released.” Id., 115 S.Ct. at 1602. Thus, the amendment affected the timing of subsequent hearings only if the Board first concluded, after a hearing, “not only that the prisoner is unsuitable for parole, but also that it is not reasonable to expect that parole would be granted at a hearing during the following years.” Id., 115 S.Ct. at 1604. 3. The Parole Board was not bound to wait three years to schedule a reinterview, but retained the discretion to tailor the frequency of subsequent rehearings to the facts of the particular case. Id. 4. In the case of a prisoner experiencing a drastic change of circumstances sufficient to alter his suitability for release on parole, there was no indication that such a prisoner would be precluded from seeking an expedited hearing from the Board. Id. Focusing on these factors, the Supreme Court concluded that California’s amendment was not an ex post facto law. III. With this background in mind, the court will examine the ex post facto challenge presented by each subclass of plaintiffs. *125 A. EX POST FACTO CLAIM OF SUBCLASS 1 Subclass 1 includes inmates who committed their crimes and were convicted between 1982-1992 receiving sentences of mandatory life, parolable life, or LIDs. The record in this case establishes that at the time members of this subclass were convicted, the frequency of parole hearings was governed by statute. Pursuant to M.C.L. § 791.234, parolable lifers/LIDs were to be interviewed on a 4 + 2+2 (etc.) schedule, and pursuant to M.C.L. § 791.244, mandatory lifers were to be interviewed on a 4 + 2 + 2 (etc.) schedule. Thus, the only question before the court is whéther the 1992 amendments, as' applied to these plaintiffs, violate the Ex Post Facto Clause of the Constitution. For the reasons discussed below, the court concludes that the 1992 amendments to the 1982 statutes governing the frequency of scheduled parole hearings are ex post facto laws as applied to the members of subclass 1. In 1992, the relevant statutes were amended and the frequency of parole review was changed significantly. The relevant provisions of the 1992 amendments read as follows: M.C.L. 791.234(6) (1992) A prisoner under sentence for life or for a term of years, other than a prisoner sentenced for life for murder in the first degree or sentenced for life or for a minimum term of imprisonment for a major controlled substance offense, who has served 10 calendar years of the sentence in the ease of a prisoner sentenced for a crime committed before [the effective date of this act], is subject to the jurisdiction of the parole board and may be released on parole by the parole board, subject to the following conditions: (a) one member of the parole board shall interview the prisoner at the conclusion of 10 calendar years of the sentence and every 5 years thereafter until such time as the prisoner is paroled, discharged or, deceased. The interview schedule prescribed in this subdivision applies to all prisoners to whom this subsection is applicable, whether sentenced before, on, or after the effective date of the 1992 amendatory act that amended this subdivision. M.C.L. 791.244(1) (1992) Subject to the constitutional authority of the governor to grant reprieves, commutation, and pardons, one member of the parole board shall interview a prisoner serving a sentence for murder in the first degree or a sentence of imprisonment for life without parole at the conclusion of 10 calendar years and thereafter as determined appropriate by the parole board, but not later than every 5 years until such time as the prisoner is granted a reprieve, commutation, or pardoned by the governor, or is deceased. The interview schedule prescribed in this subsection applies to all prisoners to whom this section is applicable, whether sentenced before, on, or after the effective date of the 1992 amendatory act that amended this subsection. Mich.Comp.Laws §§ 791.234(6) and 791.244(1) (1992). Pursuant to these amendments, the interview schedule for inmates in subclass 1 was changed from 4 + 2 + 2 (etc.) to 10 + 5 + 5 (etc.). However, the 1992 amendments had no effect on the board’s parole jurisdiction or on any of the requirements governing eligibility for early release. Looking first to the Supreme Court’s recent Morales decision for guidance, it is clear that the case establishes no controlling rule of law resolving the precise issues before this court. In fact, in footnote five of the Morales opinion, the majority states that “we express no view as to the constitutionality of any of a number of other statutes that might alter the timing of parole hearings under circumstances different from those present here.” Id., 115 S.Ct. at 1603. However, the same general test for assessing ex post facto laws applied in Morales applies here and requires this court to determine whether the 1992 amendments to Michigan’s parole laws changing the frequency of scheduled parole hearings create a “sufficient risk of increasing the measure of punishment attached to [plaintiffs’] covered crimes.” Id. Considering the factors emphasized by the Morales court in applying this “sufficient risk” test to California’s amendment, the *126 court finds that the Michigan amendments are materially distinguishable in multiple ways. First, and perhaps most importantly, the Michigan amendments apply to a large class of prisoners: all mandatory lifers, par-olable lifers and prisoners serving long, indeterminate sentences. Thus, this court cannot conclude, as the Morales Court did, that the likelihood of parole for prisoners in this subclass is remote. Certainly, early release is less likely for some members of subclass 1 than for others. In a recent Michigan Court of Appeals ease, People v. Luciano Chano Lino, 213 Mich.App. 89, 539 N.W.2d 545 (1995), the court looked at parole statistics for prisoners serving parolable life sentences and noted that a majority of such defendants are never granted parole. However, these statistics do not support defendants’ claim that plaintiffs’ chances for parole are “remote,” as compared to the narrow class of double murderers involved in Morales. Based on the record in this case, the court concludes that the plaintiffs in subclass 1 are not, oh the whole, a class of “prisoners who have no chance of being released.” See Morales, — U.S. at -, 115 S.Ct. at 1605. The second important distinction between the California and Michigan amendments is that there was no case specific inquiry conducted before these plaintiffs were consigned to the less frequent hearing schedule. Rather, the delayed hearing schedule was applied automatically across the board. A third difference is that Michigan’s amendments postpone both the initial consideration hearing (from the fourth year to the tenth year) as well as the date for subsequent hearings (from every two years thereafter to every five years thereafter). Hence, the change instituted by the Michigan amendments was more severe than in California. At most, affected California inmates would have to wait an additional two years for subsequent parole interviews, while Michigan prisoners were stripped of the right to establish a record through numerous pre-eligibility hearings, and then were required to wait five years for subsequent re-interviews. Because these significant distinguishing factors prevent Morales from controlling in this case, the court has also sought guidance from more factually similar cases decided in the Circuit Courts of Appeal. The two most helpful cases are Akins v. Snow, 922 F.2d 1558 (11th Cir.1991), cert. denied, 501 U.S. 1260, 111 S.Ct. 2915, 115 L.Ed.2d 1079 (1991) and Roller v. Cavanaugh, 984 F.2d 120 (4th Cir.1993); cert. granted, — U.S. -, 113 S.Ct. 2412, 124 L.Ed.2d 635 (1993), and cert. dismissed (1993). In both of these cases, the courts considered amendments to parole board rules that did not change the terms of parole eligibility for inmates, but did change the frequency of hearings to determine suitability for parole, and in both cases, the courts concluded that the amendments were ex post facto laws. In Akins, the Eleventh Circuit held that an amendment to a Georgia parole rule increasing the period of time between parole reconsideration hearings was an ex post facto violation. In reaching that conclusion, the court emphasized that under Georgia law, “[a] parole reconsideration hearing is both in law and in practice an important component of a prisoner’s parole eligibility.” Id. at 1565. The Akins court reasoned that: [sjince the [parole board] is required to hold some type of parole reconsideration hearing before granting parole, an inmate is effectively ineligible for parole between two reconsideration hearings. Because an inmate is not paroled without a parole reconsideration hearing, the hearing must be considered an essential part of parole eligibility. Id. at 1562 (emphasis in original). The Fourth Circuit adopted this reasoning in Roller, holding that an amendment to the rule governing frequency of parole hearings in South Carolina was an ex post facto violation. Although South Carolina’s parole rules were distinguishable from Georgia’s in that South Carolina prisoners were technically eligible for parole without a hearing, the court focused on the close nexus between the availability of hearings and eligibility for parole, and concluded that the change in an inmate’s entitlement to hearings was an ex post facto violation because it “substantively disadvantaged” the prisoner by affecting his eligibility for parole. The Roller court commented: *127 Defendants also point out that Roller becomes and remains “eligible” for parole just as he would have under the law in effect when he committed his crimes. The defendants do not, however, identify any benefit flowing from the status of “eligibility” other than the right to be heard and considered for parole. Eligibility without consideration is a cold comfort. Roller, 984 F.2d at 123. Accordingly, the Fourth Circuit concluded that South Carolina applied its new statute to “alter the conditions of ... [Roller’s] pre-existing parole eligibility.” Id. at 124. The significance of the relationship between parole hearings and early release, emphasized in both Akins and Roller is consistent with the Supreme Court’s comment in Morales that “[i]f a delay .in parole hearings raises ex post facto concerns, it is because that delay effectively increases a prisoner’s term of confinement, and not because the hearing itself has independent constitutional significance.” Id., — U.S. at -, n. 4, 115 S.Ct. at 1603, n. 4. Applying this reasoning here, this court believes that the nexus between parole hearings and parole eligibility under the Michigan parole system is a crucial factor in determining whether the Michigan amendments create a “sufficient risk” of increased punishment under Morales, and therefore violate the Ex Post Facto Clause as applied to plaintiffs. The record in this case reveals that in the normal course of things, a favorable recommendation for the early release of inmates similarly situated to plaintiffs occurs as part of a regularly scheduled interview, and is based on the record compiled during previous regularly scheduled interviews. Thus, it is clear that some nexus between regularly scheduled interviews and early release exists. However, the court recognizes two aspects of the Michigan Parole statutes potentially diluting this nexus. First, despite the change in regularly scheduled hearings in 1992, the Michigan parole board retained the discretion to schedule hearings more frequently than that required under §§ 791.234 and 791.244. This retained Board discretion was also part of California’s statute and was cited favorably by the Supreme Court in Morales. Id., 115 S.Ct. at 1604. Second, in Michigan a parole hearing is not required in order for a prisoner to be eligible for parole. See M.C.L. § 791.235(1). Thus, there is no legal nexus between the decrease in regularly scheduled parole hearings and eligibility for parole. However, despite this lack of a legal nexus, the record in this ease suggests that the relationship between regularly scheduled parole hearings and an inmate’s opportunity for early release is, in fact, quite significant. Because the Michigan amendments at issue have only been in effect since 1992, meaningful statistics about their effect on early release are unavailable. However, the court notes that defendants were unable to provide the court with any examples of inmates released without a hearing under M.C.L. § 791.235(1) or any cases where the board exercised its discretion to interview an inmate prior to his next regularly scheduled interview following the enactment of the 1992 amendments. Hence, the court concludes ■ that the lack of a legal nexus between regularly scheduled hearings and parole is insignificant. Moreover, other evidence in the record supports the conclusion that regular parole hearings play an important role in determining whether an inmate is suitable for early release from prison. A former parole board member testifying before the court gave her opinion that interviews occurring prior to the date of parole eligibility were very important, because these interviews give the prisoner a chance to create a record and to receive guidance from the board about meeting the requirements for parole eligibility. The same witness stated her opinion that inmates who had been interviewed previously received a more thorough examination by the board. Perhaps an even more significant aspect of the 1992 amendments is the delay in regularly scheduled hearings after an inmate becomes eligible for parole. Under the 1982 law, an eligible prisoner would continue to have hearings every two years, as opposed to the five year schedule established by the 1992 amendments. In support of their claim that this delay created a sufficient risk that *128 certain prisoners would be confined for longer periods of time, plaintiffs provided the court with examples of individual inmates likely to be harmed by the change. These examples include individuals with strong parole records who were recognized as “model prisoners” by the parole board. Some of these individuals had already received favorable votes for release from a minority of board members and were well on their way to early release when the 1992 amendments were enacted. At the time they were last passed over for parole, it was understood by the board and the inmate that another review would be conducted in two years. However, when the 1992 amendments took effect, these interviews were automatically delayed for an additional three years. This court finds that in at least some of these cases, a sufficient risk exists that a favorable recommendation for early release would have occurred during one of the next two regularly scheduled interviews under the 1982 statutes, and did not occur because of the 1992 amendments. Unlike the California statute in Morales, Michigan’s amendments postponed subsequent interviews without any inquiry into an inmate’s likelihood of receiving a favorable recommendation at his next review. Without this inquiry, the board’s retained discretion to conduct a hearing before five years is virtually meaningless. In the words of the Fourth Circuit, this court concludes that “ [e] ligibility without consideration is a cold comfort.” See Roller, 984 F.2d at 123. For these reasons, this court concludes that the 1992 amendments to Michigan’s parole statutes decreasing the frequency of scheduled parole interviews are ex post facto laws as applied to the members of subclass 1. Although it is impossible to predict exactly which inmates in this subclasses would have been released early under the 1982 interview system, and a majority of these inmates may never have been released early, this court finds that the opportunity for early release is not so “remote” for these inmates that the 1992 amendments created “only the most speculative and attenuated risk of increasing the measure of punishment attached to the covered crimes.” See Morales, — U.S. at -, 115 S.Ct. at 1605. Rather, this court concludes that the 1992 amendments to Michigan’s parole laws produce a “sufficient risk of increasing the measure of punishment attached to the covered crimes” for a significant number of subclass 1 members. Id., 115 S.Ct. at 1603. For these reasons, the court GRANTS plaintiffs’ motion for summary judgment and DENIES defendants’ motion for summary judgment as to the prisoners in subclass 1 serving mandatory life, parolable life or long, indeterminate sentences who committed their crimes and were convicted between 1982-1992. B. EX POST FACTO CLAIM OF SUBCLASS 2 Subclass 2 consists of those inmates who committed their crimes and were convicted between 1977-1982 receiving parolable life sentences or LIDs. These plaintiffs have argued that because the 1982 statute establishing a 4 + 2 + 2 (etc.) hearing schedule was applied to them retroactively, the ex post facto analysis for these inmates should also be based on a comparison between the 1982 and 1992 statutes. In other words, plaintiffs claim that retroactive application of the 1982 amendment effectively “grandfathered” these subclass members. This court is aware of only one case, In re Jackson, 39 Cal.3d 464, 216 Cal.Rptr. 760, 703 P.2d 100 (1985), in which a court elected to review an ex post facto challenge as if a retroactively applied post-commission statute had been in effect at the time the prisoner committed his offense. This court rejects Jackson’s reasoning and applies the long-recognized rule that an ex post facto analysis must be based on the laws in effect at the time a crime was committed. See Weaver v. Graham, 450 U.S. 24, 30-31, 101 S.Ct. 960, 965-66, 67 L.Ed.2d 17 (1981). This is true despite the fact that the 1982 laws in question were retroactively applied. Plaintiffs’ argument that “a benefit subsequently conferred cannot be taken away” would be relevant in a due process challenge, but is entirely irrelevant for determining whether a law *129 violates the Ex Post Facto Clause, which is based on an individual’s right to rely on the laws in place at the time an action is taken. 7 Hence, the question this court must decide for subclass 2 is whether the 1992 parole amendments are ex post facto laws when compared to the rules governing frequency of parole hearings in effect when the members of this subclass committed their crimes (1977-1982). For the reasons discussed below, the court concludes that M.C.L. § 791.234(6) (1992) is an ex post facto law as applied to these plaintiffs. In 1977 the Michigan Department of Corrections codified a prison regulation pursuant to the Administrative Procedures Act (“APA”). This regulation, published in the Michigan Administrative Code (MAC), reads as follows: Rule 710(2) If release is denied, the board shall furnish the resident written notice: (a) Stating with particularity the grounds on which its determination is based. (b) Including, whenever possible, a statement of what the resident should do to better qualify for parole. (c) Except for those residents who are within 18 months of sentence expiration, setting a new hearing date, to be no more than 12 months from the minimum eligibility date or previous pass-over date. MAC r. 791.7710 (“Administrative Rule 710”) (emphasis added). By its own terms, this rule does not apply to inmates who are not eligible for parole, but as discussed previously, this court concludes that this Administrative Rule was intended to apply to members of subclass 2. 8 Thus, pursuant to this rule, members of this subclass were entitled to review after seven years of incarceration, three years later, and then every year thereafter (7 + 3 + 1 + 1, etc.). The ex post facto challenge presented by subclass 2 is complicated by the fact that Administrative Regulation 710 is not a statute. Strictly speaking, Article I, § 10 of the Constitution prohibits only ex post facto laws. Of course, at the time the Ex Post Facto Clause was established, administrative regulations did not exist. Hence, this court must decide whether administrative regulations such as this are “laws” for ex post facto purposes. There are many reasons why an administrative regulation properly codified under the APA, such as Rule 710, should be considered a “law” for purposes of an ex post facto challenge. Michigan law provides that state regulations promulgated and implemented under the authority of the APA which are codified in the Michigan Administrative Code after public comment have the same legal force as statutes passed by the legislature. See Clonlara, Inc. v. State Board of Education, 442 Mich. 230, 239, 501 N.W.2d 88 (1993). Moreover, evidence of the force and effect of Administrative Rule 710 is found in Phillips v. Prison Warden, 153 Mich.App. 557, 396 N.W.2d 482 (1986), leave to appeal denied, 428 Mich. 859 (1987), in which the court issued a mandamus to force the parole board to conduct an annual parole interview under Rule 710. Because administrative rules promulgated under the APA following an opportunity for public comment are well-recognized as having the force and character of laws, this court considers them “laws” for purposes of this ex post facto challenge. The Sixth Circuit has recently embraced this conclusion, holding in Kellogg v. Shoemaker, 46 F.3d 503 (6th Cir.1995) that an administrative regulation promulgated by Ohio’s parole agency had “the same effect as legislation for the purpose of ex post facto consideration.” Id. at 509. Based on this precedent, there is no doubt that Administrative Rule 710 must be considered law for purposes of plaintiffs’ ex post facto challenge. Comparing Administrative Rule 710 to the 1992 amendment implementing a 10 + 5 + 5 (etc.) hearing schedule for members of subclass 2, the court concludes that the amend *130 ment violates the Ex Post Facto Clause. At the time members of this subclass were convicted and sentenced, they were entitled to rehearing every year after being passed over for parole, while under the 1992 amendment, subsequent hearings are scheduled only every five years. Thus, virtually all of the same reasons leading this court to grant summary judgment for plaintiffs in subclass 1 apply here as well. This court therefore concludes that the statutory change decreasing regularly scheduled parole review hearings effectively increases the punishment attached to plaintiffs’ crimes by creating a sufficient risk that subclass 2 inmates will be released later under the 1992 system than they would have under Administrative Rule 710. Because the 1992 amendment to M.C.L. § 791.284 is an ex post facto law as applied to members of subclass 2, their motion for summary judgment is GRANTED and defendants’ motion for summary judgment as to subclass 2 is DENIED. C. EX POST FACTO CLAIM OF SUBCLASS 3 The court must now determine whether the members of the remaining subclass are also victims of an ex post facto violation. The answer to this question depends on a comparison between the 1992 statutory amendments and the rules governing frequency of parole hearings in effect when subclass 3 plaintiffs committed their crimes. Subclass 3 consists of those inmates who committed their crimes and were convicted between 1977-1982 receiving mandatory life sentences, and those inmates who committed their crimes and were convicted prior to 1977 receiving mandatory life, parolable life or LIDs. When members of this subclass were convicted, the frequency of parole board hearings was determined by internal policies and procedures of the MDOC or the parole board rather than by statute or administrative regulation. The record in this case shows that it was the practice of the parole board to interview mandatory lifers convicted prior to 1982 after ten years, and every three years thereafter (10 + 3 + 3, etc.), and to interview parolable lifers/LIDs convicted prior to 1977 after seven years, and every three years thereafter (7 + 3 + 3, etc.). When M.C.L. §§ 791.234 and 791.244 were amended in 1982 to provide for an initial hearing after four years and a rehearing every two years thereafter, the amendments were applied retroactively to members of this subclasses. However, as discussed above, the fact that members of subclass 3 benefitted from retroactive application of the 1982 parole laws is insignificant for purposes of this ex post facto challenge. Having concluded that the members of subclass 3 were not “grandfathered” by the 1982 statutes set forth in M.C.L. §§ 791.234 and 791.244, the court must determine whether the internal policy directives and procedures in effect when these subclasses committed their crimes are ex post facto laws. Plaintiffs have argued that these policy directives and procedures had the force and character of laws. Because these directives provided more frequent interviews than the 1992 amendments, plaintiffs assert that the 1992 amendments violate the Ex Post Facto Clause as applied to this subclass. For the reasons discussed below, the court concludes that the internal procedures and policy directives of the MDOC and the parole board in effect when members of subclass 3 committed their crimes are not “laws” under Article I, § 10. Hence, these plaintiffs suffered no ex post facto violation and their motion for summary judgment is DENIED, while defendants’ motion for summary judgment as to subclass 3 is GRANTED. Two important considerations lead this court to conclude that the internal policy directives and procedures at issue are not laws for ex post facto purposes. The first consideration is whether these policies and procedures have the force and character of law, and the second consideration is based on the policies behind the Ex Post Facto Clause. In determining whether a policy or rule should be considered a “law” for ex post facto purposes, courts have considered many factors. In Bailey v. Gardebring, 940 F.2d 1150 *131 (8th Cir.1991), cert. denied, 503 U.S. 952, 112 S.Ct. 1516, 117 L.Ed.2d 652 (1992), the Eighth Circuit held that parole regulations issued by the Commissioner of the Minnesota Department of Corrections were not “laws” for ex post facto purposes because they did not have the same force and character of laws. In reaching this conclusion, the court emphasized that the regulations in question, which changed the frequency of parole hearings, were simply procedural aids to the body vested with discretionary authority granted by the state legislature. The court also noted that the Commissioner’s decision to promulgate such regulations was, itself, an exercise of discretion, and had no effect on the parole board’s ultimate discretion to release eligible inmates on parole.. Id. at 1156. . Similarly, in Smith v. United States Parole Com’n, 875 F.2d 1361, 1367 (9th Cir.1988), the Ninth Circuit held that “the operative factor in assessing whether a directive constitutes a ‘law’ for ex post facto purposes is the discretion that the [parole board] retains to modify that directive.” Based on these principles, the Sixth Circuit has held that administrative guidelines governing federal parole are not “laws” subject to the Ex Post Facto Clause. See Ruip v. United States, 555 F.2d 1331 (6th Cir.1977). In reaching this conclusion, the Ruip court concluded that so long as the decision to grant parole remains within the absolute discretion of the parole board, regulations which “would best effectuate the purpose of Congress in establishing the board and its parole system” were simply “guideposts” assisting the board in exercising its discretion. Id. at 1335. In Prater v. United States Parole Com’n, 802 F.2d 948, 952 (7th Cir.1986), the Seventh Circuit held that the rule against ex post facto laws applied to statutes and to administrative regulations, but not to “a mere change in enforcement methods, priorities, or policies, written or unwritten ...” Id. at 954. In reaching this conclusion, the court looked to the policies behind the Ex Post Facto Clause, noting that one important goal of the clause is to protect citizens from retroactive increases in punishment because “behavior is influenced not only by whether a proposed course of action is criminal but also by the severity of the punishment proscribed for it.” Id. at 953. The court went on to conclude that changes in the administration and enforcement of statutes have little impact on such public expectations because “[t]he decision whether to commit a crime (or which crime to commit) is unlikely to be much influenced by the details of the criminal justice system.” Id. Based on the -principles set forth in these cases, this court concludes that the internal policy directives and guidelines established by the parole board and the MDOC governing parole hearings for subclass 3 members are not “laws” for ex post facto purposes. Plaintiffs have argued that the policy directives implemented by the MDOC were binding on the parole board, giving them the force and character of laws. However, plaintiffs have not shown that the board lost all discretion to modify these directives, or to influence the MDOC to modify them. Perhaps more significantly, this court believes that the important policies behind the Ex Post Facto Clause would not be furthered by including internal policies and directives such as these, because it is highly unlikely that they could play any significant role in influencing public expectations. There is no reason to believe that the public at large could ever rely on such directives, because unlike administrative regulations promulgated under the APA, agency directives are not presented to the public for comment, and are not published in the Michigan Administrative Code. Moreover, this court believes that extending the scope of the Ex Post Facto Clause so far would only have the undesirable effect of discouraging the formulation and publication of internal agency policies. See Prater, 802 F.2d at 953. For these reasons, this court concludes that the internal policy directives of the parole board and the MDOC governing the frequency of parole hearings at the time members of subclass 3 committed their crimes are outside the scope of the Ex Post Facto Clause. Therefore, subclass 3 plaintiffs suffered no ex post facto violation when the 1992 amendments were enacted. IV. In summary, this court concludes that the 1992 amendments to Michigan’s parole laws *132 governing the frequency of parole hearings are ex post facto laws as applied to members of subclasses 1 and 2, consisting of plaintiffs who committed their crimes and were convicted after 1982 receiving mandatory life, parolable life or long, indeterminate sentences, and plaintiffs who committed their crimes and were convicted between 1977 and 1982 receiving parolable life or long, indeterminate sentences. Accordingly, plaintiffs’ motion for summary judgment is GRANTED and defendants’ motion for summary judgment DENIED as to plaintiffs in subclasses 1 and 2. The court declares M.C.L. §§ 791.234(6) and 791.244(1) (1992) to be invalid as retroactively applied to these plaintiffs, and orders defendants to reinstate a parole hearing schedule equivalent to that in place at the time these plaintiffs committed their crimes. The court further concludes that these 1992 amendments are not ex post facto laws as applied to subclass 3, consisting of plaintiffs who committed their crimes and were convicted between 1977-1982 receiving mandatory life sentences, and those who committed their crimes and were convicted prior to 1977 receiving mandatory life, parolable life or long, indeterminate sentences. As to these inmates, plaintiffs’ motion for summary judgment is DENIED and defendants’ motion for summary judgment is GRANTED. Footnotes [1] Although parole eligibility is determined as of the date of sentencing for some purposes, the Ex Post Facto Clause looks to the punishment annexed to the crime at the time the offense was committed. See Weaver v. Graham, 450 U.S. 24, 31, 101 S.Ct. 960, 965, 67 L.Ed.2d 17 (1981); Kellogg v. Shoemaker, 46 F.3d 503, 509 (6th Cir.1995). [2] In accordance with the court's decision in this case, and for the sake of clarity, this opinion will separate plaintiffs in to three (3) subclasses instead of the six (6) subclasses identified in the September, 1994 Report and Recommendation. [3] Inmates serving mandatory life sentences are not eligible for parole. However, they are eligible for early release by way of reprieve, commutation or pardon by the governor. [4] In fact, it seems as if Rule 710 was routinely ignored by the board and was not applied to any inmates. The practice between 1977 and 1982 was to interview subclass 2 members on a 7 + 3 + 3 schedule based on internal policy directives. [5] The annual reinterview did not begin until after an inmate was eligible for parole, which occurred after 10 years of incarceration for members of subclass 2. See M.C.L. § 791.234. [6] Therefore, beginning in 1982, retroactive application of M.C.L. §§ 791.234 and 791.244 established a 4 + 2 + 2, etc. hearing schedule for members of all plaintiff subclasses. [7] It is well established that none of the class members have a cognizable liberty interest in the parole procedures, policies or laws that predated the 1992 amendments. See Sweeton v. Brown, 27 F.3d 1162 (6th Cir.1994), cert. denied, - U.S. -, 115 S.Ct. 1118, 130 L.Ed.2d 1082 (1995) (holding that Michigan law provides no substantive right to parole). [8] See p. 6, supra.