OPINION
KELLEHER, Justice.
This case is before us on appeal from an order of a Superior Court justice granting the application for postconviction relief of the petitioner, Maurice Lerner (Lerner), an individual presently incarcerated at the Adult Correctional Institutions (ACI). The trial justice determined that the granting of parole-eligible status to Lerner and the subsequent revocation of that status violated the ex post facto provision of the United States Constitution. The respondents, Matthew Gill, assistant director of adult services at the Adult Correctional Institutions, and Dennis J. Roberts II, Attorney General of the State of Rhode Island, vigorously [1354]*1354oppose the Superior Court’s characterization of the events involved in this dispute. In order to clarify the issues that are presently faced by this court, a summary of the factual history that precipitated this litigation is necessary.
The gangland-style murders of which Lerner was accused occurred in the Silver Lake section of Providence on April 20, 1968. See State v. Patriarca, 112 R.I. 14, 18-21, 308 A.2d 300, 305-07 (1973). Nearly two years later, on March 27, 1970, he was convicted of two counts of murder and one count of conspiracy to commit murder. See State v. Lerner, 112 R.I. 62, 308 A.2d 324 (1973). Lerner was sentenced on September 14, 1970, to a term of ten years for the conspiracy charge and two life terms, one for each of the murders committed. All of these sentences were to run consecutively.
The issues of this case revolve around the provisions in the General Laws regarding murder and parole. For purposes of this case, the statute establishing the penalty for murder remains essentially the same as at the time of the commission of the crimes. The penalty for murder at that time was imprisonment for ten years to life. The maximum penalty faced by Lerner, therefore, for the crime of murder was life imprisonment.
The parole statutes, however, require a more detailed analysis. On the date the murders were committed, a prisoner sentenced to life imprisonment was not eligible for parole until he had served at least twenty years of his sentence. This provision was altered in 1970 when the General Assembly determined that ten years should be the minimum period served by a life prisoner before becoming parole eligible. This statute expressly repealed prior law and became effective upon its passage on April 30, 1970. As of this date, Lerner had been convicted but not yet sentenced.
Another provision that deserves mention at this point concerns the parole of prisoners subject to more than one sentence. General Laws 1956 (1981 Reenactment) § 13-8-10 provides that before being eligible for parole, a prisoner confined under consecutive sentences must serve at least one-third of the total of his sentences whereas a prisoner serving concurrent sentences becomes eligible for parole after having served a term equal to one-third of the maximum term imposed. If either calculation results in a term of years greater than ten, § 13-8-13 controls, and the prisoner becomes eligible after serving ten years. The question of parole eligibility for prisoners sentenced to more than one consecutive life term was not answered by the General Assembly until 1981, however, well after Lerner’s criminal acts, conviction, and sentencing. At that time, the Legislature specifically stated that such a prisoner must serve ten years consecutively on each life sentence before becoming parole eligible. General Laws 1956 (1981 Reenactment) § 13-8-13(b), as amended by P.L.1981, ch. 36, § 1. It is important to note that this statute applies only to crimes occurring after May 7, 1981; therefore, any possible application to the case at bar is eliminated. This assertion does not imply, however, that persons committing crimes preceding that date should be eligible for parole prior to serving ten years on each life sentence. The subsection specifically cautions against this inference.
Lerner bases his ex post facto claim on the change in the interpretation of the above statutes by the parole board. He contends that since the statutes were initially construed in his favor by deeming him parole eligible in 1979, officials could not later revise the interpretation and revoke that eligibility. To do so, he argues, violates thp prohibition against ex post facto laws found in Art. I, sec. 10, of the United States Constitution and art. I, sec. 12, of the Rhode Island Constitution. In addition, he questions the propriety of the reinterpretation on due-process grounds. He maintains that the manner in which his parole hopes were raised and later shattered was so fundamentally unfair as to violate the due-process clause of the Fourteenth Amendment. He also claims his due-process rights [1355]*1355were infringed upon when a majority of the justices of this court, at the request of the Governor, delivered an advisory opinion concerning the meaning of the above statutes. See In re Advisory Opinion to the Governor, R.I., 421 A.2d 535 (1980). Lerner characterizes this action as judicial ex post facto lawmaking. Before dealing with the merits of these claims, it is necessary to detail those activities of state officials that Lerner cites as unconstitutional.
The initial inquiry into Lerner’s parole status was held in 1976. Bradford W. Southworth, who at that time was director of the State Department of Corrections, wrote a letter to then-Attorney General Julius C. Michaelson, specifically requesting advice about the parole-eligibility date of Lerner, an inmate sentenced to two consecutive life sentences. At this time, Attorney General Michaelson construed the set of statutes discussed above as granting parole-eligible status to Lerner in 1979, after only ten years of incarceration.1 He reasoned that the ten-year minimum found in § 13-8-13 was to apply to all prisoners sentenced to life since the General Laws did not specifically provide for individuals receiving sentences consecutive to a life sentence.
Although there is no evidence in the record that Lerner was formally advised of his status in 1976, it cannot be disputed that he believed that he would be eligible for parole in 1979. He testified that in 1976 he was transferred from the Maximum Security section of the prison to Minimum Security, where he would soonafter receive such privileges as work release and furloughs. Lerner claims that he relied on the statements of state officials regarding his eligibility status in making several important family decisions. His wife and children moved from their residence in Brookline, Massachusetts, to a home in Cranston, Rhode Island, in order to take advantage of the increased visitation and furlough privileges that he believed were dependent upon his parole-eligible classification. In addition, he testified that his mother started a business in Rhode Island to provide her son with a job upon his release. This was allegedly done to increase Lerner’s chances of being granted parole by demonstrating his ability, if released, to support himself and his family in the state of Rhode Island.
The most convincing proof that Lerner was once considered eligible for parole is his actual appearance before the parole board on two occasions. He was denied parole each time. In fact, it was these appearances before the parole board that prompted an assistant attorney general to question Lerner’s eligibility. This action led the chairman of the parole board, Joseph Gal-kin, to request a clarification of the parole statute from Attorney General Roberts. The question once again posed was whether Lerner was eligible to appear before the board after serving only ten years of his two consecutive life sentences or whether he must wait twenty years.
Attorney General Roberts expressed the opinion that Lerner should serve no fewer than twenty years before becoming eligible for parole, contrary to the view of his predecessor. Such a construction, according to the correspondence of October 1979 from Mr. Roberts to Mr. Galkin, comports with the court’s imposition of two life sentences to be served consecutively as opposed to concurrently. Upon Mr. Galkin’s request for a clarification of this important issue, the Attorney General reiterated his previous position. He emphasized that any other conclusion would negate the effect of consecutive sentences for life and would be inconsistent with the relevant statutes.
On December 11, 1979, Mr. Galkin wrote a memo to John J. Moran, director of the Department of Corrections, advising him of Lerner’s revised status. Lerner testified that in January of 1980 he was told that he would have to return to the Maximum Security section of the prison. He was apparently shown the communications that Mr. Galkin and Attorney General Roberts had [1356]*1356exchanged as proof of the change in his status. Lerner remained at Maximum Security until April of 1980, when the classification board determined that he could be returned to Minimum Security. After the receipt of the advisory opinion to the Governor, Lerner was returned to Maximum, where he stayed for six months. Thereafter, Lerner was again transferred to Minimum, where he remains today.
Lerner first pursued his constitutional claims before the United States District Court for the District of Rhode Island in the form of a petition for a writ of habeas corpus. A hearing was held in October of 1981, in which testimony was presented detailing the facts previously discussed. The District Court judge stayed Lerner’s federal action until such time as he had exhausted his state remedies. Accordingly, Lerner filed an application for postconviction relief in Providence Superior Court. The record indicates that the trial justice revived Lerner’s eligibility status after consideration of the transcript of evidence adduced in the Federal Court2 as well as memoranda and oral arguments of counsel. The restoration of eligibility was based on the trial justice’s belief that the 1979 change in Lerner’s eligibility status violated the ex post facto provision of the Federal Constitution.
Ex Post Facto Clause
In asking that the trial justice’s order be affirmed, Lerner advances several alternative arguments. His main reliance, however, is upon the ex post facto clause that is to be found in both the United States and the Rhode Island Constitutions. The proviso in each Constitution is almost identical to that in the other. U.S.Const., Art. I, § 10; R.I. Const., art. I, § 12. Our conclusion, which differs from that of the trial justice, will be based for the most part on the various pronouncements by the United States Supreme Court and the ongoing developments that have occurred in the area of administrative law.
The authoritative case that must begin any thorough discussion of ex post facto laws is Calder v. Bull, 3 U.S. (3 Dall.) 385, 1 L.Ed. 648 (1798). In that case, Justice Chase clearly enumerated what types of law the ex post facto clause prohibited. Although the pronouncement that follows occurred early in the Court’s history, it remains the foremost point of reference when considering ex post facto controversies. The following were considered ex post facto laws:
“1st. Every law that makes an action done before the passing of the law, and which was innocent when done, criminal; and punishes such action. 2d. Every law that aggravates a crime, or makes it greater than it was, when committed. 3d. Every law that changes the punishment, and inflicts a greater punishment, than the law annexed to the crime, when committed. 4th. Every law that alters the legal rules of evidence, and receives less, or different, testimony, than the law required at the time of the commission of the offence, in order to convict the offender.” 3 U.S. (3 Dall.) at 390, 1 L.Ed. at 650.
Over a century later, the Court placed an additional restriction on the application of the ex post facto clause. In Ross v. Oregon, 227 U.S. 150, 33 S.Ct. 220, 57 L.Ed. 458 (1913), the plaintiff argued that the judicial construction of a statute that had been in effect at the time his acts took place, but had not previously been construed, constituted an ex post facto violation. In rejecting this claim, the Court stated:
“But that provision of the Constitution, according to the natural import of its terms, is a restraint upon legislative power and concerns the making of laws, not their construction by the courts.” Id. at 161, 33 S.Ct. at 222, 57 L.Ed. at 463.
Further, in quoting New Orleans Water Works Co. v. Louisiana Sugar Refining Co., 125 U.S. 18, 30, 8 S.Ct. 741, 747, 31 L.Ed. 607, 612 (1888), the Court stated:
[1357]*1357“The prohibition is aimed at the legislative power of the State, and not at the decisions of its courts, or the acts of administrative or executive boards or officers, or the doings of corporations or individuals.” 227 U.S. at 162, 33 S.Ct. at 222-23, 57 L.Ed. at 463.
It is important to keep this limitation in mind since it gives rise to a decisive issue in this case. The respondents maintain that the interpretation of a statute by an administrative agency, here the parole board, does not constitute a law within the meaning of the ex post facto clause. Lerner, of course, takes the opposite view. The arguments each puts forth will be analyzed in more detail below. As an aid to this analysis, it is helpful to keep in mind the purposes served by the ex post facto clause.
In Weaver v. Graham, 450 U.S. 24, 101 S.Ct. 960, 67 L.Ed.2d 17 (1981), the Court discussed various reasons for the existence of the ex post facto clause in the Constitution. First, legislative acts should give fair warning of their effect and allow individuals to rely on their meaning until explicitly changed. Second, the ban on ex post facto laws is necessary to prevent lawmakers from abusing their power through arbitrary action. Id. at 28-29, 101 S.Ct. at 964, 67 L.Ed.2d at 23. A third purpose, discussed in Warren v. United States Parole Commission, 659 F.2d 183 (D.C.Cir.1981), must be added to these considerations. Since a primary purpose of the criminal-law system is the deterrence of potentially wrongful conduct, an ex post facto law, which enacts a penalty after an action has been taken, fails to meet this goal, for it is impossible to deter acts already committed.
These purposes should be remembered when considering whether a law is constitutional or not under the ex post facto clause. If these goals are not in some way enhanced, a court should be most reluctant to invoke the ex post facto prohibition. Although we do not plan to restrict ourselves to an analysis based solely on these criteria, as some scholarly commentators have suggested (see Ex Post Facto Limitations on Legislative Power, 73 Mich.L.Rev. 1491 (1975)), we shall be mindful of them as we proceed to the case at hand.
Before proceeding, however, it is helpful to narrow our inquiry. A brief examination of the types of laws, specified in Calder, that violate the ex post facto provision indicates that only the following one of the four possibly applies to this case, that is, a law “that changes the punishment, and inflicts a greater punishment, than the law annexed to the crime, when committed.” Calder v. Bull, 3 U.S. (3 Dall.) at 390, 1 L.Ed. at 650. The respondents argue that there had been no change in the punishment that could have been imposed on Lerner in April 1968 and raise several points in support of this contention. Since we believe that the issue of whether an interpretation of a statute amounts to the enactment of a law is dispositive of Lemer’s ex post facto claim, we shall not address all of respondents’ contentions.
As indicated previously, Ross v. Oregon, supra, limited the application of the ex post facto clause to legislative acts. Ross explicitly excluded from its reach decisions of courts, acts of administrative or executive boards or officers, or the activities of corporations or individuals. If this were the extent of the Court’s holding, our inquiry could end here since it is the activities of the parole board and not the Legislature which are involved in this litigation. The Ross Court, however, went on to caution:
“But whilst thus uniformly holding that the provision is directed against legislative, but not judicial, acts, this court with like uniformity has regarded it as reaching every form in which the legislative power of a State is exerted, whether it be a constitution, a constitutional amendment, an enactment of the legislature, a by-law or ordinance of a municipal corporation, or a regulation or order of some other instrumentality of the State exercising delegated legislative authority.” 227 U.S. at 162-63, 33 S.Ct. at 223, 57 L.Ed. at 463-64.
[1358]*1358Lerner claims that the parole board acted in a quasi-legislative capacity in determining its eligibility policy regarding prisoners serving consecutive life sentences. The respondents argue that since the General Assembly did not delegate rulemaking authority to this administrative agency, its interpretations were not laws in any sense of the word.
To resolve this dispute, we feel it is helpful to turn to 2 Davis, Administrative Law Treatise, § 7:8 at 36 (2d ed. 1979), where a distinction is drawn between an interpretive rule and a legislative rule. A legislative rule is the product of an exercise of delegated legislative power to make laws through rules whereas an interpretive rule is any rule an agency issues without exercising the delegated legislative power to make law through rules. The validity of a legislative rule depends upon whether it is within the power granted by the Legislature, issued pursuant to proper procedure, and reasonable as a matter of due process. Once the validity of such a rule is established, it is as binding on a court as a valid statute. Interpretive rules, on the other hand, do not have the force of law. Courts may substitute their judgment for that of the administrative agency in deciding whether or not to enforce an interpretive rule. Although a court may choose to defer to an agency’s judgment, it is not required to do so. See Niles v. Boston Rent Control Administrator, 6 Mass.App.Ct. 135, 374 N.E.2d 296 (1978); General Electric Credit Corp. v. Small, 584 S.W.2d 690 (Tex.1979).
To determine whether a rule is to be classified as legislative or interpretive, one must consider the power assigned to the administrative agency. If a statute expressly delegates power to interpret and define certain legislation to an agency, regulations promulgated pursuant to that power are legislative rules having the force of law. Batterton v. Francis, 432 U.S. 416, 97 S.Ct. 2399, 53 L.Ed.2d 448 (1977). If the lawmaking branch has not conferred such authority upon the agency promulgating the rule, the promulgation is interpretive and is not to be considered controlling by the courts. General Electric Co. v. Gilbert, 429 U.S. 125, 97 S.Ct. 401, 50 L.Ed.2d 343 (1976); Skidmore v. Swift & Co., 323 U.S. 134, 65 S.Ct. 161, 89 L.Ed. 124 (1944).
It is important to note at the outset that the parole board’s interpretations of the statute in question possessed few, if any, characteristics of a formal rule. If a label were to be placed on the parole board’s 1976 determination that concerned when a prisoner sentenced to two consecutive life sentences would be eligible for parole, its determination would be classified as “interpretive” and not “legislative.” The General Assembly has not delegated authority to the parole board to issue rulings on the meaning of parole-eligibility statutes. Sections 13-8-1 et seq. These rulings do not have the force of law and do not constitute binding authority. The interpretations represent no more than an agency’s opinion regarding the application of a statute to a particular situation.
The significance of classifying a rule as interpretive, rather than as legislative, for ex post facto purposes has not been conclusively determined. It is clear that a ruling made under delegated legislative authority is treated differently by courts from an interpretive rule, which has no such authority behind it. The question is whether this distinction should extend to the area of ex post facto law. In view of the history and purpose of the ex post facto clause and current case law, we must respond in the affirmative.
As noted earlier, Ross v. Oregon, 227 U.S. 150, 33 S.Ct. 220, 57 L.Ed. 458 (1913), held that the ex post facto clause applied solely to legislative action, not judicial, administrative, or individual. A Legislature could not avoid the commands of this constitutional prohibition merely by delegating authority to another instrumentality of the state. The purpose of this restriction was aptly expressed by the court in Rodriguez v. United States Parole Commission, 594 F.2d 170 (7th Cir.1979).
[1359]*1359“When Congress has delegated to an agency the authority to make a rule instead of making the rule itself, the resulting administrative rule is an extension of the statute for purposes of the clause. What Congress cannot do directly, it cannot do by delegation.” Id. at 173.
As discussed previously, one of the major purposes of the ex post facto clause is the prevention of legislative abuse. When the Legislature vests in an agency the authority to make binding law through its rulemaking power, the arbitrary abuse of that authority in the hands of the agency must still be prevented. If the agency has no such authority, however, the danger of legislative abuse is not present and the ex post facto clause need not be invoked. This was particularly true of Rhode Island’s parole board.3 Although the Legislature created the parole board, it did not vest the board with official authority to promulgate rules in interpretation of statutes. Apparently the board on its own chose to make its interpretive determination absent a legislative mandate. Its decisions did not have the force of law and therefore cannot be considered a legislative abuse. To hold otherwise would ignore the precepts upon which the ex post facto prohibition is based.
A review of recent case law also leads to the conclusion that no ex post facto violation occurred in this case because the parole board had no authority to make law and did not do so. Since there is no Rhode Island precedent dealing with ex post facto infringements through changes in parole law, it is necessary to look to other jurisdictions for guidance. It must be remembered, however, that such cases are not controlling and need not be followed.
Lerner, in making his ex post facto argument, places great emphasis on Love v. Fitzharris, 460 F.2d 382 (9th Cir.1972), since its factual pattern is somewhat similar to what happened when the parole board first considered Lerner’s eligibility status. Love had been sentenced to serve two consecutive terms for violations of California’s Health and Safety Code § 11500.5. This section provided that an individual convicted under § 11500.5 must serve at least two and one-half years of his sentence before becoming eligible for parole.4 The general parole statute, Cal.Penal Code § 3049 (West 1982), provided that a prisoner became eligible for parole upon serving one-third of his minimum sentence. The State Department of Corrections sent Love a formal document entitled “Notice of Legal Status” advising him that he would be eligible for parole in three years and four months. The department determined that in the case of consecutive sentences, the parole statute controls and therefore Love would be required to serve one-third of the combined minimum sentence of ten years on his two convictions. The department altered its position shortly thereafter by concluding that § 11500.5 did apply to persons under consecutive sentences for narcotics violations. The corrected “Notice of Legal Status” sent to Love advised him that he would have to serve five years (two and one-half years per conviction) before becoming eligible for parole. The United States Court of Appeals for the Ninth Circuit held that this alteration in Love’s status by the Department of Corrections constituted an ex post facto violation.
At first glance, the sweeping language used by the Love court might seem to encompass the situation presented here. Apparently ignoring the distinction between legislative and interpretive rules, the court stated:
[1360]*1360“Absent a court pronouncement on the matter, the interpretation of the relationship between the statutes placed upon them by the administrative agency charged with their enforcement has the force and effect of law.” Love v. Fitzharris, 460 F.2d at 385.
For reasons that follow, we respectfully decline to subscribe to the rationale and conclusions found in Love. Love, unlike Lerner, was sent formal notice of his status upon his arrival at prison. Although we do not suggest that form alone determines whether a rule has the force of law, it can be indicative of the authority possessed by the agency that issues it. As noted earlier, it is difficult to conceive of Rhode Island’s parole-board determination in regard to Lerner’s parole-eligibility status as being the equivalent of a rule of any kind.
A further difference between Love and Lerner can be drawn from their respective statutory schemes. In Love the Department of Corrections had to choose between the application of the general parole statute or the particular section under which the petitioner was convicted. Love did not involve a difference of opinion over the thrust of the eligibility statute, but the California officials were confronted with the application of alternate statutes. In our opinion, the application of a completely different law to a prisoner’s sentence more closely approximates the type of abuse the ex post facto clause was intended to prevent than a difference of opinion about the reach of a statute that had never been subject to judicial scrutiny.
Apart from these particular differences, we cannot subscribe to the broad generalization advanced by the Love court. First, by ignoring any distinction between legislative and interpretive rules, Love fails to comport with the current views expressed in the area of administrative law. According to Love, any agency interpretation of a statute carries with it the force of law. We have already determined that the force of law given to a rule depends upon the presence or absence of express legislative authority to make that rule. An interpretive decision such as that made by the parole board, which does not have this authority, does not have the force of law.
The position we take here finds support in the reasoning of other courts that have dealt with the relationship between parole changes and the ex post facto clauses. Rodriguez v. United States Parole Commission, 594 F.2d 170 (7th Cir.1979), involved the application of parole guidelines to a prisoner who committed his offense prior to the issuance of these criteria. The petitioner claimed, and the court agreed, that the application of these rules to his case violated the ex post facto clause. In so holding, the court determined that the regulations were equivalent to a statute for ex post facto purposes. This finding was based on the fact that Congress explicitly delegated authority to the parole commission to promulgate rules; the parole guidelines were an indirect product of congressional action. Similar reasoning can be found in Geraghty v. United States Parole Commission, 579 F.2d 238 (3d Cir.1978). In that case, the court determined that the ex post facto clause applies to a retroactive increase in punishment brought about by rulemak-ing, “the administrative equivalent of legislation.” Id. at 266. The court expressed its disfavor at allowing a Legislature to escape constitutional limitations on its power through delegation.
In Lerner’s situation, it is clear that the Legislature was not involved in the parole board’s eligibility decision, either directly or indirectly.5 The board’s determinations in [1361]*1361regard to his eligibility date were made without the express authority of the legislative branch to do so. Further, in making its decisions, the parole board did not act in a legislative manner. The process by which a conclusion was reached appeared informal. The board sought advice from one attorney general and followed his advice but subsequently received other advice from the present attorney general. The conflicting opinions have led to this litigation. There is no record of formal hearings, debate, or discussion of the issue. In addition, once the decision was made, no statement of policy, decree, or formal letter was issued. The parole board could not have been less legislative in its actions. The board’s interpretations of the parole statute are in no way the result of a delegation of the General Assembly’s legislative prerogatives. For the above reasons, we reject Lerner’s ex post facto claim and uphold the actions of the parole board.
Due Process
When Lerner was before the trial justice, he also challenged the propriety of the change in his eligibility status on due-process grounds. Although the trial justice was well aware of the grounds upon which Lerner based his due-process claim, in granting relief he rested his grant solely on the ex post facto question. Since the due-process issue was properly raised at the trial level, we shall consider it here.
There are several facets to Lerner’s due-process claim. Initially, he refers us to Bouie v. City of Columbia, 378 U.S. 347, 84 S.Ct. 1697, 12 L.Ed.2d 894 (1964), and then claims that the parole board should be considered in the same light as the judicial body faulted by the United States Supreme Court in Bouie for engaging in what was described by the Court as the equivalent of ex post facto lawmaking. In Bouie the Court was concerned with a 1960 lunch-counter “sit-in” and a South Carolina anti-trespassing statute that made it a crime to enter the property of another after receiving notice from the property owner not to do so. The record indicated that there was no notice prohibiting entry into the drugstore where the lunch counter was located, but the South Carolina Supreme Court construed the statute so that it applied not only to the act of entry but also to the act of remaining on the premises after being told to leave. Mr. Justice Brennan, in writing for the majority, after first explaining that due process requires a criminal statute to give fair warning of the conduct that it seeks to make criminal, observed that the right to fair warning can be denied not only by “vague statutory language but also from an unforeseeable and retroactive judicial expansion of narrow and precise statutory language.” Id. at 352, 84 S.Ct. at 1702, 12 L.Ed.2d at 899.
The parole board, when it first gave Lerner eligibility status and then subsequently took it away, was not engaged in judicial legislation that was nullified in Bouie. The board’s change in viewpoint is similar to one exhibited by an individual who receives conflicting advice after consulting with two or more attorneys. Our examination of § 13-8-13, the statute that speaks of parole eligibility for a prisoner who may be serving a life sentence, is nowhere so precise as the South Carolina antitrespassing statute, especially as it concerns consecutive life sentences. Lerner gains nothing from the ruling in Bouie.
In further pursuance of his due-process claim, Lerner next charges that the change in his eligibility status was “fundamentally unfair” and refers us to Breest v. Helgemoe, 579 F.2d 95, 101 (1st Cir.1978), where the First Circuit Court ruled that a New Hampshire trial court could alter a life sentence imposed following a first-degree murder conviction so as to comply with a statute that required a certificate indicating whether or not the homicide was “psycho-sexual in nature.” If so certified, the prisoner’s parole status is severely affected. The Circuit Court, while noting that the trial justice had a duty to impose a sentence [1362]*1362that complied with the certification statute, expressed the thought that the power of a sentencing court to correct an invalid sentence
“must be subject to some temporal limit. * * * After a substantial period of time, therefore, it might be fundamentally unfair, and thus violative of due process for a court to alter even an illegal sentence in a way which frustrates a prisoner’s expectations by postponing his parole eligibility or release date far beyond that originally set.” Id. at 101.
The remarks offered by the First Circuit Court about limiting the power of a trial court to correct an “invalid sentence” were pure dicta. It should be noted that many courts expressly disagree with the comment in Breest and have held that an illegal sentence may be altered or corrected at any time. United States v. Bynoe, 562 F.2d 126; 129 (1st Cir.1977); Burns v. United States, 552 F.2d 828, 831 (8th Cir.1977); People v. Favors, 42 Colo.App. 263, 264, 600 P.2d 78, 79 (1979); Coles v. State, 290 Md. 296, 303, 429 A.2d 1029, 1032 (1981); State v. Burkhart, 566 S.W.2d 871, 873 (Tenn.1978). In State v. Fry, 61 Hawaii 226, 230, 602 P.2d 13, 16 (1979), the court held that a seven-year delay between the original erroneous sentencing and its correction was not so unreasonable as to amount to a denial of due process. Although the dicta in Breest is interesting, it is too speculative to justify a decision on Lerner’s behalf. We see nothing in the holding in Breest which supports Lerner’s due-process claim.
Lerner also attempts to reinstate his parole eligibility by having this court rule that the state is estopped from denying him that status. He claims that he relied upon the parole board’s original position when he and his family made certain decisions regarding their future. As a general rule, courts are reluctant to invoke estoppel against the government on the basis of an action of one of its officers. McCoy v. State, 277 A.2d 675, 676 (Del.1971); Austin v. Austin, 350 So.2d 102, 105 (Fla.Dist.Ct.App.1977); McLaughlin v. Berle, 71 A.D.2d 707, 708, 418 N.Y.S.2d 246, 248 (1979). In Ferrelli v. Department of Employment Security, 106 R.I. 588, 261 A.2d 906 (1970), this court expressed a similar view toward the application of estoppel against the defendant on the basis of the conduct of one of its agents. In that case, we held that in proper circumstances a public agency may be es-topped from denying representations made by its agents causing an individual to act to his detriment in reliance upon those statements. The facts and circumstances of each case must be closely scrutinized to determine whether justice requires the imposition of estoppel. Schiavulli v. School Committee of North Providence, 114 R.I. 443, 448-49, 334 A.2d 416, 419 (1975).
In the present case, Lerner asserts that he and his family acted on the assumption that he would be eligible for parole in 1979. His wife and children moved from Massachusetts to Rhode Island to take advantage of the increased visitation and furlough privileges that they believed a parole-eligible inmate enjoys. This belief is erroneous. Although work-release status is limited only to prisoners who are eligible for parole, the Legislature at its January 1977 session, with the passage of P.L.1977, ch. 93, made furlough privileges available to an inmate sentenced to life imprisonment after he has served eight years of his sentence. This change is now embodied in G.L.1956 (1977 Reenactment) § 42-56-18. Thus, an inmate may, within the discretion of the classification board, enjoy the benefit of furloughs regardless of whether or not he is eligible for parole. His family’s move to a nearby Cranston home was not so fruitless as Lerner would have us believe. Assuming that it were, however, testimony given at the hearing in Federal Court, which is now part of our record, indicates that no grave loss was suffered in this move. Mrs. Lerner testified that her children have been doing well in Rhode Island schools, and aside from her daughter’s temporary health problem, they have not suffered any ill effects from this move. The only economic loss apparently suffered resulted from the fact that [1363]*1363the Massachusetts apartment the Lerners vacated belonged to Lerner’s parents. Mrs. Lerner testified that this arrangement had represented a definite financial savings for her. By moving to Rhode Island, she and her children could no longer take advantage of this savings.
In addition to this move by his wife and children, Lerner testified that his mother had opened a business in Rhode Island for the express purpose of creating an employment opportunity for him should he be paroled. We cannot accept this action as a basis for applying estoppel against the state for two reasons. First, in establishing a business enterprise solely to help her son, Lerner’s mother was taking a risk that he would not be paroled even if granted eligibility status. Second, according to Lerner, the business opened in March of 1978 and as of the date of his testimony in Federal Court, October 1981, the business had expanded to three different locations in Rhode Island. Although the financial condition of the business was never actually a subject of inquiry, such rapid expansion indicates that the business is flourishing. No economic harm appears to have resulted from this venture.
From these facts, therefore, it is difficult to see how the misstatements of the parole board induced action that changed Lerner’s position to his disadvantage. Even if we accept his contention that these events occurred solely on the basis of his imminent parole eligibility, it is apparent that none have inflicted a grave hardship on either Lerner or his family.
In determining whether estoppel is an appropriate device to use against the government, we must not only consider the problems encountered by the petitioner, but we must also be mindful of the public interest involved. United States v. Wharton, 514 F.2d 406, 412-13 (9th Cir.1975); Beacom v. Equal Employment Opportunity Commission, 500 F.Supp. 428, 435 (D.Ariz.1980). In this case a prison inmate seeks to estop the parole board from denying his 1979 parole-eligibility date. The public has a strong interest in whether and when an inmate is deemed eligible for parole. First, the parole-eligible inmate may enjoy the benefit of work release not available to other inmates. This generally involves the prisoner’s participation in activities outside the prison. It is important for the public safety and welfare that this benefit be enjoyed only by those inmates who come within the reach of the pertinent statutes. In addition, bestowing parole eligibility upon one who has not as yet been properly punished for his crime to the extent intended by the Legislature works an injustice upon the victims of the crime or their families as well as the citizens of the state.
We have before us, therefore, a case in which the public interest greatly outweighs any possible injury suffered by the petitioner. If the estoppel doctrine is to be invoked against the government only in those instances in which justice and right so require, it is clear that Lerner cannot command its imposition.
We come to that facet of Lerner’s due-process argument in which he claims that he acquired a protected liberty interest at the moment in 1979 when the parole board first told him that he was eligible for parole. This argument is meritless.
The Supreme Court has twice held within recent times that the due-process clause does not give an inmate a constitutionally protected liberty interest simply because the state provides the possibility of parole. Jago v. Van Curen, 454 U.S. 14, 102 S.Ct. 31, 70 L.Ed.2d 13 (1981); Greenholtz v. Inmates of the Nebraska Penal and Correctional Complex, 442 U.S. 1, 99 S.Ct. 2100, 60 L.Ed.2d 668 (1979). The state is under no constitutional obligation to operate a parole system, and even when it does, the mere possibility of parole does not a fortiori result in a protectable expectation of release. Rather, the state statute must be phrased in such a way that it creates a legitimate claim of entitlement and not just a unilateral hope for parole. Greenholtz, 442 U.S. at 7-12, 99 S.Ct. at 2103-06, 60 L.Ed.2d at [1364]*1364675 — 78. Thus, whether a state statute provides a protectable entitlement is a question that must be decided on a case-by-case basis.
In seeking to determine the existence or nonexistence of the protectability of Lerner’s eligibility status, we now turn to the relevant statutes, that is, §§ 13-8-8, -9, -10(a), -12, and -13. Section 13-8-8 subjects any prisoner who is required to serve a sentence in excess of six months at the Adult Correctional Institutions to the jurisdiction of the parole board. Any person, with the exception of a prisoner known as a “lifer” or a “habitual criminal” who is serving a single sentence, may, because of the provisions of § 13-8-9, be paroled after having served one-third of the sentence. If “such prisoner” is confined while simultaneously serving two or more sentences, such an individual’s parole eligibility is determined, as noted earlier, by the provisions of § 13-8-10. A prisoner serving consecutive sentences is eligible for parole after serving a term equal to one-third of the total amount of the sentences, and a prisoner serving concurrent sentences is eligible for parole once the prisoner has served one-third of the maximum sentence imposed. A habitual criminal can be paroled by virtue of § 13-8-12 after serving at least five years of the twenty-five-year enhancement period described in § 12-19-21.6
In March 1979, the time Lerner claims he became parole eligible, his source of relief was to be found in § 13-8-13. He was the beneficiary of the 1970 legislation that had reduced from twenty to ten years the time a prisoner serving a life sentence had to serve before reaching parole eligibility.
The Rhode Island Senate Journals for the January 1970 session indicate that on March 5, 1970, Senate bill S-473 was introduced and referred to the Senate Judiciary Committee. The bill, which was entitled “An Act Permitting Inmates under a Life Sentence to be Eligible for Parole in Ten Years,” would have amended § 13-8-13. The proposed legislation was accompanied by an explanation prepared by the Legislative Council which read, “The act permits inmates under a life sentence to be paroled after ten years.” The bill emerged from the Judiciary Committee on April 22, 1970, as S-473, Substitute A. The substitute, which was entitled “An Act Permitting an Inmate Under a Life Sentence or Lengthy Sentence to be Eligible for Parole in Ten Years,” made its way through the requisite legislative channels and became law on April 30, 1970.
The substitute bill gave two types of relief to different groups of prisoners. To those serving an indeterminate sentence, to wit, life, the waiting period for eligibility was reduced from twenty to ten years, and to those prisoners who were serving sentences making them ineligible for parole in less than ten years under §§ 13-8-9 and -10, eligibility arose after a period of ten years. It is obvious that the Legislature, when affording relief to the lifer serving a single sentence, believed that the same opportunity should be afforded a prisoner such as an individual who was serving one or more lengthy definite terms such as two consecutive thirty-year sentences for robbery and rape respectively.
Earlier, in the October 9, 1980 advisory opinion to the Governor, a majority of the justices of this court who responded to the chief executive’s inquiry were of the opinion that the General Assembly never intended to extend the ten-year beneficence contained in S-473, Substitute A, to a prisoner who was serving consecutive life sen[1365]*1365tences. In re Advisory Opinion to the Governor, R.I., 421 A.2d 535, 536 (1980). However, as everyone knows, an advisory opinion merely represents the individual opinions of the justices of this court and is not a binding precedent. Romeo v. Cranston Redevelopment Agency, 105 R.I. 651, 656, 254 A.2d 426, 430 (1969).
Today, a majority of this court are of the same opinion. It is clear that §§ 13-8-9 and -10 address themselves to prisoners serving a sentence or sentences the length of which was known at the time of imposition. The lifer and the habitual criminal are specifically exempted from § 13-8-9, and it is obvious that the concurrent and consecutive provisos of § 13-8-10 refer to those sentences whose terms are definite. There is nothing in the language of § 13-8-13 which would justify the conclusion that an individual such as Lerner who has been found guilty of committing two murders should receive the benefit of the same parole-eligibility provisions as are indicated for one who is serving a life sentence for having committed one murder. Thus, Lerner was not and is not entitled to be considered for parole until he has served at least twenty years at the ACI.
Our examination of the state’s parole law makes it clear that in the spring of 1979 Lerner had no justifiable expectation of parole in ten years. At that time G.L.1956 (1969 Reenactment) § 13-8-14 stated that a prisoner would not be paroled unless it appeared to the board that (1) the prisoner was deserving of such parole because of his good conduct while in prison and that such prisoner had shown a disposition to reform and (2) the prisoner would be able to secure employment upon parole and thus not become a public charge. These guidelines constituted little, if any, limitation upon the parole board’s discretion, for as we observed in State v. Ouimette, 117 R.I. 361, 367 A.2d 704 (1976) (after noting the provisions of § 13-8-14),
“The Parole Board, because of its special expertise, has been granted an extraordinarily broad amount of discretion to make decisions regarding parole release. These decisions, in reality, are based on predictions of the future behavior of prospective parolees. The scope of these ‘predictions’ is limited by the conditions outlined in § 13-8-14. * * * As can be readily seen, the board could find a number of reasons why a prisoner did not measure up to these standards.” 117 R.I. at 369-70, 367 A.2d at 709.
In Greenholtz v. Inmates of the Nebraska Penal and Correctional Complex, 442 U.S. 1, 99 S.Ct. 2100, 60 L.Ed.2d 668 (1979), the Supreme Court, in considering a Nebraska statute that was considerably more restrictive than § 13-8-14, rejected a prisoner’s contention that the language of the statute created a “proteetible expectation of parole,” pointing out that the existence of a mere statutory expectation did not mandate repeated adversary hearings in order to continue a convict’s confinement. Id. at 14, 99 S.Ct. at 2107, 60 L.Ed.2d at 679-80. Our statute creates no greater expectation of liberty than its Nebraska counterpart. Thus, we find that Lerner had no protecta-ble entitlement to parole.
The state’s bestowal of work release upon Lerner did not give him the requisite protected liberty. An examination of § 42-56-21 clearly indicates that those eligible for work release “may be permitted to work” in the community under such conditions and restrictions as the director of the Department of Corrections or his designate may impose. The quoted language clearly demonstrates the degree of discretion that is vested in the authorities at the ACI. Even if Lerner had in fact been eligible for parole in March 1979, there was not, and there is not at the moment, any obligation on the part of the prison authorities to consider him a permanent participant in the ACI’s work-release program.
The respondents’ appeal is sustained, the judgment appealed from is vacated, and the case is remanded to the Superior Court with direction to dismiss Lerner’s petition.