OPINION
HORN, Judge.This military pay case, originally assigned to The Honorable Philip R. Miller, is before this court as a result of a reversal by the United States Court of Appeals for the Federal Circuit, Ulmet v. United States, 822 F.2d 1079 (Fed.Cir.1987), of an opinion by Judge Miller, Ulmet v. United States, 10 Cl.Ct. 522 (1986). The Federal Circuit Court’s judgment, was issued as a mandate to the United States Claims Court, to conclude the proceedings in this case. Plaintiff, Lieutenant Colonel Oliver Donovan Ulmet, who had participated in “active duty for training” periods subsequent to his involuntary release from continuous active duty in the Army Reserve, successfully argued to the Federal Circuit that his accumulated time spent on active duty for training should be counted in determining his eligibility for treatment under 10 U.S.C. § 1163(d) (1982), the “sanctuary” provision, *682and for retirement benefits under 10 U.S.C. § 3911 (1982).
On September 23, 1987, pursuant to the decision and mandate by the Federal Circuit Court, the Chief Judge of the United States Claims Court vacated the original Claims Court judgment. Subsequently, the plaintiff’s case was assigned to this judge. Following the return of the case to the United States Claims Court, each party filed a dispositive motion in this court. The plaintiff later filed a Motion to Return to Active Duty and to Compel the Defendant to Calculate and Pay Damages. Based upon his reading of the Federal Circuit Court’s Opinion, the plaintiff sought back pay, reinstatement to active duty, and active duty credit towards retirement.
For some time, activity in this case was suspended in order to allow the defendant to consider the possibility of petitioning for a writ of certiorari from the United States Supreme Court. During that time, Congress passed an amendment to the statute under which this plaintiff claims he is entitled to relief. 10 U.S.C. § 1163(d) (1982) (as amended by the Act of December 30, 1987, Pub.L. No. 100-224, 101 Stat. 1536). The amended statute specifically bars individuals, who are factually, in similar circumstances to this plaintiff, from the claimed benefits under the “sanctuary” provision.
Following the amendment to 10 U.S.C. § 1163(d) (1982), the defendant filed a Motion to Dismiss on August 2, 1988. The defendant cites the amendment as the basis for filing the Motion to Dismiss so late in the action. The defendant has raised three arguments in support of its motion. First, defendant argues that the amendment was in fact only a technical amendment and that it illuminated the clear legislative intent of the original unamended version of the statute at issue in the Federal Circuit and that the Federal Circuit did not have the benefit of the technical amendment to clarify the real meaning of the earlier statute when it issued its decision in Ulmet 1. Second, the defendant contends that this court can entertain its Motion to Dismiss without requiring the defendant to petition the Federal Circuit to recall its mandate in this case. Finally, the defendant argues that this court must apply the law in effect when it reaches its decision, so that the amended statute should be applied in this case, to preclude the relief sought by the plaintiff.
Plaintiff, on the other hand, opposes the application of the amended statute to his case and requests the award of retirement benefits. Plaintiff contends that the defendant should be stopped from raising its present arguments because the Federal Circuit has already decided the plaintiff’s claim utilizing the unamended version of the “sanctuary” provision in its analysis. Plaintiff also adds an argument against retroactive application of an amended statute, amended while a case is pending, on remand, following an appeal to the Federal Circuit.
Based on the many filings with the court, the oral argument on the Motion to Dismiss, and for the reasons discussed below, the court, hereby, DENIES the Defendant’s Motion to Dismiss and GRANTS, in part, the plaintiff’s requested relief outlined in his Motion to Return to Active Duty and to Compel the Defendant to Calculate and Pay Damages, as amended by the subsequent series of filings submitted by the plaintiff.
BACKGROUND
Plaintiff, Oliver Donovan Ulmet, entered active duty as an enlisted person in the United States Army on June 6, 1958, and was appointed a Reserve commissioned officer of the Army in 1966. On September 12, 1973, as part of a reduction in strength, pursuant to Army Regulation 635-100 and 10 U.S.C. § 681(a), he was involuntarily released from continuous active duty. When involuntarily released, plaintiff had attained the rank of Captain, and had accrued 15 years, 3 months, and 7 days of active service. Plaintiff was eligible for and received a $15,000.00 severance or readjustment payment, pursuant to 10 U.S.C. § 687 (repealed 1981), for transition to a civilian career.
*683From 1973 to 1985, the plaintiff continued to serve in the Army Reserve, and participated in 15 periods of “active duty for training.”1 During these active duty for training periods, which ranged from 2 days to over 1 year, the plaintiff became skilled as a Training Requirements Analysis System Manager and attained the rank of Lieutenant Colonel (LTC). Plaintiff contends that his participation in the so called “active duty for training” enabled him to amass 18 years, 1 month and 20 days of continuous active service time in the Army.
On September 7, 1983, after LTC Ulmet claimed he had completed over 18 years of active service, including the “active duty for training” periods, LTC Ulmet requested that he be retained on continuous active duty in order to complete 20 years of service for retirement purposes. He was denied this appeal and released from active service on September 31, 1983. On a subsequent tour of active duty in April 1985, LTC Ulmet requested to be assigned to various extended tours, which he was also denied.
In the previous action in the United States Claims Court, before The Honorable Philip R. Miller, LTC Ulmet sought back pay, reinstatement to active duty status, and active duty credit for retirement starting on September 31, 1983. Judge Miller held that time served on “active duty for training” should not be included to trigger the “sanctuary” provision, 10 U.S.C. § 1163(d) (1982) and granted the defendant’s Motion for Summary Judgment, while denying the plaintiff’s cross-motion. Ulmet v. United States, 10 Cl.Ct. 522 (1986). The United States Court of Appeals for the Federal Circuit reversed Judge Miller’s decision “in light of the plain language of [the ‘sanctuary’ provision], and in the absence of legislative history to the contrary.” Ulmet v. United States, 822 F.2d 1079, 1087 (Fed.Cir.1987). The court of appeals held that LTC Ulmet’s “active duty for training” service time was to be included in the time required for “sanctuary” under 10 U.S.C. § 1163(d) (1982), and retirement under 10 U.S.C. § 3911 (1982),
The Federal Circuit in Ulmet I applied the statutory interpretation test articulated by the United States Supreme Court in Consumer Product Safety Comm’n v. GTE Sylvania, Inc., 447 U.S. 102, 100 S.Ct. 2051, 64 L.Ed.2d 766 (1980). “ ‘The starting point for interpreting a statute is the language of the statute itself’ and that, ‘absent a clearly expressed legislative intention to the contrary, that language must ordinarily be regarded as conclusive.’ ” Ulmet v. United States, 822 F.2d at 1082 (quoting Consumer Product Safety Comm’n v. GTE Sylvania, Inc., 447 U.S. at 108, 100 S.Ct. at 2056). The Federal Circuit, reasoned as follows:
Section 1163(d) of Title 10 U.S.C., commonly referred to as the “sanctuary provision,” in part, states that:
[A] member of a reserve component who is on active duty and is within two years of becoming eligible for retired or retainer pay under a purely military retirement system, may not be involuntarily released from that duty before he becomes eligible for that pay, unless his release is approved by the Secretary.
10 U.S.C. § 1163(d) (1982). By its express language, the benefits of section 1163(d), the sanctuary provision, inure to a member of a reserve component who is on active duty, and “within two years of becoming eligible for retired or retainer pay under a purely military retirement system.”
*684Section 3911 of Title 10 U.S.C. sets forth the retirement system for regular or reserve commissioned officers:
The Secretary of the Army may, upon the officer’s request, retire a regular or reserve commissioned officer of the Army who has at least 20 years service computed under section 3926 of this title, at least 10 years of which have been active service as a commissioned officer.
10 U.S.C. § 3911 (1982).
It would appear clear, therefore, that for LTC Ulmet to be eligible for sanctuary, he had to be within 2 years of the 20-year retirement requirement. Differently stated, his eligibility for sanctuary depended upon the completion of 18 years of service as computed under section 3926.
Section 3926(a)(1), in pertinent part, provides that:
(a) For the purpose of determining whether an officer of the Army may be retired under section 3911 ... his years of service are computed by adding—
(1) all active service performed as a member of the Army____
10 U.S.C. § 3926(a) (1982).
Section 3926(a)(1) leaves no doubt that, for retirement purposes under section 3911, an officer’s years of service are determined by adding “all active service performed.” Section 101, which sets forth the definitions applicable to Title 10, defines “active service” as simply “service on active duty.” 10 U.S.C. § 101(24) (1982). Section 101 also specifically defines “active duty” as follows:
“Active duty” means full-time duly in the active military service of the United States. It includes full-time training duty, [and] annual training duty....
10 U.S.C. § 101(22) (1982).
In the instant case [Ulmet /], therefore, by the express language of these statutory provisions, if LTC Ulmet has accumulated 18 years of active service, that is, “active duty,” then he is eligible for “sanctuary” under 10 U.S.C. § 1163(d). Any question as to what is included within the meaning of “active duty” is explicitly resolved by the statutory definition which specifically includes “full time training duty.” A careful reading of these governing statutory provisions leaves no doubt that the active duty training tours of LTC Ulmet are to be included in determining the 20 years of service required for retirement under 10 U.S.C. § 3911.
Ulmet v. United States, 822 F.2d at 1082-83.
The court of appeals had been urged by the defendant to apply 10 U.S.C. § 1163(d) in accordance with the definitions applicable to its predecessor statute, the lump-sum readjustment provisions of section 265 of the Armed Forces Reserve Act of 1952 (the Act), Pub.L. No. 476, § 101(b), 66 Stat. 481, when enacted, and not in accordance with the definitions applicable to section 265 as codified by the Act of July 9, 1956, Pub.L. No. 676, § 265, 70 Stat. 517, 518. “At the time of the enactment of section 265, active duty was defined as ‘full-time duty in the active military service of the United States.’ Armed Forces Reserve Act of 1952, Pub.L. No. 476, § 101(b), 66 Stat. 481.” Ulmet v. United States, 822 F.2d at 1084. However, as codified in 1956, definition of active duty in the Act included active duty for training. See Act of August 10, 1956, Pub.L. No. 1028, § 101(22), 70A Stat. 1, 5 (1956) (codified at 10 U.S.C. § 101(22) (1982)). The defendant, citing the Supreme Court, had argued that Congress did not intend to make any substantive changes in the law when it was codified. See Cass v. United States, 417 U.S. 72, 94 S.Ct. 2167, 40 L.Ed.2d 668 (1974) (cited in Ulmet v. United States, 822 F.2d at 1084). The defendant argued in support of an interpretation of the law applied by the Armed Forces since the enactment of that law.2
*685As is clear from the above quoted language of the Federal Circuit Court’s Opinion, the appellate court did not agree with the defendant’s arguments and applied the post 1956 codification definition of active duty. As a result, the court of appeals ruled that active duty for training periods are to be included as part of plaintiff’s total active service duty for the purpose of qualifying the plaintiff for “sanctuary” under section 1163(d) and retirement under 10 U.S.C. § 3911 (1982). In addition, the court rejected the defendant’s legislative history arguments by ruling that: “[njothing in the “legislative history evidences a clear legislative intent as to the statutory provisions at issue.” Ulmet v. United States, 822 F.2d at 1084. The court of appeals also rejected the defendant’s argument that plaintiff’s acceptance of $15,000.00 as readjustment pay, pursuant to 10 U.S.C. § 687 (repealed 1981), should preclude his present request for “sanctuary” and retirement benefits. Ulmet v. United States, 822 F.2d at 1085. The court of appeals noted, however, that in accordance with 10 U.S.C. § 1174(h)(1) (1982), deductions from any back pay award to the plaintiff would be required until the total amount deducted equals the readjustment pay already received. Ulmet v. United States, 822 F.2d at 1085. Finally, the court of appeals determined it would be a “gross inequity” for the' defendant to argue on appeal that plaintiff had failed to substantiate 19 of the active duty for training days, thus leaving him 8 days short of qualifying for “sanctuary,” pursuant to section 1163(d).
Subsequent to the Federal Circuit Court’s June 30, 1987 decision and August 31, 1987 mandate returning jurisdiction over this action to the United States Claims Court,3 the Chief Judge of the United States Claims Court, vacated the original Claims Court judgment on September 23, 1987.4 The reopened case was assigned to this Judge for further proceedings, following the retirement of Judge Miller.
Plaintiff’s Motion to Return to Active Duty and to Compel the Defendant to Calculate and Pay Damages was filed on October 6, 1987. At the same time, plaintiff reported that his wife was in need of immediate, medical treatment and that she might suffer physical harm if left untreated. In order to discuss this apparent need for medical attention and to ascertain the status of the defendant’s decision regarding whether or not to petition the United States Supreme Court for a writ of certio-rari with respect to the decision of the Federal Circuit in Ulmet I, this court held a status conference on October 20, 1987. At the conference, counsel for the defendant informed the court that permission for interim, medical benefits for the plaintiff’s wife had been favorably endorsed by the Office of the Chief of Staff of the Army and was being considered by the Assistant Secretary and Secretary of the Army. Recognizing the importance of defendant’s decision regarding whether to petition the Supreme Court for a writ of certiorari, this court declined to grant plaintiff’s request to cut short defendant’s time to make its decision.
After finally having been informed of the defendant’s decision not to petition the Supreme Court, and finding that the parties had been unable to resolve their differences regarding disposition of the case, the *686court held another status conference on March 30,1988. At this second conference, a schedule was set by the court, with the agreement of the parties, for the submission of additional memoranda on the calculation of damages owed to the plaintiff, based on the Federal Circuit’s decision. That schedule, however, was lengthened by a motion for an extension of time and a Motion to Dismiss filed by the defendant, as well as a series of motions filed by plaintiff, consisting of a motion to supplement plaintiff’s memorandum, a motion for leave to respond and, finally, a motion for sanctions and to hold the defendant in contempt for filing the Motion to Dismiss.
Given the protracted history of this case, on September 23, 1988, after the defendant’s motion was fully briefed, the court held an oral argument on the Defendant’s Motion to Dismiss as well as a conference to discuss the numerous motions filed by the plaintiff to date.
At the oral argument, the defendant argued for the dismissal of the case, based on the legislative action amending section 1163(d), which had occurred on December 30,1987, following the Federal Circuit mandate, but before a judgment by this court.5 The defendant claimed that the plaintiff’s complaint should be dismissed and argued that Congress has demonstrated by the Act of December 30, 1987, Pub.L. No. 100-224, 101 Stat. 1536, its intent that time spent in “active duty for training” should not be considered in determining applicability of the “sanctuary” statute to a retirement computation. Defendant maintained that not only can this court entertain its Motion to Dismiss, without flouting the mandate from the circuit court, but that the plaintiff could not base his military pay claim on 10 U.S.C. § 1163(d) (1982) (as amended by the Act of December 30, 1987, Pub.L. No. 100-224, 101 Stat. 1536).
Plaintiff, however, argued against the dismissal of his claim for three main reasons. First, plaintiff argued that the doctrine of equitable estoppel should preclude this court from honoring defendant’s motion, because the motion is based on an argument that essentially ignores the Federal Circuit Court’s decision in Ulmet I. Second, plaintiff contends that application of the “sanctuary” provision, as amended, to this case, would be an improper, retroactive application of a statute in violation of the “Due Process and Takings Clauses of the Fifth Amendment, and the prohibition against bills of attainder in Article I, Section 9, clause 3 of the United States Constitution.” Finally, plaintiff argues that the defendant mislabelled the plaintiff’s tours of duty as “active duty for training” when, in fact, he was serving on “active duty.” Therefore, according to the plaintiff, even if the amended “sanctuary” provision should be applied, the plaintiff should be entitled to his requested relief. The plaintiff also pursued positions taken in the pending Motion to Return to Active Duty and to Compel the Defendant to Calculate and Pay Damages, as amended by the series of later filings from the plaintiff since October of 1987. Plaintiff requested that he be put in the position he would have been in, but for the improper release. Plaintiff seeks back pay, allowances, benefits, and correction of his military records. It was also clear to the court at the oral argument that the parties had found it impossible to resolve the differences between them and that the parties still even disagreed on the date as of which the plaintiff is to be constructively considered released from active duty. Therefore, should the court deny the Defendant’s Motion to Dismiss, the parties remain in disagreement as to the method of calculating the award to which plaintiff would be entitled.
DISCUSSION
I. Defendant’s Motion to Dismiss
There are two issues in this case which the court must address before discussing plaintiff’s requests) for relief. First, at this time, can this court entertain the Defendant’s Motion to Dismiss, which was filed after, an opinion by the court of appeals holding that plaintiff is entitled to certain claimed relief. Second, if this court *687can address the Defendant’s Motion to Dismiss, should this court apply the law in effect when it issues its decision, as argued by the defendant in this case to be the amended “sanctuary” statute, or should it apply the unamended statute, the law in effect when the circuit court issued its decision, and reversed the judgment entered by the United States Claims Court in Ulmet I.
As to whether the court can entertain defendant’s motion at this time, the court notes that the normal procedure by which a party to an action files a Motion to Dismiss a complaint in this court is by way of Rule 12(b) of the Rules of the United States Claims Court. RUSCC 12(b), 28 U.S.C. (Supp. IV 1986). The purpose of RUSCC 12(b) is to promote the expenditious presentation of defenses and objections early in the case. The defendant, following receipt of the complaint, and prior to filing its responsive pleading, has the option of presenting to the court in the form of a motion any of the four defenses enumerated in RUSCC 12(b).
(b) How Presented. ... [T]he following defenses may at the option of the pleader be made by motion: (1) lack of jurisdiction over the subject matter, (2) lack of jurisdiction over the person, (3) insufficiency of process, (4) failure to state a claim upon which relief can be granted. A motion making any of these defenses shall be made before pleading if a further pleading is permitted. No defense or objection is waived by being joined with one or more other defenses or objections in a responsive pleading or motion.
RUSCC 12(b), 28 U.S.C. (Supp. IV 1986). RUSCC 12(c) permits that “after the pleadings are closed, but within such time as not to delay the trial, any party may move for judgment on the pleadings.” RUSCC 12(c), 28 U.S.C. (Supp. IV 1986).
This case presents an unusual procedural situation. The defendant in the instant action is not proceeding specifically under one of the four enumerated grounds listed in RUSCC 12(b), nor did defendant cite to RUSCC 12(c) as the grounds for its motion. Moreover, the defendant has filed its Motion to Dismiss subsequent to the filing of its answer and subsequent to the filing of cross-motions for summary judgment which had been decided by Judge Miller and then appealed to and decided by the Court of Appeals for the Federal Circuit. The Federal Circuit clearly had endorsed the plaintiff’s complaint as one which properly states a claim for relief and, in fact, had indicated that plaintiff was entitled to certain forms of relief, to be defined by the Claims Court on remand. The Chief Judge of the United States Claims Court then vacated the earlier judgment entered by Judge Miller and assigned the case to this Judge for further action.
When considering a Motion to Dismiss, the court’s factual inquiry essentially is framed by the allegations contained in the complaint.
When a federal court reviews the sufficiency of a complaint, before the reception of any evidence either by affidavit or admissions, its task is necessarily a limited one. The issue is not whether a plaintiff will ultimately prevail, but whether the claimant is entitled to offer evidence to support the claims. Indeed it may appear on the face of the pleadings that a recovery is very remote and unlikely but that is not the test. Moreover, it is well established that, in passing on a motion to dismiss, whether on the ground of lack of jurisdiction over the subject matter or for failure to state a cause of action, the allegations of the complaint should be construed favorably to the pleader.
Scheuer v. Rhodes, 416 U.S. 232, 236, 94 S.Ct. 1683, 1686, 40 L.Ed.2d 90 (1974). The court, therefore, is limited to deciding the Defendant’s Motion to Dismiss within the context of the well established principle that the allegations of fact outlined in the complaint be accepted and construed in a light most favorable to the plaintiff. White Mountain Apache Tribe v. United States, 8 Cl.Ct. 677, 681 (1985). The United States Supreme Court has stated that, when reviewing a Motion to Dismiss on the pleadings, the court must accept the allegations in the complaint and supporting affidavits as true. Gardner v. Toilet Goods Ass’n, 387 U.S. 167, 172, 87 S.Ct. 1526, *6881529, 18 L.Ed.2d 704 (1967). With these general principles regarding this court’s procedure for reviewing a Motion to Dismiss and supporting documentation in mind, as well as the unusual circumstances of this case, which caused the defendant to file a Motion to Dismiss at this late date, the court finds that plaintiff’s version of the facts should be accepted as true for the purposes of deciding the instant motion.
The Federal Circuit, in Ulmet I, was not inspired by the defendant’s suggestion that plaintiff might not be eligible for “sanctuary” under 10 U.S.C. § 1163(d), prior to the amendment, because following a recalculation by the defendant of plaintiff’s service time, plaintiff had failed to substantiate 19 days of his claimed service time, leaving him 8 days short of the 18 year requirement for “sanctuary.” The Federal Circuit said that such a suggestion by the Army, if accepted by the court, on appeal, would constitute a “gross inequity.” Ulmet v. United States, 822 F.2d at 1087.6
The complex issue raised by the Defendant’s Motion to Dismiss is whether this court must follow the mandate of the Federal Circuit Court of Appeals when Congress has passed legislation amending the relevant statute, subsequent to the issuance of the circuit court’s decision, but prior to the final judgment of this court, if such legislation deals with the same issue addressed in the Federal Circuit Court’s Opinion. This issue requires analysis of two well established legal doctrines: the Mandate Rule and the law of the case doctrine. Although the application of these principles, would generally weigh in favor of a decision to follow the Federal Circuit’s mandate, there are certain exceptions to the law of the case doctrine, at least one of which, the defendant argues, would permit this court to revisit this plaintiff's claim.
The so called “sanctuary” provision, 10 U.S.C. § 1163(d) (1982), at issue in this case as well as in other pending related United States Claims Court cases,7 creates a “sanctuary” from involuntary releases from active duty for reserve officers who are within two years of qualifying for retirement. See 10 U.S.C. § 3911 (1982). Quoted in applicable part, prior to the 1987 amendment, the “sanctuary” statute prevented “[a] member of a reserve component who is on active duty and is within two years of becoming eligible for retired pay ... [from being] involuntarily released from that duty before he becomes eligible for that pay, unless his release is approved by the Secretary.” 10 U.S.C. § 1163(d) (1982). Following the amendment to the statute, which the defendant argues should be applied by this court, section 1163(d) prevents: “[a] member of a reserve component who is on active duty (other than for training) and is within two years of becoming eligible for retired pay ... [from being] involuntarily released from that duty before he becomes eligible for that pay, unless his release is approved by the Secretary.” Id. (as amended by the Act of December 30, 1987, Pub.L. No. 100-224, 101 Stat. 1536) (emphasis added). Otherwise stated, the issue is whether or not this court should apply the amended statute which was enacted after the mandate by the Federal Circuit, but before the judg*689ment of this court. The issue is particularly acute in this case because, in effect, the result of applying the amended statute would be the dismissal of the plaintiffs claims.
It is generally understood that upon remand from a circuit court, the trial court must proceed in accordance with the mandate and the law of the case as established by the appellate court. See In re Sanford Ford and Tool Co., 160 U.S. 247, 255, 16 S.Ct. 291, 293, 40 L.Ed. 414 (1895). Although the majority in In re Roberts, 846 F.2d 1360 (Fed.Cir.1988) (en banc), stated that “[ujnlike the authority to reconsider its own rulings, a district court is without choice in obeying the mandate of the appellate court.” Id. at 1363 (emphasis in the original), the dissenting opinion in In re Roberts, 846 F.2d at 1366-68, is, however, also instructive:
Although it is clear that the supervening law should be applied in the subsequent proceeding, such application is far from automatic. When a ... court has been expressly bound by an appellate court’s mandate to apply the superseded law, it is conventional for an appellate court to relieve the trial court of its obligation.
Id. at 1368. The lower court must generally follow an appellate court’s mandate on issues to which the appellate court has specifically spoken. See Gindes v. United States, 740 F.2d 947, 949 (Fed.Cir.), cert. denied, 469 U.S. 1074, 105 S.Ct. 569, 83 L.Ed.2d 509 (1984). The Mandate Rule directs that: “an inferior court has no power or authority to deviate from the mandate issued by an appellate court.” Briggs v. Pennsylvania R. Co., 334 U.S. 304, 306, 68 S.Ct. 1039, 1040, 92 L.Ed. 1403 (1947) (citing Himely v. Rose, 9 U.S. (5 Cranch) 313, 3 L.Ed. 111 (1809); The Santa Maria, 23 U.S. (10 Wheat.) 431, 6 L.Ed. 359 (1825); Boyce’s Executors v. Grundy, 34 U.S. (9 Pet.) 275, 9 L.Ed. 127 (1835); Ex parte Sibbald v. United States, 37 U.S. (12 Pet.) 488, 9 L.Ed. 1167 (1838)); see also Northern Helex v. United States, 634 F.2d 557, 560, 225 Ct.Cl. 194, 197 (1980).
It is a long standing practice of federal trial courts, on remand, to refuse to reopen what has been decided.
In the absence of statute, law of the case, as applied to the effect of previous orders on the later action of the court rendering them in the same case, merely expresses the practice of courts generally to refuse to reopen what has been decided, not a limit to their power.
Messenger v. Anderson, 225 U.S. 436, 444, 32 S.Ct. 739, 740, 56 L.Ed. 1152 (1912) (citations omitted). The decision of an appellate court on an issue of law becomes the law of the case on remand. “The purpose of the law-of-the-case principle is to provide finality to judicial decisions.” United States v. Turtle Mountain Bank of Chippewa Indians, 612 F.2d 517, 521, 222 Ct.Cl. 1, 8 (1979). Otherwise “there would be no end to a suit if every obstinate litigant could, by repeated appeals, compel a court to listen to criticisms on their opinions, or speculate of chances from changes of its members.” Roberts v. Cooper, 61 U.S. 467, 481, 15 L.Ed. 969 (1857). The strong public policy upon which the doctrine hinges is that no litigant deserves an opportunity to go over the same ground twice, or to relitigate the same issue following the final decision of an appellate court. In fact, the Court of Claims described the law of the case doctrine, albeit not inflexible under certain prescribed conditions, as a rule that “a decision by the court on a point in a case becomes the law of the case unless or until it is reversed or modified by a higher court.” Raylaine Worsteds, Inc. v. United States, 146 F.Supp. 723, 726, 137 Ct.Cl. 54, 58 (1959).
Under the law of the case doctrine, a decision on an issue of law made at one phase of the litigation becomes a binding precedent to be followed in successive stages of the same case.
When a case has been once decided by [the Supreme Court] on appeal, and remanded to the Circuit Court, whatever was before this court, and disposed of by its decree, is considered as finally settled. The Circuit Court is bound by the decree as the law of the case; and must carry it *690into execution, according to the mandate. That court cannot vary it, or examine it for any other purpose than execution; or give any other or further relief; or review it, even for apparent error, upon any matter decided on appeal; or in-termeddle with it, further than to settle so much as has been remanded.
In re Sanford Fork & Tool Co., 160 U.S. at 255, 16 S.Ct. at 293 (citations omitted).
Not only is the lower court bound by the higher court’s explicit words, but the lower court is also committed to any implication derived from the higher court’s opinion. W.L. Gore & Assocs. v. Garlock, Inc., 842 F.2d 1275, 1278 (Fed.Cir.1988); see also Kori Corp. v. Wilco Marsh Buggies & Draglines, Inc., 761 F.2d 649, 657 (Fed.Cir.1985). The lower court is free only to decide anything not covered by the mandate. 7A J. Moore, Moore’s Federal Practice U 0.404[10] (2d ed. 1985). ■ “When the [reversal and] remand is general ... the [lower] court is free to decide anything not foreclosed by the mandate.” Id.
If the situation arises in which there is some doubt as to the correctness of the law of the case as established on appeal, arguments in support of departure from the appellate court’s judgment issued as a mandate generally should be addressed to the appellate court. See In re Roberts, 846 F.2d 1360 (Fed.Cir.1988) (en banc). Such arguments should be in the form of either a petition for rehearing, or if the time for the filing of a motion for rehearing has passed, by motion for recall of the mandate, or on appeal from the judgment of lower court after the remand of the proceedings. 7A J. Moore, Moore’s Federal Practice If 0.404[10] (2d ed. 1985).
Turning to the Mandate Rule, it is broader than the law of the case doctrine. Under the Mandate Rule, to the extent that the issues of law are determined by the mandate, the Claims Court must apply this law in whatever proceedings are permitted. See In re Sanford Fork & Tool Co., 160 U.S. at 255, 16 S.Ct. at 293. However, it is also understood that:
In the case of a remand for further proceedings, the mandate constitutes the law of the case only on such issues of law as were actually considered and decided by the appellate court, or necessarily to be inferred from the disposition on appeal. In the course of subsequent proceedings directed or permitted by the mandate, the [lower] court otherwise will apply the law as it reads it, subject to correction on a second appeal.
7A J. Moore, Moore’s Federal Practice, ¶ 0.404[10] (2d ed. 1983).
Notwithstanding the compulsory nature of the Mandate Rule and the lower court’s requirement to execute the mandate in accordance with the law of the case as established by the higher court,8 there are well recognized exceptions to the law of the case doctrine. The Federal Circuit Court of Appeals recognizes three exceptional circumstances in which the doctrine would be precluded from its normal operation. The first exception is when the evidence in a subsequent trial was substantially different from the former trial upon which the appellate decision was based. The second exception is when the controlling authority has since made a contrary decision of law applicable to the case and the third is when the appellate decision was clearly errone*691ous and would work a substantial injustice if upheld. Kori Corp. v. Wilco Marsh Buggies and Draglines, Inc., 761 F.2d 649, 657 (Fed.Cir.1985) (citing Central Soya Co. v. George A. Hormel & Co., 723 F.2d 1573, 1577 (Fed.Cir.1983)); Northern Helex Co. v. United States, 634 F.2d 557, 561, 225 Ct.Cl. 194, 200 (1980). Because application of any of these exceptions to the law of the case doctrine could alter the results of the decision of an appellate court, the exceptions should be narrowly construed and only utilized when the situation, specifically and unequivocally or unquestionably requires an exception’s application. See Leg-gett v. Badger, 798 F.2d 1387, 1389 n. 2 (11th Cir.1986) (cited in United States v. White, 846 F.2d 678, 685 (11th Cir.1988)).
The defendant claims that this court can act without the defendant requesting a recall from the Federal Circuit of its mandate in Ulmet I, which represents the law of this case. In support, defendant cites to a proposition developed by applying Standard Oil Co. of Calif. v. United States, 429 U.S. 17, 97 S.Ct. 31, 50 L.Ed.2d 21 (1976), and the second exception to the law of the case doctrine. The defendant, relying on a United States Supreme Court per curiam decision in Standard Oil Co. of Calif. v. United States, 429 U.S. 17, 97 S.Ct. 31, 50 L.Ed.2d 21 (1976), contends that this court may entertain a Motion to Dismiss, even after receiving a mandate from the appellate court reversing the Claims Court’s original judgment in this case. In Standard Oil, after the Supreme Court had summarily affirmed the district court’s ruling to enjoin Standard Oil from engaging in practices determined to be violative of section 3 of the Sherman Act, codified as amended at 15 U.S.C. § 3 (1976), the plaintiff filed a motion for a recall of the Supreme Court’s judgment issued as a mandate,9 and for leave to allow the district court to entertain a motion pursuant to Rule 60(b) of the Federal Rules of Civil Procedure. Fed.R.Civ.P. 60(b), 28 U.S.C. (1976). Federal Rule of Civil Procedure 60(b), to which the United States Claims Court Rule counterpart can be found at RUSCC 60(b), 28 U.S.C. (Supp. IV 1986), allows for the recall by a court of a final judgment on several enumerated grounds.10 The Supreme Court held that the district court could entertain such a motion without a recall of the Supreme Court mandate or the need to request leave from the Court. In so holding, the Supreme Court noted that the long standing requirement of appellate leave before a district court could reopen a case adds an unreasonable and unnecessary delay and expense to the litigation and “also burdens the increasingly scarce time of the appellate courts.” Standard Oil Co. of Calif. v. United States, 429 U.S. at 19, 97 S.Ct. at 32. In Standard Oil, the Court recognized:
that in the past both [it] and many Courts of Appeals have required appellate leave before the District Court could reopen a case which had been reviewed on appeal. The requirement derived from a belief that an appellate court’s mandate bars the trial court from later disturbing the judgment entered in accordance with the mandate. It has also been argued that the appellate-leave requirement protects the finality of the judgment and allows the appellate court to screen out frivolous Rule 60(b) motions.
In our view, the arguments in favor of requiring appellate leave are unpersuasive. Like the original district court judgment, the appellate mandate relates to the record and issues then before the court, and does not purport to deal with possible later events. Hence the district judge is not flouting the mandate by acting on the motion. Furthermore, the *692interest in finality is no more impaired in this situation than in any Rule 60(b) proceeding. Finally, we have confidence in the ability of the district court to recognize frivolous Rule 60(b) motions.
Id. at 18-19, 97 S.Ct. at 31-32 (citations and footnote omitted). The defendant in this action would have this court apply the Supreme Court’s logic in Standard Oil to the instant case involving the mandate from the Federal Circuit Court of Appeals in Ulmet I.
Although some of the language in Standard Oil is broad and seems to comport with the defendant’s concern over appellate judicial economy, the Standard Oil holding also contains restrictive language which requires consideration of whether the lower court judge would be flouting the mandate by acting in a particular case. The Court found that in Standard Oil, the lower court could address plaintiff’s motion to reopen the case based on alleged misconduct by government counsel and by a material witness, which had occurred during the trial in the lower court. The express language in Standard Oil also specifically addresses a Rule 60(b) motion to relieve a party from the final judgment entered in a case, and to a situation in which the record speaks for itself, in which, following the issuance of the decision of the appellate court, no new factual event is added to the existing record when the post-mandate Rule 60(b) or similar motion is filed in the lower court. In the spirit of promoting judicial economy, and because Standard Oil appears broad enough, the court will address the issues raised by the defendant’s motion.
In Ulmet I, the Claims Court “allowed” the defendant’s motion for summary judgment, “denied” the plaintiff’s cross-motion, and ordered “[t]he clerk ... to dismiss the complaint.” Ulmet v. United States, 10 Cl.Ct. at 525. On appeal, before the Federal Circuit, the Claims Court decision was “reversed”. Ulmet v. United States, 822 F.2d at 1087. Following the Federal Circuit’s judgment reversing the Claims Court’s decision, which was issued as a mandate on August 31, 1987, the Chief Judge of the United States Claims Court “vacated” the first Claims Court judgment on September 23, 1987.
Because the judgment was reversed and vacated, there is no longer an outstanding final judgment in this case in the United States Claims Court.
A final judgment makes a difference. It marks a formal point at which considerations of economy, certainty, reliance, and comity take more strength than they have before the judgment. A court’s decision to reconsider a prior ruling before the case becomes final, however, is ultimately a matter of “good sense”.
Arizona v. California, 460 U.S. 605, 644, 103 S.Ct. 1382, 1404, 75 L.Ed.2d 318 (1983) (Brennan, J., concurring in part and dissenting in part) (citation omitted). “In such a case as this, when a party seeks reconsideration of questions decided at an earlier stage of a single, continuing litigation, the law allows courts more discretion than in a case in which the party wants to upset a final judgment.” Id. (citations omitted). In the case of a remand for further proceedings, or as in this case when the Federal Circuit’s reversal of the Claims Court decision on a point of law results in further proceedings at the trial court level, no final judgment exists. The mandate constitutes the law of the case only on such issues of law as were actually considered and decided by the Federal Circuit, or are necessarily to be inferred from the disposition on appeal. In the course of subsequent proceedings directed or permitted by the mandate, the Claims Court, otherwise, will apply the law as it reads it, subject to correction on a subsequent appeal. The Federal Circuit’s mandate in this case reversed the Claims Court’s earlier judgment in Ulmet I, which dismissed the plaintiff’s complaint, and focused on the correct application of the “sanctuary” statute, prior to its amendment. This court concludes that, as long as it does not deviate from the mandate and the law of the case established by that mandate, the absence of a final judgment and concerns over appellate judicial economy, allow this court to entertain the Defendant’s Motion to Dismiss the complaint, which was rein*693stated for consideration by this trial court by virtue of the appellate decision and the mandate.
The defendant’s argument, based on the second exception to the law of the case doctrine, is that Congress, the governmental branch granted the constitutional power to legislate concerning the armed forces, is the “controlling authority,” which has changed the law applicable to this case.11 Therefore, according to the defendant’s view, the applicable law was changed by the controlling authority, and the new version must be applied in this case at this time, in conformity with the second exception to the law of the case doctrine.
At the time LTC Ulmet originally filed his case, Congress had enacted certain laws, pursuant to which he claimed relief. The Federal Circuit indicated that under the version of the statute it reviewed, plaintiff should recover. Although it is an established principle that Congress controls the plaintiff’s right to military pay, see Norman v. United States, 392 F.2d 255, 183 Ct.Cl. 41 (1968), cert. denied, 393 U.S. 1018, 89 S.Ct. 622, 21 L.Ed.2d 562 (1969), retroactive application of congressional action, in a case such as the instant one in which Congress is silent regarding retroactivity of the changed terms of the statute, is generally not inferred. See Sea-Land Service, Inc. v. United States, 493 F.2d 1357, 1369, 204 Ct.Cl. 57, 78 (1974); cf. Amen v. City of Dearborn, 718 F.2d 789, 793 (1983).
It is this court’s belief that exceptions to the law of the case doctrine should he invoked only with great care. Of relevance to any inquiry is whether the Congress, specifically and unquestionably intended the application of the amended statute to the case. See 133 Cong.Rec. H11509-10 (daily ed. Dec. 16, 1987) (remarks of Rep. Byron) 133 Cong.Rec. S18344 (daily ed. Dec. 17, 1987) (remarks of Sen. Glenn). If Congress intended the application of the amendment to cases such as plaintiff's, it could have, as it has in the past, amended the statute at issue and specified a date from which the amendment was to apply. See, e.g., Pub.L. No. 97-252, §§ 1001-1006, 96 Stat. 730 (1982) (codified at 10 U.S.C. § 1408(c)(1)); Federal Question Jurisdictional Amendment Act of 1980, Pub.L. No. 96-486, 94 Stat. 2369 (codified as amended at 28 U.S.C. § 1331 (1982)).
The second principle which the defendant has raised in support of its contention that this court should apply the amended “sanctuary” statute is the well established rule that a court is to apply the law in effect at the time it issues its decision. See Bradley v. School Board of Richmond, 416 U.S. 696, 94 S.Ct. 2006, 40 L.Ed.2d 476 (1973). The defendant reasons that since there is no final judgment in this case, and the ease is still pending, the court is required to apply the newly amended statute. 10 U.S.C. § 1163(d) (1982) (as amended by the Act of December 30, 1987, Pub.L. No. 100-224, 101 Stat. 1536).
In 1801, the Supreme Court established a principle that when, subsequent to the issuance of a judgment and before the decision of an appellate court, a law intervenes and changes the rule which governs the case, the new law must be applied. United States v. Schooner Peggy, 5 U.S. (1 Cranch) 103, 110, 2 L.Ed. 49 (1801). In Schooner Peggy, the Court held that an appellate court was required to apply an intervening controlling law, created by a treaty executed after the lower court’s judgment. Id. at 110. Quoting Chief Justice Marshall:
[i]t is in the general true that the province of an appellate court is only to en-quire whether a judgment when rendered *694was erroneous or not. But if subsequent to the judgment and before the decision of the appellate court, a law intervenes and positively changes the rule which governs, the law must be obeyed, or its obligation denied. If the law be constitutional ... I know of no court which can contest its obligation. It is true that in mere private cases between individuals, a court will and ought to struggle hard against a construction which will, by a retrospective operation, affect the rights of parties, but in great national concerns where individual rights, acquired by war, are sacrificed for national purposes, the contract, making the sacrifice, ought always to receive a construction conforming to its manifest import; and if the nation has given up the vested rights of its citizens, it is not for the court, but for the government, to consider whether it be a case proper for compensation. In such a case the court must decide according to existing laws, and if it be necessary to set aside a judgment, rightful when rendered, but which cannot be affirmed but in violation of law, the judgment must be set aside.
Id. at 110. The Supreme Court made particular reference in its opinion to the appellate court reviewing a lower court decision (not the situation in this case). The Court specifically stated that the appellate court must apply the law in effect at the time it renders its decision. Additionally, this court notes Chief Justice Marshall’s concern that any new law enacted by the controlling authority and applied by an appellate court, following judgment of the lower court, must undergo constitutional inquiries prior to its application.
This principle has been discussed and expanded on by the Supreme Court more recently in a trilogy of cases to which the parties have directed the court’s attention.12 Following Schooner Peggy, it was unclear whether a new law was automatically to be applied to then pending cases, or whether the new law was to be given ef-feet, unless there was a clear indication that it was not to apply to pending cases.
In Thorpe v. Housing Auth. of Durham, 393 U.S. 268, 89 S.Ct. 518, 21 L.Ed.2d 474 (1968), a more expansive reading of Schooner Peggy was adopted. In Thorpe, the petitioner was a month-to-month tenant in a federally assisted public housing program. Following an election making her the president of the tenant’s association, the petitioner received a lease cancellation notice. She refused to vacate the premises without first being informed of the reason for the notice. The Housing Authority brought an eviction action and the North Carolina Supreme Court affirmed the lower court’s eviction order. After the Supreme Court had granted certiorari to consider the First Amendment right implicated in the case (organizational affiliations), the Department of Housing and Urban Development (HUD) issued a circular requiring local housing authorities, which received federal funding, to give a tenant a notice of and explanation for evicting the tenant, and to allow the tenant time to reply. The Supreme Court returned the case to the North Carolina Supreme Court and, on remand, the state court affirmed its earlier holding. In once again reversing and remanding the case, the Supreme Court held that the state supreme court was required to apply the law as enacted following the HUD circular. The Court states: “that an appellate court must apply the law in effect at the time it renders its decision.” Id. at 281, 89 S.Ct. at 526 (footnote omitted). In addition, in a footnote, the Court quoted an earlier decision in which it said that “[a] change in the law between a nisi prius and an appellate decision requires the appellate court to apply the changed law.” Id. at 281 n. 38, 89 S.Ct. at 526 n. 38 (quoting Ziffrin Inc. v. United States, 318 U.S. 73, 78, 63 S.Ct. 465, 469, 87 L.Ed. 621 (1943)). The Court also noted the manifest injustice exception to the general rule and stated that, although the appellate court is to apply the law in effect at the time it renders *695its decision, it should not do so if such a change would result in a manifest injustice.
Clearly, the Thorpe language anticipated the situation in which, following a lower court decision and before the appellate court ruling, the controlling law is changed. Therefore, pursuant to the Supreme Court’s guidance, the appellate court must apply the new law, unless application of that new law would result in a manifest injustice. In the instant case, the Supreme Court’s guidance is not directly on point because the law at issue was changed after the appellate decision. Nonetheless, this court feels that the manifest unjustice concept, as further developed by the Supreme Court, is applicable to a broader range of cases, including the instant action.
In the second case in the trilogy, Bradley v. School Board of Richmond, 416 U.S. 696, 94 S.Ct. 2006, 40 L.Ed.2d 476 (1973) the Supreme Court repeated the Thorpe language and rule. Bradley is important, however, for its application of the manifest injustice exception to the doctrine that the appellate court is to apply the law in effect when it renders its decision. The Court articulated the exception in terms of a test requiring courts to focus their attention on: (1) the nature and identity of the parties, (2) the character of the rights involved, and (3) the nature of the result the application of the changed law would have on those rights. Id. at 717, 94 S.Ct. at 2019. Additionally, the Court added two new exceptions to the Schooner Peggy principle. The Supreme Court articulated the rule as follows: a federal court is to apply the law in effect at the time it renders its decision, “unless doing so would result in manifest injustice or there is statutory direction or legislative history to the contrary.” Id. at 711, 94 S.Ct. at 2016. In so holding, the Court in Bradley, rejected the contention that a change in the law was to be applied to pending cases only if the new law specified to do so. Id. at 715, 94 S.Ct. at 2018.
In the instant case, the defendant argues that neither the manifest injustice exception applies, nor do the two other exceptions recently articulated by the Supreme Court apply. Specifically, the defendant claims that the plaintiff would not suffer a manifest injustice if this court were to apply the amended “sanctuary” statute because plaintiff does not have a “vested” right to the claimed benefits flowing from the “sanctuary” provision. Additionally, the defendant states that the amended statute does not specifically direct the application of the old, rather than the new statute, to pending cases, and that there is no legislative history directing application of the old statute. The defendant points to the legislative history of the amended statute which refers to the amendment as a mere corrective measure to change the statute to reflect Congress’ original intent. 133 Cong.Rec. H11509-10 (daily ed. Dec. 16, 1987) (remarks of Sen. Glenn). The defendant concludes his argument with the notion that subsequent legislation declaring the intent of earlier legislation is to be given considerable retroactive weight in statutory construction. Red Lion Broadcasting Co. v. FCC, 395 U.S. 367, 380-81, 89 S.Ct. 1794, 1801-02, 23 L.Ed.2d 371 (1969).
In response to the defendant’s analysis of Bradley and its application to this case, the plaintiff cites to and argues that Bennett v. New Jersey, 470 U.S. 632, 105 S.Ct. 1555, 84 L.Ed.2d 572 (1985), restricted the Bradley holding. In Bennett, the Supreme Court noted the limiting language used in Bradley. The Court said that a federal court is to apply the law in effect when it renders its decision even if the intervening or new law does not expressly state it applies to pending cases. However, such is not the case and the court would not apply the intervening change to a pending case if to do so would “ ‘infringe upon or deprive a person of a right that had matured or become unconditional.’ ” Id. at 639, 105 S.Ct. at 1560 (citation omitted). The Court added that the limitation on the application of the Schooner Peggy rule “comports with another venerable rule of statutory interpretation, i.e., that statutes affecting substantive rights and liabilities are presumed to have only prospective effect.” Id. (citations omitted). The plaintiff claims that this principle and statement by the Court dictates that the amended “sanctuary” *696statute should not be applied in this case. The defendant, counters with the argument that the plaintiff does not have a vested substantive right at stake and, therefore, the amended statute may be applied.
The defendant claims that not only does the manifest injustice exception not apply, but that the legislative history exception suggests that the court should apply the amended “sanctuary” statute. The defendant’s contention raises the issue as to whether Congress, in its 1987 amendment, amended the “sanctuary” provision to apply to past duty. The defendant states that the amendment was one of many technical corrections and cites to the statements of a member of the House of Representatives 13 and the Senate14 to support the idea that the amendment was intended to be applied in this case and other pending similar cases, and to overcome any presumption that such a statute affects an individual’s substantive rights. Bennett v. New Jersey, 470 U.S. at 632, 105 S.Ct. at 1556.
The defendant’s contention fails due to the fact that the remarks of the members of Congress are primarily neutral as to their retroactive application to this and other similar pending claims. See Wilson v. United States, 16 Cl.Ct. 765 (1989). But more important, if Congress had intended application of the amended statute to pending cases, it could have explicitly stated so in the amendment. Moreover, the plaintiff in the instant case strikes this court as precisely the type of plaintiff foreseen in the manifest injustice exception to the doctrine which otherwise requires a court to apply the law in effect when it issues its decision. This court, therefore, concludes that it is proper to apply the unamended “sanctuary” provision to decide plaintiff’s claims and that the amended “sanctuary” statute should not be applied retroactively in the present case. At the time when the plaintiff completed the required service period to claim “sanctuary,” and in accordance with the Federal Circuit’s decision, he was entitled to inclusion under the statute as it read at the time he retired.
The defendant’s argument concerning the 1987 amendment as an expression of Congress’ original intent with respect to the language of section 1163(d) in effect in 1985 when plaintiff was released from the service is not persuasive, absent more specific direction in the amended statute itself. We note that “the views of a subsequent Congress form a hazardous basis for inferring the intent of an earlier one.” Wilson v. United States, 16 Cl.Ct. 765, 768 n. 2 (1989) (citing United States v. Price, 361 U.S. 304, 313, 80 S.Ct. 326, 331-32, 4 L.Ed.2d 334 (1960); Jefferson County Pharm. Ass’n v. Abbott Laboratories, 460 U.S. 150, 165 n. 27, 103 S.Ct. 1011, 1021 n. 27, 74 L.Ed.2d 882 (1983)). Moreover, the Supreme Court recently has stated that “[rjetroactivity is not favored in the law. Thus, congressional enactments and administrative rules will not be construed to have retroactive effect unless their language requires this result.” Bowen v. Georgetown Univ. Hosp., — U.S. -, -, 109 S.Ct. 468, 471, 102 L.Ed.2d 493 (1988) (citations omitted).
It is also worth noting that the Court of Claims stated that *697Alyeska Pipeline Serv. Co. v. United States, 624 F.2d 1005, 1013, 224 Ct.Cl. 240, 254 (1980) (citations omitted). While the plaintiff proclaims that this rule requires only prospective application of an amendment to a statute, the defendant argues that this rule represents the “old law,” and that the principles discussed in Bradley above, represent the new law and new legislative history indicating Congress’ position that the statute at issue should be applied retroactively.
*696‘[t]he first rule of [statutory] construction [is] that legislation must be considered as addressed to the future, not to the past.’ This rule governs unless the words in the statute ‘are so clear, strong, and imperative, that no other meaning can be annexed to them, or unless the intention of the legislature cannot be otherwise satisfied.’
*697This court determines that in view of Bowen, Alyeska Pipeline represents good precedent. As discussed above, there is no unequivocal legislative history indicating that the amendment to the “sanctuary” statute should be applied retroactively in the instant case. While it may be true that the plaintiff, a former member of the armed services, does not have an automatic or property right to retirement benefits, Al-berico v. United States, 783 F.2d 1024, 1027 (Fed.Cir.1986), there can be no disagreement that the amended “sanctuary” statute, if retroactively applied in this case, would negatively affect the plaintiff’s rights created by the unamended statute, in effect when the plaintiff retired, and as that statute was interpreted by the Federal Circuit. Moreover, it is clear that retroactive interpretations of statutes are not favored and that the amended statute did not, by its own terms, direct retroactive application of the provisions contained therein, or indicate that the amendment represented what had always been the law.
Based on the above discussion, Defendant’s Motion to Dismiss is, hereby, DENIED. The balance of this opinion deals with the plaintiff’s various claims as interpreted by applying “sanctuary” statute to his case.15
II. Plaintiffs Requests for Relief.
After the issuance of the Federal Circuit’s August 31, 1987 mandate returning jurisdiction over this case to the United States Claims Court, the Chief Judge of the United States Claims Court vacated the original judgment in this case and reassigned the case to this Judge. The plaintiff, subsequently, on October 6,1987, filed “Plaintiff’s Motion and Memorandum in Support of Motion to Return to Active Duty and to Compel the Defendant to Calculate and Pay Damages.” The plaintiff requested an order from this court returning him immediately to active duty. Moreover, plaintiff said that he was entitled to credit toward retirement eligibility for the time he claimed he should have been on active duty since September 30, 1983.
The plaintiff also requested an order compelling the defendant to calculate and pay the damages which he claims are due to him. The plaintiff argued that he was entitled to receive back pay and compensation for benefits and allowances he claimed he should have been receiving since September 30, 1983. To set a value on the claimed compensation for deprived benefits and allowances, the plaintiff argues that the court should use the values set by the Army for these benefits and allowances detailed on a “Personal Statement of Military Compensation,” which is prepared by the Department of the Army’s Finance and Accounting Center.
The plaintiff also argued that, the defendant should not be permitted to deduct any civilian pay which he had earned since the September 1983 date. As support for this proposition, the plaintiff cited to Carter v. United States, 509 F.2d 1150, 206 Ct.Cl. 61 (1975) and Casey v. United States, 8 Cl.Ct. 234 (1985). The plaintiff argues that these cases are examples of cases in which the *698Court of Claims awarded “back pay, allowances, and benefits for wrongful discharge,” including “back pay for all time periods the plaintiff was ready, willing, and able to be on duty,” with “no mention of deduction for other income.” The plaintiff further argues, citing to 5 U.S.C. § 5534 (1982), that even if he had been on full-time active duty, during the period determined to be the constructive active duty period for which he should be entitled to a back pay adjustment, the defendant would have permitted him to earn money from sources outside the service, such as from part-time employment or “odd-jobs.” As stated by the plaintiff, “the [defendant’s] offset argument is premised on the incorrect assumption that [plaintiff’s] full-time active duty pay represents the maximum amount he could have earned, while it is actually just the minimum.”
On October 16, 1987, the defendant filed its response to the plaintiff’s October 6, 1987 filing and requested the court to issue a “Call” to the Assistant Secretary for Manpower and Reserve Affairs, pursuant to 28 U.S.C. § 2507 (1982). In response to the plaintiff’s claim that he is entitled to return to active duty, the defendant correctly states that the Federal Circuit decision did not rule that plaintiff should return to active duty. As the defendant argues, and as this court finds from its own reading of the appellate decision in Ulmet I, the Circuit Court concluded that under 10 U.S.C. § 1163(d), prior to the 1987 amendment, the plaintiff should be allowed to receive credit for sufficient active duty service in order to qualify for “sanctuary.” The court of appeals stated that the time computed to satisfy the “sanctuary” provision was to be applied toward determining his 20 years of service required for retirement pursuant to 10 U.S.C. § 3911 (1982).
The defendant further contends that, pursuant to Army Regulation 635-100, ¶ 3-31, the plaintiff was solely entitled to remain on active duty until the last day of the month following the month in which he completed 20 years of active federal service. As such, the defendant claims that even under the decision of the court of appeals in Ulmet I, the plaintiff only would have the right to have his military records corrected to show that he remained on active duty so as to qualify for retirement; he would not have the right to be returned to active duty.
Citing to Laningham v. United States, 5 Cl.Ct. 146 (1984), the defendant states that it has a right to receive a credit for any civilian earnings the plaintiff received during the constructive, active duty period at issue. The defendant argues that 5 U.S.C. § 5534 (1982), upon which the plaintiff relies to support his alleged right not to deduct outside income, provides only a limited exception to the prohibition of dual compensation while an individual is employed by the federal government, and does not apply to service members on continuous active duty. To prevent the necessity for discovery on this issue, the defendant requested a sworn affidavit from the plaintiff itemizing all civilian pay earned during the constructive, active duty period at issue.
The plaintiff filed a reply to the defendants response on October 26, 1987. The plaintiff characterizes the defendant’s position as inequitable and one which would deny the plaintiff’s alleged right to return to active duty and, instead, would retroactively retire the plaintiff as of the date on which he would have completed 20 years of active duty. The plaintiff states that
[s]ince [his] retirement pay would be 52.5 percent of his active duty pay, this would greatly reduce his damage award, by tens of thousands of dollars to date. Moreover, this would be a double penalty, because the amount of retirement pay [he] would receive for the rest of his life should be increased by 2.5 percent for each year of active duty over twenty that he served.
The plaintiff claims that the defendant’s attempt to retroactively retire him is an “ad hoc damage control effort, entirely contrary to established Army retirement procedures.” The plaintiff adds that these “retaliatory” measures would deprive the plaintiff of over two years of back pay for active duty and the same period credited toward his* retirement pay.
*699The plaintiff attempts to support his allegations with what he considers to be inconsistencies in the Army’s application of Army Regulation 635-100, 113-31, to his case. The plaintiff claims the regulation cannot now be applied to retroactively retire him as of the date he completed the 20 year retirement requirement because he has never been “relieved” from active duty, as required by the regulation. Moreover, the plaintiff states that the same regulation allows the Army to elect to retain him on active duty subsequent to the 20 year mark. The plaintiff argues that, since this regulation allows for the retention of officers under voluntary retention programs, and the Army has retained some officers in the past, “this Court could ... deem the Army to have retained [him] as long as it refused to obey the law.”16 Additionally, the plaintiff directs the courts attention to 113-33 of the Army Regulation, cited by the defendant, which plaintiff argues would require that the Army provide him with six months notice prior to his scheduled release date. The plaintiff argues that he was never given the required six months notice, nor was he ever released in accordance with the Army Regulation cited by the defendant.
As to the issue of civilian pay credit toward the amount which the Army would be required to compensate the plaintiff for back pay for active duty, the plaintiff attempts to distinguish the Laningham case, along with the cases cited therein, cited by the defendant in its response dated October 16, 1987. The plaintiff contends that there is a distinction between disability pay at issue in the cases cited by the defendant and the active duty back pay at issue in this case. Second, the plaintiff argues that under the proper test set forth in Laningham to determine the civilian pay off-set, the small amount of income which he made in the interim, should be, by itself, sufficient proof that he could have earned that same amount while on full-time active duty.
In conclusion, the plaintiff requests: (1) that he be returned immediately to active duty; (2) alternatively, that he begin to receive all retirement pay and benefits immediately; (3) that the Army be ordered to calculate within 30 days, the back pay, allowances, and benefits to which he is entitled for active duty service since September 30, 1983 through the date that he is returned to active duty or retired in accordance with Army Regulations; (4) that he receive credit for active duty service toward retirement benefits for time served since September 30, 1983, through the date that he is returned to active duty, or retired, in accordance with Army Regulations; (5) that the Army specify what information it wishes him to provide the Army with regarding his civilian earnings; and (6) that he receive a retroactive promotion to the rank of colonel.
Before discussing the merits of the arguments described in the filings detailed above, there have been several additional filings in this case, which require some discussion at this time. On March 21,1988, the plaintiff filed a Motion to Enforce Mandate. With the exception of notifying the court that the plaintiff has requested that he be constructively retired as of March 9, 1988, the March 21,1988 filing did not raise any new legal issues in this case. The plaintiff requested the same relief as before and, added a request for “tangible and intangible retirement benefits.” The motion, however, did inform the court of some problems the plaintiff was having in receiving medical benefits from the defendant.
Following receipt of this motion, due to the stated urgency outlined by the plaintiff concerning a need for medical benefits, the court ordered a conference in this case to be held March 30, 1988. During the conference, as memorialized, in an Order issued on April 4,1988, the court directed the *700parties, to file a Joint Status Report informing the court of whether they had resolved their disagreement regarding the calculation of plaintiffs back pay for active duty. The court also ordered the defendant to file a memorandum addressing two specific entitlement issues.
The defendant, subsequently, on April 4, 1988, filed a Defendant’s Status Report, purportedly in response to the court’s earlier direction, representing to the court that “[t]he parties have agreed upon LTC Ul-met’s Active Federal Service and ... are computing his entitlement upon this basis.” Subsequently, on April 13, 1988, the defendant filed a memorandum, in response to the court’s order, addressing the two entitlement issues.
The defendant first addressed the retirement date issue in this case and then focused on plaintiff’s request for compensation for other allowances and benefits. As to the first issue, plaintiff’s retirement date, the defendant stated that the parties now agree that the plaintiff had 18 years and 26 days of active service, as of September 30,1983, the date he was released from active duty (REFRAD). The parties agreed that, since the appellate court’s decision set aside plaintiff’s September 30, 1983 REFRAD, the plaintiff is entitled to receive constructive service credit as if he had continued serving on active duty beyond September 30, 1983. The defendant characterized the retirement date issue as a question concerning the proper termination date for awarding constructive credit for such active duty service to the plaintiff. The defendant then proceeded to reargue the contentions originally set forth in earlier filings. According to the defendant, the applicable law and regulations require that the plaintiff would have been'mandatorily released on October 31, 1985. Therefore, plaintiff would have been retired effective October 31, 1985 and, in accordance with the regulation and at plaintiff’s request, he could have been placed on the voluntary retired list, as of November 1, 1985.
As to the plaintiff’s request for other allowances and benefits to be calculated in accordance with his “Personal Statement of Military Compensation,” the defendant states that the plaintiff’s award of constructive active duty credit from September 30, 1983, through October 31, 1985, should be limited to the following:
1. Basic pay pursuant to 37 U.S.C. § 204 (1982).
2. Basic allowance for quarters at the “with dependants rate” pursuant to 37 U.S.C. § 403 (1982).
3. Basic allowance for subsistence pursuant to 37 U.S.C. § 402 (1982).
4. Variable housing allowance pursuant to 37 U.S.C. § 403a (1982). .
5. Pay for accrued leave pursuant to 37 U.S.C. § 501 (1982).
6. Reimbursement for actual medical expenses the plaintiff incurred during the properly calculated period of constructive active duty. See 10 U.S.C. § 1074(a) (1982).
From the total of these awards, the defendant argues that it should be permitted to mitigate its damages or off set:
1. The plaintiff’s civilian earnings during the period of constructive active duty.
2. The plaintiff’s military earnings for any military service performed after the period of constructive service commences.
3. The separation pay which plaintiff received pursuant to 10 U.S.C. § 687 (repealed (1981). See 10 U.S.C. § 1174(h)(1) (1982).
4. Other government entitlements or benefits, if any, received by plaintiff for which, upon his constructive reinstatement to active duty and retirement, he would be required to reimburse the government, e.g., certain Veterans Administration benefits.
5. Federal income tax withholding.
Any other intangible or speculative benefits which members of the armed forces serving on active duty or in a retired status may receive, the defendant argues, are not compensable.
In response, the plaintiff filed a memorandum in which he claims entitlement to $292,937.52 for back pay allowances, and benefits through March 1, 1988, his chosen *701retirement date. Plaintiff, once again, argues for the calculation of a greater period of active service time as well as for a valuation of his other active service benefits, in accordance with the “Personal Statement of Military Compensation,” first brought to the court’s attention in plaintiff’s October 6, 1987 motion. Second, the plaintiff disagrees with the defendant’s contentions regarding deductions from his total award.
On April 27, 1988, the defendant finally filed the calculation of plaintiff’s entitlement, as prepared by the United States Army Finance and Accounting Center. On May 16, 1988, plaintiff filed a Defendant’s Eeply Memorandum in which the defendant once again contends that it had properly computed the plaintiff’s entitlements based upon a proper retirement date of October 31, 1985. For the first time, in this filing, however, the defendant raises a jurisdictional issue with respect to plaintiff’s claim for benefits calculated using his “Personal Statement of Military Compensation.” Defendant argues that, since plaintiff failed to cite any statute or regulation documenting his right to those benefits, the court can infer that no such statute or regulation exists and conclude that it lacks jurisdiction to grant his requested relief. As to plaintiff’s request for compensation for medical expenses incurred while on constructive, active duty, the defendant states that he is entitled only to expenses relating to his own personal care and not for medical expenses incurred with respect to services performed on his spouse or dependents. The defendant contends that there is no statute or regulation requiring reimbursement for medical benefits of dependents of military personnel and cites to Lord v. United States, 2 Cl.Ct. 749 (1983), as support for this contention.
Before addressing each of the plaintiff’s entitlements individually, there are three threshold issues which the court must resolve: the plaintiff's retirement date, whether the plaintiff is entitled to compensation for the loss of other intangible benefits for the active duty period prescribed by the court, and whether the defendant is entitled to offset the compensation which the plaintiff earned as a civilian during the constructive active duty period to be prescribed by the court.
This court agrees with the approach developed by the United States Court of Appeals for the District of Columbia Circuit:
Judicial relief provided to military servicemen who have been wrongfully discharged from service has been premised upon one central principle: making the injured men “whole”. Courts attempt to return successful plaintiffs to the position that they would have occupied “but for” their illegal release from duty. In order to grant such relief, courts have not been reticent to apply the legal fiction of “constructive service”; as appellants have never been lawfully terminated from active duty, they are deemed to have served during the time of their illegal release. Accordingly, it is fully consistent with the holding in our previous opinion in this case to award appellants retroactive reinstatement to the positions they held on their respective dates of separation, with full active duty pay allowances and other benefits of service, including active duty credits for retirement purposes.
Dilley v. Alexander, 627 F.2d 407, 413 (D.C.Cir.1980).
In the .instant case, in order to implement the findings of this court the Secretary of the Army is ordered to correct the plaintiff’s service records in accordance with this opinion, which is also consistent with the Federal Circuit’s mandate in Ulmet I. While it is true that 10 U.S.C. § 1552 (1982), the provision by which the Secretary of the Army “may” correct military records is permissive, rather than mandatory, in this case, the court finds that the Secretary of the Army should compute the plaintiff’s retirement relief in accordance with the “sanctuary” provision found in 10 U.S.C. 1163(d) (1982). See Dilley v. Alexander, 627 F.2d at 413. Therefore, in accordance with creation by the courts of a legal fiction referred to as a constructive, service period, the plaintiff’s corrected records should reflect that he constructively served during the period following his original re*702lease date until his retirement date, as determined below. Once plaintiff’s military records have been corrected, pursuant to 10 U.S.C. § 1552(a) (1982), the Army is authorized, and directed, under 10 U.S.C. § 1552(c), to reimburse the plaintiff for any amounts owed.
The plaintiff is required to establish that he is entitled to money which he would have received but for the defendant’s action. See United States v. Testan, 424 U.S. 392, 96 S.Ct. 948, 47 L.Ed.2d 114 (1976); United States v. King, 395 U.S. 1, 89 S.Ct. 1501, 23 L.Ed.2d 52 (1969). Then, but only then, can this court order the defendant to reimburse the plaintiff for the claimed relief. See, e.g., Shaw v. United States, 357 F.2d 949, 963, 174 Ct.Cl. 899, 920 (1966).
Both parties agree that the plaintiff had 18 years and 26 days of active federal service in accordance with the definition of the term “active service” in 10 U.S.C. § 101(24) (1982), as of September 30, 1983, the date he was released from active duty. Both parties also agree that the plaintiff should receive constructive, service credit, as if he had continued serving on active duty beyond September 30, 1983. At issue is the calculation of the proper period of time for awarding constructive credit for such active duty service. To this issue the defendant argues that the applicable law and regulations mandate that the plaintiff would have been mandatorily released on October 31, 1985, and had he requested so, he could have been placed on the voluntary retired list, as of November 1, 1985.
The plaintiff, however, in all of his filings to date, requests that the court order a constructive service period beyond the 20 years required by the “sanctuary” statute, as interpreted by the Federal Circuit in Ulmet I. The plaintiff contends that he is entitled to a retirement date, pursuant to 10 U.S.C. § 3911 (1982), of March 1, 1988, the date he formally requested to be retired from the service.
When reconstructing service records in the past the Court of Claims constructively retained on duty officers who were invalidly separated if “there was very good reason to believe that the officer would have been retained if the authorities had recognized, at the time of his separation, the invalidity of separating him on the ground that was actually taken.” Merriott v. United States, 163 Ct.Cl. 261, 264-65 (1963) (citing Appell v. United States, 128 F.Supp. 153, 130 Ct.Cl. 688 (1955); Boruski v. United States, 155 F.Supp. 320, 140 Ct.Cl. 1 (1957); Egan v. United States, 158 F.Supp. 377, 141 Ct.Cl. 1 (1958)). The Court of Claims also found that it had no reason to believe that a service member should have been constructively released in a case in which the defendant did not prove or offer to prove that, if the plaintiff's case had been properly considered, the service would nevertheless have released him. Shaw v. United States, 357 F.2d 949, 962 n. 22, 174 Ct.Cl. 899, 920 n. 22 (1966).
The “sanctuary” statute, 10 U.S.C. § 1163(d) (1982), as interpreted by the Federal Circuit in Ulmet I, entitled the plaintiff, “who [was] on active duty and [was] within two years of becoming eligible for retired pay or retainer pay ... not [to] be involuntarily released from that duty before he [became] eligible for that pay, unless his release [was] approved by the Secretary.” Id. According to the defendant, once plaintiff qualified for retired pay, 10 U.S.C. § 681(a) (1982), Army Regulation (AR) 635-100, ¶ 3-31, mandated plaintiff’s release from active duty. Moreover, the decision issued by the Federal Circuit in Ulmet I, did not go so far as to order that the plaintiff could choose his own release date without regard to the needs of the Army or the regulatory framework which governs release from the service.
Since the plaintiff had 18 years and 26 days of active service on September 30, 1983, he would have reached the 20 year mark on September 3, 1985. Therefore, both parties would agree that the plaintiff should receive constructive, active duty service at least for the period beginning September 30, 1983, and ending October 31, 1985.17 While the defendant claims that *703October 31, 1985 is the date upon which 10 U.S.C. § 681(a) (1982) and AR 635-100, 113-31, mandate the plaintiffs release, the plaintiff contends that he was not subject to what the defendant describes as the “mandatory release provisions” of AR 635-100, ¶ 3-31, which govern the release of Reserve Commissioned Officers from the service.
Section 681(a) of Title 10 of the United States Code grants the Secretary of the Army authority to release a reservist from active duty service: “Except as otherwise provided in this title, the Secretary may at any time release a reserve under his jurisdiction from active duty.” 10 U.S.C. § 681(a) (1982). The defendant argues that this subsection would permit the Secretary of the Army to invoke AR 635-100, 113-31a (August 1,1982) and mandatorily retire the plaintiff on October 31,1985. AR 635-100, 113-31a states, in pertinent part, that:
Commissioned officers will be mandatorily relieved from active duty upon completion of the services specified in (1) ... below____ Officers who are qualified for retirement ... may apply for voluntary retirement to be effective not later than the first day of the month following the scheduled release date specified below.
(1) Upon completion of 20 years active federal service, on the last day of the month following the month in which such service is completed.
AR 635-100, 113-31a (August 1, 1982).
The plaintiff, however, argues that 10 U.S.C. § 681(a) does not give the Secretary broad authority to release reservists. The plaintiff states that 10 U.S.C. § 1163(d) (1982), is an exception to the Secretary’s authority granted by section 681(a) and that section 681(a) does not “mandate” the release of a reserve officer. Plaintiff adds that the Secretary has never invoked section 681(a) to release the plaintiff, either in September of 1985, or at any time since. As to the defendant’s reliance on AR 635-100, 113-31a the plaintiff raises two arguments. First, the plaintiff contends that the Army could have invoked an exception to the regulation to allow him to remain on active duty beyond October 31, 1985. Second, the plaintiff cites to AR 635-100, ¶ 3-33 (July 15, 1980), in support of the notion that the Army, in order to properly invoke ¶ 3-31a, was required to provide him with six months notice prior to his mandatory release and voluntary retirement. Since, no such notice was ever provided, according to the plaintiff, not only is the court allowed to constructively change his military records to reflect military service through October 31, 1985, but the court should also conclude that the plaintiff was constructively in the Army until the date upon which he voluntarily chose to retire on March 1, 1988.
In the present case, the court finds that if the authorities had recognized, at the time of his release, that the plaintiff was eligible for inclusion under the “sanctuary” provision, the plaintiff would have been retained on active duty beyond September 30, 1983. However, the court also finds that it does not appear from the history of this case, that the plaintiff necessarily would have been retained on active service beyond the 20 year retirement mark. Furthermore, the court concludes that the defendant has not only offered to prove, but has successfully shown, that the plaintiff would have been given the opportunity to choose between voluntarily retiring at the end of October 1985, and being involuntarily released from active duty, only to request later, voluntary retirement status in accordance with the regulations.
The court concludes that, with one exception, the defendant’s calculation of the plaintiff’s retirement date, is correct. Applying section 1163(d), plaintiff had “sanctuary” from the Secretary’s invocation of section 681(a) until the plaintiff qualified for retirement pay. Once the plaintiff, a reserve officer on active duty, see 10 U.S.C. § 101(22) (1982), qualified for retirement pay under the umbrella of the “sanctuary” *704provision, 10 U.S.C. § 1163(d) (1982), there was no longer a valid exception to bar the Secretary from invoking section 681(a). Therefore, once the section 1163(d) exception had been satisfied, the Secretary was entitled to act, pursuant to section 681(a), and release the plaintiff from active duty, in accordance with AR 635-100, II 3-31a.
As to the mandatory nature of the Army Regulation in question, AR 635-100, ¶ 3-31a, the court concludes that, because the plaintiff has not properly raised the applicability of any of the enumerated exceptions to that regulation, the regulation would call for the mandatory retirement of LTC Ulmet, upon completion of 20 years of active duty. The court also concludes, that the Army should be presumed to have acted in good faith and in accordance with its regulations, including AR 635-100, 113-33 cited by the plaintiff, when processing the plaintiff’s retirement. When a court looks to correct a plaintiff’s military records, the court can only proceed in accordance with the strong presumption that the Army would have faithfully discharged its duties and the burden is upon the plaintiff to prove otherwise. See Cooper v. United States, 203 Ct.Cl. 300, 304 (1973).
The court, therefore, orders the defendant to correct the plaintiff’s military records to reflect:
(1) proper notification, pursuant to AR 635-100, ¶ 3-33, 6 months prior to the plaintiffs REFRAD date;
(2) a constructive REFRAD date of October 31, 1985; and
(3) in accordance with AR 635-100, U 3-31a, following the plaintiff’s mandatory release from active duty, the court presumes retirement pay eligibility beginning November 1, 1985.
The second main area of disagreement between the parties is over the proper manner in which to calculate the value of the intangible benefits of active duty to which plaintiff is entitled following the court’s reconstruction of plaintiff’s military records. Plaintiff declares that he should be compensated for all benefits which he might have received had he been on active duty during the constructive period of active duty, September 30, 1983 to October 31, 1985. Plaintiff argues that he is entitled to recover compensation for benefits (at the values set forth for such benefits) as enumerated on a standard form which the Department of the Army provides to its officers to show them their “real pay value.” The form is entitled, “Personal Statement of Military Compensation,” and is issued by the United States Army Finance and Accounting Center in Indianapolis, Indiana. Plaintiff states in a sworn declaration that the Army encourages officers “to use these forms, for example, when applying for mortgages, to show the ‘true value’ of Army compensation.” The introductory paragraph of this form states:
This statement is provided you at the direction of the US Congress, which has approved its format and contents. Although we are generally aware of our overall compensation, there is no one place where we can see it all together. Our annual W-2 Statement of Taxable Wages shows only basic pay, incentive pay and special pay. Neither our W-2 nor our Leave and Earning Statement (LES) shows the indirect compensation items that increase our spendable income. There are many possible ways to compute total compensation value; we have tried to show these values in a way that will mean the most to you.
Officers’ direct compensation includes basic pay, basic allowance for quarters, basic allowance for subsistence, and housing allowance. To direct compensation, this form adds a variety of items labeled indirect compensation. “The added value shown [for indirect compensation] is an estimate of the amounts you would have to pay each year for major entitlement programs if the Government did not provide them for you.” Id. These programs and privileges include medical care, death and survivor programs, retirement, commissary, exchanges, tax advantages, and morale, welfare and recreation activities. Essentially, plaintiff claims he should now receive compensation for the benefits he would have received had he actually been on active duty, regardless of any statutory *705entitlement thereto. Defendant, however, argues that, due to plaintiff’s failure to cite any statute or regulation entitling plaintiff to relief, this court lacks jurisdiction to grant the relief he now seeks.
A member of the armed services, whose military record is corrected, is entitled to be paid for “the loss of pay, allowances, compensation, emoluments, or other pecuniary benefits, or for the repayment of a fine or forfeiture.” 10 U.S.C. § 1552(c) (1982). This court has previously held that, while section 1552 entitles members of the armed forces, who have been incorrectly released, to “other pecuniary benefits,” “[t]he privilege of patronizing a military commissary or exchange is not a pecuniary benefit,” and, therefore, it is not relief that this court can grant. Lord v. United States, 2 Cl.Ct. 749, 756 (1983). The court in Lord added:
[mjoreover, the plaintiff ... has not provided the court with any data on the basis of which the commissary and exchange privilege for the period involved here could be evaluated. Thus, the claim is wholly speculative. This is another reason why plaintiff’s claim for the value of the commissary and exchange privilege during the pertinent period must be denied.
Id. (citing Dilley v. Alexander, 627 F.2d 407, 414 (D.C.Cir.1980)).
The present plaintiff offers to the court dollar figures from the “Personal Statement of Military Compensation” which supposedly represent the amount which various military privileges would have been worth to him while serving on constructive, active duty. The numbers are presented without basis of proof and without documentation regarding plaintiff’s particular situation. The court agrees with the reasoning included in the decision of the Court of Appeals for the District of Columbia Circuit in Dilley: plaintiff is entitled to reimbursement for “all pay and allowances that [he] would have received but for [his] illegal separation, to the extent that such sums can be calculated with a reasonable degree of certainty.” Dilley v. Alexander, 627 F.2d at 414. Other than the amounts for basic pay, basic allowance for quarters, basic allowance for subsistence, and variable housing allowance, the amounts claimed by plaintiff listed on the “Personal Statement of Military Compensation,” however, fall far short of this standard.
A simple reading of the instructions included with the “Personal Statement of Military Compensation” evidences the speculative nature of the values associated with each of the military benefits enumerated on the form. In this court’s view, estimates are not entitlements and are not compensable. See Lord v. United States, 2 Cl.Ct. at 756. The report of the House Committee on Appropriations, which discussed providing the form as a way to resolve the problem “that military members do indeed undervalue their regular military compensation,” did not suggest that the form should be used to calculate dollar values of military compensation for cases in litigation. H.R.Rep. No. 1317, 96th Cong., 2d Sess. 27-28 (1980). There is no indication that the form was intended to provide improperly discharged military personnel with substantive proof of indirect compensation.
Although the form serves as a means by which to inform officers and others generally of the “real pay value” of military work, this form effects no changes in the statutory scheme of compensation. See 10 U.S.C. § 1552(c) (1982). As the form itself states, regarding some of these items of indirect compensation, “[t]heir true worth will be different for each person, depending on use. You should think seriously about how much you may have saved by using these programs over the last year.” As the form acknowledges, it is not possible to determine with any degree of certainty what the actual value is of these programs and privileges.
The third area of disagreement between the parties concerns whether the defendant is entitled to deduct from any award to the plaintiff for constructive active duty time, money actually earned by the plaintiff, during that same constructive service period, as a civilian. The defendant claims it is entitled to offset from plaintiff’s recovery *706for constructive active duty pay, plaintiff’s civilian income, earned during his constructive active duty. In support of its argument, defendant cites to Craft v. United States, 589 F.2d 1057, 218 Ct.Cl. 579, 599 (1978); Diamond v. United States, 427 F.2d 1246, 192 Ct.Cl. 502 (1970); Conn v. United States, 407 F.2d 879, 187 Ct.Cl. 319 (1969); Motto v. United States, 360 F.2d 643, 175 Ct.Cl. 862 (1966); and Dilley v. Alexander, 627 F.2d 407 (D.C.Cir.1980).
Plaintiff began discussing this issue, in his October 6, 1987 filing, by claiming that “there is no outside income for the Army to deduct.” Plaintiff added that even if there was any outside income, principles of equity should prevent the defendant from deducting this income from the plaintiff’s award in this case. After persistent requests by defendant for specific information about plaintiff’s earnings, plaintiff finally acknowledged some civilian earnings during the time of plaintiff’s constructive, active duty. However, plaintiff argues, relying on Laningham v. United States, 5 Cl.Ct. 146 (1984), that defendant cannot offset plaintiff’s civilian earnings against the back pay award because any civilian earnings actually received during the period of constructive, active duty could have been earned even if he had been on active duty. In Laningham, the court held that civilian earnings by the plaintiff could not be offset against plaintiff’s back disability pay if “such income could have been earned by an employee had he been serving in the position he held before his disability.” Id. at 158.
To prove that he could have earned the money even if he had been on active duty, the present plaintiff points to the smallness of the amount which he earned, claiming that that “should by itself be sufficient proof that he could have earned that same amount while on active duty.” Under a sworn declaration attached to plaintiff’s filing dated April 26, 1988, plaintiff declares that he “had obtained part-time employment without any difficulty in the past while on active duty____ In my experience, obtaining permission to do this sort of part-time work while on active duty is and was routinely granted.”
In Silver v. United States, 551 F.2d 295, 213 Ct.Cl. 388 (1977) (cited in Laningham v. United States, 5 Cl.Ct. 146, 158 (1984)), the Court of Claims stated that:
[t]his court has uniformly held that a claimant’s outside earnings are to be deducted from an award of back pay if such earnings would not have been received by him had he been rendering the Government the employment services called for by the position for which he is found to have been improperly denied compensation.
Id. at 297, 213 Ct.Cl. at 390. (citations omitted). This is the same rule stated in the other cases cited by defendant.18 In the instant case, the defendant has failed to respond to plaintiff’s sworn statement that he could have held such jobs and would not have earned such money even while on active duty. Defendant, therefore, should not be entitled to deduct plaintiff’s civilian earnings from his back pay for constructive, active duty, provided the work activity engaged in was consistent with plaintiff’s military service under applicable statutes and regulations, and provided plaintiff submits in writing to the defendant, with copies to the court, a detailed description of outside employment during the relevant time period, specifying dollar amounts received and type of activity for which such compensation was received.
Having concluded that the plaintiff should receive constructive, active duty, basic, back pay for the period October 1,1983 through October 31,1985, the parties agree that, once the proper REFRAD date is determined, the plaintiff is entitled to: (1) a basic allowance for quarters at the with dependant's rate, pursuant to 37 U.S.C. § 403 (1982); (2) a basic allowance for subsistence, pursuant to 37 U.S.C. § 402 (1982); and (3) a variable housing allowance, pursuant to 37 U.S.C. § 403a (1982).
*707Additionally, both parties recognize plaintiffs right to accumulated annual leave for his constructive, active duty time. The parties disagree, however, on the amount of time with which plaintiff should be credited in order to receive back pay compensation for annual leave. Defendant acknowledges in its Memorandum filed April 13, 1988, that plaintiff should be allowed back pay for accrued leave, pursuant to 37 U.S.C. § 501 (1982). Defendant states, in a table, attached to its Response to the Court’s April 4, 1988 Order, that plaintiff is entitled to 38 days of accumulated annual leave, however, no explanation of the basis for that number is given. The plaintiff proposes that he should be given credit for “at least 62.5 days” of accrued leave.
While the parties agree that 37 U.S.C. § 501(b)(3) allows plaintiff payment for accrued annual leave, as noted in that subsection, “the number of days of leave for which payment is made may not exceed sixty, less the number of days for which payment was previously made under this section after February 9, 1976.” 37 U.S.C. § 501(b)(3) (1982). The number of days for which plaintiff is to be credited is calculated under 10 U.S.C. § 701(a) (1982), at a rate of 2V2 days per month, with a maximum allowance of 60 days. Since the court has not been informed of whether payment had previously been made to plaintiff under 37 U.S.C. § 501 (1982), at this time, the court requests the parties to submit to the court a final calculation of the amount due plaintiff for accumulated annual leave, in accordance with the constructive, active duty period calculated earlier and the statutory rules discussed immediately above.
Plaintiff’s retirement benefits are also dependant upon the court’s determination of plaintiff’s constructive release date. Plaintiff qualifies for retirement as a commissioned officer of the Army, who has at least 20 years of service, computed under 10 U.S.C. § 3926 (1982). Upon his retirement date, now constructively set as October 31, 1985, the plaintiff completed 20 years, 1 month and 26 days of active service. Thus, pursuant to 10 U.S.C. § 3911 (1982), plaintiff’s retired pay is calculated using Formula A of 10 U.S.C. § 3991 (1982). The parties agree that this is the appropriate formula. According to this formula, plaintiff’s retired pay is his retired pay base, multiplied by “the retired pay multiplier” prescribed in 10 U.S.C. § 1409 (1982), for the years of service credited to him under 10 U.S.C. § 1405 (1982).
Section 1405(a)(3) indicates that plaintiff’s years of service include the years of service with which he would be entitled to be credited under 10 U.S.C. § 1333 “if he were entitled to retired pay under section 1331 of this title.” 10 U.S.C. § 1405(a)(3) (1982). The plain language of the statute, as well as the legislative history of the statute, makes it clear that Congress intended that the plaintiff must qualify for retired pay under section 1331 in order to have years of service under section 1333 included in his total years of service. In explanation of the statute, a Senate committee report restated the clause, now at (a)(3): “All years of service ... with which he would be entitled to be credited for the purposes of the computation of retired pay based on [10 U.S.C. § 1333] if he were entitled to retired pay computed under that section, rather than one of the enumerated sections.”19 S.Rep. No. 1472, 85th Cong., 2d Sess., reprinted in 1958 U.S.Code Cong. & Admin.News 2465, 2489. One of those enumerated sections is Formula A, the one under which plaintiff is entitled to retired pay, presently codified at 10 U.S.C. § 3991 (1982) (Formula A).
Without offering any information to show that he is entitled to retired pay under section 1331 or even acknowledging that such a demonstration is necessary, plaintiff argues that he is entitled to include in his total years of service his years *708of service calculated under section 1333. Also without stating why, defendant responds simply that plaintiff is not entitled to retired pay under section 1331. As the case has proceeded, however, both parties seem to agree that any right to retired pay, to which the plaintiff might have been entitled would be under section 3991, rather than under section 1331. Among other requirements under section 1331, a person seeking to qualify for credit for years of service under section 1333 must be 60 years of age and must not be entitled, under any other provision of law, to retired pay from an armed force, unless he performed active duty (other than for training) after June 26, 1950, and before July 28, 1953. 10 U.S.C. § 1331(a), (c) (1982). Plaintiff does not meet these criteria and, therefore, is not entitled to be credited for years of service under section 1333.
Plaintiff claims that the retired pay calculation should use “a base year of 360 days ... not a calendar year of 365 days.” Plaintiff’s only basis for this claim is his citation to 10 U.S.C. § 1333 (1982). Section 1333 does specify that the number of days spent in various categories of service are to be added together and then divided by 360 to determine a retiree’s years of service for the purpose of computing the individual’s retired pay under the chapter on retired pay for non-regular service. However, as already stated, section 1333 does not apply to plaintiff in determining his years of service.
The final area of dispute remaining between the parties refers to reimbursement for medical expenses incurred by the plaintiff during the period of constructive, active duty and subsequent to the constructive REFRAD date.20 Plaintiff claims entitlement to compensation for various medical expenses incurred by himself and his dependents during the period of wrongful discharge. Plaintiff argues that he would have been able to receive medical care at no cost had he been properly retained in active service.
Defendant responds that plaintiff is entitled only to compensation for necessary medical expenses he incurred for his own medical care, while constructively serving on active duty, under AR 40-3, Medical, Dental, and Veterinary Care, implementing 10 U.S.C. § 1074(a) (1982), but that dependents of service members are authorized, but not entitled, to receive medical treatment at military medical facilities. 10 U.S.C. § 1076(b) (1982).
Reimbursement for medical expenses is subject to the same standard as other compensation awarded by this court. The plaintiff must demonstrate that a statute or regulation mandates ¿uch payment to the plaintiff. E.g., United States v. Tes-tan, 424 U.S. 392, 96 S.Ct. 948, 47 L.Ed.2d 114 (1976). Under 10 U.S.C. § 1074(a) (1982), members of the uniformed services who are on active duty are “entitled to medical and dental care in any facility of any uniformed service.” Id. Dependents of active service members are “entitled, upon request, to the medical and dental care prescribed by section 1077 of this title [10 U.S.C.] in facilities of the uniformed services, subject to the availability of space and facilities and the capabilities of the medical and dental staff.” 10 U.S.C. § 1076(a) (1982).
Plaintiff, therefore, is plainly entitled to compensation for his own medical expenses incurred during the period of his constructive, active duty. Compensation for the medical care of his dependents during the constructive active duty time would be available to the extent that the medical care was of the type listed in section 1077, and to the extent that the medical care would have been available under the re*709striction, “subject to. the availability of space and facilities and the capabilities of the medical and dental staff.” Id.
Also, pursuant to 10 U.S.C. § 1074(b) (1982), servicemembers entitled to retired pay “may, upon request, be given the medical and dental care prescribed by section 1077 of this title [10 U.S.C.] in facilities of the uniformed services, subject to the availability of space and facilities and the capabilities of the medical and dental staff.” 10 U.S.C. § 1076(b) (1982). As discussed by the court in Lord v. United States, 2 Cl.Ct. 749, 756-57 (1983), these provisions regarding retired personnel and their dependents fail to establish with sufficient specificity a dollar amount for which plaintiff can recover in a proceeding before this court. Plaintiff, therefore, is not entitled to compensation for the medical bills claimed for reimbursement, incurred after his constructive REFRAD. He is entitled, however, to be compensated for his own expenses during the time of his constructive, active duty, and for the medical expenses of his dependents to the extent that he can demonstrate the services would have been available to them.21 It should also be noted that any services received by plaintiffs family during plaintiff’s constructive, active duty time is subject to the fee provided for in 10 U.S.C. § 1078 (1982). This fee can appropriately be set off against, or deducted from, any recovery for such expenses.
There also are several, outstanding off-set issues in addition to the civilian earning issue raised by the defendant. Initially, the parties disagreed as to the correct amount to offset from the plaintiff’s award due to his receipt of $15,000.00 as a readjustment payment, pursuant to 10 U.S.C. § 687 (repealed 1981). Presently, the parties agree that, in accordance with 10 U.S.C. § 1174(h)(1) (1982) and Department of Defense Pay Manual, II 40422(b) (1987), the maximum amount of the readjustment payment that the defendant should be permitted to offset from the plaintiff’s recovery is 75% of $15,000.00 or $11,250.00.
The final three items for which the defendant claims he should be allowed an off-set or deduction from plaintiff’s award in this case are: (1) the plaintiff’s military earnings for any military service performed by him after the start of the period of constructive service, September 30,1983; (2) other government benefits, if any, received by the plaintiff for which, upon his constructive reinstatement to active duty and retirement, he would be required to reimburse the defendant; and (3) federal income tax withholding.
As to the first, the plaintiff’s earnings for military service performed since his improper release from active duty, the plaintiff objects to defendant’s stated intent to deduct the pay plaintiff received for an active duty tour from October 27 to November 7,1986. Plaintiff notes that this tour occurred after plaintiff’s constructive, retirement date and claims he should receive the higher, active duty payment, instead of his retired pay during that twelve day period. In accordance with the plaintiff’s reasoning, however, the defendant should be allowed to deduct from the plaintiff’s award for back, retired pay, twelve days of retired pay for which the plaintiff already has received active duty pay.
With respect to the remaining two deductions, the government benefits received by the plaintiff for which he is required to reimburse the defendant and the federal income tax withholding, the court finds in favor of the defendant. See Laningham v. United States, 5 Cl.Ct. 146, 158-59 (1984).
CONCLUSION
Consistent with the foregoing, the Defendant’s Motion to Dismiss is, hereby, DENIED and the Plaintiff’s Motion to Return to Active Duty and to Compel the Defendant to Calculate and Pay Damages is, hereby, GRANTED in part. The court’s findings, explained more fully above, are summarized as follows:
*710(a) Plaintiff is entitled to the correction of his military records to reflect active duty from September 30,1983 to October 31, 1985, and is entitled to back payment for that period of time, including basic pay, basic allowance for quarters at the with dependents rate, basic allowance for subsistence, the variable housing allowance, and annual leave pay, all at the appropriate rates applicable to the location where the plaintiff was assigned to duty prior to his improper release.
(b) Plaintiff is entitled to receive back retirement pay beginning the first day of plaintiffs constructive retirement, November 1, 1985.
(c) The plaintiff is entitled to be reimbursed for medical expenditures associated with medical services performed for the plaintiff, himself, during the constructive, active duty period. The plaintiff is not entitled to direct reimbursement from the defendant for all medical expenses associated with his spouse and dependents. The court also orders the defendant to process through the CHAM-PUS insurance service and reimburse the plaintiff for medical expenses associated with the medical bills attached to his April 26, 1988 filing, as well as for those bills for which the plaintiff serves copies upon the defendant within ten days of the issuance of this opinion, which are properly subject to reimbursement under applicable statutes, regulations and guidelines.
(d) The defendant may not set off the plaintiff's civilian earnings, which he received during the period of constructive, active duty, from the award outlined in (a), (b) and (c) above, unless such earnings were received for wo.rk inconsistent with plaintiff’s military service under applicable statutes and regulations.
(e) The defendant may set off against the gross award described in (a), (b), and (c):
(i) plaintiff’s retired pay earnings received during his constructive retirement while performing active duty service.
(ii) $11,250.00, 75% of the $15,000.00 separation pay received by the plaintiff at the time of his involuntary release from the Army in 1973;
(iii) any other debts the plaintiff owes to the defendant for government entitlements or benefits received by the plaintiff which, upon his constructive reinstatement to active duty and retirement, he would be required to reimburse the defendant, including the type of debts discussed by the Claims Court in Laningham v. United States, 5 Cl.Ct. 146, 157 (1984);
(iv) all applicable federal income taxes required to be withheld on the gross recovered amount.
IT IS ORDERED that plaintiff is entitled to judgment in an amount which, when calculated, will reflect the net of items (a) through (e) above.
Since a computation is required to determine the precise amounts to which the plaintiff is entitled, pursuant to items (a) through (e), above, the court, hereby, issues a “Call” on the defendant, addressed to the Secretary of the Army, pursuant to RUSCC 34(d)(1)(B) and 28 U.S.C. § 2507(a), to make appropriate computations in schedule form, consistent with this opinion, and to file the schedule with the court, on or before the expiration of 30 days from the date of the issuance of this Opinion.
It is further ordered that within 15 days after the clerk has served notice of the filing of the computed schedule, each party shall file with the court its acceptance or rejection of the computation. Any rejection^) shall be accompanied by a specific statement including the reasons for the rejection, along with any appropriate citations to authorities. The court anticipates and sincerely hopes that the parties to this action will strive to work together in harmony to determine the remaining dollar calculations necessary to determine the appropriate, final award.
Following the receipt of the filings, described immediately above and the final calculation of the amounts due to the plaintiff, the court will issue an Order requesting the Clerk of the United States Claims Court to enter judgment in favor of the *711plaintiff for the specific, net dollar amount consistent with this Opinion. Subsequent to the issuance of the Order, the court will schedule a status conference with the parties to discuss the multiple outstanding motions in the case, if they still remain unresolved, particularly, the plaintiff’s motions for sanctions against the defendant and, plaintiff’s motion for interest on judgment, pursuant to 28 U.S.C. § 1961(c)(2).
. Active duty for training is one of three classifications for Army personnel. The other two categories are inactive duty for training and continuous active duty. As described in Ulmet I,
[ajctive duty for training (ADT) is a tour of duty used for training members of the reserve components to provide trained units and pre-trained individuals to fill the needs of the Armed Forces in times of war or national emergency and such other times as the national security requires. ADT generally is under orders that provide for automatic return to non-active duty status when the period of training is completed. ADT includes annual training tours, special tours of active duty for training, school tours and the initial tour performed by non-prior service enlistees.
Ulmet v. United States, 10 Cl.Ct. 522, 523 (1986).
. One of the defendant’s arguments, rejected by the Federal Circuit, was based on a settled proposition that courts must accord due weight to an agency’s interpretation of a statute it administers. Ulmet v. United States, 822 F.2d 1079, 1086 (Fed.Cir.1987) (citing Zenith Radio Corp. v. *685United States, 437 U.S. 443, 450, 98 S.Ct. 2441, 2445, 57 L.Ed.2d 337 (1978); American Lamb Co. v. United States, 785 F.2d 994, 1001 (Fed.Cir.1986)). The court of appeals, quoting language from one of its own earlier opinions, rejected this portion of defendant’s argument on the basis that "a court may reject an agency interpretation that contravenes clearly discernible legislative intent." Ulmet v. United States, 822 F.2d at 1086 (quoting American Lamb Co. v. United States, 785 F.2d 994, 1001 (Fed.Cir.1986).
. A mandate is:
[a] precept or order issued upon the decision of an appeal or writ of error, directing action to be taken, or disposition to be made of case, by inferior court. Official mode of communicating judgment of appellate court to lower court, directing action to be taken or disposition to be made of cause by trial court.
Black’s Law Dictionary 867 (5th ed. 1979).
. As noted earlier, the Federal Circuit reversed the original United Claims Court decision. To reverse is "to overthrow, vacate, set aside, make void, annul, repeal, or revoke; as to reverse a judgment." Black’s Law Dictionary 1185 (5th ed. 1979).
. Act of December 30, 1987, Pub.L. No. 100-224, 101 Stat. 1536.
. The court notes that in the Motion to Dismiss, the defendant has even failed to raise the possible lack of the 8 days of active duty service. Moreover, the Federal Circuit clearly opted in Ulmet I to give the plaintiff the benefit of the doubt regarding the 8 days under equitable principles. Therefore, even if this court was inclined to utilize the unamended statute and to insist that when qualifying for a privilege such as retirement benefits that the statutory requirements be clearly met by a potential beneficiary, it would be inclined to credit the plaintiff with the 8 days.
. The following is a list of the pending related cases of which this court is aware, currently filed in the United States Claims Court: LTC James A. Green v. United States, No. 763-87C; CWO Stephen J. Weissenberger, No. 764-87C; LTC Howard A. Harvey v. United States, No. 788-87C; LTC Fred R. Rowzee, Jr. v. United States, No. 277-88C; LTC Eugene D. Minretta, Jr. v. United States, No. 278-88C; LTC Juan M. Mata v. United States, No. 501-88C; LTC Jerald D. Nelson v. United States, No. 556-88C; LTC Larry D. Jordan v. United States, No. 557-88C; MAJ David V. Olson v. United States, No. 673-88C; LTC John F. Mitchell v. United States, No. 155-89C. On May 22, 1989, Judge Merow of the United States Claims Court issued an opinion in the case of LTC Albertis Wilson v. United States, No. 484-87C.
. In Amen v. City of Dearborn, 718 F.2d 789, 793 (1983), the United States Court of Appeals for the Sixth Circuit affirmed a district court ruling to apply a jurisdictional statute which was amended after the district court’s first decision and subsequent to the appellate court's remand, 28 U.S.C. § 1331 (1982), to a pending case, without the appellate court first being requested to relieve the trial court of its obligation under the mandate. The circuit court held that the district court did not go beyond the mandate on remand because the mandate instructed the district court to make a jurisdictional determination in accordance with the statute. The court of appeals did not just reverse the district court’s jurisdictional conclusion (which had resulted in the dismissal of the complaint), but the court of appeals also remanded the case for a new jurisdictional determination. The second reason for the Sixth Circuit’s decision to permit the district court to revisit the jurisdictional question at issue was because “Congress stated unequivocally its intent ‘that [the statute] shall apply to any civil action pending in Federal court on the date of enactment. By putting congressional intent into immediate effect, this provision will eliminate on-going jurisdictional battles, thus saving valuable court time.’ ” Id.
. The Supreme Court noted the incorrect use of the term mandate in the Standard Oil decision, but continued to use it in order to avoid confusion with the district court’s opinion.
. Rule 60(b)(3), upon which the movant in Standard Oil motioned the Supreme Court for a recall and leave, states that ”[o]n motion arid upon such terms as are just, the court may relieve a party or his legal representative from a final judgment, order, or proceeding for the ... (3) fraud (whether heretofore denominated intrinsic or extrinsic), misrepresentation, or other misconduct of an adverse party.” Fed.R.Civ.P. 60(b)(3), 28 U.S.C. (1976).
. The defendant also footnotes an argument that the new evidence exception to the law of the case doctrine applies because the amended statute is substantially new evidence which would materially change the outcome of the case. This argument fails because the substantially new evidence exception appears to contemplate a situation in which a case is remanded to the lower court and new facts are developed following the remand, after which, on further appeal, the appellate court changes its original opinion based on the new evidence, see W.L. Gore & Assocs. v. Garlock, Inc., 842 F.2d 1275, 1278 (Fed.Cir.1988). This exception does not appear to foresee a situation in which the Congress amends the applicable statute after the mandate and remand are issued.
. Thorpe v. Housing Auth. of Durham, 393 U.S. 268, 89 S.Ct. 518, 21 L.Ed.2d 474 (1968); Bradley v. School Board of Richmond, 416 U.S. 696, 94 S.Ct. 2006, 40 L.Ed.2d 476 (1973); Bennett v. New Jersey, 470 U.S. 632, 105 S.Ct. 1555, 84 L.Ed.2d 572 (1984).
. Representative Byron said that the “purely technical amendment” was to "prevent a windfall never intended by a small group of reservists who may come on active duty for training and then as a result of a recent court decision, may remain in an active status until they qualify for retirement benefits,” 133 Cong.Rec. H11509-10 (daily ed. Dec. 16, 1987) (remarks of Rep. Byron).
. Senator Glenn said that the amendment "restores the original definition of 'active duty* in section 1163(d) of title 10 which excluded any active duty for training periods____ The correction restores the original intent of the section and is appropriate.” 133 Cong.Rec. S18344 (daily ed. Dec. 17, 1987) (remarks of Sen. Glenn).
. In opposition to the defendant's motion, the plaintiff also argues that his classification by the Army as being on "active duty for training” was ambiguous and incorrect and that he was, in fact, not really in "training” at all, but rather, he was actually on active duty "designing training systems for others.” Having fully considered the defendant's arguments outlined in its Motion to Dismiss, and finding that each argument does not lead to the conclusions which the Department of the Army offers, namely that application of the 1987 amended version of 10 U.S.C. § 1163(d) is appropriate in this case, the court does not deem it necessary to enter the maze of classification issues which this argument raises. See Wilson v. United States, 16 Cl.Ct. 765, n. 2 (1989).
. In a footnote, the plaintiff contends that: The Army might argue that it would never have exercised its ability to retain [him], but this allegation is self-serving and based on an impermissible rationale for making such an administrative decision, i.e., it would be arbitrary, capricious, and an abuse of discretion for the Army to deny [him] the opportunity to serve on active duty simply as a means to retaliate against him for exercising his legal rights.
. In accordance with the defendant’s claimed invocation of AR 635-100, the plaintiff was enti*703tled to remain on active duty until October 31, 1985, the end of the month following the month in which he completed 20 years, on which day he would have been mandatorily released from active duty.
. The parties’ references to 5 U.S.C. §§ 5533, 5534 (1982) are irrelevant inasmuch as those statutes only concern dual employment and compensation within the government, and not civilian employment outside of government, concurrent with military service.
. The Military Retirement Act of 1986, Pub.L. No. 99-348, § 106, 100 Stat. 682, 691, amended 10 U.S.C. § 1405 by replacing the list of enumerated sections with the general phrase, "a provision of this title providing for such computation [of the years of service of a member of the armed forces] to be made under this section.” When amended in 1986, section 3991 was one of the enumerated sections.
. The court notes that in the plaintiffs October 26, 1987 filing, he requests a constructive, retroactive promotion to the rank of Colonel with back pay commensurate. Since that filing, neither party has addressed this issue and the plaintiff, in particular, has failed to cite any legal or factual basis for this retroactive promotion. The Court of Claims “repeatedly made it clear that deciding who gets a promotion is not [its] responsibility, absent proof of a legal right to it which has been violated." Sanders v. United States, 594 F.2d 804, 816, 219 Ct.Cl. 285, 307 (1979) (citations omitted). Therefore, plaintiff is denied any retroactive promotion, which he merely alleges would have occurred prior to becoming eligible for retirement.
. The court also notes the defendant’s suggestion that plaintiff may be able to apply for reimbursement for some of the other incurred medical expenses from the Civilian Health and Medical Program of the Uniformed Services (CHAMPUS).