OPINION
MAYER, Circuit Judge.
This is a rehearing in banc of an appeal from the United States District Court for the Western District of Louisiana, No. 86-1080 (Oct. 8, 1986), granting the government’s motion for summary judgment and dismissing, because of laches, Ronald J. Cornetta’s claim for reinstatement and back pay stemming from his allegedly unlawful discharge from the United States Marine Corps. We vacate our earlier judgment and opinion of the panel, 831 F.2d 1039 (1987), and reverse the judgment of the district court.
I. Background
The district court granted summary judgment on the allegations set out in the complaint, so the facts are not now in dispute. After graduation from the United States Naval Academy, Cornetta was commissioned a second lieutenant in the United States Marine Corps. In his early assignments, he was described by his rating officers as an officer with outstanding growth potential, and in September 1968 he was promoted to the rank of first lieutenant. In 1970, he volunteered for duty in the Republic of Vietnam where he flew combat support missions as a helicopter copilot. On four occasions, he participated in the rescue of wounded and stranded soldiers under hazardous flight conditions and hostile fire for which he was decorated in recognition of his superior airmanship and devotion to duty.
Cornetta was promoted to captain shortly after he returned from Vietnam. For the next six years, his superiors consistently reported that he was qualified for promotion and rated him “excellent,” “excellent-to-outstanding,” or “outstanding” in terms of his growth potential and overall value to the Marine Corps. He was awarded the Navy Achievement Medal for 1974, and in 1975 he was named “Training Air Wing Helicopter Instructor of the Year.”
In 1977, Cornetta was assigned to a logistic support unit near San Francisco. His rating officer in this assignment gave him an unfavorable Officer Fitness Report (OFR) covering less than a two-month period, which recognized his limited background and experience in logistics and that this was the primary reason for the low rating. The report included the remark that “within the context of the assignment to the Logistic Support Unit” he was not qualified for promotion. According to Cor-netta, “This unfavorable report was prepared, processed and incorporated into [his] naval records in a manner that violated substantive and procedural regulations governing Marine Corps Officer Fitness Reports.... ”
At his next assignment, Cornetta’s rating officer reported that he was an officer of “excellent” value to the Marine Corps and judged him to be qualified for promotion. Shortly thereafter, however, he was passed over for promotion to major. The 1977 San Francisco OFR was part of the file the selection board considered in passing him over.
Cornetta appealed to the Board for the Correction of Naval Records for removal of the 1977 OFR and reversal of the nonselection for promotion. See 10 U.S.C. § 1552. The board found that the presence of the 1977 OFR constituted an injustice and that its inclusion in his record “may have substantially prejudiced his promotional opportunity.” In accordance with the board’s recommendation, the Secretary of the Navy ordered its removal. Instead of setting Cornetta’s nonselection for promotion aside, the Secretary ordered that his record be presented to the next selection board as if he were an officer in the promotion zone for the first time. If the second selection [1375]*1375board did not select him for promotion, however, his first passover would be counted for purposes of mandatory separation from the service. See 10 U.S.C. §§ 627, 632.
Soon thereafter, Cornetta was passed over for promotion again and was considered to have been twice passed over. Accordingly, on May 31, 1979, he was honorably discharged. Following his separation from the Marine Corps, he served three years on active duty as a lieutenant j.g. in the United States Coast Guard Reserve.
On May 20, 1986, Cornetta filed this suit in the district court alleging that his discharge from the Marine Corps was unlawful. To bring his case within the district court’s jurisdiction under the Little Tucker Act, he limited his claim for back pay to $10,000. See 28 U.S.C. § 1346(a)(2). With his complaint, Cornetta served a discovery request on the government seeking production of records and documents which purportedly would show that his delay in filing suit was “neither unjustified nor unreasonable” and that the government would suffer no prejudice if the suit were to proceed.
The government did not produce the documents; instead it filed a motion for summary judgment because of laches. In response, Cornetta said the motion was premature because the materials he sought to discover were relevant to both the excuse for his delay in filing suit and whether the government was prejudiced by the delay. Accordingly, he moved to compel discovery and stay proceedings on the laches motion until the requested information was produced.
Cornetta separately argued that the government had not been prejudiced by his delay. He said the government had suffered no detriment to its ability to defend against his claim because it was based on “a legal error patent on the face of the administrative record.” Moreover, he claimed that the government could not establish economic prejudice because “the potential for back pay recovery was fixed at an insubstantial amount by virtue of the jurisdictional limit of the Little Tucker Act, as well as by substantial offsets for Coast Guard earnings, private sector earnings and severance payments, the exact amounts of which would be established by the pending discovery request.” Finally, Cornetta argued that the government’s attempt to establish prejudice by virtue of his potential receipt of retired pay was unsustainable as a matter of law.
The district court granted the government’s motion for summary judgment and dismissed the case. It concluded that Cor-netta’s delay of nearly seven years from the date of his discharge was unreasonable, and that the government had been prejudiced by the delay. The conclusion of prejudice was not bottomed on defense prejudice, such as loss of records or fading memories, but on the potential recovery of $10,000 in back pay “for work he has not performed and which possibly was performed by another officer,” and increased retirement benefits should Cornetta prevail.
Cornetta appealed here. Before a panel of this court, the government conceded that it did not claim defense prejudice, but argued that the potential recovery of $10,000 in back pay and increased retirement benefits was sufficient to establish economic prejudice. The panel affirmed the judgment of the district court.
We accepted Cornetta’s suggestion for rehearing in banc to reconsider the scope of the laches defense in military pay cases, and the novel idea that potential receipt of increased retired pay is prejudicial to the government. 837 F.2d 473 (1988).
II. Discussion
A. Development of Laches in Military Cases
The doctrine of laches emerged in an era when equity courts were not bound by statutes of limitations. 2 J. Pomeroy, Equity Jurisprudence §§ 418-19 (5th ed. 1941). It was premised on the maxim vigilantibus non dormientibus aequitas subvenit, equity aids the vigilant not those who slumber on their rights:
[1376]*1376The doctrine of laches is based upon grounds of public policy, which require for the peace of society the discouragement of stale demands. And where the difficulty of doing entire justice by reason of the death of the principal witness or witnesses, or from the original transactions having become obscured by time, is attributable to gross negligence or deliberate delay, a court of equity will not aid a party whose application is thus destitute of conscience, good faith and reasonable diligence.
Mackall v. Casilear, 137 U.S. 556, 566, 11 S.Ct. 178, 181, 34 L.Ed. 776 (1890); see also Hayward v. National Bank, 96 U.S. (6 Otto) 611, 617, 24 L.Ed. 855 (1878).
Because laches is an equitable defense, it has traditionally been unavailable in actions at law brought within the applicable statute of limitations. See, e.g., Clark v. Amoco Prod. Co., 794 F.2d 967, 971 (5th Cir.1986). Based on the twin goals of limiting monetary consequences to the government and compelling the speedy resolutions of disputes, however, it has been applied to claims for back pay by government personnel brought even before the limitations period has run, notwithstanding that these are actions at law, not equity. See Henry v. United States, 153 F.Supp. 285, 139 Ct.Cl. 362, 366-70 (1957).
The Supreme Court has never spoken to laches in the context of the dismissal of a military officer. It did, however, consider the doctrine appertaining to the dismissal of a civilian government employee in Arant v. Lane, 249 U.S. 367, 371, 39 S.Ct. 293, 294, 63 L.Ed. 650 (1919). There, an employee wrongfully discharged sought a writ of mandamus restoring him to his position. The Court said that mandamus was “generally regarded as not embraced within statutes of limitation applicable to ordinary actions, but as subject to the equitable doctrine of laches.” Id. So a public official must promptly take action challenging his discharge “to the end that if his contention be justified the Government service may be disturbed as little as possible and that two salaries shall not be paid for a single service.” Id. at 372, 39 S.Ct. at 294. Having waited more than twenty months to file his claim, during which another was appointed to the position, the Court held Arant’s petition for mandamus barred by laches.
Two years later, the Court held the claims of two customs inspectors for the pay of their former offices because of improper discharge barred by their delay in filing suit. Nicholas v. United States, 257 U.S. 71, 42 S.Ct. 7, 66 L.Ed. 133 (1921); Norris v. United States, 257 U.S. 77, 42 S.Ct. 9, 66 L.Ed. 136 (1921); see also Stager v. United States, 57 Ct.Cl. 116 (1922), aff'd mem., 262 U.S. 728, 43 S.Ct. 519, 67 L.Ed. 1203 (1923). Strictly speaking, these cases were not premised on laches, however, but on the theory that the employees had effectively abandoned their offices by their delay in bringing suit. The Court said Nicholas had by his “lack of diligence [for three years] evidenced an abandonment of his title to the office, and of his right to recover the emoluments thereof.” 257 U.S. at 76-77, 42 S.Ct. at 9. In the companion case, the Court similarly ruled that Norris had lost his right to recover the office because “during the period of eleven months after his suspension ... [he] took no steps to vindicate his right to the office, nor to recover the compensation incident to the same.” 257 U.S. at 80, 42 S.Ct. at 11. As in Arant, the earlier mandamus and laches case which the Court cited, in Nicholas a replacement had been appointed, and in Norris the position had been abolished, another reason Norris could not prevail. But, again, these were civilians, not military; there are no military laches cases from the Supreme Court.
We therefore decline the government’s invitation first presented to us at this in banc to hold on the authority of these cases that a military officer who does not diligently pursue his rights abandons “title” to his office and is barred from judicial relief regardless of prejudice to the government. In the first place, prejudice was apparent in those eases. And we doubt that if ever confronted with this question the Court would have viewed the soldier as equivalent to the civilian worker. For it has said, “ ‘By enlistment the citizen becomes a sol[1377]*1377dier.... He acquires a new status, with correlative rights and duties; and although he may violate his contract obligations, his status as a soldier is unchanged.’ ” Bell v. United States, 366 U.S. 393, 402, 81 S.Ct. 1230, 1235, 6 L.Ed.2d 365 (1961) (quoting In re Grimley, 137 U.S. 147, 152, 11 S.Ct. 54, 55 34 L.Ed. 636 (1890)); see also Parker v. Levy, 417 U.S. 733, 743, 94 S.Ct. 2547, 2550-51, 41 L.Ed.2d 439 (1974); Schlesinger v. Councilman, 420 U.S. 738, 757, 95 S.Ct. 1300, 1312-13, 43 L.Ed.2d 591 (1975). Indeed, military officers hold presidential commissions with consent of the Senate, see 10 U.S.C. § 531; these early Supreme Court cases did not contemplate that an officer could “abandon” his commission by mere delay in reclaiming it.
Even in civilian cases, the Supreme Court might have a different view today. Though the Court has not spoken in recent years on the question of laches in civilian employee disputes with the government, it has discussed the requirements in disputes between private employers and employees. Albemarle Paper Co. v. Moody, 422 U.S. 405, 424, 95 S.Ct. 2362, 2374-75, 45 L.Ed.2d 280 (1975), reviewed a claim that Title VII plaintiffs were not entitled to back pay because they had delayed five years in bringing their case. The Court required a showing of specific prejudice before laches barred their claims.
The earliest military cases in the Court of Claims, whose precedents are binding on us unless changed in banc, South Corp. v. United States, 690 F.2d 1368, 1370 (Fed.Cir.1982), apparently did not recognize the laches defense. In fact, Todd v. United States, Ct.Cl.Rep. No. 1 (Serial Set 871) (Mar. 6, 1856), recommended the plaintiff be paid for work he did as a Navy purser during the War of 1812. The court first applied laches to the military in the 1920’s, but then in only a couple of aberrational cases. See Plunkett v. United States, 58 Ct.Cl. 359, 370 (1923); Chamberlain v. United States, 66 Ct.Cl. 317, 320 (1928). After Chamberlain, the concept of laches in military cases lay dormant for over forty years, although it continued in the civilian context, but with a requirement that prejudice to the government be shown. See, e.g., O’Brien v. United States, 148 Ct.Cl. 1, 4 (1960); Henry, 139 Ct.Cl. at 369.
With Cason v. United States, 461 F.2d 784, 788, 198 Ct.Cl. 650 (1972) (Cason I), the court resurrected laches in military suits. Two things are notable about Cason I, and its sequel Cason v. United States, 471 F.2d 1225, 200 Ct.Cl. 424 (1973) (Cason II). First, they unreservedly say that in military cases laches cannot be invoked without a showing of prejudice, apart from payments for work not done, resulting from a claimant’s delay. In Cason I, the case was dismissed on the basis of laches. But on reconsideration, the court held that laches did not bar the claim because the government had not adequately shown prejudice. According to the court, “We suggested certain areas of possible prejudice in our initial opinion in this case, but the defendant has made no attempt to substantiate that it was in fact in any specific manner prejudiced by plaintiff’s delay.” Cason II, 471 F.2d at 1229.
Second, the Cason cases recognized differences between civilian and military service that “must be taken account of” when applying the laches doctrine. Id.; see 461 F.2d at 789 (Davis, J., dissenting).
In the wake of Cason II, however, the laches defense has gradually engulfed the early cases. The government, by far the stronger party, now enjoys a presumption of prejudice when delay is thought to be too long, which military claimants have little ability to rebut. The object of the suits, the remedy of back pay, has become the focus of the prejudice inquiry, and this case, holding the potential receipt of retired pay to constitute prejudice, sets out an element of prejudice likely to exist in virtually every case. We think it is time to revisit the laches doctrine.
B. Delay
We begin with the first requirement of the laches defense: delay by the claimant. Of course, the mere fact that time has elapsed from the date a cause of action first accrued is not sufficient to bar [1378]*1378suit; the delay must be unreasonable and unexcused. See, e.g., Gava v. United States, 225 Ct.Cl. 676, 679 n. 2 (1980). Laches must be applied “ ‘apart [from] and irrespective of' ” the statute of limitations. Pepper v. United States, 794 F.2d 1571, 1573 (Fed.Cir.1986) (quoting Brundage v. United States, 504 F.2d 1382, 1384, 205 Ct.Cl. 502 (1974)). So, for example, while the Soldiers’ and Sailors’ Civil Relief Act, 50 U.S.C.App. § 525, tolls the running of the limitations period it does not suspend time to be considered in the laches calculation. Deering v. United States, 620 F.2d 242, 245, 223 Ct.Cl. 342 (1980). Here, Cor-netta’s post-discharge service in the Coast Guard stopped the running of the six-year statute of limitations, but it did not affect the period to be considered in determining whether laches affects his claim.
Cornetta waited nearly seven years from the time he was separated from the Marine Corps to file suit. He says his unanswered discovery requests go to the justification for the delay, so, absent a record, we cannot say now if the district court erred in its summary conclusion that the delay was unreasonable. This is open for consideration on remand.
C. Prejudice
“Of course, delay alone does not constitute laches.” Brundage, 504 F.2d at 1386; see Gutierrez v. Waterman S.S. Corp., 373 U.S. 206, 215, 83 S.Ct. 1185, 1191, 10 L.Ed. 2d 297 (1963) (“test of laches is prejudice to the other party”); Gardner v. Panama R.R. Co., 342 U.S. 29, 31, 72 S.Ct. 12, 13, 96 L.Ed. 31 (1951) (“where no prejudice to the defendant has ensued from the mere passage of time, there should be no bar to relief”). Even lengthy delay does not eliminate the prejudice prong of the laches test. See, e.g., Tyler v. United States, 600 F.2d 786, 789, 200 Ct.Cl. 387 (1979).
There are two types of prejudice that may stem from delay in filing suit. First, the government may be unable to mount a defense. See, e.g., McGahey v. United States, 553 F.2d 105 [Table], 213 Ct.Cl. 717, 719 (1977). “Defense prejudice” may include loss of records, destruction of evidence, fading memories, or unavailability of witnesses. This is not an issue here in view of the government’s concession. The second type, economic prejudice, centers on consequences, primarily monetary, to the government should the claimant prevail. See, e.g., Bailey v. United States, 171 F.Supp. 281, 144 Ct.Cl. 720, 722 (1959). The government here relies solely on economic prejudice premised on: (1) a presumption of prejudice and the hypothesis that a replacement may have been paid for work Cornetta would have done absent his discharge; (2) the liability to pay Cornetta back pay should he prevail; and (3) potential liability for increased retired pay.
D. Presumption of Prejudice
Notwithstanding any intimations in our earlier cases to the contrary, see, e.g., Pepper, 794 F.2d at 1575; Deering, 620 F.2d at 246; Brundage, 504 F.2d at 1386, we reject the notion that the government can rely on a presumption of prejudice, or shift the burden to plaintiff to show lack of prejudice if delay is long, to support the affirmative defense of laches. The idea that prejudice could be presumed did not take hold until the 1960’s. Gersten v. United States, 364 F.2d 850, 852, 176 Ct.Cl. 633 (1966), a civilian case, stated that “the longer the delay the less need is there to show, or search for, specific prejudice, and the greater the shift to the plaintiff of the task of demonstrating lack of prejudice.” So stated, the presumption of prejudice gets stronger as the delay gets longer, a sort of “sliding” presumption. It is difficult to conceptualize and equally difficult to apply.
When Cason II brought laches to the military, this presumption was also mentioned, though not applied. 471 F.2d at 1229. Indeed, in nearly all cases in which the court has said that prejudice may be presumed in suits for military pay, it has, in fact, found substantial actual prejudice. See, e.g., Brundage, 504 F.2d at 1387; Deering, 620 F.2d at 246. While frequently stated, presumption of prejudice has been infrequently applied. But see Pepper, 794 F.2d at 1575. And rightfully so, [1379]*1379because exalting the elapsed time over the prejudice component comes perilously close to making laches a one dimensional defense, contrary to its very raison d’etre.
The presumption has been premised largely on the theory that if a claimant prevails the government will be required “to pay two salaries, over a long period of time, when it has received the services of only one person.” Brundage, 504 F.2d at 1386. In military cases, however, this “second-man” theory lacks an adequate factual predicate. Unlike civilian employees, military officers are part of a pool, not an individualized structure; they hold commissions but are assigned duties as needed. A fired civilian employee can be replaced directly from the work force; in the military there is little if any lateral hiring. Military vacancies are filled, if filled they are, through a process that begins with recruitment and progresses through training and subsequent promotions. When a service member is discharged, officers already on board are assigned to perform his duties; the presumed replacement is fictional. Civilian workers have positions with prescribed duties; military officers do what they are told to do because of the needs of the service. It does not generally follow that the military must pay a “second salary” when it discharges an officer.
Nor is it necessarily true that the military will be forced to recruit another to “replace” the one discharged. Military manpower planning is based on varied and varying factors like “force changes, available inventory, accession and separation predictions, fiscal constraints, manpower ceilings, etc.,” and looks to long-term projected needs. Department of Defense, Manpower Requirements Report — 1988 1-5 (1987) {Manpower Report). The contention that the discharge of a particular officer will cause the military to recruit a new one is tenuous at best.
Another problem with the “second-man” theory is that it is grounded on the assumption that vacancies in the military will promptly be filled. Cf. Pepper, 794 F.2d at 1575 (prejudice stemming from the payment of two salaries is presumed without any showing that they would be paid). The assumption clearly fails, however, when an officer is discharged during a general reduction in military strength. Cf. Deering, 620 F.2d at 246. And even if the service is increasing its overall manpower, the number of officers at a discharged officer’s rank or function might be declining. Indeed, the service may simply decide that any vacancy created by a departing officer need not be filled. As the Manpower Report explains, “During peacetime it is neither necessary nor desirable to fill all the positions in all the units.” Accordingly, the “second-man” theory is an inadequate basis on which to ground a presumption of prejudice. If an actual replacement is procured from outside the service, perhaps, for example, in a medical or scientific specialty, that may be a different story. But the government will need to prove it; no presumption will be entertained.1
The general rule is that laches is an equitable doctrine that must be considered in light of the facts of each case. See Bott v. Four Star Corp., 807 F.2d 1567, 1576, 1 USPQ2d 1210, 1216 (Fed.Cir.1986). “Laches ... differs from the statute of limitations in that it offers the courts more flexibility, eschewing mechanical rules.” Waddell v. Small Tube Prods., Inc., 799 F.2d 69, 79 (3d Cir.1986); see also Holmberg v. Armbrecht, 327 U.S. 392, 396, 66 S.Ct. 582, 584, 90 L.Ed. 743 (1946). To the extent that prejudice is presumed, a court loses [1380]*1380the flexibility to consider the facts of each case and to weigh the equities on both sides.
Importantly, too, under Federal Rule of Civil Procedure 8(c), laches is an affirmative defense. See also Foster v. United States, 733 F.2d 88, 90 (Fed. Cir.1984). Accordingly, the burden of proving prejudice rests with the defendant. We see no reason why the burden of proving the absence of prejudice should be visited on the plaintiff through the device of a presumption. Cf. Bott, 807 F.2d at 1575, 1 USPQ2d at 1216.
Presumptions depend on considerations of fairness and public policy. See Panduit Corp. v. All States Plastic Mfg. Co., 744 F.2d 1564, 1581, 223 USPQ 465, 476 (Fed. Cir.1984). In the military pay arena, however, those considerations militate against a presumption of prejudice. In the usual case, the government is both the stronger party and the party best able to set out facts relevant to the prejudice inquiry. The claimant is at a decided disadvantage. It is simply inequitable, and it is equity we are talking about, to force the military claimant before even reaching the merits of the case, to rebut the presumption and then disprove actual prejudice, when the government holds the evidence.
The inequity is sharply illustrated in the present case. Citing the presumption, the district court granted summary judgment on the face of the complaint. Cornetta was denied access to the documents he claimed were relevant to his delay and to government prejudice. Without access to the information, he had no opportunity to show that the government was not prejudiced. So the government had its cake and ate it too; it raised the presumption and denied Cornetta any information that might rebut it.
We conclude, therefore, that a presumption of prejudice is unsupportable. If the government invokes the affirmative defense of laches, it has the burden to show that it was prejudiced by a claimant’s tardiness in filing suit. Prejudice may not be presumed from the length of a claimant’s delay.2
E. Back Pay
1.
We likewise do not subscribe to the government’s contention that Cornetta’s potential receipt of back pay if he is successful on the merits is sufficient to support a laches bar. Earlier cases have relied on this to establish prejudice, see, e.g., Deering, 620 F.2d at 246; Brundage, 504 F.2d at 1387, but we conclude it is unsound. See Goodman, 606 F.2d at 808 (rejecting claim “that the cost of litigation or the payment of lost wages by itself could constitute prejudice within the contemplation of the laches defense”); Powell v. Zuckert, 366 F.2d 634, 638 (D.C.Cir.1966).
Equating back pay with prejudice makes a subject of a plaintiff’s claim, his request for monetary relief, the focus of the prejudice inquiry. The greater the damages sustained, the less likely the ability to be heard. Accordingly, he is put in the unseemly position of “minimizing” his injury just to get his case before the court. Like a sailor caught between Scylla and Charybdis, a discharged officer is forced to choose between substantially limiting his potential recovery or giving up the right to have his claim heard at all. But “ ‘[t]he general rule is that when a wrong has been done, and the law gives a remedy, the compensation shall be equal to the injury.’ ” Albemarle Paper Co., 422 U.S. at 418, 95 S.Ct. at 2372 (quoting Wicker v. Hoppock, 73 (6 Wall.) U.S. 94, 99, 18 L.Ed. 752 (1867)).
[1381]*1381If back pay constitutes prejudice then virtually every suit could be said to be presumptively “prejudicial” because most successful military claimants receive back pay. Cason II, 471 F.2d at 1229. Delay then becomes the only relevant factor, and delays well short of six years have been held sufficient to satisfy the unreasonable delay prong of the laches test. E.g., Brundage, 504 F.2d at 1382 (three years and eight months). Congress, however, set out a six-year limitations period in 28 U.S. C. § 2401. If the potential receipt of back pay is deemed to satisfy the prejudice prong of the laches test, we will have judicially created an unpredictable, free floating, de facto statute of limitations, su-perceding the six-year one mandated by Congress. That is not the role of the judiciary.
Congress must be credited with appreciating the consequences of back pay claims when it passed section 2401 and the Soldiers’ and Sailors’ Civil Relief Act, 50 U.S. C.App. § 525. See Boone v. Lightner, 319 U.S. 561, 575, 63 S.Ct. 1223, 1231, 87 L.Ed. 1587 (1943). These statutes reflect congressional judgment about when claims will be too “stale” to pursue against the government, and how much money to permit a claimant to recover; they define the remedy and specify the time within which to pursue it. Courts should be cautious in intruding on this legislative judgment.
Having given its consent to be sued, the government must reimburse claimants for damages they have sustained. Relying on back pay for prejudice reverses the equation; it is the government’s wrongful discharge of an officer that occasions the payment in the first place.
Using this as a measure of prejudice is also a waste of judicial resources. Instead of devoting time to the merits of a claim, courts must spend substantial time determining the salary, allowances and other benefits to which a claimant would be entitled should he prevail. Because money earned following discharge and items such as severance pay must be offset against the potential recovery, see, e.g., Cason II, 471 F.2d at 1229; Chappelle v. United States, 168 Ct.Cl. 362, 366 (1964), a court must first ascertain and then subtract post-discharge earnings from the total potential recovery. Only then does it engage in the calculus of whether the resulting sum is sufficiently prejudicial, with erratic results. Compare Deering, 620 F.2d at 246 (potential receipt of back pay over six years sufficient to establish prejudice) with Chap-pelle, 168 Ct.Cl. at 366 (no prejudice where claim for back pay was limited to seventeen-month period), Yerxa v. United States, 11 Cl.Ct. 110, 123 (1986) (no prejudice where claim was limited to $27,310.46), aff'd mem., 824 F.2d 978 (Fed.Cir.1987), and Park v. United States, 10 Cl.Ct. 790, 793 (1986) (approximately $19,000 in back pay does not constitute “significant monetary prejudice”). Incongruously, considerable resources of the litigants and court must be spent on the question, essentially, of damages before liability is even considered, a question which will not even have to be reached if the case fails on the merits.
It is more in keeping with the judicial role to reach the merits and give the litigant his day in court than to first spend untold hours deciding how much he might recover should he prevail. Cf. Hill v. W. Bruns & Co., 498 F.2d 565, 568-69 (2d Cir.1974) (“We have come a long way since the days of Baron Parke, who, it will be recalled, was proud of the fact he had decided all cases on procedural grounds”). We therefore hold that potential receipt of back pay alone is insufficient to show prejudice to the government and to that extent overrule earlier cases like Brundage and Grisham v. United States, 392 F.2d 980, 983, 183 Ct.Cl. 657 (1968), to the contrary.
2.
Even under the cases on the books, potential awards of back pay had to be substantial before they could constitute economic prejudice. See Deering, 620 F.2d at 246 (prejudice established where claim for compensation covers a “substantial period” and “would significantly increase the Government’s liability to pay for unren-dered services”); Brundage, 504 F.2d at 1386 (prejudice can be established by pay[1382]*1382ment of two salaries “over a long period of time”). The $10,000 in back pay sought by Cornetta corresponds to only three months and one week of active duty pay and allowances. This is not significant because it is an amount that inherently attends the litigation process. If he had filed his claim at the first opportunity, the day after he was discharged, there is no doubt it would have taken more than three months for any court or board to dispose of his case, not to mention the time for any appeal. At least this much back pay results from any successful case of this sort. Because the government’s maximum liability here is fixed at three months’ pay, however, Cor-netta’s delay in filing did not increase its potential liability.
Earlier cases recognized this. See Carter v. United States, 213 Ct.Cl. 727, 728 (1977) (no prejudice where plaintiff limited his claim for back pay to a period of less than one year); Chappelle, 168 Ct.Cl. at 366 (no prejudice where a claim for back pay, filed five and one-half years late, was limited to a seventeen-month period because government would be liable for no more in back pay than if claim had been timely). Because no claimant can avoid it, the potential recovery of an amount of back pay equal to the amount that would accrue for the time it would take to establish entitlement is not an element of prejudice to the government no matter when suit is brought.
F. Retired Pay
Likewise, the potential receipt of increased retired pay is insufficient to establish economic prejudice. Retired pay is reduced pay for reduced current services. Hotinsky v. United States, 292 F.2d 508, 510, 154 Ct.Cl. 443 (1961) (an officer draws retired pay because he is “still an officer” in service of country); Lemly v. United States, 75 F.Supp. 248, 249, 109 Ct.Cl. 760 (1948) (an officer receiving retired pay “is still an officer in the service of his country even though on the retired list”); see United States v. Tafoya, 803 F.2d 140, 142 (5th Cir.1986) (retired pay is “ ‘current pay’ designed in part to compensate [an officer] for his continuing readiness to return to duty should his country have need to call upon him”); Costello v. United States, 587 F.2d 424, 427 (9th Cir. 1978) (“retirement pay does not differ from active duty pay in its character as pay for continuing military service”); Berkey v. United States, 361 F.2d 983, 987 n. 9, 176 Ct.Cl. 1 (1966) (retired pay “has generally not been considered a pension, grant, or gratuity, but as something the serviceman earns and has earned”). According to the Supreme Court,
[Military retired pay differs in some significant respects from a typical pension or retirement plan. The retired officer remains a member of the Army, see United States v. Tyler, 105 U.S. [(15 Otto)] 244 [26 L.Ed. 985] (1882), and continues to be subject to the Uniform Code of Military Justice.... In addition, he may forfeit all or part of his retired pay if he engages in certain activities. Finally, the retired officer remains subject to recall to active duty ... “at any time.” ... These factors have led several courts, including this one, to conclude that military retired pay is reduced compensation for reduced current services.
McCarty v. McCarty, 453 U.S. 210, 221-22, 101 S.Ct. 2728, 2735-36, 69 L.Ed.2d 589 (1981) (emphasis added);3 see also Tyler, 105 U.S. (15 OHO) at 245.
Unlike an award of back pay where the government in effect pays for unperformed service, receipt of retired pay by an officer means the government will pay for the reduced service it then receives. This is no detriment; the government gets what it pays for.
Moreover, using potential retired pay as a basis for finding prejudice ignores the fact that officers will earn no more as retirees if they prevail on the merits then [1383]*1383they would have earned had they been permitted an uninterrupted military career. Accordingly, the payment of retired pay simply places an officer in the same position he would have been in had he not been wrongfully discharged. The delay does not put the government in any different position than if the claimant had filed suit immediately.
Even if there is a slight aura of a pension about retired pay, it is entirely speculative whether or not a service member will retire or will continue on active duty even after he is eligible for retirement. More than that, a service member may die, be separated from the service, or voluntarily leave the service before he is eligible for retired pay. Indeed, even after retirement the amount of pay a retiree will receive, if any, is uncertain, depending on longevity and even the type of post-retirement work pursued. See, e.g., McCarty, 453 U.S. at 222 & n. 14, 101 S.Ct. 2728, 2736, & n. 14. Prejudice should not be predicated on benefits that are not certain to accrue, that are in a real sense, speculative. Cf. American Western Corp. v. United States, 730 F.2d 1486, 1489 (Fed.Cir.1984).
A court will have to guess about the future if it relies on retired pay to establish prejudice. And since retired pay is tied to both length of service and rank while serving, see 10 U.S.C. §§ 3911, 3929, 3991, virtually every successful claim for reinstatement and back pay may have an effect on an officer’s potential retired pay. Cf. Moody v. United States, 650 F.2d 290 [Table], 222 Ct.Cl. 636, 639 (1980). Accordingly, potential receipt of retired pay is not prejudicial.
G.
We are not sanguine about the continued vitality of the laches defense in these military cases, particularly within the six years from the accrual of a cause of action normally allowed by the statute of limitations. But we hesitate to be more comprehensive in this case because the record is insubstantial in significant aspects. Specifically, as observed above, this case does not involve defense prejudice, so we cannot confidently decide whether or to what extent the government’s handicap in defending a suit because of a claimant’s delay is sufficient to bar suit. Similarly, there is no record of nonmonetary, perhaps reliance based, prejudice, e.g., note 1, supra, which might be factored into the laches calculus. And, indeed, this case is not one that was brought within the normal six-year period. All of this is open for consideration on an appropriate record, and nothing we have said in this opinion should be construed as suggesting our view on these issues.
III. Conclusion
We vacate our earlier judgment and opinion of the panel, reverse the judgment of the district court, and remand for further proceedings consistent with this opinion.
REVERSED AND REMANDED.