PLAGER, Circuit Judge.
This appeal is another chapter in the long-fought battle to determine responsibility for injuries sustained over the years by individuals working with asbestos. In these particular cases, plaintiff companies seek indemnification from the United States Government for the companies’ liabilities to shipyard workers for injuries caused by exposure to asbestos. Before us is an appeal from an order of the United States Claims Court (Nettesheim, J.) entered June 1, 1989, and reported as Keene Corp. v. United States, 11 Cl.Ct. 146 (1989). The United States moved in the trial court to dismiss eight suits1 brought by Keene Corporation, Eagle-Picher Industries, UNR Industries, Fibreboard Corporation, H.K. Porter Company, Inc., Raymark Industries, Inc. and GAF Corporation.2 The trial court granted the motion as to all plaintiffs except GAF Corporation.3 Keene Corporation (Keene), Eagle-Picher Industries (E-P), and UNR Industries (UNR) all appealed pursuant to 28 U.S.C. § 1295(a)(3) (1988). E-P and UNR appealed jointly, and their appeal was consolidated with Keene’s appeal for purposes of oral argument.4
The trial judge’s decision was based on her reading of a statute, 28 U.S.C. § 1500, that pertains to the Claims Court’s jurisdiction. The judge believed the statute denied the court jurisdiction, and obligated her to grant the Government’s motion to dismiss. We believe the statute does not so dictate, and for that reason we reverse and remand for further proceedings consistent with this opinion.
I.
Much of the background of the present case is detailed in the trial court’s opinion, Keene Corp., supra, and need not be recited here. In short, the present appeals raise the question of the proper application of 28 U.S.C. § 1500 (1988) to each of these cases. Section 1500 was enacted in 1868, and was most recently before us in Johns-Manville Corp. v. United States, 855 F.2d 1556 (Fed.Cir.1988) (per curiam), cert. denied, — U.S. -, 109 S.Ct. 1342, 103 L.Ed.2d 811 (1989), aff'g Keene Corp. v. United States 12 Cl.Ct. 197 (1987). The statute deals with the jurisdiction of the Claims Court, and bars that court from hearing claims when, under the terms of [656]*656the statute, those claims are pending in other courts:
The United States Claims Court shall not have jurisdiction of any claim for or in respect to which the plaintiff or his as-signee has pending in any other court any suit or process against the United States or any person who, at the time when the cause of action arose, was, in respect thereto, acting or professing to act, directly or indirectly under the authority of the United States.
In the Johns-Manville case two issues involving the interpretation of § 1500 were addressed. First, the Claims Court had applied the statute to bar the plaintiffs’ suit on the grounds that there were pending district court cases raising the same claims. On appeal plaintiffs argued that since their third-party district court suits against the Government were based on different theories of relief than their direct action Claims Court suits, they were different “claims” and thus not subject to the § 1500 bar. This Court held otherwise (thereby affirming the Claims Court), concluding that the term “claim” is defined by the operative facts alleged, not the legal theories raised.
For example, the fact that one set of operative facts may create liability both in tort and contract does not mean that a recital of such facts states two separate and distinct “claims” as that term is used in § 1500. We explained that this construction of the term “claim” serves the underlying purpose of § 1500, which is “to prohibit the filing and prosecution of the same claim against the United States in two courts at the same time.” Johns-Manville at 1562.
The second issue addressed by this Court is whether third-party complaint cases that are stayed in the district court are “pending” within the terms of § 1500. We held, based on a plain meaning interpretation of the statute, that “ ‘pending’ includes cases which have been filed but stayed.” Johns-Manville at 1567.
The case now before us raises a third issue regarding the meaning of § 1500. In the present action, on November 16, 1988, with the Johns-Manville case in hand, the Government moved for summary judgment against the seven parties involved at the trial level. The Government asserted that at the time of filing in the Claims Court, each plaintiff had one or more suits pending in another court. Proceedings on the Government’s motion were stayed while the appellants in the Johns-Manville case sought certiorari in the Supreme Court. The Supreme Court declined further review. Keene Corp. v. United States, — U.S. -, 109 S.Ct. 1342, 103 L.Ed.2d 811 (1989).
Thereafter, on June 1, 1989, the Claims Court granted the Government’s summary judgment based on lack of jurisdiction. Keene Corp. v. United States, 17 Cl.Ct. 146. With the exception of GAF Corporation, the Claims Court determined that each plaintiff had an earlier-filed suit that involved the same claim, i.e. same operative facts, as that in the Claims Court action, and that those earlier-filed suits were “pending” at the time the Claims Court suit was filed.
Keene, UNR, and E-P appealed to this court, arguing that the phrase “has pending” does not mean pending at the time the Claims Court action is filed. They contend that § 1500 should not bar jurisdiction if the earlier-filed suit, though still pending at the time the Claims Court action is filed, is dismissed before the Claims Court entertains the § 1500 motion to dismiss. Keene further argues that in its case, the earlier-filed suit does not involve the same operative facts as its Claims Court suit and that, however, the issue of “has pending” is resolved, § 1500 thus does not bar Claims Court jurisdiction. In view of our disposition of the jurisdiction issue, we need not address this additional issue.
II.
A. Earlier-Filed Suits In The District Courts
1. UNR Industries
In re All Maine Asbestos Litigation, Master Asbestos Docket (D.Me., filed July 21, 1982), is an omnibus consolidation of [657]*657225 suits brought by present or former shipyard workers or their representatives seeking recourse for injury due to exposure to asbestos manufactured or supplied by the many named defendants. UNR is one of those defendants. The defendant manufacturers and suppliers, in turn, filed third-party actions for contribution or indemnification against the United States. These third-party suits were initiated in two different complaints, Model Third-Party Complaint Against the United States of America “A” (“Model Third-Party Complaint A”) and Model Third-Party Complaint Against the United States of America “B” (“Model Third-Party Complaint B”). Keene Corp., 17 Cl.Ct. at 149-50.
Eventually, all of the claims of the Model Third-Party Complaints A and B were dismissed. While the claims within each Model Third-Party Complaint were not all dismissed at the same time, for purposes of this opinion it is sufficient to note that by March 12, 1987, all of the claims of the Model Third-Party Complaint A were dismissed, In re All Maine Asbestos Litigation (BIW Cases), 655 F.Supp. 1169, 1171 (D.Me.1987),5 and that by July 16, 1986, all of the claims of the Model Third-Party Complaint B were dismissed.6 In re All Main Asbestos Litigation (PNS Cases), Master Asbestos Docket (D.Me. July 16, 1986).7
2. Eagle-Picher Industries
a. In re All Maine Asbestos Litigation, Master Asbestos Docket (D.Me., filed July 21, 1982)
E-P’s involvement in In re All Maine Asbestos Litigation is similar to that of UNR. Thus all of E-P’s third-party complaints against the Government, filed on July 21, 1982, were dismissed no later than March 12, 1987. Keene Corp., 17 Cl.Ct. at 149-50.
b. Albert Lopez, et al v. Eagle-Picher Industries, Inc. v. United States, No. C-84-155M (W.D. Wash., filed Feb. 3, 1983)
On February 3, 1983, E-P filed in the District Court for the Western District of Washington ten third-party complaints against the United States seeking indemnification for its liabilities resulting from asbestos related injuries that its employees allegedly sustained while employed at the Puget Sound Naval shipyard. The case Albert Lopez, et al v. Eagle-Picher Industries, Inc. v. United States, No. C-84155M (W.D. Wash., filed Feb. 3, 1983) (“Lopez”), was treated as a test case; E-P’s third-party complaint was dismissed on May 19, 1986 for failure to state a claim.8 Lopez v. Johns Manville, 649 F.Supp. 149 (W.D.Wash.1986). The remaining nine third-party complaints were dismissed on June 30, 1986. Keene Corp., 17 Cl.Ct. at 152.
3. Keene Corporation
a. Miller v. Johns-Manville Bldg. Prod., et al, No. 78-1283E (W.D.Pa., filed June 1, 1979)
In this action (“Miller”), the personal representative of the estate of a laborer allegedly injured in 1943 from asbestos exposure sought damages from nine asbestos suppliers, including Keene Building Products. Keene’s insurance company, in turn, initiated a third-party action against the United States and Celotex Corporation seeking contribution or indemnification for [658]*658any resulting liabilities. Keene Corp., 17 Cl.Ct. at 153. For purposes of this opinion, it is accepted that this third-party complaint was dismissed on May 13, 1980.9
b. Keene Corp. v. United States, No. 80-CIV-0401 (GLG) (S.D.N.Y., filed Jan. 22, 1980)
This suit {“Keene (SDNY) ”) was a direct action by Keene against the Government seeking indemnification, contribution or apportionment for amounts that Keene spent or may spend in defending and settling thousands of asbestos related personal injury actions. Keene Corp., 17 Cl.Ct. at 154. The district court dismissed this suit on September 30, 1981 holding the pleadings inadequate to invoke Federal Tort Claims Act jurisdiction. Keene Corp. v. United States, No. 80-CIV-0401 (S.D.N.Y.), aff'd, 700 F.2d 836 (2d Cir.), cert. denied, 464 U.S. 864, 104 S.Ct. 195, 78 L.Ed.2d 171 (1983).
B. Claims Court Suits
On January 16, 1984, UNR filed a direct action in the Claims Court, UNR Industries, Inc., No. 16-84C, seeking damages from the Government for breach of express and implied asbestos related contracts. Keene Corp., 17 Cl.Ct. at 150-51.
On March 25, 1983, E-P filed a direct action in the Claims Court, Eagle-Picher Industries, Inc., No. 170-83C, charging that the Government is contractually liable to E-P for the costs that E-P incurred in litigating and settling claims for asbestos-caused injuries. Keene Corp., 17 Cl.Ct. at 150.
3.Keene Corporation
a. Keene I
On December 21, 1979, Keene filed suit in the Court of Claims, Keene Corp. v. United States, No. 579-79C (amended petition filed May 1, 1981) {“Keene I”), charging the Government with breach of warranty in an asbestos related contract. Keene Corp., 17 Cl.Ct. at 153-54.
b. Keene II
On September 25, 1981, Keene filed a second suit in the Court of Claims, Keene Corp. v. United States, No. 585-81C (“Keene II”), alleging that the Government violated its fifth amendment rights by recouping from injured workers the payments that the Government made to those injured workers under the Federal Employees’ Compensation Act (“FECA”). Keene argued that since it had paid compensation to the injured workers either through settlement agreements or because of judgments in favor of the them, the recoupment by the Government of amounts that it has paid to injured workers under FECA is an unconstitutional taking of Keene’s property without just compensation, because the injuries to the workers were caused by the Government’s actions. Keene alleged that this taking further had the effect of increasing the amounts of judgments and settlements that Keene was required to pay and impaired its contract rights. Keene Corp., 17 Cl.Ct. 154-55.
III.
The central question in this appeal is the meaning of § 1500; specifically, since the statute bars Claims Court jurisdiction when there is a “pending” claim in another court, the issue is, pending when? In the art of statutory construction, there are as many formulations of how to do it as there are those doing it. The Supreme Court devotes considerable print to this10; the law review [659]*659commentators have made this one of the hot topics in current legal learning.11
A reading of some of the literature would seem to suggest that there was a time when the courts, in the exercise of their independent judgment, treated a statute as Vladimir Ashkenazy plays Tchaikovsky — a vehicle for expressing one’s own creativity. However that may be, today there is widespread agreement that the courts have a responsibility, perhaps a constitutional duty, to implement the will of the Congress as expressed in the enacted legislation. Toward that end, courts have moved away from the traditional recitation of constructional rules — rules that lend themselves to post-hoc rationalization — and toward structural constraints on judicial readings.
The point of beginning, then, is the statute. We first look to the statute for any express definitions of the key term or terms, and if none are found, then to see whether the terms can fairly be said to have a plain, non-ambiguous, meaning. The key phrase in our case is “has pending.” A reading of § 1500 reveals that the statute offers no express definition of the phrase. As to the plain meaning of the term “pending,” Black’s Law Dictionary 1021 (5th Ed.1979) defines it as “[bjegun, but not yet completed; during; before the conclusion of....” Webster’s Third New International Dictionary (1986) defines “pending” as “not yet decided: in continuance: in suspense....” These definitions support a conclusion that if a case is finally dismissed, either voluntarily or for lack of jurisdiction, prior to the relevant time for determining “pendingness,” it is no longer “pending.” That is, a finally dismissed action is “complete” or “concluded” from the perspective of the dismissing court. This plain meaning of the term “pending” does not, however, answer the question, when do we decide “pendingness” — when must a claim be pending to invoke the jurisdictional bar of § 1500. Regrettably, the legislative history of § 1500 fails to shed much light on this issue.
Section 1500 is a relatively ancient statute, and the legislative history concerning it is limited. As originally enacted in 1868, what is now 28 U.S.C. § 1500 read:
Sec. 8. And be it further enacted (italics in original), That no person shall file or prosecute any claim or suit in the court of claims, or an appeal therefrom, for or in respect to which he or any assignee of his shall have commenced and has pending any suit or process in any other court against any officer or person who, at the time the cause of action alleged in such suit or process arose, was in respect thereto acting or professing to act, mediately or immediately, under the authority of the United States, unless such suit or process, if now pending in such other court, shall be withdrawn or dismissed within thirty days after the passage of this act.
Act of June 25, 1868, 15 Stat. 77 (emphasis added). The language in this original version of the statute is plain concerning our issue: the relevant time for determining the existence of a § 1500 bar is at the time the Claims Court suit is filed or prosecuted.
In 1868, an adjudication of a district court claim against an agent or officer of the United States had no res judicata effect on a judgment in a subsequent Court of Claims suit against the United States itself based on the same operative facts. See Schwartz, Section 1500 of the Judicial Code and Duplicate Suits Against the Government and Its Agents, 55 Geo. L.J. 573, 576-78 (1967) (citing Matson Nav. Co. v. United States, 284 U.S. 352, 52 S.Ct. 162, 76 L.Ed. 336 (1932)).
[660]*660Senator Edmunds, the author of the bill that became Section 8 of the Act of June 25, 1868, explained:
The object of this amendment is to put to their election that large class of persons having cotton claims particularly, who have sued the Secretary of the Treasury and the other agents of the Government in more than a hundred suits that are now pending, scattered over the country here and there, and who are here at the same time endeavoring to prosecute their claims, and have filed them in the Court of Claims, so that after they put the Government to the expense of beating them once in a court of law they can turn around and try the whole question in the Court of Claims. The object is to put that class of persons to their election either to leave the Court of Claims or to leave the other courts. I am sure everybody will agree to that.
81 Cong. Globe, 40th Cong., 2d Sess. 2769 (1868) (emphasis added). Thus, concerning sequential suits on the same operative facts in different courts against different plaintiffs, § 1500 was the means for achieving the same result as that provided by the doctrine of res judicata.
Senator Edmunds statement that “[t]he object [of § 1500] is to put that class of persons to their election either to leave the Court of Claims or to leave the other courts” would seem to be aimed at claimants with actions being prosecuted at the same time in both the Court of Claims and the district courts. Despite the plain language of the original statutory text, this statement indicates that Senator Edmunds believed that, under the statute, jurisdiction for purposes of prosecuting a claim could be settled sometime after a Court of Claims suit was filed, that is, while both suits were pending. Alternatively, this statement could have been made merely in reference to the part of the original statute that afforded the litigants thirty days from passage of the act during which plaintiffs could dismiss their other actions and preserve their Court of Claims jurisdiction. This explanation would make Senator Edmund’s statements consistent with the plain meaning of the statute.
To a large extent, Senator Edmund’s concerns have become moot. Today, § 1500 is rarely used as originally envisioned. Section 1500 was enacted to deal with a specific concern over a rapid increase in the volume of complicated cotton claims cases, resulting from land seizures during the Civil War, involving suits both in the Court of Claims against the Government and in the district courts against the Secretary of the Treasury or his agents. Schwartz, 55 Geo. L.J. at 574-577. Without the force of § 1500, a full adjudication of the merits of the same claim, with differing results, could be realized in the two different courts. Today, while there may exist some situations in which a claimant may sue an officer or agent of the Government (in his/her official capacity) separate from the Government, there does not appear to exist the wholesale problem that the Government experienced in 1868.
Certainly, to the extent cases do arise where the same claim is otherwise allowably brought against an agent or officer of the Government and the Government itself, § 1500 may still be necessary to serve the function of a res judicata statute,12 as originally envisioned. However, in the past 42 years13, the Government has invoked § 1500 in at most only three single instances to bar Claims Court jurisdiction where the same claim had previously been filed [661]*661against an agent or officer of the Government in district court.14 In most of the cases arising under § 1500, the Government uses § 1500 to bar Claims Court jurisdiction where the Government itself is the sole defendant in both the co-pending Claims Court and district court actions, and an agent or officer of the Government as such is not a named defendant. And it is out of these types of cases that the issue before us today arises — a type of case unknown to the framers of § 1500.15
Yet, in actions against the Government in which the district courts and the Claims Court have concurrent jurisdiction,16 the doctrine of res judicata will protect the government from having to defend itself on the merits against a second claim brought by the same party sequentially on the same legal theories in both courts. Similarly, res judicata will protect the Government against second suits in those actions of which the district courts and the Claims Court have exclusive jurisdiction because of different legal theories or causes of action,17 but where the actions, though based on different legal theories, involve the same set of operative facts. Today, the intended res judicata effect of the original statute is necessary only in a quite limited set of circumstances, and thus § 1500 for these purposes has been rightly criticized as an anachronism.18
Between its enactment in 1868 and 1948, several changes were made in the statute, mostly minor, and with little indication of legislative intent in regards to the phrase “has pending.” In 1874, a few changes, none affecting this issue, were made in the phraseology of section 8 of the Act of June 25, 1868 as that section was incorporated in the Revised Statutes of 1874, section 1067. The changes in phraseology were not meant to change the meaning of the statute. Remarks of Representative Butler, 2 Cong. Rec., 43d Cong., 1st Sess. 129 (1873). Section 1067 of the Revised Statutes of 1874 was later adopted without change as section 154 of the Judicial Code of 1911. Act of Mar. 3, 1911, ch. 231, § 154, 36 Stat. 1138.
However, in 1948 Congress changed the language of the statute in two ways. First, the language was changed to include suits pending against the United States as well as suits pending against agents of the United States. Act of June 25, 1948, ch. 646, 62 Stat. 942. It can be presumed that this change reflected that district courts by then had jurisdiction over some claims against the United States. See 28 U.S.C. § 1346. Second, and more importantly, Congress also changed the language at the beginning of the statute from “No person shall file or prosecute in the Court of Claims, or in the Supreme Court on appeal therefrom ...” to “The Court of Claims shall not have jurisdiction of_” (Emphasis added). The effect of this change was to eliminate a key phrase — the “shall file” language. However, the legislative [662]*662history is silent on the reason for the change, other than the statement that “[c]hanges were made in phraseology.” Act of June 25, 1948, ch. 646, 62 Stat. 942. Finally, in 1982, with the establishment of the Claims Court, § 1500 was amended to apply its terms to that court. 28 U.S.C. § 1500 (1988).
The plain meaning rule, although often lumped with other rules for statutory construction, also has important structural effect beyond its rhetorical value. Honestly applied, it tells a court what not to look at — legislative debates, committee reports, newspaper commentary and other “aids” to policy development. The meaning of the law is what the words say it is. In cases in which a plain meaning exists, it is hard to quarrel with that construct.19 This, however, is not such a case.
Here we have neither plain meaning — the statute simply does not address when an earlier-filed case must be pending to invoke a § 1500 bar20 — nor clear evidence of what the Congress intended on this issue. Compare, for example, our recent decision in Amerikohl Mining, Inc. v. United States, 899 F.2d 1210 (Fed.Cir.1990), a case involving a similar unexplained language change — between the time the draft version went to conference committee and the time it was finally enacted — in a Congressional act governing federal court jurisdiction. The original language in both the House and Senate versions stated that judicial review under the particular statute at issue was to be “only” in the District Court for the District of Columbia, but the word “only” was deleted in the final enacted version. 30 U.S.C. § 1276(a) (1982). We held that even without the word “only” the plain meaning was clear: “... it appears from the plain meaning of the language in section 1276 (a)(1) that Congress intended the District Court for the District of Columbia to be the exclusive forum ... [and] nothing in the legislative history indicates that Congress intended an interpretation contrary to the plain meaning of these words.” Amerikohl Mining, Inc., 899 F.2d at 1213. We declined to speculate about why the word “only” was removed, and applied the law as written.
As we have noted, the language that remains in § 1500 after the 1948 revision does not provide us with a plain answer to the question. And as we said in Johns-Manville, “[t]he remaining legislative history is devoid of any evidence indicating the legislature’s intended purpose of section 1500.” 855 F.2d at 1561. As in Amer-ikohl, we will not attempt to extrapolate the Congress’ intent in changing the language of the statute; however, unlike in that case, we do not have a plain meaning.
IV.
The answer to the conundrum posed by § 1500, thus, must be found by ferreting out its underlying policy and applying it to the situation before us. The Supreme Court has recently stated: “It is a well-settled canon of statutory construction that, where the language does not dictate an answer to the problem before the Court, ‘we must analyze the policies underlying the statutory provision to determine its proper scope.’ ” Bowsher v. Merck & Co., Inc., 460 U.S. 824, 831 n. 7, 103 S.Ct. 1587, 1592 n. 7, 75 L.Ed.2d 580 (1983) (quoting Rose v. Lundy, 455 U.S. 509, 517, 102 S.Ct. 1198, 71 L.Ed.2d 379 (1982)). The statement by Senator Edmund, the sponsor of § 1500, explaining the purpose of the statute, is “an authoritative guide to the statute’s construction.” Bowsher v. Merck & Co., Inc., 460 U.S. 824, 832-33, 103 S.Ct. [663]*6631587, 1592-93, 75 L.Ed.2d 580 (1983) (citing North Haven Board of Education v. Bell, 456 U.S. 512, 527, 102 S.Ct. 1912, 1921, 72 L.Ed.2d 299 (1982)).
Recently the Claims Court in a well-reasoned opinion reviewed the purpose of § 1500:
Section 1500 clearly precludes the Claims Court’s exercise of jurisdiction over proceedings with similar claims against the United States that were filed previously and remain pending in other courts. The section was enacted over a century ago to avoid the maintenance of suits against the United States in the Court of Claims after a claimant failed to receive satisfaction from suit against the United States elsewhere. At that time a judgment in another court had no res judicata effect in a subsequent suit against the United States in the Court of Claims. [Citation omitted]. The legislative history and cases indicate that section 1500 was created for the benefit of the sovereign and was intended to force an election when both forums could grant the same relief, arising from the same operable facts. Johns-Manville v. United States, 855 F.2d 1556, 1564 (Fed.Cir.1988). The current purpose served by this section is to relieve the United States from defending the same case in two courts at the same time. Id. at 1562; Dwyer, 7 Cl.Ct. at 567.
Connecticut Dept. of Children & Youth Serv. v. United States, 16 Cl.Ct. 102, 104 (1989) (emphasis added).
An important purpose, then, of § 1500 — and one that remains important— is to conserve Government resources. In order to achieve this purpose, the Government contends that § 1500 is a jurisdictional statute — like the diversity jurisdiction statute — and that like other jurisdictional statutes, the jurisdictional determination must be made by looking to the existence or non-existence of jurisdiction at the time a complaint is filed. Appellants, on the other hand, argue that if there are no other suits pending at the time the Claims Court considers a § 1500 motion, then jurisdiction is not barred; since the purpose of § 1500 is to relieve the United States from defending the same claims in two courts at the same time, that purpose would only be subverted if a claimant had another claim presently pending in another court, not formerly pending.
Both parties agree generally, in accord with Senator Edmund’s expressions, that the purpose of § 1500 is to conserve the Government’s limited resources. Both parties propose answers that would achieve that result. When there are two reasonable alternative constructions of a statute, it is the court’s duty “ ‘to find that interpretation which can most fairly be said to be imbedded in the statute, in the sense of being most harmonious with its scheme and with the general purposes that Congress manifested.’ ” FBI v. Abramson, 456 U.S. 615, 625 n. 7, 102 S.Ct. 2054, 2061 n. 7, 72 L.Ed.2d 376 (1982) (Blackmun and Brennan, J.J., and O’Connor and Marshall, J.J., dissenting in separate opinions) (quoting NLRB v. Lion Oil Co., 352 U.S. 282, 297, 77 S.Ct. 330, 338, 1 L.Ed.2d 331 (1957)).
While the Government’s theory has a bright-line advantage, it fails in two respects. First, the analogy of the diversity jurisdiction statute to § 1500 does not withstand analysis. The purpose behind the diversity jurisdiction statute, 28 U.S.C. 1332 (1988), is to insure an impartial tribunal for citizens of different states. Pease v. Peck, 59 U.S. (18 How.) 595, 15 L.Ed. 518 (1855); Burt v. Isthmus Dev. Corp., 218 F.2d 353 (5th Cir.), cert. denied, 349 U.S. 922, 75 S.Ct. 661, 99 L.Ed. 1254 (1955). Section 1332 creates jurisdiction in a federal court where none would otherwise exist, i.e. no subject matter jurisdiction. The purpose behind § 1500, on the other hand, is to save the Government from having to defend the same suit in two different courts at the same time. Section 1500 takes away jurisdiction even though the subject matter of the suit may appropriately be before the Claims Court. Operationally these two statutes focus on different points in time. The former says, “ordinarily your subject matter cannot be here, but if you satisfy the requirements of this statute, then you may go ahead and proceed in federal [664]*664court.” The latter says, “though your subject matter may appropriately be here in the Claims Court, if you come within the proscriptions of this statute, then you may no longer proceed here.”
Second, the Government’s theory fails because it too early cuts off the recourse of litigants who, either because of subject matter or other circumstances, may be found entitled to pursue their claims only in the Claims Court, and who would not otherwise violate the policy behind § 1500. The facts of the present case offer a good example. UNR had third-party complaints against the Government pending in federal district court. Unsure of whether the district court had jurisdiction, and facing a running of the statute of limitations, UNR filed the same claim in the Claims Court. More than two years before the Claims Court entertained and acted on the Government’s § 1500 motion to dismiss, the district court dismissed UNR’s third-party complaints for lack of jurisdiction.21 By the time the Claims Court entertained and acted on the Government’s § 1500 motion, UNR had no earlier-filed suits pending and had yet to have its day in court. We believe the statute does not operate to preclude UNR from having its day.
Our earlier cases dealing with § 1500, particularly those decided since 1948 when the statute was amended to remove the “filing” language, do not explicitly address the “when filed” question. In those cases in which the question is implicitly involved, the results are generally consistent with this approach. For example, in British American Tobacco Co. v. United States, 89 Ct.Cl. 438 (1939),22 the plaintiff had filed the same claim in a district court and later the same day in the Court of Claims. At the time the Court of Claims entertained the Government’s motion to dismiss, the district court action was no longer pending — it had been decided. The Court of Claims dismissed the action because the district court action constituted a final adjudication on the merits. This is in full accord with our ruling today. A plaintiff is not entitled to two bites at the apple. As the British American Tobacco case illustrates, when a district court case is no longer pending because it received final adjudication on the merits, the plaintiff is not then entitled to proceed in the Claims Court — the doctrine of res judicata accomplished exactly what § 1500 was originally devised to protect against.
In Wessel, Duval & Co., Inc. v. United States, 124 F.Supp. 636, 129 Ct.Cl. 464 (1954), at the time the Court of Claims entertained and acted on the Government’s motion to dismiss, the same claim was pending in a district court, and so the Court of Claims dismissed the action before it. The holdings in Boston Five Cents Sav. Bank, FSB v. United States, 864 F.2d 137 (Fed.Cir.1988), City of Santa Clara v. United States, 215 Ct.Cl. 890 (1977), and Casman v. United States, 135 Ct.Cl. 647 (1956), are inapposite because the pending district court claims in those cases involved a type of relief, i.e. declaratory judgment, that was not available in the Claims Court.
In Brown v. United States, 358 F.2d 1002, 175 Ct.Cl. 343 (1966), the Court of Claims considered whether § 1500 barred its jurisdiction when a claim in an earlier-filed district court suit was dismissed for lack of subject matter jurisdiction. The Court of Claims had originally dismissed the claim because at the time it first entertained and acted on the Government’s motion to dismiss, the claim was still pending in the district court. Id. 358 F.2d at 1004. However, the district court later dismissed the claim for lack of subject matter jurisdiction and the plaintiff moved for a rehearing in the Court of Claims. Under these circumstances, the Court of Claims denied the Government’s § 1500 motion to dismiss, stating:
[665]*665Our earlier order of dismissal was predicated on the fact that the other “claim remains pending in the said District Court.” That is no longer true, and the claim is no longer “pending in any other court.” In this situation, we do not believe that 28 U.S.C. § 1500 requires us to deprive the plaintiffs of the only forum they have in which to test their demand for just compensation. * * * * Section 1500 was designed to require an election between two forums both of which could presumably grant the same type of relief. [Citations Omitted]. But Section 1500 was not intended to compel claimants to elect, at their peril, between prosecuting their claim in this court (with conceded jurisdiction, aside from Section 1500) and in another tribunal which is without jurisdiction. Once the claim has been rejected by the other court for lack of jurisdiction, there is no basis in the policy or wording of the statute for dismissal of the claim pending here.
Id. 358 F.2d at 1004-05.
Under the facts of the present case, the reasoning in Brown applies with even greater force. In Brown, the Court of Claims held that § 1500 did not bar jurisdiction even though the earlier-filed suit was dismissed after the Court of Claims suit was filed, and even though the Court of Claims had already once dismissed the suit because the district court action was still pending. In the present case, all of the earlier-filed suits were dismissed before the Claims Court entertained and acted on for the first time the Government’s § 1500 motion to dismiss.
In Connecticut Dept. of Children & Youth Serv. v. United States, 16 Cl.Ct. 102 (1989), the Claims Court was confronted with a situation similar to that in Brown. In this case, the plaintiff originally filed suit in the district court, but then filed the same complaint in the Claims Court because it realized that the district court might not have jurisdiction and feared that the statute of limitations might run before the district court considered the matter. Unlike Brown, however, the district court suit had not been dismissed by the time the Claims Court entertained and acted on the § 1500 motion to dismiss. Given its predicament, the plaintiff asked the court to stay its action until the subject matter jurisdiction of the district court was resolved. The Claims Court correctly noted that under our holding in Johns-Manville, supra, the grant of a stay would not escape the bar of § 1500. It went on to note, however, that “[sjhould the federal district court lose its hold on the pending matter, finding it has no jurisdiction over the refund claim, section 1500 would no longer apply and this court would then be freed from the jurisdictional bonds that preclude action at this time.” 16 Cl.Ct. at 105.
The Government argues that the Connecticut Dept, of Children holding is limited to circumstances in which the statute of limitations has not otherwise run. This, it deems, is the import of the Claims Court’s language immediately following the above quote: “Then, and should the limitations period not have passed, this court should properly take and adjudicate the relevant issues. Brown, 175 Ct.Cl. at 348, 349, 358 F.2d at 1004, 1005.” This, however, was not a condition of the Brown holding. In Brown the Court of Claims stated: “The plaintiffs could undoubtedly file a new petition, without any bar through Section 1500; it does not seem fair or make sense to insist that this must be done—with the limitations difficulties it may well entail. ” 358 F.2d at 1005 (emphasis added). In fact, the court in Brown suggested that it is when the statute of limitations does run that § 1500 should not bar jurisdiction when the earlier-filed suit is dismissed and the Claims Court action was filed before the running of the limitations statute. Section 1500 does not impose a statute of limitation. So the Court of Claim’s first statement, i.e. “plaintiffs could undoubtedly file a new petition, without any bar through Section 1500,” merely acknowledges that the same claim would not be pending in two different courts at the same time.
V.
We hold then that when an earlier-filed district court case is finally dismissed [666]*666before the Claims Court entertains and acts on a § 1500 motion to dismiss, § 1500 does not bar Claims Court jurisdiction even though the dismissal may have occurred after the filing of the Claims Court action. Under today’s decision the § 1500 bar can only be invoked if the Government can show that the earlier-filed suit is still pending at the time the Claims Court entertains and acts on the jurisdictional question.
The Government warns that any interpretation of § 1500 that varies its focus from the time of filing will unduly burden the United States by forcing it to defend an earlier-filed district court complaint and a Claims Court complaint on the same basic claim at the same time, and would in effect give a plaintiff “two bites at the apple.” The Government further warns that such interpretation will require the Claims Court to undertake a burdensome analysis to determine whether the earlier-filed suit is pending. The decision, however, does not fuel these worries.
A claimant is entitled to its day in court. That does not mean, however, that a claimant is entitled to tie up Government resources by forcing it to simultaneously defend itself in two courts. And this is what § 1500 protects against. If an earlier-filed action is finally dismissed before the Claims Court entertains and acts on the jurisdictional question, then the Claims Court action will be the only suit pending and the plaintiff will have but one day in court.23 If at the time the Claims Court entertains and acts on the jurisdictional question the earlier-filed action is still pending, i.e. not finally dismissed, then § 1500 will bar jurisdiction. Plaintiff will then either have to proceed with the earlier-filed action or come back to the Claims Court, if a statute of repose does not prevent, should its district court action be dismissed. The determination that the Claims Court must make is straightforward. Under this rule it is not possible for a plaintiff to prosecute an action both in the Claims Court and in another court at the same time. This is what § 1500 precludes.
VI.
The dissent suggests three simple rules to be derived from § 1500 in lieu of the rule here announced. Rule 1 would key the application of § 1500 to the date of filing of the suit in the Claims Court, so that the same case previously filed and still pending in another court on the date the Claims Court suit is filed would automatically bar the Claims Court case, regardless of whether the previously-filed case is ultimately dismissed in a manner that denies the plaintiff its day in court, or whether the Government is called upon to invest any significant resources in defending the Claims Court action. That is not an unreasonable reading of the statute, but for the reasons stated we find that, given the inherent ambiguity of the statute, both sound policy and legislative history support a different reading.
Rule 2 would apply the same result to the reverse sequence of filing. That is, if the suit is first filed in the Claims Court, and then later in another court, the second filing divests pro tanto the Claims Court of jurisdiction. Rule 2 requires overruling Tecon Engineers, Inc. v. United States, 343 F.2d 943, 170 Ct.Cl. 389 (1965), which holds that § 1500 on its terms only applies to Rule 1 cases, and not to the reverse sequence cases.
It may well be that Tecon does result in an anomalous situation, one not to be lightly attributed to a rational legislature. Cf. Griffin v. Oceanic Contractors, Inc., 458 U.S. 564, 575, 102 S.Ct. 3245, 3252, 73 L.Ed.2d 973 (1982) (“interpretations of a statute which would produce absurd results are to be avoided if alternative interpretations consistent with the legislative purpose are available”). However, if, as the dissent argues, the time of filing is the critical jurisdictional determinant, Rule 2 cannot logically follow. For in a reverse [667]*667sequence case, at the time the Claims Court takes jurisdiction there are no other cases pending.
The dissent tries to avoid this logic trap by arguing that jurisdiction is a continuing concept, and that if at some later time requisite events occur, such as the filing of a suit in another court, the Claims Court will be divested of jurisdiction. This adopts at least in part our theory: it is not simply a question of whether at the time of filing there is a pending case, but whether there are two cases being litigated that invokes the bar of § 1500. In short, we reach the same result as the dissent under Rule 2, but we arrive there by applying a consistent rule.
Rule 3 simply restates the principle of res judicata as it applies to these cases; there is no disagreement here.
VII.
UNR filed its Claims Court action on January 16, 1984. Its earlier filed district court cases, i.e. Model Third-Party Complaints A & B, were all dismissed either as of March 12, 1987, or October 20, 1988 when UNR voluntarily dismissed all of its third-party actions that had been stayed through bankruptcy proceedings.
E-P filed its Claims Court action on March 25, 1983. One of its earlier-filed district court cases, Lopez, was dismissed on May 19, 1986. Its other earlier-filed district court cases, i.e. Model Third-Party Complaints A & B, as well as some other cases, were dismissed no later than March 12, 1987.
Keene filed its first Claims Court action on December 21, 1979 and its second Claims Court action on September 25, 1981. The earlier-filed district court case that preceded both Keene I and Keene II, i.e. Miller, was dismissed in 1980. Keene’s other earlier-filed district court case, Keene (SDNY), which only preceded Keene II, was dismissed on September 30, 1981.
The Claims Court entertained and acted on the Government’s § 1500 motion to dismiss UNR, E-P and Keene on June 1, 1989. Since none of those parties had the same claims pending in another court on June 1, 1989, we hold that § 1500 does not bar Claims Court jurisdiction.
Keene argues that one of its earlier-filed actions, i.e. Miller, does not involve the same “claim” as its Claims Court action {Keene I) and that § 1500 therefore does not bar Claims Court jurisdiction. We here decide that Miller was not pending so as to bar Claims Court jurisdiction because that case was dismissed before the Claims Court entertained and acted on the Government’s § 1500 motion to dismiss. Thus, we need not reach the question of whether Miller involved the same claim as Keene I.
CONCLUSION
Accordingly, we hold that 28 U.S.C. § 1500 does not bar jurisdiction of the Claims Court when the earlier-filed claim has been finally dismissed after the Claims Court action is filed but before the Claims Court entertains and acts on the Government’s § 1500 motion. The Order of the Claims Court dismissing these cases for lack of jurisdiction is therefore reversed and the case is remanded for further action consistent herewith.
REVERSED and REMANDED.