This is a qui tam case raising a question of whether the relator can proceed after the government has settled.

The government bought cruise missiles from Boeing. A division of Northrop was at relevant times manufacturing a navigational guidance device for the missiles. The missiles are designed to be carried under the wings of B-52 or B-1 bombers, dropped from the wings as the planes fly, and then to fly independently toward their targets at low altitudes. The part at issue, a "flight data transmitter" manufactured by Northrop, consisted substantially of three gyroscopes, an accelerometer, and electronic components enclosed in a box. Fluid is used to damp movement inside the box. The fluid is supposed to stay sufficiently liquid to perform its damping function down to 65 degrees below zero Fahrenheit, according to the contract specifications. The missiles were of course manufactured for supply to the United States, which paid for them.

Two things are alleged to have been false about Northrop's claims for payment for the flight data transmitters. One is that the required tests were not performed, and the certifications that they had been were false. That has been established by evidence. Barajas was largely responsible for faking the tests, and has testified that he did so, and Northrop has pleaded guilty to a crime on account of his falsifications when he was its employee. Northrop and the government settled the qui tam case, based on the fraudulent certifications of tests, and it was dismissed. Northrop paid the government $8 million to settle the civil case, on top of $17 million in fines and penalties in the criminal case. Barajas was awarded $864,000 of the $8 million settlement, as relator.

The second alleged fraud is that the fluid would gum up at 50? below zero Fahrenheit instead of 65? below, a very material difference from the contract specification of 65? below in subarctic winter conditions where temperatures between those two figures are not uncommon. This second falsehood, that the damping fluid freezes at 50? below, has never been proved, although Northrop was indicted for it. Northrop pleaded guilty only to the false claims, not the bad fluid. The government and Northrop settled the qui tam case based on the false claims alone.

Barajas testified that he thought the flight data transmitters were all fine, and passed the tests whenever the tests were honestly performed. He testified that among the reasons he faked the tests was that the flight data transmitters always passed, so he thought the tests were a waste of time. Barajas read about the tendency of the fluid to gum up in the cold in a newspaper, which reported a story based on the indictment. But he has never proved that it does, because his case was dismissed on the theory that his claim would be barred regardless. In this decision, we assume for purposes of analysis that the fluid does gum up at 50 below, but we do not know that and it has never been proved.

The qui tam statute provides that if the action was brought "by a person who planned and initiated" the fraud against the government, then the court "may ... reduce the share" he receives as relator, or shall dismiss him with no share if he is convicted of a crime in connection with the wrongdoing. Thus the statute has the peculiarity that a person may, as an employee of a government contractor, cause his employer to unknowingly defraud the government, and then as a relator, become wealthy based on his own dishonesty as an employee. Barajas was not convicted of a crime for his fraud, but he was responsible for the fraudulent certifications that the flight data transmitters had been tested, upon which his qui tam action was based.

Because of this statutory peculiarity, the district court held extensive evidentiary hearings to determine how much to award Barajas, and made findings of fact. The court found that Barajas systematically faked many of the tests. Barajas was one of two people principally responsible for testing the flight data transmitters, and "admitted that beginning in 1983 he began to systematically fake many of the FDT test results by recording made up numbers on the test charts rather than actually performing the required series of tests." Barajas was told by his supervisors to falsify some test results on an intermediate test, but nobody ever told him to falsify any of the final tests. He testified that he falsified many final acceptance tests "on his own initiative due to the time pressure to complete his work and his belief that the [test] equipment provided by Northrop was faulty." The tests Barajas faked entirely on his own initiative accounted for two thirds of the settlement. He did not tell his supervisors that he was falsifying the final acceptance tests. The district judge found that Barajas " 'planned and initiated' the violations of the acceptance tests he committed." The district judge found that low quality testing equipment, a lax atmosphere at Northrop, and instructions by supervisors to falsify some intermediate tests, encouraged Barajas to commit his fraud. The judge found some mitigation of Barajas's wrongdoing because "Barajas seems to have had legitimate doubts about the utility of performing the Acceptance Test, since the FDTs appear never to have failed." Thus Barajas skipped the testing and filled in fake numbers, not because he thought the fluid would gum up in temperatures warmer than specified, but because he thought the guidance systems were always up to specification so it was a waste of time to do the testing work.

These facts are amplified from what we understood in the earlier appeal, U.S. ex rel. Barajas v. Northrop Corp., 5 F.3d 407, 408 (9th Cir.1993), because the ligation has proceeded further in district court and there was an evidentiary hearing and findings of fact.

After getting fired, Barajas sued Northrop under the qui tam provisions of the False Claims Act. The government investigated and elected to take over the action, pursuant to 31 U.S.C. § 3730(b)(4)(A). The government's amended complaint alleged that Northrop had falsified test results and falsely claimed that required tests had not been performed, so its claims for payment were false. The government's amended complaint did not allege that the fluid would freeze at a warmer temperature than specified, just that the tests had been faked.

Six months after it took over the civil case, the government indicted Northrop for criminal violations of the False Claims Act. The indictment, unlike the civil complaint, alleged that the damping fluid would freeze at temperatures warmer than specified, as well as that tests were faked. Northrop pleaded guilty and admitted that the tests had been faked, but Northrop did not admit that the fluid would freeze above the specified temperature. The government accepted its plea agreement on this basis, and Northrop was penalized $17 million.

In the civil case, Barajas successfully moved to sever and pursue separately his damping fluid claims, which the government elected not to pursue. Barajas's amended complaint alleged both that the damping fluid froze at too high a temperature and that the tests were faked.

Barajas, the government, and Northrop settled. The government sought to limit Barajas's qui tam relator's share of the settlement to a minimal amount, because he had planned and initiated the fraudulent tests that were the basis of the recovery. Its challenge to Barajas's legitimacy as a qui tam relator occasioned the evidentiary hearing described above.

The government used an ordinary broad form release. The government released Northrop from "any and all ... claims under the False Claims Act." It is unquestioned in this litigation that the government itself is barred by its release from pursuing any claim based on the alleged tendency of the damping fluid to gum up at 50? below, because it expressly released those claims and its case was dismissed with prejudice. But Barajas's release reserved whatever right he had to pursue a qui tam action based on that inadequacy of the fluid:

Qui tam plaintiffs agree to dismiss with prejudice all portions of the original complaint and the First Amended and Severed Qui tam Complaint ... except for the claims based on the allegations that the DC-200 damping fluid used in the [transmitters] did not meet Air Force cold temperature performance requirements.

Northrop did not concede that there were any such claims or that if there were, Barajas could reserve and pursue them personally. But if there were and he could, he plainly did.

After the settlement, the district court dismissed Barajas's action on the ground that he was not the "original source" of the damping fluid allegations, because they were publicly disclosed in the indictment that preceded his complaint and he learned about them by reading the newspaper. We reversed, holding that factual inquiry was necessary to determine whether Barajas had played a hand in triggering the criminal case. Barajas I, 5 F.3d at 411. We said that if the district court determined that Barajas was an original source under the statute, then it would have to determine whether he could pursue his claim despite the settlement. On remand, the district court dismissed Barajas's claims based on the res judicata effect of the settlement. None of the questions raised in this appeal address any of what we resolved in the earlier appeal, and no claim is made that the district court in any way deviated from the mandate in the earlier appeal.

I. Res judicata in qui tam actions.

Barajas argues that res judicata cannot apply to his claim that the lubricant would gum up when cold, because it has not actually been adjudicated. He cites no authority for this proposition. It cannot be correct. The dismissal of the government's false claim action, based on the fake test reports, is the putative bar. He appears to be arguing that because the government intervened after he filed, that somehow affects the bar. The date of rendition of the judgment is controlling for purposes of res judicata, not the dates of commencement of the action creating the bar or the action to be affected by the bar. Restatement 2d of Judgments § 14 (1980). That the government intervened subsequent to Barajas's filing makes no difference. Cf. United States v. City of Green Forest, 921 F.2d 1394 (8th Cir.1990).

Res judicata bars relitigation of all grounds of recovery that were asserted, or could have been asserted, in a previous action between the parties, where the previous action was resolved on the merits. Costantini v. Trans World Airlines, 681 F.2d 1199, 1201 (9th Cir.1982). It is immaterial whether the claims asserted subsequent to the judgment were actually pursued in the action that led to the judgment; rather, the relevant inquiry is whether they could have been brought. C.D. Anderson & Co. v. Lemos, 832 F.2d 1097, 1100 (9th Cir.1987).

The government has filed1 an untimely amicus curiae brief, in which it now supports Barajas, despite its earlier hostility to his claim for compensation. The government suggests that res judicata should not be applied to qui tam actions, because it would disable private qui tam relators from pursuing claims that the government was disabled from pursuing "for lack of resources." This argument is unpersuasive. Though the point is not critical, it is hard to imagine "lack of resources" disabling the government but not a private qui tam relator. The government cites no authority (except for United States v. American Heart Res. Found., 996 F.2d 7 (1st Cir.1993), which is not in point) for this surprising proposition. The proposition is purely a policy argument, and as a policy argument it is too weak a reason to upset well established law on res judicata.

Congressional policy as manifested in the statute is to use qui tam cases to recover three dollars for every dollar taken from the government by fraudulent claims for payment. Most of the dollars necessarily must come from settlements. Yet the government's argument, if adopted, would make settlements infeasible. The government's ability to settle cases depends on its power to give defendants releases that will protect them from more litigation based on the same invoices. If the government cannot give a defendant such a release, then the defendant might as well go to trial.

II. Separate claims.

Barajas's and the government's more serious argument is that what they call the "cold temperature allegations" are distinct from the "false testing allegations," so a judgment on the latter could not bar the former. The "cold temperature claims" are that the lubricant would gum up at 50? below instead of the specified temperature, 65? below. The "false testing allegations" are that Northrop certified tests where the results were faked or the tests were never made. The government argues that substitution of inferior fluid in the gyroscopes took place in Massachusetts, false testing in California, and that the California Northrop employees did not know that the fluid was not up to specification.

To determine whether Barajas's claim that the fluid was out of specification is barred by res judicata, we focus upon "whether the two suits arise out of the same transactional nucleus of fact," though we also consider other factors as necessary. Costantini v. Trans World Airlines, 681 F.2d 1199, 1201-02 (9th Cir.1982). We have applied the doctrine of res judicata "on the ground that the two claims arose out of the same transaction," without reaching other factors cited in Costantini. International Union of Operating Engineers v. Karr, 994 F.2d 1426, 1430 (9th Cir.1993).

Application of "same transactional nucleus of fact," is quite clear here. While Barajas is plainly correct that it is one thing to have fluid that gums up in the cold, and another to lie about whether the fluid was tested for gumming up, both wrongful acts arise out of the same attempt to get paid for flight data transmitters not up to specifications. The recovery in a qui tam case is not for each false statement or bad act done to the government; it is for "a false or fraudulent claim for payment or approval." 31 U.S.C. § 3729(a)(1). Northrop was entitled to get paid for the flight data transmitters only if they were tested according to specifications and if the fluid stayed liquid in cold temperatures according to specifications. The government recovered the easy way, on the claim Northrop conceded, because Barajas admitted he had faked the tests. Thus the government did not have to prove the more technically complex claim, that the fluid was not up to specifications. And proving it would have been a waste of time, because the government could and did recover the money it had paid on Northrop's false invoices, based on the testing fraud. It did not matter to the settlement and judgment whether Northrop's invoices were false for two reasons or one reason. The material fact was that the invoices were false. The false invoices for the flight data transmitters were the "transactional nucleus."

Additionally, Barajas could not have a claim separate from the government's. A qui tam relator has Article III standing to sue only as a relator, on behalf of the government. His standing is in the nature of an assignee of the government's claim, see United States ex rel. Kelly v. Boeing Co., 9 F.3d 743, 748 (9th Cir.1993). Thus there is one claim, the government's, pursuable either by the qui tam relator on behalf of the government, or by the government on its own behalf. Once the government recovers the money it paid on a false invoice, plus penalties, or releases its claim, there is no more to be recovered by anyone, because only the government can have a claim for a false claim made upon the government. The government plainly could not recover $3 million on a $1 million invoice based on faked tests, and then recover another $3 million on the same invoice because the goods supplied had not been up to specifications. In a qui tam case, only the government has a dog in the fight. Barajas did not buy any cruise missile parts, so he could not get defrauded by fake test results or bad damping fluid.

Our application of "same transactional nucleus" has consistently been broad enough to embrace claims at least as distinct as those in this case. In International Union of Operating Engineers v. Karr, 994 F.2d 1426 (9th Cir.1993), we held that a subsequent action for an accounting, brought by an employee benefit trust, was barred by the trust's previous two actions alleging delinquent payments, which had been settled. In Western Systems, Inc. v. Ulloa, 958 F.2d 864, 871 (9th Cir.1992), we applied Section 24 of the Restatement of Judgments, which says that two events are part of the same transaction if they are related to the same set of facts and could conveniently be tried together. Under the Restatement view, different theories supporting the same claim for relief must be brought in the initial action, because a single "transaction," or nucleus of operative fact, gives rise to only one claim by one person against another. Restatement (Second) of Judgments § 24 cmt. c (1980).

III. Reservation of the claim.

Barajas argues that the settlement expressly reserved the claim that the fluid would gum up in cold temperatures, and only settled the claim that the tests were falsified. A settlement can limit the scope of the preclusive effect of a dismissal with prejudice by its terms. International Union of Operating Engineers v. Karr, 994 F.2d 1426, 1432 (9th Cir.1993); Prestin v. Mobil Oil Corp., 741 F.2d 268, 273 n. 6 (9th Cir.1984). Barajas did indeed reserve whatever right he had to litigate the claim that the fluid was not up to specifications despite the government's full and complete settlement and release. But reservation of a nullity could not entitle him to recover. The question is whether, the government's claim being over, there could be any claim left for Barajas to pursue.

There was only one claim that was false, Northrop's claim for payment on its invoices for the flight data transmitters. At most the same claim was false for two reasons, because Northrop personnel in California faked the tests and because Northrop personnel in Massachusetts substituted inadequate damping fluid. That the Californians, including Barajas, did not know what the Massachusetts personnel were doing, if that is true, does not matter. The time that matters is when the government released "any and all" claims, and at that time the parties knew about and could litigate the cold temperature claims if they so chose.

The government argues that the claims that the fluid would gum up in the cold are separate from the claims that Northrop falsely certified that it had tested the flight data transmitters, as is evidenced by Northrop's resistance to discovery on the fluid claims. But they relate to the same invoices, the same payments of money the government made.

The government argues that Northrop should be estopped from asserting res judicata because it resisted discovery regarding the cold temperature claims. The conclusion does not follow. Northrop was entitled to resist discovery on claims that the government elected, with its eyes open, not to assert, without losing the benefit of its release and judgment. Had the government wanted to do more discovery before settling the cold temperature claims, it could have refused to release them. It did release them.

That Barajas could not pursue the cold temperature theory separately while the government pursued the claim is wholly irrelevant. The government, not Barajas, was defrauded. Barajas's Article III standing arose from his position in the nature of statutory assignee of the government's claim to recover as a victim of the fraud. United States ex rel. Kelly v. Boeing Co., 9 F.3d 743, 746 (9th Cir.1993). The government had an express statutory power to "settle the action with the defendant notwithstanding the objections" of Barajas, 31 U.S.C. § 3730(c)(2)(B). It exercised that power. That left no claim of which Barajas could be the assignee.

Barajas, as relator, has already been awarded his share following a settlement and dismissal of a qui tam case by the government. The government having settled the case, there is no claim left for Barajas to pursue. The allegation that the damping fluid would gum up in temperatures warmer than specified was not a different claim from the one that the tests were falsified. They were alternative theories for the same claim, that Northrop had presented a false claim to the government for payment for the flight data transmitters.

AFFIRMED.

I respectfully dissent. The majority interprets the settlement agreement between Northrop and the Government to embrace Barajas's cold fluid claim. On its face, the Government's release of "any and all ... claims under the False Claims Act" may extend to the cold fluid claim, but this language must be interpreted against the background of the Government's civil suit against Northrop. See Miller v. Fairchild Indus., Inc., 797 F.2d 727, 734-35 (9th Cir.1986). The Government did not adopt the cold fluid claim in its civil complaint, alleging only that Northrop falsified its testing records. Throughout the litigation, both Northrop and the Government maintained the cold fluid claim was distinct from the falsification claim. Moreover, in the negotiations leading to the settlement agreement, both parties acknowledged that the cold fluid claim was not covered by the settlement.

Because the surrounding circumstances show Northrop and the Government did not intend their settlement to dispose of the cold fluid claim, the settlement agreement does not preclude Barajas's qui tam action based on that claim, irrespective of whether res judicata principles apply. If, as the Government argues, res judicata does not apply to qui tam claims, Barajas is free to assert any claims not adopted by the Government, including his cold fluid claim; if, as Northrop argues, res judicata does apply here, Barajas may nonetheless bring his claim because it is not precluded by the settlement agreement. See International Union of Operating Eng'rs v. Karr, 994 F.2d 1426, 1429 (9th Cir.1993) (noting that the res judicata effect of a settlement agreement may be limited by the terms of the agreement).