PATRICK E. HIGGINBOTHAM, Circuit Judge:

In this qui tarn action challenging Lockheed’s performance of a government contract, the plaintiff, James Mayfield, appeals from the district court’s grant of Lockheed Martin’s motion for summary judgment. We affirm.

In 1993 Lockheed bid on an Evaluation, Testing, and Analysis contract, a research- and-development project of NASA. The research contract followed on the heels of an Engineering Support Contract, for which Lockheed had been the dominant contractor since 1987. 1

Lockheed’s bid for the research contract responded to NASA’s Request for Proposal, an encyclopedic document detailing the anticipated scope of work to be performed. The RFP described a hypothetical workforce designed by NASA that it required bidders to use to calculate their bids. NASA projected a need for 13,411,404 man-hours — give or take 15% — to complete the research contract’s projected tasks. It is undisputed that those tasks required less-skilled labor and lower staffing levels than Lockheed used for the engineering contract. The research contract’s base term was to run from January 1994 through September 1998, with a contract value during that period of roughly $510,187,031. Lockheed won the contract. Its Best-and-Final-Offer, in government contracting parlance BAFO, promised 13,-411,995 labor hours at a cost of $425,341,412.

As we will explain, the structure of the research contract itself answers most of the questions in this case. It was a cost-plus-award-fee (“CPAF”) contract. According to the Federal Acquisition Regulations, under a CPAF a contractor is paid a *257 base fee fixed at the inception of the contract, plus an award fee. 2 The award is based on the government’s unilateral evaluation of the contractor’s performance at regular intervals, adjusted in accordance with the work performed under the contract, a compensation arrangement designed to enhance incentive for timely and sound performance. 3

The research contract provided this incentive through its “Award-Fee Plan,” NASA’s periodic evaluation of the quality of Lockheed’s technical work and the project’s actual costs in relation to the contract’s value. The research contract provided that Lockheed might earn an award fee of as much as $31,667,248 over the life of the contract. The actual fee was to be calculated at ten fee-evaluation periods and paid out in semi-annual installments.

The award paid in any given period was discretionary and determined by NASA. NASA evaluated Lockheed’s performance under a 100-point system that corresponded to the maximum allowable award fee available in a period. Lockheed’s “cost performance” with respect to contract value constituted 30 grading points; the other 70 points were awarded for technical performance. An overall score of 60 or below was a “failing” grade, in which case Lockheed would receive no award in that period. The research contract’s base fixed-fee amount was zero; if Lockheed failed each award period, it earned nothing.

The research contract’s Award-Fee Plan was not the only variable in the contract. Because the precise level of effort and the nature of the work to be done on a CPAF-type contract are, by definition, both unknown at the outset, neither NASA nor Lockheed knew how much the research contract’s labor requirements would cost. 4 For example, NASA considered — and ultimately rejected — exercising the RFP’s “Space-Station Option,” which would have transitioned Lockheed’s existing engineering contract projects into research contract projects. Prior to signing the research contract with Lockheed, NASA demurred on the space station, instead instructing Lockheed to finish the engineering contract’s space-station-related tasks, but to charge the hours to the research contract. This agreement was representative of many uncertainties inherent in the structure of the research contract and, says Lockheed, is why the research contract was overbudget from the start.

Appellant and qui tarn relator James Mayfield was a Specialist at Lockheed responsible for preparing the cost estimates for the research contract project. Lockheed reported its accrued costs and anticipated future costs to NASA monthly and quarterly using “533” forms. Mayfield was concerned that the “wrap rate” 5 pro *258 jections reported to NASA in the 533s greatly underestimated the actual cost of labor needed to fulfill the research contract task assignments.

Mayfield’s concern was not unfounded. Early on, Lockheed failed to meet its baseline projections, incurring cost overruns of 13.7% and 18.1% in the first two evaluation periods. NASA and Lockheed both took steps to remedy the situation: Lockheed revised the skill mixes it applied to certain tasks and laid off workers, and NASA ordered Lockheed to submit a “Cost-Recovery Program” that would bring the research contract’s costs back to Earth. 6 Mayfield questioned the accuracy of several 533s and was eventually fired in February 1995.

He responded by bringing a wrongful-discharge action in Texas state court, which granted summary judgment for Lockheed in 1996. The Texas Court of Appeals subsequently affirmed. 7 This suit, filed in the Southern District of Texas, followed. The district court granted summary judgment for Lockheed, holding that the wrongful-discharge judgment was a res-judicata bar to some of Mayfield’s qui tam claims and that his remaining claims were barred by the False Claims Act’s “original source” rule. 8

This court reversed, holding that res judicata was not a bar and that Mayfield was an independent source of the cost information at issue. 9 On remand the district court in a careful opinion granted summary judgment in favor of Lockheed, ruling that: (1) there was no evidence that Lockheed submitted a false bid within the meaning of the FCA; (2) the parties had established a post-contract relationship that precluded FCA liability; (3) Lockheed’s 533 cost projections were neither false nor material and therefore not false claims; and (4) NASA ratified Lockheed’s facilities overcharges.

Mayfield appeals, arguing that Lockheed swindled the United States out of $22,491,500 in award fees and $10,902,978 in illicit facilities costs, totaling $33,394,478 in damages, swelled by the FCA’s treble-damages provision to $100,183,434. 10 Mayfield asserts that Lockheed made false claims in its submission of (1) the bid, (2) the 533s, and (3) the claims for certain rental costs. Mayfield has standing to bring all three claims. 11 We address each in turn.

*259 II

Mayfield first attacks Lockheed’s underbid as an actionable “false bid” because “Lockheed knew from day one that it could not deliver the hours under the contract for the amount represented in its BAFO.” He argues that Lockheed’s allegedly knowing misrepresentation in the BAFO was a fraudulent act rendering each subsequent claim for payment made by Lockheed during Mayfield’s employment a false claim within the meaning of the FCA. Lockheed responds that Mayfield has produced no evidence to suggest that the BAFO was fraudulent, and explains that the cost overruns resulted from NASA’s mid-contract alterations of the research contract’s task assignments.

The FCA imposes civil liability on any person who “knowingly presents ... to an officer or employee of the United States Government ... a false or fraudulent claim for payment or approval” or “knowingly makes [or] uses ... a false record or statement to get a false or fraudulent claim paid or approved by the Government.” 12 In statutory terms, Mayfield’s argument is that submission of the BAFO to NASA constituted the fraudulent presentment of a claim under section 3729(a)(1), and that Lockheed subsequently used the 533 forms completed pursuant to the research contract to fraudulently obtain award fees in violation of section 3729(a)(2).

Mayfield’s argument that Lockheed intentionally undervalued the BAFO is essentially a fraud-in-the-inducement claim. The archetypal fraud-in-the-inducement case is United States ex rel. Marcus v. Ness, 13 in which the Supreme Court held that FCA liability was triggered when contractors colluded to inflate artificially the bid price on a public-works project, thereby inducing the government to pay more than it otherwise would have. 14 Even though the contractor’s subsequent claims for payment made under the contract were not literally false, since they derived from the original fraudulent misrepresentation, they, too, became actionable false claims. 15

Hess stands for the proposition that, to succeed on a fraud-in-the-inducement theory under the FCA, Mayfield must prove (1) that Lockheed had no intention to perform the research contract according to the terms of the BAFO; and (2) that Lockheed obtained payments under the research contract to which it was not legitimately entitled. 16 Mayfield has not creat *260 ed a genuine issue of material fact as to either element.

Without more, a contract underbid is not a false claim. 17 For FCA liability, there must be a nexus between the underbid and a request for payment that the contractor would not have been entitled to absent the contract. 18 That nexus is absent here.

The record shows, and Mayfield doesn’t deny, that the research contract was doomed to run over-budget from the start. The terms of the RFP required that all bidders design their proposals using the hypothetical skill mix NASA demanded and, accordingly, Lockheed’s projected labor costs were derived from NASA’s model. The research contract mandated roughly equivalent projects to those included in the engineering contract, but included a lower skill mix. Put otherwise, the research contract was designed in part to accomplish the same tasks as the engineering contract using lower-cost, lower-skilled labor. Lockheed willingly complied with these optimistic — perhaps unrealistic — requirements and submitted a bargain-basement-priced bid. Lockheed concedes that it “hedged its bets by bidding the lower cost elicited by the RFP.”

Mayfield infers incorrectly from this admission that Lockheed “underbid the [research contract] even though it intended to charge NASA its real cost from the very beginning.” This statement — and May-field’s entire theory of fraud-in-the-inducement liability — ignores the realities involved in projecting labor costs under CPAF-type contracts. Under contracts like the research contract, labor requirements over the life of the contract are determined by the government, not the contractor. 19 The central purpose of a CPAF-type contract is to afford the government flexibility in requisitioning labor in an unpredictable environment. 20

The inherent variability in the research contract’s structure was reflected by Lockheed and NASA’s joint behavior. On the one hand, Lockheed’s low projected labor costs were quickly outstripped by the actual skill mixes required to perform research contract tasks; on the other, NASA’s tinkering with the Space-Station Option and other tasks requiring highly skilled workers exacerbated Lockheed’s cost overruns. Indeed, NASA admitted in a Cost Performance Evaluation that 40% of the initial cost overruns were “government directed.” Mayfield’s argument is essentially the erroneous proposition that a contractor’s FCA liability can arise as a result of the government’s behavior during the performance of a CPAF-type contract. This assertion is antithetical to fraud and not surprisingly lacks support in the caselaw.

Neither can Mayfield demonstrate that Lockheed used the research contract to dupe NASA out of payments that it could not have obtained absent the contract. This case is unlike United States ex rel. *261 Mayman v. Martin Marietta Corp., 21 where an intentional underbid allowed the contractor to recoup the shortfall on the contract value via a separate R&D agreement with the government unrelated to the original contract. This “intent to recover additional monies above the contract price” was the key to the court’s finding of FCA liability. 22

Here, in contrast, there are no analogous extra-contractual payments Mayfield can point to indicating that Lockheed finagled fees it had no right to under the research contract. The payments Lockheed received were exclusively research contract-based, a product of the contract’s Award-Fee Plan.

Ill

Nor are the 533 cost projections actionable. Mayfield argues that even if Lockheed’s RFP bid was not a knowing misrepresentation of its intent to perform, Lockheed still fraudulently manipulated the data reported in the 533 forms so that the research contract cost projections would meet NASA’s targeted cost at completion. This argument fails because there is no nexus between the 533s and Lockheed’s award fees, as the 533s were only cost projections that did not independently entitle Lockheed to award payments under the research contract’s fee-award system. In short, the 533s were immaterial.

The civil False Claims Act “interdicts material misrepresentations made to qualify for government privileges or services.” 23 No majority decision of this circuit has addressed the proper standard for assessing the materiality of a false statement under the FCA’s civil-liability provisions, although at least five judges of this court suggested that false statements, in order to be material, must “actually affect” the government’s decision to pay. 24 The United States as amicus urges a different standard, suggesting that a material statement need only have “potentially affected” the decision to pay.

Under either standard, the 533 cost projections are immaterial here. The 533 projections were not submitted nor used for the purpose of calculating Lockheed’s award fee payment, rather, by their terms they were for planning purposes only. The NASA contracting officers testified accordingly, explaining that the award fee is based only on Lockheed’s meeting contractual requirements, which is determined solely by NASA, looking only to Lockheed’s past performance, not to their cost projections. Indeed, the research contract itself provides that the 533 projections are non-binding.

Mayfield insists that the 533 projections were material to Lockheed’s award. He directs us to the deposition of Gail Skow-ron, who, he says, testified that incorrect 533 projections would affect Lockheed’s award fee “to some degree.” This over *262 states Skowron’s testimony. Skowron did not testify that the 533 projections were material to the award fee. Rather, she merely acknowledged that Lockheed had a contractual duty to provide these 533 projections, and that Lockheed’s failure to provide accurate projections might be characterized as non-compliance with a contract term. Such non-compliance, Mayfield argues, was material to the award fee.

Mayfield has lost sight of his target. There is no allegation before us that Lockheed lied about whether it was in compliance with the contract, a lie that would have been material to the fee award, or so its seems. Instead, Mayfield alleges that Lockheed lied in its 533 cost projections. It is undisputed that these cost projections could not have affected (or potentially affected) Lockheed’s award fee. NASA independently reviewed Lockheed’s progress and unilaterally determined the fee Lockheed earned during a given period independent from the cost projections Lockheed submitted on the 533s. The 533s were planning tools that predicted future performance. They reflected cost overruns that were built into the research contract from the start. Viewed in the best light for Mayfield, the evidence suggests that the 533 reports were negligent miscalculations of the true labor requirements demanded by the research contract. Unreasonable or incorrect cost projections, however, cannot fend off summary judgment. 25 Neither the original bid nor the 533s were actionable false claims.

Mayfield’s remaining theory is that for nine months Lockheed billed NASA for an office building, LP4, that was not contemplated by the research contract, concealing from NASA the fact that LP4 was actually excess capacity. The alleged overpayments total $10,902,978.

LP4 was a building leased by Lockheed under the old engineering contract. The government terminated the engineering contract before its expiration date as it was a level-of-effort contract that had exhausted the allocated hours before the contract period ended. With this early ending Lockheed was stuck with the building because it had leased LP4 for the entire term of the contract. At that point, since this was a termination of convenience, Lockheed could have claimed reimbursement for the remaining lease period as well as other charges. Instead, as Mayfield himself testified, the Lockheed chairman struck a deal with the Deputy Director of the Johnson Space Center. The Deputy Director agreed that NASA would reimburse Lockheed for the LP4 lease under the research contract if it was awarded to Lockheed. Indeed, in its very first bill, Lockheed pointed out the LP4 deal, noting that “pursuant to oral guidance during the negotiation,” we are billing in excess of costs.

Under these unchallenged facts, Lockheed did not “knowingly” present a “false” claim. 26 Most of our sister circuits have recognized that “[wjhere the government and a contractor have been working together, albeit outside the written provisions of the contract, to reach a common *263 solution to a problem, no claim arises.” 27 Or, stated more broadly, “[i]f the government knows and approves of the particulars of a claim for payment before that claim is presented, the presenter cannot be said to have knowingly presented a fraudulent or false claim.” 28

Nor is Mayfield alleging collusion between a government agency and a contractor. 29 According to his own testimony, Mayfield’s claim is based on a separate agreement negotiated between the chairman of Lockheed and a high-ranking government official, an agreement not made to enrich Lockheed at the expense of the taxpayer, but, in essence, for the satisfaction of a debt. Any fallout from this agreement is a matter caught by contract law, not the punitive provisions of the FCA. Mayfield has not created a genuine fact issue on the question of whether Lockheed’s claims for the LP4 lease were knowing or false. Nor has he done so for the bid and 533s. The judgment of the district court is

AFFIRMED.

. NASA terminated the engineering contract ahead of schedule because it had consumed the labor hours it purchased. If Lockheed's bid succeeded, it might be able to dovetail its preexisting engineering contract work structure at the Johnson Space Center into production for the potentially lucrative research contract.

. See FAR §§ 16.305; 16.405.2(a).

. Id. § 16.405-2(a).

. The only immutable limitation was the research contract's "Limitation of Funds” clause, which provided that costs would be limited according to the constraints of the NASA budgetary process. Both parties anticipated, correctly, that NASA’s funding amount would increase yearly. See FAR § 16.306(b)(1) (CPAF contracts are appropriate in open-ended R&D-type situations).

.A "wrap rate” is calculated by dividing the total cost of the contract (excluding incidentals like travel, fees, and burdened material) by the number of man-hours ordered by NASA. The figures reported by Lockheed in its BAFO provided for 13,411,995 labor hours at a cost of $425,341,412.20, yielding a wrap rate of $31.71/hour. Mayfield alleges that that submission, inter alia, is a false claim because at the time Lockheed knew that the actual wrap rate needed to perform under the research contract was actually $39.46/hour— 27.5% higher than the bid wrap rate, meaning *258 that Lockheed underbid the research contract by around $114 million.

. It seems that NASA’s practice was to renegotiate the contract and allow for an increase in the total contract value. NASA, however, refused to renegotiate the research contract’s baseline costs, instead ordering Lockheed to tighten its belt.

. See Mayfield v. Lockheed Eng’g & Sci. Co., 970 S.W.2d 185 (Tex.App.1998) ("Mayfield I”).

. See United States ex rel. Mayfield v. Lockheed Martin Eng’g & Scis. Co., 186 F.Supp.2d 711 (S.D.Tex.2002); see 31 U.S.C. § 3730(e)(4) (disallowing federal jurisdiction in any false-claims action based upon public disclosure of the transactions in a civil dispute unless the action is brought by the “original source” of the information).

. See United States ex rel. Laird v. Lockheed Martin Eng’g & Sci. Servs. Co., 336 F.3d 346 (5th Cir.2003) (“Mayfield II”).

. 31 U.S.C. § 3729(a)(7).

. The Supreme Court’s recent opinion in Rockwell International v. United States, - U.S. -, 127 S.Ct. 1397, 167 L.Ed.2d 190 (2007) held that the statutory phrase “information on which the allegations are based” refers to the information on which the relator’s allegations are based rather titan the information on which the publicly disclosed allegations are based. The Court rejected "contrary conclusion of some lower courts,” citing our prior decision in this case United *259 States ex rel. Laird v. Lockheed Martin Eng. & Science Servs. Co., 336 F.3d 346, 354 (5th Cir.2003). Although Rockwell overrules our prior precedent, Mayfield has standing nevertheless. It is plain from the record and the parties’ concessions at oral argument that Mayfield has direct and independent knowledge of the information on which the allegations are based.

. 31 U.S.C. §§ 3729(a)(l)-(2).

. 317 U.S. 537, 63 S.Ct. 379, 87 L.Ed. 443 (1943).

. Cf. United States ex rel. Schwedt v. Planning Research Corp., 59 F.3d 196, 199 (D.C.Cir.1995) (submission of false progress reports about the status of a computer project fraudulently induced the government to pay for extraneous components).

. See Hess, 317 U.S. at 543-44, 63 S.Ct. at 384. ("This fraud did not spend itself with the execution of the contract. Its taint entered into every swollen estimate which was the basic cause for payment of every dollar paid by the P.W.A. into the joint fund for the benefit of respondents. The initial fraudulent action and every step thereafter taken, pressed ever to the ultimate goal — payment of government money to persons who had caused it to be defrauded.”).

. See United States ex rel. Willard v. Humana Health Plan of Tex., Inc., 336 F.3d 375, 384 (5lh Cir.2003); see also United States ex rel. *260 Bettis v. Odebrecht Contractors of Cal., Inc., 393 F.3d 1321, 1325 (D.C.Cir.2005).

. See Chicago Coll. of Osteopathic Med. v. George A. Fuller, Co., 719 F.2d 1335, 1347-48 (7th Cir.1983).

. See Harrison v. Westinghouse Savannah River Co., 176 F.3d 776, 787-88 (4th Cir.1999).

. See FAR § 16.306(b)(2); Bettis, 393 F.3d at 1325.

. See FAR § 16.301-2 ("[C]ost-reimbursement contracts are suitable for use only when uncertainties involved in contract performance do not permit costs to be estimated with sufficient accuracy to use any type of fixed-price contract.").

. 894 F.Supp. 218 (D.Md.1995).

. Bettis, 297 F.Supp.2d at 282 n. 17; see also Schwedt, 59 F.3d at 199 (fraudulent progress reports induced government to purchase programs it otherwise would have declined).

. United States ex rel. Thompson v. Columbia/HCA Healthcare Corp., 125 F.3d 899, 902 (5th Cir.1997) (emphasis added); see also United States v. Southland Management Corp., 326 F.3d 669, 679 n. 3 (5th Cir.2003) (Jones, J., concurring); United States ex rel. Wilkins v. N. Am. Constr. Corp., 173 F.Supp.2d 601, 618-30 (S.D.Tex.2001) (Rosenthal, J.); United States v. Data Translation, Inc., 984 F.2d 1256, 1267 (1st Cir.1992) (Breyer, C.J.).

. United States v. Southland Management Corp., 326 F.3d 669, 679 n. 3 (5th Cir.2003) (Jones, J., concurring).

. See United States ex rel. Butler v. Hughes Helicopters, Inc., 71 F.3d 321, 326 (9th Cir.1995) (later modifications in derogation of original contract demands might allow a breach-of-contract claim, but do not give rise to FCA liability); United States ex rel. Hopper v. Anton, 91 F.3d 1261, 1267 (9th Cir.1996) (negligent misrepresentations or systems errors are not actionable false statements under the FCA).

. 31 U.S.C. § 3729(a)(1); 31 U.S.C. § 3729(b).

. Southland Management Corp., 326 F.3d at 682; United States ex rel. Becker v. Westinghouse Savannah River Co., 305 F.3d 284, 288-89 (4th Cir.2002); United States ex rel. Lamers v. City of Green Bay, 168 F.3d 1013, 1019-20 (7th Cir.1999); United States ex rel. Butler v. Hughes Helicopters, Inc., 71 F.3d 321, 326-27 (9th Cir.1995).

. United States ex rel. Durcholz v. FKW, Inc., 189 F.3d 542, 545 (7th Cir.1999).

. See Southland Management Corp., 326 F.3d at 682 n. 8; Lamers, 168 F.3d 1013 at 1019-20.