The opinion of the court was delivered by: Savage, J.
In this action presenting issues relating to the False Claims Act ("FCA")*fn1 statute of limitations that have not been decided by the Third Circuit Court of Appeals and have divided other circuit and district courts, we hold that the tolling provision in § 3731(b)(2) does not apply to a relator when the government has not intervened, and the limitations period in 31 U.S.C. § 3731(b)(1) is triggered by the earlier filing of the claim rather than the later payment. The decision with respect to the triggering point is confined to the peculiar framework of the federal grant program.
Plaintiff Robert P. Bauchwitz ("Bauchwitz") alleges that the defendants William K. Holloman ("Holloman"), Eric B. Kmiec ("Kmiec"), Cornell University Medical College ("Cornell University"), and Thomas Jefferson University ("Thomas Jefferson"), misrepresented the findings of their DNA research when they applied for National Institute of Health ("NIH")*fn2 research grants and did not correct the misrepresentations on subsequent progress reports and renewal applications. These misrepresentations, Bauchwitz asserts, resulted in false claims in violation of the FCA.
Central to this case is scientific research performed by Holloman and Kmiec involving the identity of the gene, protein and activity within the cell of the fungus Ustilago maydis ("U. maydis").Holloman is a scientific researcher in Cornell University's Department of Microbiology. Kmiec had collaborated with Holloman as a graduate student from 1979-1984 at the University of Florida where he worked under Holloman's supervision, and as a scientific researcher in Thomas Jefferson University's Department of Pharmacology and at the Kimmel Cancer Center where he assisted Holloman on research partly funded by NIH grants from 1991 to 1999. When he was a student in the early 1980's, he and Holloman co-authored several articles in which they claimed to be the first to isolate a protein associated with nucleated cells from the eucaryotic fungus U. maydis.*fn3 They called the protein "Rec1," and claimed that the Rec1 protein was derived from the REC1 gene and had activity that allowed for the specific recombination of strands of DNA, called recombinase.*fn4 In the DNA research field, this was a significant revelation.
Bauchwitz worked in Holloman's lab from 1987-1990 as a graduate student. At that time, Bauchwitz and others working at Holloman's lab isolated the genes for REC1 and REC2, and obtained sequences from them. The results were that the REC1 DNA sequence would not specify the protein elements common to known DNA recombinases, but the REC2 gene had a sequence consistent with a recombinase.
After Kmiec left Holloman's lab, they continued to collaborate and co-author articles. One of those articles, "The Rec2 gene encodes the homologous pairing protein of Ustilago maydis"("the 1994 Article"), was published in the November 1994 issue of Molecular and Cellular Biology.*fn5 The authors claimed that the protein they had isolated from Ustilago maydis was a Rec2 protein, not a Rec1 protein, which was derived from the REC2 gene.
After reading the 1994 article sometime around November 28, 1994, Bauchwitz had suspicions about the reported findings. He had already been dubious of Holloman and Kmiec's research. His doubts dated back to when he had worked in Holloman's lab in the late 1980s. At that time, Bauchwitz suspected, for several reasons, that Holloman was engaging in scientific misconduct. First, no one else in Holloman's lab was able to replicate Kmiec and Holloman's findings that the Rec1 protein was derived from the REC1 gene and had recombinase activity. Second, when Holloman was having difficulty obtaining approval of grant applications, another graduate student in his lab was able to achieve results*fn6 that were contradicted by prior results in other labs as well as Bauchwitz's initial findings, which enabled Holloman to obtain funding. Third, Bauchwitz's own sequencing results obtained in the late 1980's on the REC1 and REC2 genes were not consistent with Holloman and Kmiec's earlier findings linking the Rec1 protein activity to the REC1 gene, and instead supported a link between the Rec1 protein and the REC2 gene.
His suspicions were so profound that after he left the lab, he continued to follow Holloman's work. In late 1990, he informed Dr. Ken Berns, Chairman of the Cornell University Graduate School of Medical Sciences' Department of Microbiology, that, over his objections, Holloman had removed unfavorable data from a manuscript that they had co-authored and that Holloman had forged Bauchwitz's initials indicating that he had approved the manuscript. Also, in 1990, he urged the Office of Scientific Integrity ("OSI")*fn7 to investigate the accuracy of Holloman's reported findings regarding the Rec1 protein activity.
After reading the 1994 article, in which the individual defendants claimed that the protein they had isolated from U. maydis was derived from the REC2, not REC1 gene, Bauchwitz suspected, even more strongly than he had previously, that Holloman and Kmiec had falsified their findings. He believed that the irreproducibility of the Rec1 protein research in the 1980s, as well as Bauchwitz's sequencing results, which were inconsistent with Holloman and Kmiec's findings linking the Rec1 protein activity to the REC1 gene, motivated Holloman and Kmiec to find a way to transform the disputed Rec1 protein activity into Rec2 protein activity. Bauchwitz alleges that the 1994 article contained two false statements to support the authors' new "Rec1 is Rec2" theory. Between December of 1994 and February of 1995, he pursued his own investigation by contacting current and former colleagues of Holloman, and current and former graduate students who worked in Holloman's lab.
As part of his investigation, Bauchwitz contacted ORI on February 6, 1995 to learn the status of the government's investigation of the defendants instigated by his first call to ORI in 1990, and to give ORI additional information based on his December 1994 phone call with Brian Rubin, a graduate student who succeeded Bauchwitz at Holloman's lab. The parties do not agree whether Bauchwitz identified himself on this call and whether he provided ORI with any details regarding alleged fraud at Holloman's laboratory.*fn8
Bauchwitz's own research, which he began in 1991 and completed in 1999, failed to replicate Kmiec and Holloman's results. He concluded that the "Rec1 is Rec2" U. maydis research was false and the defendants used it to obtain NIH grants. But, he argues that he did not reach this conclusion until 2002, after he read another article authored by Holloman*fn9 ("the 2001 article") that he believed falsified the rec2-1 mutant gene sequence to appear capable of producing protein. Bauchwitz maintains that he did not learn that the defendants obtained any of the specific grant funding at issue in this case*fn10 until September 23 and November 8, 2002, when he received responses to his FOIA requests that his counsel had submitted on June 10, 2002.*fn11
The NIH Grants at Issue and the False Statements
Bauchwitz is alleging that grant applications, progress reports and financial status reports ("FSRs") submitted to NIH by the defendants between October 31, 1991 and September 19, 2006 contained false*fn12 or misleading statements pertaining to three categories of false statements.*fn13 Specifically, he accuses the defendants of (1) fabrication and/or falsification of the amino acid sequence of the Rec1 protein; (2) falsification of the DNA sequence of the rec2-1 mutant gene; and (3) falsification of data images relating to Rec2 protein activity. With respect to all defendants, he contends that the false statements were published in, or based in substantial part upon, conclusions or findings reported in the 1994 article (categories 1 and 3). With respect to the Cornell defendants only, he contends that grant 2 RO1 GM42482-12A2 contains false statements by citing the 2001 article (category 2).*fn14 These false statements, according to Bauchwitz, were material to the NIH grants at issue in this case.*fn15
A grant, a form of federal assistance, is an exercise of Congress' spending power.
31 U.S.C. § 6501(4)(A)-(B) (2003). A grant is defined as money paid or provided by the United States to an entity for a specified purpose. 31 U.S.C.A. § 6501(4)(A). Unlike a contract where the government pays for goods or services, a grant is monetary assistance to a non-federal entity authorized by statute to meet needs that Congress deems in the public interest. Id.
The grants in this case are discretionary ones that were awarded on a competitive basis.*fn16 NIH announces opportunities for funding for specific topics or research goals. To obtain these funds, entities submit "competing new applications."*fn17 The applications must meet legislative and regulatory requirements, and published selection criteria established for the particular program. After conducting a formal review process that includes peer review of the competing applications, the agency determines which applications best address the program requirements and are most worthy of funding.*fn18
When an agency awards a grant, it issues a Notice of Grant Award ("NGA"). The NGA sets out the terms, the project period,*fn19 the total project amount, the amount authorized for each year, the annual budget, and the budget period.*fn20 To accept the grant, the recipient either signs and returns the NGA, or draws funds from the designated Department of Health and Human Services ("HHS") payment system. Funds are committed and available on the day the grant is awarded. Once the NGA is signed or money is drawn, the NGA and the grant terms are binding on the grantee and the government.*fn21
A grant may be approved for a multi-year period, known as the project period. Under the project period system of funding, when an initial NGA is awarded, the project is programmatically approved for support in its entirety, but is funded in annual increments known as budget periods. Funds for each subsequent budget period are paid on a non-competitive basis provided funds are available, the grantee has achieved satisfactory progress and the grant continues to be in the best interests of the government.*fn22 An NGA constitutes an "obligation,"*fn23 and the agency is required to submit its annual budget proposal to Congress, which includes the agency's "obligations."*fn24 Thus, the agency is only committed to funding the grant for the current budget period due to dependency upon the annual Congressional appropriations process. In this way, although the initial NGA declares the granting agency's intention to "provide continued financial support throughout the life of the grant," the projections of future funding levels are not guarantees that the project will be funded beyond the end date of the current budget period as shown in the NGA.*fn25
To obtain an NGA for a subsequent budget period under a multi-year grant, a grantee must submit an annual progress report known as a "non-competing continuation application"*fn26 The progress report, which must be submitted two months before the beginning date of the next budget period, requires a description of: (1) the progress made over the past year; (2) any significant balance of funds to be carried over to the next budget period and how they will be spent if permitted to be carried forward; and (3) any significant changes in key personnel expected during the next budget period. For a non-SNAP grant,*fn27 it must also contain a detailed budget for the next budget period.*fn28 Once funds for a subsequent budget period are awarded, a new NGA will be issued setting the new budget period and the amount of new funding that the government is obligated to provide.
In addition, a Financial Status Report ("FSR") must be submitted at the end of each twelve-month budget period in the case of certain grants, and at the end of a competitive segment in a project period in non-SNAP grants.*fn29 The FSR is a budget reconciliation form submitted to the government sixty to ninety days after the end of the report period and is an accounting of how the grantee spent the funds it received during the report period. It is not a report of results of the research project and is not a request for payment.*fn30
Once the project period has ended, the grant is either renewed or closed. 45 C.F.R. §§ 74.71-74.73. To renew grants, grantees must submit a "competing continuation application." Renewal applications compete in the same manner as initial grant applications.*fn31 If the grantee chooses not to renew or the grant is terminated, the grantee must file a final FSR*fn32 and a final progress report. The final FSR covers the entire project period. Like the annual one, the final FSR is a budget reconciliation form submitted after the end of the project period that does not report results and does not request payment. The final progress report includes, at a minimum, a summary of progress in light of the originally stated goals, a list of significant results and a list of publications.*fn33
Bauchwitz filed his original complaint under seal on June 30, 2004. The government investigated the case while the complaint remained under seal. At the request of the United States Attorney's Office, ORI conducted a scientific review of the allegations set forth in Bauchwitz's complaint. Because the research at issue had taken place so many years earlier and because it did not view the statements at issue as intentionally false, ORI concluded that it did "not believe that evidence is available" to prove that any of the three claims alleged by Bauchwitz are false.*fn34 On August 31, 2005, after its fourth motion for an extension was denied, the government elected not to intervene.
On April 19, 2006, because Bauchwitz did not prosecute the action for over seven months after the government had declined to intervene, the action was dismissed. On April 19, 2007, the dismissal was vacated upon Bauchwitz's motion. An amended complaint was filed on May 16, 2007, adding claims based on allegedly false statements made in connection with a third federal grant submitted by the Thomas Jefferson defendants. It also "identifie[d]... the specific statements made by each... defendant[ ] that the Relator contends were false or fraudulent, as well as the specific grant applications and progress reports in which such false or fraudulent statements were made."*fn35
The defendants moved to dismiss the amended complaint. The grounds included plaintiff's failure to state a claim pursuant to Rule 12(b)(6) because he is alleging a scientific dispute and not a fraud, lack of subject matter jurisdiction because plaintiff is not an "original source" of the support for his allegations, and the claims are barred by the statute of limitations. Pursuant to notice, the motion to dismiss was converted to a motion for summary judgment. The parties then conducted discovery limited to the issue of the statute of limitations.
Summary Judgment Standard
Summary judgment is appropriate if there is no genuine issue as to any material fact and "if, viewing the facts in the light most favorable to the non-moving party, the moving party is entitled to judgment as a matter of law." Ebbert v. DaimlerChrysler Corp., 319 F.3d 103, 108 (3d Cir. 2003) (citing Fed. R. Civ. P. 56(c)). Because the statute of limitations is an affirmative defense, the defendants bear the burden of proving that the statute of limitations bars Bauchwitz's claims. Richard B. Roush, Inc. Profit Sharing Plan v. New England Mut. Life Ins. Co., 311 F.3d 581, 585 (3d Cir. 2002). Thus, because the defendants are moving for summary judgment on statute of limitations grounds, the burden of demonstrating that there are no genuine issues of material fact begins and remains with them. Ebbert, 319 F.3d at 108.
Although all reasonable inferences are drawn in the non-movant's favor, Intervest, Inc. v. Bloomberg, L.P., 340 F.3d 144, 159-60 (3d Cir. 2003), an inference based upon speculation or conjecture does not create a material fact. Robertson v. Allied Signal, Inc., 914 F.2d 360, 382 n.12 (3d Cir. 1990). Thus, "[w]here the record taken as a whole could not lead a rational trier of fact to find for the non-moving party, there is no 'genuine issue for trial.'" Matsushita Elec. Indus. Co. v. Zenith Radio Corp., 475 U.S. 574, 587 (1986) (citation omitted).
The False Claims Act Statute of Limitations
The FCA prohibits "any person from making false or fraudulent claims for payment to the United States." Graham County Soil & Water Conservation Dist. v. United States ex rel. Wilson, 545 U.S. 409, 411 (2005); 31 U.S.C. § 3729(a). Any person found liable for violating the FCA is subject to a civil penalty of $5,000 to $10,000 per violation and treble damages. 31 U.S.C.A. § 3729(a) (West Supp. 2008); Hutchins v. Wilentz, Goldman & Spitzer, 253 F.3d 176, 181 (3d Cir. 2001).
An action under the FCA may be commenced in one of two ways. The attorney general may sue on behalf of the United States government; or, a private individual, known as a relator, can bring a qui tam action. 31 U.S.C.A. § 3730(a), (b)(1); Graham County, 545 U.S. at 411-12 (citing Vermont Agency of Natural Res. v. United States ex rel. Stevens, 529 U.S. 765, 769-72 (2000)). Because the relator brings the action on behalf of the government, he must give the government notice of the action. The government has sixty days from the filing of a qui tam complaint to elect to intervene in the action, and, for good cause shown, can petition the court to permit it to intervene at a later date. Graham County, 545 U.S. at 412; § 3730(b)(2) and (c)(3).
A civil action under the FCA must be brought within six years of the violation or within three years of the date when the government learned or should have learned the facts material to the violation, whichever is later. Id. §§ (b)(1), (2). In no event may an action be brought after ten years of a violation. Id. Specifically, the FCA statute of limitations provides:
(b) A civil action under [the False Claims Act] may not be brought -
(1) more than 6 years after the date on which the violation of [the False Claims Act] is committed, or
(2) more than 3 years after the date when facts material to the right of action are known or reasonably should have been known by the official of the United States charged with responsibility to act in the circumstances, but in no event more than 10 years after the date on which the violation is committed, whichever occurs last.
31 U.S.C.A. § 3731(b) (2003).
The critical difference between § (b)(1) and (b)(2) is that under § (b)(1), the statute of limitations begins to run when the violation occurs, whereas under § (b)(2), it begins to run when the appropriate person learned or should have learned facts putting him on notice that a violation occurred. A conflict arises from the interplay between the unusual procedure allowing a private party to bring a qui tam action on behalf of the government and the language of the tolling provision, which appears to relate only to the government. It is this conflict that raises the issues confronting us in this case.
The Tolling Provision - 31 U.S.C. § 3731(b)(2)
The three-year tolling provision permits suit to be brought after the six-year period where the fraud was not discovered during or until late in that period. In most cases, the three-year discovery period expires within six years of the violation. In that event, subsection (b)(1), with its longer limitations period, applies. If the fraud is discovered early in the six-year period, subsection (b)(2) will not be implicated. For example, if the fraud is discovered within one year of the violation, the three-year tolling period does not come into play because the six-year period in § (b)(1) would apply. Where the fraud is not discovered until after six years or late in the six-year period, subsection (b)(2) extends the limitations period. If the fraud is not discovered until seven years after the violation, the limitations period is extended for three years after the discovery. If it is discovered five years afterwards, the period is extended three years, effectively moving the limitations bar to eight years after the violation.
In determining whether Bauchwitz's claims are timely, we must answer three questions. First, when did the violation occur with respect to each grant to trigger the running of the applicable limitations period under § 3731(b)(1)? Second, does § 3731(b)(2), the tolling provision, apply to private relators when the government has not intervened? Third, if it does, when does the limitation period start running - when the relator learned of the violation or when the government did?
Accrual of Action Under § 3731(b)(1)
In applying § 3731(b)(1), the FCA speaks of a "violation." Is the violation the filing of the claim or is it the payment? There is a lack of unanimity as to whether the statute of limitations begins to run when the false claim is filed or when the government pays the claim. Compare United States ex rel. Karvelas v. Melrose-Wakefield Hosp., 360 F.3d 220, 225 (1st Cir. 2004) (stating that the "statute attaches liability, not to the underlying fraudulent activity or to the government's wrongful payment, but to the 'claim for payment'") (quoting United States v. Rivera, 55 F.3d 703, 709 (1st Cir. 1995)), with Jana, Inc. v. United States, 41 Fed. Cl. 735, 742-43 (Fed. Cl. 1998) (stating that if the government makes payment on a submitted false claim, the FCA violation occurs on the date payment was made, rather than on the date the claim was submitted).*fn36
Section 3729 does not define the words "false claim." It does define "claim" as "any request or demand... for money." 31 U.S.C. § 3729(c). Setting out what constitutes a violation of the FCA, it reads: "any person who... (1) knowingly presents, or causes to be presented, to an officer... of the United States Government... a false or fraudulent claim for payment or approval; (2) knowingly makes, uses or causes to be made or used, a false record or statement to get a false or fraudulent claim paid or approved by the Government...." 31 U.S.C. § 3729(a).
Bauchwitz argues that the six-year statute of limitations period begins to run on the date the government paid the claim. He contends that is when the final payment was made, which, he asserts, was at the end of each project period after the grantees submitted their final FSRs.*fn37 The defendants, on the other hand, argue that the statute of limitations begins to run at the time the grantees submitted the request for a grant*fn38 containing a false statement, not when the grant applications were approved and paid by government. According to the defendants, their position is consistent with the language of the FCA and the FCA's purpose of preventing fraud on the government by "attacking the activity that presents the risk of wrongful payment."*fn39
Both the plain language of § 3729(a) and statements made by the Supreme Court and the Third Circuit support the principle that the application for payment, rather than payment of the claim, triggers the accrual of an action. The language of § 3729(a) focuses on the means and not the end. Liability arises from the use of fraudulent submissions intended to cause the government to issue payment. The statute does not fix liability on the receipt of payment. In fact, payment is not a prerequisite to liability. Payment need only be sought or approved in reliance on the false representations. In other words, liability begins with the false statement that is intended to induce payment. See United States v. Neifert-White Co., 390 U.S. 228, 230 (1968).
The Supreme Court, in analyzing the applicability of § 3731(b)(1) to retaliation claims under § 3730(h) of the FCA, made clear that federal statutes of limitations start running when the cause of action accrues. Graham County, 545 U.S. at 418. Although it did not address the issue of whether the application for payment or the actual payment itself triggers the running of the limitations, it did use language that suggests that the period starts when the claim is made rather than when payment is issued. It said, "the language in § 3731(b)(1) [ties] the start of the time limit to 'the date on which the violation of section 3729 is committed.' In other words, the time limit begins to run on the date the defendant submitted a false claim for payment." Id. at 415. This language imparts that the cause of action accrues before payment and is keyed to the "claim for payment."
Although the Third Circuit has not ruled on the issue, it has intimated that the trigger date is when the claim is made. In United States ex rel. Malloy v. Telephonics Corp., 68 F. App'x 270, 273 (3d Cir. 2003), when applying the six-year limitation period under § 3731(b)(1), the court treated the date that the defendant made the claim, not the date of payment, as the start date for calculating the limitations period. The court noted, "Malloy concedes that the claim accrued when Telephonics filed the original false claim...." Id. The government attempts to minimize the import of the court's language by characterizing the statement as the relator's concession and not the court's position. Contrary to the government's argument, the court's analysis cannot be reduced to meaningless verbiage. Although it was not the court's holding, it is meaningful because the court would not have used that starting point in its analysis of the applicability of the statute of limitations merely because the relator did not challenge it. In other words, it would not have misapplied a legal principle even if the parties had. Otherwise, its entire analysis and the result would have been flawed. Therefore, the Malloy court's approach endorses, albeit implicitly, the principle that a § 3729 action accrues when the claim is made.
Prior to Malloy, the Third Circuit in Hutchins v. Wilentz, Goldman & Spitzer, 253 F.3d 176 (3d Cir. 2001), signaled that an FCA violation is complete at the time the claim is made. In considering what constitutes a false claim, it did not decide what established a violation for purposes of applying the statute of limitations. Nor did it rule out that submissions of false statements for approval of payment were false claims under the FCA. Indeed, the court held that the FCA "prohibits fraudulent claims that cause or would cause economic loss to the government." Hutchins, 253 F.3d at 179 (emphasis added).*fn40 See also United States ex rel. Sanders v. American-Amicable Life Ins. Co., 545 F.3d 256, 259 (3d Cir. 2008) (stating that the FCA "cover[s] instances of ...