On December 24, 2014, the United States District Court for the Eastern District of Virginia set aside a $101 million False Claims Act (FCA) jury verdict against Gosselin World Wide Moving, N.V. (Gosselin).1 The Court held that FCA liability cannot be imposed based on anticompetitive conduct alone because the FCA is not an all-purpose anti-fraud statute, and liability requires deception or deceit in connection with a specific, identifiable claim for government payment. The decision also provides important guidance on the suitability of using statistical modeling to estimate damages under the FCA.
The U.S. Government contracts with American freight carriers to transport the household goods of U.S. servicemen and women to and from Germany.2 The American carriers in turn contract with local agents, such as Gosselin, to move the goods to final destinations within Germany once they reach Europe. In November 2000, in response to a drastic drop in shipping rates, Gosselin and other local agents (comprising 70% of the local agent market) met to discuss how to address this pricing pressure. They signed an agreement3 to work under a bundled rate for all services necessary to transport the goods, the purpose of which was to raise the prices that carriers paid for their services. The local agents agreed that they would no longer contract directly with the American carriers, but rather would work through Gosselin or another company.
Government's FCA Theory
Originally filed by a qui tam relator, the United States intervened in this action as to all claims relating to the military shipping program. The government argued that Gosselin violated the FCA4by engaging in a fraudulent course of anticompetitive conduct (i.e., the bundled rate agreement) that inflated the shipping rates paid to all carriers under every shipping contract from 2001-2002. The government did not contend, nor was any evidence presented at trial, that Gosselin made any false statements or certifications, express or implied, or that Gosselin failed to comply with any contract provisions, statutes, or regulations that were a term or condition for payment. Instead, the government argued that Gosselin's anticompetitive conduct alone was sufficient to establish liability under the FCA and that all claims paid were "factually false" because they were based on a shipping rate that was "artificially inflated beyond a competitive price."5
The jury returned a $101 million verdict, finding Gosselin knowingly caused 58,950 false claims. The verdict included $33.6 million in compensatory damages. Defendants filed a motion for judgment as a matter of law and alternatively moved for a new trial.
The Court granted defendants' motion for judgment as a matter of law and conditionally granted the motion for a new trial, in the event that the Court's primary holding is vacated or reversed on appeal. There were three key components to the Court's holding.
First, the Court addressed the government's theory that FCA liability may be premised on anticompetitive conduct alone,6 calling this theory "unprecedented and untenable." Because the evidence did not show that Gosselin made any false statement or failed to comply with any term or condition of payment, nor did it show that a specific, identifiable claim for payment by a specific carrier was inflated because of Gosselin's conduct, the verdict must be set aside. The Court rejected the government's assertion that the seminal case of Marcus v. Hess, 317 U.S. 537 (1943) and its progeny supported its theory of liability, finding each case the government cited involved "an element of deception that is absent in this case." In contrast here, Gosselin openly and explicitly informed the carriers of the bundled rate agreement.
The Court further rejected the idea that liability could be imposed because Gosselin was aware of the government's "expectations" that the military shipping program would be "fair and competitive [and] free of collusion." Those "expectations" were not embodied in any contractual, regulatory, or statutory term or condition for payment and "it is incumbent upon the government to be clear as to precisely what is expected of those involved in the procurement process." It would be patently unfair to premise FCA liability on the government's "post-hoc, general, and vague declarations of what was necessary to preserve the 'integrity' of the procurement process."7
Second, the Court found that the government failed to produce evidence sufficient to satisfy the FCA's required materiality element. Gosselin's conduct was not material to the government's awards of military shipping contracts because it did not pertain to any term or condition of payment to the carriers that submitted claims for payment. The government's own lack of requirements made this eminently clear: "the government made the decision not to require the disclosure of such conduct or to obtain representations or certifications that allowed it to assume such conduct did not exist; and in this sense, the government itself has defined, through the scope of its required certifications, what competition related information was material for its purposes."
Finally, the Court set aside the compensatory damages award, finding the statistical model used to calculate damages unreliable under Daubert. Without questioning the expert's credentials or the suitability of regression modeling in certain cases, the Court found that the model used to calculate damages in this case had several flaws.8 The Court first questioned whether regression analysis could be used to predict prices set under the "opaque, unusual and complex price setting mechanism" for carrier shipping rates which were not dictated by traditional market forces like cost and demand. The Court was also unable to conclude that the model accounted for all of the major explanatory factors, noting the absence of key variables like ocean shipping rates, which accounted for approximately 30% of costs. Finally, the model's inability to accurately predict shipping rates during the benchmark period called into question its reliability.
The Court's decision serves as support for the proposition that liability under the FCA must be tethered to the presence of identifiable claims for payment submitted to the government. As the Court pointed out, the FCA is neither an antitrust statute nor an all-purpose anti-fraud statute. Instead, it is "a fraud-based statute that requires deception or deceit in connection with" a false claim for government payment. This decision may support arguments to defeat broad interpretations of FCA liability advanced by the government or private relators.
The Court's decision is also notable for its treatment of statistical modeling used to extrapolate damages under the FCA. In recent decisions, federal courts have been receptive to the use of statistical techniques to not only calculate damages, but also to establish FCA liability.9 As private relators and the government continue to pursue broader theories of FCA liability, these types of statistical techniques will become increasingly prevalent. It is helpful, therefore, that the Court's analysis provides guidance for challenging the reliability of statistical techniques. For example, the case may support arguments that statistical models are unsuitable for extrapolating damages in cases involving markets that are not driven by traditional economic measures such as supply and demand. The case may also provide grounds for attacking the reliability of statistical models under Daubertwhere modeling lacks key explanatory variables or is incapable of accurately forecasting real-world outcomes.