General Dynamics Corp. v. United States

Justice Scalia delivered the opinion of the Court. We consider what remedy is proper when, to protect state secrets, a court dismisses a Government contractor’s prima facie valid affirmative defense to the Government’s allegations of contractual breach. I In 1988, the Navy awarded petitioners a $4.8 billion fixed-price contract to research and develop the A-12 Avenger carrier-based, stealth aircraft. The A-12 proved unexpectedly difficult to design and manufacture, and by De *481 cember 1990, petitioners were almost two years behind schedule and spending $120 to $150 million each month to develop the A-12. Petitioners informed the Government that the cost of completing the contract would exceed the contract price by an “'unacceptable’” amount. McDonnell Douglas Corp. v. United States, 567 F. 3d 1340, 1343 (CA Fed. 2009); see McDonnell Douglas Corp. v. United States, 182 F. 3d 1319, 1323 (GA Fed. 1999). They proposed restructuring the contract as a cost-reimbursement agreement and offered to absorb a $1.5 billion loss. The Department of Defense had lost faith in the project, however, and Rear Admiral William Morris, the Navy’s contracting officer for the A-12 agreement, terminated the contract for default on January 7, 1991. By that point, petitioners had spent $3.88 billion attempting to develop the A-12, and the Government had provided $2.68 billion in progress payments. A few weeks after terminating the contract, the Navy sent petitioners a letter demanding the return of approximately $1.35 billion in progress payments for work never accepted by the Government. The parties later entered into a deferred payment agreement covering this amount. Petitioners filed suit in the Court of Federal Claims (CFC) to challenge Admiral Morris’s termination decision under the Contract Disputes Act of 1978, 92 Stat. 2388, as amended, 41 U. S. C. § 609(a)(1). The Federal Circuit has recognized a governmental obligation not to mislead contractors about, or silently withhold, its “superior knowledge” of difficult-to-diseover information “vital” to contractual performance. GAF Corp. v. United States, 932 F. 2d 947, 949 (1991). Petitioners asserted that the Government’s failure to share its “superior knowledge” about how to design and manufacture stealth aircraft excused their default (and also asserted other claims not relevant here). Uncovering the extent of the Government’s prior experience with stealth technology proved difficult. The design, *482 materials, and manufacturing process for two prior stealth aircraft operated by the Air Force — the B-2 and the F-117A — are some of the Government’s most closely guarded military secrets. “‘[N]eed-to-know’ or [special] access controls beyond those normally provided for access to Confidential, Secret, or Top Secret information” apply. 32 CFR § 154.3(x) (2010); see App. 384-385. The Government nevertheless granted 10 members of petitioners’ litigation team “access to the Secret/Special Access level of the B-2 and F-117A programs.” Id., at 385. Four of those ten individuals received access to even the most sensitive aspects of the programs. See ibid. That neither satisfied petitioners’ thirst for discovery nor prevented the unauthorized disclosure of military secrets. In March 1993, Acting Secretary of the Air Force Michael Donley asserted the state-secrets privilege to bar discovery into certain aspects of stealth technology beyond petitioners’ “need-to-know” authorizations. At a deposition that month, a former Navy official’s responses to questions by petitioners and the Government revealed military secrets neither side’s litigation team was authorized to know. Copies of the unclassified deposition were widely distributed and quoted in unsealed court filings until Government security officials discovered the breach a month later. A July 1993 deposition caused further unauthorized disclosures of military secrets. These disclosures led Acting Secretary of the Air Force Merrill McPeak to file a declaration with the CFC. He warned that further discovery into the extent of the Government’s superior knowledge “would present a continuing threat of disclosure of . . . military and state secrets” surrounding the “weight, profile or signature, and materials involved in the design and construction of ‘stealt[h]’ . . . aircraft and weapons systems.” Id., at 633, 635. Even relatively straightforward and innocuous questions, in his opinion, “would pose unacceptable risks of disclosure of classified, special access information,” id., at 636, including the *483 potential disclosure of covert Government programs, id., at 637. The CFC took Secretary McPeak’s concerns seriously and terminated discovery relating to superior knowledge. It later decided that the extent of the Government’s superior knowledge was a nonjusticiable question. Both sides had enough evidence to “present a persuasive case” on the superior-knowledge issue, but the CFC worried that, “wit[h] numerous layers of potentially dispositive facts” hidden by the privilege, its superior-knowledge rulings “would be a sham,” McDonnell Douglas Corp. v. United States, 37 Fed. Cl. 270, 280, 284-285 (1996), and one that would threaten national security, see id., at 281-282. In 1996, for reasons not relevant here, the CFC converted the termination into a less-Government-friendly termination for convenience and awarded petitioners $1.2 billion. McDonnell Douglas Corp. v. United States, 35 Fed. Cl. 358. The Federal Circuit reversed, 182 F. 3d, at 1332, and left it for the CFC to reconsider on remand whether the need to protect military secrets precluded discovery into the superior-knowledge issue, id., at 1329-1330. After a 6-week trial, the CFC sustained the default termination, McDonnell Douglas Corp. v. United States, 50 Fed. Cl. 311, 326 (2001), and reaffirmed that the parties could not safely litigate whether the Government’s superior knowledge excused petitioners’ default, id., at 325. The Court of Appeals reversed the default termination, but agreed that the state-secrets privilege prevented adjudicating whether the Government’s superior knowledge excused the default. See McDonnell Douglas Corp. v. United States, 323 F. 3d 1006, 1024 (CA Fed. 2003). It rejected petitioners’ assertion that the Government could not pursue a claim against a party and then use the state-secrets privilege to completely pre-empt defenses to that claim; the Court of Appeals believed United States v. Reynolds, 345 U. S. 1, 12 (1953), had already “rejected” this “very argument.” 323 F. 3d, at 1023. Litigants cannot complain, the Court of Appeals held, *484 when the state-secrets privilege trumps a defense “in [a] purely civil matter, suing the sovereign on the limited terms to which it has consented.” Ibid. On remand, the CFC again found petitioners had defaulted. McDonnell Douglas Corp. v. United States, 76 Fed. Cl. 385, 430 (2007). The Court of Appeals affirmed, see 567 F. 3d, at 1356, and we granted certiorari to review its state-secrets holding, 561 U. S. 1057 (2010). II Many of the Government’s efforts to protect our national security are well known. It publicly acknowledges the size of our military, the location of our military bases, and the names of our ambassadors to Moscow and Beijing. But protecting our national security sometimes requires keeping information about our military, intelligence, and diplomatic efforts secret. See Haig v. Agee, 453 U. S. 280, 307 (1981); Martin v. Mott, 12 Wheat. 19, 30-31 (1827). We have recognized the sometimes-compelling necessity of governmental secrecy by acknowledging a Government privilege against court-ordered disclosure of state and military secrets. In Reynolds, three civilian contractors died during a test flight of a B-29 bomber. Their widows filed wrongful-death suits against the Government and sought discovery of the Air Force’s accident-investigation report. Federal discovery rules, then as now, did not require production of documents protected by an evidentiary privilege. See 345 U. S., at 6; Fed. Rule Civ. Proc. 26(b)(1). We held that documents that would disclose state secrets enjoyed such a privilege; the state-secrets privilege, we said, had a “well established” pedigree “in the law of evidence.” 345 U. S., at 6-7. The penultimate paragraph of Reynolds rejected the widows’ assertion that if the Government invoked the state-secrets privilege it had to abandon the claim to which the thereby privileged evidence was relevant. That was, the widows observed, the price paid in criminal cases. If the *485 Government refuses to provide state-secret information that the accused reasonably asserts is necessary to his defense, the prosecution must be dismissed. See id., at 12; Jencks v. United States, 353 U. S. 657, 672 (1957). The penultimate paragraph of Reynolds said that this was a false analogy. A like abandonment of the Government’s claim is not the consequence “in a civil forum where the Government is not the moving party, but is a defendant only on terms to which it has consented.” 345 U. S., at 12. Both petitioners and the Court of Appeals rely upon this statement to support their differing positions. We think that Reynolds has less to do with these cases than the parties believe — and its dictum (of course), less still. Reynolds was about the admission of evidence. It decided a purely evidentiary dispute by applying evidentiary rules: The privileged information is excluded, and the trial goes on without it. That was to the detriment, of course, of the widows, whom the evidence would have favored. But the Court did not order judgment in favor of the Government. Here, by contrast, the CFC decreed the substantive result that since invocation of the state-secrets privilege obscured too many of the facts relevant to the superior-knowledge defense, the issue of that defense was nonjusticiable, and the defense thus not available. See 37 Fed. Cl., at 284-285. And that was so even though petitioners had brought forward enough unprivileged evidence to “make a prima facie showing.” Id., at 280. While we disagree, for reasons set forth below, with the CFC’s disposition of the remainder of the case, its perception that in the present context the state-secrets issue raises something quite different from a mere evidentiary point seems to us sound. What we are called upon to exercise is not our power to determine the procedural rules of evidence, but our common-law authority to fashion contractual remedies in Government-contracting disputes. See Priebe & Sons, Inc. v. United States, 332 U. S. 407, 411 (1947). And *486 our state-secrets jurisprudence bearing upon that authority is not Reynolds, but two cases dealing with alleged contracts to spy. In Totten v. United States, 92 U. S. 105 (1876), the administrator of a self-styled Civil War spy’s estate brought a breach-of-contract suit against the United States. He alleged that his testator had entered into a contract with President Lincoln to spy on the Confederacy in exchange for $200 a month. After the war ended, the United States reimbursed expenses but did not pay the monthly salary. We recognized that the estate had a potentially valid breach-of-contract claim but dismissed the suit. The contract was for “a secret service,” and litigating the details of that service would risk exposing secret operations and other clandestine operatives “to the serious detriment of the public.” Id., at 106-107. “[Pjublic policy,” we held, “forbids the maintenance of any suit... the trial of which would inevitably lead to the disclosure of matters which the law itself regards as confidential, and respecting which it will not allow the confidence to be violated.” Id., at 107. Six years ago, we reaffirmed that “public policy forb[ids]” suits “based on covert espionage agreements.” Tenet v. Doe, 544 U. S. 1,3 (2005). Such suits threaten to undermine ongoing intelligence-gathering and covert operations — two vital aspects of national security — through inadvertent exposure of espionage relationships. Id., at 11. Rather than tempt fate, we leave the parties to an espionage agreement where we found them the day they filed suit. We think a similar situation obtains here, and that the same consequence should follow. Where liability depends upon the validity of a plausible superior-knowledge defense, and when full litigation of that defense “would inevitably lead to the disclosure of" state secrets, Totten, supra, at 107, neither party can obtain judicial relief. As the CFC concluded, that is the situation here. Disclosure of state secrets occurred twice before the CFC terminated discovery. See *487 37 Fed. CL, at 277-278. Every document request or question to a witness would risk further disclosure, since both sides have an incentive to probe up to the boundaries of state secrets. State secrets can also be indirectly disclosed. Each assertion of the privilege can provide another clue about the Government’s covert programs or capabilities. See Fitzgerald v. Penthouse International, Ltd., 776 F. 2d 1236, 1243, and n. 10 (CA4 1985). For instance, the fact that the Government had to continue asserting the privilege after granting petitioners access to B-2 and F-117A program information suggests it had other, possibly covert, stealth programs in the 1980’s and early 1990’s. It seems to us unrealistic to separate, as the CFC did, the claim from the defense, and to allow the former to proceed while the latter is barred. It is claims and defenses together that establish the justification, or lack of justification, for judicial relief; and when public policy precludes judicial intervention for the one it should preclude judicial intervention for the other as well. * If, in Totten, it had been the Government seeking return of funds that the estate claimed had been received in payment for espionage activities, it would have been the height of injustice to deny the defense because of the Government’s invocation of state-secret protection, but to maintain jurisdiction over the Government’s claim and award it judgment. Judicial refusal to enforce promises .contrary to public policy (here, the Government’s alleged promise to provide superior knowledge, which we could not determine was breached without penetrating several layers of state secrets) is not unknown to the common law, and the traditional course is to leave the parties where they stood when they knocked on the courthouse door. *488 “In general, if a court will not, on grounds of public policy, aid a promisee by enforcing the promise, it will not aid him by granting him restitution for performance that he has rendered in return for the unenforceable promise. Neither will it aid the promisor by allowing a claim in restitution for performance that he has rendered under the unenforceable promise. It will simply leave both parties as it finds them, even though this may result in one of them retaining a benefit that he has received as a result of the transaction.” 2 Restatement (Second) of Contracts § 197, Comment a, p. 71 (1979); see, e. g., Worlton v. Davis, 78 Idaho 217, 222-223, 249 P. 2d 810, 814 (1952). These cases differ from the common-law cases that we know, in that the unenforceability did not exist at the time the contract was formed, see 2 Restatement (Second) of Contracts §179, Comment d, at 18, but arose because of the Government’s assertion of the state-secrets privilege that rendered the promise of superior knowledge unadjudicable. We do not see why that should affect the remedy. Suit on the contract, or for performance rendered or funds paid under the contract, will not lie, and the parties will be left where they are. The law of contracts contains another doctrine that relates to the CFC’s concern about the reliability of its judgment “without numerous layers of potentially dispositive facts,” 37 Fed. Cl., at 284-285. The Statute of Frauds, which has been with us since the 17th century, reflects concerns about the reliability of oral evidence. See Valdez Fisheries Development Assn., Inc. v. Alyeska Pipeline Serv. Co., 45 P. 3d 657, 669 (Alaska 2002); 9 R. Lord, Williston on Contracts §21:1, pp. 170-172 (4th ed. 1999 and 2010 Supp.). It assumes a valid, enforceable agreement between the parties but nevertheless leaves them without a remedy absent reliable evidence — a writing. See 1 id., § 1:21, at 82 (4th ed. 2007 and 2010 Supp.); 9 id., §21:5, at 192. So also here, it is preferable to leave the parties without a remedy rather than risk the *489 “potential injustice,” Valdez Fisheries, supra, at 669, of misjudging the superior-knowledge issue based on a distorted evidentiary record. The Government suggested at oral argument that where the parties stood at the time of suit was that petitioners had been held in default, liable for the ensuing consequences. See Tr. of Oral Arg. 48-49; see also Brief for United States 32, n. 9, 34-35. That had been the declaration of the contracting officer, pursuant to Chapter 9 (entitled “Contract Disputes”) of Title 41 (entitled “Public Contracts”). See 41 U. S. C. § 605. It was “final and conclusive . . . unless an appeal or suit is timely commenced.” § 605(b). We regard that, however, as merely one step in the contractual regime to which the parties had agreed. It has no more bearing upon the question we are discussing than would a provision in a private contract that declaration of default by one of the parties is final unless contested in court. The “position of the parties” in which we will leave them is not their position with regard to legal burdens and the legal consequences of contract-related determinations, but with regard to possession of funds and property. Ill Neither side will be entirely happy with the resolution we reach today. General Dynamics (but not Boeing) wants us to convert the termination into one for convenience and reinstate the CFC’s $1.2 billion damages award. See Brief for Petitioner in No. 09-1298, pp. 58-61. The language of the A-12 agreement does not give us that option. It authorizes a court to convert a default termination into a termination for convenience only if it “determine^] that the Contractor was not in default, or that the default was excusable.” 48 CFR §52.249-9(g) (2010). Our opinion does not express a view on those issues. It holds them nonjusticiable. Moreover, state secrets would make it impossible to calculate petitioners’ damages. A termination for convenience *490 ordinarily entitles a contractor to recover its incurred costs of performance, reasonable termination expenses, and a reasonable profit on the work performed (or an offset to account for the contractor’s expected losses had the contract been performed to completion). See § 52.249-2(g). The CFC’s $1.2 billion award to petitioners in 1996 simply reflected their actual costs incurred minus progress payments received. The CFC decided it could not calculate petitioners’ expected losses (or profits) without deciding the extent to which the Government’s alleged failure to share its superior knowledge contributed to petitioners’ cost overruns — a nonjusticiable question. See 37 Fed. CL, at 285. Absent proof of the Government’s superior knowledge, and of how the sharing of that would have made this a profitable contract, the $1.2 billion award might represent an undeserved windfall. The Government, for its part, wants a return of the $1.35 billion it paid petitioners in progress payments for work which it says it never approved. But the validity of that claim depends upon whether petitioners are in default on their contract. If they are not, termination for convenience of the Government would entitle them to retain those progress payments (unless, of course, they would have incurred a loss on the entire contract). Neither the question whether they are in default nor the question whether performance of the entire contract would have left them with a loss can be judicially determined because of the valid assertion of the state-secrets privilege. We leave the parties where they are. As in Totten, see 92 U. S., at 106, our refusal to enforce this contract captures what the ex ante expectations of the parties were or reasonably ought to have been. Both parties “must have understood,” ibid., that state secrets would prevent courts from resolving many possible disputes under the A-12 agreement. The Government asked petitioners to develop an aircraft the design, materials, and manufacturing process for which would be closely guarded military secrets. See Contract *491 Schedule H-l, App. 73-75; Contract Security Classified Specification, id., at 129-135. The contract itself was a classified document at one point. See Contract Schedule H-l, ¶8, id., at 75. Both parties — the Government no less than petitioners — must have assumed the risk that state secrets would prevent the adjudication of claims of inadequate performance. We believe, moreover, that the impact of our ruling on these particular eases (which we think produces rough, very rough, equity) is probably much more significant than its impact in future cases, except to the extent that it renders the law more predictable and hence more subject to accommodation by contracting parties. They can negotiate, for example, the timing and amount of progress payments to account for the possibility that state secrets may ultimately render the contract unenforceable. The Government’s concern that contractors will raise frivolous superior-knowledge defenses designed to goad the Government into asserting the state-secrets privilege is misplaced. To begin with, the rule we announce today applies only when the superior-knowledge defense is supported by enough evidence to make out a prima facie case. Moreover, Government contractors — especially cutting-edge defense contractors of the sort likely to operate in the state-secrets field — are repeat players. Even apart from the judicial sanctions available to punish bad conduct, see Fed. Rules Civ. Proe. 11, 26(g), they have strong incentive to behave rather than risk missing out on the next multibillion-dollar defense contract. And finally, while we anticipate that the rule we set forth will ordinarily control Government-contracting disputes that become nonjusticiable because of state secrets, what we promulgate today is not a statute but a common-law opinion, which, after the fashion of the common law, is subject to further refinement where relevant factors significantly different from those before us here counsel a different outcome. *492 The foregoing analysis assumes that the Government generally has an obligation to share its superior knowledge, see GAF Corp., 932 F. 2d, at 949; the parties have not challenged that assumption. The Government argued below, however, that it does not have that obligation with respect to “highly classified information,” and does not have it when (as was the case here) the agreement specifically identifies information that must be shared. Brief for United States 52. The Court of Appeals did not address those questions (it had no reason to, given its disposition of petitioners’ appeals), and we did not grant certiorari to decide them. Those issues (and whether they can safely be litigated without endangering state secrets) therefore remain for the Court of Appeals to address on remand. * * * In Reynolds, we warned that the state-secrets evidentiary privilege “is not to be lightly invoked.” 345 U. S., at 7. Courts should be even more^hesitant to declare a Government contract unenforceable because of state secrets. It is the option of last resort, available in a very narrow set of circumstances. Our decision today clarifies the consequences of its use only where it precludes a valid defense in Government-contracting disputes, and only where both sides have enough evidence to survive summary judgment but too many of the relevant facts remain obscured by the state-secrets privilege to enable a reliable judgment. We vacate the judgment of the Court of Appeals and remand the cases for further proceedings consistent with this opinion. It is so ordered. Footnotes [*] Of course, this does not mean the nonjusticiability of one aspect of a ease will necessarily end the entire litigation. If, for example, the Government asserts two justifications for its default termination, and if state secrets deprive the contractor of a prima facie valid defense to only one of those claims, the eourt can still adjudicate the validity of the other.