concurring in Parts ILB and II.E., and dissenting in Parts II.C and II.D:
This court has twice thought it “passing strange” that “the prices charged to the government for specific goods could be confidential” commercial information or “trade secrets” under the Freedom of Information and Trade Secrets Acts. McDonnell Douglas Corp. v. NASA, 180 F.3d 303, 306 (D.C.Cir.1999); see McDonnell Douglas Corp. v. Widnall, 57 F.3d 1162, 1167 (D.C.Cir.1995). We have never decided that fundamental question, however, because the government has never litigated whether such prices are subject to those statutes — and hence to the National Parks test — in the first place.1 For the same reason, we do not do so in this case. Nonetheless, the analysis adopted and result reached in Parts II.C and II.D of the court’s opinion come perilously close to a per se rule that line-item prices — prices the government agrees to pay out of appropriated funds for goods or services provided by private contractors — may never be revealed to the public through a Freedom of Information Act (FOIA) request. Because, if such prices are covered by the statutes at all, a bar on their disclosure should be the exception rather than the rule, I respectfully dissent.
I. Vendor Pricing Contractor Line Item Numbers (CLINs)
In Part II.D of its opinion, the court denies the Air Force authority to disclose prices for certain line items in its contract with appellant McDonnell Douglas (referred to in the record by the name of its parent company, Boeing). These are not mere offer or bid prices; they are prices that the government agreed to pay, and that it did pay, for specified services that it purchased from the company. Disclosure of such information permits the public to evaluate whether the government is receiving value for taxpayer funds, or whether the contract is instead an instance of waste, fraud, or abuse of the public trust. (I hasten to add that no such allegations are even hinted at here.) Such disclosure thus comes within the core purpose of FOIA: to inform citizens about “what their government is up to.” United States Dep’t of Justice v. Reporters Comm. for Freedom of the Press, 489 U.S. 749, 773, 109 S.Ct. 1468, 1481, 103 L.Ed.2d 774 (1989); see NLRB v. Robbins Tire & Rubber Co., 437 U.S. 214, 242, 98 S.Ct. 2311, 2326, 57 L.Ed.2d 159 (1978) (“The basic purpose of FOIA is to ensure an informed citizenry, vital to the functioning of a democratic society, needed to check against corruption and to hold the governors accountable to the governed.”).
It is therefore not surprising that, even applying the National Parks test, the vast majority of courts have permitted the release of agreed-upon contract prices, including line-item prices. See Gregory H. McClure, The Treatment of Contract Prices Under the Trade Secrets Act and Freedom of Information Act Exemption D *1195Are Contract Prices Really Trade Secrets, 31 Pub. Cont. L.J. 185, 196-97 (2002). That is overwhelmingly the result in the district courts of this circuit. See Center for Public Integrity v. United States Dep’t of Energy, 191 F.Supp.2d 187, 194-95 (D.D.C.2002) (collecting cases). It is also the result reached by the only other circuits that have decided FOIA cases regarding such prices. See Pacific Architects & Eng’rs Inc. v. United States Dep’t of State, 906 F.2d, 1345, 1347-48 (9th Cir.1990); Acumenics Research & Tech. v. United States Dep’t of Justice, 843 F.2d 800, 808 (4th Cir.1988). This circuit has ruled on the releasability of line-item prices on only one previous occasion, and in that case decided that the prices could not be disclosed. See NASA, 180 F.3d at 307.2 As discussed in Part II below, however, that case did not consider facts and arguments like those raised by the Air Force here.
The dispositions reached by the majority of courts are unsurprising, not only in light of the core purpose of FOIA, but also because of the nature of the National Parks test and of the way we review its application. As the court correctly recites, disclosure of information of the kind at issue here is barred only if it is “likely ... to cause substantial harm to the competitive position of the person from whom the information was obtained.” National Parks & Conservation Ass’n v. Morton, 498 F.2d 765, 770 (D.C.Cir.1974). In the case of line-item prices, such harm will result only if a competitor is able to “reverse-engineer” from the winning bidder’s price to the sensitive strategic information upon which it is based, such as the contractor’s profit margin and cost data.3 The ability to “pinpoint precisely a rival’s pricing strategy” is obviously not required to satisfy this test, Op. at 1193, but the objector must demonstrate that a rival will be able to derive strategic information that is “likely” to cause it “substantial” competitive harm. National Parks, 498 F.2d at 770.
Moreover, the objector must accomplish this demonstration within the constraints of our scope of review. First, it is the opponent of disclosure — not the requester — who bears the burden of proving whether substantial competitive harm is likely to result. See Occidental Petroleum Corp. v. SEC, 873 F.2d 325, 342 (D.C.Cir.*11961989); National Parks & Conservation Ass’n v. Kleppe, 547 F.2d 673, 679 n. 20 (D.C.Cir.1976) (National Parks II). Second, in evaluating whether the objector has met that burden, we defer to the judgments of the agency, and may overturn them only if they are arbitrary or capricious. See CNA Fin. Corp. v. Donovan, 830 F.2d 1132, 1153-55 (D.C.Cir.1987). In particular, we must defer to the agency’s expert prediction regarding the likely impact of disclosure upon the marketplace. See id. at 1155-56 (holding that an agency’s forecast of “what likely would ensue upon release of information” is the type of judgment that “courts traditionally leave largely to agency expertise” and “need not be supported by record evidence”).
Applying National Parks and the appropriate standard of review, it is plain that the line-item prices at issue here do not qualify for protection from disclosure. The sole reason my colleagues offer for reaching the opposite conclusion is their acceptance of appellant’s argument that, because the disputed prices are “composed predominantly of the costs of materials and services it procures from other vendors,” disclosure of those prices would permit derivation of the percentage “by which McDonnell Douglas marks up the bids it receives from subcontractors.” Op. at 1190. That argument, in turn, rests on the validity of McDonnell Douglas’ assertion, in its appellate brief, that its competitor “most likely obtained quotations from the same vendors” with “the same or nearly the same prices” and could thus derive McDonnell Douglas’ mark-up simply by subtracting those known vendor prices from the line-item prices contained in the contract. Op. at 1190 (quoting McDonnell Douglas Br. at 20).
But McDonnell Douglas’ assumption that its competitor “likely” obtained the same price quotations from the same vendors is simply not sufficient to carry its burden of proof on the issue. The contractor’s letter to the Air Force makes clear that this is indeed nothing more than an assumption: “It is reasonable to assume,” McDonnell Douglas said, “that our competitors obtained similar pricing from various vendors to support these tasks.” McDonnell Douglas Letter to Air Force, Aug. 3, 1998, at J.A. 20 (emphasis added). That is the entirety of the contractor’s submission on this point. Id. The Air Force, however, directly contradicted that assumption, declaring that “it is not uncommon for a vendor to quote different prices for the same basic effort to different prime contractors.” Air Force Final Administrative Decision Letter, June 23, 2000, at J.A. 74 (hereinafter Final Decision Letter). Although my colleagues disparage this declaration — contending that the Air Force “provided no actual evidence, nor did it claim special knowledge based upon its experience, to support this proposition,” Op. at 1190 — such disparagement is unfair for two reasons.
First, in context, the Air Force’s declaration appears to be precisely the “actual evidence” based on considerable “experience” with such procurement contracts that my colleagues require. In contrast to McDonnell Douglas’ acknowledgment that its assertion was based on nothing more than a “reasonable assumption],” the Air Force’s statement that “it is not uncommon” for vendors to quote different prices reads as a declaration of empirical fact. Cf. CNA 830 F.2d at 1155 (upholding agency response to objector’s affidavits, despite the agency’s lack of “independent evidence” to support its determination).
Second, and equally important, my colleagues’ approach stands the burden of proof on its head. Because identical vendor pricing is the premise of McDonnell Douglas’ argument that it will be harmed by disclosure, it is the contractor’s burden *1197to prove that such pricing is a fact — not the government’s burden to disprove it. See National Parks II, 547 F.2d at 679 n. 20. There is no such evidence in McDonnell Douglas’ submissions below; instead, the company offers precisely the type of “conclusory and generalized allegations” of harm that we have previously found “unacceptable.” Public Citizen Health Research Group v. FDA, 704 F.2d 1280, 1291 (D.C.Cir.1983). Yet it is the Air Force that my colleagues upbraid for this lacuna, not the party that bears the burden and responsibility for filling it.4
Lacking empirical support in the record, the opinion of the court turns to economic theory to support its conclusion: “In a competitive market for subcontracted work, a rational subcontractor would quote each prime contractor the lowest price consistent with covering its costs.” Op. at 1191. This is a theory of the court’s own invention: McDonnell Douglas did not make the argument in its submissions to either the Air Force or this court. Nor is there any basis for my colleagues’ assumption that there is, in fact, “a competitive market for subcontracted” military procurements. Indeed, all indications are to the contrary.5 The market at issue here is hardly one in which a myriad of buyers face a myriad of sellers. After all, the line-item prices we are considering are not for supplying widgets, but for overhauling landing gear and maintaining auxiliary power units and engine thrust reversers on military aircraft whose “primary mission ... is aerial refueling of other military aircraft.” Air Force Br. at 2 n.2; •see Final Decision Letter, at J.A. 73-74. We know that there were only two competitors for the prime contract, id. at J.A. 72, and McDonnell Douglas advised the district court that there are only “a limited number of subcontractors who can perform the work required in these types of contracts,” McDonnell Douglas Corp. v. United States Dep’t of Air Force, 215 F.Supp.2d 200, 208 (D.D.C.2002). Indeed, that was the basis for the company’s assertion that “competitors will have likely gone to the same subcontractors” and obtained the same prices. Id. In short, competitors for the prime contract did not obtain their prices from a faceless market, but rather from the same, relatively few vendors.
In this kind of imperfect, oligopolistic/ol-igopsonistic market, we cannot assume that the court’s “rational subcontractor” theory of identical prices would hold. In fact, unlike McDonnell Douglas, the Air Force did address the theoretical question and explain why differential rather than identical pricing was not uncommon: “Ongoing relationships between the vendor and different primes for other business,” the government said in its Final Decision Letter, “could affect the vendor’s decision *1198on the price to charge.” J.A. 74. This opinion regarding the nature of the market — one in which subcontractors align with and provide better prices to their preferred primes to,ensure other subcontracts — is surely plausible.6 And in light of the Air Force’s expertise and experience in this area, it is one to which we must defer. See CNA, 830 F.2d at 1155 (citing Federal Power Comm’n v. Transcontinental Gas Pipe Line Corp., 365 U.S. 1, 81 5.Ct. 435, 5 L.Ed.2d 377 (1961), and FCC v. National Citizens Comm. for Broad., 436 U.S. 775, 98 S.Ct. 2096, 56 L.Ed.2d 697 (1978)).
In . short, unless we reverse the burden of proof and deny the Air Force the deference it is owed, there is no basis for overturning its conclusion that disclosure of the prices it paid for McDonnell Douglas’ services is unlikely to cause substantial harm to the contractor’s competitive position.
II. Option Year Prices
Today’s opinion also bars the Air Force from disclosing line-item prices for option years contained in the contract with McDonnell Douglas. Op. at Part II. C. It does so based solely upon McDonnell Douglas’ argument that: “[/]% the event the Air Force does decide to rebid the contract, its competitors will be able to use that information to underbid it.” Id. at 7 (emphasis added). But McDonnell Doug-, las’' contention that it will be hurt “in the event” the Air Force decides to put the contract out for rebid is insufficient to meet appellant’s burden of showing that disclosure is “likely” to cause “substantial harm” to its competitive position. National Parks, 498 F.2d at 770. Since McDonnell Douglas does not contend that it will be injured if the contract is not rebid — i.e., if the government exercises the options — to prevail the contractor must establish that it is at least likely that there will be a rebid. This is just another way of restating the threshold requirement of our National Parks test: that the contractor must “actually face competition.” National Parks II, 547 F.2d at 679; accord Niagara Mohawk Power Corp. v. United States Dep’t of Energy, 169 F.3d 16, 18-19 (D.C.Cir.1999).7
McDonnell Douglas has wholly failed to satisfy that threshold requirement. The appellant proffered no evidence whatsoever regarding the probability of a reb'id. Indeed, it did not even assert that rebidding was likely. Rather, its submission to the Air Force said nothing more than: *1199“Should a recompetition occur, Boeing’s competitors, armed with its pricing for unexercised optionsf,] would be in a position to undercut Boeing’s prices.” McDonnell Douglas Letter to Air Force, Aug. 3, 1998, at J.A. 20-21 (emphasis added).
The government, by contrast, contends that rebidding is not likely. “[0]ption years of contracts,” the Air Force explains, “are usually exercised.” Air Force Br. at 19. And that is particularly so for contracts “to service military aircraft which are critical to USAF’s core mission.” Id. The Air Force’s contention is backed by its regulations, which instruct it to “take into account the Government’s need for. continuity of operations and potential costs of disrupting operations” in deciding whether or not to exercise an option. 48 C.F.R. § 17.207(e); see also id. § 17.202(d) (providing that options may be included in service contracts in “recognition of (1) the Government’s need in certain service contracts for continuity of operations and (2) the potential cost of disrupted support”).8 And it is further backed by the facts of this very case, as the Air Force “has exercised the past four option years for this contract each time they have come up.” Air Force Br. at 20.
Notwithstanding the logic and evidence supporting the government’s claim that McDonnell Douglas has failed to satisfy the threshold requirement for application of National Parks, my colleagues rightly conclude that we cannot resolve this appeal on that basis because, although the Air Force vigorously advances the claim here, it did not rely on it below. But while we cannot affirm the agency’s determination on that ground, we need not reverse it. When the government asserts a basis for decision that might well prevail if it were raised by the agency rather than its lawyers, we may remand to permit- the agency an opportunity to reconsider its rationale. See, e.g., Northeast Maryland Waste Disposal Auth. v. EPA, 358 F.3d 936, 949-50 (D.C.Cir.2004); Yukon-Kuskokwim Health Corp. v. NLRB, 234 F.3d 714, 718 (D.C.Cir.2000). Hence, even, if the government’s entire argument rested on the limited likelihood of a rebid, nothing more than a remand for further consideration would be in order.
The Air Force’s rationale for permitting disclosure does not, however, rest only on McDonnell Douglas’ failure to establish that rebidding is likely to occur. It also rests on appellant’s failure to prove it likely that it will suffer substantial competitive harm even if the contract were rebid. As noted, McDonnell Douglas’ theory is that it will be harmed-because, “in the event the Air Force does decide to rebid the contract, its competitors will be able to use [the options] information to underbid it.” Op. at 1188. The Air Force persuasively counters this theory by explicating its two significant flaws.
First, McDonnell Douglas has not demonstrated the likelihood that competitors would be able to use its option prices to underbid it. As McDonnell Douglas notes, if the Air Force were to decide not to exercise an option, it would conduct a new competition for the work. McDonnell Douglas Br. at 15. This would free McDonnell Douglas from the option agreements, and permit it to price the work differently. Recognizing this, appellant does not contend that its option prices would themselvés enable competitors to underbid it, but rather that competitors would be able to use that “pricing information to ‘deduce support hours, overhead factors, and profit factors for the op*1200tions’ ” — that is, they could use the option prices to reverse-engineer the factors that McDonnell Douglas would employ in pricing a new bid. Id. (quoting McDonnell Douglas Letter to Air Force, Sept. 3, 1998, at J.A. 27).
In response to McDonnell Douglas’ reverse-engineering argument, the Air Force offered a detailed exposition — on a line-item by line-item basis — explaining why “[i]t does not appear that release of the CLIN prices and the CLIN Option Year Matrix would reveal any confidential piece of information, such as a risk assessment or profit multiplier, that would place Boeing at a competitive disadvantage.” Final Decision Letter, at J.A. 71; see. id. at J.A. 72-76 (CLIN by CLIN analysis); id. at 76 (concluding that “many of the basic assumptions underlying Boeing’s arguments that a competitor can derive competitively sensitive rates and factors from Boeing’s CLIN prices are ... equally unfounded for the base year and the option years”).9 Moreover, the Air Force noted, “in a contract which contains options extending far into the future, in this case a total of nine years, the value of the option prices as a predictive tool for Boeing future pricing is significantly diminished since such option prices almost certainly include significant adjustments for risks of unknown contingencies.” Id.
The government’s explanation for rejecting McDonnell Douglas’ reverse-engineering claim is more than reasonable. Indeed, it is particularly powerful because, while McDonnell Douglas objects to disclosure of .the option-year line items, it has withdrawn its objection to disclosure of many of the same line items in the contract’s base year. Air Force Br. at 18 n.8. This leaves unexplained how a competitor’s knowledge of McDonnell Douglas’ option-year prices could materially improve its competitive position when the competitor already knows the prices appellant agreed to for the base year.
Given the reasonableness of the government’s rationale, and the fact that the burden of proof rests on McDonnell Douglas, we cannot overturn the Air Force’s decision without engaging in the same kind of line-item by line-item analysis that the Air Force did. The opinion of the court, however, does not undertake that task. Indeed, it does not address the Air Force’s reverse-engineering argument at all.10 Unless we are going to take the unprecedented step of adopting a per se rule that the release of line-item prices is always harmful and hence never permitted,11 the *1201failure to rebut the government’s arguments requires affirmance of its decision.
The government’s decision letter also exposes a second flaw in McDonnell Douglas’ competitive injury theory. Even if disclosure of option prices would significantly increase the probability that a competitor could submit a lower bid, McDonnell Douglas would not be substantially harmed by such a bid unless it were sufficiently attractive to win the contract away from the incumbent contractor. And in the circumstances of this case, the Air Force contends that such a possibility is unlikely because price “would be only one of the evaluation factors for award” of the new contract:
[I]t is important to remember that any Air Force recompetition for performance of KC-10 CLS for the option years would be conducted as a best value source selection, where price/cost would be only one of the evaluation factors for award, together with technical excellence and past performance. Based on the procedures followed in most of the other recent best value source selections conducted ... for [contractor logistical support] requirements, price/cost likely would be no more important an evaluation factor than both technical excellence and past performance. The fact that price/cost would not be the controlling evaluation factor or even the most significant evaluation factor in any such recompetition greatly reduces any potential competitive advantage a competitor could gain through release of Boeing’s option year CLIN prices and renders Boeing’s allegations of substantial competitive harm in this regard largely unfounded.
Final Decision Letter, at J.A. 76. In response to the government’s argument, McDonnell Douglas failed to show — or even to try to show — that disclosure of its option prices would enable competitors to submit bids at prices sufficiently low as to offset the advantages appellant gains from the other evaluation criteria.
My colleagues nonetheless dismiss the government’s argument on the ground that, because we previously rejected it in McDonnell Douglas v. NASA as “too silly to do other than to state it,” 180 F.3d at 306, we are bound to reject it again. Op. at 1189. But this is not the argument that we rejected in NASA. And it is not silly at all.
The argument that we rejected in NASA was the agency’s claim “that underbidding due to the disclosure would not occur because price is only one of the many factors used by the government in awarding contracts.” NASA, 180 F.3d at 306. It is indeed foolish to argue that, just because non-price factors are also relevant, a competitor would not try to bid a lower price than McDonnell Douglas. But the Air Force does not make NASA’s argument here. The Air Force’s argument is not that “underbidding ... would not occur,” but that underbidding is not likely to cause substantial harm — because competitors are not likely to be able to bid sufficiently below McDonnell Douglas’ price to overcome appellant’s nonpriee advantages. My colleagues render the two arguments the same only by assuming that the NASA court did not intend the word “underbidding” to have its ordinary meaning: “to bid less than (a competing bidder).” Mer-eiam-WebsteR’s Collegiate Dictionary 1287 (10th ed.1996). Instead, they insist that NASA intended the word to mean: “making a bid more attractive overall, not with respect only to price, so it will be chosen by the Government.” Op. at 1189. Nothing in NASA, however, supports this embellishment of the dictionary definition.
Nor is there anything silly about the argument that nonprice factors make it unlikely that disclosure of McDonnell *1202Douglas’ option prices will permit competitors to defeat it in a recompetition. As the opinion for the court acknowledges, the Air Force’s Request for Proposals (RFP) lists six criteria against which the government will evaluate bids — logistics, maintenance/repair/modifications, management, safety/fire protection, quality, and cost/ price — and states that each “will be weighted equally.” J.A.. 179.12 The court’s opinion goes on to discount the significance of this “best value” method of contractor selection on the ground that, “[bjecause price is the only objective, or at least readily quantified, criterion among the six criteria for awarding government contracts, submitting the lowest price is surely the most straightforward way for a competitor to show its bid is superior.” Op. at 1189. But whether or not low price is the most “straightforward” way to show superiority, the RFP makes clear that it will not be enough. So, too, does the entity that will make the final determination of superiority: The Air Force states in no uncertain terms — both as a matter of policy and of its experience with “most of the other recent best value source selections conducted” — that “price/cost would not be the controlling evaluation factor or even the most significant evaluation factor in any such recompetition.” Final Decision Letter, at J.A. 76. And while the governing statute requires the Air Force to consider price, 10 U.S.C. § 2305(a)(3)(A)(ii), it also expressly authorizes the Air Force to do just what the RFP and Final Decision Letter say it will do: consider nonprice factors as “approximately equal in importance to cost or price.” Id. § 2305(a) (3) (A) (iii) (II).
That the Air Force would prefer other factors over price is hardly surprising. The contract at issue here is not to supply cafeteria food, but to service planes that “will be flown by American military personnel on highly dangerous missions.” Air Force Br. at 23. One would hope that for such a contract, considerations of safety, quality, and confidence in an incumbent contractor would at least be the equal of price. And as we have just seen, one need not simply hope that were the case: It says so right in the RFP. My colleagues are therefore wrong to suggest that “[wjhether price will be but one of several factors to be weighted equally in any future RFP ... is necessarily somewhat speculative.” Op. at 1189. The Air Force has made clear — in both its RFP and Final Decision Letter — that price will be just such a factor, and that is a determination we should not second-guess. See CNA, 830 F.2d at 1155. Indeed, the only thing that is speculative is whether, in light of the announced evaluation criteria, a competitor would be likely to displace McDonnell Douglas as the prime contractor even if it had the advantage of seeing appellant’s option, prices. In the absence of even a proffer of evidence to support that conclusion, such speculation is plainly insufficient to satisfy McDonnell Douglas’ burden of proof. See Public Citizen, 704 F.2d at 1291.
In dismissing the government’s non-price factors argument and failing to address its reverse-engineering contention, my colleagues come perilously close to treating a contractor’s claim of “underbidding” as a talisman that bars disclosure of any line-item price — whether related to an option year or to a base year. But as the author of NASA made clear, that is an “overreading of our opinion,” which “did not” hold “that government disclosure of line item pricing would invariably violate *1203the Trade Secrets Act.” McDonnell Douglas Corp. v. NASA, No. 98-5251, slip op. at 2 (D.C.Cir. Oct.6, 1999) (Silberman, J., concurring in denial of rehearing en banc). Because McDonnell Douglas has failed to show that disclosure of option-year prices would violate the Act in this case, I would affirm the decision of the Air Force.
III. Conclusion
For the foregoing reasons, I conclude that McDonnell Douglas’ line-item and option-year prices fail to pass the National Parks test for nondisclosure, and that we therefore should respect the Air Force’s determination to release that information. To return to the point I made at the start, however, there remains the underlying question of whether the National Parks test is properly applied to agreed-upon prices (and at least exercised options) at all. That is an important question because — although reverse-engineering analysis applied under that test will often, as it does here, permit disclosure of agreed-upon prices — application of National Parks may bar disclosure of such prices in the very situation in which the public interest in disclosure is at its apogee.13
The archetypal case is that of the toilet seats that the Navy purchased for $640 apiece in the early 1980s.14 Whether the story is apocryphal or not,15 it serves to make the point: assuming that a low wholesale price for such seats is well-known, the only explanation for the high price paid by the government is the company’s profit margin. And notwithstanding that there would be a great public interest in determining “what the[] government is up to” in paying such a profit margin, Reporters Comm., 489 U.S. at 773, 109 S.Ct. at 1481,16 that is precisely the case in which a “hard to reverse-engineer” argument would fail and disclosure would be barred under National Parks.
This counter-intuitive result should cause us to think hard about whether it makes sense to regard prices actually paid by the government as trade secrets “of any person” under the Trade Secrets Act, 18 U.S.C. § 1905, or as confidential commercial or financial information “obtained from a person” under Exemption Four of FOIA, 5 U.S.C. § 552(b)(4). Cf. Restatement of *1204ToRTS § 757 emt. b (1939) (excluding “the amount or other terms of a secret bid for a contract” from the definition of “trade secret”); Public Citizen, 704 F.2d at 1291 n. 30 (noting that “competitive harm in the FOIA context” is limited to “harm flowing from the affirmative use of proprietary information” by competitors) (emphasis added; internal quotation marks omitted). It is indeed “passing strange” to regard an agency’s agreement to expend a specified amount of public funds as a corporate secret rather than a governmental decision — a category that is not encompassed by either statutory provision. National Parks, itself, did not concern disclosure of prices agreed to by the government, but rather of a firm’s internal financial information obtained in a government audit. 498 F.2d at 770. And it is not at all obvious that a test designed to evaluate the latter is appropriate for the former. See Reporters Comm., 489 U.S. at 773, 109 S.Ct. at 1481 (holding that disclosure of “[ojfficial information that sheds light on an agency’s performance of its statutory duties falls squarely within [FOIA’s] statutory purpose”). Moreover, even if agreed-upon prices do fall within Exemption Four, they may represent a case in which that exemption and the Trade Secrets Act should not be regarded as coextensive ■ — ■ and hence a case in which the government would have discretion to permit disclosure.17
But these are questions for another day. The only question for today, because it is the only question that the parties have litigated, is whether McDonnell Douglas has satisfied its burden of proving that the requested disclosures are likely to cause substantial harm to its competitive position. Because I conclude that appellant has failed to make that case for any of the information that the government has decided to release, I concur in my colleagues’ decision that the “over and above work” prices may be disclosed, but respectfully dissent from their determination that the vendor-pricing and option-year line items may not.
. See Widnall, 57 F.3d at 1167 ("Although the idea that a price charged to the government for specific goods or services could be a ‘trade secret’ appears passing strange to us, we agree with the government .that it is not open to us to attempt to decide that issue at this stage.... [T]he Air Force has never stated its position on McDonnell Douglas’ claim that even exercised option prices are trade secrets.”).
. The other two cases cited by my colleagues, Op. at 1193, are not apropos. Because of the "unusual posture” in which the Widnall case came to this court, we declined to rule on whether the prices at issue could be disclosed and instead remanded to the Air Force for clarification of its own position regarding release. 57 F.3d at 1167. In Gulf & Western Industries, Inc. v. United States, the requested disclosure was not of line-item contract prices, but rather of a government audit containing the objector's "profit rate” and "actual costs for units produced.” 615 F.2d 527, 529-30 (D.C.Cir.1979).
. See Pacific Architects, 906 F.2d at 1347-48 (affirming disclosure because a competitor would not be able to calculate the contractor's profit margin from its unit prices since there were too many “fluctuating variables”); Acumenics, 843 F.2d at 808 (affirming agency decision to disclose awarded unit prices because "there are too many unascertainable variables in the unit price calculation for a competitor to derive accurately [the contractor's profit] multiplier”); North Carolina Network for Animals, Inc. v. United States Dep’t of Agric., 924 F.2d 1052, 1991 WL 10757, at *3 (4th Cir. Feb.5, 1991) (unpublished table decision) (holding that release of sales and price information is required where it "reveals nothing useful about the dealer's pricing structure,” such as "the dealer's sources and costs of acquisition, customer information, [or] profit margin”); see also 1 James T. O’Reilly, Federal Information Disclosure § 14:83, at 705 (3d ed.2000) ("Where disclosure of contractor bid price details is at issue, the existence of variable factors that make it difficult to disaggregate a price figure would tend to support an agency denial of confidentiality for aggregate bid amount.”).
. I agree with the court that the Air Force's declaration regarding subcontractor pricing is one of "empirical fact.” Op. at 1190 n.4. What I disagree with is the court’s refusal, in the absence of further evidence, to accord that declaration any deference — notwithstanding the Air Force’s considerable experience with such matters and the fact that the burden lies on the opponent of disclosure. CNA, 830 F.2d at 1155, does not hold that such a declaration is unworthy of deference. Compare Op. at 1190 n.4. Nor does Occidental Petroleum, 873 F.2d at 342, support the proposition that the Air Force should bear the burden of production, or that the Air Force's declaration would be insufficient to satisfy that burden even if it should. Compare Op. at 1191 n.5.
. See Sherri Wasserman Goodman, Legal Dilemmas in the Weapons Acquisition Process: The Procurement of the SSN-688 Attack Submarine, 6 Yale L. & Pol'y Rev. 393, 395-96 (1988) (noting multiple reasons why defense procurements do not conform to the "classical economic model of a competitive market”); see also William E. Kovacic, Antitrust Analysis of Joint Ventures and Teaming Arrangements Involving Government Contractors, 58 Antitrust L.J. 1059, 1091-97 (1989).
. See, e.g., Robert Strauss & Joseph J. Dyer, Enforcement of Teaming Agreements, Procurement Lawyer, Fall 2001, at 5 ("Subcontractors ... are leery of devoting significant resources (including key engineering talent, intellectual property, and bid and proposal costs) without a formal relationship being established with prime contractors through teaming agreements.'1); U.S. Gen. Accounting Office, Report to the Subcommittee on Acquisition and Technology, Committee on Armed Services, U.S. Senate, Best Practices: DOD Can Help Suppliers Contribute More to Weapon Systems Programs (Mar. 1998) (encouraging the trend among defense prime contractors to select preferred suppliers for subcontracting work based on criteria other than lowest price, such as quality and past performance); see also Kovacic, 58 Antitrust L.J. at 1095-97 (describing the anticompetitive effect of the proliferation of prime and subcontractor cooperative relationships in the defense industry).
. In National Paries II, this, court held that, to establish that their commercial or financial information fell within Exemption Four, the appellant National Park concessioners would have "to prove that: (1) they actually face competition, and (2) substantial competitive . injury would likely result from disclosure.” 547 F.2d at 679. The court found that the concessioners did not face meaningful competition for the renewal of their park concessions because denials of renewal were rare, although they did face day-to-day competition from businesses located outside the park. Id. ‘ at 681-82.
. A contract solicitation may not even contain an option clause unless the "contracting officer has determined that there is a reasonable likelihood that the option will be exercised.” 48 C.F.R. § 17.208(c)(4).
. For example, the Air Force rejected McDonnell Douglas’ claim that a competitor could derive its hourly rate to perform cyclical (C-Check) inspections by dividing the line-item price by the known standard hours to perform the tasks, because (inter alia) títere are no standard hours for such inspections;— a point demonstrated by the "significant variance, from task to task, in the number of C-Check hours proposed.by the two competitors.” Final Decision Letter, at J.A. 72.
. There is one exception: The court affirms the Air Force’s rejection of McDonnell Douglas' claim that a competitor could derive its labor mark-up factors for "over and above work” with respect to the base year. Op. at Part II.E. Yet, the court does not explain why that determination should be any different with respect to the option years.
.As noted above, such a rule would be contrary to that adopted by other courts of appeals. See cases cited supra note 3. Gulf & Western, cited by my colleagues, is not to the contrary. In that case, a competitor would not have had to reverse-engineer anything: the sought-after information was not line-item contract prices, but rather government audits and annual financial statements that directly revealed the objector’s "profit rate,” along with its "actual costs for units produced, actual scrap rates, break-even point calculations and actual cost data.” 615 F.2d at 529-30 (internal quotation marks omitted).
. By contrast, there is no indication that the RFP at issue in NASA contained such a provision, without which "equal” consideration of non-price factors would not have been authorized. See 10 U.S.C. § 2305(a)(3)(A)(iii)(I)-(II).
. Although the FOIA requester in this case is a competitor of McDonnell Douglas, the same test — and the same result — applies when the requester is a watchdog organization dedicated to eliminating government waste. See National Archives & Records Admin. v. Favish, — U.S. —, —, 124 S.Ct. 1570, 1580, 158 L.Ed.2d 319 (2004) (noting that the decision to disclose “does not depend on the identity of the requester").
. See Mike Ward, It’s Your Information: How a Federal Law Has Turned Citizens into Giant Slayers, Austin American-Statesman, Oct. 6, 1996, at HI (noting that a FOIA request “uncovered the now-famous case” of the Pentagon’s “buying of gilt-priced toilet seats”). Such examples are not limited to military procurements. See, e.g., Justin Blum, Energy Contract Used to Repair D.C. Schools; No-Bid Agreement Pays Utility Millions, Wash. Post, Apr. 23, 2001, at A1 (noting that a FOIA request by the newspaper revealed that the D.C. public school system was charged exorbitant fees for cost-plus maintenance contracts).
. See Lawrence J. Korb, Editorial, Toilet Seats: Defense Replies, Wash. Post, Apr. 20, 1985, at A22 (stating that the $640 was for a "molded plastic cover for the entire lavatory,” not for each seat).
. See Michael Weisskopf, Radical Retooling to Be Urged For Pentagon Buying Machine: Commission Will Recommend Off-the-Shelf Shopping, Wash. Post, Feb. 22, 1986, at A2 (reporting that discovery of the $640 toilet seats led to recommendations for significant reform of the procurement process); Wayne Biddle, Price of Toilet Seat is Cut for Navy, N.Y. Times, Feb. 6, 1985, at D15 (detailing the contractor's decision to cut prices in response to the controversy).
. Cf. Widnall, 57 F.3d at 1165 n. 2 (noting that, although there may be reasons to reconsider the circuit's view that the two statutes are coextensive, the Air Force did not seek reconsideration); cf. also Chrysler Corp. v. Brown, 441 U.S. 281, 292-93, 99 S.Ct. 1705, 60 L.Ed.2d 208 (1979) (explaining that the “congressional concern” for the privacy interests of government contractors and other private entities "was with the agency’s need for confidentiality,” and that "FOIA by itself protects the submitters' interest in confidentiality only to the extent that this interest is endorsed by the agency”).