This appeal involves the application of subsection (b)(5) of the Freedom of Information Act, 5 U.S.C. Sec. 552 (1988), known as the "deliberative process" exemption, to certain cost estimates prepared by Navy officials in the course of the Navy's selecting homeports for ships in a new battleship group. We affirm the district court's finding that subsection (b)(5) protects the estimates from mandatory disclosure.

* * *

In the mid-1980s the Navy initiated a search for "homeports" for the 130 ships that it proposed to add to the U.S. fleet. One "battleship group," consisting of a refurbished USS Wisconsin and five ancillary vessels designed to carry guided missiles (a cruiser, a destroyer and three frigates), was slated for the Gulf of Mexico. The Navy initially planned to deploy the Gulf Coast group from a single U.S. harbor, and convened a special study team, composed of representatives of various command and technical units, to evaluate seven finalist sites on operational, logistic, environmental and other criteria. The team's final report, "Gulf Coast Battleship Surface Action Group: Preferred Alternative Home Port Evaluation," also included the material disputed here--cost estimates for each site, including the costs of land, ship berthing, dredging, buildings and facilities, and utilities. After completion of the study, the Secretary of Defense decided (for reasons unknown to us) not to designate a single Gulf Coast homeport for the battleship group, but rather to make the allocation as part of a broader plan, assigning 29 vessels among nine different harbors. Joint Appendix ("J.A.") 30, 269.

Appellant Randy Quarles is a reporter for Newhouse News Service, writing primarily for newpapers in Mobile and Huntsville, Alabama. Before the Pentagon made its final decision in the matter, he submitted a FOIA request to the Navy for materials relating to the search process. In addition to 7,600 pages of other responsive materials, J.A. 312, the Navy released what it characterized as "the truly factual information in the [study team's] report--such as the composition of the battleship task force, physical dimensions of the ships, personnel compliment [sic] for each ship, photographs of the proposed ports, etc." J.A. 32. The Navy excised the rest--analysis, conclusions and cost estimates. Quarles sought judicial review, but failed to persuade the district court to force disclosure of the portions withheld. See Memorandum Opinion, No. 85-3395 (D.D.C. July 29, 1988). He filed a timely appeal to this court, limited to the Navy's deletion of the cost estimates.

* * *

Exemption 5 excludes from FOIA's broad disclosure requirements "inter-agency or intra-agency memorandums or letters which would not be available by law to a party other than an agency in litigation with the agency." 5 U.S.C. Sec. 552(b)(5) (1982). It incorporates the rule of discovery, antedating the Act, that " 'confidential intra-agency advisory opinions ... are privileged from inspection.' " EPA v. Mink, 410 U.S. 73, 86, 93 S. Ct. 827, 835, 35 L. Ed. 2d 119 (1973) (quoting Kaiser Aluminum & Chemical Corp. v. United States, 157 F. Supp. 939, 946, 141 Ct. Cl. 38 (1958)). The exemption is meant "to encourage the frank discussion of legal and policy issues" in government deliberation. Wolfe v. HHS, 839 F.2d 768, 773 (D.C.Cir.1988) (en banc) (quoting S. REP. 813, 89th Cong., 1st Sess. 9 (1965)).

To receive the protection of Exemption 5, a document must first be pre-decisional. See NLRB v. Sears, Roebuck & Co., 421 U.S. 132, 151-52, 95 S. Ct. 1504, 1516-17, 44 L. Ed. 2d 29 (1975). Material that follows the adoption of a policy is thought to represent interpretation of it, and the courts have ordered disclosure of post-decisional documents "to prevent bodies of 'secret law' from being built up and applied by government agencies." Schwartz v. IRS, 511 F.2d 1303, 1305 (D.C.Cir.1975). Although it may sometimes be difficult to peg a document on the decisional continuum, see Sears, Roebuck, 421 U.S. at 152 n. 19, 95 S. Ct. at 1517 n. 19, the cost estimates sought by Quarles plainly pass this first test. The team report was "prepared in order to assist an agency decisionmaker in arriving at his decision," see Renegotiation Board v. Grumman Aircraft, 421 U.S. 168, 184, 95 S. Ct. 1491, 1500, 44 L. Ed. 2d 57 (1975); it was completed in May 1985, while the basing decision was not made until July of that year. See J.A. 40, 269.

What is disputed is the requirement that the document be deliberative in character. See Wolfe, 839 F.2d at 774. Courts have typically required disclosure of "purely factual material," see, e.g., Mink, 410 U.S. at 88, 93 S.Ct. at 836, presumably because the prospect of disclosure is less likely to make an advisor omit or fudge raw facts, while it is quite likely to have just such an effect on "materials reflecting deliberative or policy-making processes," id. at 89, 93 S.Ct. at 837. See also Dudman Communications Corp. v. Dep't of the Air Force, 815 F.2d 1565, 1569 (D.C.Cir.1987); Playboy Enterprises, Inc. v. Dep't of Justice, 677 F.2d 931, 935 (D.C.Cir.1982); Mead Data Central, Inc. v. Dep't of the Air Force, 566 F.2d 242, 256 (D.C.Cir.1977) (Mead Data I ).

Even when requested material is found to be factual, the courts have held it exempt where they were convinced that disclosure "would expose an agency's decisionmaking process in such a way as to discourage candid discussion within the agency and thereby undermine the agency's ability to perform its functions." Dudman, 815 F.2d at 1568. With this second step, courts have sometimes allowed the withholding of factual summaries prepared by decisionmakers, see Lead Industries Ass'n v. OSHA, 610 F.2d 70, 85 (2d Cir.1979); Washington Research Project v. HEW, 504 F.2d 238, 250-51 (D.C.Cir.1974); Montrose Chemical Corp. v. Train, 491 F.2d 63 (D.C.Cir.1974), as of other factual materials. See Wolfe, 839 F.2d at 774 (protecting log showing routing path and dates of FDA recommendations, i.e., facts about the inner workings of the deliberative process itself); Dudman (protecting factual material in early drafts of an official history on grounds that disclosure would unduly reveal editorial judgments). But this additional inquiry has been used only to expand Exemption 5's protections. Otherwise, "the fact/opinion distinction 'offers a quick, clear, and predictable rule of decision,' for most cases." Wolfe, 839 F.2d at 774 (quoting Mead Data I, 566 F.2d at 256). This is one such case.

Quarles would have us characterize the cost estimates as fact. Numbers have a surface precision that may lead the unsophisticated to think of them as fixed, and of course some are--Waterloo was fought in 1815. But cost estimates such as these are far from fixed, as anyone knows who has had two contractors bid on a home improvement or has compared budget estimates with final costs of a government project. They derive from a complex set of judg ments--projecting needs, studying prior endeavors and assessing possible suppliers. They partake of just that elasticity that has persuaded courts to provide shelter for opinions generally.

Quarles argues nonetheless that the Navy has not shown that harm to the decisionmaking process would result from disclosure of the estimates. Appellant's Reply Brief 2. This claim must fail. The district court expressly found that disclosure of the information sought "could chill discussion at a time when agency opinions are fluid and tentative." Mem.Op. at 6 (quoting Wolfe, 839 F.2d at 776). The court thus credited the Navy's representation that "[t]o disclose ... details of the Homeporting Study Team's analyses and estimates would likely make future researchers highly reluctant to express their candid opinions, particularly where such opinions may later prove controversial." J.A. 32-33 (affidavit of Lt. Henry C. Shelley, Jr.). If such findings are necessary to support nondisclosure under Exemption 5, see Bureau of National Affairs v. United States Dep't of Justice, 742 F.2d 1484, 1496-98 (D.C.Cir.1984); Arthur Andersen & Co. v. IRS, 679 F.2d 254 (D.C.Cir.1982); Brinton v. Department of State, 636 F.2d 600 (D.C.Cir.1980) (all upholding Exemption 5 withholding of predecisional advisory materials without inquiry into possible harms of disclosure), these are clearly sufficient. Mead Data Central, Inc. v. Dep't of the Air Force, 575 F.2d 932, 936 (D.C.Cir.1978) (Mead Data II ) (finding cost analysis of contract proposal covered by Exemption 5).

We need not stretch our imaginations to vindicate the district court's conclusion. If a study team found significant potential savings in deploying the battleship group at a finalist port otherwise deemed satisfactory, disclosure of the finding surely would tend to undermine the acceptability--among taxpayers--of forsaking that option. Yet the Navy might have reasons for rejecting the seemingly cheaper option, perhaps strategic, see J.A. 269 (Defense Department press release noting importance of dispersing naval forces), perhaps political, see J.A. 174 (Quarles affidavit recounting charges that final decision was "purely political"). Anticipating this scenario, high officials might be inclined either not to call for cost estimates, or to call only for fuzzy ones expressed as wide ranges. Cf. Mead Data II, 575 F.2d at 936. The increased tendency to fudge would increase the sloppiness (or decrease the precision) of government decisionmaking. In an ideal world not only officials but also the public would receive such analyses. But the deliberative process exemption rests on the premise that if both (nominally) can get them, neither will: too many analyses will be stillborn or wishy-washy.

Disclosure of the estimates before decision would pose additional risks. See Coastal States Gas Corp. v. Dep't of Energy, 617 F.2d 854, 866 (D.C.Cir.1980) (noting one purpose of Exemption 5 "to protect against premature disclosure of proposed policies before they have been finally formulated or adopted"). Such a release might unleash forces that would effectively preclude any alternative other than the one supported by the cost projections, just as deliberate leaks often skew the decisionmaking process. Again, the prospect of such disclosure would increase the incentives to lower officials' fudging such estimates, or to higher-ups' not even calling for them.

Quarles relies on two cases in which courts ordered the disclosure of property valuations made by government appraisers, GSA v. Benson, 415 F.2d 878 (9th Cir.1969); Tennessean Newspapers, Inc. v. FHA, 464 F.2d 657 (6th Cir.1972), but both are readily distinguishable. First, although we do not rule on Exemption 5's application to appraisals of existing property, we note that they seem to involve fewer judgment calls than estimates of what construction will cost. Second, there were special features at work in each of the cases. In Benson the court relied heavily on an agency regulation requiring disclosure in the absence of "compelling reason," 415 F.2d at 880, and in Tennessean Newspapers the agency had already disclosed the appraisal, withholding only the name of its author. See Hoover v. United States Dep't of Interior, 611 F.2d 1132, 1140 n. 10 (5th Cir.1980) (distinguishing Benson and Tennessean ). In any event, we note that where the government has persuasively asserted a resulting harm, courts have found appraisals covered by Exemption 5. See Hoover, 611 F.2d at 1135; Martin Marietta Aluminum, Inc. v. GSA, 444 F. Supp. 945, 949-50 (C.D.Cal.1977).

* * *

Finding the homeport cost estimates covered by Exemption 5, we affirm the decision of the district court.