Washington Post v. United States Department of Defense

766 F. Supp. 1 (1991)

THE WASHINGTON POST, Plaintiff,
v.
UNITED STATES DEPARTMENT OF DEFENSE, Defendant.

Civ. A. No. 84-3400-LFO.

United States District Court, District of Columbia.

May 30, 1991. As Amended June 26, 1991.

Peter E. Scheer, Onek, Klein & Farr, Washington, D.C., for plaintiff-intervenor Scott Armstrong.

Vincent M. Garvey, M. Susan Carlson, Attys., Dept. of Justice, Washington, D.C., for defendant.

MEMORANDUM

OBERDORFER, District Judge.

On December 21, 1989, the Special Master submitted his report. The parties filed initial responses, and at a status conference on June 21, 1990, they were permitted to brief the issues identified by the Special Master more fully. Those briefs have now been submitted and argument upon them was heard on May 8, 1991. Accordingly, the issues identified by the Special Master are now ripe for consideration.

I.

This matter concerns a Freedom of Information Act (FOIA) request filed by the Washington Post in 1981 on behalf of plaintiff-intervenor R. Scott Armstrong, then a reporter for the Post. Specifically, the Post requested that the defendant Department of Defense (DOD) release information concerning the failed attempt in April, 1980, to rescue American hostages held in the United States embassy in Teheran. In the first phase of this litigation, the Court considered the Post's request for four documents: the three after-action reports reviewing the mission and the "Holloway Report" assessing the lessons learned about current counterterrorist capabilities and future needs in that area. After the DOD reviewed and re-reviewed these four documents, an Order of June 10, 1988 granted the DOD summary judgment in this phase of the litigation.

In contrast to the four documents in the first phase, the phase of the litigation currently at issue involves some two thousand documents totalling more than 14,000 pages and constituting the "actual working files for the planning and review of the rescue mission on 24 April 1980, as well as the following intelligence efforts to determine the location of the hostages and monitor the political efforts to obtain their release." Rice Declaration ¶ 4. Not surprisingly, the DOD has withheld or partially withheld numerous documents in the working files, in large part due to concern for the national security. On October 30, 1985, in support of its decision to withhold this information, the DOD submitted a motion for summary judgment, accompanying declarations, and twelve volumes of supporting materials. Faced with the expense of opposing that motion, the Post declined to pursue its request further. Armstrong, however, was permitted to intervene and carry on. See Order of April 15, 1986.

In light of the large number of documents involved, earlier experience with administering a FOIA request for a similar quantity of sensitive documents (see Nishnic v. U.S. Dep't of Justice, 671 F. Supp. 771 (D.D.C.), aff'd, 828 F.2d 844 (D.C.Cir. 1987); Nishnic v. U.S. Dep't of Justice, 671 F. Supp. 776 (D.D.C.1987)), anticipated difficulties with sampling either by the defendant or through random selection, an Order of January 14, 1988 appointed Kenneth C. Bass, III, Special Master. See In re: Dep't of Defense, 848 F.2d 232, 239 (D.C.Cir. 1988). The Order charged Bass with responsibility for "(1) examining in camera, and selecting for the Court, a representative sample of the 2000 or more documents and portions of documents that are the subject of the defendant's second motion for summary judgment and (2) summarizing for the Court the arguments that each party has made, or could make, about the claimed exemptions with respect to these 2000 documents...." Order of January 14, 1988, at 2.

In fulfillment of these responsibilities, the Special Master reviewed all of the documents in the working files, identified in consultation with the parties the primary legal issues in dispute, selected documents representative of those issues, and on December 21, 1989, submitted a report summarizing both the contents of those documents and the arguments that the parties either did or could raise about them. See Report of the Special Master at 5-11. That report is divided into three components. First, there is the Report of the Special Master itself which "explains the process that resulted in the selection of the documents." Id. at 1. Second, the Report contains descriptions of the 28 documents selected by the Special Master as representative of the legal issues in dispute. Finally, the Special Master submitted a nineteen-page summary of the arguments that "the parties have made or could make regarding release of the sample documents." Special Master's Summary of Arguments at 1.

Of most immediate importance here are the legal issues identified by the Special Master. Limiting himself to the material withheld under the FOIA's national security exemption, the exemption claimed for the vast majority of information withheld, the Special Master identified five categories of documents based upon the "legal positions of the parties as set forth in the pleadings, the Court's rulings in this case, the reported decisions of courts in other FOIA cases and the Special Master's own experience in both government service and private practice." Id. at 17. Those categories are:

1. "Re-classified documents"—documents that apparently were not classified at the time of their generation but were later classified, perhaps after receipt of the FOIA request, and are being withheld on the basis of that subsequent classification.

2. "Non-official" releases—documents containing classified information that has been publicly disclosed by "non-official" sources such as information contained in Col. Beckwith's book Delta Force.

3. "Changed Circumstances"—documents containing classified information which intervenor contends can no longer be classified because the subject matter has subsequently been released in another forum such as appropriations hearings, budgetary submissions or DOD publications.

4. "Open secrets"—information which intervenor contends cannot properly be classified because "everyone knows" the secret as a result of officially released information and common sense, e.g., Foreign Broadcast Information Service (FBIS) origination of translations of foreign broadcasts or the "fact of" involvement of certain intelligence community entities or capabilities in the rescue mission.

5. "Inadvertent release"—the unintentional release by the government of classified material, as DOD apparently contends occurred with regard to document A-50, a transcript of a press conference given by General Pustay on a "background" basis to representatives of numerous news organizations.

Id. at 17. In constructing these categories, the Special Master contemplated that the Court's rulings on these documents would "be relatively easily applied to the balance of the documents" and would guide the DOD's review of the rest of the working files. Id. at 20.^[1]

In response to the Special Master's Report, Armstrong submitted four volumes of materials in the public domain that he expected the working files to contain and a memorandum identifying and describing the information within those materials. Then, as mentioned above, the parties filed responses to the Special Master's Report, several status conferences were held, and after a hearing on June 21, 1990, the parties were permitted to brief more thoroughly the issues identified in the Special Master's Report. See Order of July 2, 1990. At the June 21, 1990 hearing, the DOD also asked that it be allowed—in light of the Special Master's Report and Armstrong's responses to that report—to reexamine the exemptions it claimed in the documents selected by the Special Master, to release the documents or portions of them that it no longer felt were exempt from disclosure, and to receive rulings upon the exemptions it continued to claim. Such rulings, the DOD hoped, would provide concrete guidance for its review of the exemption claims in the remaining documents. In the process of this reexamination,^[2] the DOD decided to release document A-50, the one document in the "inadvertent release" category, thereby mooting the fifth category identified by the Special Master. The other four categories are, however, now ripe for consideration.

At the hearing on May 8, 1991, the Special Master suggested that to facilitate the goal of providing the DOD with concrete guidance on how to review the remaining documents in the working files the Court should not only resolve the broad legal issues raised in his report, but also rule on specific exemptions in four documents. It has not, however, been possible to do so in this memorandum because, as developed below, many of the DOD's justifications are now insufficient, but could with further explanation become sufficient. See supra 9-15. Moreover, because those documents contain information that could be properly classified, the discussion of them has been necessarily vague and possibly cryptic, perhaps thereby defeating the goal of providing concrete guidance to the DOD. It is therefore necessary to conduct further proceedings with regard to the four documents selected by the Special Master at the last hearing. Accordingly, the accompanying order will require the DOD to submit appropriate public, and if necessary in camera, declarations justifying the continued withholding of certain information in those documents, and a hearing will be set at which the Government should have appropriate personnel available to testify in camera concerning the continued withholding of the information in the documents identified in the accompanying order.

II.

As noted above, the Special Master's Report focused upon the information withheld by the DOD under Exemption 1 of the FOIA, which allows agencies to withhold information "in the interest of national defense or foreign policy." 5 U.S.C. § 552(b)(1).^[3] To be withheld under that exemption, "a document must be classified in accordance with the procedural criteria of the governing Executive order as well as its substantive terms." Lesar v. United States Dep't of Justice, 636 F.2d 472, 483 (D.C.Cir.1980) (footnote omitted).

The governing order here is Executive Order 12356, and it has two primary substantive criteria. First, the information must fall within one of categories enumerated by the order. Those categories include:

(1) military plans, weapons, or operations;

(2) the vulnerabilities or capabilities of systems, installations, projects, or plans relating to the national security;

(3) foreign government information;

(4) intelligence activities (including special activities), or intelligence sources or methods; [and]

(5) foreign relations or foreign activities of the United States....

Executive Order 12356 § 1.3(a), April 2, 1982, 47 Fed.Reg. 14874, 15557, reprinted in 50 U.S.C.A. § 401 note (1991) [hereinafter, "E.O. 12356"]. Second, information that concerns one or more of these categories may be classified only if "its unauthorized disclosure, either by itself or in the context of other information, reasonably could be expected to cause damage to the national security." Id. § 1.3(b).

E.O. 12356 also requires that classification follow certain procedures. First, information may only be classified by an individual with "original classification authority." Id. § 1.3(b). Furthermore, if the information is classified or reclassified after a FOIA request has been received by an agency, it must be classified by "the agency head, the deputy agency head, the senior agency official designated under Section 5.3(a), or an official with original Top Secret classification authority." Id. § 1.6(d). Properly classified documents must also have certain markings which identify the classified information and show its classification level, the identity of the original classification authority, the agency and date of origin, and the date or event for declassification. See id. § 1.5.

Substantive review of classification decisions is quite deferential. Because the major substantive question under Exemption 1 involves an evaluation of the national security, Congress instructed courts to "accord substantial weight to an agency's affidavit concerning the details of the classified status of [a] disputed record." S.Conf. Rep. No. 93-1200, 93d Cong., 2d Sess. 12 (1974), U.S.Code Cong. & Admin.News 1974, 6285. Acknowledging this congressional command and recognizing that judges "lack the expertise necessary to second-guess ... agency opinions in the typical national security FOIA case," our Court of Appeals has, in the absence of any contradictory evidence, required little more than a showing that the agency's rationale is logical. Halperin v. CIA, 629 F.2d 144, 148 (D.C.Cir.1980); accord Military Audit Project v. Casey, 656 F.2d 724, 738 (D.C. Cir.1981); Hayden v. National Security Agency/Central Security Service, 608 F.2d 1381, 1386-87 (D.C.Cir.1979), cert. denied, 446 U.S. 937, 100 S. Ct. 2156, 64 L. Ed. 2d 790 (1980).

In this case, the DOD has for the most part satisfied this lenient standard. The threat to the national security from disclosure of information contained in the working files of the hostage rescue attempt mission is evident even to a judicial eye. Those files contain intelligence planning documents for a difficult and important counterterrorist mission which will inevitably serve as a model of "do's and don't's" for future missions with similar objectives and obstacles. According to the DOD, the documents therefore provide a veritable blueprint of United States counterterrorist operations, including strategies, assessments and capabilities of military systems, and intelligence activities and capabilities. See Rice Declaration ¶¶ 13-15. The release of such information would likely "permit hostile nations or groups to avoid or neutralize American counterterrorist efforts," resulting in "a geometric increase in the risk to American citizens and interests world-wide." Id. ¶ 10. The working files also contain information about agreements with foreign governments in connection with the mission. These agreements were reached under an express or implied promise of confidentiality, and the DOD contends that disclosure of such information "would violate a U.S. agreement with the foreign government and inhibit future exchanges of information." Id. ¶ 16. Similarly, the working files also contain information provided by and evidence of cooperation with foreign governments, disclosure of which the agency contends would "violate expectations of confidentiality in diplomatic exchanges and inhibit future exchanges with officials of those and other foreign governments." Id. ¶ 17. These rationales easily meet the deferential standard for reviewing agency assessments of the threat to the national security. Thus, for the most part, DOD has satisfied its burden of showing that the information withheld from the working files has been properly classified according to the substantive criteria of E.O. 12356.

Review of the agencies' action under the FOIA does not, however, end here. In the first place, in addition to showing that the agency's classification decisions meet the substantive criteria of E.O. 12356, the DOD must also demonstrate that its determinations have satisfied the order's procedural criteria. Furthermore, because the FOIA requires the agency to disclose "any reasonably segregable portion" of a document containing exempt material, see 5 U.S.C. § 552(b), the DOD must justify its decision to withhold each segment of information by "describ[ing] the withheld information and the justification for withholding with reasonable specificity." Salisbury v. United States, 690 F.2d 966, 970 (D.C.Cir. 1982). Finally, because "the burden is on the agency to sustain its decision," 5 U.S.C. § 552(a)(4), the DOD must also show that the information withheld is "not controverted by either contrary evidence in the record nor evidence of agency bad faith." Military Audit Project, 656 F.2d at 738 (footnote omitted). These are the concerns upon which the five categories identified by the Special Master focus.

III.

The Special Master noted that it was not clear from the DOD's declarations that it had fully complied with the procedural requirements of E.O. 12356. See Special Master's Summary of Arguments at 2-7. First, many of the documents withheld in whole or in part under Exemption 1 were not marked with the identity of the original classification authority and the agency and office of origin as required by § 1.5(a) of the order. See id. at 6-7. Second, it was not clear to the Special Master from the agency's declarations whether the decisions to classify or reclassify material after the instant FOIA request were made "personally and on a document-by-document basis by ... an official with original Top Secret classification authority" as required by § 1.6(d) of that order. See id. at 3-5. DOD concedes that many documents were not properly marked and has undertaken to correct the markings on all 2,000 documents in the working files. See Hop-good Declaration ¶¶ 7-8. It maintains, however, that it did comply with § 1.6(d)'s requirement of personal review by an appropriately senior official when the working files were first reviewed for release in 1985.

The Special Master deduced from the markings on many of the documents that they were classified no earlier than 1982, the date on which E.O. 12356 was promulgated and well after the Post's request was filed. See Special Master's Summary of the Arguments at 3-5. He then noted that it was not clear from the DOD's declarations that these new classification and reclassification decisions were made "personally and on a document-by-document basis" by an original Top Secret classification authority. See id. at 4-5. Specifically, Major General W.H. Rice, the Top Secret Classification authority supervising the review of the working files, declared that

I have specifically instructed my action officers to consider whether changed circumstances, the lapse of time, official disclosure, or other events have altered the proper classification status of any of the information in these documents. This process resulted in information being declassified and released. I now certify that the information still withheld is currently, properly classified pursuant to one or more specific provision of Executive Order 12356.

Also in the course of the review portions of some of the documents which were not appropriately marked as classified at the time of origination have been upgraded to Confidential or Secret. Many of the documents which required proper markings in conformity with Executive Order 12356 were the actual working papers, i.e. handwritten notes or messages, of the group planning the rescue mission. I certify that this material should have been, and now is, properly classified.

Rice Declaration ¶ 6 (emphasis in original). Notably absent from this declaration is a statement that the newly classified or reclassified documents were personally reviewed by Major General Rice on a document-by-document basis. Accordingly, the Special Master determined that "the record presents an issue of compliance with the Executive Order." Special Master's Summary of Arguments at 5.

In response, the DOD filed declarations from Major General Rice and his chief action officer, Colonel James N. Roberts. According to these declarations, throughout the review process Rice and the action officers assigned to help him process the working files were cognizant of § 1.6(d)'s requirement that Rice, as the Top Secret classification authority, personally review all reclassification and new classification decisions. The issue was specifically discussed when the review was organized in April of 1985. See Roberts Declaration ¶ 7. Indeed, the need for personal review by Rice proved to be nearly ubiquitous: "[A]ction was taken in 1985 to downgrade, declassify, or upgrade portions of virtually every document in the sample." Hopgood Declaration ¶ 7. Accordingly, Rice was forced to make "all determinations concerning the documents or portions of the documents reviewed in connection with this action through 18 October 1985, including classification, declassification, downgrading and upgrading," based upon not only the recommendation of his action officers, but also based upon his "personal examination of the documents on a document-by-document basis." Second Rice Declaration ¶ 3.

This process satisfies the requirements of E.O. 12356. Rice, a Top Secret classification and declassification authority, reviewed each document or portion of document classified and personally decided whether the information was properly classified. Section 1.6(d) requires no more. Armstrong does not argue that § 1.6(d) always requires a Top Secret classification authority like Rice to act alone. Instead, he contends that in this case Rice delegated too much authority to his action officers and did not exercise enough independent judgment. However, the record does not support this contention. According to the Roberts Declaration, the working files were originally reviewed by action officers whose work was in turn reviewed by a quality control team. See Roberts Declaration ¶ 12. Each officer was, however, "instructed to annotate any recommendation for Major General Rice's decision on changing the level of classification by marking the recommended new level in the margin." See id. ¶ 11. Before the bi-weekly releases required by the Order of June 18, 1985, Rice reviewed the documents at issue, questioned his officers' recommendations, and often reversed them. See id. ¶¶ 13-14. Thus, far from indicating that Rice merely rubberstamped the recommendations of his action officers, Roberts' declaration fully corroborates Rice's statement that he personally reviewed and reassessed every classification decision made after this FOIA request was filed

Accordingly, the DOD must be deemed to have satisfied the procedural requirements of § 1.6(d) of Executive Order 12356.

IV.

The three remaining categories of documents identified by the Special Master deal with the question of whether information already in the public domain may be classified and withheld on the ground that its subsequent disclosure by the government would threaten the national security. As the Special Master implies, such information might constitute the sort of "contrary evidence" controverting an agency's determination that according to our Court of Appeals prompts a more searching review of the agency's justification under Exemption 1. See, e.g., Salisbury v. United States, 690 F.2d at 970; Military Audit Project, 656 F.2d at 738. The DOD responds that the presence of information in the public domain is not relevant to the evaluation of its Exemption 1 claims unless such information is identical to the information being considered and has been officially disclosed by the government. This legal conclusion, the Department of Defense continues, has been established for purposes of this case by the rulings on the after action and Holloway reports. The DOD's characterization of this Court's rulings is, however, inaccurate, and its reading of the applicable precedent sweeps too broadly.

It is a matter of common sense that the presence of information in the public domain makes the disclosure of that information less likely to "cause damage to the national security." E.O. 12356 § 1.3(b). The Executive Order itself recognizes this fact: Under § 1.6(c), previously declassified and disclosed information may be reclassified only "if it is determined in writing that ... the information may reasonably be recovered." In other words, if the information has already been disclosed and is so widely disseminated that it cannot be made secret again, its subsequent disclosure will cause no further damage to the national security. As a consequence, according to our Court of Appeals, suppression of "already well publicized" information would normally "frustrate the pressing policies of the Act without even arguably advancing countervailing considerations." Founding Church of Scientology of Washington, D.C., Inc. v. National Security Agency, 610 F.2d 824, 831-32 (D.C. Cir.1979) (footnotes omitted); accord Afshar v. Dep't of State, 702 F.2d 1125, 1130 (D.C.Cir.1983); Lamont v. Dep't of Justice, 475 F. Supp. 761, 772 (S.D.N.Y.1979); see also Phillippi v. Central Intelligence Agency, 655 F.2d 1325, 1332 (D.C.Cir.1981) (Mikva, J., concurring).

It does not, however, follow that agencies must always disclose information already in the public domain. The FOIA contains no such requirement, nor does Executive Order 12356. Moreover, our Court of Appeals has treated agency justifications for withholding information already in the public domain with the same deference it accords to all agency justifications under Exemption 1. Thus, information in the public domain may be withheld if the agency asserts in its declarations that the information being withheld is not exactly the same as the information being withheld and that release of the more detailed information in the document poses a threat to the national security. See Salisbury v. United States, 690 F.2d at 971-72. Classification and withholding of information already in the public domain may also be justified on the ground that although the information withheld is exactly the same as information in the public realm, revealing the context in which the information is discussed would itself reveal additional information, release of which is harmful to the national security. See, e.g., Afshar, 702 F.2d at 1130. Similarly, if the information in the public domain was not officially disclosed, official confirmation or acknowledgement of that information may be harmful to national security. For example, with unofficial sources of information, "unresolved doubt may still remain in the minds of the United States' potential and actual adversaries...." Military Audit Project, 656 F.2d at 744; accord Abbotts v. Nuclear Regulatory Comm'n, 766 F.2d 604, 608 (D.C.Cir.1985). The fact of an official confirmation of information already known may also be significant: While a country may be able to countenance unacknowledged covert activities within its territories, once those activities are officially confirmed it may be forced by domestic political pressure to retaliate diplomatically or otherwise. See Afshar, 702 F.2d at 1131; Phillippi, 655 F.2d at 1332-33. Finally, where confidential sources are involved, there is some value in keeping an agency's promise not to reveal the identity of a source even when the source's identity is already known. See Military Audit Project, 656 F.2d at 740. In sum, there are numerous reasons why an agency might determine that disclosure of information already in the public realm "reasonably could be expected to cause damage to national security."

It does not, as the DOD suggests, follow from the availability of such justifications for withholding that information in the public domain need only be considered if it has been officially disclosed and it is identical to the information being withheld. That is certainly not the law of the case. In rulings on phase one of this litigation, this Court did hold that "otherwise exempt information" does not lose that status merely because such information has been previously disclosed in unofficial sources. See Memorandum of September 22, 1986 at 3; see also Memorandum of June 10, 1988 at 8 (holding that information released by the Iranians does not constitute an official disclosure). However, this Court never held that the presence of information in the public domain could be ignored by an agency when considering whether an official release of such information would harm the national security. To the contrary, where the harm from official release of information was not apparent, the Court required the DOD to explain its justification for classifying the material. See Sealed Transcript, May 9, 1988, at 6, 14. Thus, it is not the "law of the case" that unofficial information in the public domain need not be considered by an agency in determining whether to withhold information under Exemption 1.

Nor is it the law of the circuit. Our Court of Appeals has never held that unofficial disclosures of information or official disclosures of similar yet not identical information may be ignored by an agency in making its classification decisions. To the contrary, although it has never explicitly required such an explanation, our Court of Appeals has only allowed the withholding of information already in the public domain based upon a specific explanation for continued withholding of that information, supported by appropriate agency declarations, of why formal release of information already in the public domain threatens the national security. See Abbotts, 766 F.2d at 608 n. 4; Afshar, 702 F.2d at 1131 n. 7, 1131-32; Salisbury, 690 F.2d at 971-72; Military Audit Project, 656 F.2d at 741; Phillippi, 655 F.2d at 1332 & n. 22; Hayden, 608 F.2d at 1388.

The decisions of this Court cited by the DOD do not require a different result. The DOD points out that in one case Judge Flannery found "dispositive ... the fact that the `disclosures' plaintiff points to ... are not official disclosures and therefore have should have no effect on the court's disposition of the Exemption 1 issue." Schlesinger v. CIA, 591 F. Supp. 60, 68 (D.D.C.1984) (emphasis in original). However, throughout his opinion, Judge Flannery emphasized that the CIA's rationale for withholding information already in the public domain was that although some facts about the CIA's involvement in the Guatemalan revolution had been revealed, the extent of its involvement had not been and that further official disclosure would prompt diplomatic retaliation detrimental to the national security. See id. at 63, 66, 68-69. Thus, Schlesinger merely follow our Court of Appeals' determination that official confirmation of publicly known facts may be withheld to avoid the threat of diplomatic retaliation. See, e.g., Afshar, 702 F.2d at 1131; Phillippi, 655 F.2d at 1332-33.

Nor does Pfeiffer v. CIA, 721 F. Supp. 337 (D.D.C.1989), provide the DOD any support. In that case, Judge Pratt found the presence of information withheld in the public domain to be "of no legal significance." Pfeiffer, 721 F.Supp. at 342. Pfeiffer, however, did not involve a decision to withhold information under Exemption 1. In that case, the CIA withheld information under Exemption 3, which allows the withholding of information specifically exempted from disclosure by statute. See 5 U.S.C. § 552(b)(3). More importantly, the statute invoked by the CIA makes the Director of the Central Intelligence Agency responsible "for protecting intelligence sources and methods from unauthorized disclosure." See 50 U.S.C. § 403(d)(3). Nothing in this statute requires the Director to determine that disclosure of those sources or methods would likely harm the national security. Accordingly, unlike a decision to withhold information under Exemption 1, under the statute at issue in Pfeiffer there was no requirement that release of such information reasonably be expected to cause harm to the national security. See Baker v. CIA, 580 F.2d 664, 667-69 (D.C.Cir.1978); see generally CIA v. Sims, 471 U.S. 159, 105 S. Ct. 1881, 85 L. Ed. 2d 173 (1985). As a consequence, Pfeiffer provides little support for the DOD's contention that under E.O. 12356 the DOD is not required to justify its determination that release of information already in the public domain could be expected to harm the national security.

Armstrong contends that in assessing the threat to the national security under Exemption 1 the DOD should start with two presumptions. First, it should presume that the information contained in books published by former high government officials is credible. In particular, Armstrong contends that accounts of individuals like Colonel Beckwith, who led the ground assault forces, and President Carter, who as commander-in-chief was responsible for the entire operation, are especially credible. According to Armstrong, each author occupied an important position with respect to the operation, their accounts corroborate each other, and their books are written with a reassuring air of objectivity. While these argument are compelling, they do not provide a basis for creating a special presumption of reliability for such books. Congress' instruction to "accord substantial weight" to agency determinations that there is a threat to the national security is clear and admits no exceptions. See, e.g., S.Conf.Rep. No. 93-1200, 93d Cong., 2d Sess. 12, U.S.Code Cong. & Admin.News 1974, 6285. As a consequence, the agency's only duty is to explain its determination in a form that responds to Armstrong's contentions and facilitates testing of the underlying rationale and its factual foundation.

Second, Armstrong argues that there should be a presumption of reliability for information in books reviewed before publication by the DOD. According to Armstrong, prepublication review of a book represents a determination that there is no information in that book whose disclosure might damage the national security. The DOD disputes this characterization. According to it, when the government preclears a book, it does not certify that every fact in that book is accurate. Indeed, it is possible that the Government might purposefully, with or without the knowledge of the author, allow misinformation to be published. See, e.g., Military Audit Project, 656 F.2d at 744 (speculating that information revealed in a book by former CIA director William Colby represented a "fallback story" meant to cover up the true purposes of a CIA operation). More generally, books precleared by the government "are received as the private product of their authors, like any other memoirs, and are accorded such respect as their content seems to deserve." Afshar, 702 F.2d at 1134. As a consequence, there is no reason to presume that books subject to prepublication review are reliable.

In summary, neither Exemption 1, Executive Order 12356, nor the case law announces a special rule for dealing with information that is already in the public domain. The presence of such information is, however, relevant to the determination, required by E.O. 12356, that disclosure of the information in question "reasonably could be expected to cause damage to the national security." E.O. 12356 § 1.3(b). By providing evidence that the information being withheld is already within the public domain, a FOIA plaintiff brings into question the government's determination that release of such information might reasonably be expected to damage the national security. Such contrary evidence, in turn, requires the Court to investigate the agency's declarations more closely and determine whether the agency has answered the questions raised by the plaintiff's evidence. Any such response must, of course, be afforded the "substantial weight" always accorded executive department determinations on such issues. However, in the absence of any response, an agency cannot meet its burden of proof under the FOIA.

With these general principles in mind, it is now time to turn to the specific categories of information in the public realm identified by the Special Master.

V.

The three categories of public disclosures identified by the Special Master—"changed circumstances," "non-official release," and "open secrets"—each present an analytically different facet of the problem of information in the public domain. Each category is treated separately below.

A.

As noted above, "changed circumstances" refers to information that has been officially released. With regard to such information, the DOD advances two broad arguments. First, it contends that Armstrong has not met his burden of demonstrating that the information in particular documents is identical to information that has been previously and officially disclosed. See, e.g., Hopgood Declaration, Enclosure 3, at 1 (discussing Document A-2). Second, the DOD argues that some information cannot be disclosed without threatening the national security because "release of such information in the context in which it is discussed would damage national security." Defendant's Response to Plaintiff-Intervenor's Second Statement of Issues at 21. Both arguments sweep too broadly.

The DOD's first argument misconstrues the proper distribution of the burdens of proof in FOIA actions. It is true that the government is not required to search the public domain for information within the scope of a given FOIA request. Such a requirement would make the government's task "virtually limitless." Lamont v. Dep't of Justice, 475 F.Supp. at 772 & n. 43. Accordingly, "a plaintiff asserting a claim of prior disclosure must bear the initial burden of pointing to specific information in the public domain that appears to duplicate that being withheld." Afshar, 702 F.2d at 1130 (citations omitted). Armstrong has, however, met that burden. He has submitted four volumes of public source materials and, in an accompanying memorandum dated January 23, 1990, he has identified the information within those sources that, he asserts, duplicates information in the working files. Armstrong does not, as the DOD suggests, have the additional burden of proving that the information submitted by him is identical to the information contained in specific passages in specific documents in the working files. Under the FOIA, the agency bears the ultimate burden of proof. See 5 U.S.C. § 552(a)(4). Nothing in the statute suggests that this ultimate burden of proof shifts to the plaintiff merely because information in the public realm is involved. Indeed, such a shift in the burden of proof would undermine the goals of the FOIA. It is well-settled that the FOIA "sets forth a policy of broad disclosure of Government documents in order to ensure an informed citizenry vital to the functioning of a democratic society." FBI v. Abramson, 456 U.S. 615, 621, 102 S. Ct. 2054, 2059, 72 L. Ed. 2d 376 (1982) (quotation and quotation marks omitted). If, however, plaintiffs had to prove the presence of specific information in a specific document, they would almost always lose no matter how meritorious their claim, because only the agency withholding the documents knows with precision the information within those documents and alone has the means of making a comparison between that information and the public information proffered by the requester. Thus, the DOD's suggested allocation of the burden of proof not only has no foundation in the statute; it would also frustrate Congress' well-settled intentions in passing the statute. The DOD must instead bear the burden of comparing the proffered information with the information being withheld, determining whether the information is identical, and, if it is not, determining whether release of the perhaps only slightly different information being withheld would harm the national security.

The DOD's second contention also sweeps too broadly. As mentioned before, information already in the public domain may be withheld because disclosure of that information in a particular context would reveal additional information detrimental to the national security. See Afshar, 702 F.2d at 1130. This justification cannot, however, be applied indiscriminately or in a conclusory fashion. See, e.g., Hayden, 608 F.2d at 1387 (noting that an agency's declarations "will not suffice if the agency's claims are conclusory, merely reciting statutory standards, or if they are too vague or sweeping") (footnote omitted). Each justification for withholding must explain the logical connection between the rationale for withholding and the information withheld. See, e.g., Salisbury, 690 F.2d at 970.

This failure to provide a sufficiently detailed justification infects two of the documents identified by the Special Master. The Special Master points out that on page one of Document A-359, the names of several agencies whose involvement in the hostage rescue attempt has been officially acknowledged are withheld. See Special Master's Description of Sample Documents at 15-16. The rationale in the Hopgood Declaration for withholding this information is that "disclosure of this information would reveal intelligence sources and methods used for planning and execution of US counter-terrorism missions. This would assist potentially hostile forces in countering future efforts or monitoring intelligence activities for warning of US counterterrorism operations." Hopgood Declaration, Exhibit B, Document A-359. The Rice Declaration and the classified declaration filed in 1985 provide similarly terse explanations. In light of the official disclosure of the involvement of some of those agencies in U.S. counterterrorism operations, these explanations are not sufficient. Their concern over the damage that disclosure of a particular agency's involvement might well be misplaced if such damage has already been done. A more particularized explanation, taking into account the presence of such information in the public realm, is necessary in order to justify continued withholding of that information. Similarly, in document B-49, evidence presented by Armstrong brings into question at least one of the rationales for withholding in the classified affidavit. See Special Master's Description of the Sample Documents at 17. Here again, in order to meet its burden of proof, the DOD must explain what justifies continued withholding of this information now that the cat is out of the bag—even if it determines that the explanation must be made in camera.

B.

The second category of public domain information identified by the Special Master, "non-official releases," concerns information that has been publicly disclosed, but in unofficial sources. The problem here, as suggested above, is that several former government officials involved in the operation have written books about the hostage rescue attempt.^[4] Here again, the DOD has failed to explain why, in light of the disclosure of such information in such an apparently credible form, further disclosure would harm the national security.

For example, in Document B-269, several code names are withheld without any apparent consideration of the fact that those code names have been revealed in books by high government officials with personal knowledge of the hostage rescue attempt. Because those books would appear to have a good deal of reliability, there would seem to be little reason to doubt their accuracy. There remains to be explained how official disclosure of these names at this time would damage the national security. Accordingly, the accompanying order will require the DOD to explain its decision to withhold these code names.

The Special Master noted that in Document I-439 the DOD failed to explain why disclosure of information disclosed in unofficial sources would damage the national security. See Special Master's Description of the Sample Documents at 12. The Hopgood Declaration justifies withholding in its entirety an intelligence report concerning information from a specific source on the ground that "disclosure of this information would reveal a particular intelligence source and method used in preparation for the hostage rescue attempt, and could preclude future use of similar activities." Hopgood Declaration, Exhibit B, Document I-439. However, the classified declaration submitted in 1985 neither described the method at issue nor explained why revealing that method would harm the national security. Furthermore, the source can be discerned from the date of the report and information in a book published by what appears to be a reliable author. See Special Master's Description of the Sample Documents at 12. Most importantly, there is no discussion of why this information cannot be segregated from the rest of the apparently factual information in the report. The classified affidavit provides another, sufficient justification for withholding the information at the end of the document, but it, too, fails to explain why that information cannot be segregated from the other, largely factual, information in the document. Accordingly, the accompanying order will require the DOD to explain its decision to withhold the information in Document I-439 and to classify one of its justifications for doing so.

C.

The third category and final category identified by the Special Master is "open secrets," information which is easily discernible through officially released information and common sense. For example, Document B-269, in addition to listing several agencies involved in the hostage rescue attempt whose participation has been officially revealed, also lists several agencies whose participation has not been officially revealed. The cooperation of the agencies in question in operations similar to the hostage rescue attempt has, however, been officially acknowledged, albeit in another context. Armed with this information, an expert, or indeed a knowledgeable amateur, analyst could deduce the agencies listed in Document B-269. Accordingly, the accompanying order will require the DOD to explain why it feels that release of this information would damage the national security.

VI.

When the DOD completes the review of the working files, it may wish once again to move for summary judgment. If it chooses to do so, it will have to provide significantly more detailed declarations in order to satisfy its burden of proof under the FOIA. See generally Vaughn v. Rosen, 484 F.2d 820 (D.C.Cir.1973), cert. denied, 415 U.S. 977, 94 S. Ct. 1564, 39 L. Ed. 2d 873 (1974).

Although the FOIA requires courts to review de novo agency decisions to withhold information, it is not necessary, nor indeed practicable, for courts to review every document withheld and every bit of information deleted. Indeed, as our Court of Appeals recognized in Vaughn v. Rosen, to do so would not only place the burden of justifying an agency's claims of exemption upon the courts; it would also practically remove plaintiffs from participating in the evaluation of their claims. See id. at 825-26, 828. Accordingly, in Vaughn v. Rosen, Judge Wilkey devised an indexing system to allow courts and plaintiffs alike to review an agency's decision to withhold information without having to actually inspect each bit of information withheld. See id. at 826-28. As later elaborated, a "Vaughn index" must "adequately describe each withheld record or deletion" by identifying the document by type and location as well as by describing the portions redacted with "as much information as possible without thwarting the exemption's purpose." King v. United States Dep't of Justice, 830 F.2d 210, 224 (D.C.Cir.1987). Such an index must also contain "a relatively detailed justification, specifically identifying the reasons why a particular exemption is relevant and correlating those claims with the particular part of a withheld document to which they apply." Mead Data Central, Inc. v. United States Dep't of Air Force, 566 F.2d 242, 251 (1977) (citations omitted); but see Keys v. United States Dep't of Justice, 830 F.2d 337, 349 (D.C.Cir.1987) (noting that an adequate Vaughn index may use general categories and generic terms—as long as those categories and terms are clearly responsive to the information to which they are applied).

The declarations provided by the DOD fail to satisfy these requirements. In the first place, the DOD's declarations do not describe the particular information deleted. For example, one of the lengthiest descriptions in the Rice Declaration is of Document B-269. That description is over 7 pages long. B-269 is, however, over two hundred and fifty pages long, and the declaration makes little attempt to describe each deletion. In describing four pages of prefatory material, the Rice Declaration merely states that:

The portions withheld on these pages and labeled with an A describe specific aspects of planning that form the basis for current counterterrorist planning. Those portions labeled with a B reveal general statements concerning U.S. intelligence capabilities and limitations that affected the intelligence effort. Portions labeled with a C reflect intelligence matters, activities, and methods.

Rice Declaration, Exhibit C, volume 1 (Document B-269). Thus, rather than attempting to describe the information contained in each redaction, the Rice Declaration summarizes the information contained in many redactions. The description in the Hopgood Declaration is similarly deficient, and it goes without saying that, because the information deleted is not clearly described, the DOD's declarations do not "specifically identify[] the reasons why a particular exemption is relevant." Mead Data Central, 566 F.2d at 251 (emphasis added).

It is clear that the type of Vaughn index necessary to satisfy the DOD's burden of proof is much more detailed and, therefore much longer, than the present declarations. To attempt, as the DOD has here, to provide a Vaughn index for each document in the working files would be an enormous task. It is not, however, necessary for the DOD to provide a Vaughn index for each document. As our Court of Appeals has recently reaffirmed, "[r]epresentative sampling is an appropriate procedure to test an agency's FOIA exemption claims when a large number of documents are involved." Bonner v. United States Dep't of State, 928 F.2d 1148, 1151 (D.C. Cir.1991) (citations omitted). Thus, if after the DOD's review of the working files and release of all documents or portions of documents that it determines to be nonexempt Armstrong still contests the exemptions claimed by the DOD, the agency should then produce a detailed Vaughn index for a sample of the working files and move for summary judgment on the basis of that index. The method for selecting that sample will be determined by the parties or by the Court when, and if, necessary.

The DOD has represented that it will be able to finish its review of the working files within a year's time. It has, however, also asked that it not be required to release documents on a biweekly basis as it was required to do in 1985. See Stipulation and Order, May 29, 1985, ¶ 5. That requirement, the DOD explains, placed considerable pressure upon it to release documents at the beginning of the review process, even though many classification decisions are best and most easily made after all related documents have been reviewed. Accordingly, it would seem wise to give the DOD six months in which to process the working files before its first release is due and then require the rest of the documents to be released in quarterly installments. Because this period may be shortened if the parties agree to pare down the number of documents to be reviewed, the final schedule for review and release of the working files will be set at the next hearing.

VII.

Armstrong urges that either the Court or the Special Master should closely supervise that review by requiring the DOD to prepare and submit to the Court a classification guide for that review. Armstrong contends that the shortcomings in the DOD's Vaughn indices and the apparent misclassification of information originally withheld but later released after the review of the Special Master's sample supervised by Major General Hopgood demonstrate that the agency has processed his request in bad faith. It is, however, clear that, while the DOD's previous releases had numerous shortcomings, for the most part its decision to classify information in the working files was justified. See supra 5-6. Moreover, the agency has agreed to conduct another review of the approximately 2,000 documents in the working files. This record hardly supports Armstrong's allegations of bad faith. Accordingly, there is no need to supervise the DOD's review of the working files.

VIII.

One final issue remains. There are approximately fifteen documents in the working files that the DOD contends are not agency records and therefore not subject to disclosure under the FOIA. The Special Master therefore included in the sample documents Document A-2, a two-page memorandum with a seventy-page transcript of a congressional hearing attached to it, as an example of this category of documents. See Special Master's Description of the Sample Documents at 23.

Under the FOIA, federal courts have "jurisdiction to enjoin the agency from withholding agency records." 5 U.S.C. § 552(a)(4)(B) (1988). Because Congress is not an agency for purposes of the FOIA, see 5 U.S.C. § 551(1)(A), our Court of Appeals has held that congressional records are not subject to the FOIA. See Goland v. CIA, 607 F.2d 339, 345 & n. 26 (D.C.Cir.1978), vacated in part on other grounds, 607 F.2d 367 (D.C.Cir.1979), cert. denied, 445 U.S. 927, 100 S. Ct. 1312, 63 L. Ed. 2d 759 (1980).

The problem here is that the records in question, although created by Congress, are no longer in Congress' possession, but rather in the possession of the DOD. Congress can, by transferring control of documents created by it to executive agencies, render such documents agency records for the purposes of the FOIA. See Paisley v. CIA, 712 F.2d 686, 694-95 (D.C.Cir.1983); Holy Spirit Ass'n for Unification of World Christianity v. CIA, 636 F.2d 838, 841 (D.C.Cir.1980), other portions vacated and remanded as moot, 455 U.S. 997, 102 S. Ct. 1626, 71 L. Ed. 2d 858 (1982); Goland, 607 F.2d at 346-47. However, the mere fact that an agency has physical possession of a document does not transform it into a congressional record. Congress can, for example, transfer documents to executive agencies merely for safekeeping. See Goland, 607 F.2d at 345-46. Consequently, "[w]hether a congressionally generated document has become an agency record ... depends on whether under all the facts of the case the document has passed from the control of Congress and become property subject to the free disposition of the agency with which the document resides." Id. at 346.

As stated above, attached to Document A-2 is a transcript from a congressional hearing attached to it. That hearing was conducted before the Senate Armed Services Committee while it was in executive session, which under the committee's rules means that no "report of the proceedings of such hearing shall ... be made public in whole or in part or by way of summary unless authorized by a majority vote of the committee or subcommittee." Senate Armed Services Committee Rules of Procedure, Rule 10(e), reprinted in 131 Cong.Rec. S2853 (daily ed. March 7, 1985). According to the Joint Chief of Staff's congressional affairs specialist, unedited versions of transcripts of such hearings are routinely released to witnesses testifying at those hearings. See Jones Declaration ¶ 3. The witnesses are asked to review the relevant transcript, make minor editing and grammatical changes, and, if the witness was "unable to provide information requested," to provide an insert for the record containing that information. Id. If any part of a witness' testimony is sensitive, that witnesses is also asked to indicate what information he or she considers to be classified. See id. ¶ 3. The memorandum in Document A-2 indicates that the attached transcript was released for this limited purpose of correction and emendation: The memorandum describes the transcript as "a working copy of the official transcript" and requests the witnesses to review the transcript

a. To insure accuracy of information provided by the hearing witnesses.

b. To provide information for the record which the witnesses were not in the position to do so during the hearing.

c. To identify classified information in the text of the testimony, if any.

Document A-2 ¶ 2.

Congress cannot be said to have abandoned control over a transcript released to witnesses from an agency for the limited purpose of correcting and emending. Indeed, this case is indistinguishable from our Court of Appeals' seminal decision in Goland. That case also involved a transcript from closed committee hearing which was transferred to an agency for limited purposes. See Goland, 607 F.2d at 347. In Goland, the transcript was transferred to the CIA "for internal reference purposes, to be used in conjunction with legislation concerning the Agency and its operations." Id. (citations omitted). Here, the documents were transferred for an even more limited purpose: correction and completion of each witness' testimony. As a consequence, Goland compels the conclusion that the transcript attached to Document A-2 is a congressional record.

Armstrong, however, contends that there must be some clear indication of Congress' intent to retain control. He bases this contention upon Judge Skelly Wright's statement in Paisley v. CIA that nothing in either the circumstances of the creation of the documents at issue in that case or the conditions attending their transfer "provides the requisite express indication of a congressional intent to maintain exclusive control over these particular records". Paisley, 712 F.2d at 695. In reviewing the case law, however, Paisley did not suggest that an "express indication of congressional intent to maintain exclusive control" is a prerequisite for a congressional record. Quite to the contrary, Paisley described the test as one that focused on two factors: "(1) the circumstances attending the document's creation, and (2) the conditions under which it was transferred to the agency." Id. at 692 (citations omitted). Nor did Paisley purport to reject our Court of Appeals' previous refusal to "direct Congress to act in a particular way in order to preserve its FOIA exemption for transferred documents." Holy Spirit Ass'n, 636 F.2d at 842. Most fundamentally, Paisley did not purport to overrule Goland. In Goland, there was no express indication from Congress of its intent to retain control over the transcripts at issue there. The Goland Court inferred Congress' intent from an affidavit by the CIA's Legislative Counsel George L. Cary swearing that the transcripts were transferred to the CIA for "internal reference purposes only." See Goland, 607 F.2d at 347 n. 43; See also id. at 345 n. 30 (describing the Cary Affidavit). The Jones declaration provided here by the DOD provides exactly the same sort of evidence of Congress' intentions.

Accordingly, DOD has shown that the transcripts in Document A-2 remain congressional records.

IX.

Like many FOIA actions, this case has raised numerous issues. It is fortunate for the Court that the Special Master identified most of those issues and focused both the parties' and the Court's attention upon them. Some of those issues have been resolved in this memorandum. It has, for example, been determined that Document A-2 is not an agency record subject to FOIA and that General Rice's use of action officers did not offend the procedural requirements of Executive Order 12356. Other issues like the question of how to deal with the presence of information in the public domain have been clarified, but a final ruling on the agency's exemption claims must await further explication of the agency's reasoning. It is, however, hoped that after the next hearing, the DOD will have ample guidance in how to review the working files, and this decade-old FOIA request will be resolved.

ORDER

For the reasons stated in the accompanying memorandum, it is this 30th day of May, 1991, hereby

ORDERED: that the Defendant's Second Motion for Summary Judgment should be, and is hereby, denied without prejudice; and it is further

ORDERED: that the Plaintiff-Intervenor Scott Armstrong's Cross-Motion for Partial Summary Judgment should be, and is hereby, denied without prejudice; and it is further

NOTICED: that Document A-2 is a congressional document and is therefore not subject to disclosure under the Freedom of Information Act; and it is further

ORDERED: that the defendant shall on or before June 13, 1991 submit public and, if necessary, in camera declarations concerning the passages in Documents A-359, B-49, B-269, and I-439 discussed in the accompanying memorandum; and it is further

ORDERED: that the plaintiff-intervenor may file a response on or before June 20, 1991; and it is further

ORDERED: that counsel shall attend a hearing on June 27, 1991, at 4:30 P.M.

NOTES

[1] The Special Master warned, however, that his sample was "designed to fairly represent the legal issues." Id. at 19 (emphasis in original). As a consequence, that sample is not proportionately representative of the 2000 documents in the working files, and it would be inappropriate to extrapolate from the percentage of mistaken exemption claims, if any, found in the sample to the percentage of mistaken claims likely to be found in the remaining documents in the working files.

[2] This process was understandably interrupted while the DOD was otherwise occupied with events in Kuwait and Iraq.

[3] Exemption 1 provides that agencies may withhold any materials that are

(A) specifically authorized under criteria established by an Executive order to be kept secret in the interest of national defense or foreign policy and (B) are in fact properly classified pursuant to such Executive order.

5 U.S.C. § 552(b)(1).

[4] See, e.g., C. Beckwith, Delta Force (1983); Z. Brzezinski, Power and Principle (1983); J. Carter, Keeping Faith (1982); H. Jordan, Crisis (1982); P. Ryan, The Iranian Rescue Mission: Why It Failed (1985); G. Sick, All Fall Down (1985); S. Turner, Secrecy and Diplomacy (1985); C. Vance, Hard Choices (1983). In addition to these materials, many documents concerning the hostage rescue attempt purportedly recovered at Desert One were published by the Iranians.