No CrossRef data available.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Published online by Cambridge University Press: 24 January 2025
Enactment of the United States Freedom of Information Act (“FOI Act”) in 1966 was a landmark event in the history of American administrative law. Among the most important of American administrative law reforms, the FOI Act has engendered a veritable cottage industry devoted to securing access to government-held information. In the sixteen-plus years since the Act went into effect Congress has twice amended the Act and is currently considering yet another round of amendments. Each occasion has engendered a new round of studies and commentaries on the Act, its purposes and effects.
The attention given to the subject by Congress pales in comparison to the attention given it by the courts. A September 1981 list shows more than 1300 decisions construing the FOI Act, and companion laws: the Privacy Act, Sunshine Act and Federal Advisory Committee Act.
1 5 USC § 552.
2 On the private FOi Act “industry” see “Lifting the Curtain From Government Secrets”, US News & World Report, 5 February 1973, 50 (“more than 15,000 lawyers and other representatives of 1,600 business and professional associations are employed at digging out hard-to-get information”); “Government Business and the People's Right to Know” (1978) 3 Media Law Reporter 20-21 (discussing the growth of FOi Act service bureaus). On the public sector counterpart to this private sector “'industry” see, eg, “'Bureaucracy's Great Paper Chase”, Time (19 December 1977). 23-24 (the FBI employs 379 full time FOi Act staffers, the Department of Defence (“DOD”), 90, the CIA, 65).
3 The Act was passed into law in July 1966 but implementation was delayed for a year in order to allow agencies time to promulgate regulations.
4 Amendments were made in 1974 and in 1976. See James T O'Reilly, Federal Information Disclosure § 3.08 (1977, updated through 1982). (“O'Reilly Treatise”)
5 See Hearings on Freedom of Information Act before The Subcommittee on the Constitution of the Senate Judiciary Committee, 97th Congress 1st Session, vols 1 and 2 ( 1981)(“1981 FOi Act Hearings”). While the 97th Congress adjourned without passing legislation, a bill (§ 774), similar to the leading bill in that Congress (§ 1730), has been introduced in the current Congress.
6 See US Department of Justice, Freedom of Information Case List (1981) 1-89. The Case List, published annually, cites all reported and unreported cases and indexes them by general topics.
7 I do not deal with these laws in this paper, though they are all part of the same family of “open government” legislation.
8 5 USC§ 552a. Essentially the Act mandates record keeping requirements for personal information of a private character, including notice to the individual about whom records are maintained, a right of the individual to obtain such records — subject to certain exemptions — and a right to challenge the contents of such records and have inaccuracies corrected. The Act also limits disclosure of personal information without the consent of the individual except where disclosure is required by the FOi Act or certain other conditions are met.
9 5 USC §662b. The Act requires that meetings of multi-membered agencies, at which agency deliberations determine agency action, be open to the public, except to the extent the meeting deals with matters exempt from disclosure under the FOi Act. (The same legislation also amended the FOi Act, as will be noted; and the APA provisions governing adjudicatory hearings for-bidding ex parte contacts in such hearings.)
10 5 USC — Appendix. The Act requires that meetings of officially established “federal advisory committees” be conducted in public except to the extent the committee deals with matters exempt from disclosure under the FOi Act.
11 See Freedom of Information Case List, supra n 6, 107-127.
12 See Swallow, “Has the Freedom of Information Act Worked — Or Has It Worked Too Well”, National Journal (15 August 1981) 1470.
13 There is a large number of reports of the burdens created for particular agencies. Much of this information consists of “horrible case” anecdotes. See, eg, 1981 FOi Act Hearings, supra n 5, Vol 1 at 106 (report of a single request to DOD requiring a search of 24 million pages to locate the requested documents, requiring some 350,000 man hours); ibid at 984 (report of single request to FBI, resulting in a court order to produce 40,000 documents a month). However, some agencies have reported specific manpower and cost data. See, eg, ibid at 105, 627 (estimated $6.8 million in annual costs for DOD; $11 million FBI). Cost to all agencies has been estimated at around $57 million for 1980, a figure that does not include judicial enforcement costs. See, ibid, vol 2 at 3.
14 One of the key features of the FOi Act was its elimination of the former limitation on disclosure to persons “properly and directly concerned” and its exception for documents that were deemed to be confidential “for good cause”. Eliminating these limitations removed any legal basis on which agencies could ask requesters to identify for whom and for what purposes the information is sought Of course, since the agencies cannot refuse to disclose non-exempt information regardless of how or by whom it will be used, requesters need not conceal their purposes. Thus, in many cases it is possible to ascertain from the request itself who is seeking information and for what use. However, it is not possible to ascertain the real parties interested or the ultimate uses of the information with any statistical precision. Estimates on these matters are necessarily rough, and may not be fully reliable.
15 Discussion of the Australian FOI Act is based on the Act as passed in 1982. A number of proposed amendments were being considered at the time this paper was completed. I have noted a few of the more important proposed changes when pertinent. No attempt is made to analyse the Australian FOI Act in full. I have merely cited major features of the Act that, more or less, parallel those of the US Act.
16 See generally R Berger. Executive Privilege: A Constitutional Myth (1974). Berger's treatment is primarily concerned with legislative access to executive secrets. hence it focusses heavily on the constitutional discussions of the privilege. The common law bases of the privilege are. however. briefly discussed in chapter 7 of his book. In general terms the executive privilege encompassed several more or less distinctive privileges: (1) A privilege for military secrets and secrets of state. one generally regarded as absolute. See United States v Reynolds (1953) 345 US 1. (2) An “informer's privilege”. a qualified privilege not to disclose the identity of informers except upon a showing that the public interest requires disclosure. See Roviaro v United States (1957) 353 US 53. (3) Assorted statutory privileges either forbidding or restricting disclosure of information given to the agency in confidence or protecting certain kinds of information gathered by the agency. such as trade secrets. financial information and the like. See. eg Federal Trade Commission Act. § 6(f). 15 USCA § 46(f); 18 USCA § 1905. (4) A qualified privilege for so-called “internal management” matters. See. eg. Appeal of the United States Securities and Exchange Commission (1955) 226 F 2d 501; cf United States v Nixon (1974) 418 US 683.
17 On old s 3 of the APA see Note “Comments on Proposed Amendments to Section 3 of the Administrative Procedure Act: The Freedom of Information Bill” (1965) 40 Notre Dame Lawyer 417.
18 See generally Johnson. The Government Secrecy Controversy (1967) 120-121.
19 O'Reilly Treatise. supra n 4. § 2.03.
20 H Cross. The People's Right to Know (1954).
21 The District Court and Court of Appeals of the District of Columbia Circuit have accounted for a majority of decisions. In part this fact reflects venue convenience for suits against Washington-based federal agencies. In part it probably also reflects a familiarity on the part of the administrative law segment of the bar with those courts. Finally, the general philosophy of those courts in matters such as this has undoubtedly been influential. Except for reverse-FOI Act cases the initiative in FOI Act review cases is. of course, determined by persons who want to overturn agency refusal to disclose records. It is a common perception of administrative lawyers that the DC Circuit has, in recent years at least, been more aggressive than most circuits in over-seeing federal agency decisions. Thus. litigants expect a more sympathetic treatment in the DC Circuit. Whether that perception is accurate is perhaps debatable. but it is the perception, not the reality, that dictates venue choices.
22 In Kissinger v Reporters Committee For Freedom of the Press (1980) 445 US 136, the Court distinguished between the “Executive Office of the President” and the “Office of the President”, the latter being limited to the President. his immediate personal staff and other executive Office staff whose sole function is to advise the President. Compare the exemptions for Cabinet and Executive Council documents in the Australian FOI Act ss 34. 35. There are some notable differences in coverage between the US and Australian Acts. The Australian Act contains a series of general exemptions for certain agencies, such as Australian intelligence agencies and various public corporations: see s 7. The US Act. in marked contrast, covers virtually all government agencies. corporations. and institutions other than Congress and the Courts. See 5 USC § 551(i) and § 552(e).
23 5 USC § 551(1)(A)&(B). The Australian FOI Act, s 5, allows access to court documents of an administrative nature.
24 5 USC § 552(a)(1). The publication provisions of the Australian FOI Acts 8. appear to be generally similar.
25 5 USC § 552(a)(2). Comparable provisions in the Australian FOI Act appear in s 9.
26 5 USC § 552(a)(3) and (b). The parallel provisions of Part III of the Australian FOi Act are rather more detailed, but broadly similar to those of the US FOI Act. One notable general difference is that the Australian disclosure provisions are applicable only to documents created or acquired by the agency after the FOI Act enactment date. except where an individual seeks access to documents concerning his personal affairs in which case access extends to records created or acquired as much as five years before the enactment date or where the requested document is necessary to understand a current document properly obtained by the requester. No such time limitation applies to the US disclosure scheme. Proposals are pending to amend the Australian FOI Act to extend the application of the disclosure provisions to all documents created or acquired after 1978, with no limit for requests for personal documents.
27 5 USC § 552(a)(4)(A). The Australian FOi Act ss 29. 30 similarly provide for waivable fees.
28 5 USC § 552(a)(6). The Australian FOI Acts 19 similarly prescribes deadlines buts 21, authorising deferral for public interest reasons, finds no counterpart in the US Act. In Federal Open Market Comm v Merrill (1979) 443 US 340 the Court found authorisation for deferral in exemption five of the US FOi Act. See infra n 48. But this was an exceptional, limited authorisation which is unlikely to be applied beyond the very special circumstances of that case.
29 5 USC § 552(a)(4)(F). No comparable provision appears in the Australian FOI Act. I am unaware of any case imposing disciplinary sanctions under the US FOI Act.
30 5 USC § 552(a)(4)(B-G). Part VI of the Australian FOI Act contains a somewhat more complex review scheme than is found in the US FOI Act. The additional complexity in the Australian FOI Act largely concerns review of exemptions for documents pertinent to national security/defence. for Cabinet and Executive Council documents and for internal working documents when a ministerial certificate that disclosure is contrary to the public interest is conclusive, and may not be overridden. See ss 33(2): 34(2): 35(2): 36(3). Although the government's determination cannot be overturned by the reviewing body. the Administrative Appeals Tribunal, it can be referred to a special Documents Review Tribunal which is authorised to review the government's claim and to render an opinion on whether the claim is reasonable. For this purpose the ORT is authorised to inspect the exempt documents in camera — a power denied the AAT in these cases. However. the ORT opinion is purely precatory: as with the AAT it may not overturn the government's claim of exemption in these cases. The current government proposals to amend the Act would abolish the ORT and allow the AAT to assume its functions — without, however, changing the ministerial prerogatives in case of these exemptions.
31 5 USC § 552(a)(4)(B). This authorisation to compel production of documents for in camera inspection extends to all documents. The Australian FOI Act, s 64, gives the AAT similar power in all cases other than those where such documents are claimed to fall within the exemptions for documents pertinent to national security/defence, Cabinet and Executive Council documents. and for internal working documents. However, the special ORT is authorised to conduct in camera inspection of such documents and if the Act is amended as now proposed this function will be assumed by the AAT. Both Acts by their terms only authorise in camera review. not compel it. See National Labor Relations Board v Robbins Tire and Rubber Co (1978) 437 214, 224.
32 5 USC § 552(b). Compare Australian FOI Act Part IV, ss 32-52. The US FOI Act has been construed to permit, but not compel withholding of exempt documents. See Chrysler Corporation v Brown (1979) 441 US 281. The Australian FOI Acts 14 explicitly grants the agency discretion to disclose “where they can properly do so or are required by law to do so”. This latter qualification appears to parallel the Chrvsler decision that, while the FOI Act does not forbid disclosure, other laws may do so. In all cases the agency's exercise of discretion is reviewable by the courts.
33 For a detailed discussion see Note, The Status of Law Enforcement Manuals Under the Freedom of Information Act (1980), 75 Nw UL Rev 734.
34 “The cases generally fall into two categories: (1) cases holding that law enforcement manuals are implicitly exempted by (a)(2)(C) (covering access/copying of staff manuals, etc) to the extent disclosure would risk circumvention of the law or impede law enforcement, see, eg, Cox v United States Department of Justice (1978) 576 F 2d 1302; (2) cases holding that such material is within (a)(2)(C) but is exempted by exemption two, applicable to internal personnel files, eg, Hardy v Bureau of Alcohol, Tobacco & Firearms (1980) 631 F 2d 653. In addition some decisions have emphasised that exemption seven, (protecting investigatory records, disclosure of which would interfere with enforcement proceedings) is at least corroborative of an exemption for such material even though it would not literally apply to manuals since they are not “investigatory records”. See ibid at 656.
35 SeeJordan v United States Department of Justice (1978) 591 F 2d 753.
36 Crooker v Bureau of Alcohol. Tobacco & Firearms (1981) 670 F 2d 1051. The court exempted two protected portions of a manual on investigative techniques. The court professed not to reject its earlier Jordan decision, but only its sweeping rationale insofar as it suggested that all law enforcement manuals must be disclosed. In fact. however. Crooker repudiated the most important, and controversial. part of Jordan.
37 See. eg, O'Reilly Treatise. supra n 4.
38 See Executive Order No 12356. 47 Fed Reg 14874 (1982). The current Executive Order, issued by President Reagan in April 1982. tightens somewhat the classification criteria and processes of previous orders. See infra nn 99 and 100 and accompanying text. However. the general criteria and procedures are not changed from those prescribed by the Carter Administration in Executive Order 12065. 43 Fed Reg 28949 (1978).
The classification scheme prescribes three levels of classification:
(1) “Top Secret” shall be applied to information. the unauthorised disclosure of which reasonably could be expected to cause exceptionally grave damage to the national security.
“Secret” shall be applied to information, the unauthorised disclosure of which reasonably could be expected to cause serious damage to the national security.
“Confidential” shall be applied to information. the unauthorised disclosure of which reasonably could be expected to cause damage to the national security.
(The only change from prior classification criteria is the elimination of the word “identifiable” qualifying “damage” in the confidential classification scheme.) Information shall be classified if it concerns:
(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;
(5) foreign relations or foreign activities of the United States;
(6) scientific, technological, or economic matters relating to the national security;
(7) United States Government programmes for safeguarding nuclear materials or facilities;
(8) cryptology;
(9) a confidential source; or
(10) other categories of information that are related to the national security and that require protection against unauthorised disclosure as determined by the President or by agency heads or other officials who have been delegated original classification authority by the President.
(The principal changes here are the addition of items 2. 8 and 9.)
Classifications remain in effect for as long as required by security considerations. (The principal change here was elimination of the process for automatic declassification after six years. However. automatic declassification determinations under prior order remain in effect unless classification is extended.)
39 S 33 of the Australian FOI Act appears generally parallel to the US provision except that a government certificate that disclosure would be contrary to the public interest is conclusive and may not be overridden by the reviewing court. By contrast there is full judicial review under the US Act. As originally drafted, exemption one of the US FOI Act was interpreted to preclude judicial review of the reasonableness of the classification. See Environmental Protection Agency v Mink (1973) 410 US 73. The 1974 amendments overruled Mink and provided for de novo review of classifications and in camera scrutiny of documents themselves where necessary to determine the reasonableness of the classifications. See Attorney General's Memorandum on the 1974 Amendments to The Freedom of Information Act (1975) 1-4 (“Attorney General's Memorandum”). However. legislative history indicates that “de novo review” in this context requires courts to give “substantial weight to an agency's affidavit concerning the details of the classified status of the disputed record”. See Attorney General's Memorandum 3. See also Military Audit Project v Casey (1981) 656 F 2d 724. 738.
40 See Ray v Turner (1978) 587 F 2d 1187.
41 The Court's decision in Environmental Protection Agency v Mink (1973) 410 US 73, that courts could not review the reasonableness of executive classifications, was explicitly a matter of statutory interpretation. The Court acknowledged that Congress had the power to overturn its ruling, but noted that such power was subject to an undefined constitutional executive privilege, at 83.
42 We know from United States v Reynolds (1951) 345 US 1, that it includes matters of national security: as to such sensitive matters the privilege is absolute. Beyond this core area, however. both the scope of the privilege and the level of protection (whether absolute or conditional) are nebulous. In United States v Nixon (1974) 418 US 683, the Court took as established that the privilege included confidential communications beyond those implicating national security interests, but ruled that the President's “generalised interest in confidentiality” had to give way to the due process rights of a criminal defendant who demanded the information in order to prepare a defence. See generally Symposium: United States v Nixon (1974) 22 UCL.AL Rev I: P Freund, “Forward: On Presidential Privilege” (1974) 88 Harvard L. Rev 13.
43 The scope of the nonconstitutional. common law, executive privilege outside the area of national security interests is also vague. See, NLRB v Sears Roebuck & Co (1975) 421 US 132: Federal Open Market Comm v Merrill (1979) 443 US 340. It is a reasonable inference from EPA v Mink, supra note 41, that the nonconstitutional privilege is broader than the constitutional privilege, but such is the nebulousness of the entire subject that we cannot do more than guess how much broader.
43 Compare the parallel exemption in s 40 of the Australian FOI Act.
44 See O'Reilly Treatise, supra n 4, at § 12.03.
45 Department of the Air Force v Rose (1976) 425 US 352, 369.
46 The 1976 amendment, enacted as a rider to the Sunshine Act, narrowed the scope of exemption three. The amendment was intended to overrule Federal Aviation Administration v Robertson (1975) 422 US 255, where the Supreme Court construed the exemption to cover statutes that gave broad discretion to agencies to withhold documents. The parallel exemption in the Australian FOI Act, s 38, appears to reflect a similar view of the required specificity for statutory exemptions.
47 See, eg, Baldrige v Shapiro (1982) 102 S Ct 1103 (Census Act within class of statutes covered by exemption 3; resolving conflicting circuit opinions). The process is unavoidably ad hoc. For a discussion of relevant criteria see O'Reilly Treatise, supra n 4, § 13.04.
48 See, Public Citizen Health Research Group v FDA (1983) 704 F 2d 1280 choosing a restrictive interpretation which limits protection to information about the “productive process” itself as opposed to more general matters of commercial confidentiality.
49 For lists of particular items that have been found to be “commercial” and “financial” under the US FOI Act see O'Reilly Treatise, supra n 4, § 14.07.
50 Confidential commercial information generated within the agency itself, as distinct from information submitted to it, is not covered by exemption four. See Federal Open Market Commission v Merrill (1979) 443 US 340, where the Court was forced to employ a rather creative interpretation of exemption five to protect the sensitive confidential strategies of the Fecjeral Reserve Board's open-market operations. The Australian FOi Act, s 44, handles this directly in its exemption for documents affecting the national economy.
51 See National Park & Conservation Association v Morton (1974) 498 F 2d 765. The Australian FOI Act, s 43, incorporates similar tests within its exemption for documents relating to business affairs.
52 For useful recent discussions see Note, “Protecting Confidential Business Information from Federal Agency Disclosure After Chrysler Corp v Brown” (1980) 80 Columbia L Rev 109; Note, “A Procedural Framework for the Disclosure of Business Records Under the Freedom of Information Act” (1980) 90 Yale L J 400.
53 Chrysler Corporation v Brown (1979) 441 US 281.
54 See 1981 FOi Act Hearings, supra n 5, vol 1 at 263 (testimony of senior counsel for Proctor and Gamble). Congress has recently amended some agency charters to forbid disclosure of material covered by exemption four, and to require advance notice to information submitters as well as opportunity to be heard on the confidential status of the information. See 15 USC § 57a(b)(2) (Federal Trade Commission); 15 USC § 2055 (Consumer Product Safety Commission). Compare s 27 of the Australian FOi Act providing for notice to submitters and an opportunity for them to argue in support of nondisclosure.
55 For example, the FDA. by regulation, refuses to notify submitters ofrequests for confidential information except where the agency is uncertain whether the information is confidential. See Pharmaceutical Manufacturers Association v Weinberger (1976) 411 F Supp 576 (affirming agency regulations against).
56 See National Labor Relations Board v Sears Roebuck & Co (1975) 421 US 132. The Australian FOI Act, s 36, incorporates a generally similar standard in its exemption for internal working documents. In one respect the Australian exemption is narrower than the US insofar as it specifically adds the condition that disclosure “would be contrary to the public interest”. Under US law this condition is assumed but not required to be established in a particular case. In another, more important, respect, however, the Australian exemption is much broader: under s 36(3), a government certificate that disclosure would be contrary to the public interest is conclusive and may not be overridden by the court.
57 Ibid. See also Australian FOi Acts 36(5).
58 Ibid. See also Australian FOi Acts 36(6). Any document incorporated by reference into an agency opinion cannot be exempt under the US FOI Act. See American Mail Line Ltd v Gulick (1969)411 F 2d 696.
59 Renegotiation Board v Grumman Aircraft Engineering Corp (1975) 421 US 168, 192. However, other provisions of the APA require opinions for formal adjudications, 5 USC § 557, and a statement of basis for “informal” rulemaking, 5 USC § 553. Even where the APA does not require a formal opinion or statement, some explanation of agency action is the minimum prerequisite of judicial review. See Citizens to Preserve Overton Park v Volpe (1971) 401 US 402.
60 See. eg, Department of Air Force v Rose (1976) 425 US 352. The parallel exemption in Australian FOI Act s 41 is not explicit on the balancing though this may be implicit in the “unreasonable” standard of s 41.
61 The Australian FOI Act, s 41, is not limited to particular files. The Supreme Court has recently diminished the significance of the limitation by giving an expansive interpretation to what qualifies as a “similar” file. United States Dept of State v Washington Post (1982) 102 S Ct l957.
62 See Attorney General's Memorandum, supra n 39, 4-13: National Labor Relations Board v Robbins Tire& Rubber Co (1978) 437 US 214. The Australian FOI Acts 37 is generally similar to the amended US FOI Act.
63 See National Labor Relations Board v Robbins Tire & Rubber Co, supra n 62.
64 See O'Reilly Treatise, supra n 4, at §§18.01-18.02. These exemptions would be largely embraced within the Australian FOI Acts 39, (documents affecting financial or property interests of Commonwealth) and s 44 (documents affecting national economy, including regulation of financial institutions).
65 See eg National Cable Television Association v FCC (1973) 479 F 2d 183.
66 See Attorney General's Memorandum, supra n 39, 22-23.
67 >Vaughn v Rosen (1973) 484 F 2d 820; cert denied (1974) 415 US 977.
68 Soucie v David (1971) 448 F 2d 1067. Compare Federal Open Market Comm v Merrill (1979) 443 US 340 where the Court construed exemption five to authorise a delay in disclosure where immediate disclosure would jeopardise government activities requiring confidentiality. On the one hand the Court seems to have incorporated a degree of equitable discretion into exemption five (in special cases); on the other hand, the very fact that it sought flexibility in the substance of a particular exemption implies that there is no general equitable discretion in enforcing the Act.
69 Environmental Protection Agency v Mink (1973) 410 US 73.
70 See supra n 39.
71 See Subcommission on Administrative Practice and Procedure of the Senate Commission on the Judiciary, Freedom of Information Act and Amendments of 1974 S Doc, No 93-82 93d Congress 2d Session 70 (1974).
72 Ibid 72 (citing data from 1972 hearings).
73 Ibid 15, 20-59.
74 Ibid 15, 60-66.
75 See supra nn 39, 46, 61 and accompanying text.
76 On these and other 1974 changes see Attorney General’s Memorandum, supra n 39. Other changes in 1974 were: modifications in exemptions 1 and 7, discussed previously; a requirement for disclosure of all material that is “reasonably segregable” from exempt material (again, incorporating existing judicial interpretation); a requirement that agencies publish periodic indexes of agency materials to which public access is mandated under 5 USC § 552(a)(2); a specific definition of “agency” (incorporating existing judicial interpretation); and a requirement that requests “reasonably describe” records requested (incorporating existing judicial interpretation of the identifiable records requirement).
77 See 5 USC § 552(c): “ … This section is not authority to withhold information from Congress”. Except where constitutional executive privilege is implicated, the right of Congress to obtain information from agencies is plenary.
78 See “Final FOIA Action Unlikely Despite Panel's Compromise”, Congressional Quarterly (May 29, 1982) 1267-68.
79 See 1981 FOI Act Hearings, supra n 5, for a detailed discussion of this and other proposed amendments. § 1730 appears in vol 1 at 30-52 and is explained in vol 2 at 1-52. In March, 1983 Senator Hatch introduced a Bill (S 774) generally similar to § 1730 but with some changes in detail. For my purposes a review of the earlier proposed legislation will adequately expose the points of important controversy.
80 See 1981 FOi Act Hearings, supra n 5, vol 2 at 3 (estimate by Constitution Subcommittee of $57 million for 1980). Some portion of the agency costs is recouped by FOI Act fees. However, charges can only be assessed for search and copying costs which the subcommittee estimated to be a trivial four per cent of the total, the rest being processing costs (reviewing documents, editing exempt material, etc).
81 lbid. Again, some of the judicial enforcement costs would be covered by court fees; however, this would be trivial compared to the unrecoverable costs of government attorneys' time involved in enforcement actions.
82 US budget outlays for fiscal year 1982 were slightly over $728 billion. Budget of the US Government Fiscal Year 1984. 98th Congress 1st session H Doc No 98-3 p 9-3.
83 Budget outlays for the USIA in fiscal year 1982 were $486 million, of which about $111 million was devoted to the VOA alone. Ibid at p 9-21 and Appendix p 1-V 141.
84 See, eg. 1981 FOI Act Hearings, supra 5, vol 2 at 431, 441-45 (statement by legislative director of ACLU listing articles and books based in whole or in part on information disclosed under FOI Act). See also Swallow, “Has The Freedom Of Information Act Worked — Or Has It Worked Too Well?” National Journal (15 August 1981) 1470, 1471 (information about campaign contributions, highway safety information. fall-out from nuclear testing, FBI surveillance of student activity, nuclear power plant safety).
85 See “Government, Business and the People's Right to Know” (1978) 3 Media Law Reporter 20-21 (discussion FOI Act service bureaus). See also Montgomery, Peters & Weinburg, “The Freedom of Information Act: Strategic Opportunities and Threats” (1978) Sloan Management Review 1-2 (use of FOI Act to obtain trade secrets and other information about competitors).
The merchandising of FOI Act services and information is not, of course, confined to obtaining business secrets. One enterprising company promotes its “Freedom of Information Kit” with an advertisement that promises: “Here's How to Find Out Which 'Enemies List' You 're On — Within IO Working Days”. Weinstein, “Open Season on 'Open Government' “ (19 June 1979) New York Times Magazine, 32. 85-86. This appeal to paranoia would seem to appeal to a large audience. See Hougan. “Pandora's Box” Harpers (August 1976) (log of FOI Act requests to CIA is an “index to the suspicions, insights, fantasies, and fears of those seeking information”: most requests seek personal files on submitter which are usually imaginary).
86 “Secrecy Mania'”, The New Republic (28 April 1982) 7-8.
87 Sometimes not even in theory. One student of the public sector, after reviewing various economic theories of “public goods”, concludes that scope of the public sector (public goods) is ultimately “defined by … the exercise of legitimate governmental decision processes” — Steiner, “Public Expenditure Budgeting” in A Blinder & R Solow (eds), The Economics of Public Finance (1974) 251. The difficulty comes, of course, in defining the realm of legitimacy where the process is used (“captured”) to produce benefits for particular private groups, at the expense of the larger public. See generally Aranson, Gellhorn & Robinson, “A Theory of Legislative Delegation” (1983) 68 Cornell Law Review I; J Mashaw, “Constitutional Deregulation: Notes Toward a Public, Public Law” (1980) 54 Tut L Rev 849.
88 The Independent Offices Appropriation Act of 1952, 31 USC § 483 admonishes federal agencies to prescribe fees to recover the value of any “work. service … benefit … license … or similar thing of value or utility performed” by the agency to “any person (including … corporations)”. The Supreme Court has held that the 1952 statute did not authorise regulatory agencies to tax regulated firms for the incidental benefits from regulation on the theory that regulation was assumedly for the benefit of the general public. National Cable Television Association v United States and federal Communications Commission (1974) 415 US 336. Moreover, there are many regulatory schemes whose essential public purpose is to confer private benefits at public expense. See, eg, Posner, “Taxation by Regulation” (1971) 2 Bell Journal of Economic & Management Science 22; Stigler, “The Theory of Economic Regulation” (1971) 2 Bell Journal of Economic & Management Science 3. Where government programmes are purposefuly redistributive in this respect, it would make no sense to tax the recipients of the benefits thus lavished upon them by the government's generosity. But both of these situations are distinguishable from the case where the government confers specific, private benefits that are separable from a general public interest objective and are not purposefully redistributive in character.
An analogy is the sale of public land resources to private individuals. No one seriously challenges, for example, the appropriateness of charging private timber firms for timber removed from public lands — even though there is some general public purpose served by periodic cutting of the forest (as a sylvicultural measure) and by the public consumption of timber.
89 The evidence and the statements of concern are scattered throughout the 1981 hearings on FOI Act amendments. See, eg, 1981 Hearings on FOi Act, vol 1 at 260-79 (statement of counsel for Proctor & Gamble, Inc); 428-53 (statement of Honeywell Inc); vol 2 at 121-210 (memorandum of Machinery and Allied Products Institute). One study of business information disclosure concluded that there was “a great deal more smoke than fire”. Ibid vol 1 at 209, 215 (statement of Professor Russell B Stevenson). See generally RB Stevenson, Corporations and Information (1980 Johns Hopkins Uni Press).
90 See, eg, “EPA Lets Trade Secret Loose in Slip-up, to Firm's Dismay”, Washington Post (18 September 1982) AI. A6 (release of trade secret formula for best-selling pesticide; attributed to inadvertent failure to black out confidential portions of document when it was copied for release).
91 See National Parks and Conservation Association v Morton (1974) 498 F 2d 765.
92 Under Chrysler v Brown (1979) 441 US 281 submitters have no right, under the FOI Act, to challenge agency disclosure. Thus the de novo review provisions of the FOI Act are inapplicable. Instead the submitter must proceed under the general review provisions of the APA, where the more deferential “arbitrary and capricious” standard applies.
93 See 1981 FOI Act Hearings, supra n 5, vol 2 at 12-13, 17. 25-29.
91 See, eg, ibid vol 1 at I 04, 110-123 (statement of Department of Defence General Counsel).
95 See, ibid vol 2 at 41-42.
96 See, ibid vol 1 at 973-1040 ( testimony of FBI Director. with lists of instances where sources have become uncooperative because of fears of disclosure under FOI Act). FOI Act critics seem to have been particularly traumatised by the efforts of Philip Agee — an ex-CIA agent who has been engaged in a campaign to expose covert CIA agents and activities around the world and to obtain CIA documents to aid in this campaign. See Agee v Central intelligence Agency (1981) 517 F Supp 1335. It is estimated that Agee's FOi Act demands cost the government more than $400,000 to process, 1342.
97 See, eg, S 1235, ibid vol 1 at 14-18 (sponsored by Senators D'Amato, Goldwater and Nickles).
98 See, eg, “Opening Federal Files” Newsweek (19 June 1978) 85 (disclosure of information in CIA file revealing belief that Israel has produced nuclear weapons inadvertently released as part of document on US atomic aid to India). See also 1981 FOI Act Hearings, supra n 5, at 984 (testimony of FBI Director).
99 See Relyea, “The Rise and Fall of the US Freedom of Information Act” (1983) IO Govern Public Rev 19, 27-28.
100 Executive Order 12356, 47 Fed Reg 14874-84 ( 1982). See supra n 38. Among other things the new Executive Order eliminates the requirement that there be “identifiable damage” to national security, substituting in its stead the standard, “reasonably could be expected to cause damage”. It also eliminates the automatic declassification process established under prior administrations. In its place the classification continues for as “long as required” unless the classifying authority establishes a specific date for termination. The new Executive Order expands the categories of documents that can be classified, including information concerning “capabilities of systems, installations, projects”, concerning “cryptology and confidential sources”. Other changes include provision for reclassifying previously unclassified information.
101 See supra n 39. The reviewing court's task is to match a particular classification against the classification criteria of the executive order. However, since exemption one is limited to classi • fications necessary in the “interest of national defense or foreign policy”, a court might in an appropriate case inquire into the question whether the criteria interpreted by the classi"•ingagency were appropriate. I am unaware of any decision so holding. Quite possibly such a decision would raise a constitutional issue, but it seems to me a permissible interpretation of what the FOi Act implies.
102 See 1981 FOI Act Hearings, supra n 5, vol 2 at 18-19.
103 Ibid at 45.
104 In 1981 The Justice Department was asked to produce a list of cases in which classifications had been overturned. Of eight cases reported, four were reversed on appeal or on rehearing; a fifth appeal was still pending but the court had stayed disclosure pending appeal; a sixth rejected the classification but still exempted the material under exemption 7; and in a seventh the agency conceded that disclosure would not be harmful. In sum, of eight cases only one was even arguably a problem. See 1981 FOI Act Hearings, supra n 5, vol 1 at 157-58. Even in that case it could not be said that it would have made any difference whether judicial review was de novo or under an arbitrary-capricious standard. Indeed, as noted earlier, the “de novo” review specified by the Act is in practice a deferential form of review given the legislative history admonishing the courts to give “substantial weight” to the agency's determination. See supra n 39.
105 This is the most prominent complaint made by enforcement agencies, who have offered fairly plausible evidence that it is a real problem. See, ibid at 852-64; 973-89 (testimony of FBI chief); ibid at 963-73 (head of CIA). A list of specific cases where information was withheld, purportedly because of a fear of FOI Act disclosure, is given in ibid at 990-1440.
106 The Papers — a “History of US Decision-Making Process on Vietnam Policy” — were classified, but were leaked to the New York Times and the Washington Post which published them after the government's unsuccessful attempt to enjoin them from doing so. See New York Times Co v United States (1971) 403 US 713. Daniel Ellsberg was indicted for divulging the Pentagon Papers, but charges were dismissed for “improper Government conduct”, N Dorsen, P Bender, B Neuborne, Emerson, Haber and Dorsen's Political and Civil Rights in the United States (4th ed 1976) 223.
107 See. eg, 1981 FOi Act Hearings, supra n 5, vol I at 847-65 (testimony of FBI Director).
108 See ibid vol 2 at I. 32-46.
109 A Tomer, Future Shock (1970) 311-315. See also H Simon, Administrative Behavior (3rd ed 1976) 279-287.
110 The literature on the subject of political and bureaucratic incentives and motivations is vast. A general overview is given in Aranson, Gellhorn & Robinson, “A Theory of Legislative Delegation” ( 1983) 68 Corn L Rev I. A more extended review is given in P Aranson, American Government: Strategy and Choice (1981).
111 See Aranson, Gell horn & Robinson, supra n 110, for an elaboration of this point. On the absence of market discipline particularly in the case of bureaucrats see, eg, A Downs, Inside Bureaucracy (1967) 29-30. See also W Niskanen, Bureaucracy and Representative Government (1971), advancing the thesis that bureaucrats seek to maximise their budgets as a means of enhancing their characteristic preferences (utility functions).
112 See, eg, Downs, supra n 111, at ch 10.
113 For a further elaboration of the difficulty of defining relevant “output” of bureaucracies in particular see Aranson. Gell horn & Robinson, supra n 110. The output definition problem in part reflects the ambiguities of purpose and function in legislative mandates of most agencies, and of the inherent conflicts among different constituent interests.
For example, is the purpose of the Interstate Commerce Commission to promote efficient transportation or to smooth the rough edges of competition through allocation of market shares? Most modern students of regulation agree that the latter is more consistent with the evidence than the former. See, eg, Posner, “Theories of Economic Regulation” (1974) 5 Bell Journal of Economic & Management Science 335. Even so there are ambiguities as to the form such allocation should take — how market shares should be determined, etc — making agency “output” almost impossible to define.
Even where there are few uncertainties about the general mandate of the agency, and the constituent interests to be served, the vagaries of defining a measurable output are vexing. For example, how does one measure the output of the Justice Department's Antitrust Divsion — by the cases filled or tried, “won” (itself an ambiguous term), number of injunctions, divestiture orders, or other “significant” sanctions, degree of reduction in concentration, reduction in other antitrust offences, etcetera?