The Freedom of Information Act is the statute that decides what the federal executive branch must show the public and what it may keep back, and its operative idea is an allocation of burden rather than a grant of permission. The enacted text supplies a formal identity of unusual precision. The Freedom of Information Act, Public Law 89-487, printed at 80 Stat. 250, was signed by President Lyndon B. Johnson on July 4, 1966, took effect one year later, and is codified at 5 U.S.C. 552. The act amended section 3 of the Administrative Procedure Act of 1946, the public information provision whose standards it displaced, and before the new text took effect Congress restated it in a June 1967 enactment, Public Law 90-23, so the section entered the Code in restated form.

The measure of the statute is the exchange it reversed. Under the provision it replaced, a person seeking a government record had to show a proper and direct concern in the subject, and an agency could withhold for good cause found, a pair of standards that left the decision effectively with the institution holding the paper. Under the statute that took its place, any person may obtain agency records without stating any reason at all, an agency may refuse only by fitting the record within one of nine enumerated exemptions, and the agency carries the burden of justifying the refusal, in its own administrative process and again in federal court if the requester sues. The right to ask changed little in the asking. What changed was who had to explain a refusal, and to what standard.

A profile of the statute therefore has to carry four things in one place. The first is the origin: a congressional campaign of eleven years, led by Representative John E. Moss of California against the opposition of the executive branch under presidents of both parties, that converted scattered complaints about secrecy into a disclosure code. The second is the inversion itself, the presumption of disclosure and the burden shift that every other feature of the statute serves. The third is the boundary structure: nine exemptions that define the only lawful grounds of refusal, among which the fifth, covering privileged internal memoranda, does work out of proportion to its length, and a coverage line that reaches executive branch agencies while leaving Congress and the federal courts entirely outside. The fourth is the amendment record, in which Congress returned to the statute repeatedly and, with one partial exception, tightened the justification an agency must supply rather than widening what a person may request.

Two cautions govern everything that follows. The exemptions are described in the form Congress enacted in 1966, because later judicial readings and later amendments altered several of them, and a profile that silently imports the altered forms misstates the statute its title names; where a later change is essential to understanding an exemption, the change is flagged with its date. And the statute is presented as a legal structure rather than as a verdict on any administration, because resistance to disclosure and pressure for it have both crossed party lines throughout the statute’s history, a pattern the signing of the act itself established.

The Regime the Statute Replaced and the Campaign Against Secrecy

A Statute Built Around Discretion

The Freedom of Information Act did not enter an empty field. It replaced a regime that gave federal agencies broad language for withholding records and gave a disappointed requester almost no means of contesting a refusal. That regime rested principally on section three of the Administrative Procedure Act of 1946, the provision devoted to public information. In operation, the provision functioned less as a disclosure law than as a withholding law. An agency controlled the record, an agency interpreted the standard governing release, and an agency decided whether the person asking had shown enough need to deserve an answer. A requester who disagreed had no judicial enforcement remedy under that provision. The practical burden sat on the requester to show need and entitlement, even though the information at issue described the actions of the government itself. Reform therefore required more than better administration. The governing text had to change.

Three phrases did much of the work. Access was limited to persons properly and directly concerned with the subject of a request, so a person unable to satisfy an agency of a sufficient personal concern could be turned away without reaching the substance of the record. Matter could be withheld in the public interest. Matter could be held confidential for good cause found. Each phrase sounded reasonable in isolation, and together the phrases formed a system in which the grounds for refusal were open ended and judgment about those grounds belonged to the institution holding the file. Public interest could be invoked against a disclosure that the same institution preferred not to make. Good cause could rest on reasons the agency itself found persuasive. Proper and direct concern could become a test of the requester rather than a test of the record. The requester had to explain a need for information before knowing precisely what the information contained, and the absence of a court remedy made a denial, for practical purposes, the end of the matter unless the agency chose to reconsider or political pressure produced a different result.

Housekeeping Authority Turned Toward Secrecy

A second legal prop supported the withholding regime. A statute dating to 1789, later codified at section 301 of title five of the United States Code, addressed the housekeeping functions of executive departments, including the internal management of departmental business and the custody and handling of departmental records. Agencies invoked that housekeeping authority as a basis for withholding information from the public. A provision about administering a department became, in agency hands, a claimed source of power to decide what the public could see. Withholding did not always rest on a considered legislative judgment that a category of information deserved protection. It could rest on authorities and interpretations accumulated over time, each step defensible as administration and the combined result amounting to a general power of refusal.

Congress addressed that prop before the larger reform arrived. An amendment enacted in 1958 stated expressly that the housekeeping statute did not authorize withholding information from the public. Reference materials of the Department of Justice and the Federal Bureau of Investigation and the history in the Department of Justice Guide to the Freedom of Information Act treat that amendment as the first statutory step toward reform. The step was narrow. It separated the management of records from the suppression of records, and it announced that authority to organize files and run a department did not carry authority to deny the public access to the information those files contained. It left section three in place, with the phrases about proper and direct concern, public interest, and good cause found, and it left requesters without a judicial remedy. Its importance lay in the principle it set and the precedent it created. Congress had begun to reclaim ground that agencies had occupied through interpretation, and the later campaign built on the recognition that each asserted source of withholding power had to be answered in statutory language.

Secrecy in the Loyalty Program Era

The legal defects of the old regime became harder to defend as executive secrecy expanded during the Cold War loyalty program era. Decisions about federal employment, loyalty, and dismissal carried severe consequences for the individuals involved and raised questions extending well beyond internal personnel management. Records about such decisions were records about the exercise of public power. When agencies withheld them, the withholding shielded the reasoning, procedures, and judgments through which the government had acted, as well as any sensitive material the records contained. An employee dismissed under a loyalty program had an obvious interest in the record of the case. Journalists, educators, and members of Congress examining the program had interests that section three forced into the language of proper and direct concern, a language poorly suited to oversight of a government wide system.

Hearings that documented agency withholding gave that conflict a concrete record. The reform record describes refusals by the Civil Service Commission over records concerning dismissed employees in the Eisenhower era loyalty program context. The episode mattered beyond the particular files. A civilian agency handling personnel matters could place its records beyond reach by relying on the discretionary standards available to it, which demonstrated that the problem did not arise only at agencies guarding military secrets or diplomatic communications. If a personnel agency could refuse records about dismissals that had ended government careers, administrative secrecy extended across the ordinary operations of the executive branch. The context also weakened the argument that agencies could be trusted to balance disclosure against the public interest without outside review. An agency asked to release records about dismissals had an institutional interest in the reputation of the program it administered. A system without judicial enforcement had no mechanism for separating a legitimate confidentiality interest from an institutional preference for quiet, and the case for a court remedy grew directly out of that structural conflict.

John Moss and the Subcommittee Investigation

The central congressional figure in the campaign was Representative John E. Moss of California. From 1955, Moss chaired the Special Subcommittee on Government Information of the House Government Operations Committee, a platform that gave the inquiry continuity and a public record. Continuity mattered because secrecy is a difficult subject for episodic oversight. An agency can answer a single inquiry with assurances and explanations of particular circumstances. A subcommittee returning to the subject year after year can compare assurances with practice, collect refusals across agencies, and show that individual explanations form a pattern. Under Moss, the subcommittee documented agency withholding through hearings that built the evidentiary foundation for legislation. Secrecy ceased to be an abstract complaint voiced by editors who wanted easier access to files. It became a documented practice, recorded in the proceedings of a House subcommittee and extending across the executive branch.

Moss also supplied a clear statement of what replacement legislation had to accomplish. In House statements in June 1966, Moss identified three changes at the core of the reform. Records had to be available to any person, not only to persons properly and directly concerned. Discrete enumerated exemption categories had to replace vague standards such as good cause and public interest. A requester had to receive a right to challenge a denial in federal district court. The three changes belonged together. A universal right of access without defined exemptions would leave agencies free to improvise new grounds of refusal. Defined exemptions without a court remedy would leave interpretation entirely to the agency invoking the exemption. A court remedy without a universal right would preserve the threshold inquiry that had excluded requesters under the old regime. Moss treated access, limits, and enforcement as parts of one design.

The length of the effort underscored the resistance that the design encountered. The Congressional Research Service, in the report titled The Freedom of Information Act: Background, Legislation, and Policy Issues, records that eleven years elapsed between the hearing held in 1955 and enactment in 1966. The count followed here is the Congressional Research Service count, although some secondary accounts describe the span as twelve years by counting differently. Eleven years of hearings, drafts, negotiation, and renewed effort separated recognition of the problem from passage of the remedy. That duration testified to more than legislative inertia. A proposal that shifted power from agencies to requesters and courts threatened established routines throughout the executive branch, and the subcommittee record made the status quo increasingly difficult to defend on its own terms. Agencies that praised openness in principle had to answer documented instances of withholding in practice, and agencies that invoked public interest had to confront the argument that the public interest included the ability to examine agency action.

A Coalition Outside Government

Congress did not carry the campaign alone. A coalition drawn from the institutions that depended on access to government information sustained pressure from outside the legislature. Newspaper editors, journalists, and press associations formed the most visible part of that coalition. For the press, access to records was not an abstract civic preference. Reporting on agency decisions required documents that agencies often preferred to summarize, delay, or withhold altogether, and a denial couched in the language of public interest left a reporter with no effective appeal and no independent account against which to test an agency statement. Press associations gave those individual experiences an organized voice, translating newsroom frustration into testimony and sustained advocacy for a statute with enforceable rights. Press advocacy also helped establish that access served the general public rather than a professional privilege for journalists alone, a point embodied in the any person standard that Moss advanced.

Educators and scientists added weight that reached beyond the news business. Educators concerned with the study and teaching of government had reason to insist that public institutions remain open to examination, because a system that conceals its records also conceals the material from which citizens learn how power operates. Scientists depended on government data and government sponsored research in fields where secrecy could obstruct inquiry and where agency control of information could shape what questions received answers. The presence of educators and scientists answered the suggestion that disclosure legislation served only reporters hunting for stories. The coalition presented access as a condition of scholarship, teaching, scientific work, and democratic oversight alike. That breadth mattered politically, and the coalition’s persistence across the eleven years recorded by the Congressional Research Service kept the subject before the public and before Congress. A shared diagnosis united the coalition: the need test forced a requester to justify curiosity, the vague standards allowed refusal without a defined reason, and no court existed in which a refusal could be tested.

Executive Opposition Across Administrations

Opposition from the executive branch matched the coalition in persistence and exceeded it in institutional weight. The Congressional Research Service record notes that no executive department or agency head supported the bill. That fact clarifies the nature of the contest. The proposal did not divide agencies into enthusiastic reformers and reluctant defenders of secrecy. The leadership of the executive establishment declined to support the measure across the board. Opposition extended through administrations of both political parties, which removed the dispute from ordinary partisan alignment and suggested that resistance arose from the institutional position of the executive branch itself. Officials who managed programs preferred control over the records of those programs, and officials who advised presidents preferred control over the timing and terms of release. A statute transferring initiative to any person who asked, backed by a court with authority to order release, reversed that control.

The stakes for the executive branch were practical as well as philosophical. An agency accustomed to withholding for good cause found faced a future in which a court could ask which exemption applied and whether the agency had sustained its justification within that exemption. An agency accustomed to screening requesters by proper and direct concern faced a future in which the identity and purpose of the requester carried no threshold weight. Executive agencies could defend particular exemptions and question particular drafting choices while the cumulative effect of the objections preserved the discretionary system intact. Sustained opposition from department and agency leadership reflected an accurate understanding of that rearrangement, and the eleven year duration of the campaign reflected the difficulty of legislating against the unified preference of the branch that would administer the law. Within the House, by contrast, a young Representative Donald Rumsfeld was a notable Republican supporter of the bill. His support illustrated that concern with congressional oversight and with the capacity of the legislature to obtain information about executive action did not map neatly onto party identity.

Passage Through Both Chambers

The vehicle for the remedy was Senate bill S. 1160, sponsored by Senator Edward V. Long of Missouri. The Senate passed the bill on October 13, 1965. The House of Representatives passed it on June 20, 1966, without a dissenting vote. Published tallies of the House action differ, so the reliable statement is the qualitative one: the House passed the bill without a dissenting vote. Unanimity at that stage carried significance after eleven years of executive resistance. Whatever disputes had attended drafting, and whatever objections agencies had pressed during the long campaign, the House recorded no dissent when the question reached a vote. The contrast between a unanimous House and an executive branch whose department and agency heads had offered no support sharpened the character of the result. Disclosure duties arrived not as an administrative reform granted by the executive but as a legislative command imposed over executive objection.

The enacted measure, Public Law 89-487, at volume eighty, page two hundred fifty of the Statutes at Large, amended the Administrative Procedure Act by replacing section three rather than merely supplementing it. Replacement mattered, because the old phrases did not survive alongside the new rights to create competing standards. The new section, codified at section 552 of title five of the United States Code, made agency records available to any person on request without a statement of reason or need, confined withholding to nine enumerated exemptions, placed the burden of justifying withholding on the agency, and gave the requester a federal court remedy in which the court decided the matter de novo. Affirmative disclosure duties formed part of the same design, requiring publication in the Federal Register and reading room availability of opinions, orders, policy statements, and staff manuals. The statute therefore attacked secrecy from two directions, obliging agencies to publish the basic materials of governance without waiting for a request and obliging agencies to answer requests for other records unless an exemption applied and the agency could defend its application.

A Signing Without Celebration

Victory in Congress did not produce enthusiasm at the White House. The Congressional Research Service describes President Lyndon B. Johnson as reportedly reluctant to sign the bill. The circumstances of the signing, reconstructed by the National Security Archive from LBJ Library documents, gave that reluctance visible form. Johnson signed the measure on July 4, 1966, at his Texas ranch, on the last day available before a pocket veto would have taken effect. No signing ceremony took place. No commemorative pens were distributed. The sponsors of the bill were not invited, and Moss, whose subcommittee work from 1955 had carried the campaign, was not present. The signing does not appear in the president’s daily diary, an omission that National Security Archive research identified in the library documents. Each detail pointed in the same direction. The president allowed the bill to become law while withholding the public embrace that presidents ordinarily confer on measures they wish to claim, and the timing left the impression of a decision deferred until the choice could be deferred no longer.

The treatment of the accompanying statements reinforced the message of the signing itself. Johnson personally weakened stronger openness language in the draft press statement, editing the public presentation of the new law toward caution. The signing statement that Johnson issued gave substantial emphasis to the exemptions and to the need to protect confidential communications. A reader of the statement encountered the limits of the statute presented with a prominence that the rights created by the statute did not receive in equal measure. The statement closed with an expression of pride that the United States is an open society, a closing line that sat in tension with the path by which the signature had arrived. Press secretary Bill Moyers later described the president as having to be persuaded at the last minute, a recollection consistent with the documented circumstances. The reluctance at the signing belonged to the same history as the opposition that preceded it, because department and agency heads had declined to support the bill during the campaign and administrations of both parties had resisted the reform across eleven years. None of that diminished the legal effect of the signature, although the absence of celebration signaled that the meaning of the new rights would be contested in implementation even after the governing text had changed.

Delay, Restatement, and Arrival

Congress built one final interval into the transition. The statute did not take effect upon signature. Its provisions became effective one year after enactment, on July 4, 1967. The delay gave the executive branch a defined period in which to prepare for duties that the branch had resisted throughout the legislative campaign. Agencies had to reckon with publication obligations, reading room availability, procedures for receiving and answering requests from any person, and the prospect of defending denials in federal district court with the burden on the agency. A year of preparation did not resolve the disagreements that the signing had displayed, but it removed surprise as a ground of complaint. When the effective date arrived, the duties had been announced in a text that replaced the familiar language of section three with enumerated exemptions and an enforceable right of access.

Before the effective date, Congress acted once more on the text. In June 1967, Congress restated the new section in an enactment numbered Public Law 90-23, so the operative text arrived in restated form. The 1966 Act replaced the old regime as a matter of enacted law, the 1967 restatement settled the form in which the replacement would operate, and the effective date then set the system in motion. The sequence formed a quiet conclusion to a lengthy contest. The measure of the achievement lay in the changed legal position of a person asking an agency for a record. The requester no longer had to establish proper and direct concern. The agency could no longer rest a refusal on public interest or good cause found at large. The agency had to point to an enumerated exemption, accept the burden of justification, and face a district court empowered to decide the question de novo if the requester sued. Every central feature of the new section answered a documented feature of the old order: the screening of requesters, the vague standards, the housekeeping claim corrected in 1958, the documented refusals of the loyalty program era, and the absence of any forum beyond the agency itself. The force of that answer depended on treating the replacement as complete rather than as an invitation to revive the habits that section three had protected.

The Inversion: Disclosure as the Default and the Burden on the Agency

The central achievement of the Freedom of Information Act of 1966 was structural rather than rhetorical. The statute did not merely encourage federal agencies to be more forthcoming, and it did not rest on an expectation that officials would voluntarily adopt a more generous attitude toward public inquiries. It rearranged the legal positions of the agency and the person seeking records. Before 1966, a requester faced an agency that controlled the records, interpreted the standards governing release, and could refuse access under broad formulations that left the requester with little practical recourse. After enactment, disclosure stood as the statutory default, withholding required justification within enumerated categories, and the agency carried the burden of sustaining a refusal. That rearrangement gave the statute its force. The right to ask mattered, but the allocation of proof and justification determined whether the right could survive an agency decision to say no.

The prior regime helps explain why such a rearrangement was necessary. Section 3 of the Administrative Procedure Act of 1946 addressed public information through qualifications that favored institutional discretion. Agencies could withhold material from persons who were not properly and directly concerned with the subject of the request. Matter could also be withheld in the public interest or held confidential for good cause found. Each formulation invited the agency to begin with its own judgment about the requester, the subject, and the wisdom of release. A person seeking information therefore had to establish more than the existence of a record and a desire to see it. The requester had to appear entitled to the record, concerned with it in the proper way, and unable to be refused on grounds of public interest or good cause as the agency understood those terms. The practical burden sat on the requester to show need and entitlement, even though the agency possessed the record and knew the reasons that might support withholding.

That arrangement also lacked an effective enforcement mechanism for requesters. The historical record for the 1946 statute is direct on the point: section 3 supplied no judicial enforcement for requesters. A denial therefore did not ordinarily lead to an independent judicial examination of whether the withholding standard had been correctly applied. The agency decided whether a person was properly and directly concerned, whether confidentiality served the public interest, and whether good cause existed, and the structure gave the disappointed requester no federal court remedy designed for the dispute. The imbalance was compounded by agency reliance on the housekeeping statute first enacted in 1789 and later codified at 5 U.S.C. 301. Agencies invoked that housekeeping authority as though authority to organize internal operations and custody of records also supplied authority to withhold information from the public. A 1958 amendment stated expressly that the housekeeping statute did not authorize withholding information from the public, and reference materials from the Federal Bureau of Investigation and the Department of Justice, together with the history in the Department of Justice Guide to the Freedom of Information Act, treat that amendment as the first statutory step toward reform. The step clarified a misuse of authority, but clarification alone did not create a general right of access, enumerated limits on withholding, or a court remedy. Those elements arrived with the 1966 statute.

The 1966 design began with a presumption of disclosure. Agency records had to be made available to any person on request. The proposition sounds simple because the statutory move was conceptually simple: access no longer depended on a preliminary demonstration that release would serve the requester in an approved way. The record was to be made available unless a recognized exemption permitted withholding. The presumption reversed the starting question in an administrative dispute. Under the older formulations, the question tended to be why the requester should receive the material. Under the Freedom of Information Act, the operative question became why the agency should be permitted to withhold it. That change in the starting question governed the rest of the design. A presumption without a disciplined set of withholding categories would have left agencies free to recreate broad discretion under new labels. A list of categories without a burden on the agency would have left requesters struggling to prove a negative about records never seen. A court remedy without fresh judicial decision would have risked converting review into deference to the initial refusal. The statute addressed the problem as a system, and each component reinforced the others.

The breadth of the phrase any person formed the first component. Representative John E. Moss of California, in House statements in June 1966, identified availability to any person as one of the three central changes, expressly contrasting that formulation with the older limitation to persons properly and directly concerned. The contrast eliminated status as a threshold test. A requester did not have to show a special relationship to the record, a professional reason for examining it, a personal stake in the underlying decision, or a need that the agency regarded as sufficient. The statute imposed no requirement for a statement of reason or need. The absence of such a requirement was not a procedural courtesy. It prevented agencies from grading requests according to the identity or purpose of the requester and then using an unfavorable grading to justify secrecy. Requests for the same record stood on the same statutory footing whether they came from a journalist, a scholar, a business, a person affected by an agency decision, or a reader with no purpose beyond curiosity. The reason for seeking a record might affect how the requester later used the material, but the statute did not make that reason the measure of entitlement.

The second component replaced vague withholding standards with nine enumerated exemptions as the exclusive withholding categories. Moss described the change in June 1966 as the substitution of discrete enumerated exemption categories for vague standards such as good cause and public interest. The importance of enumeration lay in exclusivity as much as in clarity. An agency could not treat good cause as a free standing reservoir of authority after 1966, and public interest could not serve as an all purpose answer to an unwanted request. Withholding had to be fitted within one of the categories Congress had specified. Enumeration therefore narrowed the field of permissible argument. The agency had to identify the kind of protected interest recognized by the statute and show that the requested material belonged within the relevant category. The requester and, if necessary, a court could then test that fit. Disagreement could focus on the text and scope of a particular exemption instead of dissolving into a general contest over whether disclosure seemed wise.

The nine categories enacted in 1966 show both the breadth and the limits of the settlement Congress reached. The first exemption covered matters specifically required by Executive order to be kept secret in the interest of national defense or foreign policy. The second covered matters related solely to internal personnel rules and practices of an agency. The third covered matters specifically exempted from disclosure by another statute. The fourth covered trade secrets and commercial or financial information obtained from a person and privileged or confidential. The fifth covered inter agency or intra agency memoranda or letters that would not be available by law to a party other than an agency in litigation with the agency. The sixth covered personnel and medical files and similar files whose disclosure would constitute a clearly unwarranted invasion of personal privacy. The seventh, as enacted, covered investigatory files compiled for law enforcement purposes, except to the extent available by law to a party other than an agency. The eighth covered matters contained in or related to examination, operating, or condition reports prepared by, on behalf of, or for the use of an agency responsible for regulating or supervising financial institutions. The ninth covered geological and geophysical information and data, including maps, concerning wells.

That enumeration preserved substantial areas of confidentiality, and an accurate account of the inversion should not describe the statute as a command that every agency paper be released. Congress recognized national defense and foreign policy secrecy, personal privacy, confidential commercial information, certain law enforcement files, financial institution supervision, and other specified interests. The inversion did not abolish secrecy. It changed the legal source and method of secrecy. Protection had to be traced to a category enacted by Congress rather than inferred from an agency view of good administration. The categories also differed in character. Some turned on the subject of the record, some on the source or status of the information, some on a stated privacy judgment, and some on the institutional function that produced the material. Those differences generated difficult questions of interpretation, but interpretation occurred inside a bounded structure. A court could ask whether material fell within a specified exemption without reopening the entire question whether agencies should generally control access to their files. Enumeration made withholding contestable because it made the asserted ground identifiable.

Why does the burden of justification matter in a records dispute?

The burden matters because it decides who must explain a refusal and who loses when the explanation falls short. Placed on the agency, it turns secrecy into a position that must be defended under an enumerated exemption, rather than a default that a requester must overcome by proving personal need or entitlement.

A records dispute contains a basic informational inequality. The agency knows what records exist, how they are organized, what they contain, which portions are sensitive, and what institutional interests may be affected by release. The requester often knows only that a category of records is believed to exist and that access has been denied in whole or in part. Placing the burden on the requester would preserve that inequality at the decisive point. The person with the least knowledge would have to demonstrate why material withheld from view should be disclosed, while the institution with complete knowledge could rely on a conclusory assertion. Placing the burden on the agency aligns the obligation to explain with possession of the relevant knowledge. The agency must sustain its action by connecting the record to an exemption and defending that connection. The requester does not have to prove entitlement from outside the file room.

The burden operated first in administration, before any judge became involved. An agency responding to a request could not treat denial as the natural outcome whenever a request reached a sensitive subject. The presumption of disclosure required the agency to consider the request under the statute and to justify withholding through an exemption. That requirement influenced the internal decision even when the requester never filed suit. Officials deciding whether to release material had to anticipate the ground on which a refusal would be defended. A weakly supported claim of confidentiality was no longer protected simply because the requester lacked a remedy or because the governing standard was phrased in terms of agency judgment. The statute made the reason for withholding part of the decision itself. Administrative convenience might explain a preference for nondisclosure, but the burden required a statutory justification, not a preference.

When administrative justification failed to persuade the requester, the statute supplied a federal district court remedy. Moss identified that right to challenge a denial in federal district court as the third of the central changes. The remedy completed the inversion because it moved final authority outside the agency whose records were at issue. A requester no longer had to accept the agency characterization of need, public interest, or good cause as the end of the matter. The district court decided the matter de novo, and the burden rested on the agency to sustain its action. De novo decision meant that the court addressed the dispute as a judicial question under the statute rather than merely asking whether the agency had offered some explanation for its refusal. The enforcement design therefore combined an external forum, fresh decision, and an agency burden. Each element limited the ability of an institution to become judge of its own secrecy claim.

Moss three changes therefore provide a concise summary of the design. The first change expanded the class of persons entitled to seek records by moving from persons properly and directly concerned to any person. The second change restricted the grounds of refusal by replacing vague standards with discrete enumerated exemptions. The third change supplied enforcement through a challenge in federal district court. No single change fully captures the inversion. Expansion of the requester class would have been incomplete without limits on withholding, limits on withholding would have been incomplete without enforcement, and enforcement would have been incomplete without the burden placed on the agency. The statute placed the burden on the agency and gave the requester a federal court remedy, with the court deciding the matter de novo. That sentence states the enforcement core in the form supported by the historical record, and the surrounding components explain why the sentence mattered. The right was not only a permission to submit a request. It was a legally supported demand that the government justify an exception to disclosure.

The statute also created a second channel that did not depend on requests at all. Alongside request based disclosure, the Freedom of Information Act imposed affirmative disclosure duties dating to 1966. One duty concerned publication in the Federal Register. Another concerned reading room availability for opinions, orders, policy statements, and staff manuals. Those duties are sometimes treated as background machinery beside the better known request process, but in structural terms they belonged to the same inversion. Certain categories of agency material were to be made available through publication or reading room access without waiting for a member of the public to identify and request each item. The affirmative channel recognized that a request system has limits. A person cannot easily request a policy statement, order, opinion, or staff manual whose existence is unknown. If access depended entirely on requests, agencies could benefit from public ignorance of the very materials that explained how the agencies acted. Affirmative disclosure reduced that problem by placing an obligation on the agency to bring specified material into public view.

The two channels served different functions and should not be collapsed into one another. The request channel allowed any person to seek agency records and placed the burden on the agency when an exemption was invoked to justify withholding. The affirmative channel required agencies to publish or make available defined materials without a request. The first channel was reactive in form, because it began when a request arrived, but it was not passive in legal effect, because the agency response was governed by a presumption of disclosure and a burden of justification. The second channel was proactive in form, because publication and reading room availability did not wait for an individual demand. Together, the channels attacked secrecy from two directions. One reduced the ability to refuse a known record without a statutory ground. The other reduced the ability to keep governing material obscure merely because no requester had discovered it. The distinction also clarifies the character of the statute. The Freedom of Information Act was not only a procedure for answering correspondence. It was a disclosure system built into the ordinary publication and recordkeeping practices of agencies.

That premise points toward the central claim about the statute. The operative change of 1966 was the burden allocation, more than the request right considered in isolation. The request right was indispensable, because without a right held by any person, the burden would have had no occasion to operate in an individual case. Nevertheless, a right to request records can coexist with a weak disclosure regime if the agency may refuse under open ended standards, if the requester must demonstrate need, or if no independent court can test the refusal. The historical problem under the Administrative Procedure Act was not that members of the public lacked the physical ability to write letters asking for information. The problem was that the legal structure surrounding the answer favored withholding. Agencies could invoke the requester status, the public interest, or good cause, and the requester had no judicial enforcement mechanism under section 3. The 1966 statute changed the legal structure surrounding the answer. That is why the burden deserves priority in an account of the reform.

The claim should be stated with care. Burden allocation did not operate alone, and the statute cannot be reduced to a single sentence about proof. The any person standard removed an important screening device. Enumeration confined the subject matter of justification. The district court supplied an institution capable of requiring justification to be made good. Affirmative disclosure bypassed requests for specified materials. The argument is comparative rather than exclusive: among those components, the burden allocation was the operative change because it determined who had to make the case for secrecy after a request was denied. It transformed disclosure from a benefit that a qualified person sought from an agency into a default that an agency had to overcome through a recognized exemption. Later congressional revisions belong to the longer history of the statute, but the analytical point was already present in the 1966 structure. Each later debate over disclosure necessarily confronted the question established at enactment: whether an agency justification was strong enough and specific enough to displace the default. The request right opened the door. The burden determined what happened when the agency tried to close it.

By the time the statute became effective on July 4, 1967, one year after enactment, the legal text had also been restated through a June 1967 enactment, Public Law 90-23. The restatement does not alter the structural account. The operative design that entered force treated disclosure as the default, confined withholding to nine enumerated exemptions, allowed any person to request records without stating a reason or showing need, imposed affirmative duties of Federal Register publication and reading room availability, and placed the burden on the agency in administration and in court, with de novo decision in federal district court. Those features marked a departure from the world of properly and directly concerned requesters, public interest withholding, and good cause refusals. The departure was not a promise that government would become transparent in every instance. It was a rule about who had to justify secrecy, under what categories, and before what authority. In a statute about information, that rule was the decisive transfer of power.

The Nine Exemptions: The First Four Categories

The exemption list doubles as a worked example of statutory architecture, and the method for reading a provision list of this kind, element by element and in enacted order, is set out in the guide to how to read a federal statute, which uses lists of this form as its teaching case.

Reading the List as a Structure

The exemption list in the Freedom of Information Act of 1966 rewards reading as a structure before reading as a catalog. The statute began from a rule of disclosure. Agency records had to be made available to any person on request, without any statement of reason or need, and the burden of justifying withholding rested on the agency in administration and in court. In litigation, the district court decided the matter de novo and the burden remained on the agency to sustain its action. Against that rule, the nine enumerated exemptions operated as exceptions, and the exemptions were exclusive categories. An agency that wished to withhold a record could not appeal to a general sense that secrecy served the public interest, and could not construct a new ground of withholding from the circumstances of a particular request. The justification had to fit inside one of the enumerated categories, stated in the statutory text, or the record had to be disclosed.

That design marked a deliberate break from the regime that preceded the Act. Under section 3 of the Administrative Procedure Act of 1946, agencies could withhold material from persons not properly and directly concerned, and matter could be withheld in the public interest or held confidential for good cause found. Representative John E. Moss of California, whose subcommittee work from 1955 onward documented agency withholding, identified the replacement of those vague standards as one of the central changes in the 1966 legislation. In statements in the House in June 1966, Moss described records made available to any person rather than only to persons properly and directly concerned, discrete enumerated exemption categories replacing vague standards such as good cause and public interest, and a right to challenge a denial in federal district court. The list therefore carried a double function. The list defined the protected interests that could justify secrecy, and the list also confined the vocabulary in which secrecy could be defended. Breadth inside a category remained possible, but breadth outside the categories had no statutory footing.

Each category was defined by the interest protected rather than by the identity of the requester or by the embarrassment that disclosure might cause. One category protected national defense and foreign policy secrecy. Another protected the internal personnel rules and practices of an agency. Another incorporated protections that Congress had enacted in other statutes. Another protected trade secrets and confidential commercial or financial information supplied by a person outside the government. The four categories operated through very different mechanisms. In the first category, the operative judgment came from an Executive order. In the second category, the operative judgment turned on the relationship between a record and an internal agency practice. In the third category, the operative judgment had been made by Congress in separate legislation. In the fourth category, the operative judgment turned on the source of information and on the character of that information in private hands. Attention to mechanism, and not only to subject matter, gives a clearer view of where disputes arise.

The baseline for the whole discussion is the wording as enacted in 1966. Precision about that baseline is essential, because later amendments altered the wording of several exemptions and altered the tools available to courts reviewing withholding. A description that imports later qualifiers into the 1966 text misstates the statute that Congress passed. Exemption 1 supplies the clearest example. As enacted, the category contained no requirement that material be properly classified, and that qualifier and judicial in camera review of classification arrived only in 1974 after EPA v. Mink. Exemption 3 supplies another example, narrowed in 1976 by the Government in the Sunshine Act after FAA v. Robertson. Where later narrowing occurred, the change is flagged briefly and the sequence is preserved, because the sequence explains the original design.

A claim of exemption also had to be sustained as to the matter withheld. The fact that a record touched a protected interest did not by itself complete the justification. The agency had to show that the withheld matter fell within the category invoked, under the wording of that category. That requirement of fit connects the first four exemptions despite their differences. Each discussion that follows identifies the interest the category protects, states the standard the agency had to meet under the enacted wording, and locates the pressure point where disputes concentrated. In each case, the pressure point grows directly out of the wording. A single adverb, a single qualifier, or a single reference to an external source of law often carried the entire weight of the argument between an agency seeking to withhold and a requester seeking disclosure.

National Defense and Foreign Policy Secrecy

Exemption 1, as enacted in 1966, covered matters specifically required by Executive order to be kept secret in the interest of national defense or foreign policy. The interest protected was the secrecy of defense and foreign policy information whose disclosure, in the judgment embodied in an Executive order, would compromise the security or the external relations of the United States. Protection of properly classified defense and foreign policy information commanded broad public acceptance, and Exemption 1 rested on that acceptance. A government that conducted military operations and diplomatic negotiations required a mechanism for keeping designated information secret, and the 1966 Congress supplied that mechanism by reference to the classification system rather than by attempting to list sensitive subjects in the statute itself. The category protected an interest, but the category also delegated the detailed definition of that interest to the Executive branch through Executive orders governing secrecy.

The standard the agency had to meet had two elements, both visible on the face of the text. The matter had to be required to be kept secret by an Executive order, and the requirement had to rest in the interest of national defense or foreign policy. The word specifically reinforced the first element. A general atmosphere of sensitivity around defense matters did not satisfy the text. The agency had to connect the withheld matter to a requirement imposed by Executive order. The second element confined the first, because an Executive order requiring secrecy for reasons unconnected to national defense or foreign policy would not supply a basis under the category. Within those limits, the enacted text contained a notable absence. The statute did not state that the matter had to be properly classified. That qualifier, together with judicial in camera review of classification, arrived in 1974 after EPA v. Mink, in amendments enacted as Public Law 93-502 over President Ford’s veto in November 1974. Under the 1966 wording, the agency case rested on the existence of an Executive order requirement tied to defense or foreign policy, without an additional statutory test of whether classification had been correctly applied to the particular matter.

The pressure point in Exemption 1 disputes followed directly from that structure. Because the statute incorporated an Executive order system, argument concentrated on the relationship between the order and the withheld matter, and on the degree of scrutiny a court could apply to an agency assertion that the relationship existed. The agency held the records, administered the classification system, and carried the statutory burden of sustaining withholding in court under de novo review. The requester had not seen the withheld matter and could contest the agency characterization only from outside the record. An agency description that was too general risked failing the requirement of specific connection to an Executive order. An agency description that was too detailed risked revealing the very information the category protected. The 1974 addition of the properly classified qualifier and in camera review addressed the scrutiny problem by giving the text an explicit correctness standard and by giving the court a procedure for examining classified material without public disclosure. The change lay in the justification required, consistent with the broader amendment pattern in which revisions hardened justification rather than redrawing the request right.

Internal Personnel Rules and Practices

Exemption 2, as enacted, covered matters related solely to the internal personnel rules and practices of an agency. The interest protected was administrative housekeeping. Agencies maintained rules and practices governing their own personnel, covering internal management of the workforce and routine procedures by which an agency organized its own operations. Disclosure of every internal instruction, however trivial or however focused on internal management alone, would have consumed agency and requester attention without illuminating agency action toward the public. The category recognized a distinction between records that govern the relationship between an agency and the public and records that govern the relationship between an agency and its own staff. At the narrow end of the category, the protected interest attracted little controversy, because a purely internal rule about internal procedure, with no bearing on public rights or public information, sits far from the central purpose of a disclosure statute aimed at agency records of public consequence.

The standard the agency had to meet turned on two relationships stated in the text. The matter had to be related to internal personnel rules and practices, and the relation had to be sole. The statute used the adverb solely, and that adverb performed the limiting work in the category. A matter related to internal personnel rules among other things, or a matter whose significance extended beyond internal management, faced an immediate textual obstacle. The agency had to sustain the position that the withheld matter related solely to the internal category. Unlike Exemption 1, the category required no reference to an external instrument such as an Executive order. Unlike Exemption 3, the category required no reference to another statute. The justification rested entirely on characterization of the matter itself, on a showing that the matter belonged to the internal personnel sphere and to that sphere alone. The burden of sustaining that characterization rested on the agency, as with every exemption, under de novo review in the district court.

The pressure point in Exemption 2 disputes was the breadth problem contained in the word solely. Read without discipline, the phrase internal personnel rules and practices could expand to cover a vast range of agency material, because almost every agency record touches personnel in some indirect sense, and almost every agency practice is carried out by personnel. An expansive reading would transform a housekeeping exception into a general management exception, and a general management exception would contradict the premise that the nine categories were exclusive and discrete replacements for vague withholding standards such as good cause. Read with discipline, the adverb solely confined the category to matter whose only relevant connection ran to internal personnel rules and practices, so that material with significance for persons outside the agency fell outside the category. Disputes therefore concentrated on characterization and on consequence. Argument addressed whether a rule was genuinely internal, whether a practice was genuinely a personnel practice, and whether a record that began as an internal instruction had acquired an external function through use in decisions affecting the public. The last form of argument proved especially persistent in principle, because agencies govern public-facing action through internal instructions, and a requester could contend that an instruction shaping treatment of the public related to more than internal personnel matters alone.

Incorporation Through Other Statutes

Exemption 3, as enacted, covered matters specifically exempted from disclosure by another statute. The interest protected cannot be stated in a single subject, because the category protected whatever interests Congress had chosen to protect in other legislation. The mechanism was incorporation. The Freedom of Information Act did not repeal or override specific secrecy provisions scattered across the statute book, and Exemption 3 preserved those provisions by referring to them. A statute governing a particular program, a particular kind of record, or a particular agency function might direct that certain matters be withheld from disclosure, and Exemption 3 carried that direction into the disclosure framework of 1966. The exemption operated as a junction between the general disclosure rule and the many particular judgments Congress had made in separate enactments. Exemption 3 was the category through which the exemption list remained open to the rest of the United States Code without ceasing to be enumerated, because each incorporated protection required a statute to support the protection.

The standard the agency had to meet had two elements parallel in form to the standard under Exemption 1. The matter had to be exempted from disclosure, and the exemption had to be imposed specifically by a statute other than the Freedom of Information Act itself. The word specifically performed the same confining function that solely performed in Exemption 2. A statute that merely permitted withholding in general terms, or a statute whose connection to the withheld matter was inferential, presented a weaker basis than a statute that specifically exempted the matter from disclosure. The agency had to identify the other statute, connect the statute to the withheld matter, and sustain that connection under the burden the Act placed on the agency in court. The justification depended on statutory interpretation twice over. A court had to construe the incorporating language of Exemption 3, and a court also had to construe the separate statute said to supply the exemption. The requester could prevail by defeating either step, by showing that the separate statute did not specifically exempt the matter or that the matter fell outside the separate statute properly read.

The pressure point in Exemption 3 disputes was dependence on the choices of Congress in other legislation, together with the risk that general withholding language in older or broader statutes might be pressed into service as a specific exemption. Before 1966, agencies had invoked the housekeeping statute of 1789 as withholding authority, until a 1958 amendment stated expressly that the housekeeping statute did not authorize withholding information from the public, a step that Federal Bureau of Investigation and Department of Justice reference materials and the Department of Justice Guide treat as the first statutory step toward reform. That history illustrated the danger Exemption 3 had to be read against. If any statute touching confidentiality could qualify, the incorporation mechanism could reimport the vague withholding authority that the 1966 Act displaced. If only statutes that specifically exempted matter from disclosure could qualify, the mechanism preserved deliberate congressional judgments without reopening the door to general claims. The amendment record shows that Congress returned to the boundary. The 1976 Government in the Sunshine Act, Public Law 94-409, narrowed Exemption 3 after FAA v. Robertson in 1975. Disputes under the category concentrated on specificity and on legislative intent in the separate statute, and the category remained distinctive because the decisive text often lay outside the Freedom of Information Act altogether.

Trade Secrets and Confidential Commercial Information

Exemption 4, as enacted, covered trade secrets and commercial or financial information obtained from a person and privileged or confidential. The interest protected was the willingness of private persons and organizations to supply information to the government, together with the private interest in the information itself. Agencies obtained commercial and financial information from persons outside the government in the course of regulation, procurement, licensing, and supervision. A submitter who faced public disclosure of a trade secret, or of financial information held in confidence, bore a direct competitive or commercial injury from disclosure, and a submitter who expected disclosure at the time of submission had reason to withhold information, to supply information in degraded form, or to resist agency collection efforts. The category addressed both concerns through a single text. Protection for trade secrets guarded the asset itself. Protection for privileged or confidential commercial or financial information guarded the relationship of confidence in which information reached the agency.

The standard the agency had to meet rested on two elements, source and character. The source element required that the commercial or financial information have been obtained from a person. Information the agency generated itself, through agency analysis or agency observation, did not satisfy that element, however sensitive the resulting agency record might appear. The character element required that the information be privileged or confidential, in addition to being commercial or financial in nature. Trade secrets formed a separate branch of the category, named independently in the text, so that a trade secret could qualify without separate proof under the commercial or financial branch, while commercial or financial information obtained from a person had to carry the additional character of privilege or confidence. Each element narrowed the category in a different direction. The source element excluded agency-created material. The subject element excluded information that was neither commercial nor financial. The character element excluded information that a submitter had treated as public or that had reached the agency without any basis for confidence. The agency invoking Exemption 4 had to sustain each applicable element under the statutory burden in litigation, and a requester could defeat the claim by defeating any single element, by showing an agency source, a noncommercial subject, or an absence of privilege or confidence.

The pressure point in Exemption 4 disputes was the character element, and in particular the meaning of confidential for information that a person had supplied to an agency, sometimes voluntarily and sometimes under compulsion. The text supplied the word without a definition, and argument concentrated on the circumstances of submission. Information supplied under an assurance of confidentiality presented the strongest agency case. Information supplied in a setting where disclosure was expected presented the weakest. Between those poles lay the contested ground, covering information supplied without express assurance, information marked confidential by the submitter without agency agreement, and information whose confidentiality the submitter asserted only after a request arrived. Within the 1966 text itself, the contest remained elemental. The submitter interest in confidence and the requester interest in records that explain agency decisions about regulated activity, public contracts, and public funds met inside a category that turned on provenance and character rather than on the importance of the record to public understanding. That design choice, consistent across the first four exemptions, left the importance of disclosure to be carried by the general rule and by the burden on the agency, while the category itself asked only whether the statutory elements were met.

The Nine Exemptions: From Privileged Memoranda to Well Data

The exemptions that closed the statutory list in 1966 addressed four distinct protective concerns in relatively compact language, and each carried a different relationship between the category of material described and the harm that disclosure was understood to threaten. Exemptions Six, Seven, Eight, and Nine covered personal privacy, law enforcement investigation, financial institution supervision, and geological information concerning wells.

Exemption Six protected personnel and medical files and similar files, disclosure of which would constitute a clearly unwarranted invasion of personal privacy. The wording linked protection to a defined class of files and to a qualifying standard. The files themselves were described by their personal character, with personnel files and medical files named as the central examples and similar files extending the description to material of a comparable personal nature. Protection did not follow from the label alone. The statute added that disclosure would constitute a clearly unwarranted invasion of personal privacy, and that phrase performed significant limiting work. The adverb clearly and the adjective unwarranted together set a threshold higher than embarrassment, discomfort, or a preference for confidentiality. An invasion of personal privacy had to be both unwarranted and clearly so before the exemption applied. The provision therefore recognized personal privacy as an interest capable of outweighing the general rule of disclosure, while framing that recognition in language that demanded a strong justification. It also distinguished private information about individuals from information about agency action itself. Personnel and medical material concerned the circumstances, health, and personal history of identifiable persons, and the exemption treated compelled public exposure of such material as a harm distinct from the institutional interests protected elsewhere in the statute.

Exemption Seven, as enacted in 1966, used language that later readers risk misunderstanding if the original text is replaced in memory by the revised text that followed amendment. The provision covered investigatory files compiled for law enforcement purposes, except to the extent available by law to a party other than an agency. Every element of that formulation mattered. The material had to be contained in investigatory files, those files had to have been compiled, and the compilation had to have occurred for law enforcement purposes. The threshold was therefore framed in terms of files assembled for an enforcement function, rather than in terms of a record by record demonstration of a particular injury from release. The qualifying clause that closed the provision preserved access where another law made the material available to a party other than an agency, which tied the exemption to the existing legal availability of the files in litigation and comparable settings. Stated at the strength of the enacted text, Exemption Seven was a categorical provision organized around the character and purpose of the file. It protected the enforcement process by protecting the assembled investigatory file, and it did so without enumerating separate kinds of harm that an agency was required to establish in an individual case.

That enacted formulation must be separated from the six harm test that governed Exemption Seven after later revision. The six harm test was a construction associated with the amendments of 1974, not language contained in the statute of 1966. The amendments of 1974 also changed the operative noun from files to records or information compiled for law enforcement purposes, and the threshold received further adjustment in 1986. Under the later framework, withholding turned on whether disclosure could be expected to produce specified consequences, described in the amended text as interference with proceedings, denial of a fair trial, invasion of privacy, exposure of confidential sources, disclosure of techniques, and danger to safety. Those categories gave courts and agencies a more granular vocabulary for evaluating enforcement claims, but that vocabulary belonged to the revised statute. Projecting it backward onto the Act of 1966 distorts the original design. The Congress that enacted the Freedom of Information Act chose a broad file based description for law enforcement material. A later Congress narrowed and restructured the provision by substituting an effects based inquiry for part of that categorical approach. The distinction is central to a statute profile, because it shows that the breadth associated with Exemption Seven in 1966 resulted from the text itself, while the more differentiated test resulted from deliberate legislative revision eight years after enactment.

Exemption Eight addressed a specialized regulatory setting in language tied directly to the function of financial supervision. It covered matters contained in or related to examination, operating, or condition reports prepared by, on behalf of, or for the use of an agency responsible for the regulation or supervision of financial institutions. The description reached beyond reports authored by the agency itself. A report prepared on behalf of such an agency, or prepared for the use of such an agency, fell within the same description, provided that the subject concerned examination, operation, or condition and that the responsible body performed a regulatory or supervisory role over financial institutions. The exemption therefore protected the informational basis of bank and comparable financial oversight. Examination and condition reports by their nature assess soundness, compliance, and operational state, and the provision treated material contained in those reports, and material related to those reports, as exempt from compelled disclosure. Unlike Exemption Six, the provision contained no qualifying phrase requiring a clearly unwarranted invasion. Unlike Exemption Seven as enacted, it was not organized around files compiled for a purpose. It was organized around a class of supervisory reports and the agency function that produced or used them.

Exemption Nine was the narrowest and most concrete provision in the group. It covered geological and geophysical information and data, including maps, concerning wells. The subject was defined by discipline and by object. The information had to be geological or geophysical in character, it could take the form of data or maps, and it had to concern wells. No broader category of energy information, land information, or resource information was described in the enacted language. That specificity set Exemption Nine apart from the more capacious provisions that preceded it in the list. Its protection attached to information whose value and sensitivity derived from knowledge of subsurface conditions associated with wells, and the text identified that concern directly rather than through a general standard. Together, Exemptions Six through Nine showed the range of drafting techniques within the nine categories. Congress used a privacy standard in one provision, a file and purpose test in another, a report and function description in a third, and a precise subject matter definition in a fourth. That variety makes the architecture of Exemption Five more striking, because the fifth exemption incorporated an external body of law by reference and protected a process, rather than describing a closed subject in self contained terms.

The threshold language of Exemption Five covered inter-agency or intra-agency memoranda or letters that would not be available by law to a party other than an agency in litigation with the agency. The opening words confined the provision to communications within the government. A memorandum or letter had to be inter-agency, meaning exchanged between agencies, or intra-agency, meaning exchanged within a single agency, before any further question arose. Material originating outside the government and material addressed to parties outside the government did not satisfy that threshold description merely because an agency held a copy in a file. The subject was therefore internal communication as such. The second part of the threshold then supplied the decisive legal test by reference to civil discovery. The memorandum or letter had to be one that would not be available by law to a party other than an agency in litigation with the agency. Congress did not list the protected contents in the text of the exemption. Instead, Congress borrowed the law governing disclosure between litigants and used that law to mark the boundary between internal agency material subject to disclosure under the Freedom of Information Act and internal agency material withheld from disclosure.

That borrowing gave Exemption Five its distinctive reach and its distinctive difficulty. Civil discovery privileges had developed to regulate the compelled exchange of material between adversaries in litigation, where each side sought information held by the other for use in a pending case. Transplanted into a public records statute, those privileges determined what any person could obtain from an agency without any litigation, any pending case, or any showing of need. The incorporation was not limited to a single privilege. The protections incorporated through Exemption Five include the deliberative process privilege, attorney work product, and the attorney-client privilege. Each privilege rested on a different rationale. Attorney-client protection concerned confidential legal communications. Work product protection concerned material assembled by or for counsel in anticipation of or during litigation. Deliberative process protection concerned the internal consultation that preceded agency decisions. By incorporating discovery law as a whole, the exemption gathered those separate doctrines under one statutory heading and made the outcome of a public request depend on doctrines fashioned for a different institutional contest. Within that incorporated body of doctrine, the Department of Justice Guide to the Freedom of Information Act identified the deliberative process privilege as the most commonly invoked privilege inside Exemption Five. That identification gives the privilege its central place in any treatment of the provision. The privilege protected the deliberative process of government by shielding material that reflected the consultative work through which an agency reached a policy or decision. The Department of Justice Guide recited three purposes that the courts have given for the protection. The first purpose was candor in internal advice, on the reasoning that officials and advisers would temper frank assessments if every draft, recommendation, and expression of doubt entered public circulation before or after a decision. The second purpose was avoidance of premature disclosure of proposals, on the reasoning that publication of options under consideration could create pressure around policies that an agency might refine, reject, or abandon after consultation. The third purpose was avoidance of public confusion from rationales never adopted, on the reasoning that disclosure of reasoning considered and discarded could mislead the public about the actual basis of agency action. The courts confined the deliberative process privilege through a two part test that required material to be both predecisional and deliberative. The predecisional element looked to timing and function in relation to agency policy. Material qualified only if it was generated before the adoption of an agency policy. A memorandum prepared to assist a pending choice, a recommendation offered while options remained open, and an evaluation written to inform a decision not yet made all bore the temporal and functional character that the test described, because each contributed to a decision that had not yet acquired final form. Material generated after adoption stood on different ground. An explanation written to state the basis of a policy already adopted did not become predecisional by describing earlier thought, and the courts treated post-decision explanations as disclosable. The standard citation for that proposition is the Supreme Court decision of 1975 in the case of the National Labor Relations Board and Sears, which recognized that explanations of decisions already made belonged to the public account of agency action rather than to the protected consultation that had preceded those decisions.

The deliberative element looked to content and character rather than to timing alone. Material qualified only if it reflected the give-and-take of consultation. Recommendations, drafts, proposals, suggestions, and expressions of judgment about the merits of a course of action embodied that consultative character, because each revealed how participants weighed alternatives and responded to one another during the formation of policy. A record that merely assembled information, without evaluating that information or recommending a choice, lacked the deliberative quality that the privilege required. The two elements operated together, and failure of either element defeated the claim. Material that was deliberative in tone but prepared after a decision failed the predecisional element. Material that was prepared before a decision but stated only facts failed the deliberative element. That combined structure gave agencies a defined claim and gave courts a defined inquiry. It also exposed the central tension of the privilege, because the material most likely to satisfy both elements was the material that most directly revealed why an agency chose one course over another, which was also the material of greatest explanatory value to a person seeking to understand government action.

The exclusion of purely factual material followed directly from the deliberative element. Facts assembled during consultation did not, without more, reflect the give-and-take of consultation. A compilation of data, a summary of conditions, a chronology of events, and a statement of measurements could inform deliberation without revealing the evaluative judgments of the participants. The courts therefore placed purely factual material outside the deliberative process privilege, even when such material appeared inside a memorandum that also contained protected recommendations. That treatment preserved a distinction between the evidence on which an agency acted and the advice through which officials interpreted that evidence. Disclosure of the factual portion allowed public examination of the factual basis of a decision while withholding the consultative judgments that the privilege covered. The distinction was demanding in application, because factual selection and arrangement could themselves imply judgment, but the governing principle remained clear. Factual material fell outside the privilege. An agency claim had to rest on deliberative content, not on the circumstance that facts had been gathered by officials engaged in a deliberative assignment.

The exclusion of adopted reasoning carried equal importance. When an agency adopted a recommendation, an analysis, or a rationale as the basis of its policy, that reasoning ceased to function as a private step toward a decision and became the explanation of the decision itself. The courts treated adopted rationales as falling outside the deliberative process privilege. A predecisional memorandum could therefore lose protection to the extent that the agency expressly adopted its reasoning or incorporated that reasoning by reference into a final action. The same principle governed post-decision explanations. Once an agency announced a policy, a statement explaining the policy represented the agency position, and withholding that explanation under a privilege designed for predecisional consultation would invert the purpose of the doctrine. The decision involving the National Labor Relations Board and Sears in 1975 supplied the standard authority for the disclosability of post-decision explanations. Together, the factual exclusion and the adopted reasoning exclusion marked the outer boundary of the privilege. Protection attached to consultation that remained internal and unadopted. Protection receded where material stated facts or where material stated the reasoning that an agency had made its own.

Why is the fifth exemption criticized more heavily than the others?

Exemption Five drew heavier criticism because it withheld agency reasoning itself, broadly defined through discovery privileges, without requiring proof that disclosure would cause demonstrable harm, allowing predecisional memoranda to shield explanations for decisions affecting the public even when release would cause embarrassment rather than injury to a protected interest.

The breadth described in that answer began with the incorporation mechanism. Other exemptions described content that Congress had identified as sensitive in itself. Classified defense material, personal medical information, confidential source material, financial examination reports, and geological data concerning wells each presented a category in which the nature of the content supplied the reason for protection. Exemption Five proceeded differently. Its coverage depended on the status of a communication within a decision making process and on the availability of that communication under discovery law. A memorandum could fall within the exemption because it was predecisional and deliberative, regardless of whether its contents concerned a sensitive subject. A candid assessment of a routine policy, a skeptical evaluation of a proposal with broad public consequences, and a recommendation that shaped a final rule could all satisfy the test if the timing and consultative character aligned. The exemption thus reached reasoning as reasoning. That reach separated Exemption Five from the subject matter exemptions and made the provision distinctive: it protected the reasoning process itself rather than a category of harmful content.

The absence of a harm requirement in the statute as enacted deepened that distinction. Exemption Six contained an internal measure of injury in the phrase clearly unwarranted invasion of personal privacy, which required a privacy invasion of a defined gravity and character. The later law governing Exemption Seven, after amendment in 1974, specified harms against which disclosure was to be measured. Exemption Five as enacted in 1966 contained no comparable requirement that disclosure cause demonstrable harm. If a memorandum satisfied the threshold of inter-agency or intra-agency communication and fell within an incorporated privilege, and if deliberative process material satisfied the predecisional and deliberative tests, the exemption applied without a separate showing that release would damage a concrete interest in a demonstrable way. Embarrassment, political inconvenience, and exposure of weak reasoning could therefore accompany a valid claim, even though none of those consequences amounted to injury of the kind that justified the subject matter exemptions. Criticism of the provision concentrated on that gap between coverage and demonstrated injury. The objection was not that deliberation lacked value. The objection was that protection extended by category of process, without proof of harm, could shield material whose release would serve public understanding and cause no genuine damage beyond discomfort to the institution that produced the material.

Frequency and litigation gave that structural objection practical force. The frequency claim requires precise statement. Exemption Five was among the most frequently invoked exemptions, and the deliberative process privilege was the most commonly invoked privilege within Exemption Five according to the Department of Justice Guide. Government wide citation rankings varied by agency and by fiscal year, and the privacy exemptions, including Exemption Six and the privacy component of Exemption Seven, frequently led government wide totals. No claim of universal primacy in invocation survives the citation record. The accurate statement is different and sufficient. Exemption Five was among the most frequently invoked exemptions, it was the most heavily litigated exemption, and it was the most heavily criticized exemption. Litigation concentrated on the exemption because its tests invited dispute. Parties disputed whether material preceded a decision or explained a decision, whether content was deliberative or factual, whether reasoning had been adopted, and whether an incorporated privilege applied at all. Each dispute required examination of the relationship between a document and a decision, a relationship that subject matter exemptions rarely placed at the center of a case in the same way.

The counter-reading prevents that criticism from becoming a claim that exemption as such lacked justification. Several exemptions protected interests with broad public acceptance. Personal privacy under Exemption Six, confidential sources and live investigations under Exemption Seven, and properly classified defense and foreign policy information under Exemption One commanded wide recognition as legitimate grounds for withholding. The serious criticism targeted the breadth of particular exemptions and the absence of a harm showing, not the existence of exemptions within a disclosure statute. Applied to Exemption Five, that distinction matters. The deliberation protection argument rested on candor, on avoidance of premature disclosure, and on avoidance of confusion from unadopted rationales, and those purposes addressed real risks in agency decision making. An official who expected immediate publication of every tentative judgment might write with caution that deprived a decision maker of frank advice. A proposal released before adoption might acquire a public life detached from the agency final position. A discarded rationale placed beside an adopted policy might confuse the record of what the agency had decided and why. The criticism of Exemption Five did not require denial of those risks. It required the further judgment that a process based privilege, applied without a demonstrable harm requirement as enacted, swept beyond the protection those risks justified and withheld reasoning that explained decisions affecting the public.

The place of Exemption Five in the nine exemptions therefore reflects a deliberate choice and an enduring tension within the Act of 1966. Congress displaced vague standards of good cause and public interest with enumerated categories, placed the burden of justification on the agency, and gave any person a right to challenge a denial in federal district court with the court deciding the matter de novo. Within that disclosure oriented structure, Congress also preserved the law of discovery privilege for internal memoranda and letters, and the courts built from that preservation a doctrine that protected predecisional deliberation while excluding facts and adopted reasoning. Exemptions Six, Seven, Eight, and Nine protected privacy, enforcement files as enacted, financial examination reports, and well data through descriptions anchored in content, function, or subject. Exemption Five protected consultation itself. That difference in object, joined to breadth of incorporation and the absence as enacted of a demonstrable harm requirement, explains the verified assessment that Exemption Five ranked among the most frequently invoked exemptions and stood as the most heavily litigated and most heavily criticized provision in the statute.

The nine exemptions reward a side-by-side reading, because the list is the statute’s whole law of refusal and each row pairs a protected interest with the criticism that interest attracts. The table states each exemption in its enacted 1966 substance, with the standard an agency must satisfy and the use each exemption sees in practice.

Exemption	What it protects	Standard for invoking it	Use in practice	Principal criticism
Exemption 1	Information required by Executive order to be kept secret in the interest of national defense or foreign policy	The record must fall within a category an Executive order commits to secrecy	Reaches material across the government wherever classified defense or foreign policy information is held	As enacted it carried no requirement that the information be properly classified, a gap Congress closed in 1974
Exemption 2	Matters related solely to the internal personnel rules and practices of an agency	The material must relate solely to internal rules and practices	Narrow in scope and confined to housekeeping material	The word “solely” invited agencies to argue that internal material with any internal sensitivity qualified
Exemption 3	Matters specifically exempted from disclosure by another statute	A separate statute must specifically exempt the material	Depends entirely on the statutes Congress has enacted elsewhere	It imports secrecy decisions made across the whole Code, so the exemption’s real content sits outside the act itself
Exemption 4	Trade secrets and commercial or financial information obtained from a person and privileged or confidential	The information must come from a person outside the agency and be privileged or confidential in character	Engaged whenever a request reaches commercial or financial information obtained from outside the agency	The meaning of confidential was contested for decades between submitter expectations and requester claims
Exemption 5	Inter-agency and intra-agency memoranda or letters that would be privileged in civil litigation with the agency	The material must be both internal to the government and covered by a recognized litigation privilege	Among the most frequently invoked exemptions and the most heavily litigated	It protects the reasoning behind decisions rather than any category of harmful content, and as enacted it required no showing of harm
Exemption 6	Personnel and medical files and similar files whose disclosure would be a clearly unwarranted invasion of personal privacy	Disclosure must clear a high privacy threshold, stated as clearly unwarranted	One of the privacy exemptions that frequently lead government-wide citation totals	The balancing it requires is fact-bound, so outcomes are hard to predict for requesters and agencies alike
Exemption 7	As enacted, investigatory files compiled for law enforcement purposes	In its original form the category turned on the file’s law enforcement character	The withholding category for records compiled by agencies with law enforcement missions	Its breadth as enacted let whole files be withheld on their character alone; the six-harm test of 1974 replaced category with consequences
Exemption 8	Examination, operating, and condition reports concerning the regulation and supervision of financial institutions	The record must be contained in or related to a financial supervision report	Confined by its terms to the reports of agencies that regulate or supervise financial institutions	It can shield the condition assessments on which supervisory failures later turn
Exemption 9	Geological and geophysical information and data, including maps, concerning wells	The information must concern wells and be geological or geophysical in character	Narrow by its terms to a single class of technical data about wells	Its single-industry specificity sits uneasily in a statute otherwise organized around general interests

Coverage Limits: The Institutions Outside the Statute

Every question about what the Freedom of Information Act requires begins one step earlier, with a question about where the statute applies at all. The disclosure duty, the right of any person to request records without stating a reason, the burden placed on the withholding institution, and the remedy in federal district court all operate inside a boundary drawn by a definition. An institution outside that boundary owes no duty under the statute, however secretive its practices may be and however public the interest in its papers may appear. An institution inside the boundary owes the duty in the form the statute prescribes, subject to the nine exemptions, and no argument about the sensitivity of government work in general removes it from coverage. Coverage and exemption are therefore different inquiries that produce different results. Exemption analysis asks whether a covered agency may withhold a particular record. Coverage analysis asks whether the institution holding the record is an agency in the statutory sense at all. Confusing the two inquiries produces much of the misunderstanding that surrounds the statute, because a requester who begins with exemption analysis has already assumed the answer to the question that decides the case.

The statute did not write its coverage definition from nothing. It incorporated the definition of agency found in section 551(1) of title 5, the definitional provision of the Administrative Procedure Act, and that incorporation carried both an expansive opening phrase and a set of express exclusions. The opening phrase reaches each authority of the Government of the United States. Read alone, that phrase might suggest a government wide sweep. Read in full, the definition narrows as it proceeds, first by describing the kinds of authorities covered and then by naming the institutions excluded. The definition is a list with a boundary, not a slogan about openness, and the boundary was part of the text that Congress enacted in 1966 rather than a qualification added afterward to limit a broader promise.

Which parts of the federal government sit outside the statute?

Congress and the courts of the United States sit outside the statute by express exclusion, along with territorial governments and the government of the District of Columbia, while the President’s immediate personal staff and purely advisory Executive Office units fall outside the agency definition as courts later applied it.

The categories of covered institutions deserve close attention because each category answers a possible escape route. The definition covers executive departments, the large departments through which the executive branch conducts the ordinary business of government. It covers military departments as a distinct category, so placement within the defense establishment does not remove an institution from coverage. It covers government corporations and government controlled corporations, so the corporate form of an institution does not remove it from coverage either. It covers independent regulatory agencies, named as a category in their own right, so the independence that marks such an agency in other areas of administrative law does not place it beyond the statute. It also covers establishments in the executive branch more generally, reaching authorities that do not fit neatly under a department, a corporation, or a regulatory commission. The cumulative effect is breadth within a branch. Congress described the executive establishment in several overlapping ways, leaving little room for an executive institution to argue that its particular organizational form places it somewhere other than inside the definition.

The breadth within the executive branch makes the exclusions more significant, not less. A definition that reaches executive departments, military departments, corporations under government control, independent regulatory agencies, and other executive establishments has plainly considered the variety of forms that governmental authority takes. Against that background, an express exclusion reads as a deliberate line rather than an oversight. Congress excluded itself. It excluded the courts of the United States. It excluded the governments of the territories and possessions of the United States. It excluded the government of the District of Columbia. Each exclusion appears in the definition itself, in the incorporated text of section 551(1). A court asked to apply the statute to an excluded institution does not weigh the public interest in disclosure against the interest in confidentiality. The court applies the definition, and the definition ends the inquiry.

The exclusion of Congress carries a particular weight because Congress wrote the statute. The institution that imposed a disclosure duty on the executive branch, that replaced the discretionary standards of the Administrative Procedure Act of 1946 with enumerated exemptions, and that gave requesters a federal court remedy against agencies did not impose the same duty on its own records. The exclusion covers Congress as an institution, and nothing in the incorporated definition distinguishes between kinds of congressional papers or between the chambers. A letter in the files of an executive department may be an agency record subject to request, while papers held by Congress itself sit outside the statute entirely. The distinction turns on custody by a covered institution, not on the subject matter of the document and not on the degree of public interest in the material. That custody based line recurs throughout the coverage design, and the congressional exclusion is its clearest illustration.

The exclusion of the federal courts carries a parallel weight. The statute assigns the district courts a central enforcement role, since a requester denied records by an agency may challenge the denial in federal district court, where the matter is decided de novo and the burden rests on the agency. The judiciary thus polices the disclosure duties of another branch while its own records remain outside the definition of agency. The arrangement is not a contradiction, because the definition and the enforcement provision do different work. The enforcement provision describes where an agency denial may be tested. The definition describes whose records may be demanded in the first place. A courthouse file and the internal papers of the courts are not transformed into agency records by the fact that a judge may order an agency to produce its records in litigation. The courts sit outside the statute as record holders and inside the statutory scheme as adjudicators, and holding both propositions at once is essential to an accurate account of the design Congress enacted.

The remaining express exclusions extend the boundary in two directions that requesters often overlook. The governments of the territories and possessions of the United States are excluded by name, so a territorial government does not become an agency under the statute by reason of its relationship to the United States. The government of the District of Columbia is excluded by name as well, so the District government does not become an agency under the statute by reason of the distinctive federal relationship to the District. Both exclusions reinforce the same structural point. Proximity to federal authority, subjection to federal law, or dependence on federal decisions does not substitute for inclusion in the definition. An institution either falls within a covered category and outside the exclusions, or the statute does not reach it, whatever other law may govern access to its records under a different regime.

The treatment of the presidency presents the most intricate coverage question in the section, because the presidency sits at the head of the very branch the statute covers. The incorporated definition, read with the legislative history as the courts later applied it, places two kinds of presidential institution outside the statute. The first is the President’s immediate personal staff. The second is a unit of the Executive Office of the President whose sole function is to advise and assist the President. The Supreme Court addressed the boundary in Kissinger v. Reporters Committee for Freedom of the Press in 1980, and the qualitative holding that emerged from the legislative history and its judicial application is that closeness to the President, in a purely advisory and assisting role, removes a unit from the agency definition. The reasoning implicit in the line is functional rather than personal. The question is what the unit does. A unit whose sole function is to advise and assist the President performs a role bound up with the President’s own deliberations, and the courts treated that role as outside the administrative apparatus that Congress subjected to the disclosure duty.

The functional character of the presidential staff exclusion matters as much as the exclusion itself. The test described in the legislative history and applied by the courts turns on sole function, a demanding formulation. A unit whose work extends beyond advising and assisting the President, into the administration of programs or the exercise of authority over persons outside the presidency, presents a different coverage question from a unit confined to advice and assistance. The statute and its history do not offer a general exemption for importance, sensitivity, or proximity to power in the abstract. They offer a boundary tied to a specific function, performed exclusively. That precision mirrors the larger design of the 1966 Act, which replaced vague standards such as good cause and public interest with discrete categories throughout the statute, in the exemptions and in the coverage definition alike. In coverage as in exemption, Congress legislated by category, and the categories, rather than a general judicial assessment of how open a particular office ought to be, decide the outcome.

Presidential records in the fuller sense are governed by a separate regime altogether. The Presidential Records Act of 1978 addresses the records of a presidency under its own framework, distinct from the agency records framework of the Freedom of Information Act. The existence of a separate statute confirms the structural reading of the coverage definition. Congress, when it turned to presidential materials, legislated separately rather than treating the Freedom of Information Act as the governing instrument. A requester seeking executive branch records proceeds under the Freedom of Information Act against a covered agency. A question about presidential records proceeds under the Presidential Records Act and its own rules. The distinction preserves the line drawn in 1966, in which the President’s immediate circle stood outside the agency definition, and it explains why arguments framed entirely within the Freedom of Information Act cannot settle questions that belong to the later and separate presidential records regime.

Coverage of an institution, however, settles only half of the analysis, because the statute attaches its duty to a particular kind of object. The Freedom of Information Act reaches agency records, and the courts settled on a reading of that concept with two principal elements. The record must be within the agency’s possession and control, and the record must be one that the agency created or obtained. Both elements do real work. Possession and control tie the duty to materials an agency actually holds and governs, so the duty follows custody. The created or obtained element ties the duty to the agency’s own documentary relationship with the material, so the statute reaches the records through which an agency conducts its business, whether the agency generated the record or received it from elsewhere and brought it into its files.

The decision in Forsham v. Harris in 1980 illustrates how the created or obtained element limits the statute even where information plainly concerns public business. Data generated by private grantees was held not to be an agency record, because the agency had not created the data and had not obtained it in the relevant sense. The Supreme Court’s reading kept the statute focused on records of the agency rather than on all information connected to federally supported activity. The distinction can produce results that frustrate requesters, since information may shape public decisions, rest on public funding, and remain outside the statute because no covered agency created or obtained it as a record. Its unit of concern is the record in agency hands, not the entire universe of information generated around government programs.

The same record centered design explains two further limits that the courts recognized as settled. The statute imposes no duty on an agency to create a record, and it imposes no duty on an agency to answer a question. A request under the statute is a request for records, framed so that an agency can locate existing material within its possession and control. A requester who asks an agency to explain a decision, to compile information into a new document, or to respond to an interrogatory has asked for something the statute does not supply, however reasonable the question may be. The agency’s obligation runs to the search for and production of records that exist, subject to the exemptions, and to the affirmative disclosure duties the statute imposes for defined categories of material such as Federal Register publication and reading room availability. Outside those duties, the statute is an instrument for obtaining records, not an instrument for compelling explanations, and the distinction between a record and an answer is one of the most consequential boundaries in daily practice under the Act.

The state and local exclusion is more complete than any of the federal boundaries, because it is not an exclusion within a federal definition but an absence of reach altogether. The Freedom of Information Act does not apply to state governments, to county governments, to municipal governments, or to any other institution of state or local authority. No provision of the statute subjects a state agency to the federal request procedure, the federal exemptions, or the federal court remedy that the statute creates against federal agencies. Access to state and local records runs through a different body of law entirely, since every state maintains its own public records law. The federal statute and the state laws may pursue comparable transparency aims, and a record denied under one regime may be available under another, but the regimes are legally distinct. A requester dealing with a state institution proceeds under the law of that state, with whatever rights, exemptions, procedures, and remedies that law provides, and the Freedom of Information Act contributes nothing to the analysis beyond the contrast its federal design supplies.

The most persistent confusion about coverage, and by the weight of commentary and practice the single most common misunderstanding of the statute, is the assumption that Congress and the federal courts are covered because they are parts of the federal government. The assumption has an understandable origin. In ordinary civic language, the federal government comprises three branches, and a statute celebrated as the federal government’s great disclosure law sounds, in that language, like a law for the whole of that government. The statute’s fame reinforces the assumption. Its name promises freedom of information without qualification, its request right extends to any person without a statement of need, and its litigation history features the courts ordering agencies to disclose material of the highest public consequence. A requester reasoning from the name, the fame, and the civic vocabulary of three coequal branches arrives naturally at the belief that legislative and judicial records stand on the same footing as executive records. The belief is natural, and it is wrong, because the statute does not legislate for the federal government in the civic sense. It legislates for agencies as a defined term, and the definition excludes Congress and the courts in express words.

The practical consequence of that misunderstanding falls on requesters at the very start of the process, before any exemption is considered and before any court could reach the merits of a disclosure dispute. A request addressed to Congress under the Freedom of Information Act invokes a statute that does not apply to the institution addressed. A request addressed to a federal court under the statute does the same. Congressional records are reached, if they are reached at all, under the rules that Congress itself maintains for access to its own materials. Judicial records are reached, if they are reached at all, under the rules that govern access within the judicial branch. Neither institution’s access regime is the Freedom of Information Act, neither applies the nine exemptions as the statute frames them, and neither supplies the statutory remedy of a de novo action against an agency under subsection (a)(4)(B). The misunderstanding therefore does more than misdescribe the law in the abstract. It misdirects effort, sending requesters to a procedure that cannot help them and obscuring the separate channels through which legislative and judicial materials may in fact be obtained.

A correct account of coverage also clarifies what the exclusions do not mean. Exclusion from the Freedom of Information Act is not a judgment that an institution’s records lack public importance, and coverage is not a judgment that an institution’s records must all be disclosed. Congress in 1966 drew an institutional boundary around the executive agencies it meant to regulate and an exemption structure for the records within that boundary, and the two halves of the design operate independently. Within the boundary, the presumption of disclosure, the exclusive exemptions, and the judicial remedy give the statute its force. Beyond the boundary, other law and other institutional rules govern, in the territories, in the District of Columbia, in the states, in the courts, in Congress, and in the presidency under its separate records statute. The coverage definition is therefore not a preliminary technicality. It is the first substantive decision Congress made in the statute, the decision about whose papers the new right of access would reach, and every other feature of the Act operates only inside the line that decision drew.

Six Major Revisions and the Pattern They Follow

Any count of amendments to the Freedom of Information Act depends on what kind of change the count is meant to capture. A count limited to major substantive revisions produces one figure. A count that also includes narrower measures and targeted restrictions produces another. The two figures do not contradict each other. Each figure answers a different question, and confusion arises only when a figure designed for one purpose is treated as a complete answer to the other.

The Department of Justice Office of Information Policy states that Congress has amended the Freedom of Information Act seven times since enactment, counting narrower measures in that total. The measures reflected in that broader count include a technical change in 1978, an enactment in 1984, a restriction concerning foreign requesters in 2002, and a rule concerning designation under Exemption 3 in 2009. For the different purpose of identifying enactments that reshaped administration and litigation, six major substantive revision acts supply the relevant sequence. The six acts are the amendments of 1974, the Government in the Sunshine Act of 1976, the Freedom of Information Reform Act of 1986, the Electronic Freedom of Information Act Amendments of 1996, the OPEN Government Act of 2007, and the Freedom of Information Act Improvement Act of 2016. Each act entered the statute at a distinct pressure point, and each act left a distinct mark on the justification an agency had to supply for withholding, on the tools available to a court reviewing a denial, or on the consequences that followed delay and defeat.

Read together, the six acts support a specific claim about direction. The major revisions worked mainly by hardening the justification an agency must supply, rather than by expanding what a person may request. The request right defined in 1966 reached agency records at the request of any person, without a statement of reason or need, and that formulation stood essentially unchanged through every major revision that followed. No major revision broadened the class of requesters beyond any person, because the class had been framed in the broadest available terms at enactment. No major revision added a general requirement that an agency create records or answer questions, a limitation that the courts preserved in the settled reading of the statute. Instead, Congress returned repeatedly to the mechanics of justification. Deadlines disciplined the pace of response. Segregability required release of portions that could be separated from exempt material. Standards for classification and for law enforcement withholding narrowed the ground an agency could occupy. A foreseeable harm standard tied withholding to an identifiable injury. Judicial tools allowed a district court to test an agency explanation against the record itself. Fee provisions and fee consequences attached a price to delay, to withholding, and to litigation loss. The pattern is one of procedural and evidentiary pressure applied to a stable substantive right.

One act complicates that pattern and requires honest placement inside it. The Freedom of Information Reform Act of 1986 rebuilt the fee structure, a change consistent with the larger attention to administration and consequences, and at the same time broadened the threshold for withholding under Exemption 7 from a standard expressed through the word could to a standard expressed through the phrase could reasonably be expected. That change widened the ground available to an agency under one exemption, and the widening ran against the direction taken by the other major revisions. The exception does not erase the pattern. The exception defines the limit of the pattern and prevents a tidy account from overstating congressional consistency. Across five decades, Congress pressed agencies toward fuller justification in most major revisions, accepted a broader law enforcement threshold in one major revision, and left the underlying request right in the form established in 1966.

How have the major revisions made withholding harder to justify?

Each major revision tightened the showing required for withholding by imposing deadlines, requiring release of reasonably segregable portions, adding foreseeable harm and classification standards, expanding judicial review tools, and attaching fee consequences to delay or defeat, while leaving the basic request right intact.

The hardening occurred through devices that reinforced each other. A deadline without a consequence does not test the merits of withholding. A substantive standard without a judicial tool leaves testing dependent on acceptance of an agency description. A harm standard without fee consequences leaves a requester bearing the full cost of forcing compliance. The six revisions added layers across all of those dimensions. Deadlines addressed time, segregability addressed the unit of release, harm and classification standards addressed the quality of the reason, in camera review addressed verification, and fee rules addressed incentives. An agency defending a denial after the sequence faced questions about timing, separation, foreseeable injury, the record itself, and costs, rather than a single question about category membership alone.

The amendments of 1974 supplied the largest single addition to that structure, and the circumstances of enactment underline the seriousness of the intervention. Congress enacted the amendments as Public Law 93-502 over the veto of President Ford in November 1974. The act imposed deadlines for agency response and addressed fees, including attorney fees in litigation. The act also codified the doctrine of reasonably segregable portions, a doctrine that the courts had articulated in Environmental Protection Agency v. Mink in 1973. Under that doctrine, exempt material inside a record did not by itself justify withholding the whole record when a portion could reasonably be separated and released. The act also narrowed Exemption 1 and Exemption 7. For Exemption 1, the act added the qualifier that material be properly classified, and the act provided judicial in camera review of classification. A claim of secrecy therefore rested on more than an assertion that an executive order covered the material, and a court gained a means of examining the classified material in chambers.

The treatment of Exemption 7 in 1974 illustrates the same movement from category toward justification. As enacted in 1966, Exemption 7 covered investigatory files compiled for law enforcement purposes, except to the extent available by law to a party other than an agency. The 1974 amendments changed files to records or information compiled for law enforcement purposes and attached a set of specified harms. The amended structure reflects six kinds of harm, covering interference with proceedings, fair trial, privacy, confidential sources, techniques, and safety. Withholding after 1974 therefore turned on a connection between the compiled material and an enumerated consequence within the exemption, rather than on placement inside an investigatory file alone. That shift demonstrated a legislative preference for harm language over status language when Congress revised an exemption in a disclosure direction, a preference that returned in stronger and more general form in 2016.

The Government in the Sunshine Act of 1976, Public Law 94-409, narrowed Exemption 3, the exemption for matters specifically exempted from disclosure by another statute. The narrowing followed Federal Aviation Administration v. Robertson, decided in 1975. Exemption 3 posed a structural risk different from the risk inside an individual withholding decision, because the exemption incorporated other statutes by reference and therefore allowed the boundary of disclosure to move through enactments framed for other purposes. A broad reading of Exemption 3 threatened to let a general statutory reference displace the disclosure rule without a specific congressional judgment about secrecy under the Freedom of Information Act itself. The 1976 narrowing responded by tightening the conditions under which another statute could supply a basis for withholding, limiting an incorporation device that could expand withholding authority through indirection.

The Freedom of Information Reform Act of 1986 arrived within Public Law 99-570 and pressed in two directions at once. The act rebuilt the fee structure, continuing the congressional practice of using fees as an instrument for shaping agency and requester incentives in administration. The same act, however, broadened the threshold under Exemption 7. The earlier formulation asked whether release could cause one of the specified harms described in the exemption. The 1986 formulation asked whether release could reasonably be expected to cause such a harm. The addition of reasonable expectation lowered the certainty an agency had to establish about a future injury and widened the range of law enforcement material that could satisfy the threshold. Candor about that change is essential to the pattern claim. The claim describes the main direction of the major revisions, not an unbroken line. The 1986 act stands as the partial countercurrent, a reminder that congressional revision has served enforcement protection as well as disclosure, and that the hardening thesis depends on the whole sequence rather than on selective attention to favorable examples.

The Electronic Freedom of Information Act Amendments of 1996, Public Law 104-231, addressed electronic records and reading rooms. Records held in electronic form raised questions about whether a disclosure statute framed around paper files would reach the places in which agencies in fact maintained information. The 1996 act answered by bringing electronic records and reading rooms inside the statutory design, adapting affirmative disclosure duties that had dated to 1966. The revision did not broaden the class of requesters or the grounds for a request. The revision preserved the reach of the established right against a change in recordkeeping practice that could otherwise have narrowed the right in operation without any amendment to the request language itself.

The OPEN Government Act of 2007, Public Law 110-175, returned to administration and to the cost of delay. The act provided request tracking, created the Office of Government Information Services as an ombudsman, and attached fee consequences to missed deadlines. The ombudsman supplied an institutional channel for resolving disputes outside a federal courtroom, which mattered in a system in which litigation expense could decide whether a legal right received testing at all. Delay often functions as a practical form of withholding, whatever the merits of an eventual exemption claim, and a consequence tied to delay reduced the ability of an agency to obtain through time an outcome that the exemption analysis might not have supported on the merits.

The Freedom of Information Act Improvement Act of 2016, Public Law 114-185, codified a foreseeable harm standard for withholding generally, codified a presumption of openness, and placed a limit of twenty five years on the deliberative process privilege inside Exemption 5. The foreseeable harm standard generalized the logic that the 1974 amendments had applied inside Exemption 7. Withholding required an agency to connect disclosure to a foreseeable harm, rather than to rest on category membership alone. Codification of a presumption of openness stated the interpretive starting point in the text of the statute, reinforcing the allocation of burden that the enforcement design had carried since 1966, with the district court deciding de novo and the burden resting on the agency to sustain withholding. The limit on deliberative process protection removed that privilege from records at or beyond twenty five years in age, placing a temporal boundary around a privilege that otherwise turned on the status of material within a decision process. The 2016 act therefore gathered several threads from the earlier revisions into general law.

The synthesis across the six acts rests on contrast as much as on accumulation. Between 1966 and 2016, the sentence granting access to any person without a statement of need remained the stable center of the statute, while nearly every surrounding mechanism changed at least once. Congress legislated about clocks, portions, classification quality, statutory incorporation, fees, electronic form, tracking, an ombudsman, harm, and the age of deliberative material. Experience under the statute exposed failures of justification, delay, verification, and incentive more readily than failures in the definition of the requester, and reform followed the observed failures. The result was a statute whose promise read much as the promise of 1966 had read, surrounded by a far more demanding apparatus for any agency seeking to defeat that promise in a particular case.

The session-by-session narrative of those six revisions, including the veto fight of 1974 and the long intervals in which Congress left the text alone, is carried in full by the history of the Freedom of Information Act amendments.

The Counter-Reading: Exemptions as Protections Rather Than Loopholes

A familiar charge treats the nine exemptions as loopholes, as if each exemption represented a defect through which the central promise of disclosure escapes. That framing fails at the level of design. The exemptions formed part of the enacted statute from the beginning, alongside the presumption of disclosure, the burden placed on the agency, and the federal court remedy. Congress in 1966 chose enumerated categories over the vague standards of the earlier Administrative Procedure Act regime, under which matter could be withheld in the public interest or held confidential for good cause found. Enumeration itself was a disclosure reform, because a finite list confined withholding to stated grounds and denied agencies a general discretion to refuse material that officials preferred to keep inside the government. A reading that treats every exemption as a betrayal of the statute therefore mistakes the structure of the compromise that made a justiciable disclosure right possible. The serious question raised by the exemptions concerns breadth, administration, and proof of harm in particular categories. The serious question does not concern the legitimacy of protecting every interest that any exemption names.

Several exemptions protect interests that command broad acceptance across otherwise divided views of government secrecy. Exemption 6 covers personnel and medical files and similar files, disclosure of which would constitute a clearly unwarranted invasion of personal privacy. Agency files contain intimate information about individuals who supplied information to the government, who sought benefits or employment, or who became subjects of administrative attention without surrendering every detail of private life to public inspection. Exemption 7 protects, among other interests, confidential sources and live investigations. Disclosure that identifies a person who supplied information under an expectation of confidentiality can dry up cooperation on which law enforcement depends and can expose the source to retaliation. Disclosure that interferes with proceedings can compromise enforcement before a case reaches a court. Exemption 1 protects properly classified information in the interest of national defense or foreign policy, a responsibility that nearly every account of statecraft treats as legitimate. The acceptance surrounding those interests does not settle disputes about any particular withholding. Acceptance does establish that the exemptions as a class cannot be dismissed as pretexts, because several exemptions guard people, sources, proceedings, and defense information against injuries that disclosure itself would cause.

The strongest form of the counter-reading emerges inside Exemption 5, the exemption that attracts the most sustained criticism. Exemption 5 covers inter-agency or intra-agency memoranda or letters that would not be available by law to a party other than an agency in litigation with the agency, incorporating privileges known to discovery practice. The Department of Justice Guide to the Freedom of Information Act identifies the deliberative process privilege as the most commonly invoked privilege inside Exemption 5, alongside attorney work product and attorney-client privileges also incorporated through the exemption. The deliberative process test developed by the courts requires material to be both predecisional, meaning generated before adoption of an agency policy, and deliberative, meaning reflective of the give and take of consultation. The Department of Justice Guide recites three purposes that the courts state for the privilege. The privilege encourages candor in internal advice. The privilege avoids premature disclosure of proposals. The privilege avoids public confusion from rationales that an agency never adopted.

Each purpose carries force when stated at full strength. Candor matters because officials and advisers who expect every tentative judgment to enter public circulation have an incentive to soften warnings, to avoid recording doubts, and to move consequential discussion into channels that leave no record at all. The quality of a final decision depends in part on the freedom with which subordinates can tell a decision maker that a proposal rests on weak evidence or carries an unexamined cost. Protection against premature disclosure matters because a proposal under consideration has a different status from a decision that binds the public. Release before adoption can trigger reliance and reaction directed at a course that an agency may abandon after consultation, and the reaction can distort the consultation itself. Protection against public confusion matters because government speaks through adopted policy, and circulation of discarded rationales as if those rationales explained government action can mislead the public about the basis on which power was exercised. The deliberation argument, at full strength, therefore does not rest on official comfort. The argument rests on decision quality and on the difference between exploration and action.

The openness argument answers at equal strength, and the answer begins with the distinctive character of Exemption 5 among the exemptions. Exemption 5 is among the most frequently invoked exemptions, is the most heavily litigated exemption, and is the most heavily criticized exemption. The distinctiveness lies in the object of protection. Most exemptions protect a category of content whose disclosure threatens a stated injury, such as an invasion of privacy, exposure of a source, or damage to defense interests. Exemption 5 protects the reasoning process itself. Material qualifies through a relationship to a decision in formation, not through sensitivity of content alone. A memorandum can fall inside the privilege because advisers exchanged views before adoption, even when the views concern a matter of large public consequence and even when disclosure of the views would reveal no secret technique, no private fact, and no defense information. The process status of the material does the decisive work.

That feature generates the central objection. Reasoning shielded by process status rather than by content sensitivity can include reasoning that explains, in substance, why an agency acted as the agency acted, particularly when deliberative material reveals the considerations that survived consultation in all but formal adoption. Secrecy for such material can leave the public with a final rule, order, or policy and with an official explanation, while withholding the internal account that would allow evaluation of whether the official explanation captures the actual grounds of decision. Embarrassment sharpens the objection. As enacted, Exemption 5 imposed no requirement that disclosure cause demonstrable harm, and the breadth of the exemption combined with the absence of a harm requirement forms the core of the criticism directed at the exemption. Material whose release would embarrass officials, expose weak analysis, or reveal disagreement inside an agency can satisfy a process test without any showing that release would injure defense, privacy, enforcement, or another interest of the kind protected elsewhere in the statute. A privilege justified by candor can therefore extend in application to material protected in effect by institutional discomfort, and the distance between the justification and the application supplies the openness argument with its strongest ground.

Boundaries developed by the courts limit that objection without dissolving the objection. Factual material falls outside the deliberative process privilege, and adopted rationales fall outside the privilege as well. In National Labor Relations Board v. Sears, decided in 1975, post-decision explanations were treated as disclosable, a holding that preserves public access to the explanation an agency gives for action after action occurs. The line between protected deliberation and disclosable explanation matters, because the line prevents Exemption 5 from swallowing every document that touches a decision. The twenty five year limit enacted in 2016 adds a further boundary, removing deliberative process protection from older records and recognizing that the candor interest weakens as the decision recedes into history. Codification of a foreseeable harm standard in the same act speaks to the harm objection in general form. Each boundary concedes something to the openness argument while preserving a core for the deliberation argument, and the pattern of concession explains why argument over Exemption 5 has remained productive rather than settled.

Assessment of the counter-reading therefore requires separation of two disputes that public argument often merges. One dispute concerns whether government may protect privacy, confidential sources, live investigations, and properly classified defense information inside a disclosure statute. That dispute commands broad agreement in favor of protection, and the exemptions serving those interests rest on injuries that disclosure can directly cause to identifiable people, proceedings, and responsibilities. Another dispute concerns how broadly particular exemptions should reach and whether withholding should require a showing of harm. That second dispute is durable because each side protects a genuine public value that the other side cannot absorb without loss. Deliberation protection guards the conditions under which officials test ideas before the government commits. Openness guards the conditions under which the public tests the reasoning of a government after commitment. A harm requirement mediates between those values by asking whether secrecy in a given case prevents an injury or merely prevents scrutiny, and breadth determines how much material a court must examine before the mediation begins. The Freedom of Information Act, as amended through 2016, embodies an ongoing negotiation over exactly those two variables, exemption breadth and harm. Neither the loophole charge nor an unqualified defense of the exemptions captures that negotiation. The durable dispute lives inside that design, at the point where a specific exemption meets a specific record and an agency must state, with enough precision to survive review, what injury disclosure would cause or what process interest secrecy would preserve.

How a Request Moves and Where the Profile Points Next

The statute’s design becomes concrete in the path of a single request. A person identifies records and asks the agency that holds them, describing the records sought with enough precision for agency staff to locate them; no reason for wanting them may be demanded, and the identity or motive of the requester plays no part in the legal analysis. The agency searches, reviews what the search produces line by line against the nine exemptions, releases what no exemption covers, and withholds the rest with an explanation the requester may appeal inside the agency and then challenge in federal district court, where the burden of sustaining the withholding sits on the agency and the court decides the matter anew. Fees, deadlines, and the administrative machinery that surrounds each of those steps belong to the statute’s operating layer, and the full procedural account, from the drafting of a request letter to the anatomy of a lawsuit, is carried by the companion profile of how the Freedom of Information Act operates inside agencies, which takes up where this profile’s structural account ends.

The statute also sits inside a family of disclosure laws enacted across the same reform era, and the family resemblance is instructive. Where the Freedom of Information Act compels the release of records to whoever asks, the financial disclosure title of the Ethics in Government Act of 1978 compels named officials to publish their own financial interests on a standing basis, and the two designs, request-driven disclosure of agency paper and automatic disclosure of officials’ interests, amount to twin instruments of the same transparency design, one working on the records of institutions and the other on the interests of officials. Reading the two statutes together shows the choice Congress made in each case about who initiates disclosure, and why the burden question, answered one way for records and another for officials, organizes both.

For a reader building a study path through the statute, the sequence that repays effort runs from the text outward. The enacted section repays a slow first reading with attention to the order of its parts: the affirmative publication duties, the request right, the exemption list, and the enforcement clause, in that arrangement, state the whole theory of the statute before any commentary begins. The amendment record then shows the theory being defended and adjusted, and the case law on the fifth and seventh exemptions shows where the theory met its hardest resistance. A structured notebook for working through the exemptions against the amendment record is available in the legislation study notebook on VaultBook, and the wider path through the transparency cluster and the process articles behind it is mapped in the US legislation study guide. Held together, the profile’s claim is the one the statute’s own history keeps confirming: the Freedom of Information Act mattered less for creating a right to ask, which in form any person might always have asserted, than for relocating the burden of justification onto the institution that refuses, and every serious fight over the statute since 1966 has been a fight over how heavy that burden should be.

Frequently Asked Questions

Q: What did the Freedom of Information Act change about disclosure?

The Freedom of Information Act replaced section 3 of the Administrative Procedure Act of 1946, a regime under which agencies could withhold records from persons not properly and directly concerned, withhold matter in the public interest, or hold matter confidential for good cause found, without judicial enforcement for requesters. The 1966 act inverted that design. Agency records became available to any person on request, without a statement of reason or need. Vague withholding standards gave way to nine discrete enumerated exemptions. The statute placed the burden of justifying withholding on the agency and gave a denied requester a right to challenge the denial in federal district court, with the court deciding the matter de novo. Affirmative publication duties accompanied the request right from the start.

Q: How many exemptions does the Freedom of Information Act have?

The act contains nine enumerated exemptions, and those categories are exclusive, meaning an agency may withhold a record only when the record falls within one of them. As enacted in 1966, the categories covered classified national defense and foreign policy matters, internal personnel rules and practices, matters exempted by another statute, trade secrets and confidential commercial or financial information, inter-agency or intra-agency memoranda protected in litigation, personnel and medical files whose release would clearly invade personal privacy, investigatory files compiled for law enforcement purposes, reports concerning regulated financial institutions, and geological and geophysical information concerning wells. Congress later narrowed or adjusted several categories, particularly Exemptions 1, 3, and 7, without adding a tenth category or abandoning the enumerated structure.

Q: Does the Freedom of Information Act cover Congress?

No. The act expressly excludes Congress from the definition of agency. The definition, incorporated from 5 U.S.C. 551(1), covers each authority of the Government of the United States, including executive departments, military departments, government corporations, government controlled corporations, independent regulatory agencies, and establishments in the executive branch. Alongside Congress, the statutory exclusions reach the courts of the United States, the governments of territories and possessions, and the government of the District of Columbia. The President’s immediate personal staff and units of the Executive Office of the President whose sole function is to advise and assist the President also sit outside the definition, according to the legislative history as later applied by the courts. Presidential records fall under the separate Presidential Records Act of 1978.

Q: Who has the burden under the Freedom of Information Act?

The agency bears the burden. Under the Administrative Procedure Act regime that preceded the 1966 act, the practical burden sat on the requester to show need and entitlement, and agencies could withhold material from persons not properly and directly concerned. Representative John E. Moss of California identified the reversal as one of the central reforms. The statute placed the burden on the agency and gave the requester a federal court remedy, with the court deciding the matter de novo rather than deferring to the agency’s own characterization of a record. In administration, the same allocation follows from the presumption of disclosure. A requester needs no reason or need for a record. An agency that withholds must justify the withholding under an enumerated exemption.

Q: Which Freedom of Information Act exemption is used most?

No single exemption leads in every setting. Government-wide citation rankings vary by agency and fiscal year, and the privacy exemptions, Exemptions 6 and 7(C), frequently lead government-wide totals. Exemption 5 stands among the most frequently invoked exemptions and ranks as the most heavily litigated and most heavily criticized. The Department of Justice Guide to the Freedom of Information Act identifies the deliberative process privilege as the most commonly invoked privilege inside Exemption 5. Exemption 5 is distinctive because it protects the reasoning process itself rather than a category of harmful content. Criticism has concentrated on that breadth and on the absence, as enacted, of any requirement that disclosure cause demonstrable harm to an identified interest.

Q: Did the president want to sign the Freedom of Information Act?

The Congressional Research Service reports that President Lyndon B. Johnson was reportedly reluctant to sign the measure. Johnson signed on July 4, 1966, at a Texas ranch, on the last day available before a pocket veto would have taken effect. National Security Archive research from Lyndon B. Johnson Library documents records no signing ceremony, no commemorative pens, and no invitation to the sponsors, and the signing does not appear in the presidential daily diary. Johnson personally weakened stronger openness language in the draft press statement. The issued signing statement gave substantial emphasis to the exemptions and to protection for confidential communications, closing with a statement of pride that the United States is an open society. Press secretary Bill Moyers later described the president as having to be persuaded at the last minute.

Q: How long did the Freedom of Information Act take to pass?

The Congressional Research Service counts eleven years between the opening hearing in 1955 and enactment in 1966. Representative John E. Moss of California chaired the Special Subcommittee on Government Information of the House Government Operations Committee from 1955, and hearings across that period documented agency withholding, including Civil Service Commission refusals involving dismissed employee records in the Eisenhower era loyalty program context. A coalition of newspaper editors, journalists, press associations, educators, and scientists supported the legislation, while executive agencies of both parties opposed it, and the Congressional Research Service record notes that no executive department or agency head supported the bill. The Senate passed S. 1160 on October 13, 1965. The House passed the measure on June 20, 1966, without a dissenting vote.

Q: Does the Freedom of Information Act apply to the courts?

No. The courts of the United States are expressly excluded from the definition of agency that the act applies. That exclusion places the federal judiciary outside the request system entirely, alongside the other named exclusions for Congress, the governments of territories and possessions, and the government of the District of Columbia. Coverage instead follows the executive side of the federal government, reaching executive departments, military departments, government corporations, government controlled corporations, independent regulatory agencies, and other establishments in the executive branch. The exclusion concerns institutional coverage rather than judicial involvement. Federal district courts play a central enforcement role because a denied requester may sue, a district court decides the case de novo, and the agency carries the burden of sustaining a withholding in that litigation.

Q: What is the deliberative process privilege under the Freedom of Information Act?

The deliberative process privilege operates inside Exemption 5, which covers inter-agency or intra-agency memoranda or letters that would not be available by law to a party other than an agency in litigation with the agency. The Department of Justice Guide to the Freedom of Information Act identifies deliberative process as the most commonly invoked privilege within that exemption, alongside attorney work product and attorney-client privileges. Courts require material to satisfy two conditions. Material must be predecisional, meaning generated before adoption of an agency policy, and deliberative, meaning reflective of the give-and-take of consultation. Factual material and adopted rationales fall outside the privilege. Courts recite three purposes for the protection: candor in internal advice, avoidance of premature disclosure of proposals, and avoidance of public confusion from rationales never adopted. The 2016 act added a limit under which the privilege does not apply to records twenty-five years or older.

Q: Must an agency release part of a record if the rest is exempt under the Freedom of Information Act?

Yes, when a portion is reasonably segregable from exempt material. Courts articulated the reasonably segregable portions doctrine in EPA v. Mink in 1973, and Congress codified segregability in the 1974 amendments. The doctrine prevents an exempt passage from shielding an entire record automatically. An agency must separate material that falls within an exemption from material that does not and release the nonexempt portion, unless the nonexempt material is so intertwined with exempt material that separation leaves nothing meaningful. Segregability fits the wider structure of the statute. Exemptions are exclusive categories, withholding requires justification by the agency, and litigation proceeds de novo. A claim therefore has to attach to identifiable material rather than to a document as an undifferentiated whole, and partial release is the ordinary consequence when only part of a record qualifies.

Q: What is a Vaughn index in Freedom of Information Act litigation?

A Vaughn index is a court-devised device from Vaughn v. Rosen, decided in 1973. In litigation over withheld records, the index requires an agency to itemize the material withheld and justify each withholding. Each entry identifies a document or portion and connects that material to the exemption claimed for it, with enough explanation for a court and a requester to test the claim without disclosure of the protected content itself. The device addresses a structural problem in exemption litigation. A requester cannot argue effectively against a justification stated only in general terms, and a court deciding de novo needs a record specific enough to evaluate. Itemized justification therefore disciplines agency claims exemption by exemption and portion by portion. The index complements segregability, because both mechanisms force withholding decisions down to the level of particular material rather than whole files described in broad categories.

Q: Does the Freedom of Information Act require an agency to create a record that does not exist?

No. The act reaches only records in an agency’s possession and control that the agency created or obtained, according to the settled reading of the courts. It imposes no duty to create records and no duty to answer questions posed in the form of a request. A requester therefore seeks existing agency records rather than commissioned research, explanations, or compilations assembled for the occasion. Forsham v. Harris, decided in 1980, illustrates the boundary on the possession side. The Supreme Court held that data generated by private grantees was not an agency record. Together, the creation and possession limits define the object of the statute. Disclosure obligations attach to records an agency holds, and material outside agency creation, obtainment, possession, or control falls outside the request right, whatever interest the material might hold for a requester.

Q: What kinds of records must agencies publish without any request under the Freedom of Information Act?

Affirmative disclosure duties date to the 1966 act itself, before any individual request enters the picture. One set of duties requires publication in the Federal Register. Another set requires reading room availability for opinions, orders, policy statements, and staff manuals, under the functions associated with subsections (a)(1) and (a)(2). Those categories share a common character. Opinions and orders show how an agency has decided matters. Policy statements show the positions an agency has adopted for future conduct. Staff manuals show instructions that guide personnel in dealings with the public. Publication and reading room access prevent such material from operating as unpublished law known only inside an agency. The 1996 Electronic Freedom of Information Act later extended the reading room concept to electronic records and electronic reading rooms.

Q: How did the 1974 amendments change the Freedom of Information Act?

The 1974 amendments, Public Law 93-502, were enacted over President Ford’s veto in November 1974 and formed the first of the six major substantive revision acts. The amendments added deadlines for agency responses and addressed fees. They authorized judicial in camera review and codified the reasonably segregable portions doctrine that courts had articulated in EPA v. Mink in 1973. They also narrowed Exemption 1 by adding a properly classified qualifier to the national defense and foreign policy category. For Exemption 7, the amendments changed the enacted phrase investigatory files to records or information compiled for law enforcement purposes and supplied the six harm test covering interference with proceedings, fair trial rights, personal privacy, confidential sources, techniques and procedures, and physical safety. Attorney fees provisions formed part of the same package, strengthening the position of requesters who litigate.

Q: What is the foreseeable harm standard in the Freedom of Information Act?

The foreseeable harm standard requires an agency to connect a withholding to harm that disclosure would foreseeably cause to an interest an exemption protects, rather than relying on the bare fit between a record and an exemption category. The FOIA Improvement Act of 2016, Public Law 114-185, codified that standard for withholding generally and also codified a presumption of openness. The same act placed a limit of twenty-five years on the deliberative process privilege inside Exemption 5. The standard belongs to the long amendment pattern that the statute’s own history displays. Major revisions worked mainly by hardening the justification an agency must supply through standards, deadlines, segregability, harm showings, and judicial tools, while the request right itself, held by any person without a stated reason, stood essentially unchanged from 1966 onward.

Q: Why does the Freedom of Information Act protect geological information about wells?

Exemption 9, as enacted in 1966, covers geological and geophysical information and data, including maps, concerning wells. Congress placed that subject in a discrete enumerated category rather than leaving well data to the general standards that had governed withholding under the Administrative Procedure Act. Placement matters under the act’s structure. Withholding requires a category, and Exemption 9 supplies the category for well related geological material in agency hands. The exemption sits apart from Exemption 4, which addresses trade secrets and commercial or financial information obtained from a person and privileged or confidential. Well information therefore receives protection through a subject specific provision written for geological and geophysical material, including the maps that record subsurface findings. Beyond that categorical design, the statute treats the protection as one of nine exclusive grounds, each construed within the presumption of disclosure and the agency’s burden of justification.

Q: Can any person, including a foreign national, request records under the Freedom of Information Act?

Yes, at the level of the basic right. Representative John E. Moss of California described access for any person as one of the three central changes of the 1966 act, replacing the earlier limitation to persons properly and directly concerned. A requester needs no statement of reason or need, and the right has stood essentially unchanged since 1966. Nationality does not remove an individual from the phrase any person. A qualification arrived in 2002, when Congress barred elements of the intelligence community from releasing records in response to requests by foreign government entities or their representatives, a measure the Department of Justice counts among the seven congressional amendments to the act. The restriction targets requests by or on behalf of foreign governments addressed to intelligence agencies rather than the individual right itself. Private requesters, including foreign nationals acting as individuals, remain within the any person design that Moss identified as the core of the reform.

Q: Did the original Freedom of Information Act set a deadline for agencies to answer requests?

No. The 1966 act set no deadline for agencies to answer requests. Time limits arrived with the 1974 amendments, which formed part of the package enacted over President Ford’s veto in November 1974, alongside fees, in camera review, segregability, narrowed Exemptions 1 and 7, and attorney fees. The twenty-working-day period arrived later, in the Electronic Freedom of Information Act of 1996, Public Law 104-231, the revision that also addressed electronic records and electronic reading rooms. Deadline discipline developed further in the OPEN Government Act of 2007, which added request tracking and fee consequences for missed deadlines. The sequence illustrates the broader amendment pattern. Congress left the request right from 1966 essentially unchanged and instead hardened agency obligations through deadlines, standards, and consequences attached to delay.

Q: How are fees structured under the Freedom of Information Act?

The fee structure took its developed form in the Freedom of Information Reform Act of 1986, enacted within Public Law 99-570, after the 1974 amendments had addressed fees for the first time. The 1986 structure differentiates among categories of requesters. Commercial use requesters bear fuller costs associated with processing a request, reflecting the private benefit drawn from the records sought. Other categories receive more favorable treatment within the structure. Fee waivers tie relief from charges to the public interest, so a request whose disclosure serves public understanding rather than private advantage can qualify for reduced or waived fees. No uniform charge applies across all requesters. Charging therefore turns on who requests and for what purpose, and the categories and waiver standard allocate the cost of the disclosure system between commercial users and requesters whose use carries a public character.

Q: What should a student remember about the Freedom of Information Act for an exam?

The examination core begins with identity and inversion. The Freedom of Information Act, Public Law 89-487, signed by President Lyndon B. Johnson on July 4, 1966, effective one year later, and codified at 5 U.S.C. 552, replaced section 3 of the Administrative Procedure Act of 1946. The act made records available to any person without a stated reason, replaced vague withholding standards with nine exclusive exemptions, placed the burden of justification on the agency, and supplied de novo review in federal district court. Coverage reaches executive branch agencies and excludes Congress and the courts. Affirmative publication duties accompany the request right. Amendments in 1974, 1976, 1986, 1996, 2007, and 2016 hardened justification through deadlines, segregability, harm showings, and judicial tools while leaving the request right essentially intact. Exemption 5, heavily litigated and criticized, protects deliberative process material rather than harmful content as such.