Freedom of Information Article

	Following is an article I wrote on the Freedom of Information Act that was published in the Summer 1999 edition of Communication Law and Policy. The footnotes are pasted at the bottom of the text, but unfortunately the numbers did not come out either in the text nor in the notes when I pasted the article.  But, hey, whose gonna visit this page anyway?

     In theory, the Freedom of Information Act (FOIA) represents an impressive level of governmental accountability to its citizenry.   Wielded as a sword, the FOIA exacts responsibility from those exercising power of any kind on many of the workings of government.  It does so by subjecting most governmental units to disclosure requirements of all records and documentation.  Such scrutiny is the product of a presumption of openness where the government theoretically exists under the sunshine to do the people’s business in the light of day.

     While there may be a great many success stories where the FOIA has served this intention remarkably well, in reality the FOIA does not always realize its purposes to the fullest extent.    The FOIA has its detractors; included among these critics are those who have the statutory obligations imposed upon them.  Agencies and agency personnel often view the FOIA as a thorn in their bureaucratic sides, taking away money and personnel from the agency whose singular focus is gazed upon satisfying the duties imposed by its enabling statute.   To this end, providing information to the public may often be a distraction, at best.  Ultimately, freedom of information can sometimes be viewed as a battle between members of the public seeking access to governmental information and the agencies which possess such material.  To a degree, this tension has acted to whittle away some of the intended effects of the FOIA and has undercut some achievement of its goals.

     As a result and despite the well meaning intentions of the FOIA, it is frequently not given its full force.  Agencies are notorious for their slowness in processing even the most minor and simple of requests.   They have been known to claim that documents are lost or that they lack the staff necessary to process requests.   Reporters have complained that agencies will often stall until the information sought is no longer newsworthy.   Many agencies, if not most, have information requests backlogged in their systems for years.  The Federal Bureau of Investigation is perhaps best known for this; a court in 1994 found over 11,000 requests in the possession of that agency that were unfilled at that time.   Importantly, agencies often read exemptions very broadly in order to deny disclosure requests,  contrary to FOIA’s intent which is to use these exceptions narrowly and presume that disclosure is proper.    Until recently, many agencies claimed that the information held in their electronic databases did not qualify as records and did not have to be disclosed in electronic form—a position that was repudiated by the passage of the Electronic Freedom of Information Act Amendments of 1996.

     Another tactic is for the entity to claim it is not an agency at all and thus out of the purview of the FOIA altogether, which only applies to the records of agencies of the federal government.    This situation is a relatively rare one, as this position is entirely untenable for the majority of entities that could never successfully defend such an assertion.  However, when the claim of non-agency status is successful, the result is far broader in effect than most or all of the other techniques that agencies may routinely use to avoid the burdens of FOIA.  If an entity can successfully declare itself a non-agency for FOIA purposes, it will be categorically exempt from the Act and will be under no duty to turn over any material in its possession at any time.  As a result, the entity will be free from the bounds of accountability and the public will be left without any recourse to ensure that the governmental entity acts in a reasonable or fair manner.

     One such entity that has used such a claim in recent years is the National Security Council (NSC). The NSC is housed in the Executive Office of the President (EOP), a location of the governmental hierarchy that generally interacts directly with the President more than the average agency.   Despite the fact that it has always treated itself as an agency subject to the disclosure requirements laid out by the FOIA, the NSC reconsidered its position in 1994 and declared that it is not really an agency subject to FOIA after all, but rather it is merely an entity that exists solely to advise and assist the President.   This has a profound effect on that entity and its requirement to abide by the Freedom of Information Act, as the FOIA neither entangles the President,   nor the President’s immediate personal staff that exist solely to advise and assist him.   Simply put, those that exist solely to advise and assist are not agencies for FOIA purposes.  As such, the NSC concluded that it has blanket immunity from the application of the FOIA and is empowered with the ability to conduct its business out of the public light.

     Over the last three decades, the courts have been asked to resolve disputes over whether a particular establishment is an agency under the FOIA.  In analyzing whether an organizational unit that interacts closely with the President, such as the NSC, should be classified as an agency, the courts use an inquiry commonly referred to as the sole function test.   Under this analysis, such an entity is not an agency for FOIA purposes if it exists solely for the objective of advising and assisting the President.   The issue is sometimes phrased in the converse as whether the entity exercises substantial independent authority.   If an entity exerts such authority, it cannot exist merely to advise and assist and should be classified as a FOIA agency.   The sole function test attempts to draw a line between entities that exist to serve strictly as presidential advisors and those with any sort of responsibility beyond aiding the chief executive officer.  Those that have this additional latter function are agencies subject to FOIA.

     Thus, the sole function test can be understood to be a limitation on the applicability of FOIA to certain entities located at positions closely intertwined with presidential decisionmaking.   When this limitation is levied upon an entity within the Executive Office of the President, which is by and large an area of the executive hierarchy generally under the obligations of FOIA, it is out of respect for separation of powers concerns.    The legislature is on thin ice when it attempts to regulate the conduct of the President’s daily operations and the inappropriateness of one co-equal branch of government forcing disclosure by another, absent certain conditions,  is embedded deep within constitutional jurisprudence.   Further, the principle that "not all public business can be transacted completely in the open, that public officials are entitled to the private advice of their subordinates and to confer among themselves freely and frankly, without fear of disclosure," is one firmly implanted in the common law.   For these reasons, FOIA does not apply to the President.    Similarly, when an entity is nothing more than a mere advisor to the President, its activities are likewise shielded.

     This paper intends to deal with this sole function test.  Using the situation which unfolded regarding the NSC during recent litigation as its primary example, this paper contends that the focus of this analysis has been obscured in recent court decisions at its foundation.  This change in focus was especially decisive when the Court of Appeals for the District of Columbia examined the issue of whether the NSC should be declared an agency.   Of course, it can not be disputed that the courts have the power and the duty to reassess their own interpretations from time to time.  But what this paper notes is that the retooling of the sole function test represents a departure that is irreconcilable with its earlier formulation.  More importantly, it is a departure that considerably broadens the aperture for which future entities that wish to avoid the rigors of FOIA compliance can slip through.

     Part II will introduce the structure, form and purposes of the National Security Council and set up some of the considerations that will become relevant when looking to whether it is an agency under the Freedom of Information Act.  Part III will review the two different statutory schemes which dictate recordkeeping and disposal requirements:  the Federal Records Act, which applies to executive agencies,   and the Presidential Records Act, which concerns the President and his immediate staff.    The two schemes are mutually exclusive, so no record falls under the demands of both statutory regiments.   The status of the NSC is thus important to the resolution of which statute to impose.  Part IV discusses the case law of agency determination for entities both within and outside the Executive Office of the President.  Part V then deals with the case of Armstrong v. Executive Office of the President,   which dealt precisely with the issue of whether the NSC should properly be considered an agency for FOIA purposes, and subject to the dictates of the Federal or Presidential Records Act. Finally, Part VI explores the application of precedent to the Armstrong case and suggests that the focus of the agency analysis has shifted in a way that contradicts both precedent and the core intent of the sole function analysis in fundamental ways.

     This issue brought about when an entity claims it is exempt from FOIA as a non-agency is not a constantly recurring theme in the area of access law.  Indeed, it is a claim that has been encountered somewhat infrequently in FOIA litigation up to this point.    However, the issue of what constitutes an agency as opposed to a presidential advisor is an extremely important issue, despite its irregularity in appearing on court dockets.  While most FOIA litigation tends to revolve around specific requests, such as whether material is a "record" or whether an exemption applies to that particular record, the situation under study here is of a far grander scale.  A successful defense of a non-agency claim acts to remove the entity from disclosure obligations in toto; it does not simply apply to an individual request or even a series of requests, it applies to all information held by the entity in question.  This represents a harsh reality for any segment of the public who wishes to gain access to any of the information held by that entity.  A non-agency finding is the equivalent in shutting off accountability wholly as it pertains to the individual entity; it does nothing less than make the entity permanently invisible and nonexistent for all purposes of the FOIA.

     Further, the individual entities that are most likely to fall through the cracks of FOIA dictates are entities located in the EOP, because they are aligned the most closely with the President and therefore are more likely to be ruled as advisors as opposed to autonomous agencies.   And it is in the EOP where some of the most important domestic and international issues are decided.  Thus, this issue is one which deals with the potential that entities where accountability is of perhaps the highest concern are removed from the FOIA completely.  Such intense ramifications make this a topic deserving of a detailed study.  Due to the change of focus in the agency analysis that has clouded the traditional inquiry of the sole function test, it may be far easier for other entities in the Executive Office, and perhaps elsewhere, to escape public scrutiny through similar claims.  This is an important concern, for what is given to the realm of secrecy is taken from the field of public discourse and deprives the citizenry of its right and duty to police its sovereign.

     At the top of the elaborate executive bureaucracy sits the Executive Office of the President,   housing several entities that work in close proximity to the President, including the White House Office.  One of these establishments encompassed by the EOP is the National Security Council.  The NSC was formed under the National Security Act of 1947   and transferred to the EOP in 1949.   The organization was formed to "advise the President with respect to the integration of domestic, foreign, and military policies relating to the national security," and to allow the military and other entities to cooperate better in handling issues of national security.    The National Security Act also created the Central Intelligence Agency, and installed the NSC as the superior authority of that entity.

     The NSC has a complex structure with a staff of approximately 150 people,  and is headed by an executive secretary appointed by the President without the need for confirmation from the Senate.   Three distinct segments comprise the NSC.   The Council consists of members who serve as a result of positions they hold outside of the NSC.    These statutory members are the President, Vice-President, the Secretary of State and the Secretary of Defense.     Also, there is a permanent staff that produces reports for the Council members and is responsible for much of the evaluatory and oversight functions of the NSC.   The third component of the NSC is the National Security Advisor, who serves as the principal executive officer for the National Security Council and reports directly to the President.

     Different presidential administrations have used the NSC in differing manners, and the organization often reflects the style of each administration.   Some presidents, such as Harry S. Truman and John F. Kennedy, used the NSC only sparingly, while President Dwight Eisenhower used the NSC with a much greater emphasis of oversight functions.    President Ronald Reagan used the NSC to perpetrate covert operations, thus avoiding the requirement of the Hughes-Ryan Amendment of informing Congress of such activities.   Reagan assumed that Hughes-Ryan only related to the CIA.    This course of action violated the Boland amendment which prohibited any funding of the Nicaraguan Contras, and eventually led to the Iran-Contra scandal.

     While the NSC has limited functions delegated to it organically through its statutory creation, several presidential directives and Executive Orders may have granted it authority in a number of separate functions sufficient to qualify it as an agency.  These potential functions are important considerations when looking to whether the NSC is an agency for FOIA purposes, or whether it exists solely to advise and assist the President, and is thus exempt from the FOIA’s grasp.    Protecting national security information is one such function.  The NSC is in charge of providing policy direction for the information security program governing the classification, declassification and safeguarding of sensitive information.    The NSC also reviews the security practices of other agencies.   In addition, the NSC is charged with developing and directing policies for the National Communications System, a group of organizations responsible for restoring communications in a national emergency, including nuclear attack.   All departments and agencies are also expected to comply with NSC guidelines in the area of emergency preparedness and crisis management.   Further, the National Security Council makes recommendations to the Department of Commerce relating to export license applications that potentially involve nuclear explosives.   Finally, the NSC has played a role in public diplomacy by coordinating and directing public affairs campaigns to generate support for national security goals.

     Different recordkeeping and disposal statutes would be imposed on the NSC and other entities depending on whether they are determined to be agencies or to serve strictly in the capacity of rendering advice and assistance to the President.   The Federal Records Act (FRA) applies to all agencies of the federal government and dictates obligations toward recordkeeping that work in conjunction with the disclosure requirements set out by FOIA.   The Presidential Records Act (PRA) applies to the President and his immediate staff and creates much more lenient requirements and offers far more discretion for the Executive.   The FRA applies to federal government agencies,  but the PRA explicitly excludes "official records of an agency" from its scope.   Therefore, the two statutes are mutually exclusive and no organization or materials are subject to both the FRA and the PRA.   The FRA would apply to documents created or received by an agency, including a White House or EOP agency, while the PRA would apply if the documents are created or received by the President or the presidential staff.    The determination of whether the NSC and other entities are agencies is thus significant because one of these different statutory constructions will apply depending on this classification.   As an agency, most or all NSC records would fall under the more demanding obligations required of the FRA.  Of course, an agency categorization will immediately invoke the disclosure requirements of the FOIA as well, because the FOIA, like the FRA, apples to all federal agencies.

     The Federal Records Act, codified across four chapters of title 44 of the United States Code, prescribes regulations for the creation, management and disposal of federal records.    Among other goals such as recordkeeping simplification,  the FRA serves the dual purposes of providing comprehensive and accurate recordkeeping and "judicious preservation and disposal" of governmental records.     The act also directs the heads of agencies to create and maintain records "containing adequate and proper documentation of the organization, functions, policies, decisions, procedures and essential transactions of the agency"  and to "establish safeguards against the removal or loss of records."    The FRA charges an Archivist  with the duty of advising and assisting   agencies and imparts to him or her the inspection authority to guarantee compliance.   Federal records may only be discarded with the consent of the Archivist and if they "do not appear to have sufficient administrative, legal, research, or other value to warrant their further preservation."

     The FRA does not explicitly recognize a cause of action for private parties; instead it is the role of the Archivist and agency head to seek redress through the Attorney General.   However, the Archivist’s compliance with certain portions of the act may fall under judicial review through the Administrative Procedure Act (APA),  which provides the courts with the authority to set aside any agency action that is "arbitrary and capricious, an abuse of discretion, or otherwise not in accordance with law."   The APA also provides private parties the right to seek review of an agency action that causes them "a legal wrong."    Thus, under the FRA, the actions of the head of an agency and the Archivist may be subject to judicial review in litigation brought by an individual citizen.

     Unlike the FRA, the PRA allows for Executive retention of far more discretion in recordkeeping and disposal.  Inspired by the Watergate scandal,  the PRA subjects all presidential records to "complete ownership, possession, and control" of the United States.    The act requires maintenance and preservation of records of incumbent and former presidents and prescribes procedures for disposal and mechanisms for guaranteeing compliance.

     Far more authority is granted to the President under the PRA than to the individual agencies covered under the Federal Records Act.   The Archivist has five years from the completion of an administration before FOIA is triggered and the documents are disclosed.    However, the President may restrict access to any document for up to twelve years after the close of his term under six enumerated exemptions,  and the Archivist is expected to uphold any restriction unless the President waives it.   While these categories allowing for withholding may be limited, a sitting President has the power to dispose of records virtually at will and, therefore, can destroy documents he wishes to have avoid disclosure.   While there are some procedural requirements that a President must satisfy before disposing of records, the Archivist, Congress and the courts are ultimately powerless to prevent improper destruction during a President’s term.   Thus, the President himself has near total discretion in releasing documents for disclosure under FOIA for at least twelve years following his term, and is capable of cheating impartation on a more permanent basis.

     The Presidential Records Act authorizes only the President to oversee compliance.    The Archivist has only an advisory function during a President’s term, although he has some minor authority after the term expires.   Like the Federal Records Act, the PRA is silent on the issue of judicial review.  The courts have determined that separation of powers issues are sufficiently sensitive to preclude review and the judiciary’s ability to overrule a decision made on records classified under the PRA is severely limited.   Unlike is the case with entities covered under the FRA, the Administrative Procedure Act will not permit judicial review, as the President is not an agency and therefore not touched by the APA.

     There are some major differences between the Federal Records Act and the Presidential Records Act.   It is these major distinctions that bestow importance to the debate concerning the NSC, as well as other potential future entities that may claim non-agency designation.   One contrast is the fact that the PRA lacks an enforcement scheme.   The FRA procedures are far more demanding in preventing wrongful destruction of documents than that of the Presidential Records Act.   Next, the PRA only affects presidential documents once the administration ends, while the requirements of the Federal Records Act and the Freedom of Information Act can force disclosure during the term of office.    Also, records may not be taken by a President after the completion of a term without the consent of the Archivist if the documents are classified under the FRA, while many past presidents have kept their records under the PRA.   Finally, the FRA requires the Archivist to report to Congress annually on recordkeeping policies and practices.   No such requirement exists under the PRA.

     Because of all these differences, the PRA could be viewed as a less demanding statute.  Record maintenance is more lenient.  Even more important, the President has near absolute power to dispose of records under the PRA,   thus circumventing eventual disclosure requirements, and the PRA does not allow judicial review either by its own terms or through the application of the APA.    It is whether an entity is classified as an agency which ultimately resolves which statutory regiment is imposed.  The next concern, then, are the factors and focus on which an agency classification turns.

     An understanding of whether the NSC should be classified as an agency requires a clear delineation of where exactly the boundaries of agency status lay.  Early precedent sets a fairly steady flow of reasoning in dealing with the issue of agency determination; a flow which may have branched in other directions and taken on a different and more complex form in recent cases.  This body of precedent served as the foundation for the determination of agency status of the National Security Council, and it was the application of this case law through a different focus that may have ultimately decided that case in a way unfavorable to access interests.

     A long line of case law clarifies the statutory definitions of what constitutes an agency and when an entity that serves as an advisor to the President, as is the case with the NSC, may still fall under the prescriptions of the FRA and FOIA.  In making this decision, the courts ask whether an entity exercises substantial authority independent of the President or whether it exists solely to advise and assist him.   While the President’s immediate staff is irrefutably exempt in its confined functions, when those functions expand beyond the limited role of advice and assistance, an agency determination is much more likely to be recognized.

     The statutory definitions are often of limited use in clarifying whether a specific entity is a FOIA agency.  The Administrative Procedure Act defines an "agency" as "each authority of the Government of the United States, whether or not it is within or subject to review by another agency,"  and enumerates a list of exceptions, including Congress,  the courts,  and other groups not germane to the NSC issue.  The FOIA originally incorporated the APA’s interpretation, but the definition was broadened in 1974 to expand the FOIA coverage to entities that the APA failed to reach.   The definition was also expanded to eliminate some of the confusion in the courts  involving issues of classifying   "the myriad organizational arrangements for getting the business of the government done."   The FOIA, as amended, expands the APA definition of "agency" to also include "any executive department, military department, Government corporation, Government controlled corporation, or other establishment in the executive branch of the Government (including the Executive Office of the President), or any independent regulatory agency."   However, these definitions fail to clarify some of the gray areas that exist between the typical executive branch agency, which is undeniably covered by the FOIA, and the President’s "immediate personal staff," which clearly is not.  Further complicating matters is the fact that the FOIA was the product of a sometimes contradictory legislative history in regard to the agency issue.   Case law casts some illumination on this shady situation.

     Over the course of the last three decades, the courts have reviewed determinations of agency status involving a handful of entities from both within and outside of the Executive Office of the President.  The conception of the modern agency classification inquiry originated in 1971 when the Office of Science and Technology (OST) was ruled an agency for FOIA purposes in Soucie v. David.   In Soucie, two citizens sought from the OST a document known as the Garwin Report, which evaluated a government program for the development of supersonic aircraft.   The OST, which was formed to serve the dual functions of evaluating federal scientific research programs, and advising and assisting the President, refused to release the document.   The trial court determined that the Office of Science and Technology was not an agency, but rather a component of the Office of the President and thus exempt from the requirements of   FOIA.

      The Court of Appeals for the District of Columbia Circuit determined on review that although the statutory definition of agency was unclear, the APA "apparently confers agency status on any administrative unit with substantial independent authority in the exercise of specific functions."   The court recognized that "[i]f the OST’s sole function were to advise and assist the President, that might be taken as an indication that the OST is part of the President’s staff and not a separate agency."   This inquiry is commonly referred to as the sole function test.   The court’s examinations revolved around the functional role of the OST,   and this focus on function has become the critical component of agency analysis under the FOIA since its inception.

      The court concluded that the Office of Science and Technology was an agency for FOIA purposes.   Because the OST had an "independent function of evaluating federal programs" and could take direct action in initiating and supporting research and awarding scholarships, its role went beyond merely advising and assisting the President.   Beyond failing the sole function test, the court noted that Congress appeared to indicate an understanding that the OST would possess this evaluatory function when acquiescing to the reorganization plan that created the entity—a fact that served as a "consideration of some weight."    Further, up to the point of the events leading to litigation, the OST considered and conducted itself as an agency in setting procedures for compliance with the Administrative Procedure Act and the FOIA.

        In 1980, the United States Supreme Court lent credence to the sole function test in Kissinger v. Reporters Committee for Freedom of the Press, in approving the proposition that "‘the President’s immediate personal staff or units in the Executive Office whose sole function is to advise and assist the President’ are not included within the term ‘agency’ under the FOIA."   In addition, the Conference Report on the 1974 FOIA amendments explicitly adopted Soucie in writing the new agency definition for the Freedom of Information Act.

     A number of cases have built on the Soucie sole function test and have fleshed out the scope of the decision, while maintaining the underpinnings of its approach.   In 1973, the United States Court of Appeals for the District of Columbia Circuit resolved a dispute over the status of another entity claiming non-agency classification in Grumman Aircraft Engineering Corporation v. Renegotiation Board, where the court fully endorsed Soucie.    The court determined that the regional branch of the Renegotiation Board was an agency and was required to release the documents sought in a FOIA request.    These documents explained decisions relating to whether fourteen companies accrued excess profits in their dealings with the government.   These Regional Renegotiation Boards were components of the National Board, which negotiates federal contracts.

     After resolving that the determinations of the Regional Boards represented a final opinion, the court found that the Regional Boards were granted "substantial independent authority."   The Regional Boards had personnel for investigating and negotiating, and these boards were expected to make formal recommendations.   As was the case of the OST in Soucie, the court believed that Congress expected the Regional Boards to behave as an agency in these regards, and the National Board had promulgated regulations for releasing some of the Regional Boards’ documentation.    Ultimately, these boards served as "the agency with which an affected member of the public dealt, and from whose decision appeal might lie."

     A contrary result was reached a year later in Washington Research Project v. Department of Health, Education and Welfare.   Although the D.C. Court of Appeals ultimately found against agency status, this decision still stressed that the focal point in the agency analysis was the authority to make decisions.   The Washington Research Project sought the disclosure of reports dealing with research funded by the National Institute of Mental Health.   These reports were put together by Initial Review Groups (IRGs) which were panels of consultants with specialized knowledge.  The reports were used to make recommendations to the National Advisory Mental Health Council.   The court ultimately held that the IRGs were not agencies but rather advisory committees within an agency, and thus the reports were exempt from disclosure as intra-agency memoranda.   Because the IRGs confined themselves to making recommendations and the authority to make grants was vested elsewhere,  the Initial Review Groups only had advisory functions, even though their recommendations were "often a crucial element in the approval process."   The real focus of the analysis, according to the court, was whether the Initial Review Groups had authority to make decisions, not whether the advice was generally accepted.   Because no such authority existed with respect to the IRGs, the court ruled that they were not agencies under the APA and subsequently unaffected by the requirements of the FOIA.

     Neither Grumman nor Washington Research reviewed a component structure of the Executive Office of the President like Soucie did with the Office of Science and Technology.  However, in 1978, the courts once again dealt with the agency status of an EOP branch when the Sierra Club demanded that the Office of Management and Budget (OMB) produce an environmental impact statement to be included with its annual budget request.      Since the last time the courts scrutinized an EOP entity for agency assignment in Soucie, the 1974 FOIA amendment was enacted that expanded the APA definition of agency to include any "… other establishment in the executive branch of the government (including the Executive Office of the President) … ."    In Sierra Club v. Andrus, the court pointed to the legislative history of the amendment as proof that Congress had intended to adopt Soucie.   The court noted that "the President’s immediate personal staff whose sole function is to advise and assist the President" was still exempted.    Because the preparatory stages of the budget also serve as "an aid to Congress as well as an instrument of presidential and policymaking control over the executive bureaucracy," and the OMB had other "management, coordination, and administrative functions," it did not satisfy the sole function test and was not exempt completely from FOIA dictates.

     In the 1980 case Ryan v. Department of Justice, the Court of Appeals for the D.C. Circuit clarified that the sole function test creates an all-or-nothing approach.   Although there may be situations where an entity both assists the President and has independent functions, a hybrid does not emerge.  The plaintiffs in Ryan were interested in monitoring federal judicial appointments for racial and gender inclusion.     They sought disclosure of Senator questionnaire responses dealing with the Senate procedures for choosing potential nominees for district court judges.    The court concluded that because the documents were solicited by and controlled exclusively under the supervision of the Attorney General, he was "acting as an independently controlling entity," even though he would use the information to make recommendations solely for the President.   Neither the legislative history of the 1974 FOIA amendment nor Soucie itself implied that an agency could still have distinct segments treated as a non-agency under the FOIA in some circumstances if that segment were able to satisfy the sole function test independently.   Taken as a whole, a specific entity is either an agency or it is not, and "[a]ny unit or official that is part of an agency and has non-advisory functions cannot be considered a non-agency in selected contexts on a case-by-case basis."

     Shortly after Ryan, the D.C. Court of Appeals similarly denied that an agency action can be insulated from FOIA requirements when that agency acts in the capacity of a presidential advisor.   In Pacific Legal Foundation v. Council on Environmental Quality, the court found unpersuasive the claims of the Council on Environmental Quality (CEQ) that it was not an agency when serving in the position of advising the President.   That case dealt with the open meetings requirements of the Sunshine Act, which incorporated the agency definition of the Freedom of Information Act.   The court forced the CEQ, an EOP entity, to open its meetings when the requirements of that act so dictated.    The court noted the similarities between the Council on Environmental Quality and the Office of Science and Technology.    As in Soucie, the court looked to CEQ’s authority to evaluate federal programs in identifying it as an agency.

     The D.C. Court of Appeals found against agency status in 1985 in Rushforth v. Council of Economic Advisors, because substantial authority did not exist in regard to the Council of Economic Advisors (CEA).   Like Pacific Legal Foundation, this was another case involving a unit of the Executive Office of the President.  Interestingly, the statutes that organized both the CEQ and the Council of Economic Advisors were essentially identical, as the Council on Environmental Quality statute was modeled directly after the one creating the CEA.   However, the court found that their functions differed substantially.   Unlike the Council on Environmental Quality, where several Executive Orders expanded the agency’s authority, no such situation existed in regards to the CEA.   Also, the CEQ coordinated federal environmental programs, issued guidelines, and promulgated regulations for instituting the National Environmental Policy Act.   The Council of Economic Advisors’ authority was far more limited, as it had never been granted any of this additional authority.   Finally, the CEQ admitted it was an agency for most purposes—it claimed it was not an agency only in its capacity of advising the President.   The CEQ differed in this regard.   On these facts, the court concluded that the CEA did not have any independent authority similar to the CEQ, nor did it have the ability to take "direct action" like the Office of Science and Technology in Soucie.   As was the situation in the earlier cases, this determination was based on the functional role of the entity beyond presidential advising.   Because the CEA did not act in this role, it could not be considered an agency.

    In another case involving the Executive Office of the President, the District of Columbia Court of Appeals declined recognizing agency status for the Task Force on Regulatory Relief set up by President Reagan that was headed by Vice-President George Bush.    The Task Force was responsible for revising current regulations and reviewing pending regulations with the purpose of recommending legislative remedies.    In the 1993 case of Meyer v. Bush, the focus of the agency examination may have departed from the consistent flow of Soucie and its progeny.   Although precedent had essentially looked only to the actual functions of the entity at issue in determining whether it classified as an agency,  the Meyer court took three factors into consideration in "refin[ing] the basic Soucie definition."   The first factor was how close the organization in question was to the President operationally.   The second factor was the nature of the delegation of power.   Finally, the court looked to whether the entity had a self-contained structure.   The Meyer court read the exemption of the Conference Report for the 1974 FOIA amendments that   "the President’s immediate personal staff … or units in the Executive Office whose sole function is to advise and assist the President" were excepted to mean that entities that closely resemble the characteristics of the personal staff of the President were more likely to avoid agency status.   The court determined that the Task Force, positioned "only a hair’s breadth" from the President, was not expected to make any decisions that were not "directly and precisely the President’s opinion."   The absence of a separate staff also served as a consideration and was as a "strong indicator" that independent authority was lacking.

     Arguably, the Meyer court took a turn of position on the agency analysis by looking not only to the functions and authority of the task force, but other considerations also, such as the structure and presidential proximity of the entity.  The court stated that "[f]unction is crucial, but … we believe that form follows function."     Under these considerations, the court found the President’s Task Force on Regulatory Relief to be only the second entity within the EOP, other than the Council of Economic Advisors, to satisfy the test "as an entity ‘whose sole function is to advise and assist’ the President."

     Judge Patricia Wald objected in a lengthy dissent.  She believed that a "careful examination of both [the task force’s] authorized and actual functions" was required, and upon such examination, it was clear that "significant independent authority to act on a continuing basis" without Presidential involvement existed.  Citing to a number of independent functions that the task force possessed, Judge Wald asserted that the task force satisfied the Soucie test.   While she believed that the scope and nature of the delegation of power was "the primary relevant inquiry," she objected to the remaining two factors the majority enumerated as being "entirely creatures of the majority’s making."    According to Wald, the delegation of authority had been satisfied by the Executive Order that created the task force and gave it authority to bind agencies to policy guidelines.

     Judge Wald disagreed with the majority’s new requirement that an agency must have a separate staff as being out of line with precedent which looks to "how an entity functions, not where it gets the resources to perform those functions."    Also, Wald said the court confused the "immediate personal staff" prong with the "sole function" exemption in looking to the "operational proximity" between the task force and the President.   How close or continuing the interaction is between the President and the entity does not relate to the sole function test and is not the proper inquiry.   According to Judge Wald, the majority’s three-part test was "clearly inconsistent with the functional inquiry this court has applied in FOIA cases for twenty years."

     By incorporating the three-part test into the sole function inquiry, Meyer may represent a shift in the focus of agency analysis.  Instead of pure function as the unit of investigation, form and proximity become more relevant considerations under Meyer’s formulation.  The reconciliation of these additional elements to the facts of a case may lead a court to arrive at a very different conclusion than it may have under the less diluted focus of Soucie.  The litigation involving the NSC serves as an excellent example of where this different application of the sole function test has blurred the focus of the agency analysis and may have had such an impact.

     In 1980, the Supreme Court declined an opportunity to expressly rule on whether the National Security Council was an agency under the FOIA, thus leaving the issue unresolved for future litigation. Kissinger v. Reporters Committee for Freedom of the Press revolved around documents belonging to Henry Kissinger, who held positions as both a presidential advisor (as Secretary of State) and as a member of the NSC.    The Court found that NSC documents had not been identified in the FOIA request filed by William Safire; rather documents relating to Kissinger’s position as advisor to the President were implicated.   As a result, the Court found it unnecessary to rule on the status of the NSC.  However, a case that was filed on the final day of the Reagan presidency did precisely deal with the issue of whether the National Security Council is an agency as understood by the Freedom of Information Act.

     The litigation was filed on January 19, 1989 by former Washington Post reporter Scott Armstrong, the National Security Archive and other researchers and historians.   The plaintiffs stumbled onto the fact that records from the Reagan administration were slated for destruction and filed three FOIA requests for all information contained on the Professional Office System (PROFS).   The PROFS was an internal communications system that allowed information to be sent to any of the computers linked to the central EOP network.   The plaintiffs requested all information from the NSC and EOP on the system from its date of installation in 1985 to the conclusion of the second Reagan term,  hoping to find evidence of early executive awareness of the activities that would later develop into the Iran-Contra scandal.   On the same day of the requests, the plaintiffs also filed suit seeking a declaration that much of the information situated on the PROFS and related backup tapes were federal and presidential records and sought an injunction against their destruction.   This disposal was planned for the following day to coincide with the change of administration.

     The case went through a series of steps and decided a number of novel issues along the way, including whether the FRA or PRA preclude judicial review,  whether private parties could file an action,  discovery issues,  and record maintenance and classification issues.     In February, 1995, the agency issue was brought before the District Court for the District of Columbia.

     The event that inspired litigation on this particular issue was a declaration made by the NSC, currently defending suit against the record requests, that it was not an agency for FOIA purposes, despite a long history of acting as an agency, admitting it was an agency and disclosing materials in compliance with the FOIA.    This declaration came as a result of a memorandum President Bill Clinton issued to National Security Advisor Anthony Lake  and NSC Executive Secretary William Itoh on March 24, 1994.   In this memo, President Clinton stated his strong support for the position of previous administrations of generally permitting public access to NSC records.   President Clinton issued three directives:  to establish procedures for public access to NSC documents from previous administrations,  to establish procedures for access to the current administration, and to develop a plan to provide copies of NSC records for the next administration.   Despite these directives, NSC Executive Secretary Itoh issued a ostensibly inconsistent memorandum   the following day that directed the NSC to immediately revoke the current FOIA guidelines.   President Clinton’s directives were interpreted to represent a new voluntary disclosure policy.   The memorandum claimed that the NSC was an entity within the Executive Office of the President whose sole function was to advise and assist the President and "[a]s a matter of law, therefore, the NSC is not an ‘agency’ as that term has been defined and construed under the Freedom of Information Act."    According to the memo, the NSC should consequently keep its records under the procedures established by the PRA, not the FRA.   Due to this turn of events, the District Court for the District of Columbia was again brought back into the fray of the Armstrong litigation and asked to resolve the issue of whether the National Security Council is an agency subject to FOIA, notwithstanding its reconsideration of the matter.

     In its decision written by the late Judge Charles Richey, the District of Columbia Circuit Court noted that, under Soucie, Sierra Club and Pacific Legal Foundation, only one additional function beyond advice and assistance needed to be established in order for an entity to be considered an agency.   In scrutinizing the situation before the court, it framed the admittedly slanted question:

"[w]hether a government entity, namely the NSC, which has historically treated itself as an agency, and has engaged in a multitude of functions independently of ‘advising and assisting the President,’ just like other components of the Executive Office of the President, which admittedly are ‘agencies’ subject to the Freedom of Information Act and the Federal Records Act, can unilaterally after many years of treating itself as an agency, suddenly change its designation without offering a reasoned explanation for the sudden change, and thus declare it is no longer an ‘agency?’"

        The court determined the answer to this question was a resounding "no."   The court made two initial inquiries to decide the issue.   First, the FOIA definition required that an entity must be an "establishment in the executive branch."   Because the NSC had a separate staff and a firm structure, meeting that portion of the determination which Meyer seemed to add to the analysis, the court ruled that the NSC did satisfy this requirement.

     Finding the staff and structure sufficient that the NSC could be considered amply established, the court then proceeded to its second prong.   This was the Soucie requirement of exercising "substantial independent authority," which goes beyond merely advising and assisting the President.   The court believed that the NSC also met this portion of the agency test.   It found that the NSC performed rulemaking and adjudicative functions, which were "indicative of a classic ‘agency’ under the APA."   For rulemaking, the court cited to the fact that the NSC published its regulations in the Code of Federal Regulations.   Also, it issued a circular on dealing with telecommunications systems.

      In addition, the NSC clearly had a number of adjudicative functions.  The NSC had authority to review and overturn declassification orders, review and make determinations for other agencies on FOIA matters and deal with government security agreements.    Also lending credence to a finding in support of agency status, the NSC exercised substantial authority in key policy areas independent of the President.    Such is the case in the NSC’s role as head of the Central Intelligence Agency.    It is also "the highest Executive Branch entity that provides review, guidance, and direction to the conduct of all national foreign intelligence, counterintelligence, special activities, and attendant policies and programs."   Further, the NSC has the functions of guarding information on national security independent of the President, creating federal telecommunications policies for emergency conditions, and has responsibility for national security preparedness policy.   Also, the NSC is vested with roles in arms verification, non-proliferation, and public diplomacy.   As a result of these various appointments, the court believed the NSC failed the sole function test for acting with substantial independent authority, and was thus properly classified as an agency.    According to the court, the change of position of the NSC was "contrary to history, past practice and the law."

     The court distinguished Meyer by the fact that the Task Force on Regulatory Relief was established solely by the President, while the NSC was of statutory creation and no single individual, including the President, could eliminate it.   Also, the court found the fact that the task force had no separate staff to be an important consideration in Meyer, while the same situation was not present in regard to the NSC.   Finally, and most importantly, the NSC had adjudicatory functions that the task force lacked.   On these facts, the court found that the NSC was far more similar to the Office of Science and Technology and the Council on Environmental Quality, which were held to be agencies in Soucie and Pacific Legal Foundation, rather than the Meyer Task Force, which was found not to be a FOIA agency.   The court held that any document created by an NSC official who solely advised and assisted the President must be subjected to the PRA, while the FRA would apply in all other situations.   Arguably, this result could be read to contradict Ryan and Pacific Legal Foundation’s refusal to recognize a dual capacity under the FOIA, and implicitly Soucie, which created a test that only allowed for identifying an entity as an agency or as a non-agency in its entirety.

     Finally, the court had trouble dealing with the NSC change of position on its agency status because it did not offer a reasoned explanation for its change.   As a result, even if the NSC did not have the type of adjudicatory and rulemaking roles that were sufficient to qualify it as an agency, the court would still have set aside the NSC decision as arbitrary and capricious under the Administrative Procedure Act.

     The government appealed the decision assigning agency status, and Armstrong cross-appealed the decision creating the exception for presidential advisors.   On appeal, the decision holding the NSC an agency was reversed.   The Court of Appeals framed a more simple and objective question than the lower court, namely "whether the National Security Council is an ‘agency’ subject to the Freedom of Information Act … that is, whether the NSC is an ‘executive department … or other establishment in the executive branch.’"   The court differentiated the issues of whether an entity exerts "substantial independent authority" or whether an entity’s "sole function is to advise and assist the President."    The court claimed that the three-part Meyer test harmonized these two different inquiries and began to apply that test, noting that each of the three factors do not weigh equally.   Instead, "each factor warrants consideration insofar as it is illuminating to the particular case."   Representing a departure from the lower court’s perspective,  and perhaps precedent generally,  the Court of Appeals believed the existence of an additional independent function or functions beyond merely advising and assisting was not necessarily fatal.   Rather, it is "the degree of the NSC’s independence in discharging them, that matters."

     The court found that the structure of the NSC was sufficiently self-contained with a firm structure, a separate budget and a staff, unlike the President’s personal staff.   The court weighed much more heavily the fact that the NSC works very closely with the President.   Determining that the well-defined structure does not outweigh the close proximity of the NSC in relation to the President, the next issue was whether there existed substantial independent authority.   Because of the weight the court attributed to operational proximity in the three-prong balancing, it stated that in order for Armstrong to prevail, his showing that the NSC performed agency functions would have to be "compelling."   The court found that neither the act that created the NSC, nor any subsequent Executive Order, had delegated it the power to act without direct supervision of the President.

     The court reached this decision by concluding that the NSC does not direct the CIA, except by relaying messages directly from the President, and does not have any real adjudicatory roles over protecting classified materials; instead it acts only as a monitor over other agencies.   Also, the court found no evidence of non-advisory authority in the NSC’s development of the emergency telecommunications policy or NSC policies and guidelines dealing with emergency preparedness and crisis management.   Nor was there any independent authority exercised in areas of nuclear non-proliferation or public diplomacy.    Ultimately, the court found that Armstrong had not satisfied his compelling burden of showing that the NSC had exerted substantial independent authority and, therefore, the NSC was not an agency under the Meyer test.

     Although the NSC behaved inconsistently in referring to itself as an agency which performed agency functions in some circumstances, the court said this was not a concession of agency status.   Also, while the classification could be set aside under the Administrative Procedure Act for lacking a reasoned explanation according to the lower court, the APA only applies to agencies and as a result is not relevant to the NSC because the court had just determined that the National Security Council acts only as presidential advisor.

     Judge David S. Tatel dissented.  He noted that, although the sole function and substantial independent authority inquiries could point in different directions, precedent seemed to require that the two be looked at "as essentially opposite sides of the same coin."   Thus, if an entity has significant functions beyond advising and assisting, then the entity possesses independent authority and is therefore an agency.    According to Tatel, the Meyer test was a way to understand the functions of an entity, not factors to be weighed against each other.
Judge Tatel disagreed with the majority about what was meant by independence, noting in his opinion that "[c]rucial to the ‘agency’ inquiry is not whether an entity is unaccountable to all other agencies within the executive branch, but whether the entity has the power to act provided no higher authority disapproves."   In the opinion of Tatel, the focus should have been on whether the entity takes action itself, and not whether the actions are taken to assist the President.   Because a number of functions of executive branch activities are taken ultimately to assist the President, an overly broad approach would allow the "‘advise and assist’ exception to swallow the FOIA rule."    Judge Tatel reiterated that only one additional function beyond advice and assistance need be found for an entity to fail the sole function test, and then proceeded to enumerate a number of areas that the NSC exercised substantial independent authority.   These included information security, telecommunications, non-proliferation, and public diplomacy.   He also believed that subjecting the NSC to the Federal Records Act was consistent with the legislative intent.

     Unlike the majority, which had no reason to proceed to the Armstrong cross-appeal on exempting high NSC officials when solely advising and assisting the President because it found the NSC lacking of agency standing, Judge Tatel did reach the issue.  However, instead of finding that the NSC should be treated as a hybrid as the lower court essentially held, he believed that an NSC member who also held another position should be exempt from the FRA and FOIA when acting in his or her non-NSC capacity,  a position that may avoid much of the tension with Ryan. The key to the situation, as it was under Kissinger,  is whether the official possesses a non-NSC position.

     Despite this vigorous dissent, Armstrong lost the case and ultimately received nothing from the National Security Council, as the Supreme Court denied certiorari in May, 1997.   Arguably, this result emerged from an altered approach; an approach which conceivably changes the fundamental focus of the agency analysis, while at the same time claiming to be wholly consistent with past precedent and the core inquiry of the sole function test.

     In a sense, the Soucie test seems to be a fairly simple way to determine agency status.  Some of the more recent cases complicate the analysis considerably, specifically the three-prong analysis of Meyer and the way it was applied in the Armstrong decision. Clearly, the point of paramount force in finding lack of agency in Armstrong finds its seed in Meyer.   Since its origin in Soucie, the analysis into whether an entity exerts substantial independent authority or exists solely to advise and assist has generally considered these questions together as a single inquiry, as a converse proposition carrying the same intended meaning.   However, Meyer found it necessary to "harmonize" the sole function and independent authority issues, treating the two inquiries as sufficiently different to require this reconciliation.   Meyer introduced the three factors, and Armstrong’s subsequent reading of Meyer seems to have more formally incorporated these factors into the controlling test, making the inquiry into sole function essentially a three-part balancing approach.   This new approach has muddied the picture considerably.  And while these changes have been interpreted to build on past precedent, in a number of ways they seem to contradict the earlier focus of agency analysis in a fundamental fashion.  Arguably, it is the different perspective brought about by the three-part-test, as well as some more subtle changes in the approach generally, that ultimately decided Armstrong.  Had the Soucie inquiry been applied in its purer form to the NSC, it appears less likely that the result in the Court of Appeals case would have been the same.

     Meyer and Armstrong  may represent a shift of position of the courts on this issue, a shift that appears inapposite to the general presumption of disclosure that freedom of information entails.   Perhaps the interest in access to governmental material by individuals has been diluted by recent court opinions.   These opinions have refined their approach and begun a trend of deciding the issue in a way less friendly to these access interests, at least this appears to be the case in regard to EOP entities.  It has been argued that the situation brought about by the Armstrong decision represents a tightening in the interpretation of the FOIA by the judiciary generally.   Such a stiffening seems to contradict the underpinnings of the Freedom of Information Act in maintaining public accountability.  Framed like this, there could be strong policy concerns raised when an entity shields itself categorically from the kind of accountability that is meant to flow from the FOIA.   These policy concerns serve as the backdrop to an examination of the Meyer/Armstrong departure from precedent, a deviation that may have undermined the importance society places on governmental openness.

     The sole function and substantial independent authority inquiries both find their source in the oft-cited case of Soucie v. David.   Soucie set forth the sole function test, the same test formally endorsed by the Conference Report for the 1974 FOIA amendments and by the Supreme Court in Kissinger.   It also set forth the substantial independent authority inquiry.  Interestingly, it did so in the context of analyzing the APA’s definition of agency, which was then the definition that the FOIA incorporated.  The court noted that "[t]he statutory definition of ‘agency’ is not entirely clear, but the APA apparently confers agency status on any administrative unit with substantial independent authority in the exercise of specific functions."    Ultimately, years of precedent were built on the subsequent readings of this portion of the Soucie opinion, which was in turn analyzing the definition of a provision that has not been in effect as the FOIA definition since 1974, because it was not broad enough.   In the years since Soucie, both the sole function and substantial independent authority language has been used by courts engaged in agency analysis, and often it seems they have used the terms interchangeably.

     According to the Court of Appeals in Armstrong, it was the differences in meaning between the sole function and substantial independent authority inquiries that required the three-part test of Meyer for reconciliation.   The inevitable question is whether these inquiries are contradictory or contrastive in any way, or if sole function and substantial independent authority instead really do carry the same intended meaning.  For if they really mean the same thing, there can be little need for reconciliation.  In situations such as what Grumman presented with the Regional Boards, where the entity is not encompassed in the Executive Office of the President and does not claim to be a presidential advisor, the sole function language clearly should be inapplicable.    Instead, the inquiry is simply whether the entity exercises substantial independent authority in specific functions.  However, in the case of an EOP entity where presidential concerns such as separation of powers are raised, the issue could properly be phrased as whether the entity exerts substantial independent authority to the extent that it does not exist solely to advise and assist the President.   Operationally, the two questions appear to be asking the same thing, merely in converse form.  It is hard to imagine an entity that has substantial authority independent of the President and yet solely exists to advise and assist him.  The two inquiries may very well be, as Judge Tatel states, "essentially opposite sides of the same coin."   The pre-Meyer cases seem to have used the language as if they were talking about the same thing.   It was not until Meyer and Armstrong that the inquiries were split into two different considerations, as opposed to viewing them as two ways to ask the same question.  Thus, it was necessary to "harmonize"   the sole function and substantial independent authority analysis using two factors that were "entirely creatures of the [Meyer] majority’s making."    Arguably, this need for reconciliation arises from the Meyer court’s attribution of a meaning to these inquiries that is different from how the case law had treated them for twenty years previous.

     Of course, it can be argued that it is the duty of the courts to advance the law.  When circumstances require, a court acts properly when it rethinks earlier interpretations or departs from them completely.  No one can argue that simply because precedent states that any particular test or factor is of the highest importance or the most relevant, or is controlling on the issue, that it shall remain so.  Indeed, the law is fluid, and it is the duty and obligation of the courts to alter doctrine when it serves the goals of equity to do so.  Likewise, it is not to be asserted here that the traditional incarnation of the sole function test is an unmovable rock that must remain true and without change to perpetuity simply because it carries precedential value.  Indeed, to refuse to allow the law to evolve when necessity dictates would be a substantial breach of duty on the behalf of the judicial system.

     The question is whether this change best serves the purpose of justice and the FOIA.  In regard to the sole function test, the addition of the inquiries that are extraneous to the original interpretation of the test complicate what was a fairly straightforward inquiry.   These additions, it is asserted here, do not best serve the purposes of the FOIA to maintain accountability.  Instead, they create leeway for an entity to argue its way out of an agency classification and cheat the public of the power to govern that supposedly is derived directly from them.  What is to be argued here is not merely that the new incarnation of the sole function test is a departure from precedent, but also that this departure is misplaced when measured against the presumption of openness that FOIA is intended to represent.

     The sole function test is as an exception from FOIA requirements that should be read very strictly.  The exemption is meant to satisfy separation of powers issues and recognizes that the President and other individuals and groups who aid him in making executive decisions should remain outside of the requirements of disclosure.   When sole function asks simply whether the entity in question acts strictly to advise and assist the President, it is recognizing that some entities may be nothing more than an advisor and is in actuality encompassed by the Presidential office with no crossover into broader matters.  Thus, a court may perform a disservice to the democratic principle when it expands its inquiry to include matters which have no real connection with the underlying issue of whether an entity does more than advise and assist.  For when it does so, it may create a more open-ended balancing that arguably makes undue room for an entity to claim FOIA inapplicability.  It could be argued that when an entity does exert any kind of authority in matters beyond mere advice and assistance to the President, it should be declared an agency regardless of whether it does not take on the form that agencies generally do or that it works in close proximity to the President.   Sole function is a strict limitation, not a test that should be used casually in order to exempt entities that may possess the feature of advising the President in addition to other endeavors.

     In finding that the NSC was an agency, the district court in Armstrong applied a less convoluted inquiry that revolved around the more established Soucie-type analysis.  Contrary to this more classic approach, it was the Meyer course that the Court of Appeals used to determine the lower court’s agency assertion incorrect in reversing that decision.   By elaborating on Soucie and its progeny, the Meyer and Armstrong courts may have effected change at the foundation of the agency analysis.   This change is realized in the three-part test which stresses form and operational proximity, in addition to delegated authority, in its examination.   However, the additions of these considerations may focus the light of an agency investigation on issues that do not relate well to the fundamental inquiry of the sole function test as developed by precedent—a focus that stresses function as the single relevant consideration.

1.  Focus on Form

     Meyer may have placed undue weight on the form of the task force in determining whether it qualified as an "establishment in the executive branch."   The Meyer focus originated from the perspective that "form follows function,"  or as the dissent states, "form dictates function."   The court noted that the lack of a separate staff was a strong indicator that the Task Force on Regulatory Relief was not an agency for FOIA purposes.   Such an inquiry places importance on a matter that is not central, and arguably entirely irrelevant, to the focus of the agency analysis under precedent.  Precedent would suggest the inquiry is into the function of the entity, not whether it has a firm structure or separate staff.  The focus should be "on how an entity functions, not where it gets the resources to perform those functions."   Such has been the principal inquiry since Soucie   and it has been followed through consistently in its progeny.   Under this focus, form appears to be a totally unrelated inquiry.

     Contrary to the Task Force on Regulatory Relief, the Armstrong court found that the NSC had a definite structure with a separate staff and satisfied that portion of the Meyer test.    Thus, this inquiry did not have the adverse affect in that case that it potentially could have had for the plaintiffs, or may have in future litigation involving other entities that lack the feature of a firm structure. However, this inquiry may still have been inappropriate.  The Armstrong court noted that the existence of a firm structure made the National Security Council "less like ‘the President’s immediate personal staff or [a] unit[ ] in the Executive Office [of the President] whose sole function is to advise and assist the President.’"   The court’s analysis amounts to an inquiry of "if it looks like a duck … it must be a duck."  Precedent would seem to indicate that the analysis should consider whether the entity actually acts like a duck.   Indeed, it appears evident that the President has a number of options available to him to create entities that look nothing like a classic agency in form and structure, yet still exercise substantial independent authority.   Advisors and assistants may come in a variety of forms, as do entities wielding significant power.  The question of whether an entity looks like an advisor in form is a misplaced inquiry.  The presumption for governmental openness dictates that the inquiry be limited to whether the entity serves only to advise and assist, not whether its form may or may not limit it in such a way.  Form may not be an appropriate indicator of whether an entity exerts independent authority.  It is thus incorrect to attribute weight to this inquiry.  By ignoring such tangent questions, Soucie represents the better test because it limits its inquiry to the relevant issue, of which a focus on form is a distraction that does little to point in the right direction, and may mislead the finder of fact from the true answer of whether the entity exists strictly to advise and assist.

2.  Operational Proximity

     Also less than relevant is the inquiry into the operational proximity between the entity and the President.  The Meyer court worked from a starting point that Congress intended by the "advise and assist" exception to imply that entities with characteristics of the President’s immediate staff were to be exempted.   Thus, Meyer’s requirement of operational proximity does seem to make sense.  However, again, operational proximity does not relate to the controlling test, namely whether an agency has substantial independent authority or merely exists solely to advise and assist the President.   The Meyer majority, according to Judge Wald,  mistook the Conference Report for the 1974 FOIA amendments, which defined agency as excluding "the President’s immediate personal staff or units in the Executive office whose sole function is to advise and assist the President."   The Meyer majority essentially ignored the "or" in the statement and attributed "proximity … in the sense of continuing interaction" to EOP branches and the sole function test.   The court seemed to have assumed that Congress therefore meant that entities outside the President’s immediate staff, that nonetheless resembled the White House staff in that it shares a close proximate relationship with the President, could therefore escape agency determination.

     The Court of Appeals in Armstrong weighed the operational proximity prong with "significantly greater weight" than the existence of a firm structure, which the NSC clearly had.    The court shouldered Armstrong with the heavy burden of proving the third prong that the NSC exercises the authority sufficient to declare it an agency.  In order to prove the NSC an agency he would have to make "a strong showing indeed."   Thus, the court determined under the particular facts of the case that operational proximity of the NSC to the President was a more significant issue than structure and form and therefore tipped the balance against Armstrong.

        On the final issue, that of delegated authority, Armstrong’s increased burden was to provide evidence of power that was "compelling."   This was one of the key impacts of the three-prong balancing as applied in Armstrong.  Because the NSC worked so closely with the President and the court attributed this factor more weight than the structure criterion, the plaintiff’s burden was elevated to compelling.  This is despite the fact that operational proximity may have little to do with whether the entity actually exercises substantial independent authority.  Again, the focus according to precedent is on the functions of the entity, not its other characteristics. The entity’s proximity to the President is a peripheral issue at best.

     Here again, earlier precedent may state the better test.  Under this inquiry and Armstrong’s handling of it, virtually all parties seeking information from EOP entities will be shouldered with the compelling burden of proving authority, because all Executive Office entities generally operate more closely with the President, relative to independent agencies.   As is the case with the NSC, the President is also an ex officio member of a number of other EOP groups.   This prong has the potential of tipping the presumption of sole functionality in favor of the governmental being in virtually all EOP cases.  This represents a blatant contradiction to what is intended to be a presumption of openness and may do more to aid EOP agencies in successfully asserting non-agency claims in the future than any other portion of the ill-advised three-part test.

3.  Nature of the Delegated Power.

     The third prong of the Meyer test, the inquiry into the nature of the delegated power, is the more appropriate issue and is fairly similar to asking whether there exists the authority to act independently.   Unlike the proximity and structure prongs, an investigation into delegated authority appears to be very much akin to the main thrust of the Soucie version of the sole function test and where the inquiry should lay.  However, the way functions were analyzed by Meyer and Armstrong may have been through the utilization of a more narrow reading of "independent" than precedent would suggest proper, by looking to the degree of independence instead of simply whether authority existed in any specific function.

     The Meyer majority believed that if the task force resolved disputes in accordance with the President’s wishes, it would not be acting in an independent manner.    According to Judge Wald, the majority’s definition of independence was too strict and narrow because under that definition, "[o]nly renegades or freelancers who ignored or disregarded the President’s orders would be seen to ‘act independently.’"   The approach taken by Meyer and Armstrong may support the notion that all EOP units are merely presidential advisors, if read too broadly, because all act closely with the President in executing his orders.

     At issue, then, is whether the fact that the NSC would not be expected to act in any manner except that which represents the President’s wishes should mean that it lacks independence.   All executive agencies, ultimately, exist to serve the will of the President.   Also, the definition of "agency" under the Administrative Procedure Act, used for many years by FOIA, includes an entity "whether or not it is within or subject to review by another agency."   Admittedly, the President would probably not be properly characterized as an agency.   However, the APA definition does suggest that subordination to another entity does not mean that an entity lacks the power to take direct control, and is therefore not an agency.  An entity can be subject to the authority of another entity and still be an agency.    Thus, the fact that the NSC must answer directly to the President should not be conclusive proof of non-agency status.  The 1974 Freedom of Information Act amendment on the agency definition may make this point as well, by its explicit inclusion of the Executive Office of the President under its statutory wing.  Nor should it matter that the NSC carries out the wishes of its President in determining the requisite level of independence.  The appropriate focus is not whether the entity takes actions that assist the President, but whether the entity acts independently in doing so.

      On this point, it is worth noting President Reagan’s assertion that he was "not fully informed" about the actions taken by the NSC in regard to diverting funds to the Nicaraguan Contras, an activity that led to the Iran-Contra scandal of the mid 1980s.    If the President could be unaware of such an important activity (assuming his assertion is accepted as accurate), how could it ever be argued that the NSC does not act independently in a meaningful way when it perpetrates such a consequential operation?

      Another difference in the interpretation of the delegated authority prong is in the threshold level of independence required to qualify an entity as an agency.  While precedent seems to suggest strongly that only one additional function need be found to fail the sole function test,  the recent cases have instead looked to the degree of independence exercised in each function.   Case law suggests that delegation of power in a number of areas is acceptable to find substantial independent authority.  Among these areas are the ability to bind agencies to policy guidelines,  to evaluate federal programs,  the existence of other "management, coordination, and administrative functions,"  or if the entity acts as an "independently controlling entity."   This raises the question of whether the NSC possesses any of this authority.  Under pre-Meyer and Armstrong precedent, if it does so in only one situation that should be sufficient for a finding of agency status.   Armstrong finds itself in disharmony with this point and may ultimately stand for the proposition that authority in any single function does not necessarily mean that the entity is an agency.  This is a proposition that seems to contradict the implied meaning of the "sole" function test.

      Due to the fact that EOP entities may all share the characteristic of having some advising and assisting duties in regard to the President, requiring multiple functions as opposed to a single function which goes beyond advice and assistance may make it easier for these entities to flee the bondage of FOIA compliance.  Because the sole function exemption seems to presuppose that the entity is in practice nothing more than an executive advisor and is shielded only because of respect for presidential decisionmaking, this may simply be a wrong interpretation that, again, defeats the presumption and intention of the Freedom of Information Act for disclosure and accountability.

4.  Other Considerations

     Separate from these more fundamental changes, other issues deal with how the NSC treated itself before the litigation and Congressional understanding of the status of the entity.  In Soucie and other previous cases, proof of an entity behaving like an agency in adopting FOIA regulations and Congressional understanding that the entity would be an agency served as considerations "of some weight."    However, the Armstrong court dismissed the NSC behavior consideration completely, believing that this point was irrelevant.  If an entity is not an agency to begin with, it does not matter if it pretends to be one.   Realistically, this position has some logic to it.  And while this consideration has gone from carrying weight to an irrelevant or moot point, this change in the weight the courts have been willing to accord this consideration does not relate to the sole function inquiry—the crucial consideration—and was never a controlling point in the early agency decisions.   Consequently, this point arguably only acts as background for a perhaps shifting view on this issue.

     For notable reasons, this change in focus could potentially have important effects on future cases, and this three-part test may have already been the critical determinative of Armstrong.   While the Meyer test may have been a way to understand the functions of an entity as opposed to factors to be weighed against each other,  this test makes operational proximity and the issue of self-contained structure parts of the controlling test when applied.  It may be very easy to imagine a scenario where these additions could lead to results entirely inconsistent with that which would be developed under the more classic formulation of the Soucie inquiry.  Indeed, such may have already been the case with the Armstrong litigation and the NSC.

     Under the three-part test, the elements are not to be weighed equally; instead, according to Judge Tatel, "each factor warrants consideration insofar as it is illuminating to the particular case."   Such an ad hoc approach makes a finding against agency status more of a possibility because the decision of which element is the most important is up to the particular case and the particular court conducting this balancing.  In Meyer, the court paid special attention to the Task Force’s lack of separate staff, while the Armstrong court was concerned enough about the close proximate relationship between the NSC and the President that the plaintiff’s burden to prove the existence of sufficient authority was elevated to the level of compelling.  Such a burden may have been utterly misplaced; it would seem more appropriate that the burden to prove an agency or non-agency status should generally lay with the entity attempting to escape the recordkeeping and disclosure requirements that Congress has dictated.  The agency alone has in its possession the information that proves it is an agency or not; requiring a plaintiff to prove such is a substantial hurdle which tips the playing field unfairly and inappropriately in an inequitable direction, especially when the entity’s assertion is contrary to its past practices and admissions.

     By abandoning the more bright-line approach in favor of an ad hoc tripartite balancing where each court can determine for itself which factor is most probative on the issue of agency status, the Meyer/Armstrong approach represents an entirely new inquiry from that of the Soucie incarnation.  This elaboration places the focus on issues like structure and proximity which have no relation to the original inquiry in any meaningful way.  It also uses a more narrow understanding of independence that undermines the presumption in favor of access which the FOIA mandates.   While policy suggests the Meyer/Armstrong elaboration improper, precedent is incongruent with it entirely.

     And despite this distinct change in focus and the very real opportunity for this approach to impact the resolution of controversies, the courts have acted on the premise that this approach is entirely in conformance with precedent.   Certainly it is beyond dispute that the courts are free to reevaluate their position and change the approach when they find it prudent to do so.  However, the courts have not really considered themselves having done so on the agency issue.  Instead, the courts have acted as though they were building on precedent in a consistent manner, while both explicit and implicit contradictions have abounded in their decisions.  Arguably, they have done this not only with elaborations that confuse the original inquiry, but also by avoiding precedent to a degree.   Because these additions are at fundamental odds with the functional inquiry that sole function has always engaged in, the new rule of law that Meyer and Armstrong add to the investigation may not be legitimated under the guise of conforming with precedent.  Instead, it explicitly contradicts precedent.  Thus, the contention that this new approach does nothing more than clarify the true intent of Soucie is flawed and serves as an inadequate justification for the imposition of the three-part test.  Meyer and Armstrong do not just fine tune the focus of the sole function test.  They virtually obliterate it.

     More importantly, the departure from the classical version of the sole function test is at odds with both the presumption for openness that the FOIA intends and the understanding of the sole function test as a limitation to be leveled only when the entity in question is empowered to serve as nothing more than a presidential advisor. The three-part test put forth in Meyer and Armstrong may make it possible for entities which have the kind of authority traditionally subjected to public scrutiny to categorically remove themselves from accountability.  It expands the sole function exception to cover more than it was ever meant to, and more than it ever should.  In addition to being fundamentally at odds with earlier precedent, the new sole function test is fundamentally at odds with some of the most important intentions of the Freedom of Information Act.

     The approach the recent courts have taken in regard to the agency issue differs from the consistent flow of Soucie and immediate progeny.  The most important of these major fundamental changes invoked since Meyer is the shift away from the functional inquiry in favor of a view that takes into consideration issues which would be considered less than relevant under earlier precedent.  While the courts have considered this change to be in accordance with precedent, the issues of proximity to the President and the structure and form of the entity do not comport well with the original approach which stressed the functional nature of the entity.

        As far as the issue of the National Security Council, it appears that this change in focus has had a profound effect.   The Armstrong court’s heavy emphasis on proximity and its decision to attribute heftier weight to that prong turned Armstrong’s burden on its head and forced him into a situation where his showing would have to be compelling on the authority prong in order to prevail.  And in analyzing this issue, the court looked to the degree of independence in each function instead of whether there was authority to take action in any function, a slightly different approach which, again, may be a change of a fundamental nature.   This new approach represents an obscuring of the original Soucie inquiry, and may not bode well for future litigants seeking access to information possessed by other arms of the Executive Office of the President.

     The importance of this change may be considerable.  The new approach of the sole function inquiry opens the door for many of the other entities embedded within the EOP to claim that they are not agencies subject to FOIA, and win.  The FOIA is a heavily litigated law.   The payoff for an agency who can convince a court that a classification of non-agency status is warranted is immense in terms of both raw finances and in freeing up human resources for other projects.  FOIA requirements would become an empty threat, a problem that others must deal with but one not worthy of even a mention in the course of daily activities.  Even if the odds of winning are only adequate, it may be worth trying the issue in court; the agency is likely to find itself embroiled in substantial FOIA litigation anyway.  With the new approach, the odds may now be far better than adequate, at least for those entities within the EOP that have at least some responsibility to the President as an advisor or assistant.  Most EOP entities work in close proximity to the President and some have staffs less concrete than the typical agency.  Nonetheless, many of these entities have substantial authority and would be properly classified as agencies.  Now they may all have a better argument to avoid agency status and skirt the requirements of the FOIA.

     But the FOIA is not the only statute that may now be avoided under this new approach.  As a non-agency, the NSC and future potential entities that may succeed on an identical claim may now avoid the Federal Records Act, as well.   The Presidential Records Act, and not the FRA, would apply to these organizations.   This result in not insignificant.  The PRA essentially allows the President to destroy documents almost at will.   Thus, not only is immediate disclosure cheated, but any embarrassing documents shedding light on improper governmental behavior may be wiped clean from the annals of history if they can be brought within the more lenient and deferential direction of the PRA.  The President may now arguably do this to a great number of documents by simply depositing them with the NSC.   A flight of other entities to non-agency status may substantially open up other options for a President who wishes to avoid disclosure on a permanent basis.  Such is the result of the refinement of the sole function test.

     It should come as no surprise that government entities would prefer to fall under the rubric of the Presidential Records Act, as opposed to the more demanding recordkeeping and disposal stipulations of the Federal Records Act, not to mention FOIA disclosure requirements.   With the recent turn of events in agency determination case law, it may now be much easier for an entity to achieve this intent, and cheat the prying eyes that accountability brings.
This desire nonetheless runs counter to the proper functioning of a democracy and creates the risk that the government will be able to successfully shield some of its inappropriate or embarrassing behavior from its electorate.   And the risk of a government pursuing some unethical goals through less-than-noble means should not be underestimated:  it was, after all, the Watergate scandal that inspired the PRA in the first place,  and the Armstrong litigation was itself initiated in the hopes of finding evidence of the Iran-Contra affair.    Regardless, the changed focus in the agency analysis of Meyer and Armstrong has severely undermined the ability of the public to effectively monitor its rulers, at least in regard to EOP entities, and may ultimately leave a legacy damaging to this most important function.  Evidently, it has already done so in the case of the National Security Council—an entity charged with the kind of national security matters that are clearly among the most important concerns a citizenry could ever imagine.
 

FOOTNOTES

   Pub. L. No. 90-23, 81 Stat. 54 (1967)(codified at 5 U.S.C. §552 (1997)).  The legislative history of the FOIA expresses the purpose of ensuring "the right of persons to know about the business of their government."  H.R.REP. NO. 876, 93d Cong., 2d Sess. 4 (1974)
   See Air Force v. Rose, 425 U.S. 352 (1976)(balancing individual privacy interests against "the preservation of the basic purpose of the Freedom of Information Act ‘to open agency action to the light of public scrutiny’").
    Among the FOIA’s success stories, the act has aided in uncovering Department of Housing and Urban Development’s funding of the projects of prominent republicans and friends, harassment by the Federal Bureau of Investigation of Dr. Martin Luther King, illegal surveillance by the Central Intelligence Agency, experimentation on federal prisoners, and such things as unsanitary conditions in food processing plants and safety violations in nuclear power plants.  KENT R. MIDDLETON, ET AL.,  THE LAW OF PUBLIC COMMUNICATION (1997)
   One source estimated that over $91 million was spent on 600,000 requests for information from the various agencies in 1991.  The State Department reported that each request it filed that year cost an average of $1,000.  "Requests, Costs Go Up in 1992 Annual Reports,"   Access Rep., March 31, 1993, at 3.  See also Eric J. Sinrod, Improving Access to Government Information in an Era of Budgetary Constraints 27 URB. LAW. 105, 109 (1994)(noting that "[m]any agencies strongly opposed the enactment of … time deadlines because of limited government resources and personnel.  These agencies feared that they would have to shift personnel from matters within their primary responsibilities to FOIA processing teams."
   See e.g., Eric J. Sinrod, Improving Access to Government Information in an Era of Budgetary Constraints, 27 URB. LAW. 105, 109 (1994)(noting the current average response time of the Immigration and Naturalization Service to a FOIA request in 1994 was eighty-five days.  Further, the INS had 12,536 pending requests at that time, of which 2,636 were more than a year old).
   Middleton, supra note 3, at 475
   Id.
   Electronic Privacy Information Center v. Federal Bureau of Investigation, 865 F.Supp. 1, 2 (D.D.C. 1994)(holding such a backlog did not excuse the FBI from processing a request in a timely manner).
   Middleton, supra note 3, at 475.
   See 5 U.S.C. §552(d) (1997)("[t]his section does not authorize withholding of information or limit the availability of records to the public, except as specifically stated in this section").  President Clinton’s administration has sought to make this presumption more explicit, a move that represents a stark departure from the access unfriendly administrations of Presidents Reagan and Bush.  Under new regulations and standards set forth by Clinton and Attorney General Janet Reno in 1993, "[t]he [Justice] Department will no longer defend an agency's withholding of information merely because there is a ‘substantial legal basis’ for doing so.  Rather, in determining whether or not to defend a nondisclosure decision, we will apply a presumption of disclosure."  Att'y Gen. Reno, Memorandum for Heads of Dep'ts and Agencies; Subject: The Freedom of Information Act (Oct. 4, 1993), reprinted in OFFICE OF INFO. AND PRIVACY, U.S. DEP'T OF JUSTICE, FOIA UPDATE 4 (1993)
   Pub. L. No. 104-231, 110 Stat. 3048 (1996).  Ironically, this law codified the decision, handed down in Armstrong v. Executive Office of the President, 1 F.3d 1274 (D.C. Cir. 1993), that the National Security Council’s electronically stored information did constitute records for purposes of the Freedom of Information Act and were unaffected by whether printed documents existed.  This decision was made at an earlier point in the litigation that is key to the issue to be covered in this article.  Because the court later ruled that the NSC was not an agency, the plaintiff in this case will never be able to take advantage of the new law that his case helped pioneer.
   See 5 U.S.C. §552 (1997)
   Indeed, the President is a statutory or ex officio member of a number of EOP entities, the National Security Council being just one of them.  R. Kevin Bailey, Did I Miss Anything?:  Excising the National Security Council From FOIA Coverage, 46 DUKE L.J. 1475, 1480 (1997)
   Armstrong v. Executive Office of the President, 877 F.Supp. 690, 708-711 (D.D.C. 1995)
   FOIA is binding on agencies, See 5 U.S.C. §552 (1997), while the President would not properly be classified as an agency.  Armstrong v. Bush, 924 F.2d 282, 288-9 (D.C.Cir. 1991).  But see Amalgamated Meat Cutters v. Connally, 337 F.Supp. 737, 761 (D.D.C. 1971)(where the court noted that the leading scholars "seem to be in agreement" that the Administrative Procedure Act construes the President to be an agency).
   Soucie v. David, 448 F.2d 1067, 1075 (D.C.Cir. 1971); H.R. CONF. REP. NO. 1380, 93d Cong., 2d Sess. 14 (1974); Kissinger v. Reporters Committee for Freedom of the Press, 445 U.S. 136, 156 (1980)
   Id.
   See, infra part IV
   The definition of agency used by the Freedom of Information Act was expanded in 1974 to explicitly include entities within the Executive Office of the President.  See H.R.REP. NO. 876, 93d Cong., 2d Sess. 8 (1974)
   See e.g., Armstrong v. Bush, 924 F.2d 282 (D.C. Cir. 1991)
   Soucie v. David, 448 F.2d at 1080-81
   See, supra note 15
   Armstrong v. Executive Office of the President, 90 F.3d 553 (D.C.Cir. 1996), reh’g denied, cert. denied, -- U.S. --, 117 S.Ct. 1842 (1997)
   44 U.S.C. §§2101-18, 2901-09, 3101-3107, 3301-24 (1988)
   44 U.S.C. §§2201-07 (1988)
   The Federal Records Act only subjects records "made or received by an agency of the United States Government" to its provisions, 44 U.S.C. §3301 (1988), while the Presidential Records Act explicitly excludes "official records of an agency."  44 U.S.C. §2201(2)(b) (1988)
  877 F.Supp. 690 (D.D.C. 1995), reversed, 90 F.3d 553 (D.C.Cir. 1996) reh’g denied, cert. denied, -- U.S. --, 117 S.Ct. 1842 (1997)
  See, infra parts IV and V
   See, supra note 13.  See also Armstrong v. Executive Office of the President, 90 F.3d at 572 (opining that, under the rule laid out by the majority, "[n]o EOP unit can be ‘independent’ of the President in the way that independent agencies outside the executive branch can").  This is because " … a unit of the EOP, is part of the immediate cadre of advisors and officials whose locus is strictly within the Executive Branch of the federal government.  As such, it is dissimilar from independent agencies, which have a more attenuated relationship to the Executive Branch. … [i]n the [Armstrong dissent], Judge Tatel warned that, by construing the ‘advise and assist’ language in broad terms, the court was laying the groundwork for excepting all EOP units from FOIA merely because the President’s involvement could somehow be traced to the various units."  Bailey, supra note 13, at 1496-97
   Pub. L. No. 235, 61 Stat. 496 (1947)(now codified at 50 U.S.C. §402)
   Reorg. Plan No. 4 of 1949, 14 Fed.Reg. 5227, 63 Stat. 1067 (1949)
   50 U.S.C. §402(a) (1988)
   50 U.S.C. §§403-3(a)(1) (1988), 403-3(c)(6) (1988)
   Armstrong v Executive Office of the President, 877 F.Supp. 690,700-01 (D.D.C. 1995)
   See 50 U.S.C. §402(c) (1988)
   Catherine F. Sheehan, Opening the Government’s Electronic Mail:  Public Access to National Security Council Records, 35 B.C. L. REV. 1145, 1152 (1994)
   Id. at 1153
   50 U.S.C. §402(a) (1988)
   Sheehan, supra note 36, at 1153
   Id. at 1156-58
   Id. at 1156-57
   Id. at 1156-58
   See TOWER COMMISSION REPORT, (BANTAM BOOKS 1987).  The Hughes-Ryan amendment orders the President to inform Congress "fully and currently" of any covert intelligence operations.   Also, any information about illegal intelligence activity is required to be reported in a "timely" manner.  50 U.S.C. §413 (1988)
   Sheehan, supra note 36, at 1158-59
   Id. at 1159
   See Armstrong v. Executive Office of the President, 877 F.Supp. 690 (D.D.C. 1995), reversed 90 F.3d 553 (D.C. Cir. 1996), reh’g denied, cert. denied, -- U.S. --, 117 S.Ct. 1842 (1997).  See also, infra part V.
   Exec.Order No. 12,356, 47 Fed.Reg. 14874 (1982)
   Exec.Order No. 12,829, 58 Fed.Reg. 3479 (1993)
   Exec.Order No. 12,046, 43 Fed.Reg. 13349 (1978)
   Exec.Order No. 12,656 §104(a), 53 Fed.Reg. 47491