PRIVATIZATION AND FREEDOM OF INFORMATION:
AN ANALYSIS OF PUBLIC ACCESS TO PRIVATE ENTITIES
IN THE UNITED STATES
CRAIG D. FEISER
A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ARTS IN MASS COMMUNICATION
UNIVERSITY OF FLORIDA
I thank Dr. Bill F. Chamberlin, Joseph L. Brechner Eminent Scholar in Freedom of
Information and chairman of my master's supervisory committee, for his hard work and
encouragement in assisting my efforts to produce this thesis. My efforts to complete this
monumental and time-consuming project would not have been possible without his
undying support and advice. I also offer thanks to committee members Fletcher N.
Baldwin Jr., Professor of Law, and Sandra F. Chance, Assistant Professor of Journalism
and Communications, for their valuable and insightful contributions in assisting with the
completion of this project. Finally, I would like to offer thanks to my many student
colleagues at the University of Florida colleges of Law and Journalism and
Communications, for their many words of encouragement and advice as I worked through
the stages of completing this thesis.
TABLE OF CONTENTS
ACKN OW LED GM EN TS ................................................ ii
ABSTRACT .................................................... v
I. PRIVATIZATION AND FREEDOM OF INFORMATION:
AN OVERVIEW.......... ...................................... 1
Introduction ............ .... ...................... .... ......... 1
Privatization and the Debate Over Public Accountability ....................... 10
II. PRIVATIZATION AND THE FREEDOM OF INFORMATION ACT ......... 34
Privatization and Freedom of Information in Federal Courts .................. 39
The Definition of "Agency" Under the Act ......................... 69
"Agency Records" Under the Act . ........ .......................... .. 70
III. FREEDOM OF INFORMATION UNDER STATE STATUTES .............. 75
No Definition of "Agency" or "Record" ..... . . . . . . . . . . . . 77
Flexible Approaches: The Function of the Entity and the Nature of Its Records .... 77
Restrictive Approaches: Focusing on Public Funds ................... 84
Restrictive Approaches: Close Nexus with the Government ................... 89
Summary .......... ................................. ......... 104
IV. STATE COURTS AND PRIVATE ENTITIES SUBJECT
TO ACCESS LAW S ....... ................................... 106
The Flexible Approaches Favoring Access ......................... 108
The Restrictive Approaches Limiting Access ....................... 135
Summary .......... ................................. ......... 147
V. AN ANALYSIS OF STATE AND FEDERAL
APPROACHES TO PRIVATIZATION . . ....... ...................... 148
V I. C O N CLU SIO N ..................................................... 169
BIBLIOGRAPHY ....... ....................................... 172
BIOGRAPHICAL SKETCH . . ....... ................................ 187
Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment
of the Requirements for the Degree of Master of Arts
in Mass Communication
PRIVATIZATION AND FREEDOM OF INFORMATION:
AN ANALYSIS OF ACCESS TO PRIVATE ENTITIES
IN THE UNITED STATES
Craig D. Feiser
Chairperson: Bill F. Chamberlin
Major Department: Mass Communication
Since the 1980s, government agencies at the federal, state, and local levels have been
continually turning to private entities to provide public services formerly provided by the
government. This privatization phenomenon has developed as many governments are
seeking more efficient and cost-effective ways to provide services to the public.
Frequently, these governments believe private entities are better equipped to handle
services such as incarceration, food service, research services and public safety.
However, one aspect of this privatization trend has been the fear by many commentators
and members of the media that previously important public information will now be
shrouded in secrecy once it is held by private entities. This fear has developed because
neither the federal Freedom of Information Act nor the access statutes in a majority of the
states explicitly cover private entities under their definitions of "agency" and "agency
records." Thus, important public information could become shielded from public view
simply because it is no longer in government hands.
This thesis is an exhaustive analysis of the privatization trend and the laws at the
federal and state levels that are meant to protect public access in the face of contracting
out services to private entities. The purpose is to discuss the Freedom of Information Act
and the access laws of all fifty states in terms of both statutory language and court
decisions interpreting the statutes. The thesis then concludes by categorizing and
comparing the laws at the federal and state levels to determine whether they are truly
protecting public access in the face of government privatization efforts. These efforts
often result in less public oversight due to the technical requirements embodied in the
"agency" and "agency records" definitions, even though the information requested
nonetheless pertains to important public services.
PRIVATIZATION AND FREEDOM OF INFORMATION: AN OVERVIEW
The private corporation cannot be entrusted with the responsibility of
maintaining and nurturing the rights of individual. Part of the basis for the
separation between the public and private enterprise is to protect the citizenry
from the tyranny of both entities. When decision-making, planning and
programming that were under the auspices of the public government are
transferred to the control of a private corporation, the city residents lose
whatever recourse they previously possessed to provide redress for their
From Shirley L. Mays, Privatization of
Municipal Services: A Contagion in the Body
Politic, 34 Duq. L. Rev. 41, 68 (1995).
This legislation springs from one of our most essential principles: a democracy
works best when the people have all the information that the security of the
Nation permits. No one should be able to pull the curtains of secrecy around
decisions which can be revealed without injury to the public interest.
Statement by the President Upon Signing Bill
Revising Public Information Provisions of the
Administrative Procedure Act, July 4, 1966, 2
WEEKLY COMP. PRES. DOC. 895 (July 11,
In recent years, the rising cost of government has caused a debate over the pros and
cons of privatizing government services. While some call privatization a panacea of
efficiency, others worry about potential public sacrifices.1 Among these worries is the
fear that unless access to information statutes can be extended to cover private entities,
privatization will undermine the public's right to know.2 This is because the federal
Freedom of Information Act and most state access statutes do not explicitly grant access
to documents that are in the hands of private entities, requiring courts to deal with
requests for information relating to the government but in the hands of private entities.3
Without legislative action or court intervention interpreting these statutes broadly, most
governments could effectively transfer their documents into the hands of private
companies and avoid the reach of freedom of information acts.4
ICompare Shirley L. Mays, Privatization of Municipal Services: A Contagion in the
Body Politic, 34 Duq. L. Rev. 41, 68 (1995); Nicole B. Casarez, Furthering the
Accountability Principle in Privatized Federal Corrections: The Need for Access to
Private Prison Records, 28 U. Mich. J.L. Reform 249 (Winter 1995) 1 i/h Joseph F.
Caponio and Janet Geffner, Does Privatization Affect Access to Government
Information?, 5 Government Information Quarterly 147 (1988) (stating that privatization
of government functions can be an efficient management tool if used properly).
2See, e.g., Matthew Bunker and Charles Davis, Privatized Government Functions and
Freedom of Information: Public Accountability in an Age of Private Governance 1-6
(1997) (unpublished manuscript accepted for publication to the 1997 AEJMC National
Convention in Chicago).
3See, e.g., Forsham v. Harris, 445 U.S. 169 (1980); Washington Research Project, Inc.
v. Department of Health, Education and Welfare, 504 F.2d 238 (D.C. Cir. 1974).
4It should be noted that state access laws are given several different names, such as
freedom of information laws, public disclosure laws, public access laws, and public
records acts. This thesis will refer to each state's act by its proper name when discussing
As privatization of government services continues into the late 1990s, some
commentators worry that the desire for government efficiency will cause information that
is important to the public to become shrouded in secrecy.5 They have called on
legislatures to come forward to amend their respective access statutes to ensure that
private entities performing government functions are subject to access. Without this
legislative action, they say the only alternative is for courts to protect the public's right to
know through expansive interpretations of the existing federal and state statutes.6
Otherwise, as governments continue to create contracts with private entities to handle
important services into the new millennium,7 citizens will find it very difficult to obtain
information related to the government because these private entities often do not fall
under the definitions of "agency" under state statutes and the Freedom of Information
5See, e.g., Comments of Jane Kirtley, Executive Director of the Reporters Committee
for Freedom of the Press, regarding increasing privatization of public hospitals, in
Minutes of Access and Privatization Workshop, Florida Sunshine Summit, University of
Florida, Oct. 17, 1997; Bunker and Davis, supra note 2; Casarez, supra note 1
(suggesting that private prison operators must be held accountable, and thus the public
must have access to private prison records).
6See generally Bunker and Davis, supra note 2; Casarez, supra note 1.
7These services include prison operations, the National Aeronautics and Space
Association (NASA), fire protection, health care, medical research, railroads and
gambling operations.. See Casarez, supra note 1; Bunker and Davis, supra note 2; Gary
Bowman, et al., PRIVATIZING CORRECTIONAL INSTITUTIONS (1993); Caponio
and Geffner, supra note 1, at 150; Forsham v. Harris, 445 U.S. 169 (1980); Railway
Labor Executives' Assoc. v. Consolidated Rail Corp., 580 F.Supp. 777 (D.C. Dist. 1984);
City of Dubuque v. Dubuque Racing Association, 420 N.W.2d 450 (Iowa 1988).
8See, e.g., 5 U.S.C. Sect. 552(f) (1997); see also Casarez, supra note 1; Bunker and
Davis, supra note 2; Katherine A. Meyer and Allan R. Adler, Agency, in Allan R. Adler,
Even in the absence of legislative action, however, many courts are providing
expansive definitions of "agency" and "agency records" that allow citizens to retain their
rights to government information. Most state courts have engaged in statutory
interpretation that allows the access statutes to reach private entities, at least in some
circumstances. Courts have done this by interpreting the definitions of "agency" and
"agency records" under their respective state statutes to mean more than just records
exclusively in the hands of traditional government entities. To varying degrees, most
state courts have allowed government access statutes to reach private entities that are
entangled or somehow connected to the government.9 The differences among these
courts depend on to what extent a private entity needs to be connected with the
government in order to be considered a public entity.
In contrast, federal courts have consistently held that private entities are not included
within the Freedom of Information Act's definition of "agency." And in most cases,
courts have held that these entities do not hold records with a sufficient nexus to
government to qualify as "agency records" under judicial analysis.10 Thus, federal
LITIGATION UNDER THE FEDERAL OPEN GOVERNMENT LAWS 190-193
9See, e.g., Robert Rivas and J. Allison DeFoor II, When is a 'Private' Document a
Public Record?, 67 Fla. B.J. 52 (Dec. 1993) (discussing Florida cases in which private
entities were subject to the Florida Public Records Act because of their connections to
government). A comparison of the various state court approaches in found in Chapter IV,
"1See, e.g., Forsham v. Harris, 445 U.S. 169 (1980); Washington Research Project, Inc.
v. Department of Health, Education and Welfare, 504 F.2d 238 (D.C. Cir. 1974); see also
Ann H. Wion, The Definition of "Agency Records" Under the Freedom of Information
Act, 31 Stan. L. Rev. 1093 (July 1979). The term "agency record" is not defined under
government privatization can have an even stronger impact on access to important
information than similar state privatization, as information becomes secret once it is
farmed out to private entities.
In short, although the subtle meanings of statutory terms are often difficult to interpret,
the interpretation has a major impact on what information can be obtained by the press
and the public, particularly in an age of continued federal and state governmental
privatization.11 As the desire for efficiency leads to continued privatization,12 it has been
argued that expansive interpretations of statutory terms are vital, otherwise the public will
be increasingly shut out of operations that have traditionally been open the public.13
the Freedom of Information Act, therefore the federal courts have created their own
definition. See Chapter II, infra.
"States such as Texas, Florida, and Indiana have increasingly and aggressively
pursued privatization initiatives in order to save taxpayers' money and to provide more
efficient services. However, these initiatives have not always been widely successful.
See, e.g., John Thompson, Texas Faces Problem With Private Prisons, Dallas Morning
News, Nov. 6, 1996, at 19A; Sunshine Law is Going to the Dogs, Some Say, Proposal
Would Privatize Rabies Vaccination Records, Florida Today, March 19, 1996, at 6B;
Comments of Kyle Neiderpreum, Freedom of Information Chair, Society of Professional
Journalists, regarding privatization in Indianapolis, in Minutes of Access and
Privatization Workshop, Florida Sunshine Summit, University of Florida, Oct. 17, 1997.
Federal privatization efforts, which have also experienced problems, include private
prison operations and agencies such as NASA. See Casarez, supra note 1; Caponio and
Geffner, supra note 1.
12See Bill Theobald, Public May Soon Get Access Boost, The Indianapolis Star, June
28, 1997, at B6; see also comments of Jane Kirtley, Executive Director of the Reporters
Committee for Freedom of the Press, regarding increasing privatization of public
hospitals, in Minutes of Access and Privatization Workshop, Florida Sunshine Summit,
University of Florida, Oct. 17, 1997.
13See Bunker and Davis, supra note 2, at 5.
This thesis will provide an exhaustive analysis of the issues involved in government
privatization of services and its effects on freedom of information. The chapters will
discuss the following areas: I) the perceived pros and cons of privatization (including a
discussion of its importance because of its effects on access to government information);
II) an analysis of how federal courts have determined when the Freedom of Information
Act should apply to private entities; Ill) an analysis of the access statutes in all 50 states,
including how these statutes define terms such as "agency" and "agency records;" IV) an
analysis of how state courts have applied their respective access laws to determine when
private entities should nonetheless be subject to access; and V) an analysis and
comparison of both the state and federal approaches, including a suggestion for a unified
approach that would better square with the spirit of public access. In short, the goal of
this thesis will be to thoroughly analyze all state and federal statutes and case law dealing
with what qualifies as an "agency" and an "agency record" under the various laws in light
of the policies behind public access to information.
By analyzing and comparing the different judicial approaches to statutory
interpretation that determine when private entities should be subject to access laws, this
thesis provides an extensive review of all the case law applying existing access laws to
government privatization efforts. Unlike the previous literature discussing the pros and
cons of privatization, this thesis will have a much broader approach by discussing all
federal and state law, and then comparing the two to determine whether the right of
public access is being restricted all around the country due to privatization efforts. In the
end, this thesis does something that has never been done in the previous literature
discussed in subsection B of this chapter it suggests a single approach to defining
"agency" and "agency records" that would better square with the spirit of public access on
both the state and federal level.
First, subsection B of this chapter will report the previous literature regarding the pros
and cons of privatization, its prevalence, and its potential effects on freedom of
information. The commentators in this chapter are essentially divided into those who
oppose privatization and those who favor it. After discussing this debate, the thesis will
move into Chapters II, III, and IV, discussing the relevant federal and state statutes and
case law applying the various definitions of "agency" and "agency records" to private
entities. At the beginning of Chapter II, the specific language of the Freedom of
Information Act will be discussed, before moving into the relevant federal case law
construing the Act. Similarly, the access laws of all fifty states are treated in Chapter III
before Chapter IV provides an analysis and categorization of relevant state case law. By
the conclusion of the thesis, the reader will have the most exhaustive discussion and
analysis of privatization, including an approach for the future, that has ever been provided
in scholarly literature. The reader will get a clear indication of the impact of privatization
on the public's right to know through the answers to three research questions:
1. Are private entities considered "agencies," and their records considered "agency
records," thus making them subject to the Freedom of Information Act under
federal case law when the government contracts with or relies on the services of a
2. Are private entities covered under the freedom of information statutes of the fifty
3. What are the various approaches taken by the state courts in determining when a
private entity should nonetheless be considered a public agency under access
statutes, and its records considered public records?
In Chapter V, this thesis will provide a comparative analysis of how state and federal
laws have dealt with both outright denials and limitations on access to information in the
hands of entities contracting with the government, but no longer technically in the
government agency's possession. Lastly, the author will suggest approaches to
determining whether access laws apply to private entities that best fit within the spirit and
purpose of the laws. This overall comparative analysis and suggestion for a unified future
approach has never been completely handled in the previous literature. The complete
discussion and analysis provided in this thesis is therefore important to show how
privatization can have varying effects on access to information depending on where in the
United States one is located, and what state or federal statute governs. It is also important
to foster the continuing debate over privatization by suggesting how access to information
can be protected in the future. The goal of this thesis is to encourage legislatures to
rethink their statutory definitions, instead of relying on the judicial system to protect
freedom of information by stretching the existing definitions of "agency" and "agency
While privatization may have obvious benefits,14 this thesis suggests that before
governments jump into privatization in the name of efficiency, journalists, their legal
counsel, and the public need to be aware of the impact privatization will have on public
participation in the process of government. This impact includes a reduction in the
public's access to information. Unless this reduction can be prevented, this thesis will
show how one essential attribute for adequate participation in the democratic process will
be lost in the future as public services continue to be performed by private entities.
The legal materials used in this thesis were found by running both computer searches
on Westlaw and Lexis and through key-word searches in legal texts.15 By using several
different methods to find these resources (including legal encyclopedias and American
Law Reports), the author can be confident that a thorough search has been performed to
locate the relevant primary sources. Law reviews, treatises, journal articles, news articles
and books were also located in addition to cases and statutes, in order to show
commentator reaction to the various approaches taken by state and federal courts.
14See, e.g., Phil Hampton, Privatization Has Value, Candidates Agree, The Press-
Enterprise (Riverside, Calif), March 11, 1996, at Bl.
15One uniform search was performed on the computer databases, which included key
terms related to Freedom of Information and privatization. The precise search used was
as follows: "FREEDOM OF INFORMATION ACCESS GOVERNMENT
DOCUMENTS PRIVATE COMPANIES ENTITIES AGENCY PUBLIC PRIVATE
RECORD JAIL HOSPITAL SERVICES CONTRACTING OUT FOIA." The computer
searches operate so that the twenty best cases are chosen based on the frequency of the
terms above. From this general search, a wealth of cases were found. Cases within the
principal cases were also checked for relevancy, and all cases were Shepardized in an
attempt to find other relevant case law. However, the author admits that it is possible that
some cases were missed due to the infrequency of the terms above or the use of other
terms in place of the ones included above.
The next subsection is meant both as an overview of the issues surrounding
privatization, including its potential effects on freedom of information, and as a review of
the literature discussing the pros and cons of privatization. Also included within this
chapter are several examples of both state and federal privatization, showing that
privatization is a continuing trend into the late 1990s. As a literature review, subsection
B divides the commentators into those who oppose privatization and those who favor it.
As a whole, subsection B will show how the previous literature has narrowly focused on
various aspects of state or federal privatization, such as the trend toward prison
privatization. This thesis will go beyond the previous literature by providing a more
extensive analysis of all access laws and their application to many instances of
privatization. It will also compare the federal and state laws, and suggest ways to protect
the public's right of access.
Privatization and the Debate Over Public Accountability
Privatization can take many forms. For example, the federal government has engaged
in "load shedding" by cutting social welfare programs and allowing private for-profit or
not-for-profit companies to take over the programs.16 Government subsidies have also
allowed consumers to choose private entities through a voucher system, such as in the
federal food stamp program.17 The majority of cases litigated over access to documents
held by private entities, however, involve the most common form of privatization, called
16See Bunker and Davis, supra note 2, at 7.
contracting out." This happens when the government contracts with a private entity to
provide a service previously performed by the government, or to provide a service for or
on behalf of a government entity.19 Services such as fire protection, transportation,
prisons, and health care have been delegated to private entities, and litigation over access
to records once in the hands of government has erupted. The debate over the value of
privatization and its impact on public access laws has thus centered around this
contracting out form of privatization.
As governments continue to have difficult financial times,20 contracting out has
become a widely considered option. But its success has been mixed. For example, in
1996 the City of Indianapolis contracted with Ameritech to provide on-line computer
access to many of the city's public records. Although both the city and the private
company were initially enthusiastic about more efficient and inexpensive access to
records,21 in the end the venture proved unsuccessful. This was primarily because
Ameritech found that the cost of the venture was too high, and the result was that the cost
of access was higher than if the records had been in the hands of government.22
"See Mays, supra note 1, at 43.
19See id; see also Bunker and Davis, supra note 2, at 7-8.
20See Mays, supra note 1, at 42, for a discussion of the financial woes of city
21See Jim Beck, City Gets Proposal to Put Records Online, The Evansville Courier,
March 28, 1996, at 5A.
22See Rebecca Buckman, Computer Retrieval of Records A New Wrinkle in Public
Access, The Indianapolis Star, March 21, 1994, at Al; see also comments of Kyle
Neiderpreum, Freedom of Information Chair, Society of Professional Journalists, in
Consequently, Ameritech announced in June 1997 that it would not seek to renew its
contract with the city, which expires in 1998.23 Even so, the city still plans to seek out
another contractor to provide the on-line services because officials believe that
privatization can be more efficient if done in an effective way.24
Privatization has continued even in the face of difficulties, because governments still
believe there are benefits of contracting out certain government services.25 In Arizona
and Texas, governments have found it financially beneficial to contract out the operation
Minutes of the Access and Privatization Workshop at the Florida Sunshine Summit,
University of Florida, October 17, 1997. Ms. Neiderpreum, who is also a reporter for the
Indianapolis Star, said Ameritech wanted the records to be a money-maker, but the on-
line service fell below its expectations. Additionally, concern was expressed over the
rising price of the records under the new system, even though their value was enhanced
by computer access. See id; see also Buckman, at Al; Beck, supra note 21.
23See Theobald, supra note 12.
24See id. In fact, the city is looking to expand the on-line services to include property
tax records, land records and traffic accident reports. See id. The government still
perceives that computer access could be more efficient and cost-effective, as well as more
convenient for the public, even though Ameritech wanted to focus its efforts in "a
different direction." See id.
25In fact, one of the most ardent supporters of access to government information has
also acknowledged that privatization can possibly be a success, because it could make
access to government information cheaper and more efficient. See Comments of Jane
Kirtley, Executive Director, Reporters Committee for Freedom of the Press, in Minutes of
Access and Privatization Workshop, Florida Sunshine Summit, University of Florida,
October 17, 1997. However, Ms. Kirtley stressed that there needs to be a mechanism to
ensure that the public can get records at the same or lower cost than when the government
was the vendor, which did not happen in the Indianapolis case. See id; see also Louis
Uchitelle, Competition Called Key to Success With Privatization, The Orange County
(Cal.) Register, April 26, 1988, at A6 (stating that competition can force private agencies
to bid for services at a lower cost than the government could provide them, helping to
realize the efficiency and cost benefits of privatization); Thompson, supra note 11
(discussing the problems Texas has faced in its efforts to privatize the state's prisons).
of state prisons.26 In Ohio, independent firms have screened candidates for big-city police
jobs, shielding the process from public access because the firms claim they do not fall
under the state access law.27 And in Illinois, the state press association fought to maintain
public access when a judge allowed Ameritech and its CivicLink project to manage all
court records.28 Members of the media from all over the state of Illinois became alarmed
when they realized that CivicLink would have exclusive control of all the records within
the first 72 hours of existence before disbursing the information to the public.29 Similar
to Ameritech's efforts to manage documents in Indianapolis and Prince George's County,
Maryland, members of the Illinois press worried that this privatization effort would
effectively cut off their rights to free and open access.
Even though it has detractors, advocates of privatization have continued to push for
more contracting out to private entities. Since the late 1980s, privatization of services has
continued as governments have sought to provide services in a more efficient and cost-
effective manner. But today advocates of privatization are stressing careful study as a
26See Barbara Croll Fought, Privatization Threatens Access, QUILL, September 1997,
28See id. These records were to be stored electronically and then sold to the public for
profit. See id.
29See id. The media successfully won their battle with CivicLink when county clerks
began refusing to sign exclusive contracts with CivicLink, and eventually legislators
banned future contracts. See id. at 9. As of late 1997, Ameritech said it was abandoning
the CivicLink project, and other large information venders belonging to the Information
Industry Association have stated that they do not favor exclusive arrangements to manage
information. See id.
way to determine which services would benefit from privatization from a cost-benefit
perspective.30 In many cases, governments have even engaged in "bidding wars" with
private companies to see who could provide services, such as incarceration and public
safety, in the most economic manner.31
30See Hampton, supra note 14. Hampton points out that in 1996, candidates for county
supervisor in Riverside County, Calif. were pushing for continued privatization efforts,
mentioning hospitals, jails, data processing, bill collecting and street sweeping as viable
areas for privatization. See id Even though research into the viable areas has been
stressed, advocates of privatization have still called for "massive privatization" because it
is still perceived that the profit-motivated private sector is more efficient than the
government. See id. The article points out that millions of dollars have been "farmed
out" to private companies to provide services, and entire units of government have been
shut down with a resulting loss of government jobs. See id. This shows that privatization
is not a fading remnant of the Reagan era, but is still going strong as governments seek to
rid themselves of public costs.
In fact, in the age of the Information Superhighway, local governments say they simply
cannot afford to put public records online and manage them, and therefore private
vendors are a necessity. See Croll Fought, supra note 26, at 10. These privatization
efforts have a profit goal as well, with one government in California bringing in as much
as $400,000 per year. See id. Such private computerized databases are resulting in higher
fees for public access, as the public now has to pay for the computers, software, upgrades
and employee time in accessing the records. See id. This has led members of the media
to stress that while putting records online is applaudable because of easier access, it
should not result in higher fees for records that have already been paid for by the public.
31See Uchitelle, supra note 25. Jails and fire departments in Florida and garbage
collection in Phoenix, Ariz. were used as examples of services where the government and
private companies were bidding "against" each other to provide the services. See id.
Although the government made the ultimate decision regarding who would provide the
services, the purpose was to allow private companies to bid "against" the cost of
government services to see who could provide the services at less expense. In some cases
the government won the "bidding war," but in others, such as a jail in Bay County, Fla., a
private corporation took over the services. See id. The article also gives examples such
as data processing, vehicle maintenance, street-light repair, hospital management and
park maintenance as services that have come under increasing privatization. See id. In
short, the article makes clear that during the Reagan years, competition was seen as the
key to more cost-efficient services, as governments were wasting money and finishing
fiscal years in the red. See id. However, lost in the debate was the effects on access to
Government studies into the benefits of privatization have continued into the 1990s as
cities experience financial troubles. For example, the city of Atlanta, even with its
expected revenue from the Olympics, still considered the prospects of privatization of
certain government services in 1996 in the face of what was called poor fiscal health.32 In
Miami, the city did not want to lose the revenue generated by the Miami Heat basketball
team, so it recruited a local businessman to facilitate talks with the team, a move that had
the effect of shielding the entire deal-making process from the public.33 The businessman
refused to grant public access to the records dealing directly with public funds and bonds,
even though such records would have been obviously public had the government itself
conducted the negotiations. Thus, the Miami government was able to "privatize" a
specific service, even though it related directly to taxpayer money. In so doing, whether
intentionally or unintentionally, the city momentarily shielded a deal-making process
involving public money from public scrutiny. Other areas in Florida and Indiana have
records previously held by the government.
32See Darryl Fears, Suggestion of Privatization Opens Up Gap in Atlanta Politics, The
Atlanta Journal and Constitution, May 4, 1996, at 2C. In fact, a Rotary Club member said
the city's fiscal health was so bad that it was nearing "financial collapse." See id. The
member "offered a simple solution for what he called the city's bloated employee rolls
and sluggish services: contract work currently handled by the city to private companies."
33See Croll Fought, supra note 26, at 11. The businessman who refused to grant the
media access to the records surrounding the new deal with the Miami Heat was P.
Anthony Ridder, chairman and CEO of Knight-Ridder, Inc., owner of the Miami Herald..
See id The Miami New Times eventually sued Ridder and obtained access to the files,
but the judge never ruled on whether Ridder was subject to public access, even though he
was helping to create records relating directly to public business. See id.
also contracted out services in the face of fiscal difficulties, providing further evidence
that privatization is far from a dying phenomenon.34
Although governments have continued to seek efficiency through privatization, its
detractors have remained equally vocal into the 1990s. Some commentators have argued
that privatization can negatively impact the public sector and public employees because
unlike government, private entities are not subject to constitutional limitations such as
due process, and they are not required to provide the benefits of civil service employment
as private employers.35 Government employees are also arguing against privatization
because they say it hurts their ability to negotiate employment contracts once civil service
requirements are removed, limits the government's oversight of vital public programs,
34For example, again in Indianapolis, a health department performing a supplemental
government function has been formed by the city as a private agency with no substantial
links to the government. See Comments of Kyle Neiderpreum, Freedom of Information
Chair, Society of Professional Journalists, in Minutes of Access and Privatization
Workshop, Florida Sunshine Summit, University of Florida, Oct. 17, 1997; see also Croll
Fought, supra note 26, for further discussion of privatization examples, which include
Florida jails, health care, and other services.
Additionally, Dade County, Fla. has developed a Geographic Information System
(GIS) digital land database and then allowed Florida Power and Light to acquire the
copyright, giving the county half the profits. See Croll Fought, supra note 26, at 9. One
editor in Miami fears that this exclusive copyright can prevent public access to the
mapping information. See Comments of Dan Keating, Research and Technology Editor,
The Miami Herald, at Florida Sunshine Summit, University of Florida, Oct. 17, 1997.
Joel Campbell, Vice President of the National Freedom of Information Coalition, has
pointed out that private companies are selling government records in states such as Utah,
Idaho and Nevada, just to name a few states. See id; see also Croll Fought, supra note
35See, e.g., Mays, supra note 1, at 45. The author cites the Fourteenth Amendment due
process requirements as an example. See id. In order for such requirements to apply to
the private sector, the author points out that a "state action" approach must be applied by
the courts. See id. at 45-46.
causes layoffs, and in fact insults the quality of their work by suggesting they are not
Others have suggested that the privatization process carries risks of contract disputes
between the government, the private contractor, and employees over the requirements of
providing a public service. They say privatization could increase public fees because
governments often pay private companies high prices to provide services and then pass
the costs on to the public, privatization hurts the status of labor unions that previously
bargained for government civil service positions, it reduces wages and benefits, and it
could result in government payoffs while divesting the taxpayer of former benefits such
as oversight.37 Still others claim the cost savings to the public are minimal at best, and
are outweighed by the loss in oversight that goes with privatization.38 Minorities and
elderly citizens also worry that the private sector will neglect the increasing commitment
government has shown in recent years to hiring African Americans and retaining older
36See Hampton, supra note 14.
37See Elizabeth Moore, Doling Out Services, Newsday, April 15, 1996, at C1. This
article provides a good overview of privatization issues and examples in New York State
and the rest of the nation, actually calling privatization a "worldwide movement." See id.
Recent successes and failures of privatization are highlighted, as are Gov. George
Pataki's plans to employ a commission to study the pros and cons of privatizing certain
services in an effort to "farm out" more state services and buildings to private entities.
See id; see also Hampton, supra note 14; Uchitelle, supra note 25.
8See id; see also Hampton, supra note 14 (suggesting that some privatized services
are still too risky when compared to minimal benefits gained. An example given is law
enforcement, because private law enforcement providers would not be subject to
constitutional guarantees such as due process and equal protection, causing risks of
abuses that could affect public health and welfare).
Americans.39 Lastly, privatization critics worry that private companies will become
wasteful and ineffective in providing important public services.40
One frequently debated area of federal privatization has been the trend toward
privatized federal prisons. Commentators have stated that due to the rising costs of
housing inmates and persistent overcrowding, the government has increasingly turned to
private prison operators since the Reagan era of the 1980s.41 In fact, as of 1996, as many
39See Uchitelle, supra note 25; see also Fears, supra note 32. The government's
commitment to hiring African Americans and retaining older and disabled Americans is
evident in its affirmative action and Americans With Disabilities Act programs. See, e.g.,
Uchitelle, supra note 25.
40See Fears, supra note 32. It has been suggested that this wastefulness comes from
increased spending and higher fees charged by private entities. See id. The article cites a
private health care company in Georgia that billed the state for such things as luxury cars
and jets. See id. In an attempt to solve the wastefulness problem, it has been suggested
that contract requirements for annual reports could be built into privatization agreements.
41See, e.g., Warren L. Ratliff, The Due Process Failure of America's Prison
Privatization Statutes, 21 Seton Hall Legis. J. 371, 375 (1997); see also Barbara J.
Auerbach, WORK IN AMERICAN PRISONS: THE PRIVATE SECTOR GETS
INVOLVED (1988) for a discussion of the issues involved in privatization of prisons, as
well as the chronological development of private prison industries in various states.
For further examples of prison privatization on the state level, along with the concerns
that follow such privatization efforts, see also John Thompson, Texas Faces Problem
With Private Prisons, The Dallas Morning News, Nov. 6, 1996, at 19A (stating that
eighteen states have private prisons, with Texas and Florida leading the pack in terms of
the number of private beds); Nick Gillespie, Accountability Ensures Justice for Private
Prisons, Las Vegas Review-Journal, September 29, 1997, at 6B (discussing abuses
committed by prison guards in private prisons that have resulted in government
cancellation of contracts, and stating that these economic consequences can prevent
private companies from providing inadequate services); Commentary, Prison
Privatization is No Panacea, The Hartford Courant, August 24, 1997, at C2 (also
discussing abuse of inmates by guards in private prisons, and how such abuses could not
be discovered absent protection under freedom of information laws); Carol DeMare,
Corrections Chief Pans Private Prisons, The Times Union (Albany, N.Y.), June 3, 1997,
at B 1 (discussing a speech by New York State Corrections Commissioner Glenn Goord in
as 74,000 inmates across the country were housed in private prisons, with an annual
growth rate of more than 30 percent expected in the next several years.42 Since federal
legislation first permitted private federal prisons,43 privatization has led to worries over
the quality of services being offered by private companies, as well as the treatment of
inmates by private operators.44
Commentators worry that private prisons will allow operators to take liberties with
prisoners that would not be allowed by the government.45 Additionally, private prison
operators that are not subject to public oversight could operate against the public's
interest by taking a more relaxed approach with such important functions as security
which he stated that private prison companies are only motivated by profit and that the
state can do an equally cost-effective job of corrections).
42See Ratliff, supra note 41, at 372. In fact, the Clinton Administration has endorsed
federal prison privatization in order to look "tough" on crime and reduce the federal
deficit. See id. at 404-405. At least twenty-one federal private prison facilities currently
exist. See id. at 405.
43See 18 U.S.C.A. Sect. 4013 (West 1996).
45See Ratliff, supra note 41, at 373; see also Ira P. Robbins, THE LEGAL
DIMENSIONS OF PRIVATE INCARCERATION 2-5 (1988) (arguing that districts
contemplating privatization of prisons must ensure that private operators are accountable
to the public to protect the contractual and constitutional rights of the public and of
inmates). Robbins detailed a study by the American Bar Association that was conducted
as a follow-up to the Association's 1986 recommendation that "jurisdictions that are
considering the privatization of prisons and jails not proceed ... until the complex
constitutional, statutory, and contractual issues are satisfactorily developed and resolved."
See id. at 6. The study focused on the successes and failures of prisons that were operated
privately, including disputes and fiscal aspects of privatization.
measures.46 This is because private prison operators have a financial motive, and may
sacrifice individual constitutional rights and prison quality in the name of the profits to be
gained by saving on management and maintaining a full house.47 And because of the
strong public demand to stay tough on crime, governments are likely to feel the pressure
to expand prison space while at the same time experiencing the difficulty and expense of
day-to-day monitoring, thus overlooking the deficiencies of private prison operators.48
Another example of federal privatization that could have a direct impact on the
public's ability to obtain important information involves the United States Department of
Energy. In January 1998, the DOE, the federal government's largest contracting agency,
refused to allow nuclear watchdog groups access to documents in the hands of its
46An Associated Press writer, writing on the 30th Anniversary of the Freedom of
Information Act in 1997, commented that private entities performing public functions
need to understand that their work will be subject to public scrutiny. See Bob Rivard, 30th
Anniversary of Freedom of Information Act Is Time to Reflect, The Associated Press
Political Service, May 12, 1997. Rivard pointed out that the escape of a convicted
murderer from a privately-run prison in San Antonio in 1996 is an example of a security
breach that "would not be tolerated" if the prison was run by the government. See id.
47See Ratliff, supra note 41, at 373-374; see also McKnight v. Rees, 88 F.3d 417, 424
(6th Cir. 1996) ("While privately employed correctional officers are serving the public
interest by maintaining a correctional facility, they are not principally motivated by a
desire to further the interests of the public at large."); Robbins, supra note 40, at 6-7;
Rivard, supra note 46.
48See Ratliff, supra note 41, at 379-380. In fact, Ratliff argues that the financial
interest of private prison operators makes privatization potentially unconstitutional
because due process requires the disinterest of the person affecting prisoners' liberties,
and private operators cannot remove their own profit interests from the decisions that
affect prisoners' constitutional rights. See id. at 385-386. At the very least, close scrutiny
of private prison operations is essential, and enabling statutes should closely proscribe the
condict of private operators. See id. at 393, 398-399.
contractors.49 The result of the pending litigation in the DOE's case could have a major
impact on government accountability for its energy operations because the DOE contracts
with private corporations and nonprofit organizations to perform more than 80 percent of
its work.50 Records such as travel expenses of top officials and financial aspects of new
weapons projects, which are important for public oversight of how its money is spent,
could be closed off to the public if the federal courts in New Mexico and California do
not hold the contractors accountable as "agencies" holding "agency records" under the
Freedom of Information Act.51 Under the prevailing federal case law, this is not likely to
happen,52 and the federal courts are not likely to force the DOE to comply with its own
guidelines requiring access to certain documents because unlike the FOIA, the guidelines
are not law.53
Additionally, the federal government has considered removing impediments to
privatizing airports. In 1996, the federal government began discussing the possibility of
49See Ian Hoffman, DOE Fights Records 'Release, The Albuquerque Journal, January
23, 1998, at Al.
"See id. In fact, even when it does not directly refuse to release records, nuclear
watchdog groups say the DOE delays the release of records for months or years. See id.
The DOE's attorney says other federal agencies, such as the Department of Defense and
NASA, also engage in delays even though they have guidelines for the release of
information. See id. The DOE's guidelines, according to the lawyer, go beyond what the
FOIA requires. See id.
52See infra, Chapter II.
5See Hoffman, supra note 49.
easing the federal restrictions on privatizing state or municipality-run airports.54 By
easing the sale of airports to private companies under federal regulations, a large amount
of important aviation information could be put out of the public's reach. All of these
federal examples of privatization highlight the need for a broader application of the
Freedom of Information Act by applying it to entities performing important public
functions and controlling important public information.
Although the debate over privatization has raged since the 1980s, only in the 1990s
has literature surfaced expressing the concern of media professionals over the potential
impact on the watchdog role of the media and the public's right to know.55 Members of
the media have realized that because most freedom of information laws do not explicitly
reach private entities, the entities may refuse to release previously public information, or
at least force the media or the public to engage in costly litigation.56 Therefore,
journalists have argued that while privatization may be desirable in itself, it should not
come without statutory or contractual provisions leaving public accountability intact.57
54See Martha M. Canan, House Committee Looks at Paving Way for Easier Airport
Sales, The Bond Buyer, March 8, 1996, at 7.
55See, e.g., Elliot Krieger, Privatization Raises Issues ofAccountability, The
Providence Journal-Bulletin, December 15, 1996, at 7B.
56See id. The article discusses the refusal by private companies to release information
because they do not believe they are accountable under access laws, citing the welfare
system in Wisconsin as an example of a private company that has been determining
eligibility standards in that state: "'And if you don't know what those standards are, if you
have no access to the welfare requirements, then how can you monitor?"' Id.
57See, e.g., Staff editorial, Private Contracts: Arbrister Bill Would Provide Greater
Public Access, The Houston Chronicle, April 29, 1997, at A18. In effect, the editorial
argues that a proposed bill making contracts between private companies and public
They emphasize that the public should be able to monitor the private company's activities,
and this monitoring should be at the same level of openness as when the government was
the information vendor.58 In other words, contracting out should not change the free flow
of information under the various access laws.
In short, members of the media have claimed that fears of substantially reduced access
to information from private companies have become a reality, thus state legislatures need
to solve the problem.59 In recent years, journalists have been more vocal in their fears
agencies totaling at least $1 million a matter of public record does not go far enough -- it
should include all contracts. See id. Other members of the media have agreed that
sophisticated checks need to be in place in order to ensure that the interests of the
government are being effectively monitored. See, e.g., Croll Fought, supra note 26, at 11.
"See, e.g., David Poulson, Bill May Ensure Public Tracks Spending, The Grand
Rapids (Mich.) Press, April 1, 1996, at C3. One example the article gives is the rise in
fees for copying documents once the copying is done by private entities. Although fees
for copying are allowed under the access statute, government agencies frequently waived
fees for small copying jobs. However, the article states that one private company has
charged a $17 an hour labor fee for small jobs. See id. Fearing that private companies
waste money and raise costs, one Michigan legislator has proposed a bill that would allow
the public to track private spending under public contracts. See id.
Some journalists have suggested that the most logical way to ensure public oversight
of private companies would be to include a provision in contracts that requires the private
company to abide by the state's freedom of information laws. See Rebecca Buckman,
Computer Retrieval of Records A New Wrinkle in Public Access, The Indianapolis Star,
March 21, 1994, at Al. Additionally, the public could gain access to such records by
requesting that the public body obtain them from the private entity on its own behalf, a
right most governments retain when they contract out services. See id. However, even
given the logic of these suggestions, journalists are quick to point out that the government
has nonetheless failed to monitor private firms on behalf of the public. See id. (citing as
an example the City of Indianapolis' failure to monitor a private company that took over
operation of municipal golf course, which did not seek public bids for construction
59See, e.g., Staff editorial, supra note 57; Don Noel, Privatization hinil, hi't Reduce
Public Information, The Hartford (Conn.) Courant, April 16, 1997, at Al l.
that "once-public information has disappeared behind the curtain of corporate secrecy" as
governments continue their privatization efforts.60 In addition to examples such as the
Ameritech's CivicLink project and the Miami Heat negotiations, journalists say they have
been denied access to tax collection records, prisons, health data, and even the states'
compiled laws.61 And journalists say to make matters worse, private entities have been
successful in gaining legislative exemptions from open records laws in order to nullify
any claim that the public should have access.62 Therefore, the media itself has been active
in seeking support for public access bills in states such as Michigan, Texas and
Connecticut to ensure that records in the hands of private entities remain open to the
Scholarly writers have also contributed literature discussing the potential impact
privatization can have on access to government information. For example, associate
professor Matthew Bunker and assistant professor Charles Davis have pointed out that by
creating, maintaining, and controlling previously public records, private companies are
60See, e.g., Noel, supra note 59.
61See id. The article again stresses that the more common trend is not to deny total
access to information, but to instead allow access at a "marked-up price." See id.
62See, e.g., Sunshine Law is Going to the Dogs, Florida Today, March 19, 1996, at 6B
(discussing the successful move by veterinarians to seal off pet records from the public --
these records would otherwise be public when held by public agencies such as the
63See Staff Editorial, supra note 57; Poulson, supra note 58; Noel, supra note 59.
Stating that public scrutiny is necessary to ensure that the process remains on the "straight
and narrow," Noel discusses a Connecticut bill containing provisions to ensure that
information in private hands remains open to the public. See Noel, supra note 59.
controlling access and are often "at odds with the very purpose of public records laws"
because instead of favoring openness, the companies are likely to favor only the bottom
line.64 Citing several disputes between private companies and the public seeking records,
the authors state that such disputes are likely to be a growing trend in the future.65
Bunker and Davis state that because legislatures have failed to ensure the statutes will
apply to private entities holding government information, it is up to the courts to clarify
the role of public records laws in the face of growing privatization.66 In fact, the authors
point out that privatization is continuing not only across many types of services, but
within specific public institutions as they privatize food services, health care, and other
services.67 In other words, privatization can include both a complete abandonment of a
service by the government, or a simple contract to perform a part of the service.
Although Bunker and Davis acknowledge the benefits of privatization (efficiency and
competition), they suggest that courts should keep public access to records in tact through
a "public function" approach, where private entities performing a formerly public
function remain accountable under the freedom of information laws.68 This approach is
64See Bunker and Davis, supra note 2, at 5.
66See id. at 6.
68See Bunker and Davis, supra note 2, at 24 (if an entity is performing a function likely
to affect the public would be accountable under access laws); see also Comments of Jane
Kirtley, Executive Director, Reporters Committee for Freedom of the Press, in Minutes of
the Access and Privatization Workshop, Florida Sunshine Summit, University of Florida,
Oct. 17, 1997 (stating that privatization can work if it is ensured that private companies
also suggested in this thesis, but although Bunker and Davis discuss the inadequacy of
statutes and court decisions in dealing with privatization, this thesis will provide an
exhaustive and detailed analysis of all 50 states and the federal system to analyze the
potential impact of the various approaches on freedom of information.
Another example of previous literature discussing the negative impact of privatization
on freedom of information was a 1995 article by communications law scholar Nicole
Casarez. Casarez argued that federal private prison operators should at least be as
accountable as government officials, to ensure contractors "in no way abuse the public
trust or prisoners' rights."69 She says the Freedom of Information Act, which does not
include private operators within its definition of "agency," should be applied to private
contractors to help accomplish this oversight.70 While private prison operators are subject
to marketplace competition that increases service quality,71 Casarez says it is the
government that should retain the ultimate authority over prison operations.72 The public
remain subject to freedom of information laws).
69See Casarez, supra note 1, at 249-250. The author describes various forms of private
prison operator abuse, including forcing prisoners to remain outside for many hours a day
in intense heat. See id.
71See id. at 260-261. Marketplace competition theoretically exists because if a private
operator does not provide a level of quality that is acceptable to the government, it will
lose out on future bids for private operation; thus, the private operator will theoretically
strive to provide the best quality possible. See id.
72See id. at 260-261. This government authority could be achieved if the government
is contractually allowed to oversee private prison operations, demand information from
private prison operators, and replace private operators for insufficient performance. See
can make sure this monitoring is being performed, but only if the Freedom of Information
Act allows access to private contractors.73
Because courts have failed to hold private entities accountable under the FOIA,
Casarez argues that the federal government can frustrate the public disclosure purposes
behind the Act by delegating services to the private sector.74 Therefore, Casarez contends
Congress must step in to enact legislation holding private prison operators accountable, or
to amend the FOIA's definition of "agency" to include entities performing "significant
agency functions."'7 The author of this thesis makes a similar argument by stating that
"agency" and "agency records" should include any entity or records relating to a
government function, even if that function is now being performed by the private sector.
73See id. at 268-269. The author discusses the application of the Act's terms "agency"
and "agency records" as being the ultimate deterrent to public access to information as the
law currently stands. These terms will be discussed in Chapter II of this thesis.
74See id. at 292-293.
75See id. at 293, 296. Essentially, Casarez argues that the federal government cannot
be trusted to monitor private prisons effectively, due to the expense and difficulty
involved. See id. at 295. Additionally, corruption plays a role when government
overseers turn a "blind eye" to abuses because of government payoffs or the promise of
private sector jobs. See id. However, Casarez admits that amending the FOIA to include
any entity performing any sort of "government function" could potentially cause a major
increase in the number of records requests as more records become available, thereby
causing an even greater burden on the already backlogged public agencies. See id.
Additionally, Casarez says that specifically defining the parameters of what constitutes
a government function could be difficult in light of the continuing government trend
toward privatization, implying that some entities could fall through the cracks unless
certain elements are satisfied. See id. at 296-297. Therefore, Casarez focuses solely on
prisons as opposed to the FOIA itself by suggesting that the legislation enabling private
prisons should make private prison records public records available through the
contracting government agency, instead of public records subject to request from the
private operator under the FOIA. See id. at 301.
But although Casarez wrote a detailed article concerning privatization as it relates to
federal prisons, this thesis takes a broader approach in discussing the federal approach in
a variety of contexts, as well as the various approaches taken by all 50 states in their
statutes and court decisions.
Although the Freedom of Information Act and some state laws may not reach private
entities, some commentators have argued that privatization can be an effective tool for the
government, if used properly. Writing for Government Information Quarterly, Joseph
Caponio and Janet Geffner discussed privatization as an economically beneficial
alternative for government, as long as the government promotes access to important
information.76 They state that this can be done through guidelines imposed by
government on its agencies when they choose to privatize.77 In fact, Caponio and Geffner
76See Caponio and Geffner, supra note 1, at 148. The authors do admit, however, that
there is agreement that some important services are "inherently governmental" and should
not be privatized. See id. The authors say these services have not been explicitly
enumerated. See id
In 1986, the first national conference on privatization opportunities was held in New
York City, with a focus on how privatization can increase efficiency without decreasing
services to the public. See Privatization: National Conference to Be Held in June,
Business Wire, Inc., April 22, 1986. At the conference, issues relating to the benefits and
difficulties of privatizing health care, correctional facilities, transportation, waste
management, energy supply, and hazardous waste management were discussed. See id.
Similar to Caponio and Geffner, the members of this conference focused on the benefits
of privatization while stressing the need for control over the process. See id.
77See Caponio and Geffner, supra note 1, at 148. These federal government guidelines
are generally articulated in the Office of Management and Budget's OMB Circular No. A-
130, which solidifies the government's commitment to privatization as an option to
improve efficiency, but stresses the importance of making sure the information continues
to reach its intended audience. See id. at 149. This circular requires that the public have
a "reasonable ability" to acquire the information, and it states that private entities should
not be able to exert "monopolistic controls" in ways that defeat the agencies'
dissemination obligations, such as through unreasonably high prices. See id. at 150.
argue that private entities could provide for more efficient distribution of information
than the government, as long as the entities are subjected to specific guidelines when
contracting for government services.78
According to Caponio and Geffner, guidelines have successfully protected access to
records held by certain government agencies wishing to contract out services in order to
save money for use in other agency operations.79 For example, NASA's Scientific and
Technical Information Facility, which handles the agency's science and technology
information database, has been privately operated for over 35 years in order to reserve
scarce civil service jobs for the agency's scientists and engineers, while at the same time
following guidelines ensuring public access to information.80 NASA's officials claim that
However, while not being specific about the federal government rules, Caponio and
Geffner state that the rules are "not cut and dried, allowing many opportunities for
exercising judgment" on the part of the contracting agencies. See id.
Additionally, in 1989, the Information Industry Association proposed before the House
Subcommittee on Government Information, Justice and Agriculture a policy framework
that would "guarantee continuing public access to government information" in the face of
technological advances and new strategies for managing government information. See
IIA Proposes Policy Framework to Guarantee Continuing Public Access to Government
Information, Information Today, May 1, 1989. The national association for information
providers also testified before the Senate Committee on Rules and Administration in
1997, stating that Congress "must institute policies that ensure citizens can access
government information from a wide variety of public and private sources." See
Testimony of Ronald G. Dunn, Information Industry Association, Hearing on the United
States Government Printing Office, 1997 WL 10571080 (May 8, 1997).
78See Caponio and Geffner, supra note 1, at 149-150. According to Caponio and
Geffner, the federal government should ensure adequate public access and guard against
"monopolistic controls" such as unreasonably high prices for access to information when
it allows its agencies to privatize. See id at 150.
the private contractor has remained responsive to the public due to significant
government oversight and guidelines, and has managed to save the government
significant expense.81 The National Oceanic and Atmospheric Administration, which
collects, manages and disseminates environmental information, has also contracted with
private companies to disseminate its information, subject to guidelines for public access.82
Additionally, the National Technical Information Service and the Federal Libraries, both
federal government agencies, have privatized their information vending services by
contracting with private entities to disseminate public information and add to its value by
creating national databases for easier access.83
Caponio and Geffner state that all of these agencies have been subject to guidelines
from the federal government imposed on them to maximize the government's policy of
cost-effective public access to information.84 Caponio and Geffner state that these
programs have been successful because of government monitoring of services, the value
added to the public information through private improvements to database technology and
cataloging of information, and the efficiency of dissemination once these improvements
are made by the private entity.85 However, Caponio and Geffner do not recognize that if
81See id. at 150-151.
82See id. at 151.
83See id. at 152-153.
84See id at 149.
85See id. However, the authors admit that information can become more expensive if
the private contractor employs more staff than was previously needed by the government
to disseminate the information, or adds value to the information and then raises charges to
the private entities refused to provide access to the information, they would not be subject
to the Freedom of Information Act because they would not be considered "agencies"
holding "agency records."86 Thus, public information could potentially be lost should
public access come in conflict with the goals of the agencies and their private contractors.
While Caponio and Geffner recognize the benefits of privatization and the importance of
maintaining public disclosure, their article is more narrowly focused than this thesis
because it does not discuss the constitutional and statutory issues involved in
privatization and public accountability. Thus, this thesis has a much broader approach
than the previous literature.
In short, the above literature shows that the debate over privatization and freedom of
information has gained prominence as public bodies continue to turn to privatization as
an option in the late 1990s. While come commentators recognize the efficiency of
privatization, others fear a loss of important public access and oversight.87 Commentators
recoup its expenses. See id. at 153. Thus, not all "management problems" can be solved
through privatization. See id. at 154.
86See Chapter HI.
87In the federal arena, the debate over privatization has reached the legal profession in
recent years as West Publishing, a contractor with the federal government for legal
information, has sought to protect its "value added" information from disclosure under
the FOIA. See John J. Oslund, West Publishing, Competitors Spar Over Bill Provision
Over Information, Minneapolis-St. Paul Star-Tribune, Feb. 10, 1995, at Dl. This has
sparked a debate over whether the government should contract with West to maintain
legal documents such as court cases and legislative bill tracking, as West could
essentially hold a monopoly over much legal information. See id. West is likely to win
this debate because the legal documents, which are openly public when held by the
government, are now in the hands of a private entity not covered by the Freedom of
Information Act. See United States Department of Justice v. Tax Analysts, 492 U.S. 136,
144-145 (1989), and Chapter II's discussion of "agency" and "agency records" under the
have argued that without oversight through the free flow of information, important abuses
and constitutional violations could be shielded form public knowledge.88 Some
commentators have argued that the appropriate method for ensuring that private entities
follow legal prescriptions is through a "state action" concept, which effectively makes
private operators responsible as if they were the government.89 One possible consequence
of this extension of the state action doctrine would be to make private entities subject to
the Freedom of Information Act or state access laws as if they are government agencies,
as long as they are performing a public function.90
Whatever the arguments for or against privatization, it is obvious that as long as
legislatures remain silent on the issue, federal and state courts will face decisions that
must balance the arguments and attempt to stay in tune with the spirit of the access laws.
All 50 states and the federal government have enacted access laws, but in most cases their
"See generally Ronald A. Cass, Privatization: Politics, Law and Theory, 71 Marq. L.
Rev. 449, 496-523 (1987) (discussing the legal difficulties that result from privatization,
such as the application of constitutional guarantees to private entities, and the application
of governmental immunity to private entities); Casarez, supra note 1.
89See Bunker and Davis, supra note 2, at 18; Cass, supra note 88, at 505; Casarez,
supra note 1; Mays, supra note 1, at 68-69; Daphne Barak-Erez, A State Action Doctrine
For An Age of Privatization, 45 Syracuse L. Rev. 1169 (1995) (discussing the need to
extend the current state action doctrine to government delegation to private entities in
order to ensure constitutional guarantees to the public such as due process and equal
protection, particularly through requiring private entities performing public services that
are not also provided by the state to abide by constitutional limitations).
90See Bunker and Davis, supra note 2, at 18-24. The adaption of the state action
doctrine would make private entities comply with the FOIA or state laws if they are
considered state actors for other purposes, because they are performing a "public
function." See id. This approach is discussed further in Chapter V of this thesis.
legislatures have not modified their access laws to ensure continued public access when
the government chooses to privatize its operations. Therefore, the courts have had to
tackle the issue in the absence of statutory amendments.
This thesis will go beyond the previous literature discussed above in that it will
provide an exhaustive analysis of the access laws in all 50 states and the federal
government as of late 1998. Instead of focusing on one narrow aspect of privatization,
such as the debate over private prison operation or the operation of certain federal
agencies, this thesis will exhaustively show how freedom of information laws throughout
the United States apply to government privatization efforts as they continue into the new
millennium. Additionally, this thesis both synthesizes and goes beyond the previous
literature by comparing and categorizing the various laws, and by suggesting two
potential approaches that could protect public access in the face of further privatization
In this way, this thesis acts as a call to legislatures to protect public access to important
information. Of course, the difficulties involved in applying old statutory definitions to
new government situations has created ambiguous and complex judicial decisions.
However, unless legislatures act to change or clarify the various statutes, the courts must
attempt to interpret the laws in a way that protects the spirit behind access to information,
a cornerstone of public involvement in the democratic process. The following chapters of
this thesis will show how both the federal and state courts have addressed public access to
PRIVATIZATION AND THE FREEDOM OF INFORMATION ACT
As the federal government began delegating services to administrative agencies at an
increasing rate, the Freedom of Information Act (FOIA) was passed in 1966 as an attempt
to better allow citizens to find out what their government is up to by ensuring that these
agencies would be subject to access.91 Because it was suggested that the Administrative
Procedure Act was actually being used to withhold information from the public,92
Congress amended the APA with the Freedom of Information Act in order to foster a
government founded on a policy of disclosure.93 As commentators and the Act's
legislative history make clear, the Act's purposes are to allow the public to be informed
91See Casarez, supra note 1, at 264-265; see also Harold L. Cross, THE PEOPLE'S
RIGHT TO KNOW: LEGAL ACCESS TO PUBLIC RECORDS AND PROCEDURES
233-234 (1953) (discussing the rise in administrative regulations during this century);
Justin D. Franklin and Robert F. Bouchard, GUIDEBOOK TO THE FREEDOM OF
INFORMATION AND PRIVACY ACTS 13-14 (1997).
The United States Supreme Court has discussed the public's ability to find out what its
government is up to as the central purpose behind the Freedom of Information Act. See
United States Department of Justice v. Reporters Committee for Freedom of the Press,
489 U.S. 749, 773 (1989). This "central purpose" concept was mentioned by Justice
Douglas in an earlier case in which he quoted an author in the New York Review of
Books for the proposition that "a democracy cannot function unless the people are
permitted to know what their government is up to." See EPA v. Mink, 410 U.S. 73, 105
(Douglas, J., dissenting) (quoting from Henry Steele Commager, the New York Review
of Books, Oct. 5, 1972, at 7).
92See Cross, supra note 91, at 227-28; H.R. Rep. No. 1497, 89th Cong., 2d Sess.
(1966); Casarez, supra note 1, at 265; Franklin and Bouchard, supra note 91, at 14.
93See Casarez, supra note 1, at 265.
about public policy, and to protect the public from a secret government, which breeds
mistrust and corruption.94 Communications law scholar Nicole Casarez has narrowed
these goals down to one word: accountability.95 However, an entity will only be
accountable to the public if it can be subjected to public monitoring under the specific
terms of the Act.
In determining whether or not the FOIA applies to private entities delegated functions
by the federal government, federal courts have had to determine whether private entities
can be considered "agencies" under the Act, and whether their records can be considered
"agency records," a term that was not given any meaning until the 1996 amendments to
the Act. Although the federal courts have fashioned their own "official control" test for
determining what is an "agency" and "agency record" under the Freedom of Information
Act, the Act itself and its legislative history provide some guidance in determining
whether private entities doing business with the government will be subject to access.
Under 5 U.S.C. 552(f)(1), the definition of "agency" is expanded beyond the entities
94See id. at 265-266; see also 112 Cong. Rec. 13,660 (1966); Kenneth C. Davis, The
Freedom of Information Act: A Preliminary Analysis, 34 U. Chi. L. Rev. 761 (1967).
95See Casarez, supra note 1, at 266. Casarez and other commentators discussed in the
last two sections of this chapter essentially fear that a lack of oversight due to the failure
of the Freedom of Information Act to reach private entities in its explicit definitions of
"agency" and "agency record," discussed in the next section, poses a real danger to the
public. Unless Congress or the courts create liberal definitions of "agency" and "agency
records" under the Freedom of Information Act, it is possible that simply by filtering
records out of their possession, federal agencies can circumvent the Act and its spirit of
open government. See id.; see also infra at 69-74.
provided for in the Administrative Procedure Act.96 Section 552(f)(1) states that the term
"agency" includes "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."97 This section was added in 1974 when Congress
amended the Freedom of Information Act, broadening the definition of "agency" to
include entities not explicitly mentioned under the Administrative Procedure Act, but
which "perform governmental functions and control information of interest to the
public."98 Once an entity falls under the Act's definition of "agency," the entity must
comply with the FOIA when it adopts a rule or guideline, or when it receives an access
For example, the term "Government controlled corporation" was meant to reach
corporations that receive federal funds and operate under federal control such as the
96See 5 U.S.C. Sect. 551(1) (1997). The APA defines an agency as "any authority of
the Government of the United States, whether or not it is within or subject to review by
another agency." See id. The APA makes "any administrative unit with substantial
independent authority in the exercise of specific functions" an "agency." See Soucie v.
David, 448 F.2d 1067, 1073 (D.C. Cir. 1971). By contrast, the amendments to the APA
that became the Freedom of Information Act gave a more specific meaning to the term
"agency," expanding the definition beyond the "substantial independent authority"
requirement of the APA to include other entities of or controlled by the government. See
5 U.S.C. Sect. 551(1) (1997).
975 U.S.C. 552(f)(1) (1997).
98See H.R. Rep. No. 876, 93d Cong., 2d Sess. 8-9 (1974).
99See 5 U.S.C. Sect. 552(a)(3) (1997); James T. O'Reilly, FEDERAL
INFORMATION DISCLOSURE 2D (1995).
Postal Service and Amtrak.1"' However, the Act's legislative history clearly states that
merely receiving federal funding will not create a "government controlled corporation"
under the Act. The entity must be chartered or controlled by the government.101
Even if the entities are not "agencies," courts could nonetheless require them to
disclose records if there is a sufficient nexus between the records and an agency to create
"agency records." In order to obtain information under the FOIA, the information must
be considered an "agency record."102 But unlike the term "agency," the term "record" was
not defined under the Act until the 1996 amendments known as the Electronic Freedom
of Information Act.103 After the 1996 amendments, Sect. 552(f)(2) defines "record" as
100See H.R. Rep. No. 1380, 93d Cong., 2d Sess. 14-15 (1974); see also Aug. v.
National Railroad Passenger Corporation, 425 F.Supp. 946 (D.D.C. 1976) (Amtrak
subject to disclosure requirements); National Western Life Insurance Company v. United
States, 512 F.Supp. 454 (N.D. Tex. 1980) (Postal Service is a public agency even though
it operates like a private business).
101See id; see also Meyer and Adler, supra note 8, at 191. The Ninth Circuit Court of
Appeals has provided a detailed discussion of the legislative history behind the Act's
definition of "agency." See Irwin Memorial Blood Bank of San Francisco Medical Soc'y
v. American Nat'l Red Cross, 640 F.2d 1051, 1052-1054 (9th Cir. 1981); see also
Casarez, supra note 1, at 268-273.
Congress itself can also choose to define what entities are subject to disclosure
requirements through the enactment of other laws, such as the Rail Passenger Service Act
making Amtrak subject to the provisions of the FOIA. See Aug. 425 F.Supp. at 950.
102See 5 U.S.C. Sect. 552(a)(3) (1997) (referring to the disclosure of "records"); 5
U.S.C. Sect. 552(a)(4)(B) (1997) (referring to "agency records" for enforcement); see
also Forsham, 445 U.S. at 178 ("...Congress contemplated some relationship between an
'agency' and the 'record' requested under the FOIA."); Katherine A. Meyer, Allan R.
Adler and Elaine English, Agency Records, in Allan R. Adler, LITIGATION UNDER
THE FEDERAL OPEN GOVERNMENT LAWS 195-200 (1997).
103See 5 U.S.C. 552(f)(2) (1997). There was originally no legislative history
construing the term "agency record." See Nichols v. United States, 325 F.Supp. 130, 134
(D. Kan. 1971), affd, 460 F.2d 671 (10th Cir.), cert. denied, 409 U.S. 966 (1972). The
including "any information that would be an agency record subject to the requirements of
this section when maintained by an agency in any format, including an electronic
format."104 However, the Act still does not define the term "agency record," and as the
next section will make clear, federal courts agree that the record must be sufficiently
linked to agency operations in order to label the information an "agency record."105
1996 amendments, commonly known as the "Electronic Freedom of Information Act,"
added the definition of "record" and were meant to "provide for public access to
information in an electronic format," but the amendments still did not provide any insight
into what constitutes an "agency record." See Pub. L. No. 104-231, 104th Cong. (Oct. 2,
104See 5 U.S.C. 552(f)(2) (1997). This arguably vague definition suggests that a
"record" should include something held in any format, but does not hint at the link
between the record and an agency that is required to create an "agency record," discussed
infra. This author has not discovered recent cases construing this new definition of
"record" to mean anything different than the previous federal cases discussed in this
115See the "agency record" discussion, supra; see also Marie Veronica O'Connell, A
Control Test For Determining "Agency Record" Status Under the Freedom of
Information Act, 85 Colum. L. Rev. 611, 614 (April 1995); Stephen D. Hall, What Is A
Record? Two Approaches to the Freedom of Information Act's Threshold Requirement,
1978 B.Y.U. L. Rev. 408 (1978). Hall argues that to achieve flexibility on the "record"
question, courts should use their equitable powers instead of being tied to rigorous
definitions. See Hall, 1978 B.Y.U. L. Rev. at 427; Franklin and Bouchard, supra note 91,
Since the passage of the Electronic Freedom of Information Act, it has been argued
that the 1996 amendments were meant to ensure that the Freedom of Information Act's
definition of "record" will be interpreted broadly. For example, Senator Patrick Leahy,
who commented on the Amendments in the Senate before EFOIA's passage, has stated
that "the legislative finding for the Electronic FOIA Amendments clarifies that the
purpose of the FOIA is to allow any person to access government information for any
purpose." See Patrick Leahy, The Electronic FOIA Amendments of 1996: Reformatting
the FOIA for On-Line Access, 50 Admin. L. Rev. 339, 340 (Spring 1998). In fact,
Senator Leahy has stated that the Amendments were meant to ensure that "agency
records" mean more than just records that directly shed light on the operations of the
government agency. See id; see also Michael E. Tankersley, How the Electronic
Freedom of Information Act Amendments of 1996 Update Public Access for the
The purpose of the next section is to analyze federal court application of the FOIA to
private entities dealing with the government. After the case law has been analyzed, the
remaining two sections of this chapter will discuss whether the law squares with the
explicit language in the FOIA and its stated purposes as expressed by Congress.
Privatization and Freedom of Information in Federal Courts
In determining whether to subject private entities to the Freedom of Information Act,
federal courts have taken a variety of approaches. While courts have had some direction
from Congress in determining what is an "agency" under the Act,106 courts must also
determine if the record in question is an "agency record," a term Congress has left
undefined.107 The purpose of the first part of this section is to analyze two federal court
factors used in determining what is an "agency" under the Act. The second part will then
Information Age, 50 Admin. L. Rev. 421, 457 (Spring 1998) ("For the first time, all
federal agencies are required to make broad categories of agency records immediately
available to the public at agency records depositories and on-line.").
However, the argument that EFOIA actually broadens any definition of "agency
record" has not been adopted by the federal courts, and it has also been argued that if
Congress wanted to broaden access, it would have been more explicit in the
Amendments. See, e.g., Martin E. Halstuk, Bits, Bytes, and the Right to Know: How the
Electronic Freedom of Information Act Holds the Key to Public Access to a Wealth of
Useful Government Databases, Santa Clara Comp. & High Tech. L.J. (forthcoming
December 1998) (arguing that Congress should be more specific by amending the
Freedom of Information Act further to ensure that its purpose of broad disclosure is
fulfilled; EFOIA simply states that records maintained in an electronic format are to be
treated the same as those records held in paper format, and it did not explicitly change the
United States Supreme Court's ruling in Reporters Committee that the FOIA is meant to
allow access to information dealing directly with government agency performance).
Thus, the EFOIA Amendments do not change the case law that determines what is an
"agency record" under the Freedom of Information Act, discussed at 51-69, infra.
106See infra at 69-70.
107See infra at 70-74.
analyze the factors used by the federal courts in attempting to define the term "agency
Determining What Is an "Agency" Under the FOIA
In construing the Freedom of Information Act's definition of "agency,"108 federal
courts have considered two factors. One factor is whether the entity has substantial
independent authority in performing a function of the government, making it the
"functional equivalent" of the government.109 The other factor is whether the government
substantially controls the entity's day-to-day operations or organizational framework.110
In seeking to determine whether one of these determinative factors exists, the court is
essentially asking to what degree the entity is performing a government function. In one
case the government is pulling nearly all of the strings; in the other case the entity is
making decisions independently for the government. Although these two factors can
overlap, they will be treated separately in this section in order to distinguish the reasoning
behind each factor.
An analysis of the case law determining when a private entity is nonetheless an
"agency" under the Freedom of Information Act reveals that either a high degree of
government involvement is necessary, or the entity must in effect become the
government. Unless the government substantially controls the operations of a private
iosSee 5 U.S.C. 552(f) (1997).
109See, e.g.,United States v. Orleans, 425 U.S. 807 (1976).
"11See, e.g., Forsham v. Harris, 445 U.S. 169 (1980); see also Shafer, supra note 70,
for a discussion of the factors determining what constitutes an "agency" under the Act.
entity, or unless the private entity is substantially acting for the government, the private
entity will not be subject to the Act. The entity must essentially be an arm of the
The "control" factor. The seminal case in determining what constitutes an "agency"
under the FOIA is the United States Supreme Court's decision in Forsham v. Harris.111 In
that case, the petitioners sought the release of raw data possessed by a group of private
physicians and scientists conducting research under federal grants from the National
Institute of Arthritis, Metabolism, and Digestive Diseases.12 Although the National
Institute supervised the grantees and had a right of access and permanent custody of their
documents, it did not exercise these rights and left the day-to-day operations of the
program to the grantees.113 The raw data ultimately led to the restriction of certain drugs
by the Secretary of Health, Education and Welfare (HEW) and the Food and Drug
Although the Court admitted that a federal grantee could potentially be considered a
federal agency, that status could only be conferred on entities subjected to substantial
federal supervision, not merely a lower level of federal oversight "necessary to assure
compliance with the goals of the federal grant.""115 Without this substantial control by the
111445 U.S. 169 (1980).
112See id. at 172.
n3See id. at 173.
114See id. at 174-175.
115See id. at 180, n.ll.
government, the Court reasoned that Congress did not intend federal funding and some
supervision alone to create an agency under the Act.116 Private grantees, which are
neither chartered by the government nor controlled by it, cannot be considered agencies
under the Act.117 Because the grantees in Forsham retained their decision-making
authority, there was no extensive, day-to-day supervision by the government necessary to
turn the grantees into a federal agency.11"
The Court in Forsham relied on an earlier decision involving a private corporation,
United States v. Orleans.119 In that case, a federally funded community action agency was
sued under the Federal Tort Claims Act.120 Although the agency was being completely
funded by the government, it was independently operated and not subject to the control of
the government.121 The government did not have the power to supervise the day-to-day
operations of the community action agency, but was instead free to contract with the
116See id. at 181. Additionally, the Court would not allow access though a broadening
of the "agency records" definition. See id. This aspect of the case will be discussed infra.
117See id. at 179 (citing H. Conf. Rep. No. 93-180, at 14-15 (1974)).
118See id. at 180. The Supreme Court essentially followed the reasoning of the United
States Court of Appeals for the District of Columbia Circuit, which held that entities
receiving federal grants do not automatically become government agencies. See Forsham
v. Califano, 587 F.2d 1128, 1135 (D.C. Cir. 1978). Even with the multi-million dollar
federal grant, which entirely funded the study, the government had not exercised its
contractual right to review or retain the raw data. See id. at 1136. This, without the
exercise of its rights to review and oversee the entity's operations, the government
essentially exercised no control over the project. See id.
119425 U.S. 807 (1976).
121See id. at 814.
agency in order to protect its investment.122 The Supreme Court held that this did not turn
acts of the agency into acts of the government for purposes of the Tort Claims Act, and
this reasoning was carried over to the FOIA definition of "agency" in Forsham.123
Lower federal courts have followed the Forsham-Orleans approach by requiring
government control of a private entity before holding the entity accountable under the
Freedom of Information Act. As long as the requisite control is present, however, courts
have been willing to hold private entities accountable as "agencies" under the Freedom of
Information Act. In Rocap v. Indiek,124 the Court of Appeals for the District of Columbia
Circuit found that the Federal Home Loan Mortgage Corporation (FHLMC) was an
"agency" under the Freedom of Information Act because it met the definition of
"government controlled corporation" under Sect. 552(f) of the Act.125 This was due to
several factors: the Corporation was chartered under federal law, it was controlled by
federal statute, its employees were federal employees, it operated solely on federal funds,
and it was subject to the complete control of federal officials.126 In essence, it was
"subject to such substantial control over its day-to-day operations" that it was clearly an
"agency" under the Act.127 While one of these factors alone would not be enough to make
122See id. at 815.
123See id. at 815-816.
124539 F.2d 174 (D.C. Cir. 1976).
125See id. at 175; see also 5 U.S.C. Sect. 552(f) (1997).
126See Rocap, 539 F.2d at 176.
127See id. at 177.
the Corporation an "agency" under the Act, the court held that when combined, the
factors meant that the FHLMC was the kind of entity Congress intended to include in the
"government controlled corporation" definition.128
In contrast, four years after Rocap the District of Columbia Circuit held in Public
Citizen Health Research Group v. Department of Health, Education and Welfare129 that a
medical foundation contracting with the Department of Health, Education and Welfare to
conduct a professional standards review for the Medicare and Medicaid programs was not
an "agency" under the Freedom of Information Act.130 The foundation was a private,
non-profit corporation organized pursuant to state law, and its operations were the
responsibility of private physicians, not the government.131 Because the foundation was
independently run by private physicians, with little review by the government, the court
concluded that Congress did not intend for these review entities to be subject to public
access.132 Although the private foundation was making decisions with direct implications
128See id. The Court noted, however, that Congress did not intend to extend the
"agency" definition to "corporations which receive appropriated funds but are neither
chartered by the Federal Government nor controlled by it." See id. at 179 (citing
S.Conf.Rep. 93-1200, 93d Cong., 2d Sess., at 14-15 (1974)).
129668 F.2d 537 (D.C. Cir. 1981).
130See id. at 538.
131See id. at 542.
132See id. at 543. The court noted that just because an organization makes decisions
that impact the government does not make it a government agency, because "each
arrangement must be examined in its own context." See id. at 544.
for the federal Medicare and Medicaid programs, it had to remain closed to public
scrutiny because it was independent of government.133
Writing in concurrence, Chief Judge Edward Tamm pointed out that each case
determining whether an entity is an agency under the Act must be examined in its own
context.134 He wrote to express his belief that the factors used to determine agency status
are not necessarily cumulative: it would not be necessary, for example, to show that an
entity has authority to make decisions for the government and that it is controlled by the
government.135 "A galaxy of factors," such as control, the nature of the entity, what
133See id. at 543-544. The court described the foundation's function as though the
physicians were acting as expert advisors to the government. See id. The court reasoned
that it would be impractical and unfair to subject groups such as this to public scrutiny
just because they render opinions for the government. See id. Statutory provisions and
regulations which govern the private group, standing alone, did not provide the degree of
government control necessary to turn the foundation into an agency under the FOIA. See
id. at 544; see also St. Mary Hospital v. Philadelphia Professional Standards Review
Organization, 1980 WL 19448 (E.D. Pa. 1980) (holding that a Professional Standards
Review Organization was not an "agency" under the FOIA because Congress intended
PRSOs to be private, and did not want the federal government to assume any day-to-day
responsibilities for running the organizations).
134See Public Citizen Health, 668 F.2d at 544 (Tamm, C.J., concurring).
135See id. The authority to make decisions for the government, or what can be termed
the "functional equivalence" factor, is discussed in the next subsection. The District
Court for the District of Columbia had previously held that the medical foundation was an
"agency" under the FOIA because it had independent authority to make decisions; it was
required to review health care and to make final and binding determinations as to whether
the care rendered was necessary and thus qualified for federal reimbursement under the
Medicare and Medicaid programs. See Public Citizen Health Research Group v.
Department of Health, Education and Welfare, 449 F.Supp. 937 (D.D.C. 1978). The
lower court reasoned that this authority to make crucial decisions made the foundation an
agency, but the federal circuit court chose instead to rely on a "control test," finding that
the lack of federal day-to-day control made the foundation a private entity not subject to
FOIA. See Public Citizen, 668 F.2d at 537.
portion of its business is done through government contract and the nature of the
function performed should be considered to ensure that the values of the FOIA are
protected, rather than applying a rigid "cumulative" approach.136
Similar to the majority in Public Citizen Health, the United States Court of Appeals
for the Ninth Circuit used a control test in Irwin Memorial Blood Bank of The San
Francisco Medical Society v. American National Red Cross137 to determine that the
American Red Cross is not an "agency" under the Freedom of Information Act.138 Even
though the Red Cross has a close relationship with the federal government, the court held
that it is not subject to the kind of "substantial control" by the government that would
136See id. at 545. In dissent, Judge Abner Mikva expressed the concern that the
majority was applying a rigid "cumulative" test to determine when an entity is an agency
under the Act. See id. (Mikva, J., dissenting). Judge Mikva did not want "independence
from the government" to be dispositive, because other factors could necessitate finding
that the values of the FOIA would be served by disclosure. See id. He wanted some
discussion about which factors should be given more weight, and which factors could
necessitate a result in certain situations. See id.
In seeking a more generalized explanation of the process to be employed, Judge Mikva
explained the difference between independent decision-making and government control
by stating that bodiesis with the delegated authority to make significant decisions are
agencies in their own right. They act in the place of a preexisting government body in the
exercise of a central function." See id. at 546. In contrast, bodiesis that are very closely
controlled by the government are also subject to FOIA, but for the opposite reason. They
possess too little independent authority to be considered entities separate from the
government agency to which they are related. Their records are equivalent to the records
of the controlling agency." See id. at 546. Judge Mikva sought a "consistent, predictable
body of law" in the agency area, in order to prevent government from farming out
functions to private entities to avoid disclosure. See id. Because the foundation in the
present case had sufficient independent decision-making authority, Judge Mikva would
have allowed public access. See id. at 547.
137640 F.2d 1051 (9th Cir. 1981).
138See id. at 1053.
make it an "agency" under the Act.139 The court reasoned that each case must be decided
on its own facts because entities could differ so drastically that one case provides little
precedential value in determining when an entity should be considered an agency.140 But
as a threshold showing, the court reasoned that "substantial federal control or
supervision" is required before an entity can be considered part of the federal
By using the "substantial federal control" criteria, the Irwin court stated that it
intended to separate entities truly controlled by the federal government from those that are
merely "quasi-public or quasi-governmental."142 To determine the proper degree of
federal control, the court stated it would be necessary to analyze the various "federal
characteristics" possessed by the entity.143 The court stated that the Red Cross does not
employ federal employees, it does not function through federal funding unless under
specific purpose grants, and while it is supervised to some extent by the government, the
Red Cross is not directed by federal officials in its day-to-day operations.144 Essentially,
139See id. at 1054-1055.
140See id. at 1054.
142See id. at 1055.
144See id. at 1056-1057. Federal supervision did exist in requiring the Red Cross to
file reports, in its use of federal buildings, and because the President appoints members to
the Red Cross' Board of Governors, but that was not the kind of control or direction that
would make the Red Cross a government agency. See id.; see also Lombardo v. Handler,
397 F.Supp. 792, 802 (D.D.C. 1975) (holding that the National Academy of Sciences is
the court reasoned that the Red Cross exists to provide voluntary aid, and it is operated in
many different countries without any substantial control by one particular government.145
It is a "close ally" of the United States government, but it is not under the direct control of
the government on a daily basis.146
The "functional equivalence" factor. This factor basically represents the opposite
situation from the control factor. Here, the entity is functioning independently but
making decisions for the government, as opposed to having its decisions made by the
not an agency under the Freedom of Information Act even with the existence of many
federal factors: it was established by and reports to Congress, it is obligated to function
for the federal government upon request, and it is funded almost completely by the
government. Even so, the Academy is not subject to significant day-to-day control by the
145See Irwin, 640 F.2d at 1057.
146See id; see also Dong v. Smithsonian Institution, 125 F.3d 877 (D.C. Cir. 1997)
(holding that the Smithsonian Institution is not an agency subject to the Privacy Act,
which borrows its definition of "agency" from the FOIA, because the Smithsonian is not
subject to the day-to-day control of the government, and it exists to do research, not to
make independent decisions for the government).
In contrast to Dong, the United States District Court for the District of Columbia
previously held that the Smithsonian Institution was an "agency" subject to the Freedom
of Information Act. See Cotton v. Mcc. Adams, 798 F.Supp. 22 (D.D.C. 1992). The
court reasoned that the Smithsonian performs government functions as a center of
scholarship and national treasures, is chartered by Congress, receives federal funding,
enjoys immunity, and employs mainly civil service employees. See id. at 24; see also
Cotton v. Heyman, 63 F.3d 1115 (D.C. Cir. 1995) (affirming the district court opinion,
but holding that the Smithsonian could reasonably believe that it was not an "agency"
subject to the Act, and therefore the plaintiffs were not entitled to attorney's fees).
The Dong and Cotton cases appear to be in conflict. See also Alexander v. Federal
Bureau of Investigation, 971 F.Supp. 603, 606 (D.D.C. 1997) (reasoning that Congress
essentially adopted the "agency" definition from the FOIA for use in the Privacy Act, but
that the Privacy Act also has explicit language which may make the "agency" definition
somewhat different; for example, the Privacy Act includes the Office of the President,
while the FOIA does not).
government. In effect, it is the functional equivalent of the federal government, and
therefore it should be an "agency" under the Freedom of Information Act. The federal
cases in this section and the one above demonstrate that if either government control or
functional equivalence is present, the entity is likely to be considered an "agency" under
the Act. These are the two factors federal courts look for in a given case, and either will
be sufficient for public access to an entity.
The earliest court decision to use this "functional equivalence" analysis was Soucie v.
David,147 decided before the 1974 FOIA amendments expanded the definition of
"agency." In Soucie, the court held that the Office of Science and Technology (OST) had
an independent government function of evaluating federal scientific programs, therefore it
was an "agency" under the Administrative Procedure Act and the Freedom of Information
Act.148 The OST functioned as a distinct government agency, and not merely as an arm of
the President, because the OST was fully transferred responsibilities from the National
Three years after Soucie, the District of Columbia Circuit again discussed the
"functional equivalence" factor in Washington Research Project, Inc. v. Department of
147448 F.2d 1067 (D.C. Cir. 1971).
149See id. at 1075. Therefore, because the OST was considered to be an "agency," a
report that it generated was an "agency record" subject to disclosure because it was
created as part of OST's functions, unless one of the Act's exemptions applied. See id. at
Health, Education and Welfare. 15 In that case, the court held that initial review groups
that pass on research applications to the National Institute of Mental Health are not
agencies under the FOIA because they do not have legal authority to make decisions.15
Instead, the court considered the groups "advisory committees" performing staff
functions, and not independent agencies under the Act.152
The court reasoned that "employing consultants to improve the quality of the work that
is done cannot elevate the consultants to the status of the agencies for which they work
unless they become the functional equivalent of the agency, making its decisions for
it."'3 In the court's view, influence on the federal government is not enough; there must
be independent authority to make decisions.154 Because the groups were not "agencies,"
their recommendations could only be subject to disclosure when they were expressly
adopted by a government agency in making its decisions.155
In short, government agencies could farm out operations to private entities and avoid
disclosure. They would only have to stay out of the day-to-day operations of the private
entity, yet they could still retain their own important decision-making authority. The
private entity would not be an agency and its records are likely not to be considered
150504 F.2d 238 (D.C. Cir. 1974).
151See id. at 246.
153See id. at 247-248.
154See id. at 248.
"agency records," unless they become government records. The "agency record"
requirement is the subject of the next subsection.
Determining What Is an "Agency Record" Under the FOIA
In addition to the requirement that a private entity petitioned under the Freedom of
Information Act be an "agency," there must be a sufficient link between the government
agency and the record to justify calling it an "agency record."156 Although the Freedom of
Information Act does not define "agency record," the United States Supreme Court has
determined that there must be "some relationship between an 'agency' [and a] 'record'
requested."157 Therefore, courts have had to determine the degree of this relationship,
leading some to fear that the "agency record" definition will be construed too narrowly,
thereby frustrating the purposes behind the FOIA."58 Questions arise as to what degree of
control by the federal government over the record is necessary before it can be considered
an "agency record."159 As this section will show, the United States Supreme Court has
156See generally O'Connell, supra note 105.
157See Forsham, 445 U.S. at 178.
158See O'Connell, supra note 105, at 615; Note, Agency Records Under the Freedom
of Information Act: A Analysis of Forsham v. Califano, 13 Ga. L. Rev. 1040, 1095 (1979)
(arguing that a narrow definition of "agency record" could have the effect of expanding
the exemptions under the Act).
159See O'Connell, supra note 105, at 617-619; United States v. Charmer Industry, 711
F.2d 1164 (2d Cir. 1983) (requiring documents to originate in a government body subject
to the FOIA); Carson v. United States Department of Justice, 631 F.2d 1008, 1015 (D.C.
Cir. 1980) (holding that records central to the function of an agency are controlled by the
agency); Berry v. Department of Justice, 733 F.2d 1343, 1348-1349 (9th Cir. 1984)
(requiring some sort of "meaningful control" by an agency over the record before it can
be considered an "agency record").
required that the record be both possessed and controlled by an "agency" under the
Freedom of Information Act, a test that can be termed "official control." This test has
also been followed by lower federal courts.
The Supreme Court "official control" approach. In determining what requested
materials constitute "agency records," the United States Supreme Court has used a two-
prong test. First, the Court has stated that the government agency must "either create or
obtain" the materials as a threshold requirement for "agency record" status.160 In other
words, it must first be determined whether the record was originally created by a
government agency, or if it was created by a private entity, it must be obtained by the
federal agency in order to meet this threshold requirement. Either creation or later
possession establishes that the record has some government purpose and is not purely
private in nature.
Second, the Court has required direct control by the government agency over the
records at the time the FOIA request is made before they can be considered "agency
records."161 Possession is a necessary requirement for "control," but it is not sufficient
unless there is some sort of possession of the record for use in the agency's official
160See Tax Analysts, 492 U.S. at 144. The Court noted that creation is not an essential
prerequisite to information becoming "agency records;" it is enough that the agency
acquire the information because "[t]o restrict the term 'agency records' to materials
generated internally would frustrate Congress' desire to put within public reach the
information available to an agency in its decision-making process." See id. (citing
Chrysler Corp. v. Brown, 441 U.S. 281, 290, n.10 (1979)).
161See Tax Analysts, 492 U.S. at 145; see also Kissinger v. Reporters Committee for
Freedom of the Press, 489 U.S. 749, 772 (1989).
duties.162 The record cannot simply be located at the agency; it must become a part of the
agency's official business.
In short, this two-prong test focuses first on "creation or possession" by a federal
agency in order to establish that the record has a governmental purpose and is not purely
private. It then moves to whether the necessary "control" is present through both
possession and use in the agency's official duties at the time the FOIA request is made.
By focusing on both an official purpose of the record163 and control over the record at the
time of the FOIA request, this two-prong test can be termed an "official control" test.
Discussing the "creation or possession" prong of the test, the Court in Forsham held
that data created by a group of private physicians could not be considered "agency
records" because the data were not at any time obtained by the Secretary of Health,
Education and Welfare.164 In order for records that are created, owned and possessed by a
private entity to become "agency records," they must first be obtained by an FOIA
agency.165 Because the FOIA agency in Forsham only had rights of access and permanent
custody of the records, and did not in fact possess the records, the data did not have the
162See Tax Analysts, 492 U.S. at 145.
163This would be established by showing either that the government created the record
itself, or that a private entity created the record and it was later obtained by the agency,
suggesting that the record was used by an agency for an official government purpose.
164See Forsham, 445 U.S. at 171.
necessary link to be considered "agency records."166 The first prong of the "agency
records" test was not satisfied because there was no creation or possession by a FOIA
The Court reasoned that federal funding and supervision alone did not make the
private group an "agency" under the Freedom of Information Act, and therefore its
records could not be considered "agency records" as long as they remained in private
hands.168 The Forsham Court therefore created a threshold requirement of creation or
possession, and in that case it was not necessary to inquire further into the nature of the
records because the FOIA agency had never possessed the records.169
166See id. at 171-173. The Court pointed out that even the Code of Federal
Regulations states that until the records are obtained by the federal agency, they are only
"records of grantees." See id. at 181 (quoting 45 C.F.R. Sect. 74.24 (1979)). Under the
FOIA itself, the records must have been in fact obtained, not merely could have been
obtained, before they can be considered "agency records." See id. at 186.
168See id at 181.
169See id. at 182. The Court also noted that "Congress has associated creation or
acquisition with the concept of governmental record" in other statutory contexts, such as
the Records Disposal Act, and in the legislative history of the FOIA itself. See id. at 183-
The Court agreed with the lower court in reasoning that a federal agency should not be
required to exercise its right of access in order to create an "agency record," because the
FOIA does not require agencies to create records. See Forsham v. Califano, 587 F.2d
1128, 1136-1137 (D.C. Cir. 1978). Only if a federal agency has created or obtained a
record, or if it has some duty to obtain, it in the course of doing its work, can the record
be considered an "agency record." See id. at 1136. But the majority did acknowledge
that obviouslysy a government agency cannot circumvent FOIA by transferring physical
possession of its records to a warehouse or like bailee." See id. at 1136, n. 19.
Additionally, the court stated: "Scientists engaged in research on federal grants must
accept the fact that any documents filed with the federal government, whether on the
scientists' own initiative or an audit or other lawful demand, are subject to FOIA." See
In dissent, Justice William Brennan preferred to weigh each case on its merits, noting
that where the nexus between the agency and the information is close, and where the
information is important to the public's understanding of agency operations, the
information should be an "agency record" under the FOIA.170 To Justice Brennan, the
issue was "the importance of the record to an understanding of Government activities"
and the extent to which the agency has treated the record as part of the regulatory process,
not actual possession.171 Because the government agencies at issue in Forsham had
significantly relied on the private study, and were "deeply involved" in the creation of the
data, Justice Brennan would have held the information to be an "agency record" under the
id. at 1137. Thus, the lower court equated "control" with "possession."
170See Forsham, 445 U.S. at 188-189 (Brennan, J., dissenting).
171See id. at 189-190.
172See id. at 191-192. Justice Brennan pointed out that the government agencies were
also "deeply involved" with the records because they "remain with the grantee only at the
pleasure of the government." See id. at 192. He feared that by relying on private grantees
to perform important government services, governments could avoid the FOIA if the
courts do not define "agency record" in a broader sense. See id.
Judge Bazelon, who dissented in the lower court opinion, also agreed that factors other
than simple possession should be considered in determining "agency record" status. See
Forsham v. Califano, 587 F.2d 1128, 1140 (D.C. Cir. 1978) (Bazelon, C.J., dissenting).
Because the government had an unrestricted right of access to the records, provided all of
the funding for the private physicians group, and extensively relied on the study data,
Judge Bazelon said there was sufficient government involvement with the data to create
agency records. See id. The funding, access, and reliance factors, taken together,
established agency records in Judge Bazelon's opinion. See id. at 1142.
Judge Bazelon also pointed out that records need not be located at the government
agency to be considered "agency records" because recordsrs may be bailed to a privately-
owned warehouse, loaned to a private entity, or may have been sold or donated to the
government but not delivered." See id. at 1144. The most determinative factor, in Judge
The second prong of the "official control" test is the requirement that the records be in
the possession of a FOIA agency at the time of the request for use in its official duties, a
prong developed in 1980 in Kissinger v. Reporters Committee for Freedom of the
Press.173 In that case, the Reporters Committee for Freedom of the Press sought to gain
access to records and notes of Henry Kissinger's official telephone conversations made
during his service as Assistant to President Nixon for National Security Affairs and as
Secretary of State.174 Holding that the records were not "agency records" under the
Freedom of Information Act at the time of their creation because the Act does not include
the Office of the President, the Court reasoned that they did not acquire "agency record"
status when Kissinger transferred them to his office at the State Department, an agency
under the FOIA.175
While the records were technically located at the FOIA agency, they were not in the
Department's official control because they were not used by the Department for any
purpose.176 Although the Forsham "creation or possession" requirement was technically
met, the Kissinger Court added the additional requirement that the possession be for
Bazelon's opinion, was the extreme reliance by the government on the data generated by
the private physicians, especially because it was a controversial study. See id.
173445 U.S. 136 (1980).
174See id. at 139-140.
175See id. at 157.
official purposes, creating the two-prong "official control" test.177 The possession
requirement must first be met, but the records cannot be mere personal records located at
an agency; they must be used by the agency in the conduct of its official duties.178
Because the records were personal to Kissinger and his donee, the Library of Congress,
the State Department did not have official possession or control over the documents at the
time of the FOIA requests.179
The Court noted that it was not deciding at what point records that relate to agency
affairs become records of that agency.180 However, noting that the records related only to
Kissinger's employment in the Office of the President, the Court ruled that something
more than the mere physical location of records in an FOIA agency was needed;
otherwise Kissinger's "personal books, speeches, and all other memorabilia stored in his
office would have been agency records subject to disclosure under the FOIA.""181 This
would create public records out of personal information. In essence, the Court was
178See id. at 151. The Court cited the Attorney General guidelines issued shortly after
the FOIA was passed, stating that the FOIA "refers, of course, only to records in being
and in possession or control of an agency..." See id. (quoting Attorney General's
Memorandum on the Public Information Section of the Administrative Procedure Act 23-
24 (June 1967)) (emphasis added). Under the requirements laid out in Forsham and Tax
Analysts, however, there must be both possession and control -- neither factor alone is
sufficient to create "agency records."
179See id. at 155.
18oSee id at 156.
s11See id. at 157.
looking for "official control" by an FOIA agency, which must include possession for use
in its official capacity.182
The Supreme Court further continued to apply the "official control" test in United
States Department of Justice v. Tax Analysts,183 which focused not only on the "creation
or possession" requirement, but also required that the records be under the agency's
control at the time of the FOIA request.184 There, the Court required the Tax Division of
the Department of Justice to make available to the public district court cases received
while litigating tax cases because the Department obtained the records from the district
courts and was in control of the records when the FOIA requests were made.185 The
Court spelled out the two prongs of the "official control" test laid out in Forsham and
Kissinger: an agency must first create or obtain the records, and it must be in control of
the requested records at the time of the FOIA request, meaning that "the materials have
182In his opinion concurring in part and dissenting in part, Justice Stevens worried that
applying a strict "possession and control" approach would allow agencies to simply
remove documents from their physical possession and therefore frustrate the purposes
behind the FOIA. See id. at 161-162 (Stevens, J., concurring in part and dissenting in
part). In fact, Justice Stevens commented that the Court's strict possession approach
"creates an incentive for outgoing agency officials to remove potentially embarrassing
documents from their files in order to frustrate future FOIA requests." See id. at 161. In
his opinion, an agency "withholds" a record when it refuses to produce a record for which
it has a legal right to possess or control. See id. at 162. In other words, Justice Stevens
would require an agency to produce records in private hands if the records could legally
be acquired by the agency. See id. at 164.
183492 U.S. 136 (1989).
184See id at 145.
185See id. at 138.
come into the agency's possession in the legitimate conduct of its official duties."186
Creation or possession of the record is not enough without official control over and use of
the record at the time of the FOIA request.
In discussing both prongs of the "official control" test, the court emphasized that it
was not restricting "agency records" to those actually created within the FOIA agency.187
The Court noted that FOIA agencies frequently use privately-generated materials in
performing their official duties, and these materials are still a part of the decision-making
process even if not created within the agency. 188 But the Court noted that agency records
would not include personal materials in an agency employee's possession; they must be
acquired by the agency itself in connection with its official business.189 In other words,
"creation or possession" alone would not satisfy the "agency record" test unless the
agency possessed the records for its own use at the time of the FOIA request.190 There
must be "official control" over the records, beyond mere physical location at the
agency.191 Both prongs of the "official control" test must be met.
186See id. at 144-145.
187See id. at 144.
188See id; see also Chrysler Corp. v. Brown, 441 U.S. 281, 293 (1979) (noting that
FOIA agencies frequently obtain potentially confidential information from private entities
in the course of doing business).
189See Tax Analysts, 492 U.S. at 145 (citing Kissinger v. Reporters Committee for
Freedom of Press, 445 U.S. 136, 157 (1980)); Forsham, 445 U.S. at 183.
190See Tax Analysts, 492 U.S. at 145.
191See id. at 143-145.
The Court in Tax Analysts held that the district court cases were "agency records"
because they were both possessed and controlled by the Department in the course of its
official duties; it did not matter that the cases were created by a nonagency under the
FOIA.192 The FOIA's goal should be to give the public access to records received by an
agency in conducting its business, regardless of the actual author of the records.193 The
Court reiterated the threshold requirement of "creation or possession," but it also
required more than just possession -- the records must be located at the FOIA agency and
be a part of its official business at the time of the FOIA request. In short, the Court
merged the precedent set by Forsham and Kissinger, solidifying the "official control"
The lower court "control" approaches. Lower federal courts have also required
some degree of control over a record by an FOIA agency in the conduct of its official
duties in order to make the record an "agency record." In Berry v. Department of
Justice,194 the Court of Appeals for the Ninth Circuit held that presentence reports
prepared by the courts for use by the Federal Bureau of Prisons and the Parole
Commission, both FOIA agencies, were "agency records" when in the possession of the
192See id. at 146-147. The Court noted that disputes over control should be infrequent,
but that it was "leaving for another day" those cases where records are "purposefully
routed ... out of agency possession in order to circumvent [an impending] FOIA request,"
or "wrongfully removed by an individual after a request is filed." See id. at 146 (quoting
Kissinger, 445 U.S. at 155, n.9).
193See id. at 147.
194733 F.2d 1343 (9th Cir. 1984).
agencies for their official use.195 The records were not "court documents," which would
not be subject to the FOIA, because they were in the possession of agencies, used in their
official duties, and subject to the free disposition of the agencies.196 Therefore, they were
"agency records" because they were within the "official control" of the agencies.
The court in Berry followed Supreme Court precedent by holding that creation alone is
not dispositive of the "control test."197 Otherwise, all documents of nonagencies would
be shrouded in secrecy.198 If the records are possessed by the agency, the court stated they
could still be subject to access as "agency records" even if they were originally created by
a nonagency because they essentially become a part of government business.199 The court
relied on an "official control" test focusing on whether the records are in the possession
195See id. at 1344.
196See id. at 1346; see also Crooker v. United States Parole Commission, 760 F.2d 1
(1st Cir. 1985) (presentence reports used by the U.S. Parole Commission are "agency
records"); Carson v. U.S. Department of Justice, 631 F.2d 1008, 1015 (D.C. Cir. 1980)
(presentence reports prepared by the probation service but in the possession of the Parole
Commission, an FOIA agency, are central to the Commission's primary function and are
therefore "agency records"); cf Goland v. Central Intelligence Agency, 607 F.2d 339, 346
(D.C. Cir. 1978) (Congress had not relinquished its control over documents even though
they were in the temporary possession of the CIA; thus, the CIA did not have sufficient
"control" over the records to render them "agency records").
197See Berry, 733 F.2d at 1348 (citing Kissinger, 445 U.S. at 157, and Forsham, 445
U.S. at 181 ("Records of a nonagency certainly could become agency records as well")).
198See id. at 1348.
199See id. at 1350. The court stressed the record's creation for agency decision-making
as important for disclosure, but it also suggested that data created outside the agency but
possessed by the agency could also be subject to access, particularly if other factors such
as federal funding are involved. See id. at 1350, n.10.
of an agency and "prepared substantially to be relied upon in agency decision-making."200
Possession by the agency is a prerequisite, but the record must also reflect how an agency
conducts its official duties.201 Because presentence reports are routinely prepared by
courts and then forwarded to the federal agencies for use in their duties, the reports were
"agency records" because they were both in possession of the agencies and within their
control.202 The Berry court therefore required both possession and control, in keeping
with the "official control" approach solidified in the Supreme Court cases ending with
Similar to the Berry court's approach, the Court of Appeals for the Seventh Circuit in
General Electric Company v. United States Nuclear Regulatory Commission204 held that
the term "agency record" covers not only documents created by a federal agency, but also
200See id. at 1349.
201See id. at 1350.
202See id. at 1350-1351.
203See also Burka v. United States Dept. of Health and Human Services, 87 F.3d 508,
515 (D.C. Cir. 1996), in which data tapes and questionnaires regarding surveys about
smoking habits and attitudes conducted by the National Cancer Institute and a private
company at the direction of the Department of Health and Human Services were "agency
records"; HHS exercised extensive control over the data and documents were created on
behalf of HHS, and the National Cancer Institute planned to take possession of the
materials and use them in its official duties.
In essence, this case can be contrasted to Forsham, where the federal agency did not
have the extensive control over the private entity study present in Burka. In Burka, a
federal agency assisted in the creation of the records, and another federal agency
exercised extensive day-to-day control over the study. See id. In short, the entities were
"agencies" under the Forsham approach, and their records would eventually be under the
official control of HHS. See id.
204750 F.2d 1394 (7th Cir. 1984).
documents submitted to the agency for use in its official duties.205 Otherwise, the court
reasoned that the purpose behind the Freedom of Information Act -- "to give the public
access to information on which the government bases action" -- would be impeded if
documents possessed by the agency and moving the agency to act were excluded from the
term "agency record."206 This reasoning resulted in the conclusion that a record compiled
during a nuclear licensing proceeding was an "agency record" used by the Nuclear
Regulatory Commission, an FOIA agency, even though the record was prepared by
General Electric, a private company.207 This was because the NRC had possession and
control over the document for its official use, which satisfied the two prongs of the
"official control" test.
Other federal courts have also used an "official control" test in determining what
records are subject to the FOIA. In Teich v. Food and Drug Administration,208 the United
States District Court for the District of Columbia held that a summary of consumer
complaints about a private manufacturer's breast implants was an "agency record" subject
to the FOIA because it was filed with the FDA, an agency under the Act, and was
205See id. at 1400.
206See id.,; see also NLRB v. Robbins Tire & Ribber Co., 437 U.S. 214, 242 (1978)
(discussing purposes behind FOIA); Weisberg v. United States Dept. of Justice, 631 F.2d
824, 827-828 (D.C. Cir. 1980) (same).
207See General Electric, 750 F.2d at 1400. The court added that even though the
Commission did not rely extensively on the report, it was still an agency record. See id.
This was because the public still had an interest in knowing what portions of the report
the Commission relied on and what exact uses it made of the report, and thus the public
should be entitled to the whole report. See id.
208751 F.Supp. 243 (D.D.C. 1990).
obtained as part of the FDA's official duties.209 This was because in requesting
documents from Dow Corning regarding breast implants, the FDA was conducting its
official business.210 It was therefore irrelevant whether the FDA planned to return the
documents, as long as they were possessed by the Agency and used in its official
business.21 The two prongs of the "official control" test were met.
In contrast, the Southern District of New York in Ciba-Geigy Corporation v.
1 A,/ein reasoned that records of a university research group, which was not an agency
under the FOIA,213 were not "agency records." The court stated that it was not shown that
209See id. at 248.
210See id; see also Rush v. Department of State, 716 F.Supp. 598 (S.D. Fla. 1989)
(holding that documents possessed by the Department of State regarding negotiations on
the status of Berlin were "agency records;" the Department was the only entity in control
of the documents since obtaining them in 1977); Ryan v. Department of Justice, 617 F.2d
781, 786 (D.C. Cir. 1980) ("Unless there is evidence of control by some other entity, [the
court] must conclude that [the agency] controls these documents"); Ni Iwi 0 Na Kupuna
O Mokapu v. Dalton, 894 F.Supp. 1397, 1411 (D. Hawai'i 1995) (inventory of Native
American remains under the Native American Graves Protection and Repatriation Act
was an "agency record" because the Secretary of the Navy had contracted for and
obtained the inventory, and it was in the federal agency's control indefinitely).
211See Teich, 751 F. Supp. at 248. Additionally, the court seemed troubled by the fact
that the FDA itself attempted to determine that the documents were not "agency records"
by promising confidentiality during the review process. See id. at 248-249. In effect, the
court stated that the FDA was attempting to evade the Act by making this information
regarding breast implants, which was of extreme importance to the public, confidential by
its own determination. See id. The court would not allow this agency determination of
confidentiality. See id.
212428 F.Supp. 523 (S.D.N.Y. 1977).
213This was because the research group had no independent authority to perform
government functions, no legal authority to make decisions, and was not under any
federal control. Its only power was to make recommendations to federal agencies, not
decisions. See id. at 526-528; see also Wolfe v. Weinberger, 403 F.Supp. 238 (D.D.C.
the records were controlled or substantially utilized by a federal agency under the Act.214
Because "agency records" reach only records owned or controlled by the government
agency and thus used in its official business, the court reasoned that "it must appear that
there was significant governmentt involvement with the records themselves in order to
deem them agency records."215
The District Court first held that government funding alone would not make the
underlying private research data "agency records."216 Otherwise, all entities gaining any
federal government support would be subject to disclosure, an effect that would "chill"
research progress.217 Additionally, the data produced by the entity were intended to be
private research data; they remained in control of the private entity and the government
did not exercise any dominion over the data.218 Lastly, while the Food and Drug
Administration relied on reports submitted by the private grantees, it did not rely on the
1975) (holding that an FDA advisory review panel played an important part in the FDA's
decision-making process, but it was still not an "agency" under the act because it was
empowered only to make recommendations, not decisions, for the FDA). The "agency"
definition under the FOIA is discussed at the beginning of this chapter, supra.
214See Ciba-Geigy, 428 F.Supp. at 528-529.
215See id at 529.
216See id at 530; see also Lombardo v. Handler, 397 F.Supp. 792, 802 (D.D.C. 1975)
(holding that substantial federal funding did not make the National Academy of Sciences
subject to the FOIA because it was not controlled by the government and its studies were
not conducted as an adjunct to government operations; FOIA was not meant to be applied
to private entities contracting with the government to conduct research).
217See Ciba-Geigy, 428 F.Supp. at 530.
218See id at 530-531.
underlying data.219 Thus, funding, access, and partial reliance were insufficient to create
"agency records" without some actual government agency possession and control in an
The Court of Appeals for the District of Columbia Circuit agreed with the Ciba-Geigy
court's reasoning in Wolfe v. Department of Health and Human Services.221 In Wolfe, the
Federal Circuit Court held that a report prepared by the President-elect's transition team
regarding the Department of Health and Human Services was not an "agency record"
because HHS did not create or maintain control over the documents, even though two
copies were physically located at the Department.222 The court reasoned that the agency
must have possession of documents in order to have "obtained" them, but there must also
219See id. at 531; see also Illinois Institute for Continuing Legal Education v. United
States Dept. of Labor, 545 F.Supp. 1229, 1233-1234 (N.D. Ill. 1982), (holding that a two-
volume briefing book on the Department of Labor prepared by the president elect's
transition staff, a nonagency under the FOIA, was not an "agency record" because the
DOL did not obtain the book as an agency. The book was only possessed by an official at
the DOL, thus there was an insufficient nexus between the agency itself and the record to
establish control; mere "influence" of the record on the agency would not be enough to
create an agency record absent actual possession of the record in an official capacity).
220See Ciba-Geigy, 428 F. Supp. at 531. The court stressed that "[m]ere access without
ownership and mere reliance without control will not suffice to convert the [private
entity's] data into agency data." See id. The research data was never turned over to the
FDA, it was prepared according to the methods of the private grantee, and the documents
only passed into the FDA's hands at a particular point in time, not on a permanent basis.
See id. And although private documents directly relied upon or memoranda expressly
adopted by an agency are subject to disclosure even if created by a private entity, in this
case the government agency did not expressly rely on or adopt the underlying data created
by a private research group. See id. at 532.
221711 F.2d 1077 (D.C. Cir. 1983).
222See id. at 1080-1081.
be some "'nexus'" between the agency and the documents other than the mere incidence
of location."223 Similar to Kissinger, the mere fact that two of the documents were
"within the four walls of the agency" was not enough to turn the documents into "agency
records" without official use by the FOIA agency.224
Although an individual had brought the documents within the four walls of HHS, the
court stated that they were not integrated into the agency's files or records.225 In other
words, they did not become official records of the agency itself, and were not used by the
agency in its official business. Physical location provided the only nexus between the
private documents and the agency, and this would not create agency records absent
possession by the agency in its official capacity, the second prong of the "official control"
test.226 Documents must still retain their private character even though they may relate to
an agency's business, as long as they remain within the exclusive control of the private
individual and do not become a part of the agency's business.227 In short, the court
22See id. at 1079-1080.
224See id. at 1080.
226See id. at 1080-1081; cf Ryan v. Department of Justice, 617 F.2d 781 (D.C. Cir.
1980) (holding that Attorney General's possession of documents was tantamount to
possession by the Justice Department, because the Attorney General possessed the
documents during his tenure in his official capacity, therefore the Department had control
over the documents).
227See Wolfe, 711 F.2d at 1081. The documents in Wolfe did not become a part of the
agency's official business because they were never integrated into the agency's files and
were not used by the agency in any way. See id. at 1080. Additionally, the court took the
approach in Forsham by ruling that access to the documents by the agency alone would
not create "agency records" unless the agency in fact exercised this ability to access to
followed the "official control" approach laid out by the Supreme Court in determining
the reach of "agency records" under the Freedom of Information Act, and both prongs
must be present to satisfy the test.228
Federal courts construing the undefined "agency record" term under the Freedom of
Information Act have therefore looked at creation or possession as a threshold
requirement, but have also required more than just physical location at the agency to
establish the second prong of the "official control" test. Some sort of control, comprising
not only possession but also use in an official capacity, by the agency is required in order
for the records to be considered "agency records." This control provides the necessary
link between the agency and the record to establish the agency's control over the record
required by the Supreme Court, allowing private entities to potentially be covered under
the Freedom of Information Act. The next two sections will discuss these federal court
records. See id; see also Illinois Institute, 545 F.Supp. at 1233-35 (documents of
transition team were not "agency records" because they had never been used by the
228See also Katz v. National Archives & Records Administration, 862 F.Supp. 476,
479-480 (D.D.C. 1994) (holding that autopsy photographs of President John F. Kennedy
were not "agency records" because even though they were created and possessed by a
government agency, they were never officially used for government business prior to their
private donation, and the Kennedy family retained control over the photographs);
International Brotherhood of Teamsters et al. v. National Mediation Board, 712 F.2d
1495 (D.C. Cir. 1983) (holding that National Mediation Board's possession of labels
bearing names of employees eligible to vote was not control over the labels because
possession was limited to one-time attachment of labels subject to court order); Martin
Marietta Aluminum, Inc. v. Administrator, 444 F.Supp. 945, 948-949 (C.D. Calif. 1977)
(holding that documents created by private consultants at the request of the agency were
agency records because they were prepared for the agency and officially used by the
approaches in light of the Freedom of Information Act's express terms, as well as the
legislative history behind the Act.
The Definition of "Agency" Under the Act
Commentators have suggested that the drafters of the 1974 amendments intended to
create a broad definition of "agency" that would include all federal organizations and
entities, beyond those the Act specifically excluded, such as Congress and the courts.229
But as the above section showed, the definition of "agency" in the Act has been subject to
judicial interpretation that does not include all entities subject to the government. Courts
have had to determine the parameters of the Act's "agency" definition, especially as
governments continue to contract out services and create a "myriad of organizational
arrangements for getting the business of government done."230 Because the FOIA
establishes the right of access, and there is no right unless an agency is involved, the
construction of this definition by the courts is crucial to the public's ability to monitor
229See O'Reilly, supra note 99; see also Taylor v. Diznoff, 633 F.Supp. 641 (W.D. Pa.
1986). One commentator has stated that a private entity can only be considered an agency
if it has statutory authority to make decisions for the contracting agency. See Anne H.
Wright, The Definition of "Agency" Under the Freedom of Information Act as Applied to
Federal Consultants and Grantees, 69 Geo. L.J. 1223, 1240 (1981). Wright was writing
about the stance taken by the Circuit Court of Appeals for the District of Columbia in
Washington Research Project, discussed supra. This author's research has not revealed
other commentators taking this stance with regard to the law in other federal circuits.
230See Washington Research Project, 504 F.2d at 245.
231See O'Reilly, supra note 99, at Section 4.02. In his treatise on freedom of
information, O'Reilly gives a thorough description of the various entities considered to be
agencies under the Freedom of Information Act, as interpreted by the federal courts. See
id; see also Richard P. Shafer, Annotation, Meaning of Term "Agency" For Purposes of
"Agency Records" Under the Act
The question discussed in the first section has been what relationship between the
agency and the record must exist to create an "agency record" under the FOIA. Federal
courts dealt with this issue by creating the "official control" test. However, one
commentator has argued that defining "agency record" too narrowly effectively expands
the Act's nine exemptions by allowing certain entities to automatically escape access
because of their private nature, thereby obstructing the Act's goals of obtaining "the
fullest responsible disclosure."232 If the government holds the information itself, or
makes a decision based on information contained in records created by a nonagency,
commentators argue that these should be "agency records" subject to disclosure.233
Regardless of whether the agency possesses the record or relies on it in the decision-
making process, disclosure would nonetheless be in keeping with the Act's purpose of
allowing oversight of the entities that "control information of interest to the public"
because in either case the record relates to a public function.234 This thesis argues that as
long as the record relates to a public, government function, it should be accessible under
the Freedom of Information Act without the need for a narrow focus on who possesses or
Freedom of Information Act, 57 A.L.R. Fed. 295 (1982), for a detailed overview of the
various cases applying the "agency" definition to various entities.
232See Wion, supra note 10.
233See id. at 1097; see also Casarez, supra note 1; Bunker and Davis, supra note 2.
234See H.R. Rep. No 876, 93d Cong., 2d. Sess 8-9 (1974); see also Wion, supra note
10, at 1097.
record such as the "official control" approach described in the first section. The
information is public and should therefore be open to the public.
However, commentators have noted that research held by independent companies, but
which forms the basis of government reports and is used for government decisions,
nonetheless falls outside of the Act under the federal courts' interpretation of "agency
records" discussed above.235 While government possession of information will usually
create "agency records,"236 without possession or significant day-to-day control by
the government agency, use or reliance on nonagency-created information will not turn it
into "agency records."237
The requirement that the subject of an FOIA request be an "agency record" can
therefore prevent access to information possessed by researchers and other private entities
regardless of the arguably public nature of the information.238 Commentators not only
fear that a narrow definition of "agency record" will create a new exemption to the
235See Wion, supra note 10, at 1102-1106, for a summary of these cases.
236See id. at 1107; see also Forsham v. Califano, 587 F.2d 1128, 1134, n.11 (D.C. Cir.
1978). Possession will generally suffice to create "agency records," unless the records
were created by a group specifically excluded from the Act by Congress. See, e.g., Warth
v. Department of Justice, 595 F.2d 521 (9th Cir. 1979). In these cases, courts have
generally required "control" of the record by the covered agency. See the Supreme Court
cases discussed supra; see also Goland v. CIA, 3 Med. L. Reptr. 2341 (D.C. Cir. 1978).
Thus, possession is a necessary requirement for "agency record" status, but it is not
sufficient in all cases. See the first section of this chapter, supra.
237See, e.g., Forsham v. Harris, 445 U.S. 169 (1980); Ciba-Geigy Corp. v. Matthews,
428 F. Supp. 523 (S.D.N.Y. 1977); Forsham V. Califano, 587 F.2d 1128 (D.C. Cir.
1978), and the other cases discussed in the first section, supra.
238See Note, supra note 150, at 1042-1043.
FOIA, but they also recognize that an overly broad meaning of the term will discourage
private contractors and grantees by subjecting their confidential information to public
scrutiny.239 Because Congress has thus far failed to define the term "agency record,"
commentators state it becomes an issue for the courts to decide whether to read the words
"agency record" narrowly by requiring "official control" over the records by a
government agency, thus allowing Congress to decide on a different meaning if it wishes,
or to form a broader judicial definition of "agency record" that could include records held
and controlled by a private entity.240 By following the more narrow approach discussed in
Section A-2, commentators argue that federal courts allow agencies to become shrouded
in secrecy once their previously public information originates in or is transferred to
entities not technically covered under the Act.241
239See id at 1054-1055 (discussing the fears of both access and privatization
240See id at 1056-1057. The author describes these two approaches, taken by the
majority and dissent, respectively, in Forsham v. Califano, as the "property" and
"government involvement" approaches. See id. The property approach relies on specific
and literal meanings of "agency records," while the "government involvement" approach
is fashioned by the judiciary to take a more flexible account of the government's
involvement with the records -- making it more likely that private entities will be subject
to the Act. See id. at 1066-1067.
241See O'Connell, supra note 105, at 612. In this note, O'Connell concludes that
"agency record" should be defined broadly by looking at the function of the record, thus
determining if it is used by the government for its functions. See id. Otherwise, the
purposes of disclosure behind the FOIA will be undermined. See id. at 615-616.
Neither Congress nor the courts have determined what is sufficient control to make
"private" documents agency records. See id. at 624. O'Connell argues that "[b]y letting
the public eye follow where agency hands reach, a control standard promotes a more
informed electorate and a more accountable government." See id. at 626. In short, if the
government produces part of the material or relies on it in its decision, that material is
part of the decision-making process and should be subject to the FOIA. See id. at 626-
Commentators have pointed out that Congress intended the FOIA's terms to be read
liberally in favor of disclosure "unless the information is exempted under clearly
delineated statutory language."242 But commentators continually argue that judicial
interpretations of the "agency" and "agency records" requirements discussed above have
effectively narrowed the Act's application, allowing government agencies to avoid the
Act's requirements by transferring records out of the government's hands.243 This thesis
essentially agrees with these fears by arguing that more liberal definitions of "agency"
and "agency record" would better square with the spirit of the Freedom of Information
This chapter demonstrates that the courts' narrow interpretation of the terms "agency"
and "agency record" under the Freedom of Information Act effectively creates secrecy
when government agencies choose to privatize important federal services and record-
keeping functions. Without substantial government control over the agency and its
records, they simply will not be subject to public access even if the information they
627. To hold otherwise would actually allow agencies to avoid disclosure by transferring
information to nonagencies. See id. at 627; see also Casarez, supra note 1; Bunker and
Davis, supra note 2.
242See 112 Cong.Rec. 13,654 (1966) (statement of Rep. Rumsfeld, a member of the
House Subcommittee that drafted the FOIA); S. Rep. No. 813, 89th Cong., 1st Sess. 5
(1965); O'Connell, supra note 105, at 620; Bunker and Davis, supra note 2.
243See, e.g., O'Connell, supra note 105, at 617-620, 626-627; Wion, supra note 10, at
1093. O'Connell argues that this can be avoided by fashioning a "control test" that
focuses on the agency's superior ability to control and use the records. See id. at 629-630.
If the agency can access the records and use them in conducting its business, this
approach would allow access even if the agency is not possessing, using or controlling the
records at the time of the FOIA request. See id.
contain is of vital public importance. The next two chapters will show that under the
varying access laws of the 50 states, privatization can also result in important government
information becoming shrouded in secrecy, depending on where in the country one is
located. The final chapter will then analyze and compare the federal and state
approaches, as well as suggest an approach that better comports with the policies behind
public access to information.
FREEDOM OF INFORMATION UNDER STATE STATUTES
All 50 states, as well as the District of Columbia, have embraced the notion of public
access to government information within their respective statutes. These statutes fall
under different chapters of state laws, and are given various names, from Public Records
laws to state Freedom of Information acts. Three states Florida, Montana, and North
Dakota have made access to government a constitutional right.244 All states not only
allow for access to government records, but provide a statutory requirement for access to
government meetings as well. And all state statutes allow for access to government
information based on how they define a public "agency" and public "records" open to
access. In most states, these terms are given specific definitions within the first few
sections of the access laws.
A primary difference among states, however, is how specifically they define the terms
"agency" and "record." While some define the terms precisely, others leave their
definitions more ambiguous and open to judicial interpretation. Others offer very little
definitional guidance at all. In many states, the more specific the definitions, the more
restriction is placed on the courts when applying the public access laws to information in
the hands of private entities. This is because, as discussed below, the state laws often
require a close relationship between the agency and its records to the government, often
244See Fla. Const. art. I, sect. 24; Mont. Const. art. II, sect. 9; N.D. Const. art. XI, sect.
6, discussed infra.
in the form of public funding or control. Unless the specific requisite government
element is present, such as a certain level of public funding or control that would create
the necessary close relationship with government, state courts have little choice but to
deny public access to documents regardless of their interest or importance to the public.
As a result, the continuing government privatization discussed in Chapter II can lead to
many records and functions becoming shrouded in secrecy in many states.
The purpose of this chapter is to analyze the access statutes of all 50 states, as well as
the District of Columbia, in order to compare their definitions of "agency" and "records."
This chapter will therefore provide perspective for the state court opinions construing
these statutes discussed in Chapter V because all state courts must follow the statutorily-
created access laws created by the legislatures. By showing how these statutes vary in
their approaches, it is easy to see that the degree of access one can gain to information in
the hands of private entities will vary depending on where in the country one is located.
While some states require specific elements to be present in order to allow access, others
are more open in their potential application to private entities. Others seem to explicitly
recognize the need to provide access to private entities, while still others choose to
provide very little guidance to the courts, seemingly relying instead on judicial
interpretation. This chapter is meant to group these states into the more specific,
"restrictive" approaches that allow access only in more narrow circumstances, as
compared to the more ambiguous, "flexible" approaches to access that give the courts
more leeway to interpret the access laws. The latter will be discussed first.
No Definition of "Agency" or "Record"
One state statute Alabama's does not seem to provide any definitions of what can
be considered a public "agency" or a public "record." Instead, the Alabama Code Article
dealing with Inspection and Copying of Records deals only with the right of the public to
access "any public writing," without defining what constitutes a "public writing."245 Nor
does the statute determine when an agency that is not technically public in nature could
nonetheless hold records subject to public access. Instead, the statute merely determines
that every citizen has a right to inspect "public writings," and it excepts only those
writings that are registration and circulation records of public schools and universities.246
Therefore, Alabama's statute seemingly allows courts nearly free reign to determine what
can be considered a "public writing," and under what circumstances records in the hands
of private entities should fall within the definition.
Flexible Approaches: The Function of the Entity and the Nature of Its Records
In 13 states, the statutes appear to allow private entities and their records to potentially
be covered under their access laws under certain circumstances. These circumstances
tend to focus on the function the entity is performing, as well as the nature and subject
matter of the records it is holding. If the entity is performing a function for the benefit of
the public, or its records relate to a government function, these states would allow public
access even if the entity is private.
245See Ala. Code Sect. 36-12-40 (1996).
For example, Alaska's statute explicitly allows private entities to be subject to access.
Under its definition of "public agency," the statute includes not only traditional political
and public subdivisions, but also an "instrumentality of the state or a municipality."247
Therefore, this section would potentially allow a private entity to become a "public
agency" subject to access, as long as it is performing a function that could make it an
"instrumentality" of the government, a term seemingly left open to judicial
interpretation.248 Even more explicit, however, is the statute's definition of "public
records," which includes not only those documents in the hands of public agencies, but
also documents in the hands of private contractors acting for public agencies which are
being held for either their "informational value or as evidence of the organization or
operation of the public agency."249 In other words, if the records relate to a public
function, they will be subject to access regardless of the public or private nature of the
agency holding the documents.
Delaware's Freedom of Information Act also appears to focus on the public function
being performed by the entity in order to subject it to public access. In its definition of
"public body," the statute includes not only those agencies created by state government,
but also those entities "appointed" or "otherwise empowered" by a governmental
entity.250 In essence, this would allow those private entities charged with a public duty or
247See Alaska Stat. Sec. 09.25.220(2) (West 1998).
249See Alaska Stat. Sec. 09.25.220(3) (West 1998).
250See Del. Code Ann. tit. 29, sec. 10002(a) (1997).
function to become subject to public access as long as they carry some power previously
held by the government, and are otherwise supported by public funding, expending public
funds, or charged to advise, report to, or make recommendations to public bodies.251
Conceivably, this would include those private entities charged with performing public
functions. Additionally, the Delaware statute's definition of "public record" is similar to
Alaska's in that it includes those records "relating in any way to public business, or in any
way of the public interest, or in any way related to public purposes."252
Also following a "public function" approach is Florida. Florida's statute is similar by
defining as "public" all records made or received "in connection with the transaction of
official business by any agency."253 The term "agency" is defined to explicitly include
private entities "acting on behalf of any public agency."254 In short, the statutes in
Delaware and Florida allow public entities to become subject to access as long as they are
answerable to the government, and the laws go even further in allowing records to
become public as long as they relate in some way to a public function. If the entity is
performing a public function and creating records in connection with that function, the
records will be subject to access.
Georgia's Inspection of Public Records statute can also explicitly apply to private
entities performing a public function. In its definition of "public record," the statute
252See Del. Code Ann. tit. 29, sec. 10002(d) (1997).
253See Fla. Stat. Ann. sec. 119.011(1) (West 1998).
254See Fla. Stat. Ann. sec. 119.011(2) (West 1998).
includes those records which are "received or maintained by a private person or entity on
behalf of a public office or agency," and it explicitly disallows the placement of records
into private hands "for the purpose of avoiding disclosure."255 Like Alaska, Delaware and
Florida, Georgia explicitly allows records to become public if they relate to a public
function, and its statute appears to recognize the potential threat that government
privatization poses to public access by spelling out the policy of continued access in the
law itself. Although a certain amount of public funding is required to create an "agency"
under Georgia's access law,256 the definition of "public record" provides for public access
to private records as long as they are kept on behalf of a public agency, thus preventing
privatization from resulting in secrecy in cases where the entity is performing a public
function by holding public records for a public body.
Louisiana's law is also explicit in its "public function" approach. This law seems to
contemplate access to records held by private entities by stating that all instrumentalities
or other entities "designated as an entity to perform a governmental or proprietary
function" will be considered "public bodies" under the Public Records statute.257 This
statute would also make those entities performing public functions subject to access, and
its lengthy definition of records includes those retained for use in the conduct of public
255See Ga. Code Ann. sec. 50-18-70(a) (1997).
256See Ga. Code Ann. sec. 50-14-1(a)(1)(E) (1997).
257La. Stat. Ann. tit. 44, sec. 1.A.(1) (West 1998).
business or under authority of any public body, thereby allowing records in private hands
to nonetheless remain public due to their subject matter.258
New York's access law also takes a specific "public function" approach. Although the
New York Freedom of Information Law's definition of "agency" appears to apply only to
"governmental entities," it nonetheless seems to include within this definition those
entities "performing a governmental or proprietary function" for the government,
presumably allowing the courts to subject private entities to access as long as they step
into the government's shoes.259 This "contracting out," as explained in Chapter I, is the
most common form of government privatization, and it could potentially create a
"governmental entity" under New York's law once a governmental or proprietary
function is created.
In Rhode Island and South Dakota, the access statutes are similarly explicit in
requiring public access to entities performing a public function or holding records
connected with the public's business. In Rhode Island, the terms "agency" or "public
body" are defined to include entities exercising governmental functions or "acting on
behalf of any public agency."260 Therefore, any entity performing a public function
through a government contract would be subject to access in Rhode Island, and its records
would be public as long as they are made "in connection with the transaction of official
258See La. Stat. Ann. tit. 44, sec. 1.A..(2) (West 1998).
259See N.Y. Public Officers Law sec. 86 (McKinney 1997).
260See R.I. Gen. Laws sec. 38-2-2(1) (1997).
business," or pursuant to the government contract.261 And although South Dakota's
access law does not explicitly include private entities under its definition of "agency," it
does focus on the nature of the public records by making those records made "in
connection with the transaction of official business" subject to access.262 Therefore, as
long as the entity is acting for the public agency or holding records for and connected
with the public entity, the private entity would likely be subject to access in Rhode Island
and South Dakota.
Wyoming also makes public any records "received in connection with the transaction
of public business"263 subject to public access, but it goes one step further than other
states in that it explicitly includes documents relating to the use and disposition of public
income, as well as "all agreements and contracts to which the state or any agency or
subdivision thereof is a party."264 In short, Wyoming's Public Records law appears to
provide for some situations where the government contracts with a private entity,
ensuring that the public nature of the records requested keeps them open to the public
regardless of whether they are in the hands of a private entity.
Montana and New Hampshire also appear to adhere to the "nature of records" and
"public function" approaches, although these statutes are much less clear than all the
states discussed above. In Montana, where the state constitution protects the public's
261See R.I. Gen. Laws sec. 38-2-2(4)(i) (1997).
262See S.D. Codified Laws sec. 1-27-9(2) (Michie 1997).
263See Wyo. Stat. sec. 16-4-201(v) (1997).
264See Wyo. Stat. sec. 16-4-201(v)(A) (1997).
right to know,265 the subject matter of the records also seems to be the focus of the state
access law, although the Public Records law is more ambiguous in that it allows "official
documents" and "public records ... of private writings" to be considered "public records,"
without defining these terms.266 Unlike Louisiana and other states focusing on the
agency's functions and the subject of records, Montana's approach seems to leave much
of the interpretation in the hands of judiciary. New Hampshire also leaves much to
judicial interpretation, in that its only relevant definition included in its Access to Public
Records and Meetings law is its definition of "public proceedings," which could subject
private entities to access by making public all transactions affecting the public, including
those of "advisory committees."267 However, unless the definitions of "public
proceedings" and transactions are read by the courts to apply to a definition of "public
records," the New Hampshire law does not provide a definition that would allow access
to records in the hands of private entities.
Oklahoma's Open Records Act is also vague in defining "record" and "public body,"
and therefore it is difficult to determine when the Act will reach private entities. The Act
appears to focus on the nature of the record's subject matter, and it could reach private
entities by including records received by "representatives" of public entities "in
connection with the transaction of public business, the expenditure of public funds or the
265See Mont. Const. art. II, sec. 9.
266See Mont. Code Ann. sec. 2-6-101 (1997).
267See N.H. Stat. Ann. sec. 91-A:1-a (1996).
administering of public property.""268 Additionally, the term "public body" is defined to
include not only entities supported by public funds, but also those entities "entrusted"
with the expenditure of public funds "or administering or operating public property."269
In other words, if a private entity contracts with the government to provide important
public services, it is possible that the entity could be considered a "representative"
performing a public function and utilizing public funds under the statute, but this is not
clear under the statute.
Lastly, Vermont's statute is also very unclear and open-ended in its definitions. The
terms "public agency" or "agency" apply only to government agencies and political
subdivisions of the state, but the terms "public record" or "public document" do include
those documents "that are produced or acquired in the course of agency business."270
Therefore, depending on how the courts interpret the statute's definition of record,
records kept by a private entity for a government agency could be considered records
"produced or acquired in the course of agency business," as long the courts determine that
the private entities essentially step into the shoes of the government when they sign the
contract. In short, the courts would have to take a "public function" approach, which is
not clearly provided for under the statute.
Restrictive Approaches: Focusing on Public Funds
Eleven state laws require a requisite level of public funding, or the spending of public
268See Okla. Stat. Ann. tit. 51, sec. 24A.3 (1) (West 1998).
269See Okla. Stat. Ann. tit. 51, sec. 24A.3 (2) (West 1998).
270Vt. Stat. Ann. tit. 1, sec. 317 (1997).
funds, in order to subject an entity to access. Under these statutes, one factor, a requisite
level of public funding, is absolutely essential in order for an entity to be considered
public. For example, Arizona's Public Records statute defines "public body" as an entity
"supported in whole or in part" by money from state or local government, or spending
funds provided by a government entity.271 Without this involvement of public funds, the
statute presumably restricts the power of courts to subject private entities to public access,
regardless of the public importance of the information involved.
Similarly, Arkansas' Freedom of Information Act requires public funding in order to
create "public records." Although the statute's definition of "public records" defines
records to include those that "constitute a record of the performance or lack of
performance of official functions," those functions must be performed by a public official
or government agency, or any other agency "wholly or partially supported by public funds
or expending public funds."272 Thus, even if a private entity is performing a public
function, the Arkansas statute would keep its records shrouded in secrecy unless public
funding is involved. This would be true regardless of the information contained in the
records, unless they directly contain information concerning a government agency.273
California's statute pertaining to Inspection of Public Records is even more restrictive
in its definition of "local agency" in that it requires the entity to be supported solely by
271See Ariz. Rev. Stat. sec. 39-121.01(A)(2) (1997).
272See Ark Code Ann. sec. 25-19-103(1) (Michie 1997).
273In other words, a private entity that is simply holding records of a government
agency would be subject to access. But if the records pertain only to that private entity,
regardless of its function, the statute would not likely apply.
public funds.274 Records must be prepared, owned, used or retained by these completely
public entities in order to be considered "public records" under the statute.275 In short, the
statute effectively ensures that a private entity will never be subject to access because by
nature private entities are not completely funded through the public; otherwise they would
be public entities.
Colorado and Michigan also require public funding under their respective statutes. In
Colorado, the Public Records statute simply defines "public records" as those held by any
local government-financed entity for use in the exercise of its functions or in connection
with the receipt or expenditure of public funds.276 Unless the entity can be considered
financed by the government, or the record is connected with public funding, the public
will not have access to the records under the Colorado statute. Michigan also requires an
entity to be "primarily funded through state or local authority" in order to subject it to
access, otherwise the entity is not a "public body" connected with the government and its
records are private.277 Thus, it is the money involved, not the function of the entity or the
information contained in its records, that is crucial under the statutes in Colorado and
274See Cal. Gov't Code sec. 6252(b) (West 1998).
275See Cal. Gov't Code sec. 6252(d) (West 1998).
276See Colo. Rev. Stat. sec. 24-72-202(6)(a)(1) (1998).
277See Mich. Comp. Laws Ann. sec. 15.232(d)(iv) (West 1997).
Although the definition of "public records" under New Mexico's law would include
those records "relating to the public's business,"278 the records must nonetheless be used,
created, received, maintained or held by a "public body" under the Inspection of Public
Records statute.279 The term "public body" is defined to require that the entity receive
any public funding, although the statute does not specify a requisite amount of funding.
This differs from North Dakota's Access to Public Records statute, which is much more
detailed in defining "organization or agency supported in whole or in part by public
funds" to include an organization or agency which has "received public funds exceeding
fair market value of any goods or services given in exchange for the public funds."280
Under this statute, the entity must be getting public support beyond the market value of
its output, or be expending public funds, in order to be considered a "public entity."281 In
short, unlike New Mexico's statute, North Dakota requires a more specific level of public
funding in order to create a "public entity. "282
Similar to North Dakota, Wisconsin's Public Records and Property statute requires a
specific amount of public funding in order to justify access to public records. Under its
definition of "authority," the Wisconsin law includes not only governmental entities, but
also those entities receiving more than 50 percent of their funding from governmental
278See N.M. Stat. Ann. sec. 14-2-6(E) (Michie 1997).
28See N.D. Cent. Code sec. 44-04-17.1(9) (1997).
281See N.D. Cent. Code sec. 44-04-17.1(11)(c) (1997).
entities and which provide public health or safety services for a governmental unit.283 In
other words, not only must the entity meet the requisite funding requirement under the
statute, but it also must be providing a specific service for the government. Although the
service requirement could be met in many cases where the entity is performing a service
for the public's benefit, such as fire protection, the funding requirement will often not be
met in cases where the entity is purely private, even though the information is of
importance to the public. Thus, Wisconsin's law allows much information to become
secret once the government contracts with a private entity.
South Carolina also focuses on entities "supported in whole or in part by public funds
or expending public funds" in order to subject them to the reach of its Freedom of
Information Act, although it is not specific in the amount of funding necessary.284 Under
its definitions of "public body" and "public record," the act requires some public support
or expenditure in order to subject private entities to public access, but it does not provide
the necessary amount of public funding.285 This is similar to Virginia's Freedom of
Information Act, which also requires the entity to be "supported wholly or principally" by
public funds in order to subject it public access, without providing the requisite amount of
283See Wis. Stat. Ann. sec. 19.32(1) (West 1997).
284See S.C. Code Ann. sec. 30-4-20(a) (Law. Co-op. 1997).
285See S.C. Code Ann. sec. 30-4-20(c) (Law. Co-op. 1997).
286See Va. Code Ann. sec. 2.1-341 (Michie 1997). The Virginia Act defines "public
body" under its meetings section, and then carries that definition over to the access to
public records section. See id. It also requires that records be prepared, owned, or in the
Lastly, in Utah, not only does the access statute specifically exclude "private
providers" from its definition of contractors with the government that would otherwise be
subject to access, 287 it also requires an entity to be funded or established by the
government to carry out public business in order to create a "governmental entity."288
Additionally, the statute explicitly excludes private entities from the reach of its
definition of "record" by requiring preparation, ownership, or possession by a
governmental entity, and by excluding records legally owned by an "individual in his
private capacity."289 In other words, if the records are owned by a private individual, they
would be private even though they contain information of importance to the public.290
Restrictive Approaches: Close Nexus with Government
The remaining 26 states require close ties between the entity and its records with a
government agency in order to subject them to public access. While these statutes differ
in how they define this close nexus with the government, all of these laws require some
specific connection with the government for public access. Many of these states follow a
somewhat "hybrid approach," requiring public funding or close government control to
possession of a public body in the transaction of public business in order to create
"official records." See id.
287See Utah Code Ann. sec. 63-2-103(5)(b) (1997).
288See Utah Code Ann. sec. 63-2-103(9)(b) (1997).
289See Utah Code Ann. sec. 63-2-103(18)(a) (1997).
290This exclusion for private individuals does not, however, include private corporate
entities. See id. Therefore, many private entities would not be excluded under this
subsection, but they could still be excluded due to a lack of public funding.
create a "public agency," but allowing records to be public if they are closely connected
with public business. Other statutes focus solely on government control, requiring the
government connection to be so close that the government essentially completely controls
the entity and its records. Still others allow a more tenuous connection with the
government, with a lesser degree of government control necessary. This connection
could be established, for instance, if the government has direct access to the records. But
in all of these statutes the courts are essentially restricted in subjecting private entities to
public access, unless the necessary specific connection with the government is present.
The states in this section are divided into the "hybrid approach" and "control approach"
states, in order to separate those states that maintain public funding as a factor in
determining whether the necessary connection exists from those that focus solely on
The "hybrid" approach. This approach differs from the "public funding" approach
discussed above in that in this approach, public funding is a factor to be considered in
determining whether the necessary government connection exists, but it is not the sole
determinative factor. For example, the public information statute in Texas appears to be
less restrictive in its focus on public funding, but it requires ownership or control by the
public agency to provide the necessary connection between the records and the
government. Under its definition of "public information," the Texas statute appears to
require a close connection with a governmental body in order to create public
information. The "public information" definition includes information collected,
assembled or maintained for a governmental body "in connection with the transaction of
official business," but the governmental body must own the information or have a right of
access to it."291 In other words, if the government has continuous power over the
information, it would be considered public even if in the hands of a private entity, as long
as it relates to the public's business.
However, under the definition of "governmental body," the Texas statute requires that
an entity be "supported in whole or in part by public funds."292 Therefore, the entity has
to either be publicly funded or its records must have a close connection with the
government in order to subject them to public access. While the statutes in the "public
funding" section focus primarily on the funding connection to government, the Texas
statute allows access if the government has significant power and control over the
information. Texas' approach could therefore be considered a "hybrid approach" in that a
private entity will only be considered public if it is funded publicly, but its records may
nonetheless be public if they are owned or controlled by a public entity.
Similar to Texas, Illinois' Freedom of Information Act seems to follow a "hybrid
approach" in determining when an entity and its records will be subject to public access.
In its definition of "public body," the Act limits the covered entities to "subsidiary
bodies" of the government which are "supported in whole or in part by tax revenue. "293
Thus, not only does the law require the entity to be close enough to the government so as
to become a "subsidiary," but it also requires an undefined amount of funding in order to
291See Tex. Gov't Code Ann. sec. 552.002(a)(2) (West 1997).
292See Tex. Gov't Code Ann. sec. 552.003(x) (West 1997).
293See Smith-Hurd Ill. Comp. Stat. Ann. sec. 140/2(a) (West 1998).
provide the necessary connection with the government. The Act's definition of "public
records" also requires a close connection with an agency as the records must be under its
ownership or control, but it also appears to include reports or studies prepared for the
public body, or documents prepared by independent consultants or contractors for the
public body.294 Thus, the Act is somewhat of a "hybrid" in that it requires a close
connection with the government in order to create a "public body," but it recognizes that
some records can be so related to government functions so as to make them, in effect,
Other statutes also follow a "hybrid approach" that requires either public funding or
control to create a public agency. Additionally, government control is necessary to create
public records. If funding or control are not present, the records will not be subject to
public access. In Iowa, the Open Records law defines "government body" to include only
those entities that are a part of the government because they are delegated government
functions, or those entities supported in whole or in part by tax revenue.295 There must be
a significant government connection, either through funding or control, in order to subject
the entity to public access, and the records must belong to such an entity in order to create
"public records" under the statute.296
294See Smith-Hurd Ill. Comp. Stat. Ann. sec. 140/2(c) (West 1998).
295See Iowa Code Ann. sec. 22.1(1) (West 1997).
296See Iowa Code Ann. sec. 22.1(3) (West 1997).
Similarly, the Kansas statute requires either government control creating an "alter ego"
of government, or a certain amount of government funding, to create a "public agency."297
However, this statute goes even further by explicitly excluding those records owned by a
private person or entity which are unrelated to government functions.298 Without funding
or control, Kansas follows Texas, Illinois, and Iowa by not allowing access to entities that
are not directly a part of government.
Kentucky's Open Records statute is even more detailed in its definition of "public
body" in that it requires one of several elements to be present in order to satisfy the
necessary connection to a government body. All of these elements provide either the
necessary government control or public funding to allow access to the records. Under the
definition of "public agency," the Kentucky law requires connection in the form of
creation by a government entity, at least 25 percent funding from a government entity, a
majority of the governing board controlled by a government entity, or a subcommittee
created by a government entity.299 Additionally, the statute's definition of "public record"
297See Kan. Stat. Ann. sec. 45-217(c)(1) (West 1996), defining "public agency" as
either a political subdivision or instrumentality of a political body, or an entity "supported
in whole or in part by public funds."
298See Kan. Stat. Ann. sec. 45-217(2) (West 1996), stating that a "public record" shall
not include "records which are owned by a private person or entity and are not related to
functions, activities, programs or operations funded by public funds..." In other words,
the statute seems to recognize that private entities could possess records that are related to
public functions, and therefore these records should be subject to access but at the same
time it requires that the records be under an agency's control in order to be considered
"public records." The statute therefore requires a sufficient government nexus for public
299See Ky. Rev. Stat. Ann. sec. 61.870(1)(h)-(j) (Baldwin 1998).
requires that the entity's record be owned, in the possession of, or in the control of a
public agency, and it adds an additional requirement that any record owned by a publicly-
financed entity pertain to a government-funded operation.300 Thus, the Kentucky statute
is detailed in its requirements that both the entity and its records are connected to and
pertain specifically to a government entity, through control or funding that creates a
public agency, and government control that creates a public record.
Pennsylvania also appears to be a "hybrid" state in its requirement that an agency be
under close government control. However, the statute states that records pertaining to
public business are subject to access. The Pennsylvania law requires its "agencies" to be
political subdivisions of the state, but its law goes even further by requiring that the entity
be "created by or pursuant to a statute which declares in substance that such organization
performs or has for its purpose the performance of an essential public function."301 In
other words, the entities have to not only be closely connected with the state, but the
statute is even more restrictive in allowing access to only those entities performing the
most important functions. However, the statute then appears less restrictive in its
definition of "public record," allowing access to records and contracts "dealing with the
receipt or disbursement of funds by an agency" or its performance of a public function.302
300See Ky. Rev. Stat. Ann. sec. 61.870(2) (Baldwin 1998), which excludes those
records owned or maintained by a body funded at least 25 percent by the public that are
not "related to functions, activities, programs or operations funded by a state or local
301See 1998 Pa. Legis. Serv. sec. 66.1(1) (1998).
302See 1998 Pa. Legis. Serv. sec. 66.1(2) (1998).