Law and Contemporary Problems: Sidney A. Shapiro and Rena I. Steinzor, The People's Agent: Executive Branch Secrecy And Accountability In An Age Of Terrorism, 69 Law & Contemp. Probs. 99 (summer2006)

Cited: 69 Law & Contemp. Probs. 99
[*pg 99]

THE PEOPLE'S AGENT: EXECUTIVE BRANCH
SECRECY AND ACCOUNTABILITY IN AN AGE OF TERRORISM

SIDNEY A. SHAPIRO^* AND RENA I. STEINZOR^**

I. INTRODUCTION

II. AGENCY AND ACCOUNTABILITY
   A. The Principal-Agent Relationship
   B. Agency Costs
   C. Tools to Reduce Agency Costs
   D. Transparency

III. A TALE OF THREE STATUTES
   A. The Freedom of Information Act
   B. The Federal Advisory Committee Act
   C. Critical Infrastructure Information Act

IV. CONCLUSION

FOOTNOTES

I. INTRODUCTION

The increase in government secrecy is an important and troubling policy trend. Although the trend predates it, the movement toward government secrecy has accelerated dramatically since the 2000 presidential election. The Bush Administration has initiated significant legislative and administrative changes that have expanded the government's authority to operate behind closed doors.¹ Vigorously defending its preoccupation with secrecy as the necessary prerogative of a strong and effective executive branch, the Administration invokes the "war on terrorism" to justify withholding information that was routinely disclosed by past Presidents.

Claims that the executive branch needs extensive secrecy to operate effectively are troublesome because of the important role transparency plays in the American constitutional system of checks and balances. When secrecy becomes sufficiently pervasive, it becomes difficult, even impossible, for Congress and the public to determine what is going on in the executive branch. Government failures are hidden and the public interest suffers. Indeed, it is not [*pg 100] an exaggeration to say that pervasive secrecy can fatally undermine the structure of our constitutional government by allowing the executive branch to withhold crucial information from the other two branches and, as important, a free press.

Nevertheless, absolute transparency is neither a realistic nor an appropriate goal. The release of some types of information can do more harm than good. Detailed information about the precise location of chemical stockpiles that could be used by terrorists to wreak havoc is one example, but there are other areas in which the public interest may favor nondisclosure, such as protecting legitimate trade secrets that have never been circulated publicly or preserving the confidentiality of information that would compromise individual privacy rights. The nation's premier open government law, the Freedom of Information Act (FOIA),² recognizes a number of situations in which Congress has determined that the public interest is best served by nondisclosure.³

The difficult public policy issue, of course, is striking an appropriate balance between openness and secrecy, and there is a considerable literature debating the merits of particular tradeoffs.⁴ In Washington, however, this careful and thoughtful debate increasingly seems beside the point as the executive branch ignores both the spirit and the letter of FOIA and similar laws. It is more and more difficult to pry timely responses from agencies and departments to legally sound FOIA requests, and more and more information is withheld entirely, often on specious grounds.⁵

Despite concerns expressed across the political spectrum by commentators and advocates,⁶ the public seems generally apathetic regarding these developments. One reason for this apathy appears to be the Administration's use of the war on terrorism as the "poster child" for its efforts to avoid accountability in policy areas that have little or nothing to do with national security. This characterization places public interest advocates in the position of making the case for open government on a more demanding field of battle [*pg 101] where officials argue that anyone requesting information must demonstrate affirmatively that disclosure will be absolutely useless to terrorists.⁷

The case for open government is based on political principles embraced by the framers of the U.S. Constitution. Transparency plays an important role in our constitutional system of checks and balances, not least because it is the condition precedent for a free press. Although thoughtful people from the framers to modern-day political scientists acknowledge that at times the government must act in secret, an underlying assumption of American democracy is that the people have a right to know about business transacted in their name.

This article seeks to bolster these arguments by applying "agency theory," a school of thought that was developed by the law and economics literature, to the question of how much secrecy is too much. Although agency theory is most often used to analyze private sector economic relationships, commentators have also applied it to the analysis of methods for holding legislators and executive branch officials accountable to the public. This paper extends this literature by using agency theory to evaluate the impact of burgeoning secrecy on the likelihood that executive branch officials will engage in faithful and forceful implementation of statutory mandates, particularly in the arenas of protecting public health, safety, and natural resources.

The paper begins by explaining why agency theory is useful in analyzing how the government should resolve conflicts between transparency and other important public policy objectives. This analysis yields the following insight: whereas wise policymaking requires a balancing of competing interests in secrecy and openness, such decisions must be made by dispassionate and authoritative officials who have no personal stake in whether the information is ultimately disclosed.

The paper then turns to an examination of the three primary statutes governing information disclosure and transparency: FOIA, the Federal Advisory Committee Act (FACA),⁸ and the Critical Infrastructure Information Act (CIIA),⁹ which was passed as part of the Homeland Security Act of 2002.¹⁰ All three statutes are involved in the trend toward greater secrecy. Judicial [*pg 102] interpretations and executive branch hostility have narrowed the application of FOIA and FACA. The CIIA, passed in the wake of the September 11, 2001 terrorist attacks, unreasonably ignores the tradeoffs between greater government secrecy and transparency in favor of maximizing government secrecy. The cumulative impact of these changes is to make it more difficult to hold executive branch officials accountable when their self-interest lies in keeping information confidential in order to prevent political embarrassment. The principle of maintaining a balance between public accountability and national security is eroded, as is the foundation of the U.S. constitutional democracy.

Congress should amend FOIA and FACA to reverse judicial interpretations that have weakened these laws by restoring the open government requirements that Congress intended to impose in enacting them. And Congress should limit the secrecy protections of the CIIA to situations in which there is a reasonable likelihood that transparency presents a tangible risk to the nation's security. In isolation, these changes will not be enough to forestall executive branch abuses. But they will begin to reverse the harmful trends we have identified by making it easier for the public and the free press to expose illegitimate secrets.

II. AGENCY AND ACCOUNTABILITY

Agency theory is commonly used to analyze contractual arrangements designed to address the risk that an agent, such as a corporate manager, will serve his or her own interests, rather than the interests of the principal. Applying agency theory to the relationship between the public and the executive branch can explain why it is considerably more difficult to control agents in political contexts than in corporate contexts and emphasizes the importance of maximizing government transparency.

A. The Principal-Agent Relationship

An organization is formed in response to collective-action problems that constrain individuals from undertaking common action that is to their mutual benefit.¹¹ Persons who form the organization, the "principals," inevitably depend on "agents" to achieve the gains from collective action, and the expected mutual gains can be lost if an agent's self-interest diverges from a principal's goals.¹² The difference between results achievable by an agent acting solely in the interest of the principal and actual results is described by [*pg 103] economists and other social scientists as an "agency cost."¹³ Principals can combat this downside and avoid agency costs if they have the tools and resources needed to prevent their agents from getting sidetracked in the pursuit of the principal's interests.

The core concepts of agency-cost problems in the corporate context¹⁴ can be extrapolated to the government arena by substituting "the people" -- defined as all of the nation's citizens -- for a corporation's shareholders. In a constitutional democracy like ours, the people incur analogous agency costs concerning the performance of the executive branch.

The people "employ" the executive branch as their agent in the two ways depicted by this diagram.¹⁵ In some circumstances, the executive branch acts [*pg 104] directly on behalf of citizens, invoking the President's authority under Article II of the Constitution.¹⁶ Article II constitutes the agency contract between the people and the President. In other instances, citizens "employ" Congress as their agent to enact legislation. The Congress in such instances is acting under the powers granted to it by Article I of the Constitution.¹⁷ Article I constitutes an agency contract between the people and Congress. Congress, in turn, performs a subdelegation, transferring power to the executive branch to implement the legislation. The agency contract involved in this subdelegation is the statute enacted by Congress.

The congressional role can be analogized to the function served by the board of directors of a corporation. The board serves as the agent of the stockholders. But the board hires managers to implement its policies. In the same manner, Congress turns to the executive branch to implement certain congressional policies. Although the managers are chosen by the board, they are still "agents" of the stockholders.¹⁸ And although, through statutes, Congress directs the executive branch, the executive branch is still the agent of the people.

Speaking of the American people as a principal, or even a group of principals, is admittedly a simplification because people have diverse interests, even with respect to collective action. Nevertheless, when legislation is enacted, it can be regarded as an expression of the people regarding the nature and scope of the collective action they seek. This characterization is useful even though the enacted legislation may imperfectly reflect the will of the citizenry (itself an agency-cost problem). The legislature in our constitutional system is "understood to serve -- even if imperfectly and not without significant moments of contestation -- as the privileged institutional expression of the capacity of the 'people' or the 'nation' to rule itself."¹⁹ Characterizing the executive branch as an agent of the people is another simplification because it does not take into account the reality that the executive branch is comprised of numerous and varied centers of political power that engage in a constant struggle for control with each other. Nevertheless, viewing the executive branch as the people's agent in carrying out the mandates conferred by the people's surrogate, that is, [*pg 105] the Congress, correctly reflects the legal roles assigned by the Constitution²⁰ and by the Administrative Procedure Act (APA),²¹ which views the executive branch as an institution that takes "final agency action" pursuant to statutes that are then subject to review by the courts.²²

B. Agency Costs

The implementation of legislation by the executive branch almost always creates the potential for agency costs.²³ Professor Peter Lindseth, who has also written of the relationship between these branches in terms of agency theory, explains, "[t]he reality of the situation will of course be that the agent enjoys some measure of autonomy -- often a significant amount, in fact [as a result of] vague delegations as well as the inherent limitations of the principal's capacity for supervision and control."²⁴ The task of a principal, Lindseth continues, "is to ensure that the agent will not pursue its own interests rather than those of the legislature -- what economists call the 'agency cost problem.'"²⁵

Much of the time, executive branch officials faithfully implement regulatory legislation.²⁶ At times, however, several factors converge to lead them off this path.²⁷ Administrators may fail to serve the public's interests, as defined by statutory mandates, because their efforts are not very effective or because they prefer not to work very hard. In addition, administrators, including the President, are vulnerable to the demands of special interests that conflict with effective implementation and enforcement of such mandates. Officials, in particular political appointees, may be tempted to yield to special-interest pleading, especially when satisfying those interests serves the officials' own goals. For example, the President must seek reelection supported by considerable fundraising. Other officials are tempted to gain security, develop future job prospects, or follow their own policy preferences. Whatever the cause, those who would benefit from the regulations will end up shouldering the costs of failing to discover when agents do not represent them faithfully. Those [*pg 106] costs can include fraud and other forms of economic waste²⁸ and can even be life-threatening.²⁹

The potential for agency costs is greater when Congress, beset by intense controversy and the clashing of politically and economically powerful forces, enacts ambiguous legislation that does not resolve significant issues but instead redirects them to the administrative process, thereby delegating substantial discretion to administrators. The likelihood of drift is greater in policy arenas where the results of an agent's actions are less obvious and government activities are harder to scrutinize.

The arena of health and safety regulation is a prime breeding ground for agency-cost problems because such regulation typically involves ambiguous delegation and a lack of transparency. Environmental policymaking, in particular, is highly contested terrain involving frequent clashes between powerful business interests and highly motivated environmental organizations.³⁰ The results of an agent's actions are difficult to discern because environmental exposures to harmful pollutants are both pervasive and poorly understood.³¹ Very often, the adverse effects of such exposures are diseases (for example, cancer), that have long latency periods, obscuring the connection between industrial practices, regulation, exposure, and health consequences. Or such adverse effects are very subtle and difficult to detect, such as the loss of a few IQ points. Executive branch officials coping with such issues are quickly submerged in highly technical debates, losing their footing in response to an onslaught of advocacy by regulated industries and public interest advocates.

The success of efforts to track such complex debates and to explain to the public why and how its agents have faltered depends on the active involvement of a free press. Reporters have a difficult time sorting through the morass of technical claims and counterclaims even when their organizations give them the [*pg 107] luxury of sticking with a story for more than one news cycle. In the wake of the tragedies that began on September 11, 2001, those conditions and opportunities have become scarce while other crises, at home and abroad, compete for the media's attention.

C. Tools to Reduce Agency Costs

An agency-cost problem can arise in any principal-agent relationship. Every principal therefore has an incentive to use tools to reduce agency costs up to the point at which the expense of using these tools exceeds the agency costs that the tools prevent. A principal can reduce agency costs by negotiating highly specific contracts that explain expectations and set standards for performance and then by monitoring the agent's actual performance under the contract. In addition, the principal can use financial incentives to align the self-interest of the agent with that of the principal. The use of these tools can be difficult with either private or public agents, but it is generally easier to reduce agency costs in the private sector than the public sector.

1. Contracting
A principal can seek to control an agent through comprehensive contracting. In economic contexts, there is an actual contract. In the public sector, as suggested earlier, a statute functions as a kind of employment contract. The contract specifies the duties of the agent, implicitly stipulating that the agent's success will be measured against these goals. In both the private and public contexts, however, the principal is subject to "bounded rationality" -- significant time, resource, and cognitive restraints that limit its capacity to anticipate and specify how the agent should act when new contingencies arise.³² Because most principals are usually subject to at least some level of bounded rationality, comprehensive and definitive contracting is normally impossible except in simple transactions.³³

As difficult as principals in the private sector find it to reduce their agency costs through contracting, the public sector encounters significantly more daunting problems for two reasons. First, the problems that Congress tackles are typically much more complex and affect a far larger and diverse set of stakeholders than the typical corporate transaction.

[T]he basic separation between polity and economy has always, even amongst the most confirmed libertarians, left a residual of activities to be undertaken by government because of the inherent difficulty that arose from the public good attributes, free riding and costly information of certain types of activity . . . . Those that can be readily handled by individual or small group bargaining don't need to be placed on the public agenda . . . .³⁴

[*pg 108]

Further, the dimensions of what constitutes success are infinitely more intricate, incorporating a broad range of intangible values and straying far from the profit motive that is at the heart of the corporate mission. In health and safety regulation, these difficulties are compounded by the need to take action to prevent harm without the luxury of definitive scientific evidence regarding how, when, and why the harm will occur.³⁵

Second, the private sector generally finds it less costly to switch agents.³⁶ Principals may find it difficult to remove corporate officers who fail to serve the principals' interests, but the private sector principal has the option of selling stock in one corporation and replacing it with stock in another corporation that has more reliable agents. By comparison, the public cannot readily or inexpensively replace executive branch agents because there are both timing and severability problems. The President runs for election only every four years and voters have no direct way of either voting for or removing members of the Cabinet or other administrators. Unless there is sufficient public outcry, the President is reluctant to fire administrators revealed to have engaged in malfeasance or nonfeasance because such behavior reflects poorly on the administration.

2. Financial Incentives
In addition to having a significantly easier time negotiating highly specific contracts that define expectations and set standards for success, private sector principals can rely to a greater extent on financial incentives to reduce agency costs.³⁷ Although the economic performance of a firm is not a perfect indicator of an agent's loyalty or ability given economic factors beyond managerial control, profitability offers private-sector principals a reasonable way to measure the performance of their agents. As the most obvious example, the compensation of a manager can be based on the economic performance of the firm. In all but the most dysfunctional firms, regular performance evaluations make it clear, from the top of the organization to the bottom and up again, what accomplishments will be rewarded and what outcomes viewed with disfavor. Even these concrete and time-tested incentives are imperfect. Turnover in executive suites may be routine, but many companies retain their leadership long after they have suffered severe financial reversals.

Congress faces greater, even insurmountable, hurdles in using financial rewards to motivate its bureaucratic agents for several reasons. Not only are the goals set for bureaucrats significantly more complex and amorphous, but it is much more difficult to measure public sector agency costs because of the [*pg 109] difficulty of specifying and measuring what constitutes good performance.³⁸ If we are uncertain about the nature and extent of the risks posed by toxic pollution, for example, how can we measure the after-the-fact success of measures designed to avoid such risks?

The closest Congress has come to mimicking private sector financial incentives is the 1993 Government Performance and Results Act (GPRA),³⁹ one of several "good government" laws designed to improve public sector managerial performance.⁴⁰ The GPRA requires agencies and departments to develop performance plans and then report on their progress in achieving the goals they have identified. Embraced by both parties,⁴¹ but most prominently used by conservatives in Congress to excoriate agencies they dislike,⁴² the Act provides a vehicle for cutting the budgets of agencies that do not perform, in theory the public sector equivalent of the performance-based bonuses offered private-sector managers.

This "cure" is often far worse than the disease. If the program is a desirable one, budget cuts, even if they somehow discipline managers, will also make it more difficult for an agency to deliver the benefits of the program to the public. As frustrated as Congress may become with agency performance, no one would suggest that the nation should stop carrying out such essential functions as processing immigration applications or collecting taxes because Congress is disgusted with the bureaucrats who administer such programs.

As a management tool, the GPRA has another significant disadvantage. If Congress enacts significant budget cuts in response to a recommendation by the President based on a performance review under the GPRA, it will be more difficult for the agency to implement its statutory mission. If the budget cuts have the impact of dismantling or hindering statutory programs, the budget cuts will constitute a de facto repeal of the statutory mandate. Even though the mandate may remain on the books, the agency will be unable to implement it in any meaningful way. This repeal, however, will have occurred without [*pg 110] Congress's actually passing a bill that repeals the mandate, further undermining the system of checks and balances and legitimate lawmaking. Since the budget process is considerably more opaque than standard lawmaking, a significant cut in an agency's budget allows the President as well as Congress to avoid accountability for rewriting popular laws.

3. Monitoring
A principal can seek to ensure the loyalty of an agent through contracting, but the principal must also monitor the performance of the agent and determine how well the agent is actually achieving the principal's interests. The effort required to monitor the agent's exercise of its contractual duties will increase in direct proportion to the incompleteness of the underlying contract because the principal will not be able to simply check whether the agent has complied with a detailed contractual term. Instead, the principal must identify the agent's actions and weigh the extent to which they are consistent with the principal's broader and often ill-defined objectives.

Such monitoring is made more difficult by information asymmetries.⁴³ The agent is often in control of information the principal needs to judge that person's performance. As a result, the agency cost problem is "most severe when the interests or values of the principal and agent [have] diverge[d] substantially, and information monitoring is costly."⁴⁴

Both private and public principals are subject to the problem of information asymmetries. The regulation of financial markets, for example, includes requirements that mandate the disclosure of certain types of financial information and regulate the conduct of persons who produce that information. These requirements assist individuals in deciding which agents to hire (that is which firms to invest in) and in monitoring the performance of those agents. The monitoring that occurs in the public sphere is somewhat more complicated. Under the United States' constitutional framework, both judges and members of Congress perform monitoring functions.

a. Judges.
Administrative law offers one crucial method of monitoring the behavior of administrative agents in light of information asymmetries. As one author notes, the "raison d'être [of administrative law] is to manage the inevitable agency autonomy that comes with the legislative delegation and thus to reduce 'agency costs.'"⁴⁵ The APA authorizes persons "adversely affected" by final agency action to challenge that action in federal court.⁴⁶ The APA also authorizes federal judges to set aside agency actions that are inconsistent with a statutory mandate⁴⁷ or that are arbitrary and capricious.⁴⁸ The Supreme Court [*pg 111] has interpreted this last provision to require an agency to offer a reasoned explanation for its actions.⁴⁹ The courts also expect an agency to make any scientific tests or other technical data on which it relied in formulating a proposed regulation available to the public.⁵⁰

The threat of an adverse judicial decision creates a strong incentive for administrators to disclose the information on which they relied and to defend their decisions as nonarbitrary public policies consistent with the agency's statutory mandate. This system is also imperfect. For one thing, a court's capacity to require an agent to implement congressional intentions depends on how clearly and precisely Congress has stated its intentions concerning how the agent is to perform.⁵¹ But Congress has difficulty writing such specific instructions.⁵² Even when congressional instructions to an agency are relatively precise, judges may still choose to interpret a statute in a manner not intended by the enacting Congress in order to fulfill their own policy and political goals. Congress always has the alternative of overruling judicial interpretations by amending the statute, but the hurdles to enacting corrective legislation are many.

Moreover, not all agency actions are subject to judicial review. The judicial review provisions of the APA do not apply when Congress has passed legislation expressly precluding such review⁵³ or when Congress does not establish legal standards under which review can take place.⁵⁴ Further, the beneficiaries of regulation can seek review only if they have standing, which requires that a plaintiff have a sufficiently direct interest in the lawsuit's outcome and that the lawsuit presents a "case and controversy."⁵⁵ In some instances, the Supreme Court has interpreted this requirement to forestall judicial review in cases in which the plaintiff sought to hold administrators accountable for questionable policies.⁵⁶

[*pg 112]

Finally, but hardly least of all, not all actions subject to judicial review will be challenged in a lawsuit. This problem is especially acute when a decision affects a very large number of people adversely, producing a large collective injury, but no person suffers an individual injury large enough to have a sufficient incentive to sue. For the same reason, individuals will not have a sufficient incentive to form a group to bring the lawsuit. The cost of organizing the group will exceed the individual benefit to any one individual.⁵⁷

b. Congress.
Congress monitors agency performance through two types of oversight. In "police patrol" oversight, Congress routinely oversees agency action, investigating whether the executive branch is faithfully implementing legislation.⁵⁸ When Congress engages in audits, investigations, and other forms of oversight, the goal is to make "it very difficult for agencies to strategically manipulate congressional decisions by presenting a fait accompli, that is, a new policy with already mobilized supporters."⁵⁹ Congress also engages in "fire alarm" oversight to spot what it misses in "police patrol" oversight,⁶⁰ depending on third parties to spot significant deviations from statutory contracts by the Branch, and then reacts to "fire alarms" pulled by disaffected interest groups.⁶¹

Significant agency costs remain despite legislative oversight. Members of Congress often have strong incentives to shirk this role. Oversight is less rewarding in terms of increasing the chances for members' re-election than other activities, such as constituent casework. Partisan concerns also play a role. For example, Republicans holding a majority of seats in both houses of Congress may slack off in order to protect a Republican President from the embarrassing disclosures in committee hearings. Finally, the efficacy of "fire alarm" oversight depends on how readily third parties can overcome information asymmetries and find out about administrative failures. The trend toward more secrecy makes it more difficult for third parties, that is citizens, to engage in this function.

D. Transparency

The importance of government transparency should now be apparent. Judicial review and legislative oversight are important methods of monitoring executive branch performance, but they leave much to be desired. In light of these imperfections, the public has a significant incentive to monitor the performance of the executive branch. Information asymmetries erect substantial barriers to such monitoring. By removing information asymmetries, laws that mandate disclosure make it feasible for the public to hold the [*pg 113] executive branch accountable. The cost of implementing transparency laws, while expensive, is likely to be far less expensive than the agency costs that occur if the executive branch fails to perform its statutory duties effectively.

Although transparency lies at the heart of reducing agency costs in the public sector, making government more transparent will not necessarily lead to significant or immediate reductions in agency costs. Even if the public is aware of executive branch failures, collective action problems may prevent people from acting on this information. Nevertheless, transparency is a prerequisite for any effort to hold the executive branch more accountable. As Congress implicitly concluded when it passed the open government statutes, excessive government secrecy invariably translates into increased agency costs.⁶² These laws, however, have been undermined by more recent legislation that -- apart from any effect on agency cost -- in fact supports such greater secrecy.

III. A TALE OF THREE STATUTES

Laws mandating executive branch transparency typically require disclosure but contain exemptions that balance the benefits of transparency against the unacceptable costs of revealing secrets that could harm the public interest. Three important statutes -- the Freedom of Information Act (FOIA),⁶³ the Federal Advisory Committee Act (FACA),⁶⁴ and the Critical Infrastructure Information Act (CIIA)⁶⁵ -- typify such laws. The first two statutes have transparent government as their primary goal, while the third embodies the most pointed, recent statutory backlash against open government.

An agency-cost analysis of these laws reveals that the trend toward greater government secrecy involves a shift away from the original FOIA model of disclosure, making it significantly more difficult for the public to monitor the executive branch. This shift has accelerated dramatically in the wake of the tragedies that began on September 11, 2001. All three branches of government have contributed to this rejection of the FOIA model. The judiciary has interpreted FOIA and FACA in ways that weaken the presumption-of-disclosure model. Congress has not only acquiesced in these changes, but, ostensibly in response to the "war on terrorism," passed the CIIA. President Bush supported passage of the CIIA, and his Administration has vigorously sought to increase the level of government secrecy through administrative [*pg 114] interpretation,⁶⁶ including interpretations of FOIA and FACA that make it easier for the executive branch to avoid outside scrutiny.

A. The Freedom of Information Act

1. FOIA Requirements
FOIA is by far the most important law mandating transparency in government.⁶⁷ The statute has been under sustained attack, and its implementation weakened substantially since President Bush took office.

FOIA is a paradigmatic disclosure statute for four reasons.⁶⁸ First, the statute established a strong baseline for disclosure by allowing access to information without requiring the requestor to demonstrate any need to know. The information covered by FOIA includes documents (broadly defined) generated by the government, or by private-sector actors, so long as they are in the possession (or custody) of the government. Second, the Act imposes strict deadlines for the government's response to information requests. Third, it imposes the burden of justifying a decision to withhold information on the government and requires that all such decisions fit under one of nine exemptions specified in the statute. Decisions to withhold must be made with respect to the status of the information at the time the request is made; that is, information that once was a trade secret may have lost this protection given its public circulation in the period since it was submitted. Finally, the Act subjects decisions to withhold information to judicial review and instructs the judiciary not to defer to an agency's interpretation of the statute but rather to make a decision regarding the legitimacy of nondisclosure de novo.

a. Disclosure baseline.
The assumption that all information must be disclosed without demonstrating a need to know was revolutionary at the time FOIA was enacted in 1966. Prior to its enactment, government agencies and departments had unrestricted authority to withhold information they either had developed or had submitted to them by a wide range of private sector actors.

FOIA mandates two types of disclosure that did not previously exist. It imposes affirmative obligations on executive branch agencies and departments to make specific records available in public reading rooms, including "final opinions, including concurring and dissenting opinions, as well as orders, made in the adjudication of cases"⁶⁹ and any other information that "the agency determines have become or are likely to become the subject of subsequent [*pg 115] requests."⁷⁰ The Electronic Freedom of Information Act Amendments of 1996 further mandate that each agency establish a website for information that FOIA requires the agency to make available in a reading room.⁷¹

Beyond this specific information, federal departments and agencies must release records they possess in response to requests from any "person."⁷² Requestors must provide a "reasonable" description of the information sought and the request must correspond to the agency's published rules concerning the "time, place, fees (if any), and procedures to be followed."⁷³ Agencies may charge fees for the resources used in searching for the information and for duplicating the records.⁷⁴

b. Time deadlines.
The government custodian of requested records must respond to valid requests within twenty days, either disclosing the record or explaining why it is exempt.⁷⁵ However, as a practical matter, these deadlines are often ignored, and, even if the requestor brings a lawsuit to enforce such obligations, the courts grant liberal extensions in response to so-called Open America motions.⁷⁶ Nevertheless, the time deadlines are judicially enforceable, which means agencies cannot ignore FOIA requests with impunity.

c. Government must justify secrecy.
FOIA addresses the tradeoffs between transparency and secrecy by creating nine exceptions to disclosure.⁷⁷ By far the most important exemptions for the purposes of this discussion are the first (exempting disclosure that "reasonably could be expected to cause damage to national security"), the third (exemption for information "specifically exempted from disclosure by another statute"), the fourth (exempting disclosure of "trade secrets and other sensitive commercial data"), and the fifth [*pg 116] (exempting "inter-agency and intra-agency memoranda . . . covered by the attorney/client privilege"). These are also the provisions most heavily litigated in the four decades since the statute became law.

When Congress created the exemptions, it retained agency discretion to disclose information even if it fits within one of the FOIA exemptions.⁷⁸ The agency does not have this discretion, however, if another statute forbids disclosure of the information subject to a FOIA request. Anyone potentially injured by such disclosure can sue to prevent disclosure in such circumstances,⁷⁹ a process known as a reverse FOIA action.⁸⁰

d. Judicial review.
Finally, a person requesting information can seek judicial review if the request is denied. Two aspects of judicial review are important to government transparency. FOIA not only imposes the burden on agencies or departments to justify applicability of the exemption,⁸¹ but a court uses a de novo scope of review to judge the validity of agency's arguments, which means the judge is not obligated to give any deference to the agency's judgment in asserting an exemption claim.⁸² This scope of review is unusual. In contrast, the APA generally requires judges to give deference to agency decisions, and factual determinations are reviewed under an "arbitrary and capricious" scope of review.⁸³

2. FOIA Retrenchment
Although FOIA remains a key element in government transparency, its role in reducing agency costs has been significantly weakened in the last several years. For a time, the courts construed FOIA in a manner that increased the amount of information available to the public, and in the few cases in which the courts narrowed FOIA, Congress passed amendments to overrule these decisions.⁸⁴ More recently, the trend has been reversed. The courts have generally acted to protect government information, and Congress has gone along.

[*pg 117]

The courts' interpretation of the exemption for "trade secrets and commercial or financial information obtained from a person and privileged or confidential" is a particularly troubling example of recent trends.⁸⁵ The courts have interpreted this exemption in the broadest possible manner, ignoring other interpretations that balance the economic reasons for protecting business information against the goal of promoting government transparency.⁸⁶ Unfortunately, Congress has chosen not to overrule these cases.

The Bush Administration has also acted to reduce the effectiveness of FOIA.⁸⁷ In October 2001, Attorney General Ashcroft issued a memorandum warning agencies that any decision to release information under FOIA that might be kept confidential by claiming an exemption from disclosure "should be made only after full and deliberate consideration of the institutional, commercial, and personal privacy interests that could be implicated by disclosure of the information."⁸⁸ The memorandum pledged that the Department of Justice (DOJ) would defend decisions not to release information "unless they lack a sound legal basis or present an unwarranted risk of adverse impact on the ability of other agencies to protect other important records."⁸⁹ The memorandum reversed a Clinton Administration policy that agencies should adopt a "presumption of disclosure"⁹⁰ and that the DOJ would defend decisions to withhold information "only in those cases where the agency reasonably foresees that disclosure would be harmful to an interest protected by that exemption."⁹¹

The change in policy appears to have substantially narrowed the amount of information available under FOIA. According to a Government Accountability Office study that surveyed 183 FOIA officials at twenty-three federal agencies, nearly one-third (thirty-one percent) reported that they were less likely to release information because of the memorandum, and another one-fourth reported that the Ashcroft memorandum had caused their agencies to change their use of specific exemptions.⁹²

[*pg 118]

The Ashcroft memorandum flouts legislative intent and previous administrative practice because it pressures federal custodians to withhold disclosure under the widest possible interpretation of the statute, promoting a "need to know" policy⁹³ that stands in sharp contrast to the statute's plain language granting an absolute right to disclosure. As noted earlier, contrary to the memorandum, decisions not to apply the nine exemptions lie within the discretion of agencies and departments.⁹⁴

Although the country has experienced a significant drift away from the original conception of FOIA by the Congress that first enacted it in 1966, FOIA continues to provide the best transparency tools available to people and a free press seeking to monitor the fidelity of the executive branch to its statutory mandates. By comparison, the courts have dealt a far more serious blow to the efficacy of the Federal Advisory Committee Act (FACA).⁹⁵

B. The Federal Advisory Committee Act

FACA was enacted in 1972 "to promote good-government values such as openness, accountability, and balance of viewpoints."⁹⁶ It mandates that a committee comprised of one or more private sector stakeholders convened to advise a government agency or department must comply with a series of open government requirements. On its face, FACA is a potentially powerful tool to reduce agency costs, but the statute has been eviscerated by judicial interpretations sought by hostile administrators.

1. FACA Requirements
FACA applies to any advisory group that is "established" or "utilized" by a federal agency and that has at least one member who is not a federal employee.⁹⁷ It requires advisory committees to provide advance notice of their public meetings, with some narrow exceptions.⁹⁸ Committees must keep minutes of their meetings, and those minutes, as well as all other committee records, must remain open to the public unless the information falls within one [*pg 119] of the FOIA exemptions and the government chooses to withhold it.⁹⁹ The law requires annual reporting on the status and budgets of the committees established under these provisions.¹⁰⁰

FACA has the potential to reduce agency costs for two reasons. First, it requires that any group of special interest representatives convened by the government in order to influence public policy must meet in the open, where the group's deliberations can be monitored to ensure that they do not interfere with the executive branch's fidelity to its statutory mandates.¹⁰¹ Equally significant, FACA requires that such panels be "fairly balanced" with different points of view, serving as a further hedge against disproportionate suasion from a specific economic sector.¹⁰²

2. FACA Retrenchment
In light of the value of these requirements in reducing agency costs, it is perhaps not surprising that this ostensibly benign piece of "good government" law has come under sustained attack. The executive branch has persuaded the courts to create significant loopholes in FACA's coverage.

a. The Public Citizen case.
In 1989, the Supreme Court decided Public Citizen v. United States Department of Justice,¹⁰³ holding that a committee of the American Bar Association (ABA), which evaluated the legal abilities of persons nominated to be federal judges, was not an advisory committee under FACA.¹⁰⁴ The implications of applying FACA broadly provoked the Court's doubts that Congress had intended for FACA to apply "any time the President seeks the views of the National Association for the Advancement of Colored People (NAACP) before nominating Commissioners to the Equal Employment Opportunity Commission, or asks the leaders of an American Legion Post he is visiting for the organization's opinion on some aspect of military policy."¹⁰⁵ The Court simply did not believe that Congress intended to intrude on the operations of private groups, such as the ABA, to this extent.¹⁰⁶ Finally, the Court observed that construing FACA to apply to the Justice Department's consultations with the ABA presented "formidable" constitutional difficulties:¹⁰⁷

The District Court declared FACA unconstitutional insofar as it applied to those consultations, because it concluded that FACA, so applied, infringed unduly on the President's Article II power to nominate federal judges and violated the doctrine of [*pg 120] separation of powers. Whether or not the court's conclusion was correct, there is no gainsaying the seriousness of these constitutional challenges.¹⁰⁸

To reach this result, however, the Court had to overcome the Act's broad language, including its definition of "advisory committee" as any committee "established or utilized" by an agency "in the interest of obtaining advice or recommendations."¹⁰⁹ According to Public Citizen, an agency "establishes" an advisory committee only if it actually forms the panel itself,¹¹⁰ and it "utilizes" an advisory committee only if the panel is so closely tied to an agency as to be "amenable to [] strict management by agency officials . . . ."¹¹¹

b. Expansion of Public Citizen.
Whether or not the Court intended this result, Public Citizen created a sizeable loophole because it enabled agencies to avoid FACA requirements altogether by simply asking a nongovernmental outside group to form an advisory committee. Two cases illustrate this unfortunate outcome.

In Food Chemical News v. Young,¹¹² the D.C. Circuit Court of Appeals held that an advisory group formed by a governmental contractor was not "established" or "utilized" by an agency for advice.¹¹³ The Food and Drug Administration (FDA) had hired the Federation of America Societies for Experimental Biology (FASEB), a federation of major bio-medical research organizations, to organize an advisory panel of scientific experts to advise it on issues relating to the safety of foods and cosmetics.¹¹⁴ Although FDA had hired the FASEB to establish a peer review panel, the court held that the panel had not been "established" by FDA, but by the FASEB. Further, although the advice of the panel had been used by FDA, the agency nevertheless had not "utilize[d]" the panel because the agency had not exerted significant influence over the committee's formation.¹¹⁵

The D.C. Circuit reached a similar result nearly a decade later in Byrd v. United States EPA,¹¹⁶ holding that a peer review panel convened by an EPA contractor, the Eastern Research Group (ERG), was not a FACA advisory committee.¹¹⁷ This time around, however, EPA had had far more involvement in the formation and deliberation of the committee than FDA had concerning the FASEB panel. EPA had hired ERG to provide a peer review of a report on the carcinogenic effects of benzene. Under the contract, the EPA determined the issues for the panel to evaluate; proposed potential members of the panel [*pg 121] and expressed its approval of the persons chosen by ERG; held a teleconference with ERG and the selected panelists, instructing them as to the nature of their duties; and sent EPA employees to attend and participate in the meeting.¹¹⁸ In the court's view, none of these activities was sufficient to trigger the application of FACA. The court interpreted prior cases as holding that participation by an agency, or even an agency's "significant influence" over a committee's deliberations, does not qualify as sufficient management and control such that the committee is "utilized" by the agency under FACA.¹¹⁹

These cases rest on a mechanistic application of "establish" and "utilize" as defined by the Supreme Court in Public Citizen without considering whether that decision should be expanded to cover a different set of facts. In Public Citizen, the Court was concerned that FACA not be extended to every occasion when the federal government asked an outside organization for advice. The Court offered two reasons why it was not Congress's intention to apply FACA to organizations like the ABA or the NAACP when they offered advice to the President. Neither reason applies when an agency hires a private contractor to form an advisory committee.

The first reason was that FACA requires a government official to be in charge of each advisory committee, and the Court could not believe that Congress intended to intrude on the operations of private groups, such as the ABA, to this extent.¹²⁰ The situation, however, is entirely different when an agency hires a private contractor to form an advisory committee. Since there is extensive regulation of private contractors by the federal government, it is highly unlikely that Congress would have been concerned that FACA might intrude on how a contractor operates an advisory committee.

The Court also justified its interpretation on the ground that the application of FACA to the ABA would unduly infringe on the President's Article II power to nominate federal judges and thereby violate the doctrine of separation of powers.¹²¹ There is no similar constitutional concern if FACA applies to private contractors who are hired to form an advisory committee by a government agency.

The reason for the hostility of the D.C. Circuit to FACA is not apparent from its decisions in Food Chemical News and Byrd. What is clear is that these judges gave no serious consideration to the public's interest in executive branch transparency and the important role that FACA can play in reducing agency costs. That court could have, and should have, distinguished Public Citizen, and held that FACA applies to private contractors.

c. The aftermath of Public Citizen.
Taking full advantage of the loophole created by the D.C. Circuit, the Bush Administration has structured a [*pg 122] number of important advisory committees in a manner to avoid the Act, including the President's Commission on Intelligence on Weapons of Mass Destruction, the President's Commission to Strengthen Social Security, and the Energy Project Streamlining Task Force.¹²² It has also successfully supported legislative exemptions for other advisory committees. For example, the Homeland Security Act gives the Secretary of Homeland Security authority to exempt any committee from FACA, whether or not the committee falls within the already established exception that allows agencies and departments to close meetings if necessary to protect national security.¹²³

Compounding these trends, the Bush Administration has argued that the Constitution bars the application of FACA to the White House. Shortly after taking office in 2000, President Bush established the National Energy Policy Development Group, chaired by Vice President Cheney, to recommend a national energy plan. The task force, composed of federal officials, apparently met with various energy producers and trade associations but made no effort to meet with environmental or other public interest groups. Judicial Watch and the Sierra Club sued the government, claiming that FACA applied to the task force. The plaintiffs sought the right to obtain information that might support their case from White House officials.¹²⁴ The White House objected that such discovery would violate separation of powers and would interfere with the constitutional prerogatives of the President and Vice President.¹²⁵ The D.C. District Court permitted some discovery and said that the White House could refuse to disclose specific documents on the grounds of executive privilege. Once such a claim was made, the court indicated it would rule whether such documents had to be disclosed.

The White House brought an interlocutory appeal of the district court's order, arguing that the Constitution barred the use of discovery altogether. The court threw out the interlocutory appeal, explaining: "The Government comes to this court . . . to block discovery, having never claimed that any of the disputed material is privileged and having never responded to the District Court's invitation to specify their objections to the disputed discovery orders."¹²⁶ The Supreme Court overruled that decision and ordered the Court of Appeals for the D.C. Circuit to consider the White House's argument.¹²⁷

On remand, the D.C. Circuit ruled that the district court lacked legal authority to compel the White House to produce any documents concerning the participation of private parties in meetings of the Energy Task Force.¹²⁸ Judicial [*pg 123] Watch based its argument that FACA applied to the Task Force on an earlier decision of the D.C. Circuit holding that FACA applied when private parties regularly attend and fully participate in government-run committee meetings so as to constitute "de-facto" members of the committee.¹²⁹ This time around, however, the D.C. Circuit narrowed its prior interpretation of FACA.

The Court did not reach the merits of the constitutional argument raised by the White House, but it did rule that "separation-of-powers considerations have an important bearing on the proper interpretation of the statute."¹³⁰ In light of the separation-of-powers problems, the Court strictly interpreted the extent to which FACA applies to the executive office of the President. It held that the participation of private persons in advisory committees appointed by the President did not come within the ambit of FACA unless a private person has an official voting role on the committee or, if the committee acts by consensus, a veto over the committee's decisions.¹³¹ Since neither Judicial Watch nor the Sierra Club had explicitly alleged that any private person played such a role, the Court determined the district court lacked any legal basis to compel the White House to respond to the plaintiff's request for documents.¹³²

The effort by the D.C. Circuit to exempt the Vice President and President from FACA is another significant departure from the plain language and clear intent of the statute. To be fair, however, the court was responding after a Supreme Court opinion that strongly indicated its sympathy for the White House's constitutional claim.¹³³ The Supreme Court had also criticized the district court's solution that the White House should assert executive privilege if it wished to raise a constitutional objection to turning over specific documents. The Court had objected to this approach because it would force the judiciary "into the difficult task of balancing the need for information in a judicial proceeding and the executive's Article II prerogatives."¹³⁴

It is a time-honored tradition that the chief executive be given significant autonomy to receive confidential advice from his cabinet and staff, but these cases present an entirely different issue: namely, whether K-Street lobbyists can lobby the executive with impunity without even acknowledging that such conversations took place. As noted earlier, there are important tradeoffs between transparency and secrecy that must be considered and resolved. The Supreme Court, however, appears to have endorsed a broad constitutional protection for the White House, precluding a more nuanced approach that would recognize the tradeoffs and seek to balance them appropriately.

[*pg 124]

C. Critical Infrastructure Information Act

As the preceding discussion demonstrates, Congress has sat on the sidelines while the White House and the courts have narrowed the applications of FOIA and FACA, demonstrating its own lack of fidelity to the original legislative intent of this legislation. Congress has also contributed to the trend toward secrecy by passing the Critical Infrastructure Information Act (CIIA) as part of the Homeland Security Act.¹³⁵ FOIA has an exemption for information specifically exempted by another statute from disclosure.¹³⁶ The CIIA is such a statute. It prohibits the government from disclosing "critical infrastructure information" that is "voluntarily submitted" to the Department of Homeland Security (DHS),¹³⁷ protecting such material from disclosure under FOIA.

The Homeland Security Act accomplished the most significant reorganization of the federal government in several decades. Dozens of new provisions cover everything from the transfer of existing agencies to the new DHS to the creation of several "directorates" to enhance domestic security. Although a handful of opponents opposed Congress's rush to judgment on such complicated legislation,¹³⁸ a large majority, confronted by intense pressure from President Bush and the public's anxiety about terrorism, persevered. The CIIA was one consequence of this telescoped and highly politicized process. Despite its importance, the bill was barely mentioned during the truncated congressional debate.

One of the hallmarks of FOIA is that the nine exemptions allowing (but not requiring) the government to withhold information from disclosure were narrowly and carefully crafted by Congress. In contrast, the CIIA not only prohibits the government from releasing protected information, but it forbids the use of such information by the government or private parties in civil litigation.

1. Overbreadth
The national concern about sequestering information that is useful to terrorists is understandable. But if one has a fertile imagination, it is possible to construct a scenario for terrorist use of a wide variety of information that is also of great value to the public. The CIIA suffers from three major defects.

a. Definition of critical infrastructure information.
The statute contains language that could result in a preposterously broad scope:

[*pg 125]
The term "critical infrastructure information" means information not customarily in the public domain and related to the security of critical infrastructure or protected systems -- (A) actual, potential, or threatened interference with, attack on, compromise of, or incapacitation of critical infrastructure or protected systems by either physical or computer-based attack or other similar conduct (including the misuse of or unauthorized access to all types of communications and data transmission systems) that violates Federal, State, or local law, harms interstate commerce of the United States, or threatens public health or safety.¹³⁹

The CIIA notes that "critical infrastructure information" is "related to the security of critical infrastructures or protected systems," but the Act omits a definition of "critical infrastructure." Instead, it refers to the definition of "vital infrastructure" set forth in the Patriot Act,¹⁴⁰ which includes "systems and assets, whether physical or virtual," the destruction of which would have a "debilitating impact on security, national economic security, national public health or safety."¹⁴¹

This definition does not condition inclusion of a facility on the likelihood that information about it is likely to be used in a terrorist attack, but rather extends the definition to any facility that, if destroyed, could have a debilitating impact on economic security. It is hard to imagine a manufacturing facility of any significance that would not meet the definition in the Patriot Act.

If interpreted expansively by secrecy-minded courts, the CIIA has the potential to shield from disclosure a broad range of private-sector communications that are of no assistance to potential terrorists. For example, the Act appears to authorize a corporate lobbyist "[to] meet secretly with DHS officials to urge changes to federal immigration or customs regulations if the lobbyist asserts that the changes are related to the effort to protect the nation's infrastructure."¹⁴² As the Heritage Foundation's Mark Tapscott has noted, "One need not be a Harvard law graduate to see that, without clarification of what constitutes vulnerabilities [of infrastructure], this loophole could be manipulated by clever corporate and government operators to hide endless varieties of potentially embarrassing and/or criminal information from public view."¹⁴³

b. Permanent secrecy.
The DHS regulations implementing the CIIA exacerbate its overbreadth. As noted, the CIIA sets out certain bare-bones conditions for information to qualify for nondisclosure.¹⁴⁴ In theory at least, the DHS is under some affirmative obligation to check submissions stamped "CII" by the private sector to ensure that they meet these conditions. Yet in rules implementing the Act, the DHS says it will look at the information only once, [*pg 126] when it first comes in.¹⁴⁵ Thereafter, no matter what changes occur in the status of the information or its import, it will remain forever secret. So, for example, a facility could be torn down, eliminating it as a target for terrorists, but any information about problems that occurred there would be withheld from disclosure and, not incidentally, from use in court, as explained below.¹⁴⁶

c. No redaction.
Unlike FOIA, the CIIA does not require the release of as much information as can be released, with the redaction of any withheld information.¹⁴⁷ Depending on how the courts interpret this omission, the Act could authorize the government to keep all of the information in a document secret, even those portions that do not themselves meet the definition of critical infrastructure information.

d. Harsh penalties.
Finally, the CIIA contains harsh penalties for disclosure of critical infrastructure information in an effort to motivate federal government employees to err on the side of nondisclosure if they have any doubts about whether information must be kept secret.¹⁴⁸ Government officials who knowingly disclose voluntarily submitted critical infrastructure information face criminal liability, with a maximum sentence of one year in prison.¹⁴⁹

2. Civil Immunity
The CIIA goes further than simply making a universe of information secret. The CIIA prohibits the use of "critical infrastructure information" voluntarily submitted to the DHS from being "used directly by such agency, any other Federal, State, or local authority, or any third party, in any civil action arising under Federal or State law if such information is submitted in good faith."¹⁵⁰

This unprecedented protection means that not only is the public unlikely to find out about transgressions of the executive branch, but that critical infrastructure information held by DHS cannot be used to hold regulated entities accountable for the violation of federal laws or for imposing tort liability. Depending on how the courts interpret this language, the Act could be interpreted to prohibit the use of information held by DHS as critical infrastructure information regardless of whether the government or a private [*pg 127] party has (or can) obtain the information from some other source, such as by subpoenaing it from the company under investigation or being sued.

This worrisome aspect of the CIIA could hamper government enforcement efforts, as well as preclude injured parties from recovering damages in tort actions. Since both types of actions deter government malfeasance, an expansive reading of the CIIA would significantly increase agency costs. An overly broad interpretation of the immunity provisions could create an incentive for companies to submit information to DHS in order to forestall its use by government or plaintiffs lawyers in litigation against the company.

3. The Consequences
The extent to which the CIIA will increase agency costs is unknown. Companies appear to be reluctant to submit information to DHS under the Act.¹⁵¹ It is no small irony that the most hopeful prospect for decreasing agency costs is corporate America's traditional suspicion that submitting information to the government voluntarily can only cause trouble for a company. But this resistance could be overcome if courts interpret the Act broadly and companies become convinced that they have something to gain from such protections.

For its part, DHS has shown little interest in narrowing the CIIA. Indeed, the opposite appears to be the case. Consider, for example, its recent proposal concerning the implementation plans of its various "components" (the agencies consolidated to form the Department) regarding the Environmental Impact Statement (EIS) mandate contained in the National Environmental Policy Act (NEPA).¹⁵² The proposal announces that any "critical infrastructure information" contained in an EIS, regardless of whether it was voluntarily submitted or even generated by the government and regardless of whether it is already in the public domain, will be withheld from disclosure.¹⁵³ Since the purpose of the law is to inform public consideration of the merits of [*pg 128] construction (and other) projects, this effort to pluck a definition from the Homeland Security Act and drop it into a statutory scheme famous for its transparency will almost certainly be challenged in court if DHS does not review the statutory language more carefully and radically change its proposal.

While it is clearly premature to predict catastrophe, there can be no question that the CIIA has the potential to overwhelm FOIA when companies invoke it broadly. An expansive reading would transform the CIIA into a radical reversal of common tort liability and open government requirements. Under this scenario, the CIIA would immunize corporations and their employees from malfeasance in routine activities, from discrimination on the basis of race in the workplace, to embezzlement, to violations of environmental laws, to negligence that harms the general public financially or physically. Not incidentally, these interpretations would also immunize corporations that proved negligent in the face of terrorist threats, allowing them to avoid accountability for endangering their fellow citizens.

IV. CONCLUSION

The executive branch serves as the public's agent for purposes of implementing and enforcing regulatory mandates. There are agency costs in this arrangement because administrators will not always produce the public health, safety, and environmental benefits intended by Congress. Deterring such behavior by executive branch agents is a critical component of our constitutional system.

Although the agency-cost problem arises in the context of both private and public sectors, it is a more difficult problem in the public sector. As a general matter, it is more likely that agency contracts in the public sector will be incomplete. In addition, principals will find it more difficult to change agents and to use financial incentives to reduce agency costs.

Since the tools regularly used in the private sector to reduce agency costs are not as effective in the public sector, the public relies on forms of monitoring unique to the public sector: judicial review and legislative oversight. Information asymmetries frustrate these efforts. The primary purpose of the nation's open government laws is to eliminate such asymmetries so that the public, through a free press and Congress, can hold the executive branch accountable for breaching its fiduciary duties.

In 1977, Secretary of Defense Donald Rumsfeld was Chief of Staff to President Gerald Ford, and Vice President Richard Cheney was his deputy. Presumably acting under their advice, President Ford vetoed amendments expanding FOIA, only to have his veto overridden by Congress. Across three decades, the concerns reflected in the Ford veto have once again risen to the forefront of government information policy.

As this article has begun to demonstrate, shrouding the government in secrecy conflicts with an essential attribute of the United States' constitutional [*pg 129] democracy. Unless there is reason to believe -- for the first time in the nation's history -- the executive branch can function effectively and ethically without scrutiny for extended periods, undermining those checks and balances will impose large costs.

FOOTNOTES

*	University Distinguished Chair in Law, Wake Forest University; Board Member and Scholar, Center for Progressive Reform.
**	Jacob A. France Research Professor of Law, University of Maryland Law School; Board Member and Scholar, Center for Progressive Reform. Professor Steinzor spent her sabbatical year, 2000-2001, as an academic fellow at the Natural Resources Defense Council, where she worked on secrecy issues, testifying before Congress on the negative implications of secrecy for environmental protection. See Securing Our Infrastructure: Private/Public Information Sharing Hearing Before the S. Comm. on Governmental Affairs, 107th Cong. 38-48 (2002) (testimony of Rena Steinzor). Professors Shapiro and Steinzor thank Professors Les Boden, Ahmed Taha, and Ronald Wright for helpful comments and Natalie J. Havlina for excellent research assistance. This work was supported in part by the Project on Scientific Knowledge and Public Policy.
1.	For a listing and description of the numerous efforts by the Bush Administration to keep government information secret, see MINORITY STAFF OF H.R. COMM. ON GOVERNMENT REFORM, SPECIAL INVESTIGATIONS DIVISION, 108TH CONG., SECRECY IN THE BUSH ADMINISTRATION (Comm. Print 2004) [hereinafter SECRECY IN THE BUSH ADMINISTRATION].
2.	5 U.S.C. § 552 (2000).
3.	See id. at § 552(b).
4.	See, e.g., Keith Anderson, Is There Still a "Sound Legal Basis?": The Freedom of Information Act in the Post-9/11 World, 64 OHIO ST. L.J. 1605 (2003); Dianna G. Goldenson, FOIA Exemption Five: Will It Protect Government Scientists from Unfair Intrusion?, 29 B.C. ENVTL. AFF. L. REV. 311 (2002); Martin E. Halstuk, Shielding Private Lives From Prying Eyes: The Escalating Conflict Between Constitutional Privacy and the Accountability Principle of Democracy, 11 COMMLAW CONSPECTUS 71 (2003).
5.	See Editorial, The Dangerous Comfort of Secrecy, N.Y. TIMES, July 12, 2005, at A20, available at http://www.nytimes.com/2005/07/12/opinion/12tue3.html?8br.
6.	Conservatives are equally concerned about these issues as progressives because they fear that concentrated government power not held accountable to the citizenry will evolve into a totalitarian state. See, e.g., Mark Tapscott, Too Many Secrets, WASH. POST, Nov. 20, 2002, at A25 (Director, Heritage Foundation's Center for Media and Public Policy), cited in SECRECY IN THE BUSH ADMINISTRATION, supra note 1, at 9.
7.	Lest the reader think this is an exaggeration, consider an article written by Professor Steinzor's clinical law students about the unpleasant experiences they had representing a community group concerned about toxic pollution at the U.S. Army's Aberdeen Proving Ground in Aberdeen, Maryland. Christopher Gozdor et al., Where the Streets Have No Name: The Collision of Environmental Law and Information Policy in the Age of Terrorism, 33 ENVTL. L. REP. 10978 (2003). To make a long story short, the article reports the Army withheld the street names from maps depicting environmental contamination, although maps with the street names shown could be purchased at any local gas station, on the grounds that the Army's maps might make it marginally easier for terrorists to attack the base because it would constitute an additional dissemination of the information in the specific context of hazardous waste. See id. at 10982-83, 10994.
8.	5 U.S.C. app. §§ 1-15 (2000).
9.	6 U.S.C. §§ 131-134 (Supp. II 2002).
10.	Homeland Security Act, Pub. L. No. 107-296, 116 Stat. 2135 (2002) (codified at 6 U.S.C §§ 101-557 (Supp. II 2002)) [hereinafter Homeland Security Act].
11.	MANCUR OLSON, THE LOGIC OF COLLECTIVE ACTION: PUBLIC GOODS AND THE THEORY OF GROUPS 15 (1971).
12.	See D. RODERICK KIEWIET & MATHEW D. MCCUBBINS, THE LOGIC OF DELEGATION: CONGRESSIONAL PARTIES AND THE APPROPRIATIONS PROCESS 24-25 (Benjamin I. Page ed., 1991).
13.	See Michael C. Jensen & William H. Meckling, Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure, 3 J. FIN. ECON. 305, 308 (1976) (offering theory of agency costs in principal-agent relationships). The concept of "agency costs" includes both the cost to the principal associated with the agent failing to pursue the principal's interests and the cost incurred by the principal to prevent the agent from deviating from the principal's interests. See, e.g., HENRY N. BUTLER, ECONOMIC ANALYSIS FOR LAWYERS 462 (1998). The term "divergence costs" can be used to describe the portion of agency costs attributable to the cost of the agent to act to further the principal's interest. See, e.g., Richard L. Revesz, Specialized Courts and the Administrative Lawmaking System, 138 U. PA. L. REV. 1111, 1140 (1990).
14.	A substantial body of literature discusses the nature of agency cost problems in the corporate context. See, e.g., John W. Pratt & Richard J. Zerckhauser, Principals and Agents: An Overview, in PRINCIPALS AND AGENTS: THE STRUCTURE OF BUSINESS 1, 3 (John W. Pratt & Richard J. Zerckhauser eds., 1991) (describing the literature).
15.	To analyze agency in the political context, Professors Kiewiet and McCubbins use a more general definition of "principal" and "agent" than is used in the economics literature. They note that the economic definition of a principal is someone who signs a binding contract with someone else-the agent-for performance of specified services. In the political context, they propose an agency relationship is established "when an agent is delegated, implicitly or explicitly, the authority to take action on behalf of another, that is, the principal." KIEWIET & MCCUBBINS, supra note 12, at 239-40 ch.2 n.1. The same definition is adopted here.
16.	See U.S. CONST. art. II.
17.	See id. art. I.
18.	When the board of directors hires company executives, it will contract with them to fulfill the interests of the principals, assuming board members themselves act as faithful agents of the stockholders when they perform this function. The company executives are therefore also agents of the stockholders, although they are hired by the board.
19.	Peter Lindseth, Agents Without Principals?: Delegation in an Age of Diffuse and Fragmented Governance 2 (Univ. of Conn. Sch. of Law Working Paper Series, Paper No. 18, 2004), available at http://lsr.nellco.org/uconn/ucwps/papers/18. It also does not matter for this purpose that enacted legislation may imperfectly reflect the preference of legislators. Although Congress may be a disparate collection of 535 individual legislators, the agency approach envisions Congress in the ultimate role it was assigned by the framers of the U.S. Constitution: acting to pass legislation, reconciling differences between the House and Senate, and referring final conference report to the President for his signature. See U.S. CONST. art. I, § 7.
20.	See U.S. CONST. art. II.
21.	5 U.S.C. §§ 551-559, 701-706 (2000).
22.	See, e.g., id. §§ 551-553
23.	The political science literature characterizes agency costs as "bureaucratic drift" to indicate administrators have failed to achieve the goals of the coalition that enacted a statutory mandate. See, e.g., Sean Gailmard, Expertise, Subversion, and Bureaucratic Discretion, 18 J.L. ECON. & ORG. 536, 537 (2002) (discussing bureaucratic drift).
24.	Lindseth, supra note 19, at 3.
25.	Id.
26.	See STEVEN KELMAN, MAKING PUBLIC POLICY: A HOPEFUL VIEW OF AMERICAN GOVERNMENT 250-54, 266 (James Q. Wilson ed., 1987) ("[T]he more important a policy is, the less important is the role of self-interest in determining that policy."); Steven P. Croley, Theories of Regulation: Incorporating the Administrative Process, 98 COLUM. L. REV. 1, 66 (1998) ("[R]egulatory outcomes ameliorate market failures and vindicate the citizenry's interests . . . more commonly than other scholars of regulation acknowledge . . . .").
27.	See, e.g., Terry M. Moe, Politics and the Theory of Organization, 7 J.L. ECON. & ORG. 106, 124-25 (1991) (special issue).
28.	See, e.g., Suzanne Goldenberg, Pentagon 'Hid' Damning Halliburton Audit, GUARDIAN, Mar. 16, 2005, at 13 (describing the Defense Contract Audit Agency's report finding the Pentagon knew that Halliburton has overcharged the United States at least $108.4 million).
29.	See, e.g., Gardiner Harris, F.D.A. Official Admits 'Lapses' On Vioxx, N.Y. TIMES, Mar. 2, 2005, at A15 (describing how the Food and Drug Administration was aware that the drug Vioxx could cause heart attacks or strokes for over a year before it made that information available to prescribing physicians); Tom Hamburger & Alan C. Miller, Mercury Emissions Rule Geared to Benefit Industry, Staffers Say, L.A. TIMES, Mar. 16, 2004, at A1 (explaining how the White House and political appointees at the Environmental Protection Agency rewrote regulations for coal-fired power plants to reduce their effectiveness in reducing mercury contamination of the human food chain).
30.	See RICHARD J. LAZARUS, THE MAKING OF ENVIRONMENTAL LAW (2004) (explaining the clash of interests in the passage of the nation's environmental laws).
31.	See ROBERT V. PERCIVAL ET AL., ENVIRONMENTAL REGULATION: LAW, SCIENCE, AND POLICY 491 (2003) ("[C]ertain conventional air pollutants have become nearly ubiquitous."); Carol Browner & Dan Glickman, Clean Water Acton Plan: Restoring and Protecting America's Waters 7-9 (Feb. 14, 1998), http://water.usgs.gov/owq/cleanwater/action/cwap.pdf (detailing the "serious water quality problems [that] persist throughout the country"); Office of Pollution Prevention and Toxics, EPA, Chemical Hazard Data Availability Study 2 (1998), http://www.epa.gov/oppt/chemtest/hazchem.pdf (explaining that EPA's ability to protect public health is impaired by lack of understanding about certain High Production Volume Chemicals to which the public is exposed in amounts greater than one million pounds every year).
32.	OLIVER E. WILLIAMSON, THE ECONOMIC INSTITUTIONS OF CAPITALISM 45 (1987).
33.	Id. at 46 (citation omitted).
34.	Douglas C. North, A Transaction Cost Theory of Politics, 2 J. THEORETICAL POL. 355, 361-62 (1990).
35.	See Howard A. Latin, The Feasibility of Occupational Health Standards: An Essay on Legal Decisionmaking Under Uncertainty, 78 NW. U. L. REV. 583, 583-87 (1983).
36.	See ALBERT O. HIRSCHMAN, EXIT, VOICE, AND LOYALTY (1970) (describing exit as a strategy to agency-cost problems).
37.	See id. at 10-11.
38.	See Oliver E. Williamson, Public and Private Bureaucracies: A Transaction Cost Economics Perspective, 15 J.L. ECON. & ORG. 306, 331 (1999) (noting that some political delegations are massively incomplete because of the difficulty of specifying in advance what action an agent is to take).
39.	See Government Performance and Results Act of 1993, Pub. L. No. 103-62, 107 Stat. 285 (codified in scattered sections of 5 U.S.C. and 31 U.S.C.).
40.	See 139 CONG. REC. S7739, 7743 (June 23, 1993) (statement of Senator Glenn) (noting that GPRA fits in a statutory constellation that includes the Chief Financial Officers Act of 1990, the Cash Management Improvement Act of 1990, the Single Audit Act of 1984, the Federal Managers Financial Integrity Act of 1982, the Prompt Payment Act of 1982, the Paperwork Reduction Act of 1980, and the Inspector General Act of 1978).
41.	For example, Alice Rivlin, director of the Office of Management and Budget under President Clinton, has explained the GPRA as a way to "reduce the [budget] deficit in an intelligent way." Alice M. Rivlin, Linking Resources to Results: Management and Budgeting in a Time of Resource Constraints; Government Performance and Results Act of 1993, THE PUB. MANAGER: THE NEW BUREAUCRAT (June 22, 1995) at 3-4.
42.	See, e.g., Annys Shin, GOP Grades the Feds, But Not on a Curve, 17 NAT'L J. 939 (1998) (reporting on GPRA "grades" awarded by prominent Republican conservatives to disfavored agencies).
43.	Pratt & Zeckhauser, supra note 14, at 4-5.
44.	Id. at 5.
45.	Lindseth, supra note 19, at 3 (emphasis omitted).
46.	5 U.S.C. § 702 (2000).
47.	Id. § 706(2)(C).
48.	Id. § 706(2)(A).
49.	See Motor Vehicle Mfrs. Ass'n of U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 40-43 (1983).
50.	See United States v. Nova Scotia Food Prods. Corp., 568 F.2d 240, 251 (2d Cir. 1977); Portland Cement Ass'n v. Ruckelshaus, 486 F.2d 375, 392-93 (D.C. Cir. 1973), cert. denied, 417 U.S. 921 (1974).
51.	See Sidney A. Shapiro & Richard E. Levy, Judicial Incentives and Indeterminacy in Substantive Review of Administrative Decisions, 44 DUKE L.J. 1051, 1068-71 (1995).
52.	See note 38 & accompanying text.
53.	5 U.S.C. § 701(a)(1); see, e.g., Block v. Cmty. Nutrition Inst., 467 U.S. 340, 347-48, 352-53 (1984) (holding Congress had intended to preclude consumers from obtaining judicial review of milk marketing orders by the Department of Agriculture).
54.	5 U.S.C. § 701(1)(2); Citizens to Pres. Overton Park v. Volpe, 401 U.S. 402, 410 (1971) (The APA prohibits judicial review when there is "no law to apply.").
55.	See U.S. CONST. art. III, § 2.
56.	See, e.g., Lujan v. Defenders of Wildlife, 504 U.S. 555, 578 (1992) (holding that individuals could not sue government administrators to claim lack of compliance with the Endangered Species Act). See generally Richard J. Pierce, Jr., Lujan v. Defenders of Wildlife: Standing as a Judicially Imposed Limit on Legislative Power, 42 DUKE L.J. 1170, 1200 (1993) (criticizing the Court's standing cases on the ground that an agency will be able to ignore legislative commands because no one who opposes agency decisions will have standing to challenge them).
57.	See OLSEN, supra note 11, at 48 (discussing the lack of financial incentives to engage in collective political action).
58.	Mathew D. McCubbins & Thomas Schwartz, Congressional Oversight Overlooked: Police Patrols Versus Fire Alarms, 28 AM. J. POL. SCI. 165, 166 (1984).
59.	KIEWIET & MCCUBBINS, supra note 12, at 33 (emphasis added).
60.	McCubbins & Schwartz, supra note 58, at 166.
61.	Id.
62.	What is unknowable is the extent to which more disclosure reduces agency costs since it depends on the extent to which collective action problems are overcome by the public. Unless one assumes that more disclosure of Executive Branch failures will lead to no response by the public, however, greater secrecy will increase agency costs.
63.	5 U.S.C. § 552 (2000).
64.	5 U.S.C. app. § 1-15 (2000).
65.	6 U.S.C. §§ 131-134 (Supp. II 2002) (passed as part of Homeland Security Act of 2002).
66.	See SECRECY IN THE BUSH ADMINISTRATION, supra note 1, at 81.
67.	See infra note 68 and accompanying text.
68.	For a general description of the Act and the opportunities it provides requestors of information, see Allen Robert Alder, American Civil Liberties Union, Step-by-Step Guide to Using the Freedom of Information Act, available at http://www.skepticfiles.org/aclu/foia.htm (last visited on October 28, 2005).
69.	Freedom of Information Act, 5 U.S.C. § 552(a)(2)(A) (2000).
70.	Id. § 552(a)(2)(D).
71.	See Pub. L. No. 104-231, § 4(7), 110 Stat. 3048, 3049 (1996) (codified as amended at 5 U.S.C. § 552(a)(2)(E) (2000)).
72.	5 U.S.C. § 552(a)(3)(A).
73.	See id. § 552(a)(3).
74.	Id. § 552(a)(4)(A).
75.	Id. § 552(a)(6)(a).
76.	See Mark H. Grunewald, E-FOIA and the "Mother of All Complaints:" Information Delivery and Delay Reduction, 50 ADMIN L. REV. 345, 348-58 (1998); Richard J. Pierce, Jr., Judicial Review of Agency Actions in a Period of Diminishing Agency Resources, 49 ADMIN. L. REV. 61, 77-84 (1997); Eric J. Sinrod, Freedom of Information Act Response Deadlines: Bridging the Gap Between Legislative Intent and Economic Reality, 43 AM. U. L. REV. 325, 347-50 (1994).
77.	FOIA exempts information from disclosure if (1) a specific Executive Order authorizes keeping the information secret in the interest of national defense or foreign policy, and the information is in fact properly classified pursuant to the Executive Order, (2) the information is related solely to the internal personnel rules and practices of a government agency, (3) the information is specifically exempted from disclosure by another statute, (4) the information constitutes trade secrets and other sensitive commercial data, (5) the information is comprised of inter-agency and intra-agency memoranda that are covered by the attorney/client privilege or compiled in preparation for a trial, (6) the information is contained in individual personnel and medical files, (7) the request involves law enforcement investigative files if disclosure would jeopardize the case or an informant, (8) the information concerns the regulation of certain financial institutions, or (9) the information is geological and geophysical data revealing oil well locations. See 5 U.S.C. § 552(b)(1)-(9).
78.	Chrysler Corp. v. Brown, 441 U.S. 281, 293-94 (1979).
79.	Id. at 318.
80.	See Stephen F. Hehl, Reverse FOIA Suits After Chrysler: A New Direction, 48 FORDHAM L. REV. 185, 187 (1979); Paul M. Nick, De Novo Review in Reverse Freedom of Information Act Suits, 50 OHIO ST. L.J. 1307, 1308 (1989).
81.	5 U.S.C. § 552(a)(4)(B).
82.	See id.
83.	Id. § 706(2)(A).
84.	RICHARD J. PIERCE, JR., SIDNEY A. SHAPIRO & PAUL R. VERKUIL, ADMINISTRATIVE LAW & PROCESS 435 (4th ed. 2004). As originally enacted, for example, FOIA exempted two types of information from disclosure: information related to a government function "requiring secrecy in the public interest" and information "relating solely to the internal management of an agency." Pub. L. No. 89-554, §552(a), 80 Stat. 378, 383 (1966) (codified as amended at 5 U.S.C. § 552 (2000 & Supp. II 2002)). In 1976, Congress intensified the requirements for information exempted by statute. See Pub. L. No. 94-409, 90 Stat. 1241, 1242 (1976) (codified as amended at 5 U.S.C. § 552 (2000 & Supp. II 2002)).
85.	See 5 U.S.C. § 552(b)(4).
86.	Mary L. Lyndon, Information Economics and Chemical Toxicity: Designing Laws to Produce and Use Data, 87 MICH. L. REV. 1795, 1825 (1989). See also Thomas O. McGarity & Sidney A. Shapiro, The Trade Secrecy Status of Health and Safety Testing Information: Reforming Agency Disclosure Policies, 93 HARV. L. REV. 837, 857-82 (1980).
87.	See Kristen Elizabeth Uhl, Comment, The Freedom of Information Act Post-9/11: Balancing the Public's Right to Know, Critical Infrastructure Protection, and Homeland Security, 53 AM. U. L. REV. 261, 263, 272-74 (2003).
88.	Memorandum from John Ashcroft, Attorney General, to Heads of All Federal Departments and Agencies, The Freedom of Information Act (Oct. 12, 2001), http://www.usdoj.gov/04foia/011012.htm.
89.	Id.
90.	Memorandum from Janet Reno, Attorney General, to Heads of Departments and Agencies, The Freedom of Information Act (Oct. 4, 1993), http://www.fas.org/sgp/clinton/reno.html.
91.	Id.
92.	United States General Accounting Office, Freedom of Information Act, Agency Views on Changes Resulting from New Administration Policy, GAO-03-981 11, 14 (Sept. 2003), available at http://www.gao.gov/new.items/d03981.pdf; see also 4 OMB Watcher #19, GAO Report Indicates Less Information Available Under Ashcroft (Sept. 22, 2003), http://www.ombwatch.org/article/articleview/1821/1/189/.
93.	See Laura Parker, et al., Secure Often Means Secret, USA TODAY, May 16, 2002, at 1A (quoting Gary Bass, executive director of OMB Watch, as stating, "We seem to be shifting to the public's need to know instead of the public's right to know."), quoted in Uhl, supra note 87, at 285.
94.	See supra note 78 and accompanying text.
95.	See 5 U.S.C. app. § 1-15 (2000).
96.	Steven P. Croley, Practical Guidance on the Applicability of the Federal Advisory Committee Act, 10 ADMIN. L.J. 111, 117 (1996); see also Jay S. Bybee, Advising the President: Separation of Powers and the Federal Advisory Committee Act, 104 YALE L.J. 51, 73 (1994) (noting congressional hearings on FACA "focused on the non-representative nature of the advisory committees, and the need to open their proceedings and reports to the public").
97.	5 U.S.C. app. § 3(2).
98.	See id. § 10 (meetings may be closed to protect national security, to avoid the release of information specifically exempted from disclosure by statute, or to protect trades secrets and confidential commercial and financial information).
99.	See id.
100.	Id. § 6(c).
101.	See id. § 10(a)-(b).
102.	See id. § 5(b)(2).
103.	491 U.S. 440 (1989).
104.	Id. at 443.
105.	Id. at 452-53.
106.	See id. at 453.
107.	Id. at 466.
108.	Id. at 466-67 (citation omitted).
109.	See Federal Advisory Committee Act, 5 U.S.C. app. § 3(2) (2000).
110.	See 491 U.S. at 452, 456-57, 461-63.
111.	See id. at 457-58.
112.	900 F.2d 328 (D.C. Cir. 1990), cert. denied, 498 U.S. 846 (1990).
113.	Id. at 333.
114.	See id. at 329-30.
115.	See id. at 333.
116.	174 F.3d 239 (D.C. Cir 1999), cert. denied, 529 U.S. 1018 (2000).
117.	See id. at 245-48.
118.	See id. at 241-42.
119.	Id. at 246.
120.	See Public Citizen, 491 U.S. at 452-53.
121.	See id. at 466-67.
122.	SECRECY IN THE BUSH ADMINISTRATION, supra note 1, at 38-42.
123.	Homeland Security Act of 2002, Pub. L. No. 107-296, § 871, 116 Stat. 2135, 2243 (2002) (codifed at 6 U.S.C. § 451 (Supp. II 2002)).
124.	Judicial Watch, Inc. v. Nat'l Energy Policy Dev. Group, 219 F. Supp. 2d 20, 23-24 (D.D.C. 2002).
125.	See id. at 44-45.
126.	In re Cheney, 334 F.3d 1096, 1110 (D.C. Cir. 2003) (emphasis omitted).
127.	Cheney v. United States Dist. Court, 124 S. Ct. 2576, 2593 (2004).
128.	In re Cheney, 406 F.3d 723, 729-31 (D.C. Cir. 2005).
129.	See id. at 726 (citing Assoc. of American Physicians & Surgeons, Inc. v. Clinton, 997 F.2d 898 (D.C. Cir. 1993)).
130.	Id. at 727.
131.	Id. at 728.
132.	See id. at 730-31.
133.	The Court of Appeals had dismissed the separation of powers concerns of the White House on the basis of United States v. Nixon, 418 U.S. 683 (1974). See In re Cheney, 334 F.3d at 1098-99.
134.	Id. at 2592.
135.	See 6 U.S.C § 131 (Supp. II 2002).
136.	5 U.S.C. § 552(b)(3) (2000).
137.	6 U.S.C. § 133(a)(1). For a thorough explanation of the statutory language see Rena I. Steinzor, "Democracies Die Behind Closed Doors": The Homeland Security Act and Corporate Accountability, 12 KAN. J.L. & PUB. POL'Y 641, 645-50 (2003).
138.	See, e.g., Gail R. Chaddock, Bush Spurs Lame-Duck Session to Action, CHRISTIAN SCI. MONITOR, Nov. 18, 2002, at 2 ("For heavens sake, we have a right to know what is in this 484-page bill, and as of this moment, we do not," Senator Byrd said on the floor of the Senate.).
139.	6 U.S.C. § 131.
140	. Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism (USA PATRIOT) Act of 2001, Pub. L. No. 107-56 (2001).
141.	See 42 U.S.C. § 5195c(e) (Supp. I 2001).
142.	SECRECY IN THE BUSH ADMINISTRATION, supra note 1, at 9.
143.	Tapscott, supra note 6.
144.	See 6 U.S.C. § 131(3).
145.	See Protected Critical Infrastructure Information, 6 C.F.R. § 29.6(e) (2005). The rule was proposed on April 15, 2003. See Procedures for Handling Critical Infrastructure Information, 68 Fed. Reg. 18524 (Apr. 15, 2003). It became effective as an interim rule on February 20, 2004, and apparently continues to operate as an interim rule. See Procedures for Handling Critical Infrastructure Information; Interim Rule, 69 Fed. Reg. 8074 (Feb. 20, 2004). The rule presumes that all information submitted satisfies the requirements for protection unless and until the department finds otherwise. 6 C.F.R. § 29.6(b).
146.	See discussion infra Part II.C.2.
147.	SECRECY IN THE BUSH ADMINISTRATION, supra note 1, at 9-10.
148.	Id. at 10.
149.	6 U.S.C. § 133(f).
150.	6 U.S.C. § 133(a)(C).
151.	Professor Steinzor represented OMB Watch, a nonprofit public interest organization that serves as a watchdog for regulatory agencies, in a FOIA lawsuit seeking information about the number of CIIA requests that the DHS had received as of March 2005. Documents released prior to settlement of the lawsuit indicated that the total number of such requests was twenty-seven.
152.	NEPA requires that all federal agencies "include in every recommendation or report on proposals for legislation and other major Federal actions significantly affecting the quality of the human environment, a detailed statement" about the action's environmental impacts. 42 U.S.C. § 4332(C) (2000).
153.	See Timely and Effective Environmental Planning in the Department of Homeland Security, 69 Fed. Reg. 33043, 33063 (June 14, 2004). The Natural Resources Defense Council submitted comments objecting to the proposed regulation on the grounds that it would limit public access to information, much of which is currently available to the public, about activities that could harm the environment. See Comments from Sharon Buccino, Senior Attorney, & Max Wilson, Legal Intern, Natural Resources Defense Council, to Environmental Planning, Department of Homeland Security, 1 (July 14, 2004), available at http://www.dhs.gov/interweb/assetlibrary/Mgmt_NEPA_NRDC_071404.pdf. The Center for Progressive Regulation also submitted comments urging DHS to rethink the proposed regulation because its sweeping secrecy provisions violate NEPA. See Comments from Rena Steinzor, Board Member and Member Scholar & Karen Sokol, Staff Counsel, Center For Progressive Regulation, to Environmental Planning, Department of Homeland Security, 1 (Aug. 16, 2004), available at www.progressivereform.org.

THE PEOPLE'S AGENT: EXECUTIVE BRANCH SECRECY AND ACCOUNTABILITY IN AN AGE OF TERRORISM