Read an Excerpt

Fixing Intelligence

Yale University Press

Chapter One

Studies on intelligence reform have become a cottage industry. Why do we need another one? The answer is simple: most of them have attempted to doctor the symptoms, not the illness. The skeptical reader will object that this claim is not self-evident. But consider the public record.

Since Congress began investigating the Intelligence Community in the mid-1970s, the issue of intelligence reform has been raised repeatedly. During the Carter administration several initiatives were taken to implement some of the ideas produced by a committee chaired by Senator Frank Church (D-Idaho), but no fundamental structural change occurred. Several times in the 1980s congressional oversight committees raised the reform issue, and Senator David Boren (D-Okla.) actually drafted legislation for several structural changes in the Intelligence Community. The House committee offered an alternative draft, but neither bill became law. In 1994 Senator John Warner (R-Va.) introduced legislation for a presidential commission to consider Intelligence Community reforms, and his bill became law. The resulting commission produced its report in early 1996. At the same time, several unofficial intelligence reform studies produced a flurry of activity and a wide variety of proposals.

Most of these reform efforts were inspired by sensational problems and episodes within the Intelligence Community, especially in the Central Intelligence Agency (CIA). The Church committee was outraged by evidence that the CIA had attempted assassinations as part of covert actions in the past, and that the army's counterintelligence units had been used to help the Federal Bureau of Investigation (FBI) keep track of antiwar movement leaders in the late 1960s and early 1970s. In the late 1980s, CIA covert actions in Central America and related activities by National Security Council staff members became the focus of renewed interest in intelligence reform. In the 1990s a number of incidents within the CIA, some having to do with personnel policies, others involving serious operational failures, and still others involving National Reconnaissance Office accountability for funds, brought the issue once again to congressional and public attention. The revelation in 1994 that Aldrich Ames of CIA's Directorate of Operations was a KGB agent produced a new and unprecedented level of concern.

After almost three decades of such episodes, no fundamental reform has occurred. Virtually all congressional investigations and reform studies have merely focused on the scandals and raised policy issues. For example, should the CIA be permitted to carry out assassinations? Should Army Counterintelligence be involved in domestic surveillance of civilians? Should CIA clandestine officers be allowed to have cover as journalists? Should the Intelligence Community budget be made public? And so on.

Almost none of the congressional committees' efforts at reform have addressed structural, organizational, and management issues. The major exception was Senator Boren's draft legislation that directed certain organizational changes, notably the creation of a national agency specializing in intelligence gleaned from photography, radar and more sophisticated imaging technologies imagery. A second partial exception was the 1996 organizational redesign proposal by the House Permanent Select Committee on Intelligence. A presidential commission of 1996 was instructed to examine the roles and missions of all the parts of the Intelligence Community-an invitation to deal with the structural issues-but it effectively declined the invitation.

One major feature of the Boren bill of the late 1980s, the creation of a national imagery agency, was actually accepted, and the National Imagery and Mapping Agency was created on 1 October 1996. A few additional but minor organizational steps have been taken in connection with the support staff for the director of central intelligence (DCI), formerly called the Intelligence Community Staff, now reorganized and renamed the Community Management Staff. Immediately after the Ames case broke, a flurry of activity centered on the traditional FBI-CIA dispute over counterintelligence turf, but interest soon abated.

With few exceptions, however, structural reform has been largely ignored. This is both strange and unfortunate. Several senators and House members proved reluctant to delve into the structural issues. The CIA and most directors of central intelligence have also resisted serious review of the structural problems. Thus a quiet and informal consensus that nothing structurally is wrong has prevailed, not only in Congress and in the Intelligence Community itself but in the presidential commission as well. Private-sector intelligence reform studies have followed suit.

The reasons why are not entirely clear, especially after Senator Boren opened the question of deeper problems in need of structural solutions. Senator Dennis DeConcini (D-Ariz.) complained that the problem was the "culture" at the CIA. Organizational cultures are normally the products of structural conditions. The accumulation of dangerously embarrassing incidents should have suggested to both administration officials and the congressional committees that they were looking at the symptoms, not the underlying ills.

Do serious structural and organizational problems really exist? One does not need access to classified information to answer yes. It would be extraordinarily surprising if there were not. That Aldrich Ames of the CIA and Robert Hanssen of the FBI could be recruited by the KGB, meet KGB case officers undetected for years, and make "dead drops" in Washington, D.C., under the nose of the FBI suggests that more than policy problems are involved. Little known to the public but notorious in military circles, failures in CIA intelligence contributed to the failed attempt in 1980 to rescue hostages in Iran and compromised operations in the Persian Gulf War. Other intelligence failures, as well as bureaucratic turf fights between the National Reconnaissance Office and other agencies, are evidence of deeper problems. Regardless of the individuals in charge, the structural arrangements and the problems they create have remained constant or have worsened.

Since the Intelligence Community structure settled into place some forty years ago, it has remained essentially unchanged. Meanwhile, technologies have advanced at a blinding pace over those decades. In the 1960s transoceanic communications were fairly limited. As a result, many technical intelligence activities had to be located in Europe and East Asia. A decade later, many of those activities had been moved to the United States, as space-based communication capabilities made possible radically different and more effective ways of gathering, processing and producing, and distributing intelligence. In the quarter-century since, communication has advanced exponentially in speed and sophistication. The communications revolution alone provides grounds for suspecting that major structural reforms in the Intelligence Community are long overdue. Some adaptations have been made, but not enough to allow full exploitation of the technologies.

The introduction of intelligence collection systems in space, for example, created radically new possibilities, and many of them were exploited in the 1960s and 1970s. As time passed, however, and as the number and variety of space systems increased, new ways to exploit them became possible that could not have been foreseen. Most such opportunities, however, require organizational changes, and very few were made. A national imagery agency should have been created at least twenty years ago, but bureaucratic turf concerns kept the very idea from being considered. Nor does the belated creation in 1996 of the National Imagery and Mapping Agency assure that it will be given all the authority required to make the most of imaging systems.

Structural problems afflict not only intelligence collection but also intelligence analysis. During the Cold War this was conspicuous in the continuing debate between the CIA and the Defense Intelligence Agency (DIA) on estimates of Soviet military capabilities. The gross underestimate of Soviet military expenditures can be explained largely as the result of competition that caused each of these agencies to be less concerned with the truth of matters in the Soviet Union than with proving the other wrong in the eyes of the Congress. The needs of executive branch policymakers, for whom this intelligence was primarily produced, tended to be secondary in both agencies' calculations. The handling of intelligence analysis on al Qaeda in the weeks and days leading up to the events of 11 September 2001 offers another example of problems in both analysis and distribution of intelligence. The FBI's jealousy over its counterintelligence turf, not only vis-à-vis the CIA but also vis-à-vis the military services' counterintelligence analysis (as well as operations), is a similar symptom of structural problems.

These examples suggest that structural problems distorted U.S. perceptions of Soviet military and economic capabilities and permitted the KGB to penetrate our own intelligence services to a degree that was preventable.

Failure to make appropriate organizational changes to exploit evolving technology has also wasted money. The costs of technical intelligence collection systems dwarf the costs of intelligence analysis and clandestine human intelligence operations. Yet efficiency has rarely been considered in the purchase of technical systems. In the 1950s and 1960s, there was not yet adequate experience with most new technologies to allow meaningful efficiency comparisons. In some cases it could also be argued that Intelligence Community technologies more than paid for themselves as they migrated into the civilian economy. For example, development of modern digital computational means-computers-occurred almost entirely as a result of the National Security Agency's research and development efforts in the 1950s. IBM and CDC essentially got their start in modern computers from National Security Agency funding, and without it, we might be two decades behind where we are in computers today.

By the 1970s and 1980s, however, these arguments for ignoring the burgeoning costs of technical collection programs were as outdated as the displaced technologies. But bureaucratic processes and organizational interests stood in the way of examining costs and reducing them. Predictably, inefficiencies mounted. Covered by secrecy, arcane organizational practices, and technical complexities, these accumulating structural-management issues have remained largely opaque to the congressional oversight committees and also to high-level officials in the executive branch. Neither DCIs nor secretaries of defense have really understood them fully.

None of the accumulating inefficiencies and missed opportunities for better and more efficient exploitation should be terribly surprising in organizations dealing with fairly rapid change in leading-edge technologies. IBM, AT&T, GMC, Chrysler, and other large industrial firms have been slow to make fundamental structural changes, too, but they did make them, in the face of serious financial difficulties resulting from business competition. Management at all levels in organizations tends to resist change, especially structural reform. But if change is necessary for AT&T, is it not also necessary for the Intelligence Community?

Critical scrutiny of the Intelligence Community is clearly justified, scrutiny of its structure, organization, and management arrangements. I am not concerned here with issues that received a lot of media attention in the 1990s, like whether or not the CIA should put agents under cover as journalists. I am not concerned with whether or not the CIA predicted the collapse of the Soviet Union. I am not concerned with proper security rules and techniques for preventing another Ames case. I am not concerned with whether more intelligence attention should be put on the Third World, terrorism, nuclear proliferation, Russia, or economic affairs. Nor am I concerned with whether the United States should engage in covert actions, or whether the United States should have an Intelligence Community. (Those last questions, however, are relevant to the structural arrangements for intelligence support to a new Department of Homeland Security.) These are issues treated elsewhere, and all of them are policy problems. They are important, but they are different from management and structural problems. Thus they will be treated here only when they appear as symptoms of structural problems. As the events of 11 September have once again forced intelligence issues onto the agenda, the initial reactions have been to treat them primarily as policy matters, ignoring the underlying structural issues.

To clarify this distinction with a metaphor, assume that the Intelligence Community is a ship which has had numerous problems on its recent voyages. It is now in harbor for repairs. Most of the ship's investors ask where it should sail next, when it should sail, what flag it should fly, what color to paint the ship. Here those questions will be set aside and primary attention focused on the ship's hull, its engines, its navigation gear; I will look for malfunctions and prescribe needed repairs. When the repairs are completed, it can then sail anywhere, any time, under any flag, with any passengers. These choices are up to executive branch policymakers and military officials served by the Intelligence Community.

Continues...

---

Excerpted from Fixing Intelligence by William E. Odom Copyright © 2004 by William E. Odom. Excerpted by permission.
All rights reserved. No part of this excerpt may be reproduced or reprinted without permission in writing from the publisher.
Excerpts are provided by Dial-A-Book Inc. solely for the personal use of visitors to this web site.

---