Untitled

Brig Gen Kenneth Newton Walker

Kenneth Walker enlisted at Denver, Colorado,

on 15 December 1917. He took flying training at

Mather Field, California, getting his commission

and wings in November 1918.

After a tour in the Philippines, he returned

to Langley Field, Virginia, in February 1925

with a subsequent assignment in December

1928 to attend the Air Corps Tactical School.

Retained on the faculty as a bombardment in-

structor, Walker became the epitome of the

strategic thinkers at the school and coined the

revolutionary airpower “creed of the bomber”:

“A well-planned, well-organized and well-flown

air force attack will constitute an offensive

that cannot be stopped.”

Following attendance at the Command and General Staff School at Fort

Leavenworth, Kansas, in 1933 and promotion to major, he served for three

years at Hamilton Field, California, and another three years at Luke Field,

Ford Island, and Wheeler Field, Hawaii. Walker returned to the United

States in January 1941 as assistant chief of the Plans Division for the chief

of the Air Corps in Washington, DC.

He was promoted to lieutenant colonel in July 1941 and colonel in March

1942. During this time, when he worked in the Operations Division of the War

Department General Staff, he coauthored the air-campaign strategy known as

Air War Plans Division—Plan 1, the plan for organizing, equipping, deploying,

and employing the Army Air Forces to defeat Germany and Japan should the

United States become embroiled in war. The authors completed this monu-

mental undertaking in less than one month, just before Japan attacked Pearl

Harbor—and the United States was, in fact, at war.

In June 1942, he was promoted to brigadier general and assigned by Gen

George Kenney as commander of Fifth Air Force’s Bomber Command. In this

capacity, he repeatedly accompanied his B-24 and B-17 units on bombing mis-

sions deep into enemy-held territory. Learning firsthand about combat condi-

tions, he developed a highly efficient technique for bombing when aircraft faced

opposition by enemy fighter planes and antiaircraft fire.

General Walker was killed in action on 5 January 1943 while leading a

bombing mission over Rabaul, New Britain—the hottest target in the theater.

He was awarded the Medal of Honor. Its citation, in part, reads, “In the face of

extremely heavy anti aircraft fire and determined opposition by enemy fighters,

General Walker led an ef

fective daylight bombing attack against shipping in the

harbor at Rabaul, which resulted in direct hits on nine enemy vessels. During

this action, his airplane was disabled and forced down by the attack of an over-

whelming number of enemy fighters. He displayed conspicuous leadership

above and beyond the call of duty involving personal valor and intrepidity at an

extreme hazard to life.” Walker is credited with being one of the men who built

an or

ganization that became the US Air For

ce.

After you have read this research report, please

give us your frank opinion on the contents. All

comments––large or small, complimentary or

caustic––will be gratefully appreciated. Mail them

to CADRE/AR, Building 1400, 401 Chennault

Circle, Maxwell AFB AL 36112–6428.

Thank you for your assistance.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Cut along dotted line

Intelligence Reform

A Question of Balance

Bansemer

AIR FORCE FELLOWS

COLLEGE OF AEROSPACE DOCTRINE,

RESEARCH AND EDUCATION

AIR UNIVERSITY

Intelligence Reform

A Question of Balance

JOHN D. BANSEMER

Lieutenant Colonel, USAF

Walker Paper No. 5

Air University Press

Maxwell Air For

ce Base, Alabama 36112-6615

August 2006

This publication includes copyrighted material.

any form without written consent from the Program on Information Resources Policy, Harvard

University, Maxwell Dworkin 125, 33 Oxford Street, Cambridge MA 02138. (617) 495-4114. E-mail:

[email protected] URL: http://www.pirp.harvard.edu

This publication has been reviewed and approved by competent personnel of the preparing com-

mand in accordance with current directives on doctrine, policy, essentiality, propriety, and quality.

Opinions, conclusions, and recommendations expressed or implied within are solely those of

the author and do not necessarily represent the views of Air University, the United States Air

Force, the Department of Defense, or any other US government agency. Cleared for public re-

lease: distribution unlimited.

Air University Library Cataloging Data

Bansemer, John D.

Intelligence reform : a question of balance / John D. Bansemer

p. ; cm. –– (Walker paper, 1555-7871 ; no. 5)

Includes bibliographical references.

ISBN 1-58566-151-1

1. Intelligence service––United States. 2. Intelligence service––United

States––Management. 3. Administrative agencies––United States––Reorganization.

4. Military intelligence––United States. I. Title. II. Series.

353.170973––dc22

This W

alker Paper and others in the series ar

e available

electr

onically at the Air University Resear

ch W

eb site

http://research.maxwell.af.mil and the AU Press Web

site http://aupr

ess.maxwell.af.mil.

Air Force Fellows

Since 1958 the Air Force has assigned a small number of care-

fully chosen, experienced officers to serve one-year tours at dis-

tinguished civilian institutions studying national security policy

and strategy. Beginning with the 1994 academic year, these pro-

grams were accorded in-residence credit as part of professional

military education at senior service schools. In 2003 these fellow-

ships assumed senior developmental education (SDE) force-

development credit for eligible officers.

The SDE-level Air Force Fellows serve as visiting military am-

bassadors to their centers, devoting effort to expanding their col-

leagues’ understanding of defense matters. As such, candidates

for SDE-level fellowships have a broad knowledge of key Depart-

ment of Defense (DOD) and Air Force issues. SDE-level fellows

perform outreach by their presence and voice in sponsoring

institutions. They ar

e expected to provide advice as well as

promote and explain Air Force and DOD policies, programs,

and military-doctrine strategy to nationally recognized schol-

ars, for

eign dignitaries, and leading policy analysts. The Air

Force Fellows also gain valuable perspectives from the ex-

change of ideas with these civilian leaders. SDE-level fellows

are expected to apprise appropriate Air Force agencies of sig-

nificant developments and emerging views on defense as well

as economic and for

eign policy issues within their centers.

Each fellow is expected to use the unique access she or he has

as grounds for research and writing on important national se-

curity issues. The SDE Air For

ce Fellows include the National

Defense Fellows, the RAND Fellows, the National Security Fel-

lows, and the Secretary of Defense Corporate Fellows. In addi-

tion, the Air Force Fellows program supports a post-SDE mili-

tary fellow at the Council on For

eign Relations.

iii

On the level of intermediate developmental education, the

chief of staff approved several Air Force Fellowships focused on

career broadening for Air Force majors. The Air Force Legislative

Fellows program was established in April 1995, with the Foreign

Policy Fellowship and Defense Advanced Research Projects

Agency Fellowship coming under the Air Force Fellows program

in 2003. In 2004 the Air Force Fellows also assumed responsi-

bility for the National Laboratories Technologies Fellows.

AIR FORCE FELLOWS

Contents

Chapter Page

DISCLAIMER . . . . . . . . . . . . . . . . . . . . . . . . . . . ii

FOREWORD . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

ABOUT THE AUTHOR . . . . . . . . . . . . . . . . . . . . xi

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

1 A NEED FOR CHANGE? . . . . . . . . . . . . . . . . . . 1

Notes

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6

2 GOLDWATER-NICHOLS AS A MODEL

FOR INTELLIGENCE REFORM . . . . . . . . . . . . 9

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3 REFORMING INTELLIGENCE:

A 50-YEAR EFFORT . . . . . . . . . . . . . . . . . . . 39

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

4 THE 9/11 REPORT AND INTELLIGENCE

REFORM LEGISLATION . . . . . . . . . . . . . . . . . 71

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96

5 THE PUSH AND PULL OF

INTELLIGENCE REFORM . . . . . . . . . . . . . . . 107

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130

6 THE QUESTION OF BALANCE . . . . . . . . . . . . . . 139

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 156

Appendix

A INTERVIEWS . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

B CHRONOLOGY OF INTELLIGENCE REFORM

PROPOSALS RELA

TED TO IC REFORM

. . . . .

159

Notes

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

163

Appendix Page

C MILESTONES IN THE INTELLIGENCE

COMMUNITY . . . . . . . . . . . . . . . . . . . . . . . . . 165

Notes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

ABBREVIATIONS . . . . . . . . . . . . . . . . . . . . . . . . 171

Table

1 Elements in intelligence community reform . . . . 151

CONTENTS

Foreword

In Intelligence Reform: A Question of Balance, Col John D.

Bansemer shares his penetrating insights into reforming the US

intelligence community (IC) to improve its performance. He of-

fers valuable guidelines for thoughtful action on this perennial

concern.

The events of 9/11/2001 resulted in national soul-searching

as we attempted to understand how such terrible events could

happen. Congress investigated these events, and the 9/11 Com-

mission studied them. Although the commission ultimately

made 41 recommendations, the ones that called yet again for

reform of the IC captured the most attention in both the press

and Congress. Why has intelligence reform been called for so

often over the years? Why has it proven so difficult to improve

the performance of intelligence agencies to anyone’s lasting

satisfaction?

In addressing these questions, there is a direction that Banse-

mer wisely does

not take, namely attempting to sort out the rela-

tive roles of intelligence failures and operational failures. Surely

futur

e studies of Hurricane Katrina will illustrate, through an

example of nearly laboratory purity, that operational failures can

occur even with nearly per

fect intelligence, hence that intelli-

gence r

eform may well need to go hand-in-hand with operational

reform to improve overall performance.

Bansemer examines the performance of the US intelligence

community by focusing on underlying tensions that are not

unique to the IC but that occur also within any large organiza-

tion. Understood this way, reform looks less like a choice among

polar preferences and more like a question of altering balances,

each of which stems from unavoidably competing interests

within an organization.

One touchstone of Bansemer’s analysis is the Goldwater-

Nichols Act (GNA) of 1986. The GNA exemplifies one way to at

tempt r

efor

m of a lar

ge organization that harbors specialized

and competing interests—in this case the Department of De-

fense (DOD) with its component military services. On the record,

Congress and the 9/11 Commission had elements of the GNA

in mind when they crafted their legislation and r

ecommenda

vii

tions, respectively. The novel insight from Bansemer’s analy-

sis is its finding that while there may be some commonality in

the symptoms (i.e., the reasons for reform) between the DOD

and the IC, the recommended course of treatment may not be

universally applicable. Crucial particulars of organizational

structures, culture, and incentives all play a role in the suc-

cess of any reform effort in improving performance.

Another key element of Bansemer’s analysis is the question of

why intelligence reform has so rarely met the expectations of the

reformers recommending change. He finds that this phenome-

non has less to do with broad organizational structure and more

with tensions among elements of the or

ganization. In the case of

the IC these tensions are heightened, relative to those in the

DOD, with its four military services by virtue of the presence in

the intelligence “community” of a larger number of much more

loosely affiliated elements with much more diverse missions be-

longing to many government departments, including State, Trea-

sury, and Homeland Security. In this “community,” no one per-

son is r

eally in charge and, short of pulling a number of

intelligence agencies out from their respective departments, no

one is ever likely to be.

At first blush the answer seems to be creating a monolithic in-

telligence establishment under a single person; however, even a

rudimentary analysis quickly finds difficulties with such an ap-

proach. Centralization may provide more control over an enter-

prise, but it risks losing the agility and mor

e tailored approaches

found in more decentralized organizations. The attractions of

centralization versus decentralization do not set up the only ten-

sion at play in the IC. Bansemer develops three more illustra-

tions from among the more important tensions: “connecting the

dots” versus “groupthink”; sharing information versus protecting

it; and the sometimes conflicting needs of the national policy

maker versus the needs of the departments. His central conclu-

sion is that, given the underlying tensions, reformers will likely

never be satisfied completely with the outcome of whatever they

may pr

opose. At any given moment, the many tensions are in an

unstable and dynamic balance.

A useful guideline for thoughtful and ef

fective action stems

om Bansemer’s findings. Pr

ecisely because they ar

e unstable

viii

FOREWORD

and dynamic, balances can be adjusted through a wide variety

of mechanisms that are available to alter the underlying ten-

sions. However, before picking up a tool, it is always better to

read the instruction manual. This leads to the most applicable

part of Bansemer’s analysis. He doesn’t prescribe solutions.

He develops instead the potential results of balance adjustments.

If one accepts that much of any reform effort is really about ad-

justing balances within organizations, then understanding the

potential impact of these readjustments becomes a critical part

of any reform effort aimed at improving performance. Bansemer

lays the foundation for improving this understanding.

In summary, there is no magic bullet for reforming an orga-

nization to improve its performance, and the changes made to

remedy yesterday’s failures may not be the ones to succeed

against tomorrow’s challenges. Instead success depends on

the ability of the organization to alter balances dynamically

and with agility—the key finding of this report.

ANTHONY G. OETTINGER

Chairman

Program on Information Resources Policy

Harvard University

FOREWORD

Colonel Bansemer is currently director of Homeland Defense

on the National Security Council staff. In his current position, he

is responsible for the formulation, coordination, and implemen-

tation of homeland defense policies with a focus on cyber secu

rity and infor

mation operations. He graduated from Roanoke

College and received his commission as a distinguished gradu-

ate of Officer T

raining School in 1988. Upon completion of ini

tial technical training, he served in a variety of staff and opera-

tional positions. In 1994 he served as a detachment commander

providing space intelligence support to Air Force Space Com-

mand. In 1999 he was assigned to the Air Staff where he was re-

sponsible for defensive infor

mation warfare doctrine and policy.

His duties included formulating policy to pr

otect Air Force infor-

mation systems and advising the director of Intelligence, Surveil-

lance, and Reconnaissance on information warfare issues. He

next attended Air Command and Staff College at Maxwell Air

Force Base, Alabama, where he was a distinguished graduate. In

2001 he assumed the r

ole of dir

ector of operations at the 23d In

formation Operations Squadron, Air Force Information Warfare

Center, Lackland Air Force Base, Texas. In 2002 he assumed

command of the squadron, which is tasked with providing the

Air Force information operations tactics, techniques, and proce-

dures. Following his selection as a National Defense Fellow in

2005, he studied at Harvard University in the Program on Infor-

mation Resour

ces Policy.

About the

Author

Lt Col John D. Bansemer

Colonel Bansemer holds master of science degrees in com-

puter science from James Madison University and military

operational art from Air University. He is married to Ellen Moore

Bansemer. They have three children: Katherine, Emily, and

Phillip.

ABOUT THE AUTHOR

xii

Abstract

On 22 July 2004 the 9/11 Commission released its report on

the events surrounding the attacks of 11 September 2001. The

9/11 Report renewed calls for reform of the intelligence commu-

nity, continuing a long series of intelligence reform efforts that

began shortly after the National Security Act of 1947 laid the

foundation of the modern IC. As reform proceeds and govern-

ment officials consider further changes, three topics remain

relevant: (1) the 1986 Goldwater-Nichols reform of the Depart-

ment of Defense and its applicability to the IC, (2) the common

findings and recommendations of past reform efforts of the IC,

and (3) the competing inter

ests inher

ent in the IC that influence

the pace and character of actual r

eform. This study explores

these topics in the context of the 9/11 Report and the subse-

quent reform efforts initiated by the executive and legislative

branches.

While there was common motivation between the latest ef-

fort to reform the IC and the earlier DOD reform effort as em-

bodied in the Goldwater-Nichols Act, it r

emains less clear if

the meas

ures taken in the DOD case ar

e equally applicable to

the IC. One reason to question the applicability of DOD reform

forts to the IC is the unique organizational context of the IC—

an interagency organization supporting multiple departments as

well as national policy makers. Reform of the IC is unlike reform

of a single cabinet-level department, for at its most basic level the

IC exists to enhance the effectiveness of multiple departments

and senior policy makers in the accomplishment of their as-

signed functions. In short, the IC serves varied interests with

sometimes shared and sometimes conflicting intelligence needs.

This organizational context suggests that successful reform re-

quir

es an on-going r

ecalibration of competing inter

ests to meet

the changing demands inherent within a dynamic national secu-

rity environment.

xiii

Chapter 1

A Need for Change?

Some of the saddest aspects of the 9/11 story are the out-

standing efforts of so many individual officials straining,

often without success, against the boundaries of the possi-

ble. Good people can overcome bad structures. They should

not have to.

—Final Report of the National Commission on

Terrorist Attacks upon the United States

Its [the 9/11 Commission’s] focus was al-Qaeda’s attack,

9/11. And I think we want to be very careful that we not

structure the intelligence community and these reforms in

such a way as to fight only that war.

—James Woolsey, former director of central intelligence,

Testimony to the Senate Governmental

Affairs Committee, “Reorganizing America’s

Intelligence Community: A View from Inside”

Background

On 22 July 2004 the Final Report of the National Commission

on Terrorist Attacks upon the United States

(hereafter referred to

as the 9/11 Report) was r

eleased to the public. The 567-page

document provided a detailed description of the 9/11 attacks

and outlined a new strategy, including recommendations in-

tended to reduce the likelihood of future attacks. The report was

notable for the unanimity of commission members, the breadth

of public interest, and the relatively quick response by Congress

and the executive branch to the recommendations.

Within two

days of publication, House and Senate leaders were promising to

reform and restructure the intelligence community (IC), and the

White House was indicating support for quick action before elec-

tions.

On 27 August Pr

es. Geor

ge W

. Bush issued executive or-

ders (EO) implementing several of the recommendations, and in

early October both chambers of Congress passed separate bills

in r

esponse to the r

eport. Finally, on 17 December, after a suc-

cessful congressional effort to synchronize the two bills, Presi-

dent Bush signed into law the National Intelligence Reform Act

of 2004. Interestingly, although the 9/11 Report contained 47

recommendations, the media and congressional spotlights fo-

cused primarily on eight of them: those dealing with restructur-

ing the IC.

The commission’s recommendations included establishing a

joint command for counterterrorism, proposed as the National

Counterterrorism Center (NCTC); creating the position of a na-

tional intelligence director (NID) to lead a restructured IC; and

improving information sharing among intelligence organiza-

tions.

The legislation subsequently enacted, and in many r

spects implemented, the core aspects of these recommendations.

Both the reform efforts and the resulting legislation were, in

some ways, reminiscent of an earlier government effort: the

Goldwater-Nichols Act of 1986, which resulted in the reform

of the Department of Defense (DOD). In fact, elements of the

Goldwater-Nichols Act appear to have guided the efforts of

both the 9/11 commissioners and Congr

ess in several key

areas of intelligence reform, including improved joint action,

altering the r

elative power of key leaders, and joint personnel

practices.

There ar

e key differ

ences as well. Unlike the DOD, which is a

single department (albeit one with four separate services), the IC

comprises 15 discrete organizations that fall under six cabinet-

level departments and encompasses a diverse mixtur

e of person-

nel, both military and civilian.

Furthermore, because of the

multidepartmental span of the IC, several congressional com-

mittees provide oversight of the community, which in itself can

influence the nature and type of reform. Additionally, such is-

sues as information sharing and analytic tradecraft are spe-

cific to the IC. However, when viewed as a model for reforming

a large government bureaucracy and not as a checklist for re-

form, Goldwater-Nichols offers valuable insight into the current

intelligence r

efor

m ef

forts.

Ultimately, as several commentators and for

mer defense of-

ficials have noted, major reform of the IC is not without trade-

fs. Centralizing intelligence may run the risk of encouraging

those aspects of gr

oupthink blamed for such pr

oblems as in

A NEED FOR CHANGE

accurate intelligence estimates that fail to incorporate alterna-

tive analyses. Conversely, decentralized intelligence without con-

comitant regimens to improve information sharing runs the

risk of failing to connect the dots. In a similar vein, moving or-

ganizational boxes, particularly across executive departments,

entails changing command and control (C

) structures with

accompanying effects, both positive and negative. Improving

information sharing, which at first glance appears relatively

innocuous, has consequences as well, including the negative

effects of information overload and potentially damaging dis-

closures of sensitive intelligence to adversaries. This is not to

say that r

eform is not needed or necessary. However, reform

may be more successful when the competing interests at stake

are understood and considered.

Finally, questions remain. Why, when over 20 studies of the

IC have been undertaken since 1947, did this particular re-

form effort develop so much traction with Congress and the

executive branch?

Was it because the events of 9/11 were so

horrific and so palpable that elected of

ficials felt they had no

choice but to act? Was it because the report established the

clear

est link yet between structural impediments in the IC and

the fatal consequences of intelligence failures? Was it because of

the strength of various 9/11 advocacy gr

oups and the prox

imity

of a presidential election? As important as these questions are,

a more fundamental one remains: Will the 9/11 Commission’s

recommendations and the resulting legislation adequately

address the problems identified in the commission’s report

as well as those in previous studies and investigations of the

US intelligence community?

This question gets to the heart of the problem with intelli-

gence reform. Adjustments designed to address one aspect of

the IC may have unintended consequences for other aspects of

the intelligence effort. Thus, an examination of the competing

tensions associated with proposed recommendations provides

a mechanism to analyze the potential impacts of r

efor

m. As

this study illustrates, r

eform of the IC involves numerous

competing tensions and associated trade-offs with significant

implications. Because of these complexities, further compli

cated by the interagency natur

e of the IC, this paper does not

A NEED FOR CHANGE

purport to offer an easy solution. In fact, there likely is no sin-

gle solution. Instead, successful reform requires the creation

and use of mechanisms that allow the dynamic recalibration

of these tensions in response to an ever-changing organiza-

tional context.

Scope and Organization

Both the 9/11 Report and statements by commission lead-

ers suggest that the commission had in mind various aspects

of the 1986 Goldwater-Nichols legislation when proposing

their recommendations. Of particular relevance are the re-

port’s discussions on joint action, unity of effort, and the re-

structuring of the IC.

Rather than analyze the direct applica-

tion of the 1986 act to intelligence refor

m, chapter 2 examines

the evolution of DOD reform and the underlying tensions in

play during the reform effort. As the chapter describes, the

DOD reform effort centered on three key areas: the failure of

joint action, the structure of the Joint Chiefs of Staff (JCS),

and the quality of joint personnel. Because each of these areas

has some commonality with the current intelligence reform ef-

forts, the problems and legislative prescriptions mandated by

Goldwater

-Nichols r

epresent a practical model of the r

eform of

a large-scale government bureaucracy. The questions ad-

dressed are: What was the legislation designed to improve?

What structural and procedural changes were made as a re-

sult of the legislation? and How effective was the legislation in

meeting its intended aims?

Chapter 3 turns to previous intelligence reform efforts to

search for commonalities in their recommendations and find-

ings. This discussion helps to identify where the 9/11 Commis-

sion’s r

ecommendations echo or dif

fer fr

om the findings of

pre-

vious studies. This analysis represents the first step in

understanding the potential ramifications of various actions

for the IC writ large. The review of over 50 years of intelligence

reform efforts illustrates, among other things, a gradual shift

toward greater centralization of intelligence, an intelligence

system that has evolved in response to changes in the inter-

national envir

onment far more than some reform advocates

A NEED FOR CHANGE

would suggest, and the continued relevance of issues that

have faced decision makers since the late 1940s.

Chapter 4 examines in detail the 9/11 Commission’s recom-

mendations and the legislative and executive branch re-

sponses to the report. Of the five sets of recommendations re-

lated to the IC, the first three groupings—improved joint

action, improved structure, and improved information shar-

ing—share elements with DOD reform efforts. The final two

sets, related to improving congressional oversight and domes-

tic intelligence efforts, are unique to the IC. Next, this chapter

reviews the measures taken by the executive branch in late

August to addr

ess the 9/11 Commission’s recommendations

and then analyzes the legislation passed in December. To bet-

ter assess the potential success of the recent reform efforts,

the chapter concludes by comparing the recent intelligence re-

form legislation with the Goldwater-Nichols Act to see how

each piece of legislation attempted to improve joint action,

alter the power relations among the primary parties, and im-

ove joint personnel practices.

Chapter 5 offers a fuller examination of the competing inter-

ests and or

ganizational exigencies that hang in the balance of

any major reform of the IC. Although any successful reform

should consider a variety of balances, determining the pr

oper

extent of centralization versus decentralization may be the

most fundamental because it affects so many of the other as-

sociated balances. As a historical r

eview shows, the IC has

gradually become more centralized over time, although certain

barriers make further centralization increasingly difficult.

After examining the arguments on both sides of the centraliza-

tion/decentralization issue, the focus shifts to three associ-

ated tensions: the tension between centralized and decentral-

ized analysis, greater information sharing and the need to

protect information, and the intelligence needs of national pol-

icy makers and the more tailored and operational intelligence

needs of the IC’s component departments, particularly the

DOD. The need to balance these competing inter

ests compli-

cates reform efforts and suggests one reason why the IC has

followed a gradual, evolutionary path rather than one mor

closely r

esembling punctuated equilibrium with dramatic

A NEED FOR CHANGE

changes followed by relatively long periods of stability. After

noting other tensions worthy of further study, this chapter

concludes with a brief examination of such further relevant

factors as internal organizational structures, organizational

context, and the quality of organizational personnel and lead-

ers that may prove key to any successful reform effort.

Chapter 6 summarizes the previous analysis in more con-

densed form. The success of the latest reform efforts depends

on multiple factors, including how well reformers diagnose the

perceived ills within the IC and whether the proposed treat-

ment effectively addresses the underlying issues. Reformers

have a variety of tools available. Legislative measures are usually

relatively permanent and broad in their focus and prescrip-

tions, while the tools of the executive branch of

fer greater pre-

cision and timeliness to adjust conditions as needed. To be ef-

fective, these tools need to be applied to multiple aspects of an

organization ranging from the organizational structure, to the

relative power of key leaders, to incentive structures, to the

underlying support infrastructure. Thus, true reform likely de-

pends primarily upon the inter

nal efforts of the IC itself. The

degree of cooperation between agencies and departments;

changes in community policies, practices, culture, and in the

underlying supporting infrastructure; and—most important—

the effectiveness of the community’s leaders and personnel all

play important roles in any r

efor

m effort. In the end, adjusting

the fulcrum to achieve the desired balances may be more art

than science; however

, such an adjustment can be informed

by potential or even likely outcomes and consequences—the

ultimate objective of this research.

Notes

1. 9-11 Commission Media Advisory, “9/11 Commission Releases Unan-

imous Final Report,” http://www.9-11commission.gov/pr

ess/pr_2004-07

-22.pdf (accessed 4 October 2004).

2. Dan Eggen and Helen Dewar, “Leaders Pick Up Urgency of 9/11 Panel;

Congress and Bush Vow to Speed Refor

ms,”

ashington Post

, 24 July 2004,

A01.

Final Report of the National Commission on Terrorist Attacks upon the

United States

(hereafter referred to as the 9/11 Report) (New York: W.W. Nor-

ton & Company, 2004), 400–418. For a discussion of the NCTC see 400–406,

A NEED FOR CHANGE

or a discussion of restructuring the IC and establishing the NID see 407–15,

and for a discussion of information sharing, see 416–18.

4. Statement of Senator Joseph Lieberman (D-CT), 108th Cong., 2d sess.,

Congressional Record (8 December 2004): S11968. In remarks on the Senate

floor, Senator Lieberman stated, “Just as, in the military, the Goldwater-

Nichols Department of Defense Reorganization Act of 1986 sought to integrate

the military services’ capabilities by strengthening the combatant command-

ers, so this legislation fosters greater jointness among the intelligence agen-

cies.” See also remarks by Rep. Jane Harman (D-CA) who stated on the House

floor, “This long-overdue legislation will modernize our capabilities, integrate

our intelligence collection and analysis efforts, unify our counterterrorist ef-

forts and promote intelligence sharing. It will promote the same jointness in

intelligence that has been the hallmark of our military’s success since the

Goldwater-Nichols Act of 1986.” Statement of Rep. Jane Harman, 108th

Cong., 2d sess.,

Congressional Record (7 December 2004): H11004.

5. The current IC comprises the Central Intelligence Agency (CIA), Defense

Intelligence Agency (DIA), National Security Agency (NSA), National Reconnais-

sance Agency (NRA), National Geospatial-Intelligence Agency (NGA), and the in-

telligence elements of the Army, Navy, Air Force, Marines, Coast Guard, Fed-

eral Bureau of Investigation (FBI), Department of State, Department of

Treasury, Department of Energy, and Department of Homeland Security (DHS).

See “The Intelligence Community in a Nutshell,” http://www.intelligence.gov/

ic_in_a_nutshell/index.htm (accessed 5 October 2004).

6. Since the establishment of the initial outlines of the IC by the

National

Security Act of 1947

, numerous investigations, reform efforts, and commis-

sions have examined the intelligence process and provided recommenda-

tions for reform. The total number of these studies ranges from somewhere

over 20 to nearly 40 depending on the metric used to describe a reform ef-

fort. This paper focuses on the reform efforts identified in Richard A. Best,

Proposals for Intelligence Reorganization, 1949–2004 (Washington, D.C.:

Congressional Research Service, 29 July 2004).

7. 9/11 Report, 403, 408, and 412. Page 408 in particular discusses

Goldwater-Nichols and joint commands as well as incentives for DOD joint

officers.

A NEED FOR CHANGE

Chapter 2

Goldwater-Nichols as a Model

for Intelligence Reform

We believe that the intelligence community needs joint analy-

sis, joint collection and joint management. It needs a whole-

sale reform, and we believe it ought to be modeled in the

Goldwater–Nichols reforms.

—Thomas Kean, chairman, 9/11 Commission,

Testimony to the House Permanent Select

Committee on Intelligence, US House of

Representatives, 11 August 2004.

The 9/11 Commission report, individual commission mem-

bers, congressional witnesses, and former government officials

explicitly cited the Goldwater

-Nichols Defense Reorganization

Act of 1986 as a model for reform of the IC.

At least two thought

processes appear to account for this. On the one hand, some r

formers seemed to view Goldwater-Nichols as a paradigm for re-

forming a large bureaucracy.

In this instance, reformers fo-

cused on Goldwater

-Nichols as illustrative of legislative

solutions that solve large-scale bureaucratic problems. Other

refor

mers, notably some members of the 9/11 Commission and

of Congr

ess, had specific elements of Goldwater-Nichols in mind

as they crafted particular elements of legislation and recom

mendations; thus, they used the legislation more as a template

for action.

Goldwater-Nichols is also notable because the tensions in-

volved resemble those addressed in the current intelligence

debate. In his book on DOD r

eorganization, Gordon Lederman

suggests that three fundamental tensions were present and

that the need to find the proper balance among these tensions

guided the efforts of reformers.

They involved (and continue

to involve) tensions of centralization versus decentralization,

functional versus geographic r

esponsibility, and specialist ver

sus generalist perspectives.

The tension between centralization and decentralization fun-

damentally concer

ns where decisions are made in an organiza-

tion. In theory, greater centralization provides more unified di-

rection at the expense of less efficient execution, whereas decen-

tralization can foster innovation and speed reaction to unfolding

events while increasing the risk of uncoordinated efforts and

diffused responsibility. The geographic/functional tension re-

sembles the tensions in the business world over whether to or-

ganize around functional or product lines. In the military con-

text, they involve decisions to organize around specific

geographic r

egions, which allow control of all the military assets

in the r

egion and a stronger understanding of regional issues

(e.g., the unified commands of US Pacific Command and US Eu-

ropean Command [USEUCOM]) versus placing all the military

assets dedicated to a particular purpose under a global com-

mander (e.g., the functional command of US Transportation

Command). Finally, tension exists between the need to master

specific skills within the profession of military arms (specializa-

tion) versus the need to understand the broad concepts of mili-

tary art to better synchronize the full potential of all contribut-

ing military arms (generalization). As Lederman illustrates,

there is no perfect solution. The balances shift to reflect a

changing environment and the guiding priorities of government

leaders, which are modulated by opposing and competing advo-

cates within the military establishment.

Beyond this common backgr

ound, the Goldwater-Nichols

legislation offers other important analogies to the present in-

telligence r

eform initiatives. In particular

, issues that

prompted intelligence reform efforts after 9/11 have much in

common with those identified in a series of reports, congres-

sional staff studies, and testimony that led up to the passage

of Goldwater-Nichols. These commonalities include failures of

joint action; structural barriers that inhibited ef

fective opera

tions, particularly dual-hatting of key personnel and inade

quate authority; and policies that deterr

ed high-quality per

sonnel fr

om pursuing joint duty.

Given the similar impetus

for intelligence r

eform and Goldwater-Nichols, the act’s poten-

tial applicability, and its use as a historical reference, an ex-

amination of the events that preceded the act’s passage serves

as a good starting point to better understand why r

efor

m was

GOLDWATER-NICHOLS AS A MODEL FOR IR

deemed necessary and what elements of the DOD the legisla-

tion was designed to improve.

Why Goldwater-Nichols?

Goldwater-Nichols arose from a confluence of multiple fac-

tors, including poor or suboptimal military performance in a se-

ries of operations (e.g., the failed rescue of Iranian hostages [Op-

eration Eagle Claw], the invasion of Grenada, and the bombing

of the Marine barracks in Beirut), increasing calls for reform

from both within and outside the DOD, and greater congres-

sional activism resulting from perceived intransigence of senior

DOD officials regarding calls for reform.

While these factors are

important for understanding the context surrounding passage

of Goldwater-Nichols, more relevant to this report is an under-

standing of the problems targeted by the legislation.

Although studies and historical analyses of Goldwater-

Nichols cite multiple issues and problems in need of solutions,

this chapter focuses on three areas commonly identified by

various r

esear

chers and commentators on the act.

These cor

issues include (1) the failure of joint action and the associated

imbalance of responsibility and authority, particularly as it re-

lated to the combatant commanders of the unified commands;

(2) the structure of the Joint Chiefs of Staff (JCS), the dual-

hatting of the service chiefs, and the effects of this organiza-

tional structur

e on the military advice provided to the pr

esi

dent; and (3) the poor quality of personnel filling joint duty

positions. Each of these areas shares features with problems

identified within the IC.

The Failure of Joint Action

Before examining the concept of joint action, a few distinc-

tions are necessary. The DOD defines joint as “activities, oper-

ations, or

ganizations, etc., in which elements of two or more

Military Departments participate.”

Joint action can thus be

partially distinguished from unified action, where activity oc-

curs under the auspices of a unified commander and includes

the “synchronization of activities with other governmental and

nongover

nmental agencies.”

The following discussion uses

GOLDWATER-NICHOLS AS A MODEL FOR IR

the term joint action to denote activities that are purely joint in

nature, did not necessarily occur under a unified commander,

and did not involve synchronization with other governmental

agencies. This distinction becomes important, because the

9/11 Report (without a supplemental definition) also uses the

term

joint action, apparently to denote action involving two or

more intelligence agencies or organizations.

The

Rescue Mission Report, the official report on the Iranian

hostage rescue mission, attributed the operation’s failure to mul-

tiple causes and highlighted 11 major problems, of which two

focused on C

issues.

The report cited the lack of unified plan-

ning, separate training that pr

ecluded full mission rehearsal,

and significant operational issues including weather as factors

that combined to create the conditions for failure. The report

also noted, “From [commander, Joint Task Force] COMJTF

downward command channels were less well defined in some

areas and only implied in others.”

A separate examination of

the event highlighted the difficulty in diagramming a chain of

command, noting such a diagram was “litter

ed with dotted

lines of coordination and instances of one unit responding to

two lines of command.”

Taken in isolation, the failure of the operation was not enough

to compel refor

m. However, two other events in October 1983

also contributed to a general sense that military reform might

be necessary: the terrorist bombing of the Marine barracks in

Beirut and the US invasion of Gr

enada.

Investigations of the Marine barracks bombing attributed

the tragedy primarily to the inadequacy of security measures.

These security conditions were, in turn, symptomatic of a larger

issue of failed unity of command. Joint doctrine describes the

concept of

unity of command as “all forces operat[ing] under a

single commander with the requisite authority to direct all

forces employed in pursuit of a common purpose.”

In the

case of Beirut, Gen Bernard Rogers, USA, the USEUCOM com-

mander in char

ge of the security situation in Beirut, had no

eal authority to revise or enforce security procedures. The

House Investigations Subcommittee found failures of commis-

sion at the lower levels of command and failur

es of omission

in the for

m of inadequate oversight by elements further up the

GOLDWATER-NICHOLS AS A MODEL FOR IR

chain of command, to include General Rogers.

In addition,

the subcommittee struggled with conflicting testimony on the

structure of command relationships and found “the lack of

clarity in the chain of command might contribute to an inabil-

ity to fix responsibility and authority.”

While General Rogers

accepted responsibility in his testimony before the subcom-

mittee, evidence indicates that he lacked the “requisite author-

ity” to alter the events that transpired. For example, USEU-

COM had offered antiterrorist training fully 11 months before

the attack, but the Marines had r

ejected the offer. This illus-

trated the relative impotence of a combatant commander to di-

rect actions within his own theater.

The disparity between responsibility and authority was dis-

cussed in the aftermath of the incident. A report by Lt Gen

John H. Cushman, USA, retired, that later influenced congres-

sional reformers specifically spoke to General Rogers’s limited

authority, citing his inability to effect “such matters as admin-

istration, discipline, internal organization, and unit training

except when a subordinate commander requests assistance.”

Adm William Crowe, then commander of the unified Pacific Com-

mand, supported this finding, stating, “The unified commander

had all the r

esponsibility, he did not have sufficient authority.

His component commanders reported to their own service chiefs

for administration, logistics, and training matters.”

A br

oad

analysis of unified command structur

es befor

e Beirut foretold

potential problems, noting that combatant commanders had

“insuf

ficient authority and influence in peacetime to ensur

their war-fighting capabilities were commensurate with their

mission.”

Another event in October further strengthened the calls for

reform. On 25 October 1983 the United States invaded Grenada,

in part to pr

event the island fr

om becoming a satellite of Cuba

as well as to pr

otect American citizens on the island. Although

the operation achieved its objectives, numer

ous C

and inter

operability problems surfaced that might have had dire conse-

quences had US for

ces faced a mor

e competent adversary. The

operation again violated the principle of unity of command, as

the ground elements of Marine and Army forces fought under

separate chains of command.

As then-Maj Gen Colin Powell

GOLDWATER-NICHOLS AS A MODEL FOR IR

said at the time of the invasion, “Relations between the services

were marred by poor communications, fractured command

and control, [and] interservice parochialism.”

On the inter-

operability front, the Army was unable to communicate with

Navy ships offshore to obtain necessary supporting fires.

Be-

sides interoperability problems, the assessments highlighted

poor joint planning, lack of service expertise in joint headquar-

ters, and lack of unity of command.

The investigations of these and other military operations

yielded a consistent theme of inadequate unity of command

and ineffective joint action. They also validated the findings of

a 1970 Blue Ribbon Defense Panel report that found “an im-

balance between the responsibility and authority of unified

commanders.”

The joint capstone document on unified ac

tion states that

unified action of the Armed Forces of the United States starts with uni-

fied direction. For U.S. military operations, unified direction normally

is accomplished by establishing a joint force, assigning a mission or

objective to the joint force commander (JFC), establishing command

relationships, assigning and/or attaching appropriate forces to the

joint force, and empowering the JFC with sufficient authority over the

forces to accomplish the assigned mission.

This description, written some 15 years after Goldwater-

Nichols came into for

ce, succinctly describes the military prin

ciples that serve, at least in part, as a prescriptive formula to

avoid the identified causes of some operational failures of the

past. The military operations outlined above suffer

ed to vary-

ing degr

ees from failur

es to establish appr

opriate command

relationships and empower commanders with sufficient au-

thorities.

Joint Chiefs of Staff Structure

The structure of the JCS was another major consideration

of reformers. Criticism of that structure came from within and

without military or

ganizations. Former secretary of defense

James Schlesinger succinctly described the problem, stating,

“The central weakness of the existing system lies in the struc-

ture of Joint Chiefs of Staff. . . . The existing structure, if it

does not preclude the best military advice, provides a substan-

tial, though not insur

mountable, barrier to such advice.”

GOLDWATER-NICHOLS AS A MODEL FOR IR

ternally, criticism came from the chairman of the JCS (CJCS),

Gen David C. Jones, USAF, who outlined the need for reform

in testimony before the House Armed Services Committee and

later in the press. His criticism was especially notable because

it came from within the military establishment: from a still-

serving general officer.

Much of his criticism focused on the overall structure of the

JCS, particularly the role of each of the service chiefs vis-à-vis

their joint and service responsibilities. Prior to enactment of

the Goldwater-Nichols legislation, the service chiefs exercised

significant control over Joint Staff action.

Likewise, military

advice emanating fr

om the JCS required agreement of all the

service chiefs, and the law required them to inform the president

of dissent.

This statutory requirement resulted in the com-

mon critique that the JCS took the “lowest common denomi-

nator” approach towards decision making.

In fact, such splits

within the JCS occurred only a “few times per year” and were

referred to the secretary of defense (SecDef).

Further pr

oblems centered on the Joint Staff. The organiza-

tional structure in place gave each of the service staffs a de facto

veto over the staf

fing process, with the r

esult that staff papers

were often watered down so as not to offend a service or to pro-

tect individual service programs or desir

es.

As the sole military

officer without concurrent service obligations, the CJCS was the

only person who could realistically be expected to rise above ser-

vice par

ochialisms. However, the Joint Staff worked for the en-

tire JCS, not just the chairman, which meant that the chairman

could not obtain independent, nonservice-centric advice and

support.

Moreover, the CJCS had no official deputy, which

made it hard to maintain continuity when the chairman traveled,

a problem compounded by the three-month rotation of the ser-

vice chiefs in the nominal deputy role.

Related to the structural problems of the JCS was the dual

role of the chiefs of the individual services as both the heads of

their service and members of the JCS. Underlying this dual na

tur

e is a fundamental tension between service prerogatives and

joint action.

In his initial testimony to the House Armed Ser-

vices Committee General Jones noted the institutional pr

essures

that each of the service chiefs felt in r

epr

esenting the inter

ests of

GOLDWATER-NICHOLS AS A MODEL FOR IR

their services while also crafting and supporting unified deci-

sions.

This pressure created a spokesman-statesman dilemma

for the service chiefs.

Gen Edward C. Meyer, USA, another sit-

ting member of the JCS, supported this view, stating that “dual-

hatting, dictated by law, confers real power with the service chief

hat and little ability to influence policy, programming, and

budget issues with the joint hat.”

The statutory requirement

that the service chiefs focus on the development of their respec-

tive services further compounded the preeminence of the service

perspective over the joint perspective.

This dominance of the military services over joint organiza-

tional structures gets to the heart of Lederman’s specialist/

generalist tension.

Wearing their specialist hats, the service

chiefs supported service needs over the broader needs repre-

sented by the generalist hat. According to reformers, multiple

factors produced this imbalance, including the imperative for

the service chief to maintain support within his service and

veterans’ groups as well as the sheer lack of time for the ser-

vice chief to execute both service and joint responsibilities ade-

quately.

The P

oor Quality of J

oint Personnel

At least two factors af

fected the overall quality of joint per-

sonnel: lack of experience and negative perceptions of the ca-

reer impact of joint duty. These factors wer

e present even

within the Joint Staff. General Jones and others pointed to in-

adequate experience and competence, which resulted at least

partially from the statutory three-year time limit on Joint Staff

membership.

A study of the Joint Staff initiated by General

Jones, the Chairman’s Special Study Group, found that “only

2 percent of officers in the [Office of the Joint Chiefs of Staff]

OJCS during 1982 had any previous Joint Staff experience.”

At the highest echelons of power, less than 60 percent of gen-

eral officers and admirals had served in joint assignments.

The 1977 Defense Organization Study of the DOD supported

these findings, stating that the “services have consistently

avoided assigning their most highly qualified officers” to the

Joint Staf

The lack of experience meant that joint person

GOLDWATER-NICHOLS AS A MODEL FOR IR

nel were likely to turn to their specialist training and back-

ground while fulfilling their duties on the staff.

Compounding the problem was the perception that joint

duty harmed an officer’s career. One reason for such a percep-

tion may have been that the services controlled promotions,

and thus officers serving on the Joint Staff may have believed

that supporting joint positions at the expense of service inter-

ests could jeopardize their promotion opportunities.

Dis-

cussing the overall quality of joint personnel, Admiral Crowe

noted, “It was unusual to find the most highly regarded offi-

cers laboring in the Joint Staff vineyard; many considered a

tour there as a hurdle on the career path.”

Further exacer-

bating the problems with the Joint Staff, only 13 percent of of-

ficers on the Joint Staff had attended joint training prior to as

suming joint duty.

Accusations even arose that the services

intentionally placed lower quality officers on the Joint Staff to

further obstruct the staf

fing process.

This underscored the tension between specialization and

generalization. The Joint Staff suffered in many respects be-

cause it consisted of specialists who had neither the experi

ence nor the training to foster a more generalist perspective. A

negative feedback cycle r

esulted, with joint personnel having

inadequate joint experience because they consciously avoided

joint assignments whenever possible. As long as service mem-

bers believed that joint duty harmed their careers, there was

little likelihood of improving the experience level on the Joint

Staff, which further diluted its effectiveness.

The preceding description should not be interpreted as a full

account of the issues leading to the passage of Goldwater-

Nichols. However

, all of the issues mentioned contributed to

an environment where reform was deemed necessary. Reform-

ers saw the system as unbalanced, with too much control

vested in the services at the expense of joint action. This im-

balance manifested itself in an overemphasis on decentraliza-

tion and specialization and a tr

end towards functional ap-

oaches. A r

ecalibration of these balances thus became a

primary emphasis of the new legislation.

GOLDWATER-NICHOLS AS A MODEL FOR IR

Goldwater-Nichols Legislation

The framers of the legislation set forth eight objectives:

1. to reorganize the Department of Defense and strengthen

civilian authority in the Department;

2. to improve the military advice provided to the President,

the National Security Council, and the Secretary of De-

fense;

3. to place clear responsibility on the commanders of the

unified and specified combatant commands for accom-

plishment of missions assigned to those commands;

4. to ensure that the authority of the commanders of the

unified and specified commands is fully commensurate

with the responsibility of those commanders for the ac-

complishment of missions assigned to their commands;

5. to increase attention to the formulation of strategy and

to contingency planning;

6. to provide for more efficient use of defense resources;

to impr

ove joint officer management policies; and

otherwise to enhance the effectiveness of both military

operations and DOD management.

Of these objectives, 2, 5, 6, and 8 related primarily to im-

proving the structure and effectiveness of the JCS and ad-

dressed conflicts r

esulting fr

om the dual-hatting of the service

chiefs. Objectives 3, 4, 5, and 8 related to improving joint op-

erations, with 3 and 4 specifically addressing the imbalance

between authority and responsibility. Objective 7 specifically

focused on the quality of the joint personnel.

Goldwater-Nichols brought about notable changes by mak-

ing the CJCS the principal military advisor to the president

and the SecDef and establishing the position of vice chair

man

(not a voting member of the JCS). This meant that the pr

eemi

nent senior military advisor was now also the leader who by

design had a generalist perspective and, in turn, removed some

of the underlying pressures on the service chiefs in their sol-

dier/statesman roles. The service chiefs could continue to ad-

vocate their service perspectives in the knowledge that the

CJCS or vice chair

man, Joint Chiefs of Staf

f (VCJCS) could

GOLDWATER-NICHOLS AS A MODEL FOR IR

modulate the competing service perspectives in offering advice

to the president or SecDef. The service chiefs still wore two

hats; however, by increasing the power and prestige of the

chairman, the legislation mitigated some of the prevalent spe-

cialist tendencies. In essence, the greater independence of the

CJCS reduced the institutional pressures to seek unanimity at

all costs, thus avoiding the lowest common denominator ap-

proach previously mentioned.

The legislation further evidenced a leaning towards the gen-

eralist perspective by the functions it assigned to the CJCS.

The act made the chairman the spokesman for the combatant

commands; he would advise the SecDef on their needs and

priorities and seek to improve the alignment of the services’

budgets with their requirements.

To create stronger founda-

tions for joint culture, the legislation made the chairman re-

sponsible for formulating joint doctrine and education and

training policy. It also placed the Joint Staff under the direct

control of the chairman, expanded the length of terms for joint

duty, and incr

eased the number of officers authorized on the

staff from 400 to 1,627.

The Goldwater

-Nichols Act partially addressed the r

esponsi-

bility versus accountability issue raised by the combatant

commanders by requiring that all units within a geographic

region be assigned to the combatant command.

The act also

laid out seven aspects of that authority, including the ability

to establish the chain of command, to select subor

dinate com-

manders, to organize forces with the command, and to con-

vene courts-martial for disciplinary purposes.

Moreover to

further preclude the possibility that a combatant commander

could operate with insufficient authority, the act required

combatant commanders to promptly notify the SecDef if they

considered their authority insufficient.

To limit the power of

the services vis-à-vis the combatant commanders, subordi-

nate commanders were placed “under the authority, direction

and command” of the combatant commander

Finally, di

ectly subordinate commanders could only be assigned by the

services with the concurrence of the combatant commander,

who was also given the authority to evaluate and suspend

subor

dinate commanders.

These changes tilted the balance

GOLDWATER-NICHOLS AS A MODEL FOR IR

of power towards the combatant commanders and away from

the services.

The legislation also considered the quality of joint person-

nel. The act describes the requirements for joint specialty offi-

cers (JSO), including education and joint assignments.

It re-

quires that 50 percent of joint duty assignments at the grade

of O-3 (captain) and above be filled by JSOs.

To address the

negative perceptions of career impacts, the act required that

the promotion rates of JSOs and officers currently or previ-

ously serving on the Joint Staff match the rate for officers of

their respective services’ headquarters staffs.

It also in-

creased the minimum duration of joint assignments and man-

dated that officers promoted to flag rank have served in at

least one joint duty assignment.

These changes increased

the likelihood that senior leaders would have broader, more

generalist perspectives than previously, when officers could

specialize within a service and have limited exposure to joint

operations and the perspectives of other services.

Legislative Effectiveness

Relatively few salient metrics exist to prove or disprove the

success of Goldwater

-Nichols in r

edressing the fundamental

imbalances in the DOD. One way to understand the impact of

Goldwater-Nichols is to examine military effectiveness in such

joint operations after the act took effect as Desert Storm (the

1991 Gulf War) and Allied For

ce (the North Atlantic Treaty

Organization [NATO] operation directed against Slobodan

Miloˇsevi´c and Yugoslavia). These operations and the degree to

which the forces operated in a joint manner in terms of uni-

fied command and discernible interoperability provide some

indications of the legislation’s ef

fectiveness.

As to the structure of the JCS, the quality of advice is, by

implication, a qualitative measure that depends in large part

on the relative confidence and satisfaction expressed by the

president and senior officials within the DOD.

Several recent

studies have examined the larger aspect of civilian-military re-

lationships since 1986; however, they provide less commen-

tary on the overall str

ength and quality of advice afforded to

GOLDWATER-NICHOLS AS A MODEL FOR IR

civilian leaders.

Fortunately, more concrete data exist on the

overall effectiveness of increasing the promotion rates of mem-

bers of the Joint Staff, with the underlying implication of in-

creased quality of personnel.

The Effectiveness of Joint Action

It is difficult to compare the pre–Goldwater-Nichols military

operations discussed earlier with the much larger operations

that have occurred since.

The post–Goldwater-Nichols opera-

tions described below involved far greater numbers of forces

and resources. That these operations are generally considered

mor

e successful than their earlier counterparts suggests that

fundamental changes occurred since the legislation’s passage.

The primary causal factor for improved operations remains a

topic of debate, with some analysts suggesting advanced tech-

nology or the relative weakness of the opposing adversaries as

the primary factor contributing to success. However, another

factor was at work: more effective C

in joint operations. This

does not imply that problems with C

and interoperability had

disappeared; however, when compared with pre-reform opera-

tions, their impact appears to have been less pr

ofound.

Operation Desert Storm. Operation Desert Storm was the

first major conflict after the implementation of Goldwater-

Nichols and secured the role and power of the combatant com-

mander in executing military operations. Multiple levels of

analysis can be applied to Desert Storm, from the combatant

commander’s control of the operation, to interservice unity

and disunity of action, to the most basic level of interoperabil-

ity. At the strategic level, Gen H. Norman Schwarzkopf, the

combatant commander, was firmly in control of the operation

and exercised significant control of forces assigned, including

control of logistics. He also successfully rebuffed interference

om the services.

For example, Schwarzkopf denied a Ma

rine request for an amphibious assault over the objections of

the Marine commandant, who tried to pursue the request

thr

ough the CJCS.

Likewise, joint considerations wer

e at

play in the modification of the air campaign. In this case, the

initial plan calling for predominantly strategic leadership at-

tacks against Iraqi political, economic, and C

tar

gets was later

GOLDWATER-NICHOLS AS A MODEL FOR IR

counterbalanced by ground preferences for attacks against

Iraqi fielded forces. The result was a more balanced, phased

air campaign that was enormously successful in shortening

the ensuing ground campaign.

An examination of the air campaign provides the best in-

sight into interservice unity. The sheer number of aircraft in-

volved (1,600–1,800) and the complexity of the air campaign

(over 40,000 sorties) attest to the complexity of the opera-

tion.

Managing this complex operation was a joint force air

component commander (JFACC) who, unlike in previous com-

bat operations, was able to unify the vast majority of the air ef-

fort, with the notable exception being the Marines, who with

held approximately 50 percent of their F/A-18 aircraft to

support specific Marine operations.

The after-action report

to Congress found that the JFACC “secured unity of effort.”

The report also recommended that the JFACC be staffed by

members from all services.

An assessment of Desert Storm

lessons by the chairman of the House Armed Services Com-

mittee found that “combining r

esponsibility and unquestioned

authority in the [commander in chief] CINC—in this case as

delegated to his Joint For

ces Air Component Commander—for

planning and deployment of all theater aircraft optimized the

achievement of campaign objectives.”

While interservice unity improved greatly on the operational

front, planning was less unified, with the services dominating

planning for operations within their particular medium. The

Air Force controlled the initial planning of the air campaign,

which (as noted above) was eventually modified to accommo-

date the needs of the other services.

Also, contrary to the in-

tent of Goldwater-Nichols, an office at Air Force headquarters

developed the initial air plan rather than designated planning

elements within Central Command (CENTCOM).

In addition,

the other services complained of Air Force dominance in near-

term planning (operations to be conducted within 24 to 48

hours), particularly as it r

elated to the pr

osecution of specific

tar

gets.

On the gr

ound side, the Army relied initially on a

special team of planners to develop the concept of the ground

fensive strategy, excluding the other service component with

a significant r

ole in gr

ound operations—the Marines.

In both

GOLDWATER-NICHOLS AS A MODEL FOR IR

the air and ground cases, initial planning was stovepiped and

relied upon ad hoc planning arrangements. Only later itera-

tions of the planning process included the views and require-

ments of the other services.

In the area of interoperability, control of the airspace was re-

markably effective, especially considering the numbers and mul-

tiple types of US and allied aircraft that supported operations. All

services shared a variety of space systems, including the newly

launched Global Positioning System (GPS) satellites, which made

important contributions to the war effort.

However, negative ex-

amples abound as well; for instance, the lack of adequate com-

munications between Air For

ce and Navy units necessitated that

the nightly air tasking order be flown to the carriers.

Tactical

communications were often “plagued by incompatibilities and

technical limitations.”

Finally, “stovepiped intelligence and

command information systems” were also prevalent and affected

support to tactical forces.

While these problems cannot be dis-

missed, some of them can be attributed to the long acquisition

and pr

ocurement timelines for many systems and the fact that

these legacy systems were often in the pipeline long before the

Goldwater

-Nichols refor

Operation Allied Force. The DOD did not completely re-

solve the issues highlighted above in the years intervening be-

tween the end of Desert Storm and the onset of Allied Force

(Yugoslavia). Like Desert Storm, Allied Force was a joint and

combined operation; however

, the operation faced even more

political constraints because of NATO alliance concerns.

addition, the operation was predominantly an air war, with

land options not even considered at the time of the conflict. As

with any operation, postevent analysis identified problems.

Much of the criticism focused on the strategic limitations cre-

ated by US and allied leaders, including shifting political pri-

orities, explicit lack of a ground option, and charges of incre-

mentalism/creeping escalation.

Criticism of the military aspects of the operation found ar

eas

for impr

ovement, including evidence of service parochialism in

the choice of targets.

This is not entirely unexpected, as in most

major military campaigns each service has decided tar

geting

efer

ences. Such evidence indicates that the specialist perspec

GOLDWATER-NICHOLS AS A MODEL FOR IR

tive had not been eliminated; however, this may never have been

the intention of the reformers. An examination of the targets ul-

timately chosen over the course of the operation reveals the in-

fluence of both Army and Air Force perspectives. In effect, the

views of one service did not predominate over those of the other,

as the reformers intended.

With regard to C

, one analysis found “problems . . . were

further aggravated by a confusing chain of command, unsuit-

able organizational structures, and a lack of staff integration

where it was needed most.”

Some of the problems resulted

from parallel chains of command: one for allied (NATO) opera-

tions and one for each participating allied country—but those

in the strictly US chain of command are harder to justify. Ac-

cording to some accounts, command relationships were estab-

lished outside of normal doctrine and assessments found

“command echelons being skipped or omitted entirely, and ei-

ther a duplication of effort or functions not being performed at

all, since one organization erroneously thought the other was

responsible for a particular task.”

The unanswer

ed question

is whether the operation succeeded in spite of poor C

relation-

ships or whether those r

elationships were not as untenable as

some critics suggest.

A RAND study faulted the lack of joint air-land planning and

the failure to integrate Army assets into the air campaign early

on; however, much of the criticism targets the civilian leader-

ship for taking gr

ound options off the table.

Inter

estingly, the

report also noted that even when a specific component option

is removed by civilian leaders (e.g., no ground troops), joint

planning can still benefit from a multiservice perspective, as

air operations planning did when they incorporated ground

expertise during the latter portions of the operation. In this

particular case, since a significant focus of the air war was on

enemy ground forces, the air planners would have benefited

from the earlier involvement of expert Army planners.

The

eport also noted that the JFC should str

ongly consider estab

lishing component commanders even when the component

has little or no immediate role in the operation.

Even with these dif

ficulties, the air campaign met its overall

objective of for

cing Slobodan Milo

sevi

c to cede to allied de

GOLDWATER-NICHOLS AS A MODEL FOR IR

mands.

The air campaign required the coordination of over

250 US Air Force and Navy aircraft combined with aircraft

from 11 other countries.

A single commander was placed in

charge of air operations and the geographic combatant com-

mander, from USEUCOM, was clearly in charge of the overall

operation. Critiques of the operation, some more pointed than

others, do not change the fact that the air war met its objec-

tives, was far more centrally controlled than similar operations

prior to Goldwater-Nichols, and effectively coordinated the ac-

tivities of hundr

eds of US and allied aircraft.

Analysis. Examination of military operations since Goldwater-

Nichols still shows ample evidence of service parochialism,

continued difficulties with interoperability, and sometimes con-

tentious command relationships. It is important to note that im-

proved jointness does not necessarily equate to the lack of inter-

service rivalry or disputes. To the contrary, competing service

viewpoints may actually result in plans that are more balanced

and joint. The effectiveness of Goldwater-Nichols in addressing

issues at lower levels (particularly tactical) is less apparent.

The legislation appears to have solidified the supremacy of the

combatant commander, but to have had less success in institu-

tionalizing lower

-level command relationships. That these

lower-level command structures take on the character of the

combatant commander may not be entir

ely bad, as they allow

that commander to tailor C

structur

es best suited to conduct

ing tasked military operations.

However

, the general public clearly believes that military ef

ficacy has improved. A comparison of public perceptions of the

military in 2004 with those prior to Goldwater-Nichols (1982)

shows a 25 percent increase in the number of people who

think of military officers as having “great prestige,” with only

4 per

cent having a negative view of military pr

estige.

One as

pect of this incr

eased r

espect manifests itself as a belief in

eater operational competence by the military. Goldwater

Nichols rightly deserves credit for some of the improved mili-

tary per

for

mance in the 1990s. While the act did not eliminate

the underlying tensions, it succeeded in r

ecalibrating them to

a degree and played an important role in the military’s success

in the conduct of major joint operations.

GOLDWATER-NICHOLS AS A MODEL FOR IR

The Strength of Joint Advice

One of the criticisms of the DOD prior to reform was that the

services, unable to overcome parochialism, took a lowest com-

mon denominator approach to policy and budget recommenda-

tions. In essence, the effort to maintain consensus meant that

none of the services would challenge the others on major service

programs. On the positive side, the Joint Requirements Over-

sight Council (JROC), an outgrowth of the Goldwater-Nichols

Act, provides a mechanism to reduce this parochialism and has

achieved somewhat favorable results. Likewise, the reform leg-

islation led to a much greater scrutiny of joint programs and a

much gr

eater emphasis on interoperability. On the negative

side, the civilian leadership must still make some of the most

difficult and contr

oversial budgetary decisions. For example, the

SecDef, despite significant r

esistance from senior Army leader-

ship, ultimately canceled the Army’s Crusader artillery program.

If the hope was that the CJCS could establish joint priorities that

each of the services would support, such an objective has not

been realized.

This might indicate that the JCS remains re-

luctant to r

ecommend cancellation of major

, multibillion-dollar

service programs, preferring to leave this unpleasant task in the

hands of civilian leadership.

Evaluating the quality of military advice to the president

and secretary of defense is even more difficult.

Part of the

difficulty is that the nature of the advice often depends upon

the personal leadership styles and preferences of those receiv-

ing it.

In one administration, the president and secretary of

defense may want multiple options, while other administra-

tions may adopt a style that looks for a single, coordinated

viewpoint from the JCS. In addition, a fundamental dynamic

is at play in the tension between of

fering professional advice

based upon years of military service and accepting the out

come when civilian leaders choose not to accept that advice.

Senior military leaders are thus in the uncomfortable position

of both trying to pr

ovide the best possible military advice and

subor

dinating themselves and their actions to civilian rule,

leaving them to decide how hard to press civilian leaders to ac-

cept their r

ecommendations.

GOLDWATER-NICHOLS AS A MODEL FOR IR

It is difficult to gauge whether military advice has improved

since Goldwater-Nichols.

Certainly there are several anecdotal

incidents of friction, including conflict between Chairman Powell

and SecDef Richard Cheney, disputes on military action in Ser-

bia, and more recent disputes between Army leadership and

the Office of the Secretary of Defense.

It would be almost im-

possible to ascertain whether Goldwater-Nichols exacerbated

these disagreements or whether such disputes are a natural

part of the national security decision-making process.

The Quality of Joint Personnel

Any evaluation of the overall quality of joint personnel de-

pends upon the measure of merit used. For example, do the

impr

oved promotion rates of joint personnel suggest that they

are better qualified than their predecessors before 1986, or is

some other measur

ement more relevant? The military promo-

tion system operates much like a meritocracy, suggesting that

indeed the best, most qualified personnel are promoted. If pro-

motion becomes the measur

e of quality, quantifiable infor

tion is readily available to assess the promotion rates of joint

personnel and thus their overall quality.

One of the major requir

ements of the legislation was that of

ficers who served on the Joint Staff and JSOs be pr

omoted at

a rate equal to or above the rate of members serving in service

headquarter staffs, while officers serving in other joint posi-

tions be promoted at the same rate as service members in the

same grade.

The promotion rates indicate that the rates for

joint officers indeed often far exceed both the service averages

and the service headquarters rates.

The exception is for offi-

cers serving in non-JSO/Joint Staff positions, whose promo-

tion rates routinely fall slightly below the overall service pro-

motion rates. In terms of length of tour (a metric that can

indicate the experience level on a staff), general officers were

serving an average of 28.8 months, while field-grade officers

served an average of 38.8 months.

Both of these time frames

exceed the mandated requirement of 24 months for general of-

ficers and 36 months for field-grade officers.

Congr

ess also mandated that the services cr

eate critical

joint duty positions (curr

ently 800 total) where joint experi-

GOLDWATER-NICHOLS AS A MODEL FOR IR

ence and training in joint matters are essential and fill these

positions with JSOs.

100

Yet currently only 47 percent of these

positions are filled by JSOs—far below the mandated 100 per-

cent.

101

In fact, this represents a downward trend since 1994,

with fewer and fewer critical positions being filled by JSOs.

102

There are multiple reasons for this gap. First, in many in-

stances a combatant command creates joint critical positions

that have special skill requirements. In those instances, the

combatant command may pr

efer an individual with the neces-

sary skill set instead of a JSO who lacks the r

equired skills.

Second, many individuals serving in JSO positions have met

joint duty assignment requirements but failed to complete re-

quired joint professional military education (JPME) training.

For instance, some individuals have served multiple joint

tours but for various reasons have never been able to attend

the JPME training and thus are not, technically, JSOs; how-

ever, such individuals are undoubtedly quite familiar with

joint procedures and operations. Third, another aspect of the

act—the requirement for general officers to have served in a

joint duty position—serves as a disincentive to fill a critical po-

sition with an officer who already has joint credit. Instead, the

inclination is to fill the position with an officer still in need of

joint cr

edit to allow that officer to be eligible for general of

ficer

rank.

103

Congr

ess’s intent to have critical positions filled by

the most highly qualified joint officers appears to remain un-

fulfilled; however

, if one assumes that longer tour lengths, on

average, equate to greater experience and quality, then this

aspect of the program appears to be working.

The legislation describing joint officer management policies is

both the most detailed and most controversial, as it had a direct

impact on the services’ pr

omotion policies and ef

fectively si-

phoned of

f the most talented service members to joint positions

for extended periods.

104

In fact, this ar

ea of the legislation saw

several modifications in subsequent years, including the al

lowance for waivers to moderate the impact on the services. How

ever, even given the controversy and necessary adjustments, the

legislation did in fact positively alter the composition of the Joint

Staf

f and other joint or

ganizations and, with the exception of fill

GOLDWATER-NICHOLS AS A MODEL FOR IR

ing of critical joint duty positions by JSOs, appears to have met

Congress’s intent to a large extent.

The Success of Goldwater-Nichols

By most accounts Goldwater-Nichols has successfully

achieved its fundamental objectives. Reviewing Congress’s

stated objectives that have some commonality with intelli-

gence reform provides a basis for analysis. The third and

fourth objectives focused on establishing “clear responsibility”

for the combatant commanders and ensuring they had the

proper authority to carry out their responsibilities. In this

case, the act has clearly been successful. In all major opera-

tions since the act, the combatant commander has been in

firm contr

ol, the services have supported the effort without

undue influence on the operations, and the operations have

been decidedly more joint and effective than those pre-1986.

This is not to suggest that interservice rivalry and interoper-

ability issues are fully resolved.

105

Rather, the balance of

power has shifted more to the combatant commanders and

away from the services, just as the act intended.

Improving the quality of the Joint Staff and personnel filling

other joint positions, the seventh objective, is a r

esounding

success story. Promotion rates are close to Congress’s original

intent, and the services’ disgruntlement with the mandates as

affecting the quality of officers available to support service re-

quirements illustrate that high-quality of

ficers are indeed fill

ing joint positions. The requirement for general officers to

serve in joint positions prior to promotion has also had the

positive effect of ensuring a broader perspective for each ser-

vice’s most senior military offices.

The act also seems to have attained its final objective of en

hancing the effectiveness of military operations. Certainly, other

factors contributed to operational successes in the 1990s and

the early years of the twenty-first century, such as technologi-

cal advances (e.g., precision weaponry and stealth); improved

command, control, and communications (C

) systems; and the

relatively poor quality of opposing forces. However, the success-

ful command and contr

ol of these large undertakings should

GOLDWATER-NICHOLS AS A MODEL FOR IR

not be minimized and suggests that jointness is a far more per-

vasive concept and operating characteristic than it was before

Goldwater-Nichols.

Taken in its totality, the Goldwater-Nichols Act successfully

adjusted various tensions, moving the DOD to a more central-

ized and generalist orientation. The CJCS and the combatant

commanders consolidated power, and the combined views of

the services began to prevail over service-specific approaches.

However, the legislation did not remove the underlying ten-

sions. Some would argue that reforms have not gone far enough,

while others argue that certain elements of the legislation have

pushed the balances too far in favor of jointness. Furthermore,

certain goals, such as improving the quality of joint personnel,

were much more amenable to legislative changes than such

others as the mor

e nebulous objective of improving the qual-

ity of advice. With those reservations understood, the success

of the act suggests that in some cases legislatively mandated

structural changes can have positive intended effects on large

bureaucracies; however, they probably do not address all the

issues with equal effectiveness. Understanding where struc-

tural change is appropriate then becomes a determining fac-

tor in overall legislative success.

Notes

1. Thomas Kean, Testimony to the House Permanent Select Committee on

Intelligence

, 108th Cong., 2d sess., 11 August 2004. “We consciously and

deliberately draw on the military example, the Goldwater-Nichols model. We

can and should learn, because that was a very successful reform in the mili-

tary two decades ago.” See also “Restructuring, From the Top . . . ,”

Wash-

ington Post

, 1 August 2004, B1, B4–B5, which details the views of seven for-

mer government of

ficials on the 9/11 r

ecommendations for intelligence.

John Deutch, former CIA director, cites Goldwater-Nichols as an “interest-

ing, though imperfect, model for centralizing the intelligence community,”

and Robert McFarlane, for

mer national security advisor, states that the

“military’s unified command structure . . . is a sound model for the new di

ector’s [of national intelligence] of

fice.”

2. US Congress, Senate,

Report of the Committee on Governmental Affairs,

United States Senate to Accompany S. 2840

, 108th Cong., 2d sess., 2004, S.

Report 359. The report examines the Goldwater-Nichols Act from the prob-

lems, to the act itself, to the ef

fects of the legislation and states, “The notion

GOLDWATER-NICHOLS AS A MODEL FOR IR

f ‘Goldwater-Nichols for the Intelligence Community’ has been a recurring

metaphor for intelligence reform since even before 9/11.”

3. 9/11 Report, 403 (see chap. 1, n. 3). The most direct correlations with

Goldwater-Nichols are recommendations to establish a “civilian-led unified

joint command for counterterrorism” and the need to centralize control and

authority under a national intelligence director, 411. See also comments by

Philip Zelikow, 9/11 Commission staff executive director, who stated, “The

national counter terrorism center is part of a Goldwater-Nichols concept for

the intelligence community. . . . This is a conception of that [unified com-

mands within the DOD] for the intelligence community, with joint mission

centers bringing people from different disciplines and agencies together

against common targets, supported by those home agencies like CIA and

DIA.” “Prospects for Change,”

NewsHour with Jim Lehrer, 23 July 2004,

http://www.pbs.org/newshour/bb/terrorism/july-dec04/commission_7-

23.html.

4. Gordon N. Lederman,

Reorganizing the Joint Chiefs of Staff: The

Goldwater-Nichols Act of 1986

(Westport, Conn.: Greenwood Press, 1999).

5. Ibid., 2.

6. Ibid., 2–4. Lederman argues that the military model is pluralistic, simi-

lar to other US democratic institutions, and that there are advocates on com-

peting ends of each tension helping to provide decision makers with the broad-

est range of perspectives.

7. 9/11 Report, 408 for structural barriers, 409 for dual-hatting.

8. Operation Eagle Claw was the name chosen for the entire rescue opera-

tion. Desert One was the rendezvous location for the first phase of the opera-

tion. Many discussions of the failed rescue use both names interchangeably.

For a viewpoint on change from within the hierarchy as expressed by the sit-

ting chairman of the Joint Chiefs of Staff, see David C. Jones, “Why the Joint

Chiefs of Staff Must Change,”

Armed Forces Journal International 199, no. 7

(March 1982): 69–71.

9. James R. Locher,

Victory on the Potomac: The Goldwater-Nichols Act Uni-

fies the Pentagon

(College Station, Tex.: Texas A&M University Press, 2002).

Locher, as the primary Senate staffer working the effort, had a particularly in-

timate view of events and provides a comprehensive review of the events lead-

ing up to and surrounding the passage of the Goldwater-Nichols Act. Leder-

man,

Reorganizing the Joint Chiefs of Staff, provides a scholarly review,

including the inherent tensions in the American military organization, an

analysis of the reorganization process, and a review of the military after the

act’s passage. An evaluation of the legislation’s impact can be found in Paul

M. Besson,

The Goldwater-Nichols Act: A T

en-Year Report Car

, P-98-2 (Cam-

bridge, Mass.: Harvar

d University Program on Information Resources Policy,

June 1998), http://www.pirp.harvard.edu/publications/pdf-blurb.asp?id=380

(accessed 18 October 2004).

10. Joint Publication 1-02,

DOD Dictionary of Military and Associated

Terms

, http://www.dtic.mil/doctrine/jel/doddict/data/j/index.html (accessed

18 October 2004).

GOLDWATER-NICHOLS AS A MODEL FOR IR

1. Ibid. The full definition of

nified action

s “A broad generic term that

describes the wide scope of actions (including the synchronization of activi-

ties with governmental and nongovernmental agencies) taking place within

unified commands, subordinate unified commands, or joint task forces

under the overall direction of the commanders of those commands.”

12. Department of Defense,

Rescue Mission Report (Washington, D.C.:

GPO, 23 August 1980), 4.

13. Ibid., 16.

14. John E. Valliere, “Disaster at Desert One: Catalyst for Change,”

Pa-

rameters

22, no. 3 (Autumn 1992): 77.

15. See US House, Committee on Armed Services,

Adequacy of U.S. Ma-

rine Corps Security in Beirut: Summary Findings and Conclusion of the Inves-

tigations Subcommittee

, 98th Cong., 1st sess., committee print no. IIA, 19

December 1983, 1. The first line of the summary document states, “There

were inadequate security measures to protect the Marine Unit from the full

spectrum of threats.”

16. Joint Publication 3-0,

Doctrine for Joint Operations, 10 September

2001, A-2.

17. US House, Committee on Armed Services,

Adequacy of U.S. Marine

Corps Security in Beirut: Report Together with Additional and Dissenting

Views of the Investigations Subcommittee

, 98th Cong., 1st sess., committee

print no. II, 19 December 1983, 43–54.

18. Ibid., 47.

19. Richard A. Gabriel,

Military Incompetence: Why the American Military

Doesn’t Win

(New York: Farrar, Straus, Giroux, 1985), 138.

20. See also, John H. Cushman,

Command and Control of Theater Forces:

Adequacy

(Washington D.C.: AFCEA International Press, 1985), 78. After re-

tiring, General Cushman wrote further on the subject of command and con-

trol of forces and reiterated the problems with limited authority. For ex-

ample, in describing the system circa 1983 he stated, “most commands are

extraordinarily complex arrangements, their charters laced with intricate

definitions of authority. And [Unified Actions Armed Forces] UNAFF . . . sub-

stantially limits the authorities of the unified commander, and is careful to

prescribe in detail the authorities retained by his Service subordinates in

peace and war.” See also Locher,

Victory on the Potomac, 157.

21. Locher,

Victory on the Potomac, 209.

22. Archie D. Barrett,

Reappraising Defense Organization: An Analysis on

the Defense Organization Study of 1977–1980

(Washington, D.C.: National

Defense University Press, 1983), 53.

23. Locher

ictory on the Potomac

, 135.

24.

Ibid., 309.

25. Besson,

The Goldwater-Nichols Act, 19.

26.

Ronald H. Cole,

Operation Ur

gent Fury: The Planning and Execution of

Joint Operations in Grenada

(Washington, D.C.: Joint History Office of the

Chairman of the Joint Chiefs of Staff, 1997), 67.

27. Locher,

Victory on the Potomac, 156.

GOLDWATER-NICHOLS AS A MODEL FOR IR

8. Joint Publication 0-2,

nified Action Armed Forces,

0 July 2001), I-5.

29. Quoted in Locher,

Victory on the Potomac, 133.

30. Jones, “Why the Joint Chiefs of Staff Must Change,” 65, 72.

31. Ibid., 62.

32. Edward C. Meyer, “The JCS—How Much Reform Is Needed?”

Armed

Forces Journal International

199, no. 8 (April 1982): 89.

33. Jones, “Why the Joint Chiefs of Staff Must Change,” 66.

34. Ibid., 66, 72.

35. Lederman,

Reorganizing the Joint Chiefs of Staff, 36.

36. Meyer, “The JCS,” 85.

37. Barrett,

Reappraising Defense Organization, 36.

38. Jones, “Why the Joint Chiefs of Staff Must Change,” 68.

39. Meyer, “The JCS,” 85.

40. Lederman,

Reorganizing the Joint Chiefs of Staff, 34.

41. Ibid., 35.

42. Ibid.

43. Jones, “Why the Joint Chiefs of Staff Must Change,” 68.

44. Lederman,

Reorganizing the Joint Chiefs of Staff, 42.

45. Ibid., 42.

46. Barrett,

Reappraising Defense Organization, 55.

47. Lederman,

Reorganizing the Joint Chiefs of Staff, 43.

48. Locher,

Victory on the Potomac, 209.

49. Lederman,

Reorganizing the Joint Chiefs of Staff, 43.

50.

Department of Defense Reorganization Act of 1986, Public Law 99-

433, 99th Cong., 2d sess. (1986), sec 3.

51. Lederman,

Reorganizing the Joint Chiefs of Staff, 77.

52. Ibid., 77–78. Note that 400 and 1,627 refer to the numbers of Joint

Staff personnel authorized by Congress. In actuality, additional personnel

supported the Joint Staff pre–Goldwater-Nichols although they were techni-

cally filling positions in other organizations.

53. Lederman,

Reorganizing the Joint Chiefs of Staff, 79.

54. Ibid., 80.

55.

Department of Defense Reorganization Act, sec. 164(c)(3).

56. Ibid., sec. 164(d)(1).

57. Ibid., sec. 164(e)(1)(A) and sec. 164 (e)(3)(A).

58.

JSO is “a term for officers who are particularly trained in and oriented

towards joint policy and procedures. In order to be eligible to become a JSO,

an of

ficer must have completed both Phase I and Phase II of the JPME [joint

professional military education] program and then complete a full Joint Duty

Assignment (JDA), usually in that order.”

Joint Officer Management,

http://www.afpc.randolph.af.mil/jom/Pages/FAQ.html (accessed 9 April

2005).

59.

Department of Defense Reorganization Act, sec. 661(d).

60.

Leder

man,

Reor

ganizing the Joint Chiefs of Staf

, 82.

61.

Ibid., 83.

GOLDWATER-NICHOLS AS A MODEL FOR IR

2. For a decidedly more negative assessment of the success of Goldwater-

Nichols, see Frank Hoffman, “Goldwater-Nichols after a Decade,” in Murray

Williamson, ed.,

The Emerging Strategic Environment (New York: Praeger,

1999), 156–82.

63. Peter D. Feaver and Richard H. Kohn, “The Gap,”

National Interest,

no. 61 (Fall 2000): 34–5. The article cites the growing politicization of the

military as potentially contributing to less trust in uniformed advice. The ar-

ticle also cites a potential trend in how advice is provided, “Contrary to the

traditional understanding of civilian control, a majority of elite military offi-

cers today believes that it is proper for the military to insist rather than

merely to advise.” See also, Peter D. Feaver and Richard H. Kohn, eds.,

Sol-

diers and Civilians: The Civil-Military Gap and American National Security

(Cambridge: MIT Press, 2001); and Eliot A. Cohen, Supreme Command (New

York: Anchor Books, 2002).

64. See Hoffman, “Goldwater-Nichols after a Decade,” for a discussion of

Operation Just Cause, the military action to remove Manuel Noriega from

Panama, and operations in Somalia. Hoffman suggests that operations in

both Panama and Somalia were little better than those prior to Goldwater-

Nichols passage.

65. See Michael R. Gordon and Bernard E. Trainor, The Generals’ War:

The Inside Story of the Conflict in the Gulf

(Boston: Little Brown and Com-

pany, 1995). The general’s power is described in the epilogue to the book,

noting that even the service chiefs could not visit without his permission,

471. See also Les Aspin and Bill Dickinson,

Defense for a New Era: Lessons

of the Persian Gulf War

(McLean, Va.: Brassey’s, 1992), 43.

66. Katherine Boo, “How Congress Won the War in the Gulf,”

Washington

Monthly

23, no. 10 (October 1991). As one defense official described it, “This

is the first time since 1945 that a military commander had the power to re-

sist the Marines’ desire for an amphibious assault.” See also Aspin and

Dickinson,

Defense for a New Era, 43.

67. Thomas A. Keaney and Eliot A. Cohen,

Gulf War Air Power Survey

Summary Report

(hereafter referred to as GWAPS) (Washington, D.C.: GPO,

1993), 38. The original Instant Thunder plan remained largely intact for the

first phase of the operation; however, the following phases broadened target-

ing efforts. Phase II focused on gaining air superiority over Kuwait, Phase III

focused on attrition of Iraqi ground forces, and Phase IV was support to a

ground attack into Kuwait.

68. US General Accounting Office,

Operation Desert Storm: Evaluation of

the Air Campaign

, NSIAD-97-134 (Washington, D.C.: GAO, 1997), 14, uses

the “over 1600” figure while the

GWAPS, 7, reports “nearly 1800.”

69.

For example, in Vietnam air operations had as many as four separate

air chains of command. In addition, the country was broken into six sepa-

rate r

outes with various services and command or

ganizations assigned to

specific routes. No single joint commander planned and executed the air ef-

fort. See Aspin and Dickinson,

Defense for a New Era, 8, and Lederman, Re-

organizing the Joint Chiefs of Staff

, 103.

GOLDWATER-NICHOLS AS A MODEL FOR IR

0. Department of Defense,

onduct of the Persian Gulf War: Final Report

to Congress

(Washington, D.C.: Department of Defense, 1991), 119–20.

71. Aspin and Dickinson,

Defense for a New Era, 10.

72. Initial planning associated with Instant Thunder was accomplished

by the USAF Checkmate office in the Pentagon. After General Powell, CJCS,

was briefed on the plan, he directed the effort be expanded to include rep-

resentatives from all the services. Gordon and Trainor,

The Generals’ War,

84. Later, the JFACC assumed control of the planning under the auspices of

the Special Planning Group or “Black Hole,” which included representatives

from all services; however, the Air Force still predominated,

GWAPS, 38.

73. Gordon and Trainor,

The Generals’ War, 76, 77. This seeming viola-

tion of intent is somewhat mitigated, however, because the request for Air

Force assistance came from the CENTCOM commander rather than having

the plan thrust on the unified commander by the service. In addition, the

initial plan only served as a template, with the remaining planning being ac-

complished by the JFACC’s designated planning cell.

74. Aspin and Dickinson,

Defense for a New Era, 9.

75. Lederman,

Reorganizing the Joint Chiefs of Staff, 100. Initial planning

of the ground campaign was conducted by a team of four Army officers re-

ferred to as the

Jedi Knights.

76. Desert Storm marked the first use of the GPS satellites. See Aspin

and Dickinson,

Defense for a New Era, 21.

77. Ibid., 24.

78. Ibid., 23.

79. Peter C. Reddy, “Joint Interoperability: Fog or Lens for Joint Vision

2010?” (Maxwell AFB, Ala.: Air Command and Staff College, March 1997), 13.

80. The concept of

combined operations is differentiated from joint opera-

tions to mean operations involving “forces or agencies of two or more allies,”

Joint Publication 1-02,

DOD Dictionary. Joint operations involve operations

from two or more military departments. See also Paul E. Gallis, “Kosovo:

Lessons Learned from Allied Force,” CRS-4 (Washington, D.C.: Congres-

sional Research Service (CRS), 19 November 1999). Gallis states, “From the

outset, Operation Allied Force was a unique type of operation, and not

guided by normal military doctrine. There appears to have been no confu-

sion on military objectives among NATO military forces. The significant dif-

ferences in perspective were rather on the national political level, where

short-term military objectives were affected by political adjustments in tar-

get lists, str

ong aversion to casualties on either side of the conflict, and con-

cerns over domestic public reactions.”

81. Benjamin S. Lambeth, “Lessons from the War in Kosovo,”

Joint

Forces Quarterly

, Spring 2002, 13, 15.

82. Ibid., 16. In this case, SACEUR, Army general Wesley Clark, argued

for more attacks against fielded forces while the combined force air compo-

nent commander

, Air For

ce lieutenant general Michael Short, ar

gued for

mor

e strategic and infrastructure attacks such as those directed against

GOLDWATER-NICHOLS AS A MODEL FOR IR

Miloˇsevi´c, his C

apparatus, military and police headquarters, and electric

ower grids.

83. Benjamin S. Lambeth,

Nato’s Air War for Kosovo: A Strategic and

Operational Assessment

(Santa Monica, Calif.: RAND Corporation, 1999),

207, http://www.rand.org/publications/MR/MR1365/ (accessed 19 Oc-

tober 2004).

84. Ibid., 210.

85. Bruce R. Nardulli et al.,

Disjointed War: Military Operations in Kosovo

(Santa Monica, Calif.: RAND Corporation, 1999), 7, http://www.rand.org/pub-

lications/MR/ MR1406/ (accessed 19 October 2004).

86. Ibid., 113.

87. The JFC in this case, Admiral Ellis, supported this view, stating, “The

lack of a Land Component Commander was doctrinally flawed and opera-

tionally dangerous. . . . The JFC was left without valuable expertise on the

land component aspect (e.g., training, qualifications, contingency, opera-

tions, logistics, force protection, etc.). It also increased confusion by compli-

cating planning and impeding an efficient operational chain of command.”

See Nardulli et al.,

Disjointed War, 114.

88. Whether air operations were solely responsible is open to debate.

See, for example, Gallis, “Kosovo: Lessons Learned from Allied Force.” Gal-

lis points to the ambiguity, stating, “Why Miloˇsevi´c decided to accept NATO

terms and withdraw his forces remains unclear. Damage caused by NATO

bombing, sustained allied unity, possible allied planning for a ground war,

a desire to preserve his forces, and desertion of Russia as a possible protec-

tor were likely principal reasons.”

89. Lambeth,

Nato’s Air War for Kosovo, 20.

90. Harris Poll 51, 6 September 2000, http://www.harrisinteractive.com/

harris_poll/index.asp?PID=494 (accessed 25 October 2004), table 1. In 1982

only 22 percent viewed the profession of military officers as very prestigious

versus 47 percent in 2004. According to the poll, “Prestige is strongly asso-

ciated with respect, public service and good work.”

91. The decision to cut the Crusader system was highly contentious,

with the Army lobbying Congress even after the DOD leadership had

reached a decision to cancel the program. See James Dao, “Rumsfeld Sets

up Showdown over Weapon,”

New York Times, 9 May 2002, A-28. “Though

Army officials waged a lobbying campaign on Capitol Hill last week to pro-

mote the Crusader after wor

d of Mr. Rumsfeld’s plans first surfaced, the

Army secretary, Thomas E. White, said today that he would work hard to

support the decision.” The amount of JCS involvement in the final decision

and their overall recommendation is unclear; however, Secretary Rumsfeld

alludes that senior military leaders were actively involved in the decision.

See Donald Rumsfeld, “A Choice to Transform the Military,”

Washington

Post

, 16 May 2002, A25. “The decision to recommend termination of the

Crusader pr

ogram was r

eached after many months of car

eful r

eview, wide-

ranging discussion and in-depth planning and analysis. This was a review

not just of the Crusader program but also of future capabilities, of the strat-

GOLDWATER-NICHOLS AS A MODEL FOR IR

gy to guide U.S. and of a framework for assessing and balancing risks. The

senior uniformed and civilian leaders of the Defense Department spent

countless hours discussing these matters in a process that started well be-

fore Sept. 11.”

92. The National Command Authorities consist of the secretary of de-

fense and the president.

93. See, for example, Cohen,

Supreme Command, 214–15. Within the larger

text, Cohen explores how four different leaders—Lincoln, Churchill,

Clemenceau, and Ben-Gurion—approached the task of leading major war ef-

forts. Cohen goes on to offer a somewhat variant take on the “normal” theory

of civil-military relations and describes how each of the leaders interacted with

and received advice from senior military officials. Chapter 7, “The Unequal

Dialogue,” provides an excellent summary of his analysis. See also Hoffman,

“Goldwater-Nichols after a Decade,” 163–66. The author suggests that military

advice since Goldwater-Nichols has been poor, aiming most of his criticism at

then-CJCS Colin Powell. Hoffman suggests that advice from a single source

(Powell) is not always superior to that originating from multiple sources and

viewpoints; however, Hoffman does not explain why the other service chiefs did

not express dissenting views, as is their right under Goldwater-Nichols, if they

believed that Powell’s military advice was clearly wrong.

94. Peter J. Roman and David W. Tarr, “The Joint Chiefs of Staff: From

Service Parochialism to Jointness,”

Political Science Quarterly, 113, no. 1

(Spring 1998): 109. The authors suggest that civilian decision makers believe

“that the military now provides higher quality and more timely advice as a re-

sult of the Goldwater-Nichols reforms.” Unfortunately, the authors do not

identify which civilian leaders they interviewed and whether these civilian

leaders had received military advice prior to Goldwater-Nichols passage to pro-

vide a basis for comparison.

95. For a brief description of the Cheney/Powell conflict see Cohen,

Supreme Command, 190. “As Powell himself indicates in his memoirs, it took

a sharp rebuke by Secretary of Defense Dick Cheney to remind him that he

was supposed to execute policy, not make it.” See also Steven Lee Myers and

Eric Schmitt, “War’s Contract Creates Tension among Chiefs,”

New York

Times,

30 May 1999, P1. In Serbia, the larger dispute was between the com-

batant commander, Gen Wesley Clark, and the Clinton administration. In

addition, see Steven Lee Myers, “U.S. Military Chiefs Firm: No Ground Force

for Kosovo,”

New York Times, 3 June 1999, A16. Myers states, “From the

start, the chiefs have expressed reservations about NATO’s strategy and ever

deeper misgivings about the prospects of an invasion.” Other accounts can

be found in Bradley Graham “White-Rumsfeld Dispute, Round 2; Ex-Army

Secr

etary Fires Back on Iraq,”

ashington Post

, 19 September 2003, Friday,

Final Edition, A23; and Jonathan Weisman, “Rumsfeld and Myers Defend

ar Plan,”

ashington Post

, 2 April 2003, Final Edition, A19. “Flar

e-ups

include [SecDef] Rumsfeld’s selection of a successor to Army Chief of Staff

Gen. Eric K. Shinseki more than a year before Shinseki’s planned retirement,

the secretary’s abrupt cancellation of the Army’s Crusader self-propelled ar-

GOLDWATER-NICHOLS AS A MODEL FOR IR

illery gun and, weeks ago, Deputy Defense Secretary Paul D. Wolfowitz’s

public dispute with Shinseki over the size of a force needed to occupy Iraq

once hostilities cease.”

96.

Department of Defense Reorganization Act, sec. 662(a)(1)-(3).

97. Department of Defense, “2003 Secretary of Defense Annual Report to

the President and the Congress” (hereafter cited as ADR), 194–97,

http://www.defenselink.mil/execsec/adr2003/adr2003_toc.html (accessed

25 October 2004).

98. ADR, 189.

99.

Department of Defense Reorganization Act, sec. 663, (a)(1)-(2).

100. Ibid., sec. 663 (d)(2)(A)-(B).

101. ADR, 190. Also, James Locher, interviewed by the author, 2 February

2005. In contrast to the possible reasons for the gap, Locher believes the

services continue to leave too many JSO positions unfilled and have not fully

supported the full intent of this aspect of Goldwater-Nichols.

102. Dr. Paul Herbert and Mr. Stuart Wilson, associates, Booz Allen

Hamilton,

Testimony before the Subcommittee on Total Force, House Armed

Services Committee, United States House of Representatives

, 19 March 2003,

http://www.house.gov/hasc/openingstatementsandpressreleases/108thco

ngress/03-03-19herbert.html#_ftnref20 (accessed 31 March 2005). The

DOD contracted Booz Allen Hamilton to perform an independent study of

“Joint Officer Management and Joint Professional Military Education” in

September 2002. The testimony by Herbert and Wilson highlighted the re-

sults of the study.

103. Ibid.

104. Lederman,

Reorganizing the Joint Chiefs of Staff, 107–8.

105. In fact, there are legitimate reasons for not wanting to completely

rid the DOD of such rivalry, particularly when it can spur innovation. For a

brief description of this and other advantages of interservice competition,

see Harvey M. Sapolsky, “Interservice Competition: The Solution, Not the

Problem,”

Joint Forces Quarterly, Spring 1997, 50–53.

GOLDWATER-NICHOLS AS A MODEL FOR IR

Chapter 3

Reforming Intelligence:

A 50-Year Effort

Within our generation and for our protection, organized In-

telligence has been forced upon us by the rapidly shrinking

world of electronics, nuclear weapons and planes which

travel at supersonic speed.

—Task Force on the Organization of

Intelligence Activities, 1955

While the context of the thr

eats described in the epigraph

above has changed, the need for organized intelligence remains.

In many respects, the emergence of multiple transnational ad

versaries, coupled with continuing threats from nation-states

and the resultant intelligence needs of policy makers, has

placed even greater demands on the US intelligence system.

As the international environment has changed over the last

50-plus years, the IC has also evolved. The extent of this evo-

lution and its suitability to meet the current and projected

challenges ar

e less clear. However, at its core the IC remains

an interagency construct gover

ned mor

e by consensus than

by administrative fiat, and this r

eality has significant implica-

tions for how the IC is dir

ected and how it acts.

An ongoing

debate has ensued that has offered many opportunities to ex-

amine the IC, assess its suitability, and recommend alter-

ations in its structure and focus. These assessments, often

performed by congressional or presidential commissions or

studies, ar

e the subject of this chapter

The IC has certainly not escaped outside scrutiny. Initia

tives to improve or restructure the IC or individual agencies

have existed almost as long as the IC itself. A series of Con

gressional Research Service (CRS) reports documents several

major efforts to examine the IC; another account indicates

that 16 federal studies and commissions have addressed the

subject of intelligence since 1991 alone.

The r

eal question is

whether this scrutiny resulted in fundamental change since

1947 or whether change has been marginal at best.

The IC has changed considerably from its rather humble ori-

gins in 1947. To be sure, the basic structural framework re-

mains the same; however, the IC has grown dramatically in

terms of the absolute number of organizations and personnel

and the resources applied to the intelligence effort. Much of

this change occurred without legislation; in fact, Congress en-

acted little significant intelligence reform legislation prior to

2004. The executive branch has been far more active, estab-

lishing new intelligence agencies, dramatically increasing the

size of the IC in ter

ms of resources, and issuing multiple execu-

tive orders. Still, despite the relative lack of substantive legis-

lation since 1947, the threat of congressional action has, in

some cases, impelled the executive branch to make desired

changes through internal restructuring, management initiatives,

or the issuance of new executive orders.

Most reform efforts have focused on one of two distinct

eas: the operations of the CIA and/or the management of the

IC. Those efforts focused on the IC have examined a variety of

community management issues, thr

ee of which are salient to

the most recent reform efforts. The first concerns the repeated

recommendations to further centralize the IC to impr

ove coor-

dination and reduce duplication among the multiple organiza-

tions within the IC. Tightly coupled to this is a second area of

commonality—the need to incr

ease the director of central intel-

ligence’s (DCI) control over the IC. This issue has both a dual-

hatting aspect, since the DCI also serves as the director of CIA,

and an authority aspect. Thus there were repeated calls to in-

crease the DCI’s authority to better manage the IC. The third

topic is analysis, including methods to improve analytic prod-

ucts for the policy maker, the proper relationship between the

policy maker and analytic community, and the role of analysis

organizations within the separate departments.

The following discussion centers on those elements of past

eforms that have examined the IC as a whole rather than the

conduct or operations of individual agencies.

Appendix B

summarizes these r

eform efforts, while appendix C highlights

several of the milestones in the IC’s evolution.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

IC Origins: The National

Security Act of 1947

The modern-day IC and the DOD share a common legisla-

tive ancestor: the National Security Act of 1947. As the legisla-

tion that lays the foundation of the modern IC, the act is no-

table more for what it does not specify than for what it does.

In a little more than one-and-a-half pages, the act established

the CIA and defined its chief functions. The CIA effectively

subsumed the functions of what at the time was the Central

Intelligence Group, and it was placed under the newly formed

National Security Council (NSC).

The primary duties of the

CIA were to advise the NSC on intelligence matters and to “cor-

relate and evaluate intelligence relating to national security”

from the intelligence departments.

The act granted the CIA

authority to receive all national-security-related intelligence

om the departments.

Despite assigning the central role in correlating national se-

curity intelligence to the CIA, and more specifically to the DCI,

the act confir

med the authority of the other departments to

continue to “collect, evaluate, correlate, and disseminate de-

partmental intelligence.”

Thus the legislation paid obvious

deference to the need for each of the departments with intelli-

gence organizations to maintain its organic capabilities and

use them for specific departmental requirements.

Early Studies of the IC

As would be expected after a major legislative change to or-

ganizations, the transition to the new IC structure involved

controversy and uncertainty as to the roles and missions of

various organizations. The newly formed CIA in particular had

to struggle with its r

oles.

One particularly difficult issue was how to obtain the neces-

sary cooperation fr

om the Departments of Defense and State

without any formal line or budgetary authority over the intel-

ligence organizations within those departments. The role of

the CIA in operational activities was also unclear. This ambi-

guity led to a study of the CIA soon after its formation.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Intelligence Survey Group (Dulles-Jackson-Correa

Report)

In 1948 the fledgling NSC commissioned a study of the CIA.

Much of the study, known as the Dulles-Jackson-Correa report,

focused on the CIA’s effectiveness in carrying out the tasks spec-

ified in the National Security Act—specifically, the coordination

of intelligence activities and the production of national intelli-

gence.

The report found the “CIA was not coordinating intelli-

gence activities in the Government; the correlation and evalua-

tion functions were not well organized, and other members of the

fledgling IC were not fully included in the estimates process.”

The report faulted the CIA and its leadership on the quality of its

intelligence estimates, citing the CIA’s “failure to take charge of

the production of coordinated national estimates.”

To address

problems with intelligence estimates, the report suggested mak-

ing stronger use of the existing Intelligence Advisory Committee,

creating an Estimates Division within the CIA, and ensuring full

involvement of participating intelligence agencies.

The report

found that “jurisdictional conflicts and duplication which the Na

tional Security Act was intended to eliminate” still r

emained.

Less clear was how the CIA was to coordinate more effec-

tively the activities of the IC. Possibly presaging the need for a

version of today’s Community Management Staff, the report

recommended the reconstitution and str

engthening of a previ-

ous staff function, the Interdepartmental Coordinating and

Planning Staf

f. However

, it also found that such coordination

could best be achieved through “mutual agreement” and the

“right measure of leadership.”

The Dulles-Jackson-Correa

report thus began a 50-plus-year debate as to whether effec-

tive coordination required greater DCI authority or simply the

proper execution of authority already provided. The report

clearly pr

esumed the latter would suf

fice.

The First Hoover Commission

During the same year, the overall efficacy of the entire execu-

tive branch of government was under review by the Commis-

sion on Organization of the Executive Branch of the Govern-

ment, which came to be known as the Hoover Commission.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

The commission completed its exceptionally broad assessment

of the entire government in 1949. While the review covered the

CIA, the analysis of the agency was less substantive in com-

parison to other aspects of the commission’s review. The

Hoover Commission was composed of various task forces as-

signed to examine various aspects of the government.

Ferdinand Eberstadt (who had performed an earlier review of

the Central Intelligence Group) led the task force on National

Security Organization that reviewed the intelligence commu-

nity as part of its examination of the entire national security

system as then configured. Although broad, the study included

several notable findings. The task force’s report, perhaps pre-

saging a central theme in intelligence reviews, found “disturb-

ing inadequacies in our intelligence system,” referring to a mis-

taken intelligence estimate in 1948.

The report recommended

that a board focused exclusively on intelligence evaluation be

created within the CIA and that the CIA establish better rela-

tions with the other intelligence organizations upon which it

depended for actual intelligence.

The report also made the

first case for greater centralization of intelligence, stating that,

“T

oo many disparate intelligence estimates have been made by

the individual departmental intelligence services; that these

separate estimates have often been subjective and biased, . . .

and that a more comprehensive collection system, better coor-

dination and more mature and experienced evaluation are im-

perative.”

Finally, the task force shed light on the root of a

problem that would become a common theme in future reviews

and reform efforts: the lack of support from other intelligence

agencies was negatively affecting the CIA’s ability to provide

sound, comprehensive intelligence estimates. Although Con-

gress took no action on the r

eports of either the Intelligence

Survey Group or the First Hoover Commission, the incoming

DCI, Gen Walter Bedell Smith, created a Board of National Es-

timates to r

eview and pr

oduce national intelligence estimates

and thus improve the estimative process.

The Second Hoover Commission

In 1955 the T

ask For

ce on Intelligence Activities, working as

part of the Second Hoover Commission, r

eleased its report on

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

the IC.

As with the First Hoover Commission, this assess-

ment was only part of a much larger effort. The task force was

initially instructed to examine only the CIA; however, the task

evolved to studying all intelligence operations. Even at this rel-

atively early time in the IC’s evolution, the task force found 12

separate agencies involved directly in intelligence, with 10 ad-

ditional agencies involved indirectly.

The task force was gen-

erally comfortable with the intelligence effort, finding it was

“pursued in a diligent and dedicated manner,” although there

were some instances of inefficiencies.

Of the nine recommendations, only one dealt specifically

with improving the function of the IC. The commission sug-

gested creating the position of executive director of the CIA to

relieve the DCI of the basic administrative tasks related to the

CIA and allow the DCI to dedicate more time to coordinating

IC efforts.

This recommendation was one of the first to high

light the demands placed on the DCI by his two roles; however,

no action was taken. In fact, the DCI at the time, Allen Dulles,

did the opposite and created a deputy for community affairs,

thereby allowing himself more time to focus on the CIA.

This

action seems especially ironic given the emphasis he had

placed on coordination and national estimates when he was

part of the Intelligence Survey Group in 1949.

Executive Branch Activity 1955–63

In response to the Second Hoover Commission’s recommen-

dation to establish an oversight mechanism, Pres. Dwight D.

Eisenhower issued Executive Order 10656 in 1956, creating

the President’s Board of Consultants on Foreign Intelligence

Activities (PBCFIA).

This board reiterated the need for the

DCI to focus more on community affairs and again recom-

mended establishing an executive director of the CIA to relieve

the DCI of “many of the management r

esponsibilities he is cur

rently called upon to discharge as head of the CIA.”

Despite

these continuing calls for an executive director and pressure

om Pr

esident Eisenhower

, DCI Dulles r

efused to cr

eate the

position. His resistance illustrates one of the problems of re-

form: the roles that personality and individual preferences of

senior leaders play in implementing suggested, or in some

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

cases mandated, reforms. In this case, DCI Dulles success-

fully resisted the reforms and continued to focus his efforts on

his preferred role as head of the CIA at the expense of the IC.

During the early 1960s, the concept of separating the two

roles of the DCI seems to have been considered by the PBCFIA

and later by the renamed President’s Foreign Intelligence Ad-

visory Board (PFIAB).

Their recommendations were never

acted upon, and it remains unclear if they simply refined ear-

lier recommendations or responded to specific events such as

the Bay of Pigs failure. What is clear is that Pres. John F.

Kennedy believed the role of leading the IC, not the manage-

ment of the CIA, was the cor

e task of the DCI.

Again, person

ality and management preferences, in this case the presi-

dent’s, played a role. President Kennedy pr

essed for greater

community coordination, and the DCI at the time, John McCone,

attempted to comply. However, he apparently failed to secure

a stronger r

ole for the DCI because of bureaucratic maneuver-

ing by or

ganizations within the IC.

No significant studies or reform proposals emerged in the

remainder of the 1960s. However, the following decade pre-

sented a significant contrast to this relatively quiet period.

The 1970s

The 1970s saw intensive studies and reforms of the IC carried

out by both the executive and legislative branches. Executive

branch scrutiny of the IC began with the Office of Management

and Budget’s (OMB) significant evaluation of the IC. Congress

also became more involved, first through a congressionally

mandated study, but later through congressional investigations

into charges of impropriety and questionable activities on the

parts of various or

ganizations within the IC. The executive

branch launched a separate study, the Rockefeller Commission,

to provide a separate investigation of the charges.

Taken together, the resulting reports offered the most in-

depth examination of the IC performed up to that time. Many

of their recommendations would be carried forward in future

efor

m efforts.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

The Schlesinger Report

In 1971 Pres. Richard M. Nixon requested that the OMB ex-

amine the IC’s organization and recommend improvements.

The report submitted by James Schlesinger, the OMB director,

offered one of the broadest and most insightful evaluations of

the IC and its efforts at that time. In language reminiscent of

future reform efforts the report found:

While a number of specific measures may help to bring about a closer

relationship between cost and effectiveness, the main hope for doing so

lies in a fundamental reform of the intelligence community’s decision-

making bodies and procedures. This conclusion is advanced in the full

recognition that reorganization will, at best, only create conditions in

which wise and imaginative leadership can flourish. In the absence of

reorganization, however, the habits of [the] intelligence community will

remain as difficult to control as was the performance of the Depart-

ment of Defense prior to the Defense Reorganization Act of 1958.

The report identified “unproductively duplicative” development

of collection capabilities and poor interagency comparisons of

capabilities, “largely unplanned and unguided” expansion of the

community with respect to allocation of resources, the growing

cost of collection capabilities at the expense of analytical capa-

bilities, and limitations on the DCI’s per

formance of his duties

because of his multiple roles.

It found problems similar to

those facing combatant commanders prior to Goldwater-

Nichols: a “lack of institutions governing the community with

the authority and responsibility to resolve issues without exces-

sive compromise.”

The report laid out several variables af

fect-

ing effectiveness and efficiency of the IC that future reform ef-

forts would revisit. These variables included the DCI’s control

over resources and over the intelligence activities of the military

services, the size and functions of the DCI’s staff, “the number

and location of national analytical and estimating centers,” and

the need for independent review mechanisms.

The Schlesinger Report then provided three options to rec-

tify the perceived problems. The first option examined estab-

lishing a dir

ector of national intelligence (DNI) and placing the

majority of collection assets under the DNI’s dir

ect contr

ol,

with the DOD only maintaining contr

ol over tactical assets.

The r

eport identified the advantages of this appr

oach as clari

fying r

esponsibility and per

mitting more effective budgetary

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

and programming decisions. The disadvantages included the

potential to further overload the DCI with responsibilities, re-

sistance from the DOD, and (perhaps most important) the like-

lihood of the DOD’s creating its own indigenous capabilities

that it could directly control.

The second option examined in-

creasing the power of the DCI and dividing the CIA, with one

part providing the staff and analytic support needed by the

DCI and the other becoming a separate agency under a sepa-

rate director. This option would have removed from the DCI

the day-to-day responsibility of managing the CIA, allowing

him time to manage community affairs. It would have pre-

vented the conflict of inter

est that resulted when the DCI was

both an advocate for CIA programs and an arbiter of community-

wide programs.

The primary disadvantages of the approach

were the DCI’s lack of direct authority over much of the IC and

decreased control over the CIA.

The third, and clearly least

favored, option was to place a coordinator of national intelli-

gence in the White House to oversee the IC and ensure the

community met the needs of national policy makers.

The final portion of the report dealt with mechanisms to im-

ove analysis and estimates—a topic still relevant today. Rec

ommendations included greater dialogue with consumers of

information, upgrades to analytic centers to impr

ove “the com-

petition of ideas,” “periodic reviews by outsiders of intelligence

products” and analytic methods employed, creation of a net

assessment gr

oup to challenge assumptions, and “stronger in-

centives to attract good analysts,” including improved career

opportunities.

In response to the report’s recommendations, the president

issued a directive calling for a limited reorganization.

The re-

organization involved implementing the report’s second option

by enhancing the power of the DCI, establishing an Intelli-

gence Resources Advisory Committee to prepare a consoli-

dated intelligence budget, and establishing a net assessment

oup within the NSC to r

eview and evaluate intelligence pr

od-

ucts and pr

oduce net assessments. The directive also consoli-

dated DOD mapping agencies under a new organization, the

Defense Mapping Agency.

Finally, in 1972 the dir

ective led to

establishment of the IC Staf

f, the pr

edecessor to the curr

ent

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Community Management Staff.

All the recommendations were

implemented through the executive branch, opening the re-

form efforts to criticism that would become more prevalent in

future years—namely, the lack of a statutory basis to support

the DCI in exercising his newly granted authority.

The executive branch’s relative activism with regard to this

particular reform effort is notable, as it does not appear to

have been driven by either direct intelligence failures or pres-

sures from Congress. The Schlesinger effort also highlights

two recurring themes in intelligence reform. First, it followed

the relatively continuous path of previous reform efforts in its

push towards greater centralization of intelligence efforts both

for improved efficiency and for improved analytic production.

Second, it emphasized one of the fundamental requirements

for greater centralization: commensurate authority to achieve

the desired r

esults. The lack of such authority, or the failure

to properly exercise the authority provided, has consistently

undermined centralization efforts—a topic addressed by the

next commission to review the IC.

The Murphy Commission

In 1972 Congress authorized the Commission on the Orga-

nization of the Government for the Conduct of Foreign Policy,

chaired by Amb. Robert Murphy.

The study of the IC was

part of a broader effort, and as such, the recommendations

and findings are much less detailed than those of some other

studies. One of the commission’s insights highlighted the im-

portance of organization vis-à-vis the effectiveness of govern-

ment, finding that “organization affects more than the effi-

ciency of government; it can affect the outcome of decisions.

Organizational patterns determine the probabilities that a de-

cision will be taken at one level rather than another, or in one

agency instead of another

.”

On issues of intelligence organization, the Murphy Commis-

sion set two standards of performance by the IC: (1) analytic

esponsiveness to decision makers with accompanying compe

tence and integrity of the intelligence product, and (2) econ-

omy in the planning, programming, and budgeting of intelli-

gence systems.

The findings identified gaps in analysis, poor

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

collection focus, and poor cost management in the develop-

ment of systems. According to the commission, the sources of

the problems included the numerous agencies and lines of

command involved in the intelligence process, the relative

strength of the DOD with respect to budget and manpower re-

sources, and a tendency to chase technology without first ex-

amining whether new capabilities were truly necessary.

The

commission further noted that the DCI lacked line authority

over organizations within the IC other than the CIA; however,

it did not recommend granting such authority to the DCI.

The Murphy Commission recommended moving the DCI

physically closer to the White House to provide more direct ac-

cess to the president and counterbalance the lack of line au-

thority over other organizations within the IC. Like previous

studies, the report argued for incr

eased centralization and

stressed the need for the DCI to emphasize management of

community affairs over management of the CIA.

The com-

mission also recommended a strengthening both of the PFIAB

to provide a source of independent advice to the president and

of the NSC Intelligence Council to improve guidance to the IC

on the intelligence r

equirements of policy makers.

In addi

tion, the commission recommended that all departments main-

tain and strengthen their internal analysis efforts—an ap-

proach more supportive of decentralized intelligence analysis

than that of similar studies.

To improve the national estima-

tive process, the commission’s recommendations focused on

more clearly reflecting analytic differences across agencies in

intelligence products and dedicating staff to the formulation of

estimates rather than forming ad hoc groups to develop the

products.

Finally, presaging findings in the 9/11 Report, the

commission warned that human intelligence was being mar

ginalized at the expense of more costly technical solutions. It

therefore recommended that the IC place renewed emphasis

and investment on using human sour

ces of intelligence.

The Church Committee

Until 1975 the executive branch oversaw nearly all signifi-

cant changes to the intelligence structur

e. This changed in

early 1975, when both the Senate and House began to inves

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

tigate alleged illegal domestic intelligence activities as well as

specific improprieties of certain CIA operations.

The Senate

formed the Select Committee to Study Governmental Opera-

tions with Respect to Intelligence Activities under the direction

of Senator Frank Church (D-ID). The committee had two

tasks: investigate charges of abuse and propose legislative

remedies to correct any abuses or deficiencies.

Over a period

of 15 months, the Church Committee held a series of hearings,

interviewed hundreds of witnesses, and ultimately released

183 recommendations encapsulated in six final reports

roughly reflecting the various areas of investigation.

The

committee called for a compr

ehensive legislative charter for

the IC to include organic acts to recognize the different intelli-

gence agencies that had been established since 1947, sort out

roles and missions for the various agencies, and establish leg-

islative restrictions on the types of activities in which intelli-

gence agencies could engage.

Although the committee focused on allegations of abuse, it

made r

ecommendations regarding the management of the IC

at large. For example, it found that the DCI had insufficient

authority, failed to pr

event unnecessary waste and duplica-

tion, and did not adequately steer the IC toward meeting the

needs of policy makers.

Other significant findings dealt with

the need to properly balance national and tactical require-

ments and the DCI’s difficulties in removing department and

agency bias in the for

mulation of intelligence estimates, par-

ticularly when the White House or the DOD applied the pres-

sure.

The committee commented on the dual roles of the

DCI, finding that the two roles created both a conflict of inter-

est in managing the community and a span of control too large

for the DCI to execute both functions adequately.

The Church Committee also found that analysis was subor-

dinated to other activities. Analysts had insufficient and some-

times negative incentive structures, were often overwhelmed

with infor

mation, and wer

e disconnected fr

om the policy mak-

ers who used their pr

oduct.

Recommendations included

changing incentive structures to allow analysts to remain in

analysis positions longer

, injecting analyst personnel from

outside the traditional confines of the IC, and assigning ana

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

lysts to perform exchange tours with other agencies and aca-

demic institutions.

The Pike Committee

The House also formed an investigative committee under the

chairmanship of Rep. Otis Pike (D-NY) to conduct its own in-

vestigations of the IC. The Pike Committee had a far more ad-

versarial relationship with the executive branch than the

Church Committee. It “failed to conduct well-documented,

complete, and effective hearings” and produced a report that

the House ultimately voted not to publish.

However, the

House published the committee’s recommendations in 1976.

The recommendations included that the DCI be given central

power over the IC, that the position of DCI be separated from

that of head of CIA “with a view to eliminating duplication in

collection and promoting competition in analysis,” that budget

totals of each agency be disclosed, and that the Defense Intel

ligence Agency (DIA) be abolished because of perceived dupli-

cation with the CIA.

However, the controversy over both the

conduct of the investigation and the publication of the final re-

port diminished the overall legacy of the committee.

The Rockefeller Commission

The executive branch performed its own investigation via a

commission chaired by Vice Pres. Nelson A. Rockefeller.

the congressional investigations, the Commission on CIA Activi-

ties within the United States (Rockefeller Commission) focused

almost exclusively on the conduct of one agency, namely the CIA.

However

, some of its recommendations reappeared in later re-

form efforts. They included the development of a Joint Commit-

tee on Intelligence in Congress, with the Joint Committee on

Atomic Ener

gy cited as one example of successful joint oversight.

The commission also recommended the public release of CIA

budget figures, presaging later calls for budgetary transparency

within the IC. Echoing the Murphy Commission, Rockefeller

Commission recommended strengthening the PFIAB to provide

additional oversight, including “assessing the quality of for

eign

intelligence estimates.”

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Although proposed for different reasons, several of the rec-

ommendations affecting the FBI and CIA remain relevant

today. The first focused on improving the analytic capabilities

of the FBI to “evaluate, analyze and coordinate intelligence and

counterintelligence collected by the FBI concerning espionage,

terrorism, and other related matters of internal security.”

second recommendation was that the CIA and FBI prepare an

agreement to describe each agency’s jurisdiction and to provide

“for effective liaison on matters of mutual concern.”

Results

In response to these investigations, Congress attempted to

pass legislation that would enact many of the recommenda-

tions, principally those made by the Chur

ch Committee. The

most comprehensive of these efforts was a bill called the “Na-

tional Intelligence Reor

ganization and Reform Act of 1978,”

which included a proposal to create a DNI who would retain

control over the CIA.

According to one analysis of the effort,

Congr

ess debated separating the two r

oles of the DCI but r

jected changes because the loss of the CIA would “deprive the

Dir

ector of a str

ong institutional base and would subject him

to more pr

essur

e from policymakers.”

None of the measures

passed.

However, both the legislative and executive branches took

some action during the Carter administration. In 1978 Con-

gress passed the Foreign Intelligence Surveillance Act (FISA).

Also, as suggested by the Church Committee final report, Con-

gress passed the Intelligence Oversight Act of 1980, establish-

ing new requirements on the IC to report significant activities

to the congressional intelligence committees.

Perhaps the

greatest legacy of the investigations was the establishment of

permanent select committees on intelligence in both houses to

enhance congressional oversight of intelligence activities.

These committees stand in stark contrast to the oversight pro-

vided prior to 1975 via the Ar

med Services committees of the

House and Senate, where few meetings were held, no notes

were taken, and relationships with the intelligence agencies

wer

e consider

ed mor

e collegial than confrontational.

How

ever

, the formation of the Senate Select Committee on Intelli-

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

gence (SSCI) in the Senate was not without controversy, and

the Senate Armed Services Committee (SASC) asserted its pre-

rogative to retain oversight of tactical intelligence programs—

a structure that continues today.

Congress also established

the precedent in 1977 of enacting intelligence authorization

bills similar to those for other executive department activities.

In the executive branch, perhaps to preempt greater legisla-

tive activism, Pres. Gerald Ford issued EO 11905 to further

clarify the responsibilities of IC member organizations, to in-

clude the reporting and identification of unlawful activity.

The executive order also created the Intelligence Oversight

Boar

d to examine intelligence activities of the IC, review inter-

nal practices of intelligence organizations, and report findings

to the attorney general and the president.

The SSCI also

worked directly with the Carter administration to formulate

EO 12036, which further prescribed the duties of intelligence

organizations, placed r

estrictions on intelligence activities,

and defined oversight pr

ovisions.

These executive or

ders ul-

timately implemented several aspects of the Church Commit-

tee’s recommendations and thus can be reasonably construed

as products of successful legislative oversight.

The 1980s

Following the Iran-Contra scandal, Congress again became

more active, proposing the separation of the DCI’s role as head

of the IC from his role as head of the CIA. During the Reagan

administration, Senator Arlen Specter (R-PA) put forth legisla-

tive recommendations in both the 100th and 101st Congresses

to split these duties through the creation of a DNI who would

be charged with coordinating the overall efforts of the IC.

A CRS study attributed the incr

eased congr

essional activity

to a number of factors including (1) the increasing expansion

of the IC, (2) the dependence of effective community coordina-

tion upon the degree of attention devoted to this role by the in-

cumbent DCI, (3) perceived intelligence failures, and (4) alle-

gations of illegality.

With the exception of the fourth

attribute, the same conditions wer

e appar

ent 15 years later as

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

intelligence reform efforts once again received substantial leg-

islative attention.

The 1990s

Reform efforts internal to the IC paralleled those of high-

level commissions and boards. For example, in 1991 DCI

Robert Gates established multiple task forces to examine both

internal CIA and community-wide processes. The findings led

to the establishment of offices to better utilize open-source in-

formation and the strengthening of the DCI staff to aid in com-

munity management efforts.

Concurrent with the DCI’s effort,

the chairmen of the congressional intelligence committees,

Senator David Boren (R-OK) and Cong. Dave McCurdy (D-PA),

introduced legislation seeking “sweeping changes in the intel

li-

gence apparatus of government” necessitated by the end of the

cold war.

The legislative effort, which was never enacted, would have

established a DNI and released the top-line budget amount.

The legislation echoed the Schlesinger Report with respect to

the DNI’s relation with the CIA. It recommended that the DNI

gain control over the analytic resources of the CIA, thereby ef-

fectively pr

oviding the DNI with a centralized intelligence func

tion while ceding daily control of the remainder of CIA func-

tions to a separate director.

The proposed legislation would

have significantly increased the authority of the head of the IC

(in this case a DNI), granting the authority to transfer person-

nel for up to 180 days and transfer funds without the consent

of department heads.

Finally, the legislation also proposed

the creation of a National Imagery Agency.

Senator Boren’s recommendations included providing a DNI

with a gr

eater r

ole in the hiring of intelligence agency heads.

The restructuring envisioned was significant, effectively pro-

ducing four major organizations: three of them devoted to the

intelligence collection disciplines of signals, imagery, and

human intelligence, respectively, and a fourth devoted to intel-

ligence analysis.

The goal of such a restructuring, like those

of most drives towards centralization, was to increase effi-

ciency and accountability. Other r

ecommendations included

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

improving the collection of human intelligence and increasing

the use of open-source information. Senator Boren also com-

mented on problems with intelligence estimates, suggesting

they were “overly cautious, caveated and consensus driven,”

and recommended augmenting the cadre of IC analysts with

outsiders from the private sector and academe.

The Aspin-Brown Commission

In October 1994 Congress continued its examination of the

IC in light of several developments that included the end of the

cold war, a perceived lack of executive branch will to make

changes, concern over the amount of resources committed to

intelligence, and general charges of mismanagement within the

component or

ganizations.

The Commission on the Roles and

Capabilities of the United States Intelligence Community, also

known as the Aspin-Br

own Commission, was given a mandate

to assess the roles of the various agencies and perform a com-

prehensive analysis of the IC, touching on both community-

wide issues and issues specific to individual or

ganizations. The

committee’s report included recommendations to improve over-

all guidance to the IC thr

ough a per

manent NSC organization;

to develop closer ties between the IC and the policy community;

to improve central contr

ol of the IC; to improve sharing of in-

formation between law enforcement and intelligence agencies;

and to improve planning, programming, and budgeting

processes to reduce waste and redundancy.

The commission found that the DCI’s authority had strength-

ened over time with regard to establishing requirements and

budget consolidation; however, the report also noted the lack of

line authority that undermined this power.

In the course of

studying the DCI’s roles, the commission examined some ap-

proaches to decentralize intelligence and others to further in-

crease centralization. The decentralized approach involved pro-

viding budgetary authority to the individual agencies and

departments and making the DCI r

esponsible only for CIA pr

grams.

The commission rejected this approach because of the

potential for greater waste and duplication as well as the possi-

bility that non-DOD activities would gradually lose the attention

of decision makers.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Another approach looked at increasing centralization by giv-

ing the DCI line authority over several major DOD intelligence

agencies. This approach was rejected because the commission

considered these organizations critical in providing support to

military operations and believed that removing them from the

DOD would jeopardize this support. Equally important was

that military personnel staffed many of these organizations. A

shift in line authority would likely have resulted in the DOD’s

retention of these personnel, with a concomitant requirement

to replace them with non-DOD civilians at substantial costs.

A different option for centralization was to create a DNI who

would oversee the entir

e community, but the commission also

rejected this because it believed that the loss of leadership

over the CIA would reduce the DCI’s overall authority.

The commission eventually chose a relatively middle course

that would not have significantly altered the centralization/de-

centralization balance within the IC. It involved strengthening

the DCI’s authority by requiring the DCI’s concurrence on the

appointment of heads of the national intelligence agencies

within

the DOD, dual-hatting these officials under the CIA, and speci-

fying by statute two deputies for the DCI—one to manage the

CIA and the other to manage the IC.

However, the commis-

sion did not believe other legal changes were required to pr

vide the DCI with the increased authority.

The commission made numerous recommendations in other

eas as well. To improve budgetary management, the com-

mission focused on organizing the budget around intelligence

disciplines, such as signals and imagery, with the goal of re-

ducing duplication and waste.

Recommendations for improv-

ing intelligence analysis suggested improving the relationships

between intelligence producers and consumers (policy mak-

ers). While acknowledging the potential for politicization, the

committee saw a greater danger in too much distance between

the IC and the policy maker, which would result in intelligence

analysis becoming “irr

elevant to the pr

ocess of gover

nment.”

Like pr

evious commissions, Aspin-Brown made suggestions to

improve the quality of analysts, to increase the use of open-

sour

ce information, and to develop stronger ties with experts

outside the IC. In contrast to both prior and futur

e r

efor

m ef

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

forts, cost-saving efforts focused on cuts in personnel rather

than in intelligence systems.

The commission also recom-

mended that the IC attempt to consolidate its personnel func-

tions to allow better rotation of staff between various agencies.

Finally, echoing the 1992 legislative proposals, the commis-

sion recommended disclosure of top-line budget amounts.

IC21: The Intelligence Community in the 21st

Century

The House Permanent Select Committee on Intelligence

(HPSCI) undertook a second investigation of the IC in an at-

tempt to reap cost savings following the end of the cold war.

Notably, the resulting report,

IC21: The Intelligence Community

in the 21st Century

(her

eafter

IC21), was written within a con

text that maintained the key issue was “opportunity, not reform”

and stated the “US national security inter

ests are less threat-

ened than at any time since 1940.”

IC21 examined the role

of the DCI along the lines of greater or lesser centralization

and str

ongly supported incr

easing the DCI’s authority, citing

a “glaring gap between his responsibilities and his authori-

ties.”

The r

eport r

ecommended increasing DCI control over

budgets, which was considered to be the “essential bur

eau

cratic lever”; creating a str

onger community management staff;

providing the DCI authority to detail intelligence personnel be-

tween agencies; and giving the DCI greater say in the appoint-

ment of agency heads.

It cited testimony by former DCIs who

expressed the importance of retaining an agency reporting di-

rectly to the director to provide another source of power. There-

fore, unlike other studies, the report did not envision creating

a separate DNI. Instead, the HPSCI chose to strengthen the

roles of deputy DCIs to provide more oversight over both the

CIA and the larger community.

IC21 did not recommend consolidation of analytic resources

within a single organization. It found that organic analytic ef-

forts pr

ovided important and tailor

ed support to their parent

departments and also aided in the competitive analysis

process. The report reiterated previous calls to allow greater

personnel exchange, cr

eate commonality in personnel systems

at the senior executive level, and—echoing Goldwater

-Nichols—

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

require rotational assignments within the community as a

prerequisite for promotion.

Improving analysis was seen as

a function of establishing greater linkages with policy makers

and collectors to understand better what information policy

makers desired. Like the Aspin-Brown Commission,

IC21 rec-

ommended reestablishing the Committee on Foreign Intelli-

gence (CFI) to provide more direct feedback and interaction be-

tween policy makers and the IC and thus improve policy

coordination.

The report also studied the existing analytic centers, which

at the time numbered seven.

101

It found that such centers

wer

e valuable when they addressed “enduring intelligence is-

sues”; however, the centers were not true interagency con-

structs and were unevenly staffed by the various agencies.

Moreover, IC personnel did not consider participation in cen-

ters career enhancing. To address these shortcomings, the re-

port recommended changes similar to the joint personnel prac-

tices of Goldwater-Nichols, including increased incentives,

impr

oved promotion opportunity, and increased participation

from IC component organizations other than the CIA.

IC21

specifically stated, “the IC personnel evaluation and pr

omotion

systems must accurately reflect and reward the performance

of employees detailed to centers.”

102

On the subject of informa

tion sharing, specifically with law enforcement agencies, the

report found that existing statutes were sufficient to allow the

necessary practices, but that specific infor

mation manage-

ment and policies should be improved to allow for sufficient

information sharing.

The Aspin-Brown report and the

IC21 report were released

in March and April 1996 respectively. The initial response

from Congress appears to have been relatively substantial in

comparison to its response to other reform efforts. In the Intel-

ligence Authorization Act for Fiscal Year 1997, Congress estab-

lished the Committee on Foreign Intelligence and the Commit-

tee on T

ransnational Thr

eats within the NSC, enhanced the

authority of the DCI over budgets and personnel, and cr

eated

two deputy director positions under the DCI and three assis-

tant DCIs (ADCI) to aid in community management activi

ties.

103

The CFI was tasked, among other functions, to estab

lish

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

“priorities (including funding priorities) among the programs,

projects, and activities that address” national security needs.

Thus, in effect, the legislation moved a key element of commu-

nity management up one level in the executive branch hierar-

chy. The additional authorities granted to the DCI were largely

cosmetic, charging the DCI to “facilitate the development of

the annual budget” and to concur in the appointment of se-

lected intelligence officials within the DOD.

104

The additional managerial layers within the DCI community

staff might have helped the DCI to deal with span-of-control

issues; however, the executive branch resisted this measure.

President Clinton was particularly concerned with Congress’s

forced establishment of two committees on the NSC staff as an

unnecessary invasion into executive branch jurisdiction, and

he supported the DCI’s resistance to the new ADCIs estab-

lished for community management. In fact, only two individu-

als were ever nominated to fill ADCI positions, and after they

left office in 2003 their positions were not filled.

105

Further at-

testing to the relatively limited effect of reform, one staff mem-

ber of the Aspin-Brown commission stated, “The IC was, in

sum, largely unaffected by the Aspin-Brown inquiry—or for

that matter, the similar r

eform proposal of IC21.”

106

The 9/11 Inquiries

After the events of 9/11, the intelligence committees of Con-

gress per

formed a joint inquiry of the IC to conduct “a factual

review of what the intelligence community knew or should

have known prior to September 11, 2001” and to identify “sys-

temic problems that may have impeded the intelligence com-

munity.”

107

In addition, Congress created the National Com-

mission on T

err

orist Attacks upon the United States,

commonly referred to as the 9/11 Commission.

108

In addition

to examining in detail the events and intelligence prior to the

attacks, both investigations led to recommendations on re-

structuring the IC that were partially enacted in subsequent

legislation.

In many instances the joint inquiry findings were remark-

ably similar to those later published in the 9/11 Report. These

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

findings included inadequate community structures, poor tech-

nological interfaces, and failures of information sharing.

109

Recommendations from the congressional inquiry included

separating the DCI’s two roles by creating a cabinet-level DNI

who possessed the “full range of management, budgetary, and

personnel responsibilities needed.”

110

Judging from the ampli-

fying text, the inquiry appears to have recommended giving

the DCI nearly complete control over budgetary, personnel,

and collection resources. The inquiry also used the Goldwater-

Nichols metaphor in recommending the application of joint per-

sonnel practices to the IC, including joint assignments of per-

sonnel to other agencies and impr

oved incentives for those

personnel who take such assignments.

111

Analyzing Past Reform Efforts

Even a cursory overview of the evolution of the IC and of the

various examinations of its organization and function leads to

several conclusions. First, governmental action has resulted

in a gradual shift towar

d greater centralization of intelligence

activities. Initially, the focus was on placing greater emphasis

on the community management role of the DCI at the expense

of dir

ect management of the CIA. Recommendations and follow-

on executive branch action have resulted in the cr

eation of

deputies and staff structures to better enable the DCI to coor-

dinate the activities of the IC.

Second, as the emphasis towards greater centralization pro-

gressed, recognition of the inadequacy of the DCI’s authority over

the IC became more prevalent. Recommendations for increasing

that authority ranged from enhancing the prestige of the DCI vis-

à-vis the other organizations (informal authority) to providing

eater budgetary or line authority over various IC or

ganizations

(formal authority). Only a few of the studies recommended ele-

vating the DCI to cabinet-level status. Many of the studies sug-

gested increases in DCI budgetary authority. EOs accomplished

this to a degree while also clarifying DCI responsibilities. Provid-

ing the DCI more line authority over various agencies was decid-

edly mor

e complicated and was never implemented.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Third, beginning in 1960, more studies advocated a com-

plete separation of the DCI’s two roles, but disagreed as to the

resources and authorities the DCI should have.

The quan-

tity and type of manpower assigned to a separate DNI were one

area of debate. Too few personnel assigned would create the

impression of a czar, whereas too many could overburden the

DNI with excessive managerial responsibilities, reintroducing

one of the primary reasons for advocating separation of roles.

Fourth, the IC has been far from static. While the amount of

comprehensive legislation mandating organizational change is

relatively slight, it has not been completely absent. Laws af-

fecting the operations and oversight of agencies wer

e supple-

mented by laws enacted to establish charters and recognize

various agencies and more recently to create additional assis-

tant directors who would aid in community management.

However, the executive branch made the more significant

changes, forming major new intelligence organizations includ-

ing all of the combat support agencies: the NSA in 1952, DIA

in 1961, and mor

e recently the NGA. Additional EOs clarified

and amplified the responsibilities and tasks of the intelligence

agencies.

Fifth, both the legislative and executive branches have im-

portant roles in adjusting and maintaining the US intelligence

apparatus. Executive branch action offers both speed and nu-

ance. If a directive has significant or unintended side effects,

executive or

ders have the added advantage that they can be is-

sued and altered far more easily and quickly than correspon-

ding legislative actions. Moreover, managerial proximity to the

affected organizations in certain situations may give the exec-

utive branch better understanding of both the problems and

potential consequences of proposed remedies.

Aside from oversight, Congress has multiple roles. First, by

threatening legislative action Congress can compel the execu-

tive branch to act. For example, President Carter’s executive

ders wer

e meant, at least in part, to pr

eclude further con-

essional action. Second, Congress can play an advisory role,

as it did in consulting with the Carter administration on the

content of various executive or

ders. Third, the executive

branch is pr

ohibited fr

om taking certain actions, such as cr

eat

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

ing a DNI. Fourth, legislation provides a degree of permanence

not found in executive orders. As the National Security Act of

1947 illustrates, the effect of legislation is often long lasting.

Thus, prudence is warranted, as what is set in motion often

continues without legislative intervention for prolonged peri-

ods. In addition, as the final check on executive power, Con-

gress can pass legislation to implement its intent more fully

when it believes the executive branch lacks the will to make

the perceived necessary changes.

Finally, the problems that faced decision makers in the late

1940s remain relevant today. Calls for improved information

sharing, greater reliance on open-source information, and bet-

ter coordination of activities have persisted since the IC was

founded. As early as 1975, the Murphy Commission discussed

the marginalization of human sources of intelligence.

113

Various

other studies highlighted the importance of coordinated intelli-

gence activities with the FBI. Numer

ous studies and commis-

sions have expressed dissatisfaction with the quality of analytic

products and methodologies. Calls for increased competitive

analysis, str

onger car

eer incentives for analysts, and augmen

tation of the analysis staff with experts from outside the IC were

common, as was emphasis on the need to maintain a close r

lationship with consumers of intelligence (policy makers) while

avoiding excessive politicization.

114

However there was less

agreement as to whether intelligence analysis should become

more centralized or remain decentralized with the departments

maintaining or even strengthening their analytic arms.

The persistence of these problems may represent less an in-

dictment of the government’s failure to correct known deficien-

cies than an indication that the pr

oblems defy simple solu

tions. The complexity of the underlying issues suggests the

existence of competing interests that further complicate re-

form efforts. How to balance these interests while having the

intended effect is the subject of chapter 5. However, before

tur

ning to the discussion of balances, chapter 4 examines the

9/11 Commission Report, ar

guably one of the most significant

commission r

eports in the last 50 years.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Notes

1. Lt Gen Michael Hayden, interviewed by the author, 18 February 2005.

More specifically, General Hayden described the IC as an organization “gov-

erned by a principle of consensus over the last fifty years.”

2. Best,

Proposals for Intelligence Reorganization (see chap. 1, n. 6). See also

Alfred B. Prados,

Intelligence Community Leadership: Development and Debate

Since 1947, CRS-89-414 F (Washington, D.C.: CRS, 27 June 1989). Best exam-

ines both legislative and executive initiatives to reorganize the IC, citing over 20

examples of various initiatives. A core focus of the Prados paper was on the ori-

gins and rationale of the DCI’s dual roles as well as proposals to alter those

roles. David E. Kaplan, Kevin Whitelaw, and Monica M. Ekman, “Mission Im-

possible,”

U.S. News & World Report 137, no. 3 (2 August 2004), http://

www.usnews.com/usnews/news/articles/040802/2intell.htm.

3. The list of reform efforts is not exhaustive and purposely excludes

studies that had as their chief focus the activities of specific agencies. If a

study had relevant commentary on the management of the larger IC, it was

included.

National Security Act of 1947, Public Law 253, 80th Cong., 1st sess.

(26 July 1947), sec. 102 (a).

5. Ibid., sec. 102 (d)(1)-(3). The intelligence departments in question were

the intelligence arms of the Departments of the Army, Navy, and newly es-

tablished Air Force, as well as the intelligence arm of the Department of

State.

6. Ibid., sec. 102 (d)(3) and sec. 102 (e).

7. Allen W. Dulles, William H. Jackson, and Matthias Correa, “The Cen-

tral Intelligence Agency and the National Organization for Intelligence,” Re-

port to the National Security Council, 1 January 1949, iv. The NSC tasked

the group, composed of Allen Dulles (future DCI), William Jackson (future

deputy DCI) and Matthias Correa, to perform “a thorough and comprehen-

sive evaluation of the structure, administration, activities, and inter-agency

relationships of the Central Intelligence Agency.” See also Best,

Proposals for

Intelligence Reorganization (see chap. 1, n. 6), 6; and Commission on the

Roles and Capabilities of the United States Intelligence Community, “Prepar-

ing for the 21st Century: An Appraisal of US Intelligence,”1 March 1996,

http://www.access.gpo.gov/int/report.html, A-10, hereafter cited as Aspin-

Brown.

8. Aspin-Brown, A-10. Thus, almost from the beginning, there were

problems with the community management r

ole. See also Dulles, Jackson,

and Correa, “Central Intelligence Agency,” 11. The criticism of the CIA was

strongly worded, “The principal defect of the Central Intelligence Agency is

that its dir

ection, administration, organization and performance do not

show suf

ficient appr

eciation of the Agency’s assigned functions, particularly

in the fields of intelligence coordination and the production of intelligence

estimates.”

Best,

oposals for Intelligence Reor

ganization

(see chap. 1, n. 6), 6.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

0. Dulles, Jackson, and Correa, “Central Intelligence Agency,” 3–6. The

Intelligence Advisory Committee included the DCI as chairman and the

heads of the intelligence staffs of the Departments of State, Army, Navy, and

Air Force, the Atomic Energy Commission, and the Joint Intelligence Group

of the Joint Staff.

11. Ibid., 3.

12. Ibid., 5.

13. Ibid., 55. The report recognized that the CIA may, in certain in-

stances, have insufficient authority. However, it suggested that the CIA

should use the NSC as an arbiter to resolve disputes and should make rec-

ommendations to the NSC to resolve disputes. Specifically, the report stated,

the “lack of authority in a specific situation should not deter the Central In-

telligence Agency from exercising its responsibility to submit recommenda-

tions [to the NSC] so that proper coordination will result.”

14. Committee on National Security Organization, “Task Force Report on

National Security Organization (Appendix G),” January 1949.

15. Ibid., 76. The report found that “relationships . . . notably to G-2 of

the Army, the FBI, the Atomic Energy Commission and the State Depart-

ment have been and still are unsatisfactory.”

16. Ibid., 76.

17. Best,

Proposals for Intelligence Reorganization (see chap. 1, n. 6), 8.

18. Commission on Organization of the Executive Branch of the Govern-

ment, “Intelligence Activities,” May 1955, 13 (hereafter the Second Hoover

Commission). This appears to be the first government document to explic-

itly use the term IC to refer to the confederation of agencies with intelligence

responsibilities. The IC consisted at that time of the CIA, NSC, NSA, FBI, in-

telligence sections of the Department of State, Navy, Army, and Air Force,

and the Atomic Energy Commission.

19. Ibid., 17.

20. Ibid., 67.

21. Ibid, 69–70. The remainder of the Second Hoover Commission find-

ings dealt with issues such as the establishment of an oversight body to pe-

riodically review intelligence organization and operations, the need for periodic

reinvestigations of intelligence personnel, and the need to exploit open-source

information more effectively.

22. Prados,

Intelligence Community Leadership, 42.

23. See Senate, Select Committee to Study Governmental Operations

with Respect to Intelligence Activities,

Supplementary Detailed Staff Reports

on Foreign and Military Intelligence, 94th Cong. 2d sess., Report 94-755,

Book IV, 23 April 1976, 45. There are multiple explanations for why Dulles

emphasized his role as director of the CIA at the expense of the DCI, includ-

ing his relationship with President Eisenhower and his supposed dislike for

confrontation.

24.

Prados,

Intelligence Community Leadership, 42.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

5. James R. Killian, Jr., board chairman, PBCFIA, to the president, let-

ter, 20 December 1956, released 17 July 1977, quoted in Prados,

Intelligence

Community Leadership

, 42.

26. Prados,

Intelligence Community Leadership, 44.

27. Pres. John F. Kennedy, “Memorandum for: Director of Central Intel-

ligence,” 16 January 1962, quoted in Prados,

Intelligence Community Lead-

ership

, 45. The extent to which later presidents demanded similar prece-

dence to the role of community management is often less clear and explicit.

For example, in recent testimony from DCI Stansfield Turner, President

Carter held a similar view to President Kennedy’s as to the priority to place

on the competing roles. As Admiral Turner stated, “At my very first meeting

with the president, before he had actually designated me as his nominee for

director of central intelligence, he gave me oral instruction that if I took this

job I was to concentrate on being the director of central intelligence, not on

being the head of the CIA.” Testimony before the Hearing of the Senate Com-

mittee on Governmental Affairs, “Reorganizing America’s Intelligence Com-

munity: A View from the Inside,” 16 August 2004.

28. Best,

Proposals for Intelligence Reorganization (see chap. 1, n. 6), 16.

29. Office of Management and Budget, “A Review of the Intelligence Com-

munity,” 10 March 1971, 1–2 (hereafter cited as Schlesinger Report).

30. Schlesinger Report, 8–10, 15.

31. Ibid., 13.

32. Ibid., 21–23. On the specific issue of oversight, the report looked

more to the PFIAB, OMB, and other organizations to provide this oversight,

not Congress.

33. Ibid., 25.

34. Ibid., 27–28.

35. Ibid., 30.

36. Ibid., 31.

37. Ibid., 45–46.

38. “Reorganization of the IC,”

Weekly Compilation of Presidential Docu-

ments

, 4 November 1971, 1482.

39. This agency was eventually combined with other organizations to

form the present-day National Geospatial-Intelligence Agency.

40. Senate, Select Committee to Study Governmental Operations with

Respect to Intelligence Activities,

Supplementary Detailed Staff Reports, book

4, 84; hereafter cited as Church Committee Report.

41.

Prados,

Intelligence Community Leadership, 46.

42. Commission on the Organization of the Government for the Conduct

of Foreign Policy, “Report,” June 1975, hereafter cited as Murphy Commis-

sion. The authorizing legislation for the commission was

Foreign Relations

Authorization Act,

Public Law 352, 92d Cong., 2d sess., 13 July 1972.

43. Murphy Commission, 21.

44.

Ibid., 91.

45.

Ibid., 97.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

6. Best,

roposals for Intelligence Reorganization

(

see chap. 1, n. 6), 18.

The commission states, “It is neither possible nor desirable to give the DCI

line authority over that very large fraction of the IC which lies outside the

CIA. If he is to effectively supervise the whole community, however, there is

only one substitute for such authority, and that is a close relationship with

the one official who does ultimately command each of the separate strands

of the community: the President,” Murphy Commission, 98.

47. Murphy Commission, 98.

48. Ibid., 99.

49. Ibid., 103.

50. Ibid.

51. Ibid., 105. See also 9/11 Report, 104 (see chap. 1, n. 1).

52. Prados,

Intelligence Community Leadership, 6.

53. Frank J. Smist, Jr.,

Congress Oversees the United States Intelligence

Community

, 2d ed. (Knoxville: University of Tennessee Press, 1994), 25. The

committee was officially formed on 27 January 1975.

54. Ibid., 77.

55. Ibid., 125.

56. Church Committee, book 1, 432.

57. Ibid., 432–33.

58. Ibid., 434.

59. Ibid., 444.

60. Best,

Proposals for Intelligence Reorganization (see chap. 1, n. 6), 24,

speaks of the adversarial nature of the committee. See also Smist,

Congress

Oversees

, 195.

61. Best,

Proposals for Intelligence Reorganization, 24.

62. House, Select Committee on Intelligence,

Recommendations of the Final

Report of the House Select Committee on Intelligence

, 94th Cong., 2d sess., 11

February 1976, H. Rpt. 94-833, 3–4. For budget recommendations, see 5.

63. Commission on CIA Activities within the United States,

Report to the

President

(Washington, D.C.: GPO, 6 June 1975), http://historymatters.

com/archive/church/rockcomm/html/Rockefeller_0001a.htm (accessed 25

July 2005) (hereafter cited as Rockefeller Commission).

64. Ibid., 15.

65. Ibid., 23.

66. Ibid., 30. The nature and extent of these agreements between the FBI

and CIA are unclear. What is clearer is that these same factors were present

25 years later. For example, the 9/11 Report detailed how effective liaison,

jurisdictional problems, and lack of information sharing were contributory

factors in the 9/11 attacks.

67.

Best,

oposals for Intelligence Reorganization

(see chap. 1, n. 6), 26.

68. Aspin-Brown, A-19. In addition, although not directly related to IC

efor

m, Congr

ess passed the For

eign Intelligence Surveillance Act of 1978 to

establish procedures for authorization of electronic surveillance.

69.

Intelligence Authorization Act for Fiscal Year 1981, Public Law 450,

96th Cong., 2d sess. (14 October 1980). See also

Intelligence Oversight Act of

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

980

96th Cong., 2d sess., Senate Report 96-730, 15 May 1980. One of the

provisions created an “affirmative duty to keep the two intelligence commit-

tees fully informed of significant activities” (i.e. those with policy implica-

tions), 7. The primary intent of the various provisions appears to have been

to establish more formal reporting congressional requirements for a discrete

set of IC activities.

70. Smist,

Congress Oversees, 7, 9.

71. Ibid., 83.

72. Executive Order 11905, sec. 4 a(3)–(5), United States Foreign Intelli-

gence Activities, 18 February 1976.

73. Ibid., sec. 6 (a)(3).

74. Smist,

Congress Oversees, 103; and Executive Order 12036, United

States Intelligence Activities, 24 January 1978, http://www.fas.org/irp/

offdocs/eo (accessed 22 November 2004). EO 12333, United States Intelli-

gence Activities, 4 December 1981, eventually superseded EO 12036 in 1983.

75. Prados,

Intelligence Community Leadership. Neither proposal passed.

76. Ibid., 7. Besides the Bay of Pigs scandal, no specific instances of fail-

ures were cited in the study, rather the report alludes to instances where in-

telligence failed to provide strategic warning. Allegations of illegality

stemmed from reports of US intelligence agency involvement in domestic ac-

tivities.

77. Commission on the Roles and Capabilities of the United States IC,

“Preparing for the 21st Century: An Appraisal of US Intelligence,” A-23.

Hereafter cited as Aspin-Brown.

78. David L. Boren, “The IC: How Crucial?”

Foreign Affairs 71, no. 3

(Summer 1992): 52–53.

79. Aspin-Brown, A-24. See also

The Intelligence Reorganization Act of

1992,

SR 2198, 102d Cong., 2d sess., Congressional Record 138 (5 February

1992): S1066, which would have created an analysis arm under the DNI.

“There is established under the Deputy Director of National Intelligence for

Estimates and Analysis an Office of Intelligence Analysis that shall be

headed by a director appointed by, and serving at the pleasure of, the Direc-

tor of National Intelligence. The Office shall be composed of analysts as-

signed to agencies within the IC and shall be responsible for preparing all

current intelligence and other analysis that is intended to be disseminated

within the Government as a whole.” Whether this organization would have

replaced the analytic arm of the CIA remained unclear.

80.

Intelligence Reorganization Act of 1992, S. 2198, sec. 202.

81. Ibid., sec. 321.

82. Bor

en, “The IC: How Crucial?” 59.

83.

For a similar proposal see William E. Odom,

Fixing Intelligence (New

Haven, Conn.: Yale University Press, 2003).

84.

Bor

en, “The IC: How Crucial?” 56.

85. Aspin-Brown, 1–2. The authorizing legislation can be found in

Intel-

ligence Authorization Act for Fiscal Year 1995

, Public Law 359, 103d Cong.,

2d sess. (14 October 1994). The duties of the commission were spelled out

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

n sec. 903, which were “to review the efficacy and appropriateness of the

activities of the United States IC in the post-cold war global environment.”

86. Ibid., 33. The NSC organization envisioned was a return to President

Ford’s CFI, chaired by the national security advisor with the deputies of

state and defense. The tasks of the CFI were to establish high-level require-

ments and priorities. Below the CFI would be a consumer’s committee to en-

sure the IC met the needs of policy makers. However, neither committee had

any oversight functions, leaving in question how effective the committees

would be in steering intelligence requirements and priorities.

87. Ibid., xix. Information sharing was focused on global threats to in-

clude terrorism, proliferation of weapons of mass destruction, and narcotics

trafficking. Budgeting recommendations included restructuring the budget

along functional lines such as signals intelligence (SIGINT), imagery intelli-

gence (IMINT), and human intelligence (HUMINT) to reduce “unnecessary

duplication, interoperability problems, and other inefficiencies,” 75.

88. Ibid., 51.

89. Ibid., 52.

90. Ibid., 53.

91. Ibid., 54.

92. Ibid., 57–58. The commission recommended that the DCI concur in

the appointment of the directors of NSA, NIMA, and NRO. The dual-hatting

involved the NSA and the National Imagery and Mapping Agency (now NGA),

creating assistant directors of central intelligence for signals intelligence and

imagery, respectively, and giving these directors program and budget re-

sponsibility for their respective disciplines.

93. Aspin-Brown, 59.

94. Ibid., xxi.

95. Ibid., 86.

96. Ibid., xxii.

97. House, Permanent Select Committee on Intelligence,

IC21: The Intel-

ligence Community in the 21st Century (Washington, D.C.: GPO, 1996), 1.

98. Ibid., 10.

99. Ibid., 11. Envisioned budgetary authority included increased pro-

grammatic control over the National Foreign Intelligence Program (NFIP), in-

cluding the ability to transfer limited funds within the NFIP. The report rec-

ommended that the SecDef appoint agency heads with the advice and

consent of the DCI. The ability to detail personnel was limited to 180 days

without the consent of the agency head.

100.

IC21, 78–79.

101. Ibid., 256. The centers at the time of the report focused on the top-

ics of arms control, nonproliferation, terrorism, counterintelligence, coun-

ternarcotics, organized crime, and overseas security. Furthermore, the cen-

ters were housed in the CIA, and, according to the report had “a distinct

‘CIA’ identity” that pr

ecluded their acceptance as community activities.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

02. Ibid., 45. Such a recommendation is similar to the joint personnel

practices established by Goldwater-Nichols that required the individual

services to reward joint duty.

103.

Intelligence Authorization Act for Fiscal Year 1997, Public Law 293,

104th Cong., 2d sess. (11 October 1996), sections 801–17. The CFI was

composed of the DCI, SecDef, and secretary of state, and chaired by the as-

sistant to the president on national security affairs. The Committee on

Transnational Threats had the same membership with the addition of the

attorney general. The act created ADCIs for collection, administration, and

analysis and production.

104. Ibid., sec. 815. The SecDef could still forward nominations despite

the DCI’s nonconcurrence as long as he/she included a statement of the

DCI’s nonconcurrence in the recommendation.

105. Best,

Proposals for Intelligence Reorganization (see chap. 1, n. 6), 37.

The

Intelligence Reform Act of 2004 eliminated the ADCI positions.

106. Loch K. Johnson, “The Aspin-Brown Intelligence Inquiry: Behind

the Closed Doors of a Blue Ribbon Commission,”

Studies in Intelligence 48,

no. 3 (2004), http://www.cia.gov/csi/studies/vol48no3/index.html (ac-

cessed 19 January 2005). Specifically, “The IC’s budget remained intact; . . .

the weaknesses in accountability went largely unaddressed; and the DCI’s

powers remained stunted.”

107. House Permanent Select Committee on Intelligence and the Senate

Select Committee on Intelligence,

Joint Inquiry into IC Activities before and

after the Terrorist Attacks of September 11, 2001

, 107th Cong., 2d sess., De-

cember 2002, Senate Report 107-351, House Report 107-792, 1.

108. Richard A. Falkenrath, “The 9/11 Commission Report: A Review

Essay,”

International Security 29, no. 3 (Winter 2004/05): 170–90. Falken-

rath asserts that the joint inquiry was essentially a failure, and this led Con-

gress to establish the 9/11 Commission in 2002, 172. Congress established

the 9/11 Commission by

Intelligence Authorization Act for Fiscal Year 2003,

Public Law 306, 107th Cong., 2d sess., 27 November 2002.

109. House Permanent Select Committee on Intelligence and the Senate

Select Committee on Intelligence,

Joint Inquiry, xv–xviii. For inadequacy of

the intelligence organization, see finding 1; for technical issues, see finding

4; and for information sharing problems, see finding 10.

110. Ibid., errata print, 2.

111. Ibid., errata print, 10–11.

112. Prados,

Intelligence Community Leadership, 43. The PBCFIA made

the initial recommendation in 1960; it was reiterated by the PFIAB in 1961.

113.

Murphy Commission, 105.

114. For a sample of analytic recommendations, see Schlesinger Report

45–46.

REFORMING INTELLIGENCE: A 50-YEAR EFFORT

Chapter 4

The 9/11 Report and

Intelligence Reform Legislation

In a world of perfect information, there would be no uncer-

tainties about the present and future intentions, capabili-

ties, and activities of foreign powers. Information, however,

is bound to be imperfect for the most part. Consequently, the

intelligence community can at best reduce the uncertainties

and construct plausible hypotheses about these factors on

the basis of what continues to be partial and often conflict-

ing evidence.

—Schlesinger Report

Intelligence r

emains a critical component of our national se-

curity process. While a world of perfect intelligence that pr

vides ample warning of future threats and complete fore-

knowledge of enemy actions and capabilities is desirable, such

a condition is not attainable. Instead, reforming intelligence

requires changes that provide incremental yet substantive im-

provements in both threat warning and intelligence assess-

ments of enemy capabilities and likely actions. A central ele

ment of refor

m ef

forts thus becomes determining the type of

adjustments, the most appropriate targets for change, and the

potential implications of such changes. Like the reform efforts

and studies that preceded it, the 9/11 Commission had the

task of reforming one of the most complex areas of the US gov-

ernment. The 9/11 Commission and the resulting intelligence

reform legislation faced a daunting task—substantively im-

proving a process that by its nature is inherently uncertain

and imperfect without negatively affecting other intelligence

activities that, by and large, operated effectively.

This chapter first examines the findings and recommenda-

tions of the 9/11 Report. Some of the findings were unique to

the IC, but others shar

ed common features with previous re-

form efforts such as Goldwater-Nichols, particularly in areas

related to joint action, to imbalances in authority and respon-

sibility, and to inadequate or

ganizational structures. Further-

more, the commission recommended Goldwater-Nichols–like

solutions when, for example, it called for improved joint action

through the creation of a “civilian-led unified joint command

for counterterrorism”—the NCTC.

Second, this chapter describes the executive branch’s re-

sponse to the report’s recommendations. In acting on the 9/11

Report, President Bush signed three EOs that increased the

authority of the DCI over the intelligence budgeting process,

established the NCTC, and implemented many of the informa-

tion sharing recommendations of the 9/11 Report. Third, this

chapter reviews the congressional response that created new

organizational structures such as national intelligence centers

and the NCTC, drawing clear analogies to DOD structures en-

abled by Goldwater-Nichols. Furthermore, Congress estab-

lished the position of DNI and sought to further clarify the new

DNI’s budgetary and personnel authority. Fourth, this chapter

compares how the two reform efforts attempted to improve

joint action, alter the power relations among primary parties,

and improve personnel practices in both of the affected insti-

tutions. This comparison is important because the Goldwater-

Nichols Act served as a model for key aspects of the intelli-

gence reform legislation.

The 9/11 Commission Report

and Its Recommendations

The 9/11 Commission report, released on 22 July 2004, dif

fers in several ways from previous examinations of the IC. First,

the commission was formed not to examine the IC or even the

broader national security system, but instead to examine the

facts and causes related to the 9/11 attacks and to report its

“findings, conclusions, and recommendations for corrective

measures that can be taken to prevent acts of terrorism.”

Sec-

ond, the r

eport r

eads mor

e like a narrative history than a tra

ditional gover

nment commission report. The first 11 chapters

provide an extensive description of events that is most notable

for the absence of subjective judgments and for the focus on re-

porting the factual evidence. Only in the final two chapters ar

the more traditional findings and recommendations to be

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

found. Even here there are differences: most reports have taken

an approach of explicitly listing findings, followed immediately

by recommendations. The 9/11 report does not explicitly high-

light findings and instead focuses on 47 recommendations that

affect multiple aspects of government.

The 9/11 Report’s recommendations were controversial and

led to some charges that the recommendations lacked congru-

ence with the factual reporting contained in the rest of the re-

port, were overly prescriptive, and failed to account for

changes within the IC since the 9/11 attacks.

In criticizing

the report’s structural recommendations, one commentator

found that faulting systemic intelligence failures misses the

larger point that “it is almost impossible to take effective ac-

tion to prevent something that hasn’t occurred previously.”

Others critiqued the report’s failure to adequately identify the

problems in need of solutions.

In other words, the report did

not adequately explain the cause and effect of intelligence fail-

ures. Despite these and other criticisms, both Congress and

the president acted upon significant portions of the report’s

recommendations.

A review of these findings suggests one reason why the com

mission chose the Goldwater-Nichols metaphor: many of its

findings resemble the criticisms of the DOD prior to the pas-

sage of Goldwater-Nichols. The report contained five sets of

recommendations, of which the first three have elements of

commonality with DOD reform efforts and are most relevant to

the present study. They address improving interagency joint

action, improving the IC’s structure, and, somewhat less di-

rectly, improving information sharing.

Improving Joint Action

Both the pre–Goldwater-Nichols DOD and the IC prior to

9/11 suffered failures of joint action, albeit with different or-

ganizational structures contributing to these failures. The IC’s

failur

es involved lack of coor

dinated effort among multiple

agencies across multiple cabinet-level departments versus the

DOD’s failures to coordinate action among separate military

services within a single department. The 9/11 Report cited the

IC’s inability to pull together all sour

ces of information, “as-

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

sign responsibilities across the agencies,” and track and re-

solve problems.

Thus, there were two aspects of effective joint

action—intelligence and planning. On the intelligence front,

the report noted the need to pool information and centralize

analytic resources to create a more accurate analytic picture.

The 9/11 Report identified the need for more effective joint op-

erational planning as a second aspect of the issue and found

current interagency efforts within various departments/agen-

cies and on the NSC staff lacking.

The proposed solution to this problem was the development

of a new type of organizational construct: “a civilian-led uni-

fied joint command for counterterrorism,” nominally titled the

NCTC, which would oversee both joint intelligence and joint

planning.

The joint intelligence aspect of such a center would

coordinate and conduct long-term analysis, net assessments,

and warning on the terrorist threat. The joint planning aspect

was less clear, particularly as regards the relationship be-

tween planning and execution, although the report did specify

that the NCTC would not direct the execution of efforts, in-

stead leaving that aspect to the tasked agencies.

Altering Power Relations

The 9/11 Report noted that there was no “quarterback”

within the IC who “called the plays,” which created conditions

where “responsibility and accountability were diffuse.”

Such

an imbalance can imperil joint action, as chapter 2 described.

To address this situation, the report recommended the NCTC

be given hiring authority for individuals in specific counterter-

rorism organizations in the FBI, CIA, and DOD.

The 9/11 Report also considered the relative power of the

leader of the IC and listed six problems: (1) structural barriers

to performing joint intelligence work, (2) lack of common stan-

dards across the foreign-domestic divide, (3) divided manage-

ment of national intelligence capabilities, (4) weak capacity (of

the DCI) to set priorities and move r

esour

ces, (5) too many jobs

(for the DCI), and (6) an organization that was too complex and

secret.

The first problem was related to the inability to “con-

nect the dots” and failur

e to per

for

m “integrated, all-source

analysis.” The 9/11 Commission r

ecommended the creation of

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

analytic centers to tackle enduring problems such as coun-

terterrorism and counterproliferation. The report attributed

the final five problems to the DCI’s lack of authority, which

undermined the DCI’s ability to oversee the IC, establish and

enforce common rules and policies, reduce duplication, and

make necessary resource decisions. It identified the three fun-

damental authorities that a true head of the IC would require:

control of the budget, hire-and-fire authority over senior man-

agers of the IC, and the ability to set common standards and

policy.

Unlike Goldwater-Nichols in which the service chiefs re-

mained dual-hatted as both the chiefs of their respective ser-

vice and members of the JCS, the 9/11 Commission recom-

mended separating the DCI’s roles and creating a national

intelligence director (NID) rather than creating other bur

eau-

cratic incentives to balance the competing interests of leaders

serving in two roles. Under the recommendations, the NID

would oversee multiple intelligence centers created in the vein

of the NCTC; would be the president’s principal intelligence

advisor; would plan, program, and budget r

esources for na-

tional intelligence; and would establish policies and standar

for the IC.

Improving Information Sharing

The third set of recommendations concerned removing the

barriers to information sharing. While DOD reform in this area

primarily focused on improving interoperability among sys-

tems, the 9/11 Report identified both “human and systemic

resistance to sharing information.”

Here the commission

noted that cultural challenges strongly favored protecting in-

formation (including sources and methods) over sharing infor-

mation. The commission found current practices “nurture

overclassification and excessive compartmentation,” with in-

centive structures strongly favoring protecting information at

the expense of sharing.

The recommendations included increased incentives to pro-

mote the need to share over the need to know and to imple-

ment better infor

mation technology strategies to better facili

tate infor

mation sharing.

The r

eport left unsaid how

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

incentives would be structured and implemented to encourage

a cultural shift favoring information sharing. It also did not

discuss whether technical constraints might exist that would

inhibit rapid implementation.

Other Issues

The final two major sets of recommendations related to im-

proving congressional oversight of intelligence and improving

domestic defense efforts. The report recommended that Con-

ess consolidate its oversight committees and structures for

both intelligence and homeland security, citing multiple juris-

dictional issues that inhibited proper “oversight, support and

leadership.”

A final set of recommendations focused on the FBI, the home-

land defense missions of US Northern Command (USNORTH-

COM), and the DHS. The report did not r

ecommend cr

eating a

domestic intelligence agency or passing legislation to improve

FBI counterterrorism intelligence efforts; instead, it recom-

mended that the FBI continue the internal reform efforts begun

after the attacks.

The r

eport was silent on why one of the or-

ganizations with a major role in the events preceding the attacks

should rely on internal reform efforts whereas the majority of

the other intelligence or

ganizations r

equired external, congres-

sional action to ensure long-term reform.

Executive Branch Response

to the 9/11 Report

Unlike most previous reform efforts and investigations, the

9/11 Report and its recommendations galvanized both Con-

gress and the president to respond quickly and substantively.

The executive branch released three EOs on 27 August 2004

that implemented key aspects of the 9/11 Report’s first three

sets of recommendations.

Improving Joint Action

Addr

essing the failur

e of joint action, EO 13354, National

Counterterr

orism Center

, 27 August 2004, established the

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

NCTC. In doing so the president, using language remarkably

similar to that in the 9/11 Report, granted the center the au-

thority to “conduct strategic operational planning for counter-

terrorism” and to serve as the primary organization for the

analysis and integration of terrorism information.

The EO

echoed the 9/11 Report’s recommendation and prohibited the

NCTC from directing actual operations.

Because of statutory limitations, the president could not

create a DNI. The EO instead placed the NCTC under the con-

trol of the DCI. To preclude information sharing difficulties,

the EO required agency heads to provide the NCTC access to

all terrorist data and to provide personnel, funding, and other

resources “after consultation with the head of the agency and

with the approval of the director of the Office of Management

and Budget.”

Altering Power Relations

The next EO, 13355, Strengthened Management of the In-

telligence Community, 27 August 2004, amended EO 12333,

United States Intelligence Activities, 4 December 1981, and

addr

essed the r

eport’s findings on the DCI’s relative lack of

power by strengthening the DCI’s budgetary authority. In lan

guage similar to that of Goldwater-Nichols, the or

der made the

DCI the “principal advisor to the president for intelligence mat-

ters” rather than the “primary advisor.”

The order added new

provisions requiring the DCI to emphasize information sharing

practices and to develop new standards to create interoper-

ability among agency information systems, and specified that

countering terrorism was “a matter of the highest national se-

curity priority.”

The order addressed the foreign-domestic in-

telligence divide that the 9/11 Commission had indicated im-

peded information sharing by tasking the DCI to manage

collection activities “whether information is to be collected in-

side or outside the United States.”

On the budgetary fr

ont, the executive or

der became aligned

with existing law and tasked the DCI to participate in the de-

velopment of the annual budgets for DOD’s military intelli-

gence pr

ograms—the joint military intelligence pr

ogram (JMIP)

and the tactical intelligence and r

elated activities (TIARA) pro-

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

gram. More significantly, the order gave the DCI the authority

to transfer funds within the NFIP with the approval of the di-

rector of the OMB and with no mention of a requirement for

concurrence from either an agency or department head.

The

ability to transfer NFIP funds without departmental concur-

rence represented a significant increase in the DCI’s authority

that would later be reduced by legislation. The DCI’s authority

was further strengthened by requiring DCI concurrence in the

appointment of the heads of any intelligence organization

within the IC.

Finally, hearkening back to Goldwater-Nichols

joint personnel practices, the EO required the DCI to establish

“an integrated professional development and education sys-

tem” to encourage service in multiple IC organizations and

“make such rotated service a factor to be considered for pro-

motion to senior positions” within the IC.

Improving Information Sharing

The third EO, 13356, specified new policies and responsibil-

ities to impr

ove infor

mation sharing. The or

der requir

ed agen-

cies to design information systems to allow the interchange of

infor

mation among agencies and state and local gover

ments.

It also requir

ed the DCI to set standar

ds for sharing

terrorism-r

elated information and suggested that the stan-

dards include many of the methods included in the 9/11 Re-

port, such as (1) creating unclassified versions of products

whenever possible by removing information on sources and

methods, (2) creating products at multiple classification levels,

(3) removing originator controls from products, (4) minimizing

the use of compartmentalization, and (5) increasing personnel

incentives for sharing information.

The order also created an

information systems council, with representatives from all the

departments with counterterrorism missions, to oversee the

establishment of an “information sharing environment” with

the apparent intent of improving the interoperability of auto-

mated systems.

These three EOs went quite far toward implementing many of

the 9/11 Report’s recommendations. In a different political en-

vir

onment such changes might have been suf

ficient; however

for a variety of r

easons Congress also felt compelled to act.

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

Legislative Branch Response

to the 9/11 Report

The legislative branch also reacted quickly. On 30 July

2004, only eight days after the report’s release, Congress held

its first of approximately 30 hearings on the 9/11 Commis-

sion’s findings.

Multiple committees, including both cham-

bers’ intelligence, judiciary, and armed services committees

and the Senate Governmental Affairs Committee, impaneled

expert witnesses to cover multiple aspects of the report and its

recommendations.

On 23 September the Senate Governmental Affairs Commit-

tee introduced intelligence reform legislation, which the Sen-

ate passed on 6 October

The House r

esponded with its own

bill on 9 September and a revised bill on 24 September that

passed on 8 October.

The two chambers then attempted to

reconcile their versions of the bill prior to the November elec-

tions but did not succeed in doing so despite heavy pressure

om for

mer 9/11 Commission members and victims’ family

advocacy gr

oups. After the election, Pr

esident Bush signaled

that passing intelligence reform legislation remained a prior-

ity. Finally, after a series of compr

omises, Congress passed the

reform legislation, and President Bush signed the Intelligence

Reform and Terrorism Prevention Act of 2004 into law on 17

December 2004.

Some Possible Reasons for Success

Why this reform effort succeeded where so many prior ones

had failed is open to debate. Multiple reasons probably con-

tributed to the passage of the intelligence reform bill, includ-

ing the impact the attacks had on the nation’s consciousness,

the continued relevance of the threat, lobbying by relatives of

the 9/11 victims and commission members, the support of the

president, and proximity to the presidential election.

Often, as in the case of the Church and Pike investigations

and Rockefeller Commission, studies are conducted in response

to a discrete set of events. By the time the investigations con-

clude, often years after the initial impetus, the same events may

no longer be as salient, may have been overtaken by mor

ess

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

ing matters, or may have been addressed through a combina-

tion of internal reforms and executive branch activity that ob-

viated the need for more concrete legislative action. Another,

decidedly more cynical, view would suggest that some commis-

sions are formed to give the appearance of action where none is

intended. Other reform efforts, such as Boren-McCurdy,

IC21,

and Aspin-Brown, produced valid findings and recommenda-

tions, but in the absence of agreed-upon and compelling rea-

sons for change. Senator Jay Rockefeller, in discussing why

previous recommendations had not been enacted, stated,

“There was no sort of galvanizing event or series of events and

the will, therefore, in the Congress, joining with the adminis-

tration, never came to be.”

In other words, with no major in-

terest group advocating reform and no notable recent failures,

the will to effect change was significantly undermined.

By contrast, although the 9/11 Report was released nearly

three years after the attacks, the reasons for reform remained

compelling after the commission had completed its work.

Less clear is why the executive branch response did not short-

circuit legislative action, as happened in pr

evious cases.

One

possible explanation is that the executive branch was far mor

supportive of legislative action. Furthermore, in contrast to

many intelligence issues, a strong victims’ advocacy group lob-

bied members of Congress and worked through the media,

further ensuring that the issue stayed in the public’s con-

sciousness.

Legislative Foundations and Focus

The nature of the legislative changes represents an amalga-

mation of earlier intelligence reform recommendations com-

bined with aspects of the 9/11 Commission’s recommenda-

tions. Senator Lieberman, in remarks on the Senate floor,

confirmed that the legislation was influenced by multiple

sources, including the 9/11 Report, the joint investigation by

the House and Senate intelligence committees, and prior stud

ies dating back 50 years.

The legislation added Goldwater-

Nichols–like initiatives with an eye toward improving the IC.

Employing Goldwater

-Nichols–like r

efor

ms and language was

decidedly intentional. The 9/11 Report, 9/11 commissioners

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

in prepared testimony, the Senate report accompanying the

legislation, and congressional statements all attested to the

relevance of implementing similar reforms for the IC.

Thus

the foundations of the legislation were a mixture of recommen-

dations from earlier reform studies, new executive orders, rel-

evant reform measures from Goldwater-Nichols, and the rec-

ommendations of the 9/11 Commission.

Although much of the press and reporting on the legislation

focused on the reform of the IC, only about one-fourth of the leg-

islation (58 pages out of the total 235) is devoted to intelligence,

with the remainder addressing such topics as FBI organization,

transportation security, border protection and immigration, and

terrorism prevention. Topics related to the IC included many of

the core recommendations of the 9/11 Commission as well as

legislation to improve the quality of analysis, improve IC person

nel practices, and create a privacy and civil liberties oversight

board. The most notable recommendations not addressed by

Congress were those related to congr

essional oversight, thus ig-

noring, for the time being, the requests of commission members

not to treat the r

ecommendations as a “Chinese menu” and pick

and choose which r

ecommendations to implement.

Improving Joint Action

The 9/11 Commission described two aspects of “joint ac-

tion”: intelligence and planning. The legislation codified these

aspects through two related organizational constructs: national

intelligence centers and national centers as embodied by the

NCTC and the National Counterproliferation Center (NCPC).

This distinction is not readily apparent without a close reading

of the legislation and a review of the colloquy between Senators

Joseph Lieberman (D-CT) and Susan Collins (R-ME) found in

the

Congressional Record.

Senator Lieberman clearly defined

the intent for national intelligence centers as addressing only

the intelligence aspect of joint action; in other words, they have

no planning function. The r

esponsibilities of these centers,

analogous to those of DOD unified commands, are to “provide

all-source analysis of intelligence, identify and propose to the

DNI intelligence collection and analysis r

equir

ements,” and to

ovide net assessments and warnings.

Thus, in theory, cen

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

tralizing, analyzing, and acting upon information from across

agencies in a coordinated fashion as well as establishing center-

related tasking (collection) requirements improves joint action.

In this respect, the national intelligence centers appear to offer

a rational construct for such centralization. As a central reposi-

tory of information for analysis and reporting, the center con-

struct is assuming, partially, one of the original functions of the

CIA—to centralize intelligence analysis and production using

intelligence provided by the separate organizations of the IC.

Congress is clearly hoping to avoid some of the initial growing

pains experienced by the nascent CIA by increasing the au-

thority of the DNI.

The NCTC and the NCPC embody a second, slightly different

organizational construct. These organizations, by virtue of

their strategic operational planning mission, have both intelli-

gence and planning functions, making them roughly equiva-

lent to expanded national intelligence centers.

The legisla-

tion defined this planning as including “the mission, objectives

to be achieved, tasks to be performed, interagency coordina-

tion of operational activities, and the assignment of roles and

responsibilities.”

Senator Lieberman likened the planning di-

rectorates of these new organizations to the J3 on the joint

staff and described how such a planning process would work

by comparing it to lanes in a highway.

In essence, the NCTC

would choose “the lane of the highway”; that is, it would des-

ignate the executing agency and the type of activity (e.g., diplo-

matic, military, law enforcement, etc.). At the same time, the

designated agency would determine how to accomplish the as-

signed task (i.e., how to drive in the lane). The joint planning

aspect of these expanded centers thus augments or potentially

even replaces the current interagency process.

Altering P

ower Relations

Congress was far more detailed and specific in addressing the

9/11 Report’s second set of recommendations concerning the

establishment of a DNI with greater authority. The legislation

effectively eliminated debate concerning the dual roles of the

DCI by cr

eating the position of DNI and attempted to clarify

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

and expand the new DNI’s authority over budget, personnel,

and tasking relative to the former DCI’s.

The budgetary authorities involve three different parts of the

budgeting process: determining and developing the annual

budget, executing the budget, and transferring/reprogramming

funds. First, the act grants the DNI the ability to “develop and de-

termine” the national intelligence program (NIP) budget versus

the previous terminology in the National Security Act of “develop

and present.”

According to Senator Collins, a cosponsor of the

final bill, the word determine is essential, as it implies that “the

DNI is the decision maker regarding the budget and does not

shar

e this authority with any department head.”

Even within

Congress there remains some disagreement as to whether this

actually represents substantial new authority.

The DNI’s role in

the DOD’s development of JMIP and TIARA budgets remains the

same—participatory not directive.

Second, the act appears to provide greater budget execution

authority by requiring the OMB to apportion funds at the

DNI’s “exclusive dir

ection.”

The DNI may also conduct audits

and evaluation of elements of the IC to monitor compliance.

The legislation clearly expands the DNI’s execution authorities

beyond the DCI’s previous authorities; however, less clear is

the real impact of these authorities, as Congr

ess plays the key

role in appropriating funds.

In the third area of budgetary authority, the legislation in-

eases the DNI’s transfer and reprogramming authority over

the NIP; however, somewhat ironically, the limitations im-

posed on this authority actually represent a decrease in au-

thority from that provided by EO 13355, which gave the DCI

unlimited authority to transfer and reprogram national intelli-

gence funds between departments and agencies.

The new

legislation imposes limitations on such transfers to $150 mil-

lion or 5 percent of a department’s overall NIP funds.

The act provides slightly greater authorities with respect to

personnel transfers by allowing the transfers for up to two

years, up fr

om the previous one-year limit, and no longer re-

quires the department head’s concurrence; however, the

transfers will be limited by futur

e procedures agreed to by the

DNI and the heads of departments.

The legislation r

equir

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

the DNI to concur in the appointment of the heads of various

organizations within the IC and prohibits a department head

from filling the position or recommending to the president an

individual who does not have the DNI’s concurrence.

The leg-

islation also authorizes the transfer of up to 100 IC personnel

to national intelligence centers as they are established. The

tasking authorities of the DNI remain largely the same as the

DCI’s with greater amplification of the DNI’s authority to

“manage and direct the tasking of, collection, analysis, pro-

duction and dissemination of national intelligence by elements

of the IC.”

The legislation eliminates the NSC’s role in arbi-

trating the DNI’s access to intelligence and grants the presi-

dent sole authority to make access decisions.

Quality of Joint Personnel

The legislation changes personnel practices within the IC—a

topic the 9/11 Commission only partially addressed.

In this

case, Goldwater-Nichols served as a direct model, as the intelli

gence legislation r

equires the DNI to create mechanisms that

“seek to duplicate for civilian personnel within the IC the joint of-

ficer management policies” established by the Goldwater-Nichols

Act.

Just as Goldwater

-Nichols recognized the need to

strengthen the quality of personnel serving in joint duty as-

signments, the intelligence legislation requir

es the DNI to

make similar adjustments to improve the quality of personnel

serving in key positions in the IC. The act requires the DNI to

develop improved incentives (left unspecified) for personnel

serving on the DNI staff, the national intelligence centers, or

in other IC management positions.

Much as joint service is

a prerequisite for promotion to the general officer ranks, the

act requires service in more than one IC organization as a

prerequisite for promotion to certain positions within the IC.

The legislation does not specify either the specific positions

or the number of positions that should carry this joint re-

quirement. The act also requires the DNI to “set standards

for education, training, and career development of personnel

of the IC.”

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

Information Sharing

The act addresses information sharing by directing the DNI

to establish uniform security standards and procedures, im-

prove information technology interoperability among IC organi-

zations, and develop policies “to resolve conflicts between the

need to share intelligence information and the need to protect

intelligence sources and methods.”

In addition, the act re-

quires the DNI to provide an annual report of any “statute,

regulation, policy, or practices that the Director believes im-

pedes the ability of the Director” to implement the information

sharing requirements spelled out in the act.

Section 1016 of the act further addr

esses 9/11 Commis

sion recommendations related to information sharing by di-

recting the establishment of an information sharing envi-

ronment (ISE) “for sharing terrorism information among all

appr

opriate Federal, State, local, and tribal entities, and the

private sector” through a mixtur

e of procedural guidelines

and technologies.

The act directs the president to estab-

lish guidelines for sharing information in “its most share-

able form, such as the use of tearlines to separate out data

from the sources and methods by which the data are ob-

tained.”

To ensure that progress is made towards imple-

menting the ISE, the act directs the president to provide a

report to Congress specifying, among other items, the de-

tails of the ISE, cost estimates to fully implement the ISE,

metrics and enforcement mechanisms, and training require-

ments.

The act also establishes a program manager for the

ISE and an information sharing council to oversee imple-

mentation of the ISE and recommend necessary technical or

procedural changes. Finally, the act directs the president to

submit annual reports on the state of the ISE and informa-

tion sharing activities across the government.

This annual

eport appears designed to pr

ovide executive accountability

in the creation and maintenance of the ISE, as well as de-

scriptions of the extent of information sharing not only across

the federal government but with state, local, and tribal offi-

cials as well.

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

Preliminary Legislative Analysis

Congress unquestionably intended to incorporate Goldwater-

Nichols–like reforms and language in the intelligence reform

legislation. One approach to estimating the potential effective-

ness of the intelligence reform legislation is to compare the ap-

plicable measures in the legislation to those in the Goldwater-

Nichols Act. Such an examination may provide a better

appreciation of how the two pieces of legislation approached

somewhat similar issues.

The use of multiple Goldwater-Nichols analogies by the

9/11 Commission and the ensuing legislation highlights the

relevance of such a comparison. Important aspects include

how each reform effort attempted to improve joint action, alter

the power relations among the primary parties, and impr

ove

personnel practices. Another area—improved information

sharing—is less applicable to Goldwater-Nichols; therefore,

the following section examines potential difficulties with im

plementing the various information sharing objectives without

refer

ence to DOD reform.

Improving J

oint Action

In some respects comparison between the DOD and the IC

may not be appr

opriate, as the concept of

joint action has dif

ferent meanings in the different contexts. Within the DOD,

joint action is associated with the planning and execution of

military operations by two or more military services. In the

context of the 9/11 Report and the ensuing legislation, joint

action involved drawing relevant intelligence from across the

government, requiring greater cooperation among intelligence

agencies, and creating a single “quarterback” to assign roles

and responsibilities.

Goldwater-Nichols attributed much of the failure of military

joint action to the r

elatively weak authority of the combatant

commanders. As described in chapter 2, the act addressed

this deficiency by vesting greater authority in the combatant

commanders, while maintaining the train, organize, and equip

functions within the services and giving the combatant com

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

manders relatively little authority over the military services’

planning, programming, and budgeting activities.

Congress created two new organizational constructs to ad-

dress two different aspects of joint action: intelligence and

planning. It drew an analogy between the intelligence con-

structs (national intelligence centers) and DOD unified com-

mands and the planning ones (e.g., the planning directorate of

the NCTC) and the J3 on the joint staff. While Congress’s in-

tent was clear, there are important differences between these

new organization constructs and the DOD organizational con-

structs they seek to partially replicate.

Likening national intelligence centers to unified commands

is reasonable if the analogy is meant to describe an organiza-

tional construct that seeks to bring together multiple semi-

independent organizations under a single leader. The underly-

ing assumption is that this single leader would more effectively

coordinate and conduct joint activities. The advantage of such

a construct is that it focuses responsibility and accountability

on a single individual. For national intelligence centers, mor

effective joint action implies greater information sharing, en-

hanced thr

eat warning, and mor

e accurate and complete analy-

sis. Under these assumptions the center construct appears to

offer a r

easonable alternative to impr

ove the quality of intelli

gence in specifically designated areas. However, the center di-

rectors face at least two potential difficulties.

First, a center dir

ector’s control over assigned personnel is

less clear than that of a unified commander. Goldwater-

Nichols granted combatant commanders greater control over

subordinate personnel, including full hire-and-fire authority,

disciplinary authority, and the ability to assess individual

performance (a key component of future promotions). Apply-

ing the lessons of DOD reform suggests the positive benefit of

permanently assigning personnel to national intelligence cen-

ters versus relying on loans or details of personnel for tempo-

rary periods. Per

manent assignments pr

omote stability and

institutional lear

ning, reduce training and orientation costs,

and grant the center supervisors the ability to assess per-

for

mance and thereby potentially influence future promo-

tions.

o be sur

e, many workers within the centers would

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

likely fully support center efforts; however, knowing that the

owning center would be writing performance appraisals

would add incentives to accord greater allegiance and effort to

the working center.

Second, the centers will likely be more heterogeneous than

similar military organizations. Within the military, each of the

services maintains a unique culture; however, at a more macro

level, military members across the services share a more generic

culture associated with the military profession of arms. The IC

has multiple cultural dividing points—military versus civilian,

analytic versus collection mind-sets, and departmental mission-

related cultures (e.g., DOD versus Department of State, FBI

versus CIA). Creating a shared culture, while beyond the scope

of the legislation, is nonetheless often considered an essential

ingredient in promoting joint action and may become one of

the biggest challenges to successful implementation.

In sum,

it appears that centers will have less authority over personnel

assigned, and face potentially greater cultural challenges than

the DOD faced after Goldwater

-Nichols.

The second aspect of joint action, that is, planning, repre-

sents a mor

e difficult comparison with existing DOD struc

tures. The analogy of the NCTC’s planning directorate to the

joint staff J3 is pr

oblematic because the J3’s role within the

DOD decidedly differs from what intelligence reformers envi-

sioned for the NCTC. In the DOD model, the vast majority of

planning is conducted by a combatant commander’s or a des

ignated JFC’s staff and executed through the combatant arms

also assigned to or controlled by the same commander.

The

role of the joint staff (and the services and some DOD agen-

cies) is to review the plans.

In contrast, based upon congres-

sional testimony, it appears Congress expects the NCTC to

play a more hands-on role in planning than simply reviewing

plans developed by other agencies or departments.

The task of joint planning represents a considerable depar-

tur

e fr

om the DOD model that the r

ecommendation purports

to emulate. It appears that the NCTC will initially have less au

thority over personnel assigned, have significantly less control

over the conduct of operations, and will face gr

eater cultural

challenges than the DOD faced after Goldwater

-Nichols. Joint

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

action in the DOD was improved because the joint organiza-

tions both planned and executed operations, with forces di-

rectly assigned to a single organization. After Goldwater-

Nichols, there were often disagreements in the conduct of

operations and the proper use of service forces among joint

force commanders and their component commanders; how-

ever, invariably, the services were forced to accede to the joint

commander’s desires.

Because the NCTC conducts planning of operations and not

their execution, there is a looser coupling between planning and

execution than currently exists within the DOD. This could pro-

duce incr

eased frictional delays in both the planning and exe-

cution of proposed operations.

Also, the legislation does not

address the nature and specificity of the plans or a depart-

ment’s or agency’s responsibilities in implementing the plans if

they disagree with portions of the plans or roles assigned. Be-

cause the center director has no control over operational ele-

ments, the overall power of the director to enforce the execution

of plans is less clear

. For example, if USNORTHCOM received

concurrent tasking requiring the use of similar resources from

the NCTC and DOD, who would r

esolve the conflict—the NSC,

the president, or some type of existing or new interagency

process? While not necessarily insur

mountable, these issues

would likely seriously complicate implementation.

In all likelihood, successful planning will require coopera-

tion among departments and agencies for successful execu

tion. It might also require yet another interagency construct—

a scenario that runs counter to the 9/11 Report, which

suggests officials should not have to overcome structural or

bureaucratic obstacles in the performance of missions.

Recognizing the potential for disagreement, Congress estab-

lished an appeals process whereby the DNI would resolve dis-

putes between the NCTC and department heads, leaving the

department or agency head the opportunity to appeal the de-

cision to the pr

esident. The pr

ocess for r

esolution and appeals

of conflicts with the individual departments r

elies on the DNI’s

power vis-à-vis the department heads. If the DNI maintains a

str

ong personal relationship with the president, there is a rea-

sonable chance the appeals pr

ocess could work; if not, the

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

NCTC’s ability to plan operations could be significantly under-

mined.

Moreover, this process places the president, who has

decidedly limited time, in a managerial position of having to

resolve disputes on resource and planning issues.

In some ways the planning function more closely resembles

the DOD prior to Goldwater-Nichols, with the combatant com-

mander in that era relying on the cooperation of the service ele-

ments to carry out operations and having reduced authority

control the conduct of those operations.

This suggests

that the NCTC may not be the best construct to combine ele

ments of planning and operations. Instead, an interagency

construct created in the vein of Goldwater-Nichols would pro-

vide the leader of such an organization greater control over the

forces and personnel charged with executing the developed

plans.

In addition a more flexible construct may be more ap-

propriate than permanent planning organizations like the

NCTC. Under this more flexible construct, notionally termed a

joint interagency task force (JITF), the department or organi-

zation with the greatest resources supporting the operation

would lead the JITF, with other organizations providing neces-

sary support as required. To follow the DOD analogy, an ini-

tial interagency process would establish a JITF and assign a

commander based upon the natur

e of the intended opera-

tions.

Thus if the operation is envisioned to be primarily a

law-enforcement operation, the JITF commander would come

om DOJ, a predominately military operation would have a

DOD commander, and so forth. The interagency process would

then give the JITF commander operational control of personnel

necessary to accomplish the mission. Thus departments and

government organizations would find themselves in various

supporting and supported r

elationships, depending upon the

overall natur

e of the task.

Further

mor

e, the JITF appr

oach

could be applied acr

oss multiple mission ar

eas—not just

counterterr

orism. Under such a construct the national intelli

gence centers would be r

oughly analogous to the joint intelli-

gence centers within the unified commands by providing intel-

ligence support to JITFs but not planning or overseeing the

execution of operations.

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

Altering Power Relations

When different parties represent competing tensions, one way

to recalibrate the underlying balance is to alter the power rela-

tionship of the parties.

The 9/11 report specified three formal

authorities essential for the head of any agency: budgetary au-

thority, hire/fire authority, and the ability to set standards.

Al-

tering the distribution of these authorities represents one way to

adjust a power relationship more formally.

Thus, by more

clearly specifying the authorities of principal office holders, Con-

gress, through legislation, or the president, through EO, can in-

crease the legitimate authority of an agency head. Another option

is to use more informal means, such as altering the r

elative sta-

tus of the parties. In this way, the actual or perceived relation-

ship with the president can increase the status or prestige of the

DCI vis-à-vis the competing parties.

In the Goldwater-Nichols example, Congress adjusted the

balance of power between the JCS chairman and the service

chiefs. The legislation incr

eased the chairman’s informal

power by making the CJCS the primary advisor to the presi-

dent, thereby in effect increasing the chairman’s status vis-à-

vis the service chiefs. In addition, the joint staff was assigned

to work for the chairman, thus removing much of the services’

control over the staff. Notably, the chairman gained little for-

mal authority in ter

ms of incr

eased power over service bud-

gets or personnel. This suggests that increases in formal au-

thority are not always necessary to alter power relationships.

The crafters of intelligence reform recognized a need to in-

crease the power of the DNI relative to two key sets of actors:

the various component organizations of the IC and the depart-

ments housing intelligence organizations. To that end, the leg-

islation focused on increasing both the formal and informal

aspects of the DNI’s authority.

The formal aspects of the DNI’s authority—budgetary, per-

sonnel, and tasking—should enable the director to exercise

mor

e complete and ef

fective contr

ol over the IC. The DNI has

many cabinet-level r

esponsibilities, such as planning, pr

gramming, and budgeting for major systems; establishment of

unifor

m policies across the enterprise; and providing broad

planning guidance.

However

, with the exception of the CIA,

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

the NCTC, and community management staff, no intelligence

organizations report directly to the DNI. Thus, unlike a depart-

ment head, the DNI is tasked to lead and coordinate organiza-

tions that are dispersed throughout the executive branch,

most of which report to other cabinet-level officials, further

complicating the DNI’s ability to manage.

The DNI maintains

hiring authority in conjunction with department heads and

has some limited abilities to shift personnel within the com-

munity; however, they are considerably reduced when com-

pared to those of a department head.

While the DNI’s formal powers are essential, a less formal

aspect of the position’s power, namely the relationship the DNI

maintains with the president, may be just as important. Cer-

tainly in the case of DOD reform the increased prestige of the

CJCS was a critical aspect of the chairman’s power base.

The same may be the case for the DNI, as the position’s formal

authority is no guarantee against disputes with the department

heads on a variety of issues ranging from programmatic deci-

sions to resource allocation to intelligence support for depart-

mental requir

ements. This is because the IC exists to support

a variety of gover

nmental customers. It is an interagency orga

nization for a reason: multiple government organizations with

diverse missions need intelligence support to better accom-

plish their assigned tasks. Therefore, department heads retain

the ability to appeal DNI decisions to the president. In these

instances, formal authority is less relevant and the status of

the DNI relative to the department head (i.e., who has the

stronger relationship with the president) may become the de-

ciding factor. This does not imply that the formal authorities

of the DNI are insignificant, but it suggests that the DNI’s

power base has multiple components, some of which are r

ela-

tively immune from legislative action.

Quality of Joint Personnel

In many r

espects the intelligence r

eform legislation most

closely mirrors Goldwater-Nichols reform in the area of joint

personnel practices. Goldwater-Nichols addressed the poor

quality of joint personnel by mandating that personnel serving

in joint positions be pr

omoted at rates at least equivalent to

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

service promotion rates, requiring completion of joint duty

prior to promotion to the general officer corps, increasing the

number of joint authorizations, and lengthening the minimum

duration of joint assignments. The act also levied congres-

sional reporting requirements on the DOD to detail how suc-

cessful the department was in meeting the requirements laid

out by the act.

Instead of providing similarly detailed requirements, the in-

telligence reform legislation directs the DNI to establish mecha-

nisms comparable to those of the Goldwater-Nichols Act. The

legislation directs equivalent promotion rates for intelligence

personnel serving in select positions and the use of incentives

to encourage joint duty.

It requires the DNI to establish

mechanisms to rotate select IC personnel throughout the com-

munity to encourage breadth and allows the DNI to determine

positions where the completion of a joint duty assignment is

required prior to promotion; however, the number and grade

level are left unspecified. Finally, the act does not specify the

minimum duration of joint assignments.

The success of replicating the DOD’s joint duty practices

rests almost entirely with the DNI and his or her staf

f. Track-

ing the promotion rates of personnel filling joint positions is

relatively easy; it is the other portions of the DOD’s joint duty

practices that should not be ignored. If Goldwater-Nichols is

to serve as an example, then the keys to success include des-

ignation of a lar

ge number of senior positions requiring joint

duty experience and specificity in policy describing joint per-

sonnel practices. To ensure a broader infusion of a joint cul-

ture, prudence would suggest that more senior positions

within the IC require joint experience for promotion rather

than fewer. The requirement that every general officer within

the military have joint experience has undoubtedly con-

tributed to a greater appreciation of the other services’ capa-

bilities and engendered a more joint perspective. Another po-

tentially important element of success is pr

escribing minimum

assignment length for joint cr

edit to prevent creating a revolv-

ing door with various joint duty assignments where individu-

als fill positions for short periods of time to r

eceive joint credit.

Finally, the DOD cr

eated joint educational opportunities to

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

further standardize the training of military personnel in joint

matters. Currently, within the IC there is no similar type of

standardized training on how to be a professional intelligence

officer.

Finally, the number of departments involved complicates es-

tablishing common policies and procedures. Congress recog-

nized this difficulty in certain instances. For example, the intel-

ligence reform act requires the establishment of uniform

policies related to personnel security and requires the president

to select a single agency to perform required security investiga-

tions.

The act also requires more standardization in analytic

tradecraft and quality across the IC.

However, if standardiza-

tion across the community is the objective, significant addi-

tional areas remain in need of policy congruence, such as train-

ing, education, and personnel practices to include hiring, firing,

and promotion. The mixture of military and civilian personnel

who comprise the IC further complicates the task, making stan-

dardization in areas such as promotion, career incentives, and

even training and education more difficult than it would be

within a single department or organization. A further limiting

factor is the power of the DNI to enforce additional standardiza-

tion not mandated by law. In such instances, the DNI may have

to rely on the cooperation of the department heads to achieve

the sought-after policy congruence.

Improved Information Sharing

Nothing in the Goldwater-Nichols Act directly parallels the

mandate to improve information sharing in the IC. The closest

resemblance is in the act’s focus on interoperability of systems

among the services. The scope of intelligence reform spanned

human as well as technical systems and therefore involved a

wider-ranging series of mandates.

Taken in total, the information sharing section of the act im-

poses some of the most detailed requirements on the executive

branch and potentially may be the most difficult to implement

considering the diversity of information infrastructures the

ISE is attempting to consolidate. Compounding the difficulty

will be dif

fering security clearance levels that ISE participants

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

will likely have, particularly when the sharing environment is

extended to the state and local levels.

Another obstacle that may arise is assessing the type of in-

formation to be shared. Although the act defines

terrorism in-

formation

, determining whether individual pieces of informa-

tion are terrorism related may not always be easy. Data that

at first appear irrelevant may in a larger context have impor-

tant terrorism-related implications. In the absence of sharing

all information among all agencies (an even greater logistical

hurdle), this may further complicate implementation.

The required annual report may, over time, become a signif-

icant tool to allow the DNI to highlight either bureaucratic in-

transigence or statutory limitations. In the event of an intelli-

gence failure in the future, the report may also limit the ability

of the DNI or a center director to claim lack of intelligence from

one or more organizations in the IC as a factor in the failure.

Thus, the annual report could offer a powerful incentive for

greater information sharing in the future.

Other difficulties could potentially include dealing with

problems of information overload, training analysts to recog-

nize and use different sources and types of information, and

training all staff to use new IT systems and pr

ograms.

Such

issues could significantly complicate implementation. Ther

are also technical and procedural challenges. Although the

DOD has made significant progress, C3 infrastructures among

the military services are still not completely inter

operable.

the DOD has not yet fully succeeded, the likelihood of creating

a standardized data-sharing and communications architec-

ture among the intelligence agencies is remote in the near

term and would likely face significant monetary, technical,

and bureaucratic impediments.

Summary

It is far too early to determine whether the recently passed

intelligence reform will have a similar impact on the IC as

Goldwater-Nichols had on the DOD. Establishment of the

NCTC and the corresponding centralization of personnel and

esources may improve elements of joint action, particularly

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

those elements related to terrorism. Additional information

sharing measures incorporated in the legislation should also

assist in this effort. The effectiveness of future operational

planning is uncertain, given the relative lack of authority of

both the NCTC director and the DNI over the departments and

agencies they are trying to direct.

With the exception of the NCTC, the legislation did not funda-

mentally alter the structure of the IC. Instead, it split the two

roles of the former DCI—head of the IC and director of the CIA—

and reassigned the personnel supporting each of the positions,

as required. There was some increase in personnel to support

community management efforts, but the general effect was to

move the DNI up one level in the IC hierarchy and remove the

CIA fr

om its position of pr

eeminence within the IC and place it

on a more equal footing with the rest of the intelligence agen-

cies. While Congress increased the DNI’s budgetary authority,

the DNI still possesses less authority than the average depart-

ment head and will likely have to continue to rely as much on

interpersonal relationships and cooperation with department

and agency heads as on budgetary authorities.

Finally, the

priority the DNI places on impr

oving both the quality and num-

ber of joint personnel with broad IC experience will determine

the ef

fectiveness of these efforts, since Congress left the bulk of

the implementation details in the DNI’s hands.

Although the reform legislation resembles Goldwater-Nichols,

it does not r

eplicate it, and its ef

fects may be less profound. Aside

from the obvious fact that the IC is fundamentally different from

the DOD in terms of organization, personnel, resources, mis-

sions, and culture, the underlying tensions at play inhibited pre-

vious intelligence reform efforts, limited the current reform ef-

fort, and are likely to have continuing effects on the IC. The next

chapter examines some of these tensions.

Notes

1. 9/11 Report, 403 (see chap. 1, n. 3).

Intelligence Authorization Act for Fiscal Year 2003, sec. 602 (see chap.

3, n. 108).

3. Richard A. Falkenrath, “The 9/11 Commission Report, a Review

Essay,”

International Security 29, no. 3 (Winter 2004/05): 181–82, 188.

Falkenrath states, “In short, while r

eor

ganizing the intelligence community

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

ay ultimately prove to do more good than harm, the history of 9/11 as told

by the commission does not make the case that such a reform is necessary

or even necessarily beneficial.” See also, Richard A. Posner, “The 9/11 Re-

port: A Dissent,”

New York Times, 29 August 2004, sec. 7, 1.

4. Posner, “The 9/11 Report: A Dissent,” 2.

5. James Locher, interviewed by the author, 2 February 2005. According to

Locher, a prime architect of the Goldwater-Nichols reform, determining the

problems and their underlying causes is the critical task of any reform effort,

an area where he did not see sufficient emphasis in the 9/11 Report. He stated

that the Senate Armed Services Committee staff spent nearly 18 months work-

ing this aspect of the DOD reform effort. See also Center for Strategic and

International Studies, “Guiding Principles for Intelligence Reform,” 21 Sep-

tember 2004, http://www.csis.org/0409_intelreformprinciples.pdf (accessed

31 March 2005).

6. 9/11 Report, 400 (see chap. 1, n. 3).

7. Ibid., 402.

8. Ibid., 403. Both the center construct and analogy to a unified command

have been suggested prior to the 9/11 Report. See Larry C. Kindsvater, “The

Need to Reorganize the Intelligence Community,”

Studies in Intelligence 47, no.

1, 2003, http://www.cia.gov/csi/studies/vol47no1/index.html (accessed 15

January 2005).

9. 9/11 Report, 404 (see chap. 1, n. 3).

10. Ibid., 400.

11. Ibid., 405. The report states, “the head of the NCTC must have the

right to concur in the choices of personnel to lead the operating entities of

departments and agencies focused on counterterrorism.” At a minimum the

report would provide this authority in the selection of the head of the CIA’s

Counterterrorist Center, the head of the FBI’s Counterterrorism Division,

and the commanders of US Special Operations Command (USSOCOM) and

USNORTHCOM. It makes no mention of firing authority.

12. Ibid., 408–10.

13. Ibid., 410.

14. Ibid., 416.

15. Ibid., 417.

16. Ibid., 417–18.

17. Ibid., 419.

18. Ibid., 425. See also Maj Gen Glen Shaffer, retired, interviewed by author,

3 December 2004. General Shaffer, as well as other commentators, stressed

that one of the most important steps within intelligence reform is fixing the

seam between domestic and foreign intelligence or

ganizations. This view is

supported by the WMD Commission (Robb-Silber

man), which found that

“the FBI has not constructed its intelligence program in a way that will pro-

mote integrated intelligence ef

forts, and its ambitions have led it into unnec

essary new turf battles with the CIA.” The Commission on the Intelligence

Capabilities of the United States Regarding Weapons of Mass Destruction,

“Report to the President of the United States,” 31 March 2005, 451.

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

9. All three executive orders are available from http://www.white

house.gov/news/orders/ (accessed 21 December 2004).

20. Executive Order 13354, National Counterterrorism Center, 27 Au-

gust 2004, sec. 3 (a)-(b).

21. Ibid., sec. 6(A)-(C)(b).

22. Executive Order 13355, Strengthened Managment of the Intelligence

Community, 27 August 2004, sec. 2. The intent of this change of wording is

unclear. Prior to Goldwater-Nichols, all of the members of the JCS were

“principal military advisors.” Goldwater-Nichols changed this by making the

JCS chairman the “principal military advisor.” The intent of legislation was

clear—to allow the JCS chairman to provide advice independent from the

service chiefs. In the case of intelligence, the DCI did not share a presiden-

tial advisory role with other members of the IC; therefore, this change may

be more symbolic than substantive.

23. Ibid., sec. 2(c) (amending EO 12333, United States Intelligence Activ-

ities, 4 December 1981, 1.5[g]).

24. Ibid., sec. 2(d) (amending EO 12333 1.5[m]).

25. Ibid., sec. 3 (amending EO 12333 1.5[n]).

26. Ibid., sec. 4(a).

27. Ibid., sec. 5(a). Goldwater-Nichols requires that officers considered

for promotion to general officer ranks have served a minimum of one joint

duty assignment.

28. EO 13356, Strengthening the Sharing of Terrorism Information to

Protect Americans, 27 August 2004, sec. 1.

29. Ibid., sec. 3. See 9/11 Report, 417–18 and specifically footnote 16

(see chap. 1, n. 3).

30. Because the report’s recommendations spanned large sections of gov-

ernment, multiple committees held hearings. The list and video feeds from the

hearings are available at http://www.cspan.org/terrorism/?Cat=Current_

Event&Code=911_Cong&ShowVidNum=10&Rot_Cat_CD=911_Cong&Rot_HT=

&Rot_WD=&ShowVidDays=365&ShowVidDesc=&ArchiveDays=100 (accessed

21 December 2004).

31. The original bill,

A Bill to Reform the Intelligence Community and the

Intelligence and Intelligence-related Activities of the United States Govern-

ment

, S. 2840, 108th Cong., 2d sess., 2004, was placed in the record on 23

September. The Governmental Affairs Committee unanimously approved S.

2845, 108th Cong., 2d sess., on 27 September 2004, and the Senate passed

the bill by a vote of 96–2.

32. The first House bill introduced was

To Implement the Recommenda-

tions of the National Commission on Terr

orist Attacks upon the United States

H.R. 5040, 108th Cong., 2d sess., 2004. The second version that ultimately

passed by a vote of 282–134 was

To Provide for Reform of the Intelligence

Community, T

err

orism Pr

evention and Pr

osecution, Bor

der Security, and Inter

national Cooperation and Coordination

, 108th Cong., 2d sess., 2004, H.R. 10.

33. The House passed the measure on 7 December by a vote of 336–75.

The Senate passed the measure on 8 December by a vote of 89–2. The law

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

as enrolled as the

ntelligence Reform and Terrorism Prevention Act of 2004

P.L. 108-458, 17 December 2004.

34. Senator Jay Rockefeller,

Congressional Record, 27 September 2004,

S9707.

35. Dana Milbank, “Deeply Divided Country Is United in Anxiety,”

Wash-

ington Post

, 4 November 2004, A28. For example, in events leading up to the

2004 election, fears of terrorism remained a significant issue for American

voters. “Nearly everyone feared terrorism. Three-quarters of voters—sup-

porters of President Bush and John F. Kerry alike—said they were worried

there will be another major terrorist attack in the United States.”

36. See, for example, the Ford and Carter executive orders created after

the Church and Pike investigations. The executive branch could reasonably

argue that action had been taken to address the problems and further con-

gressional action was not required. Congress could, of course, have pressed

its case; however, without executive branch support such efforts would

likely have been further complicated.

37. Senator Joseph Lieberman,108th Cong., 2d sess.,

Congressional

Record

, 27 September 2004, S9704. Lieberman, in supporting the legisla-

tion, stated, “Our legislation is based not only on the comprehensive work of

the 9/11 Commission I have described but on the earlier work of the joint

House and Senate Intelligence Committees in their inquiry into matters of

intelligence, on the expertise of scores of experts who have been thinking

about this subject for decades, and on critical reports, as Senator Collins in-

dicated, that date back not 10 or 20 years but 50 years, making similar rec-

ommendations to the ones we have made. It is a tragedy that it took 9/11 to

shake us out of our bureaucratic lethargy to be on the edge of doing what

should have been done 50 years ago.”

38. Senate,

Report of the Committee on Governmental Affairs, United

States Senate to Accompany S. 2840

, 108th Cong., 2d sess., 2004, S. Report

359, 5–8.

39. John Lehman, Testimony to the House Government Reform Commit-

tee, 3 August 2004. “Our recommendations are not a Chinese menu. They

are a whole system. If all of the important elements are not adopted, it

makes it very difficult for the others to succeed.” This failure has not gone

unnoticed in Congress with Representatives Shays and Maloney publicly

calling for congressional reform. See Christopher Shays and Carolyn Ma-

loney, “Congress, Reorganize Thyself,”

Washington Post, 22 December 2004,

A27.

40. Unnamed congressional staffer interviewed by the author, 8 March

2004. It was only through this conversation that the distinction between a

national intelligence center and the NCTC became appar

ent.

41. For the complete colloquy, see

Congressional Record, 8 December

2004, S11964–S11975.

42. See also

Intelligence Reform and Terrorism Prevention Act of 2004,

sec. 1023. The establishment of national intelligence centers is at the dis-

cretion of the DNI. Furthermore, the legislation makes no mention of the re-

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

ated net assessments or warning requirements for national intelligence cen-

ters as Senator Lieberman described in the colloquy (see

Congressional

Record

, 8 December 2004, S11968). Such tasks appear logical for a national

intelligence center; however, their absence in legislation potentially provides

the DNI some discretion as to where these functions will be placed within

the IC. The “center” construct is not a new idea. The

IC21 study identified

seven intelligence centers in existence at the time of the report (1996), but

found that the centers were not true community constructs due to poor rep-

resentation from other organizations outside the CIA (

IC21, 269). See also

Kindsvater, who suggests using centers to manage the IC around

mission/issue and again stresses the need for community-wide participa-

tion in the centers. In creating the national intelligence center construct,

Congress appears to have adopted most aspects of Kindsvater’s recommen-

dations. However, the Kindsvater proposal diverges from the congressional

approach when it comes to the placement of the national intelligence cen-

ters within the organizational hierarchy. See

Congressional Record, 8 De-

cember 2004, S11968, for the analogy to DOD unified commands.

43.

Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1022.

Whether the president will choose to establish an NCPC remained unclear

as of 21 March 2005. The legislation orders the establishment of the NCPC

within 18 months; however, it allows the president to waive the require-

ments of the section if he so desires. If the distinction between national in-

telligence centers and the NCTC/NCPC continues (i.e., the strategic plan-

ning function remains in the NCTC and NCPC), then a generic title for the

latter set of organizations would be helpful. The terms

supercenters or mega-

centers

and other similar terms have obvious drawbacks as they call to mind

unrelated commercial entities. The term

Joint Intelligence and Planning Cen-

ter

(JIPC) seems to be both more descriptive and less open to potential

ridicule.

44.

Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1021,

adding sec. 119(j)(2) to

USC 50, 402.

45. Senator Lieberman,

Congressional Record, 108th Cong., 2d sess., 8

December 2004, S11972.

46. Ibid., S11972. This also appears to be an intrusion in executive

branch activities as it dictates, to some extent, how the executive branch

chooses to execute policy.

47. For comparison see

Intelligence Reform and Terrorism Prevention Act

of 2004

, sec. 1011, amending the National Security Act, sec. 102A(c)(B) ver-

sus previous

USC 50, sec. 403-3(c)(1)(A).

48. Senator Collins,

Congressional Recor

, 108th Cong., 2d sess., 8 De-

cember 2004, S11969.

49. Ibid., S11947. For example, Senator Specter stated during the final

debate, “The DNI is left with the power to ‘develop and deter

mine’ the na

tional intelligence program budget, which is effectively the same authority

that the current DCI is given over the National Foreign Intelligence Program

budget by executive order.”

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

100

ntelligence Reform and Terrorism Prevention Act of 2004

sec. 1011,

amending the

National Security Act, sec. 102A(c)(5)(B).

51. See EO 13355, sec. 3, which amends EO 12333 sec. 1.5(o)(1). The

legislation removes the word

foreign throughout the amended National Secu-

rity Act of 1947

. For example, the National Foreign Intelligence Program be-

came the National Intelligence Program.

52.

Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1011,

amending sec. 102A(d)(5)(A)(iv) of

USC 50.

53. Ibid., sec. 1011, amending sec. 102A(e)(2)(A).

54. Ibid., sec. 1014.

55. Ibid., sec. 1011, amending sec. 102A(f)(1)(A)(ii).

56.

USC 50, 403-4(a). The previous language stated, “To the extent rec-

ommended by the National Security Council and approved by the president,

the Director of Central Intelligence shall have access.” The only additional

limitations are those provided by law. For the new language, see

Intelligence

Reform and Terrorism Prevention Act of 2004

, sec. 1011, amending sec.

102A(b).

57. 9/11 Report, 409 (see chap. 1, n. 3). The most specific reference to

joint personnel practices states, “A common set of personnel standards for

intelligence can create a group of professionals better able to operate in joint

activities, transcending their own service-specific mind-sets.” What is less

clear is whether Goldwater-Nichols–like incentives for joint duty were envi-

sioned, whether there were perceived consequences with joint duty, and

whether the quality of joint personnel was deemed inadequate.

58. Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1011,

amending sec. 102A(l)(3)(C).

59. Ibid., sec. 1011, amending sec. 102A(l)(1)(A)(i)–(iii). The act specifies

that the incentives could take the form of financial incentives, bonuses, or

other awards, as the DNI deemed appropriate.

60. Ibid., sec. 1011, amending sec. 102A(f)(3)(A)(ii).

61. Ibid., sec. 1011, amending sec. 102A(g)(1)(A)–(F).

62. Ibid., sec. 1011, amending sec. 102A(g)(4).

63. Ibid., sec. 1016 (b)(2).

Terrorism information is defined in the act as

“all information, whether collected, produced, or distributed by intelligence,

law enforcement, military, homeland security, or other activities relating

to— (A) the existence, organization, capabilities, plans, intentions, vulnera-

bilities, means of finance or material support, or activities of foreign or in-

ternational terrorist groups or individuals, or of domestic groups or individ-

uals involved in transnational terrorism; (B) threats posed by such groups

or individuals to the United States, United States persons, or United States

inter

ests, or to those of other nations; (C) communications of or by such

groups or individuals; or (D) groups or individuals reasonably believed to be

assisting or associated with such gr

oups or individuals.”

Intelligence Refor

and Terrorism Prevention Act of 2004

, sec. 1016 (a)(4).

64. Ibid., sec. 1016 (d)(1).

65. Ibid., sec. 1016 (e)(1)–(11).

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

101

6. Ibid., sec. 1016 (h). The first report is due two years after the enact-

ment of the act.

67. 9/11 Report, 400 (see chap. 1, n. 3).

68. Unfortunately, the center’s ability to perform such assessments will

likely be hampered by the multitude of different personnel appraisal sys-

tems used by the various agencies participating in the center. For example,

within the military services it can be difficult for supervisors to use appraisal

systems from other services and thus accurately assess a subordinate’s per-

formance. In addition, each of the services often uses signaling devices

within performance appraisals to help promotion boards differentiate be-

tween star performers and average ones. Unfamiliarity with such terminol-

ogy makes the job of supervisor rating a subordinate from another service

that much more difficult. It is likely that similar difficulties will exist in joint

intelligence organizations, as supervisors of subordinates from a different

intelligence agency must learn the vagaries and unique characteristics of

multiple personnel appraisal systems.

69. James Q. Wilson, “Thinking About Reorganization” in Roy Godson,

Ernest R. May, and Gary Schmitt,

US Intelligence at the Crossroads: Agen-

das for Reform

(McLean, Va.: Brassey’s, 1995), 30. Wilson takes the need to

reform culture one step farther, stating, “The central task facing anyone pro-

posing to reorganize a Type Two agency [to include intelligence organiza-

tions] is to think through exactly what operating culture we wish to produce

among rank-and-file employees and then to design an organizational struc-

ture that will increase the chances of that culture being created and sus-

tained.” For an earlier example of implementing culture change within one

part of the IC, see Douglas J. MacEachin, “The Tradecraft of Analysis” in

Godson,

US Intelligence at the Crossroads, 72–74.

70. A final set of participants—the SecDef and CJCS—can, and often do,

have a great deal of influence on the final nature of the plan. See Cohen,

Supreme Command, 239 (chap. 2, n. 63), in describing the planning for op-

erations in Afghanistan. Cohen states, “The plan was, to be sure, General

Franks’ [the combatant commander] plan. But it was a plan hammered out

in months of grilling by Rumsfeld and his immediate subordinates; it was a

plan that was approved—grudgingly, in some cases—by the uniformed

chiefs of the services, who had no responsibility for war planning, but who

exercised an important influence on military opinion, and, potentially, on

public opinion.”

71. For a complete description of joint campaign planning, see figure 3-

3 in Joint Publication 5-00.1,

Joint Doctrine for Campaign Planning, 25 Jan-

uary 2002, 3–5. Chapter 3 of this document describes the planning process

in depth. Further

more, it is the J-7, not the J-3, that has this review task,

“OPLANs submitted to the Chairman of the Joint Chiefs of Staff for review

e r

eferr

ed to the J-7, which conducts and coor

dinates the final plan r

view” (3–17). See also Joint Publication 5-0,

Doctrine for Planning Joint Op-

erations

, 13 April 1995, 1–6: “The combatant commanders are responsible

for the development and production of joint operation plans.” After the com-

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

102

atant commander drafts a plan, the plan passes through a review process.

“Those joint operations tasked in the JSCP [Joint Strategic Capabilities Plan]

for CJCS approval are reviewed by the Joint Staff, Services, and DOD agen-

cies. All other plans are approved by the CINCs,” 1–8.

72. In both Desert Storm and Allied Force the air component command-

ers had different views on the proper employment of airpower. In both in-

stances there was some push towards a middle ground; however, in neither

case was the component commander’s initial desires fully implemented. See

chap. 2.

73. 9/11 Report, 406 (see chap. 1, n. 3). To its credit the 9/11 Report

recognized this potential stating, “The proposed NCTC would be given the

authority of planning the activities of other agencies. Law or executive order

must define the scope of such line authority.”

74. Ibid., 399. “Some of the saddest aspects of the 9/11 story are the

outstanding efforts of so many individual officials straining, often without

success, against the boundaries of the possible. Good people can overcome

bad structures. They should not have to.”

75. In the present situation the president may largely determine the

power of the DNI. If the president outwardly supports the DNI and routinely

favors the DNI’s positions over those of the department heads, then both the

real and perceived power of the DNI will grow correspondingly; however, if

the opposite occurs, a primary source of power for the DNI (the support of

the president) will be significantly undermined. This suggests that at least

some aspects of the DNI’s power base will be variable as opposed to the rela-

tive stability of purely positional power and will be based upon each presi-

dent’s behavior and actions towards the DNI and department heads.

76. Prior to 1986, the lack of a true joint force commander with control

over forces assigned and the resulting problems were evident in both the

failed Desert One operation, with each of the service’s wanting a prominent

role, and the lack of a joint air campaign in Vietnam.

77. Shaffer interview. General Shaffer described the concept as a

“Goldwater-Nichols at the interdepartmental level.”

78. The advantages of such an approach include flexibility and synchro-

nization. Flexibility is obtained by establishing JITFs in response to antici-

pated government action. This allows the government to tailor JITFs in

terms of size, organizational composition, and relative permanence in re-

sponse to specific mission requirements. Second, JITFs solve a fundamen-

tal problem of the NCTC construct—the lack of control over personnel

tasked with accomplishing a mission. The JITF commander would have op-

erational control over for

ces and thus could plan operations based upon the

for

ces assigned and then oversee execution of the plan. In these ways the

JITF construct would much more closely follow the DOD joint constructs

than the NCTC. The primary disadvantage of the JITF appr

oach is that the

task forces are temporary in nature. Thus a degree of organizational learn-

ing must occur each time a JITF is established. The DOD partially mitigates

this learning curve through the development of doctrine that specifies how

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

103

oint forces are organized and clearly establishes command relationships.

Training and exercises provide an additional source of learning. A JITF con-

struct would need to develop doctrine and training opportunities so that the

multiple government departments and organizations could establish an

agreed upon framework for the establishment and operation of JITFs.

79. Shaffer, interview.

80. Michael V. Hayden, interviewed by the author, 18 February 2005. In

short, the codification of authorities remains a central issue if the IC is to

become less personality dependent and consensus driven. See also Richard

M. Emerson, “Power-Dependence Relations,”

American Sociological Review

27 (1962): 31–41. Emerson’s paper describes four balancing operations to

alter power-dependence relations: withdrawal, extension of power network,

coalition formation, and emergence of status. While Emerson was focused

on social power, aspects of the balancing operations are applicable to bu-

reaucratic power adjustments, particularly coalition formation.

81. 9/11 Report, 410 (see chap. 1, n. 3). See also Kindsvater, who de-

scribes the need for hire/fire authority, personnel transfer authority, and

budgetary authority as essential elements to successful management of the

IC and the intelligence centers.

82. John R. P. French and Bertram Raven, “The Basis of Social Power,”

Group Dynamics, D. Cartwright and A. F. Zander, eds. (Evanston, Ill.:

Row, Peterson Publishers, 1960), 259–69. The authors set forth five bases of

social power: reward, coercive, legitimate, referent, and expert. Of the power

bases, legitimate power most clearly equates with power formally vested in

an office by statute, executive order, or policy. However, individuals holding

an office also have a degree of reward and coercive power based upon the

authorities granted by Congress or the president.

83. Hearing of the Senate Governmental Affairs Committee, “Reorganiz-

ing America’s Intelligence Community: A View from the Inside,” 16 August

2004. The variance in how presidents interact with the DCI was remarkably

evident during an exchange with Senator Lieberman and three former DCIs.

Senator Lieberman asked each whether they personally briefed the presi-

dent, and how often they spoke to him. In response, former DCI William

Webster stated he briefed President Bush on most days but that President

Reagan had taken the brief from the national security advisor. Former DCI

James Woolsey stated that President Clinton normally read the brief and

that he had two “substantive meetings” with the president in two years. Fi-

nally, Adm Stansfield Tur

ner stated that President Carter pr

eferred to read

the President’s Daily Brief but that he had a half-hour session with him

three times a week.

84. The DNI is tasked with managing and directing the “tasking, collec-

tion, analysis, production and dissemination of national intelligence.” See

Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1011, amend-

ing NSA, sec. 102A(f)(1)(A)(i)–(ii) for a mor

e complete description of the DNI’s

tasking and other authorities.

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

104

5. Joan Dempsey, interviewed by the author, 30 November 2004.

Dempsey stressed that budgetary authority is essential; however, the ability

to move personnel throughout the enterprise and have organizations directly

report to the DNI remain important elements to the DNI’s authority.

86. For example see Cohen,

Supreme Command, 191 (see chap. 2, n. 63).

“The Goldwater-Nichols legislation had made the chairman of the Joint Chiefs

of Staff the primary military adviser to the president, and in Powell’s hands

the chairmanship became an exceptionally powerfully position and one which

shut off many other sources of military advice to civilian authority.”

87.

Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1011,

amending the

National Security Act, sec. 102(A)(l)(1)(A). Joint positions speci-

fied include the staff of the DNI, the staff of the national intelligence centers,

the staff of the NCTC, and “other positions in support of the IC management

functions of the Director.”

88. Joan Dempsey, interviewed by the author, 30 November 2004.

89.

Intelligence Reform and Terrorism Prevention Act of 2004, sec.

3001(c)(1).

90. Ibid., sec. 1019(a), which requires the assignment of an individual or

entity to “be responsible for ensuring that finished intelligence products pro-

duced by any element or elements of the IC are timely, objective, indepen-

dent of political considerations, based upon all sources of available intelli-

gence, and employ the standards of proper analytic tradecraft.”

91. Aris A. Pappas and James M. Simon, Jr., “The Intelligence Community:

2001–2015,”

Studies in Intelligence 46, 1 (2002), http://www.odci.gov/csi/

studies/vol46no1/index.html (accessed 10 April 2005). Information overload

problems include not only processing demands for the data, but the larger

human limitations to “usefully synthesize, analyze, and simply understand

what we have.”

92. Although interoperability has shown steady progress there are many

examples where work remains. For example, there is no common e-mail di-

rectory service for all DOD personnel. Likewise, the services use different

network security architectures, employing different firewalls, intrusion de-

tection systems, and other security solutions. There are also procedural

hurdles. The creation of interoperable communication systems does not

guarantee that the users of the system will share common communication

protocols. In other words, the users can talk to each other but they are using

different terminology that can create confusion.

93. See, for example, Stephen Daggett,

The US Intelligence Budget: A

Basic Overview

(Washington, D.C.: CRS, 24 September 2004), 4. “On its

face, this [legislative changes to DCI budgetary authority] gives the DCI con-

siderable authority over NFIP budgets. The statutory language, however

gives the secretary of defense operational control over the major defense in-

telligence agencies, even though tasking of the agencies may be issued by

the DCI. So, as a practical matter, it is the secretary of defense who has the

main authority to establish priorities.”

THE 9/11 REPORT AND INTELLIGENCE REFORM LEGISLATION

105

Chapter 5

The Push and Pull of Intelligence Reform

At our first public hearing on March 31, 2003, we noted the

need for balance as our government responds to the real

and ongoing threat of terrorist attacks. . . . This balancing is

no easy task, but we must constantly strive to keep it right.

—9/11 Report

As the epigraph above shows, the 9/11 Commission recog-

nized the need to grant the government more power in periods

of heightened threats while continuing to protect civil liberties.

This is one of the many balances that the US government

must consider as it evolves to address a wide variety of threats

in the international security environment. Just as there is a

need to maintain government-wide balances, the individual

departments and agencies must operate in the face of often

opposed requirements.

The consideration and interplay of competing interests

within

the government is not a new phenomenon. As described in

chapter 2 and by other commentators on the Goldwater-

Nichols Act, reform of the DOD involved recalibrating the

balances within the system. In a similar vein, reforming the IC

may be thought of as a recalibration of fundamental balances

that represents inher

ent and competing interests within the IC

rather than as a restructuring exercise. Understood this way,

reform requires the action of governmental actors, who adjust

these balances using a variety of more or less permanent and

dynamic recalibration mechanisms.

One balance, the degree of centralization versus decentral-

ization, is common to both defense and intelligence reform ef-

forts. Associated balances unique to intelligence reform in-

clude (1) balancing the need to share information with the

need to protect information; (2) organizational adjustments

that support “connecting the dots” while maintaining diversity

of analytic opinion; and (3) determining the proper allocation

of finite intelligence r

esour

ces to meet the needs of both na

107

tional policy makers and the departments, particularly sup-

port to military operations. Still other balances, while not ex-

amined in detail in this study, are likely to affect recalibration

efforts; for example, determining the proper amount of power

to vest in the government while protecting the civil rights of

citizens, as noted above. Similarly, reformers will have to con-

sider the personnel requirements of the IC’s component de-

partments as well as the staff needed to support analytic cen-

ters and other centralizing organizations. Each of these and

many other fundamental tensions are important in their own

right, because changes within the IC, structural or otherwise,

could have unintended consequences on the operation of

other aspects of the IC.

The centralization/decentralization balance will be a primary

focus of the present study because it interacts with and thus

affects many of the other balances. This chapter examines

possible reasons why the trend towards greater centralization

may now be approaching some real barriers, inhibiting further

centralization. Next, it reviews some of the associate balances

to provide further insight into the types of issues confronting

leaders within the IC as they attempt to implement reforms.

The chapter closes with a discussion of some of the unique

challenges that face r

eformers because of the interagency na

ture of the IC and how the Goldwater-Nichols Act addressed

somewhat analogous challenges.

Aspects of Recalibration

Recalibration attempts share several important features.

First, the interrelationships among many of the balances mean

that adjusting one may cause changes in another, complicating

any r

ecalibration and causing intended or unintended ripple ef

fects as changes are made. For example, as organizations be-

come more centralized there is often a corresponding trend to-

wards more generic skill sets within the workforce. By contrast,

decentralization most often encourages greater specialization as

dispersed elements focus more exclusively on their unique

tasks. In the specific example of intelligence, greater centraliza-

tion r

esulting from consolidation of analytic resources around

THE PUSH AND PULL OF INTELLIGENCE REFORM

108

enduring problems should, in theory, support greater informa-

tion sharing, but this could occur at the expense of more nar-

rowly focused analytic capabilities, trained to support specific

departmental requirements.

Second, the potential consequences of adjustments are not

inevitable. While some forces encourage movement along a

continuum, measures can be taken to limit or even reverse the

movement. For example, more centralization does not make

greater generalization unavoidable although that may be the

natural course. Instead, measures such as retaining strong in-

dependent analytic shops within the constituent IC organiza-

tions may allow departmental specialization to continue. Simi-

larly, intelligence centers of the type envisioned by the 9/11

Commission would centralize analytic capabilities but might

also encourage a different type of specialization—one that is

threat-based rather than department-based.

Third, recalibrations can be accomplished by a variety of

mechanisms ranging from legislation to executive orders to

changes in internal processes and procedures. As noted previ-

ously, legislation offers a degree of permanence but lacks the

precision and responsiveness of internal mechanisms that pro-

vide more dynamic adjustments. Thus, a trade-of

f exists among

the types of mechanisms chosen to r

ecalibrate a balance.

Centralization versus Decentralization

At a fundamental level, determining the amount of centraliza

tion within an organization means deciding how much relative

autonomy to grant to subunits within the structure. The IC ex-

hibits a fundamental tension between a desire to maintain con-

trol over a vast enterprise to reduce duplication and improve co-

dination and a desir

e to allow the various departments to

maintain relatively autonomous organizations to obtain the tai-

lored intelligence support they require. Thus, the centraliza-

tion/decentralization debate can be recast in terms of cohesion

and autonomy.

Moreover, this balance has a significant effect on many of

the other balances. For example, the degree of centralization

can af

fect the development and implementation of organiza-

THE PUSH AND PULL OF INTELLIGENCE REFORM

109

tional standards, which may, in turn, affect the valence of such

other balances as the degree of specialization/generalization

or information sharing/protection. Because this balance plays

a primary role, it is important to understand how the IC has

evolved with respect to this balance and outline the relative

advantages of both centralization and decentralization before

examining why the trend towards centralization may be near-

ing some concrete barriers that make further centralization in-

easingly difficult.

The National Security Act of 1947 was bor

n from a system in

which national intelligence was largely decentralized, with the

most significant efforts occurring independently within the

Departments of State, War, and Navy.

The passage of the act

began the first of many steps towards increasing centralization

within the IC. Initially, the DCIs, particularly Allen Dulles, fo-

cused almost exclusively on their role as director of the CIA.

However, as early as 1956, advisory bodies within the execu-

tive branch began issuing recommendations for the DCI to in-

crease managerial emphasis on IC affairs. In early 1962 Pres-

ident Kennedy directed the DCI to place greater weight on

community management over leadership of the CIA. The 1971

Schlesinger Report, the 1972 Murphy Report, the 1976 Church

Committee Report, the 1994 Aspin-Br

own Commission, and

the 1996

IC21 report all supported increased centralization

and recommended enhancing the DCI’s power over the IC to

impr

ove IC management.

The movement towards greater centralization within the IC

is not unusual and follows similar trends in other areas of the

government. Since World War II, the armed forces have be-

come more centralized, first with the creation of the Depart-

ment of Defense and then via the Goldwater

-Nichols Act. In a

similar vein, the consolidation of such pr

eviously independent

agencies as the Federal A

viation Administration, Federal High

way Administration, Federal Railr

oad Administration, and US

Coast Guar

d resulted in the formation of the Department of

Transportation.

More recently, the DHS was formed to create

“a single, unified homeland security structure” to centralize the

her

etofor

e dispersed elements of gover

nment involved in such

THE PUSH AND PULL OF INTELLIGENCE REFORM

110

functions as border and transportation security and emer-

gency preparedness and response.

Various benefits associated with centralization have encour-

aged this trend. The fundamental functions of training and

equipping can often benefit from greater centralization. Cen-

tralization can provide greater coherence through increased

standardization across education and training programs—a

useful element to increase the overall quality of the analytic

force or other relevant intelligence functions. Increased cen-

tralization in the planning, programming, and budgeting process

also offers the chance to reduce costs and duplication by giv-

ing the decision maker an enterprise-wide view of the types,

capabilities, and costs of systems.

Such a comprehensive per-

spective increases the likelihood that duplicative systems can

be identified prior to full funding and that funded systems will

support the broadest possible customer set.

On the analytic front, centralization, if supported by strong in-

formation sharing, can help create a more accurate and com-

plete pictur

e of large, complex issues and threats that span more

than one agency or department. Establishing more uniform per-

sonnel systems may assist in the movement of personnel among

organizations in the community and ensure that compensation

and other incentives are distributed fairly acr

oss the enterprise.

Furthermore, consolidating authority within a single position or

organization can increase legitimate sources of power.

For ex-

ample, gr

eater statutory or formal authority to enforce new stan-

dards and practices may do more to attain objectives such as in-

creased information sharing than relying too much on informal

mechanisms such as cooperation and consensus, as was done

previously. Finally, centralization focuses responsibility, author-

ity, and accountability often to a single person for the conduct of

organizational activities.

Yet strong incentives for decentralization also remain. Good

intelligence depends on the productive blend of multiple, highly

specialized skills, and it is the IC’s component departments

that select and train the practitioners of those skills. Thus,

while analysts may benefit from some standardized and cen-

tralized training, they also likely r

equire more tailored training

specific to their departmental analysis r

esponsibilities. While

THE PUSH AND PULL OF INTELLIGENCE REFORM

111

centralization can reduce duplication in the acquisition process,

the individual organizations may have less of a voice in advo-

cating their specific requirements. Thus, given limited resources,

systems tailored to the unique requirements of individual or-

ganizations might lack community-wide funding support. Fur-

thermore, duplication is not entirely bad, and attempting to

eliminate duplication completely has definite drawbacks. Re-

dundant capabilities enable the larger intelligence system to

respond more ably in times of crisis through greater capacity.

Decentralization can serve the analytic function by increasing

the likelihood that indigenous analytic resources will focus on

the unique pr

oblems and challenges facing a specific depart-

ment or agency. In this way DOD intelligence analysts can sup-

port the operational requirements of the military, while other

departmental and agency analysts can do the same for their

respective organizations. Furthermore, redundant departmen-

tal analysis shops should promote a diversity of analytic opin-

ion on broad topics while reducing the likelihood that “group-

think” pathologies may for

m. Thus the quest for greater

autonomy has a rational basis and may be less opportunistic

than some IC critics assume.

Although the recent legislation continues the trend toward

centralization, this trend may soon r

each a point where the

mar

ginal gains achieved from additional centralization are out-

weighed by the associated costs. For example, a further step

in the pr

ogression towards greater centralization of the IC

would be to create a cabinet-level department encompassing

the major intelligence organizations (i.e., the CIA, NSA, NGA,

and National Reconnaissance Office [NRO]), while allowing the

remaining departments to maintain separate analysis organi-

zations focused on their specific requirements.

While possi-

ble, such a solution would have to overcome many obstacles.

First are the costs associated with personnel. Currently

military personnel represent a significant cross section of the

IC. It is unlikely that the DOD would be willing to lose a lar

number of its intelligence r

esources to a centralized intelli-

gence structure without some form of dispensation or ability

to contr

ol the personnel in support of unique DOD require-

ments.

Alter

natives include detailing military personnel for

THE PUSH AND PULL OF INTELLIGENCE REFORM

112

extended periods or authorizing the IC to hire new personnel,

but both of these options create significant costs for either the

DOD or the government at large. There are also difficulties in

supporting the personnel requirements of intelligence centers.

Even without a complete transfer of personnel, the require-

ments to support such analytic centers as the NCTC could be-

come burdensome for those departments—for example,

State—that maintain relatively small intelligence organiza-

tions. Whether the government has adequate resources to

continue to support both departmental analytic organizations

and centers depends to some degree on the personnel require-

ments levied by the DNI to support the NCTC or other centers.

A second difficulty is the requirement of the various depart-

ments for indigenous analytic elements to support their re-

spective operational missions. Even the most dedicated re-

formers have not suggested eliminating the component

departments and agencies, which will continue to demand

specialized support. The difficulties surrounding further cen-

tralization of intelligence analysis r

eflect the balance between

generalists and specialists. The CIA was formed in part to cen-

tralize the analysis of intelligence fr

om across the IC and thus

began a small shift towards the generalist pole. The NCTC con-

tinues the trend towar

ds centralization but focuses the cen-

tralization on a somewhat narrower area; therefore, its overall

effect on the specialist/generalist balance remains unclear. In-

stead of shifting the specialist/generalist balance, centers may

instead change the type of specialization from a departmen-

tally oriented one to a threat-based one. A decrease in con-

gressional funding or an increase in the number of analysts

required to support the centers might prompt a greater shift

towards the generalist perspective or a change in the nature of

specialization. In either case, specialist support to the depart-

ments would likely bear the brunt of the impact.

Another obstacle is associated with the placement of the

combat support agencies (CSA) curr

ently housed within the

DOD. The pr

ospect of moving the CSAs under a DNI and out

from under the DOD was discussed during a House Armed

Services Committee hearing on the 9/11 Report. One witness,

. John Hamr

e, a for

mer deputy secr

etary of defense, consid

THE PUSH AND PULL OF INTELLIGENCE REFORM

113

ered such an option feasible, if not desirable; while another

witness, Gen William Odom, a former NSA director, thought

otherwise.

Odom believed that the DOD would replicate the

units and associated intelligence capabilities lost to a DNI to

maintain the support the military services deemed necessary.

Others argue that such replication is too expensive, and thus

the DOD and the DNI would have to cooperate. The answer

probably lies somewhere in the middle.

The DOD experience with close air support (CAS) provides

insight into how events might transpire. The DOD could have

chosen to consolidate all air assets into a single service: the

US Air For

ce. Instead, the services continue to maintain air

capabilities to better support their specific requirements; for

example, the Army and Marines maintain indigenous air ele-

ments to provide tailored air support to their associated

ground forces. Ground forces prefer the current system for

several reasons. On the practical side, dedicated air support

assets can routinely train and exercise with the ground forces

they ar

e tasked to support, providing valuable experience and

expertise for actual combat. In addition, both the Army and

Marines str

ongly believe that owning indigenous assets is the

only way to ensure that necessary support will be available

when requir

ed.

This second reason illustrates a fundamen

tal tenet in most organizations: direct control over supporting

assets is nearly always preferable to reliance on cooperative ef-

forts fr

om an outside entity.

Thus, there is some reason to believe that should the DOD

lose more control over the CSAs, such as NSA and NGA, the de-

partment would attempt to replicate the inherent capabilities of

these organizations in some fashion.

The 1971 Schlesinger Re-

port discussed this possibility in examining the potential rami-

fications of transferring control of intelligence assets to a DNI,

stating, “there would remain the serious and continuing prob-

lem of finding ways to meet the intelligence needs of Defense

[the DOD] without, at the same time, causing the Services to r

constitute their own intelligence activities, even at the expense

of other programs.”

Even the cost issues raised by Dr. Hamre

may not fully deter such ef

forts. In fact, Congress might be a

willing participant in allowing the duplication to pr

oceed. Be

THE PUSH AND PULL OF INTELLIGENCE REFORM

114

cause Congress maintains separate oversight committees for in-

telligence and the armed services, the DOD would likely have

strong support from its oversight committee to fund service-only

intelligence systems if the department could reasonably show

that intelligence support to military operations had suffered as

a result of organizational restructuring or reduced priority for

departmental requirements.

Second, the CAS example illustrates that Congress, perhaps

recognizing the advantages of indigenous support assets, is

not averse to funding multiple and sometimes redundant sys-

tems to support the requirements expressed by the military

services. Thus, while the new systems funded may not be as

capable as the original systems, some funding of new systems

would probably be authorized, creating additional redun-

dancy. Finally, as the recent debate over intelligence reform

legislation illustrated, the armed services committees ar

e, and

likely will remain, staunch supporters of the military services,

which increases their likelihood of acting any time intelligence

support to the military is threatened.

Associated Balances

As the fulcrum is shifted on the centralization/decentraliza-

tion axis, it often affects other balances. The first is related to

creating conditions that allow the successful assimilation of dis-

ete data into a coherent and accurate analytic pictur

e without

creating conditions that impede the diversity of opinion.

Balancing the need to share with the need to protect is a

second tension that involves facilitating greater access to in-

formation while still protecting critical elements of information

that could hinder future intelligence efforts. A third, and by no

means final, trade-off is associated with balancing the intelli-

gence needs of the national policy maker with the needs of the

individual departments.

“Connecting the Dots” versus “Groupthink”

During the debate surrounding the 2004 intelligence reform

efforts, Dr. John Hamre of the Center of Strategic and Interna-

tional Studies expr

essed concer

n over what he called the “con

THE PUSH AND PULL OF INTELLIGENCE REFORM

115

nect the dots” versus “groupthink” problem. In essence, the

problem revolves around how to centralize intelligence so that

the IC can develop a more accurate and complete threat pic-

ture without creating conditions where analysts fail to fully

consider alternative analyses of the available information. In

theory, as centralization increases, the diversity of analytic

opinion can be short-circuited or impeded. Centralization, as

a general rule, helps solve the connect the dots problem by

consolidating both data and the analysis. On the other hand,

decentralized analytic organizations create a natural mecha-

nism to achieve diversity of opinion but make it more difficult

to consolidate infor

mation and thus connect the dots.

However, other ways to address this balance do not require

wholesale structural changes to analytic organizations. For ex-

ample, additional incentives to encourage greater information

sharing or changes to current policies that encourage informa-

tion protection are two of the mechanisms available. Other pre-

viously suggested approaches include improved dialogue with

consumers of infor

mation, periodic external reviews of intelli-

gence products and analytic methodologies, and the creation of

net assessment gr

oups to challenge assumptions.

However

encouraging a diversity of opinion does not have to rely exclu-

sively on independent analytic groups, and in any event, such

groups cannot fully substitute for internal changes to analytic

organizations. Instead, a renewed focus on analytic tradecraft

and standar

dized analyst training may offer plausible alterna-

tives to the maintenance of multiple analytic organizations. At

the same time, separate analytic organizations serve a valuable

purpose by providing tailored support to the owning department

or agency—a more compelling reason for their continuation

than simply avoiding groupthink.

The intelligence reform legislation appears to incorporate a

mixture of both approaches discussed above—increased cen-

tralization via structural changes and new initiatives to improve

the quality of analysis. By r

etaining departmental analytic bod

ies, Congr

ess avoided shifting the centralization/decentraliza-

tion balance too far towards either pole. As a new organizational

construct, the NCTC holds pr

omise of improving the consolida-

tion of terr

orism-r

elated data and thus aiding in mor

e holistic

THE PUSH AND PULL OF INTELLIGENCE REFORM

116

assessments of those data. Additionally, information sharing

should be simplified, because all organizations can now report

terrorism-related information to a single organization represent-

ing the entire IC. This contrasts with the situation described in

the 9/11 Report, where the sheer number of disparate organi-

zations involved in both collection and analysis impeded infor-

mation sharing. In some ways simply knowing with whom to

share the information may be a key aspect of the problem and

one that the NCTC should help address.

However, there are potential dangers to this consolidation.

First, unless safeguards are put in place, alternative view-

points could either be suppr

essed or in the worst case not

even generated (i.e., the groupthink problem).

Thus an im-

portant aspect of center implementation may be the relation-

ships the centers continue to maintain with the analytic or-

ganizations of the department. A second danger of continued

consolidation is that formation of the NCTC and any other fu-

ture centers may drain the analytic resources of the depart-

ments and agencies within the IC. This r

esembles the issue

that the military services raised in response to joint duty re-

quir

ements for service personnel. As long as the requir

ements

are not too onerous, they should not create a significant prob-

lem. However, some departmental intelligence or

ganizations

are relatively small, and their ability to support both multiple

centers and departmental analytic requirements could be re-

duced if the demand for analysts gr

ows significantly.

Another approach Congress used to achieve balance was

embedded in two sections of the act. In the first section, Con-

gress required the DNI to appoint an individual or entity that

would be responsible for ensuring that “elements of the IC

conduct alternative analysis (commonly referred to as ‘red-

team analysis’) of the information and conclusions in intelli-

gence products.”

Left unsaid is how such analysis would be

incorporated into products or presented to decision makers. In

a second section, Congr

ess r

equir

ed the DNI to appoint another

individual or entity to addr

ess issues of analytic integrity.

Both portions of the law appear to have been designed to guard

against some of the negative ef

fects of greater centralization

while maintaining a diversity of analytic opinion.

THE PUSH AND PULL OF INTELLIGENCE REFORM

117

The congressional solutions centered on creating new struc-

tures to centralize intelligence analysis around enduring threats

and providing new independent review mechanisms. While these

changes may be effective, internal mechanisms to improve an-

alytic tradecraft and training of analysts may be equally essen-

tial.

Thus, creating environments where analysts are encour-

aged to challenge assumptions, to avoid mirror-imaging and

other deleterious activities, and to integrate outside sources of

expertise may reduce the likelihood that destructive group

processes would form.

Information Sharing versus Information Protection

The preceding discussion shows that the connect-the-dots

oblem is related to how the information sharing/information

protection balance is calibrated. If the balance is shifted to en-

courage gr

eater information sharing among agencies, it be-

comes more likely that the necessary information to build an ac-

curate analytic picture will be available. Unfortunately, at the

same time pr

otecting infor

mation, particularly sour

ces and

methods, becomes more difficult. As with other balances, there

e legitimate inter

ests on both sides. Greater information shar-

ing can not only help establish more accurate analytic pictur

es,

but also assist in direct support of operations (e.g., military,

counterintelligence, homeland security). Within the military,

war fighters have almost constantly called for more detailed, ac-

curate, and timely information. This has required changes in in-

telligence processes to protect sources and methods while pro-

viding greater access to information for operators.

Although access to such information can fundamentally im-

prove the quality of analysis and the effectiveness of opera-

tions, there are also dangers. The security infrastructure es-

tablished to protect information has evolved in response to

what was initially a serious counterintelligence threat from the

Soviet Union. A series of spy cases from John Walker to

Aldrich Ames to Robert Hanssen clearly illustrates the dangers

of failing to protect information, with damage ranging from

loss of counterintelligence investigative techniques to loss of

valuable sour

ces and methods and, in the worst cases, to the

loss of human life.

Ther

e is also the need to protect intelli-

THE PUSH AND PULL OF INTELLIGENCE REFORM

118

gence processes from disclosure, as such disclosures could

enable an adversary to influence these processes.

Therefore,

even though the environment has changed, there remain com-

pelling reasons to protect information.

Increasing the number of individuals who have access to in-

formation also increases the likelihood of compromise, either

intentional or unintentional (through poor security practices).

On the other hand, increased sharing may not only make

analysis more effective, but may also save lives through better

warning of impending threats. According to this line of reason-

ing, it does no good to protect information if overly restrictive

otective measures prevent timely warning of imminent

threats or if those tasked to counter a threat fail because of in-

adequate or incomplete intelligence.

As with any balance, a guiding objective here is to increase

the benefits of recalibration. In this instance, the shift towards

greater information sharing would continue until the costs as-

sociated with the compromise of information became too great.

The risk of compr

omise remains a primary reason why complete

information sharing is unlikely. As compromises increase, the

quality and quantity of infor

mation collected would begin to de-

grade sharply, effectively negating any benefits gained from in-

creased infor

mation sharing through the loss of r

eliable and

necessary information. Because that risk is real, a key aspect of

any successful recalibration will be the design of security proce-

dur

es that continue to protect essential elements of information

while allowing for greater dissemination among analysts and

operators. This is exactly the approach the legislation has taken

in recommending the use of tearlines and other procedures to

assist in the information sharing process.

Another focus area for recalibration centers on providing

greater incentives to share information while reducing real

and sometimes culturally imposed disincentives.

Disincen-

tives such as statutory and personnel penalties for violating

security r

equir

ements ar

e relatively straightforward to set and

enfor

ce. However, like most disincentives, punishments often

result in a focus on self-protection, which in turn may result

in manifestations of unintended behavior

In this case, the

disincentives imposed discourage disclosur

es of infor

mation

THE PUSH AND PULL OF INTELLIGENCE REFORM

119

not only to unauthorized individuals (intended behavior) but

also to individuals who may have legitimate needs (unintended

behavior). Currently there are very few incentives to share in-

formation and creating such incentives requires an entirely

different approach.

In addition to the challenge of determin-

ing the nature of the reward (time off, monetary or institu-

tional awards, etc.) is the real problem of measuring how data

is being shared, who is withholding information, and who is

sharing infor

mation. Thus, the development of metrics to as-

sess pr

ogress towards information sharing may become an im-

portant implementation issue. An agency’s responsiveness to

a request for information, the amount of data provided, and

audits by community managers to determine the amount of

information withheld represent partial but incomplete solu-

tions to the problem of measurement.

The combination of procedural and incentive changes

should assist adjustments along the information sharing con-

tinuum. Moreover, both mechanisms provide relatively greater

opportunities for precise and dynamic adjustments than

purely legislative changes can accomplish. For example, if in-

formation sharing policies result in unacceptable disclosures

of information, such policies can be more rapidly adjusted to

better pr

otect information. Finally, cultural changes within

the IC will pr

obably be needed to move from a need-to-know to

a need-to-share mind-set.

Explicit means such as procedural

and incentive adjustments should have some ef

fect on behav-

ior; however, long-lasting changes will likely require more pro-

found changes to the underlying cultural makeup. In the pre-

sent case, multiple influences shape organizational culture to

include the normative behavior of peers with regard to infor-

mation sharing, actions taken by managers to encourage or

discourage behavior

, and the per

ceived r

ecipr

ocity of infor

tion sharing among participating or

ganizations. The IC may be

well served by examining the pr

os and cons of the DHS ap

oach to information sharing, which continues to seek ways

to develop procedures and incentive structures to encourage

the sharing of information not only across the federal govern-

ment but also with state and local gover

nments.

THE PUSH AND PULL OF INTELLIGENCE REFORM

120

In short, as with nearly every balance, there is no single so-

lution. Instead, greater information sharing will likely require

technical changes to communication architectures, revised

policies and processes, adjustments to incentive structures,

and changes in organizational culture.

Intelligence Needs of the National Policy Maker

versus the Departments

A third balance is associated with the sometimes competing

and sometimes shared intelligence needs of the national policy

maker and the departments, particularly the DOD.

Because

intelligence resources for both collection and analysis are finite,

there are limits on the amount of intelligence that can be col-

lected, pr

ocessed, and analyzed, which necessitates a balance

in resource allocation. Further complicating this problem, na-

tional security and r

egional military priorities are now less syn-

chronized because of the absence of a unifying threat such as

existed prior to the end of the cold war.

This loss of synchro-

nization implies that the demands of policy makers and the de

partments are more likely to conflict than previously.

The national versus tactical balance r

epr

esents a distinction

perpetuated, in part, by the division of the intelligence budget

into separate national and departmental components.

Al-

though some intelligence can be characterized as either na-

tional or departmental/tactical, considerable overlap and am-

biguity remain. This ambiguity further complicates the

balancing of competing demands for these resources. A more

accurate distinction in today’s strategic context may be be-

tween long-term, enduring requirements and near-term, gran-

ular ones.

An assessment of the possible future actions of a

nation-state, its military capabilities, and its industrial capac-

ity would fall into the long-term category, while determination

of the position and disposition of enemy forces in support of

current military operations would be characterized as near

ter

m. Such a distinction may mor

e accurately convey the dif-

fering types of intelligence and the valid requirements of both

poles of this balance. However, even this distinction does not

esolve along neat or

ganizational lines, with the DNI r

esponsi-

ble for meeting enduring intelligence needs and the depart

THE PUSH AND PULL OF INTELLIGENCE REFORM

121

ments responsible for near-term intelligence. For example, the

DOD has both long-term and near-term intelligence require-

ments in much the same way as it has both national and tac-

tical/departmental ones. Long-term enduring intelligence is

needed to support senior defense policy makers and to assist

the services’ force modernization and planning functions,

while near-term granular intelligence is needed to support on-

going military operations.

One argument for placing greater weight on national intelli-

gence requirements is that the resulting intelligence forms the

basis for national security policy making.

According to this

gument, national intelligence requirements should trump

others because the actions of the departments are predicated

on the decisions of national policy makers. Without solid intel-

ligence, national policy makers may commit or fail to commit

government resources as necessary. For example, accurate in-

telligence assessments of an adversary country’s intentions

and capabilities become a critical factor in determining the

gover

nment’s response and the associated instruments of

power to employ. Thus, strong national intelligence may actu-

ally pr

eclude the use of military force if the thr

eat is deemed

to be less urgent than originally assumed.

A counterargument is that the activities of the departments

have national security implications that in turn affect the na-

tional decision-making context. Incomplete or inaccurate in-

telligence can r

esult in increased casualties, unintended col-

lateral damage, or less effective military operations, all of

which can have significant implications for the further execu-

tion of national policy. This argument states that adequate

support to military operations or other department-specific

operations is essential to avoid reducing the options of the na-

tional decision maker.

For example, as military casualties in-

crease, the American public may reassess the costs of US in-

volvement against the interests served. As these costs begin to

be per

ceived as outweighing the benefits, public opinion may

shift towar

d disengagement, making continued involvement

politically harder to justify. The bombing of the Chinese em-

bassy during Operation Allied For

ce in 1999, for instance,

complicated further military operations and r

equir

ed mor

THE PUSH AND PULL OF INTELLIGENCE REFORM

122

sustained effort by national leaders to maintain NATO com-

mitment to the operation.

By some accounts this balance has shifted from the situation

prior to the 1991 Gulf War, where too much emphasis was

placed on national requirements, to excessive emphasis on sup-

port to military operations, to a relatively balanced situation in

2004.

After the Gulf War new structures such as joint intelli-

gence centers, deployed and integrated organic intelligence as-

sets such as unmanned aerial vehicles, and upgraded commu-

nications architectures all helped ensure greater support to

military operations.

However, there were concerns that sup-

port might be shifting too much in favor of departmental inter

ests. The

IC21 HPSCI staff study of 1996 cautioned, “Thus, a

balance needs to be struck. Urging an increased emphasis on

SMO [support to military operations] without looking across the

board at all IC requirements runs the risk of leaving many other

ongoing policy needs partially or completely unfulfilled.”

Actions to adjust this balance have largely fallen to internal

IC or

ganizations, such as those supporting the assistant direc-

tor of central intelligence for collection (ADCI/C). Furthermore,

the r

ecalibrations that have occurred over the last 15 years sug

gest that both sides have strong advocates within the IC and

that each has been at least partially effective in adjusting the

underlying balance to address competing intelligence needs.

In the end, legislative measures may not be the best way to

fect changes in this particular balance because they suffer

from both their relative lack of precision and their inability to

make dynamic adjustments.

Legislation might change orga-

nizational structures and authorities in an attempt to shift the

balance in favor of national requirements over tactical or de-

partmental ones. This would give greater control over intelli-

gence resources to a DNI, and a dramatic recalibration might

be possible. However, as discussed earlier, such a recalibration

could impel the departments to develop organic intelligence

capabilities if they per

ceive intelligence support fr

om the

oader IC as inadequate.

The relative imprecision of congressional action may mean

that the best mechanisms for adjusting this balance r

eside

within the IC, such as coor

dinating bodies like the National In

THE PUSH AND PULL OF INTELLIGENCE REFORM

123

telligence Collection Board (NICB).

An internally oriented ap-

proach has several advantages. First, the bulk of the 9/11 Re-

port’s findings, as well as studies on prewar intelligence about

Iraq, found far less fault with how the IC has balanced the

competing interests of the national decision makers and the

departments than with other aspects of the IC’s operations.

This suggests that significant recalibration may not be re-

quired at this time. Second, internal processes offer far greater

precision and responsiveness. Interagency approaches can

shift priorities by weighing the competing interests of national

decision makers and the departments against current reali-

ties.

Thus the balances can be adjusted in near real time in

response to world events, cr

eating a far mor

e responsive and

dynamic process.

Other Balances

Other significant balances not examined in this study are

nonetheless important as reform efforts progress. First, both

Congress and the 9/11 Commission recognized the sensitive

balance between granting the government additional powers

so that it could better protect its citizens and infringing upon

the privacy and civil liberties of those same citizens and of “US

persons.”

The distribution of r

esources within the commu

nity involves another balance. Her

e the DNI, the department

heads, the president, and Congress will have to balance the

need for a strong analysis community with the need to acquire

and maintain intelligence collection capabilities. Given the IC’s

finite resources, it is unlikely that the full range of either ana-

lytic or collection requirements can ever be filled. Determining

the best allocation of resources thus becomes yet another bal-

ancing act.

An earlier section outlined the competing demands associated

with staffing both interagency centers such as the NCTC and the

IC’s constituent organizations. Determining the proper relation-

ship between the intelligence analyst and the policy maker r

ep-

esents yet another balance. Close pr

oximity to the policy maker

ovides the analyst with better insight into the policy maker’s

intelligence needs and r

equir

ements, but carries the danger that

THE PUSH AND PULL OF INTELLIGENCE REFORM

124

intelligence analysis can be politicized and altered to conform to

the policy makers’ preferences.

Balancing in an Interagency Context

In addressing the balances, Congress and the 9/11 Commis-

sion chose in large part to apply matrix-like organizational con-

structs.

This is not surprising, given the interagency nature of

the IC, with its dual-hatting of key officials and dual responsi-

bilities of several senior managers. Thus the heads of CSAs such

as NSA and NGA had and continue to have the dual responsibil-

ities of providing sufficient support to military operations while

meeting the intelligence requirements of national policy makers.

Creation of the NCTC further strengthens the trend toward

matrix-like organizations by drafting personnel fr

om across

the IC. Applying a business analogy to the NCTC, counter-

terrorism intelligence is the product line that relies on the re-

sources of the various functional organizations of the IC (FBI,

CIA, DIA, etc.) to succeed. Given the distribution of intelligence

resources within the US government, creating a new centralized

structure would require either considerable expansion of the IC

in terms of manpower and resources or the transfer of organic

intelligence r

esour

ces from the departments to the centralized

structure. Neither option is particularly appealing.

However, matrix organizations have their own drawbacks,

several of which are relevant to the newly reorganized IC.

Some matrix organizations tend to drift towar

d disorder

, “where

people do not recognize a ‘boss’ to whom they feel responsi-

ble.”

For example, personnel assigned to the NCTC may feel a

dual allegiance: to the NCTC and to their home organization.

Without explicit agreements between the agencies that provide

the manpower and the NCTC, the pr

oblems of dual allegiance

could impact overall effectiveness. The DOD addresses this

problem via specific command relationships that grant opera-

tional control over forces to JFCs and grant administrative con-

trol to the services.

In essence, service personnel normally

know which boss provides their daily tasking and which one

provides their administrative support. The DOD has expended

considerable ef

fort on defining command relationships, devoting

THE PUSH AND PULL OF INTELLIGENCE REFORM

125

much of its 118-page capstone document, Unified Action Armed

Forces

, to an explanation of command relations and the various

types of control of military forces.

Similarly explicit procedures

and terminology to describe the relationships between the vari-

ous component entities could yield great benefit to the IC, in

that both leaders and staff operating in joint capacities—

whether in the NCTC or on the DNI’s staff—would better under-

stand their relationships with their owning organization and

their assigned organization.

Power struggles resulting from the competing interests of

the constituent organizations within the IC represent a second

potential pitfall.

Departmental or agency inter

ests may not

coincide with those of the DNI or the NCTC. The Goldwater-

Nichols Act addressed the potential of power struggles by ad-

justing the power relationships between the combatant com-

manders and the services. By providing greater authority to

the combatant commanders, the act short-circuited a repeat

of the struggles the combatant commanders had previously

had with the services. Such power struggles ar

e inevitable;

therefore, it will be important to define the power relationships

of the various leaders within the IC. The incr

eased budgetary

authority of the DNI or the still unclear authorities of the

NCTC director may not suf

fice to balance the power relation

ships with the constituent intelligence organizations; this is

one area that may require continued recalibration by Congress

or the pr

esident.

Finally, organizational agility will suffer if consensus among

all participating organizations is required to reach decisions:

either the decision-making process will be unduly cumber-

some or the implementation of the resulting decisions will be

excessively delayed.

This was a criticism of the conduct of

the pre-1986 JCS, where a consensus approach both slowed

the staffing process and resulted in the lowest-common–de-

nominator approach to policy recommendations. To address

this, the Goldwater

-Nichols Act placed gr

eater power in the

joint staf

f, thereby removing the services from joint staff ac-

tions, and made the CJCS the principal advisor to the presi-

dent on military matters, ef

fectively reducing the incentive to

each consensus.

THE PUSH AND PULL OF INTELLIGENCE REFORM

126

The degree of coordination and consultation expected and

required from the individual intelligence agencies with regard

to the NCTC or the DNI’s execution of community manage-

ment activities poses a danger to the IC. Greater operational

friction could result if individuals temporarily assigned to the

NCTC feel obligated to seek permission from their owning

organization before taking an action. Likewise, pressure to

reach consensus could slow the decision-making process and

dilute the quality of overall decisions. Again, the IC could miti-

gate such situations by clearly and explicitly articulating com-

mand relationships, to include the respective responsibilities

and authorities of participating organizations for both opera-

tional and administrative matters.

Summary

Reforming an organization as large and diverse as the IC de-

fies easy solutions. If anything, solutions that address only one

aspect of the underlying issues may actually hinder reform be-

cause such solutions fail to account for the competing balances

involved within the larger system. Greater centralization may be

a good solution until it starts to degrade the ability of the vari-

ous departments to obtain the type of tailor

ed intelligence r

quired to support their unique missions. Centralizing control

over intelligence resources and assets provides an excellent way

to reduce duplication and control costs. However, if organiza-

tions that depend on intelligence for operational success are dis

satisfied with the amount of support they receive, they will be

more likely to build indigenous capabilities to better control ac-

cess to intelligence and to tailor that intelligence to their unique

operational needs. Centralizing intelligence analysis via entities

like the NCTC should help impr

ove infor

mation sharing and ad

dress the connect–the–dots problem; however, without proper

safeguards a diversity of analytic opinion could be hindered or

suppressed. Similarly, greater information sharing is a worthy

objective until valuable intelligence sources and methods are

compromised, which would result in less accurate and reliable

information to analyze and report. Finally, balancing the intelli-

gence needs of the departments with those of national decision

THE PUSH AND PULL OF INTELLIGENCE REFORM

127

makers remains an enduring issue and one that may require

continued and dynamic recalibration, because it requires both

long-term and short-term responses. Long-term activities in-

clude determining the appropriate structure and resource re-

quirements to satisfy both competing needs. In the shorter

term, the balance must be struck within the geostrategic con-

text prevailing at the time.

Both the legislative and executive branches have taken meas-

ures to adjust the balances described in this chapter. Their ac-

tions focused primarily on one part of the organizational con-

text: the structure. While critically important, structure is

nor

mally only one contributor to a successfully functioning

organization. The success of intelligence reform will probably

depend on other factors, such as the design of new internal or-

ganizational structures and modification of existing ones, the

underlying support infrastructure, and the quality of IC’s per-

sonnel and leaders.

Put another way, legislative measures and

executive orders have established the framework within which

the IC must work to ef

fect the desired reform. With the struc-

tural framework established, how the IC structures itself inter-

nally, the changes it makes to existing policies and pr

ocedures,

incentive systems, and the underlying infrastructure, and the

effectiveness of its leaders and quality of its workfor

ce will be-

come key determinants to success.

Again, there are analogies with DOD reform. Goldwater-

Nichols established an initial framework that for

med the basis

of continued reform efforts within the DOD. From this new

baseline, the DOD conceived new joint organizational con-

structs and coupled these organizations with doctrine and pol-

icy that explicitly articulated responsibilities, functions, and

command relationships. The act altered the structure of the

DOD principally by changing the authority relationships of

key actors—increasing the authority of the combatant com-

mands and changing the services’ role with regard to the JCS.

Contextual changes also occurr

ed. New policy was developed

in the for

m of joint doctrine, directives, and instructions ad-

dressing a broad range of topics from personnel to operations

to planning.

Incentives, particularly those r

elated to joint

duty, wer

e modified first by the actual legislation and then by

THE PUSH AND PULL OF INTELLIGENCE REFORM

128

the personnel systems of the individual services until much of

the stigma associated with joint duty was removed. Renewed

emphasis was given to system interoperability and greater

standardization of C

systems.

Additionally, leadership and the quality of personnel played

a role. The leaders, particularly service leaders, had discretion

as to how wholeheartedly they supported the legislation.

While there were undoubtedly skeptics, it appears that a large

majority of senior leaders made good–faith efforts to implement

reform and that their efforts were largely successful. All of

their efforts could have been for naught without a strong team

to implement the r

eforms. Thus, at least some of the credit for

the military’s success in military operations since 1986 should

be attributed to the quality and training of personnel involved.

The preceding discussion is not meant to marginalize the sig-

nificance of the Goldwater-Nichols legislation. Instead, it sug-

gests that legislative efforts do not guarantee complete success.

Therefore, the success of the latest round of intelligence reform

may depend much mor

e on the early steps that occur within the

IC. Key determinants will be the internal organization of the

NCTC and any other centers and their r

elationships with the

contributing organizations within the IC. The length of assign-

ment, the quality of personnel assigned, and the authority of a

center’s director over individuals assigned from outside agen-

cies are just a few of the elements that will likely play a role in

the ef

fectiveness of any center’s operations.

A related issue concerns the changes made to the under-

lying support structures. The Goldwater-Nichols experience

suggests the importance of explicit governing policies, particu-

larly policy related to establishing responsibilities, functions,

and command relationships. This is perhaps even more cru-

cial for the IC than for the DOD, given the interagency nature

of the IC at large and of the NCTC specifically. In other policy

areas, security regulations may need to be modified to facili-

tate gr

eater infor

mation sharing. Changes in personnel incen

tive systems ar

e almost certainly essential. Goldwater-Nichols

offers examples for encouraging and promoting joint/intera-

gency duty; other incentive systems r

elated to information

sharing and impr

oved analytic tradecraft may need to be ad

THE PUSH AND PULL OF INTELLIGENCE REFORM

129

justed as well.

For example, the IC may need to adjust and

standardize personnel systems to better measure performance

across the enterprise and reward desired behavior consis-

tently. Similarly, the supporting infrastructure, particularly

the communications infrastructure, will likely require alteration

to support greater information sharing. While simple in the-

ory, execution could be decidedly more difficult given the

number of departments and agencies involved.

Finally, leadership will play a decisive role. The leaders of

the top-level or

ganizations, such as the DNI and the head of

the NCTC, will be faced with implementing a multifaceted ap-

proach to enable success. Designing sound internal struc-

tures, creating a strong mission focus and vision, developing

and adjusting policies and procedures, modifying support in-

frastructures, and—perhaps most difficult of all—achieving

the necessary cooperation from multiple department heads,

with only limited authority, are just some of the tasks await-

ing these new leaders. The intelligence reform legislation rep-

resents another step in the evolution of the IC. Ultimately,

whether the legislation achieves its intended ef

fect r

ests, in

lar

ge part, on implementation by the larger IC.

Notes

1. Michael V. Hayden, interviewed by the author, 18 February 2005. This

debate can thus be thought of either in terms of the structural considera-

tions (more or less centralization/decentralization) or in terms of likely re-

sult (more or less cohesion/autonomy).

2. The Central Intelligence Group (CIG) was officially formed by President

Truman in 1946 and had started the movement towards centralization; how-

ever, the bulk of the resources and efforts were present in longer-standing or-

ganizations. The CIG was replaced by the CIA with the passage of the

Na-

tional Security Act of 1947.

See Prados, Intelligence Community Leadership,

3 (see chap. 3, n. 2).

3. George Nesterczuk, “Successful Start for the Department of Homeland

Security Requires Management Flexibility,” Heritage Foundation Back-

grounder no. 1572, 19 July 2002, 2, http://www.heritage.or

g/Research/

HomelandDefense/BG1572.cfm (accessed 5 January 2005).

4. Department of Homeland Security, “The Department of Homeland Se-

curity,” June 2002, 2; and “History: Who Became Part of the Department,”

http://www.dhs.gov/dhspublic/interapp/editorial/editorial_0133.xml (ac-

cessed 10 April 2005).

Schlesinger Report, 9 (see chap. 3, n. 29).

THE PUSH AND PULL OF INTELLIGENCE REFORM

130

. This reference is to power associated with position or associated gov-

erning rules, policy, or law.

Congressional Record, Senate, 108th Cong., 2d sess., 8 December

2004, S11970. Increasing accountability for the IC appears to have been a

central concern of Congress. For example, Senator Collins states, “This pro-

vision [increased tasking authority] is another example of Congress’s intent

to create a strong DNI with sufficient authority to manage and be account-

able for the Intelligence Community, including those elements within the

Department of Defense.”

8. Harvey M. Sapolsky, “Equipping the Armed Forces,” in G. Edwards

and E. Walker, eds.,

National Security and the U.S. Constitution (Baltimore,

Md.: Johns Hopkins University Press, 1987), 129. Sapolsky suggests that

continued centralization within the DOD is not without drawbacks. Decen-

tralization may encourage duplication, but this same duplication compli-

cates the task of the adversary who must be able to counter a wider variety

of systems and forces.

9. John Hamre, “Testimony to House Armed Services Committee,” 108th

Cong., 2d sess., 11 August 2004, FDCH e-Media Transcripts, http://www.ap

digitalnews.com/partnerships.html. Dr. Hamre suggests the option, which he

did not favor, of moving the NSA, NRO, and NGA under the DNI to give the DNI

institutional strength. See also William Odom,

Fixing Intelligence, 25 (see chap.

3, n. 83). General Odom suggests a slightly modified version, with the IC re-

organized by function under a DCI (SIGINT, IMINT, HUMINT, and counterin-

telligence). The effect would be roughly the same as if the large CSAs (NSA,

NRO, NGA) were removed from under the DOD.

10. See, for example, William Odom, “Testimony to the House Armed Ser-

vices Committee,” 108th Cong., 2d sess., 11 August 2004. General Odom, a

former director of the NSA, stated in his testimony, “If you pull those agen-

cies out of the Defense Department, I can tell you the Defense Department

will abandon them because two-thirds of the personnel in NSA belong to the

military. And they will take them and recreate their own units.”

11. House Armed Services Committee, “The Implications of the 9/11

Commission’s Recommendations for the Department of Defense,” 108th

Cong., 2d sess., 11 August 2004.

12. This mind-set is still present long after Goldwater-Nichols. The

Marines routinely successfully withhold a large subset of their assets from

direct tasking by the JFACC instead keeping those assets to support Marine

ground operations. For example, in the Gulf War, the Marines withheld 50

percent of their FA-18s and all of the AV-8B Harriers. See Gordon,

The Gen-

erals’ War

, 311 (see chap. 2, n. 65). As a general rule, the use of Army air

assets is contr

olled entirely by the joint force land component commander

(JFLCC) to ensure necessary support to ground operations.

13.

illiam Odom, “T

estimony to the House Per

manent Select Commit

tee on Intelligence,” 108th Cong., 2d sess., 4 August 2004. “I think you’ll see

the services creating their own new NSA, their own new imagery agency.

You’re going to just end up with a big mess.” See also Charles E. Allen, “Tes-

THE PUSH AND PULL OF INTELLIGENCE REFORM

131

imony to the House Permanent Select Committee on Intelligence,” 108th

Cong., 2d sess., 4 August 2004. Allen spoke about the need for both national

intelligence support and for indigenous support from the military services,

stating, “At the same time we expect the Department of Defense and the

military services to also continue to build their own organic support.”

14. Schlesinger Report, 28 (see chap. 3, n. 29).

15. Harold Ford, “The Primary Purpose of National Estimating,”

Studies in

Intelligence

, 1991, 72, http://www.cia.gov/nic/PDF_GIF_anal_meth/trade

craft/purpose_of_estimating.pdf (accessed 23 March 2005). In describing con-

ditions prior to the Pearl Harbor intelligence failure, Ford states, “The many

significant off-chance bits of the puzzle . . . were scattered around and among

these many entities and were never shared, brought together, and assessed.

Tragedy ensued because there was no mechanism to collect and cooperatively

examine every scrap of pertinent all-source intelligence.”

16. Schlesinger Report, 45–46 (see chap. 3, n. 29). See also Church Com-

mittee, book 1, 443–44 (see chap. 3, n. 40). The report discusses such op-

tions as expanding the time analysts can remain in analysis vice manage-

ment positions and injecting personnel into the analyst community from

outside the traditional confines of the IC. Recommendation 34 describes ex-

panding analysts’ tours of duty in other organizations and academic institu-

tions to “broaden both their analytic outlook and their appreciation for the

relevance of their analysis to policymakers.” Also John Gannon, interviewed

by the author, 30 November 2004. In describing steps to improve intelli-

gence estimates and analytic products, Gannon stressed the importance of

integrating outside sources of expertise and stated that a key difficulty will

likely be determining who can best answer the analytic question at hand.

The Aspin-Brown, 86 (see chap. 3, n. 7), also suggested increasing ties with

experts outside the IC.

17. Balancing the need to accurately convey dissenting viewpoints in ana-

lytic products while providing a decision maker with actionable information

is yet another tension affecting the operations of the IC and one that this re-

port will not cover in great detail. The groupthink problem is to a great deal

dependent upon how conflicting analytic opinions are incorporated into final

products; therefore, the answer may not just be decentralized analytic

groups. At one extreme all dissenting opinions could be listed within a prod-

uct or report providing a decision maker with little actionable information

with which to base a decision. On the other hand, the consolidating body

could mar

ginalize dissenting views to such an extent that such views are ei

ther not explicitly expressed or excessively discounted to the decision maker

so that the overall effect is the appearance of unified analytic opinion. Ford

may have stated it best, “It is incumbent on an Estimate’s producers . . . to

ensure that dissenting or differing judgments are presented sharply and un-

ambiguously. Split decisions, clearly stated, are worth far more to policy

making consumers than is coor

dinated mush.” For

d, “The Primary Purpose

of National Estimating,” 77.

THE PUSH AND PULL OF INTELLIGENCE REFORM

132

ntelligence Reform and Terrorism Prevention Act of 2004

sec. 1017

(see chap. 4, n. 33). For a more negative view of red teams, see Sherman

Kent, “A Crucial Estimate Relived,”

Studies in Intelligence 36, no. 2 (1992):

111–19. Kent, who served as chairman of the Board of National Estimates

from 1952 to 1967, stated, “Since at least World War I, intelligence services

from time to time set a group of individuals apart and instruct them to think

of themselves as the enemy’s general staff. Their task as a red team is to

ponder and act out the way the enemy will respond to situations as they de-

velop. The idea seems to be that by the creation of an artificial frame . . . you

will get a more realistic appreciation of the enemy’s probable behavior. . . .

It does not necessarily follow.”

19.

Intelligence Reform and Terrorism Prevention Act of 2004, sec. 1019

(see chap. 4, n. 33).

20. Gannon interview. Separate organizations or oversight entities may

not be the entire answer. In describing important elements to competitive

analysis, Gannon stressed the importance of internal processes, such as

getting analysts to question their assumptions that were at least equally im-

portant to improving analytic rigor. In other words, competitive analysis may

not need to rely on separate analytic teams, rather internal measures, if em-

ployed correctly, may be equally effective. See also Ford, “The Primary Pur-

pose of National Estimating,” 76: “the business of analysis is more art than

a science and the skills required of the estimators are in large part a matter

of experience and ’feel.’”

21. Federal Bureau of Investigation, “Affidavit in Support of Criminal Com-

plaint, Arrest Warrant and Search Warrants,” press release, 20 February

2001, 3–4, http://www.fbi.gov/libref/historic/famcases/hanssen/hanssen

.htm (accessed 7 January 2005). According to the affidavit, Hanssen was al-

leged to have “compromised numerous human sources. . . . Three of these

sources were compromised by both Hanssen and former CIA officer Aldrich

Ames, resulting in their arrest, imprisonment, and, as to two individuals, ex-

ecution.” In addition Hanssen was alleged to have “compromised numerous

FBI counterintelligence investigative techniques, sources, methods and oper-

ations.”

22. Jennifer Sims, “What is Intelligence? Information for Decision Mak-

ers,” in Roy Godson, Ernest R. May, and Gary Schmitt,

U.S. Intelligence at

the Crossroads: Agendas for Reform

(Washington, D.C.: Brassey’s, 1995), 8.

A further need to protect information is that compromised information can

provide adversaries with important insights into the policy maker, 12.

23.

Intelligence Reform and Terrorism Prevention Act of 2004, sec.

1016(d)(1)(see chap. 4, n. 33).

24.

See Kaplan, “Mission Impossible” (see chap. 3, n. 2). John Gannon,

former chairman of the National Intelligence Council, is quoted as saying,

“The r

ewar

d systems ar

e all based on pr

otection. . . . Y

ou don’t come into

work saying how can I share better today.”

25. J. Richard Hackman,

Leading Teams: Setting the Stage for Great Per-

formances

(Boston: Harvard Business School Press, 2002), 135.

THE PUSH AND PULL OF INTELLIGENCE REFORM

133

6. Lee Hamilton, “Testimony to the House Select Committee on Home-

land Security,” 108th Cong., 2d sess., 17 August 2004. Hamilton speaks to

the lack of incentives stating, “As we say in our report, there is this very

strong urge, which every one of you has encountered, to overclassify. Look,

a document comes before a person who has the authority to classify or not

classify. He or she looks at that document, there is no incentive for that per-

son to make it public. The incentive is to classify it and protect himself from

the possibility of the information getting out and causing a problem. So they

stamp it, “secret,” and we pile up enormous amounts of information, ware-

houses full of information that are secret because the incentives are all on

the side of classification.”

27. See “Testimony of William Crowell, Markle Foundation, before the

House Government Reform Committee,” 108th Cong., 2d sess., 24 August

2004, FDCH Transcripts. “Decisions about sharing intelligence in the govern-

ment today are still made largely in the context of a system of classification

that was developed during the Cold War. During the Cold War, the use of in-

formation was dominated by a culture of classification and tight limitations

on access in which information was shared only on a need-to-know basis.”

28. David Broder, “Shaping our Defense,”

Washington Post, 13 January

2005, A21, describes some information sharing practices at the DHS, in-

cluding large signs in the DHS operations center that read, “Our Mission: To

Share Information.” While signs alone will not effect greater information

sharing, such seemingly small steps may, over time, create underlying con-

ditions that gradually result in the desired cultural shift.

29. Dulles, “The Central Intelligence Agency,” 66 (see chap. 3, n. 7). In

one of the earliest references to national intelligence, the Dulles-Jackson-

Correa report refers to NSC Intelligence Directive No. 3, Annex 9, which de-

fines

national intelligence as “integrated departmental intelligence that cov-

ers the broad aspects of national policy and national security, is of concern

to more than one Department or Agency, and transcends the exclusive com-

petence of a single Department or Agency or the Military Establishment.”

30. Pappas and Simon, “The Intelligence Community: 2001–2015,” 4 (see

chap. 4, n. 91).

31. At the top level the larger intelligence budget is comprised of the

newly renamed National Intelligence Program (NIP), JMIP, and TIARA. For a

more complete discussion on the intelligence budget, see Stephen Daggett,

“The U.S. Intelligence Budget: A Basic Overview” (Washington, D.C.: CRS, 24

September 2004).

32. Maj Gen Paul J. Lebras, interviewed by the author, 3 December

2004. This delineation has the added advantage that it alters the distinction

between national and DOD (support to military operations) r

equirements.

Other departments and organizations, particularly the Department of Home-

land Security and the newly for

med NCTC, also have significant near

-ter

granular intelligence requirements.

33.

IC21, 241 (see chap. 3, n. 97). This study describes support to mili-

tary operations as involving three pillars: “support to the defense policy-

THE PUSH AND PULL OF INTELLIGENCE REFORM

134

aker; support to force modernization and planning; and support to the

warfighter.”

34. Richard J. Best,

Intelligence Community Reorganization: Potential Ef-

fects on DOD Intelligence Agencies

(Washington, D.C.: CRS, 21 December

2004), 3. Best defines

national intelligence as intelligence that is of concern

to more than one department or agency and provides the basis for national

security policy making.

35.

IC21, 242 (see chap. 3, n. 97). In addressing the balance between na-

tional and support to military operations, the study’s authors state, “More-

over, the term [

support to military operations] suggests that the primary

focus on intelligence should be on the actual need to use force (i.e., ‘fight a

war’), when we continue to believe the successful foreign and national secu-

rity policy is designed to preclude such an event if at all possible.”

36. Lebras, interview. Major General Lebras suggests that not losing

forces in combat is a national priority and thus in many respects military in-

telligence requirements are an extension of national policy.

37. “Testimony of William Odom, former director of the National Security

Agency, before the House Permanent Select Committee on Intelligence,”

108th Cong., 2d sess., 11 August 2004, FDCH e-Media Political Transcripts.

In speaking about conditions prior to 1990, General Odom stated, “One of

my major problems was getting the civilian force inside NSA to agree to do

these real-time tactical support missions. Move them outside of OSD and

they will abandon them entirely. We’ll go right back to the prioritization

problem where if you wanted to shift civilian assets to deal with a crisis on

Saturday, they say, ’Well, we can’t do that until Monday morning, until the

Intelligence Committee . . . meets and reestablishes the collection priority.’

It took years to overcome that.” See also “Testimony of Charles Allen, Assis-

tant Director of Intelligence for Collection before the House Permanent Se-

lect Committee on Intelligence,” 4 August 2004, FDCH e-Media Political

Transcripts. In speaking about conditions in the 1990s up to 2004, Mr. Allen

stated, “I think right now we have a good balance. I think in the 1990s there

was a swing where we were thinking so much about military support some-

times we’ve lost some of our capabilities to collect information to analyze

against these global threats, asymmetric means of attack and the kinds of

global coverage that was missing [when the balance shifted to the DOD] that

has to be done.”

38.

IC21, 250 (see chap. 3, n. 97).

39. Ibid., 246.

40. Stansfield Turner, testimony to the Senate Governmental Affairs

Committee,

Reorganizing America’s Intelligence Community: A V

iew from the

Inside

, 108th Cong., 2d sess., 16 August 2004. For

mer DCI Turner, in tes-

timony, argued for a less formal process to recalibrate this balance stating,

“But we do want to have a system that ensur

es that [a] military commander

gets the best support we can possibly afford to give within the limits of also

putting the number one priority on terrorism. And I don’t know that you can

write that into a law. I think that’s got to be ironed out by presidents and

THE PUSH AND PULL OF INTELLIGENCE REFORM

135

ecretaries of Defense and national intelligence directors as to how they bal-

ance that out individually.”

41. For a good discussion of the drawbacks of granting the DCI greater

authority of these resources, see Aspin-Brown, 52–53 (see chap. 3, n. 7).

42. See the description of the ADCI/C on the CIA’s public web site http://

www.intelligence.gov/1-management.shtml (accessed 12 January 2005).

The NICB is “the senior IC advisory board and forum for achieving consen-

sus on national collection issues. It serves as the IC’s overarching mecha-

nism for developing and implementing cross-discipline, cross-agency coordi-

nation of intelligence collection. The Board provides the ADCI/C an

integrated forum for collection decisions on current and standing require-

ments that are timely and meet the needs of the DCI.”

43. See Allen,

Testimony to the House, who describes the operations of

the NICB, “Senior managers who sit on the Collection Board are empowered

by their respective collection disciplines to speak with authority for agency

heads and to initiate changes in their collection capability as directed by the

board. These managers have developed into a cooperative team, taking on

any intelligence problem and working closely together to address it. In short,

we’re problem-centric, collaborative and task-oriented. We address our criti-

cal intelligence issues across the spectrum, from crises to enduring hard

targets.”

44. Ibid. In testimony Mr. Allen stated, “I convened a Collection Board

meeting every day to manage the collection surge during Operation Iraqi

Freedom. We continue to meet twice a week on Iraq to ensure our collection

capabilities are focused appropriately and the collectors are seamlessly in-

tegrated with each other and with field elements.”

45. See, for example, House, Permanent Select Committee on Intelli-

gence,

9/11 Commission Recommendations: Counterterrorism Analysis and

Collection, Requirement for Imagination and Creativity

, 108th Cong., 2d sess.,

4 August 2004. Rep. Porter Goss said, “Because this is really the critical de-

bate before our country; it’s the balance between civil liberties and the pro-

tection of American citizens. That is the debate, and that debate is going to

go on for a long time. And my guess is society will evolve and we’ll never get

it exactly right; we’ll always be off the mark a little bit one way or the other.”

The distinction between a US citizen and a US person is a subtle, yet impor-

tant one. EO 12333 defines a

US person as “a United States citizen, an alien

known by the intelligence agency concerned to be a permanent resident

alien, an unincorporated association substantially composed of United

States citizens or permanent resident aliens, or a corporation incorporated

in the United States, except for a corporation directed and contr

olled by a

for

eign government or governments.” In essence, the use of the term

US per

son

broadens the pool of individuals guaranteed protection.

46.

Sims, “What is Intelligence?” 11. “Intelligence must live with an

irony: the relevance and timeliness of intelligence are enhanced if intelli-

gence analysts work side by side with those they serve, while such proxim-

ity may be the greatest single source of temptations to privatize and politi-

THE PUSH AND PULL OF INTELLIGENCE REFORM

136

ize intelligence.” For some counters to the arguments to the inverse rela-

tionship between objectivity and proximity to policy makers, see Abram

Shulsky and Gary Schmitt, “Intelligence Reform: Beyond the Ames Case,” in

Godson,

U.S. Intelligence at the Crossroads, 53: “The absence of a tie to de-

partmental policymakers does not, and cannot, mean ‘independence’: intel-

ligence always works for, and is subordinate to, some policymakers.” “The

question of objectivity and bias is much more complicated than it might ap-

pear at first glance. A truly open mind is difficult to retain under any cir-

cumstances, and mere bureaucratic arrangements cannot guarantee it.” See

also MacEachin, “The Tradecraft of Analysis” in Godson,

U.S. Intelligence at

the Crossroads

, 74: “Fundamentally, the only guard against politicization

lies in the integrity of the supervisors and analysts.” See also Paul Wol-

fowitz’s commentary, 79. “In my own experience, the most successful input

of intelligence for policy has been when intelligence professionals have been

close to the policy process.”

47. Joan Dempsey, interviewed by the author, 30 November 2004. This

is not the first time the IC has relied on matrix-like organizations as various

intelligence centers. Arguably the precursors to the NCTC and associated

national intelligence centers have been a part of the IC for nearly 20 years.

48. Stanley M. Davis and Paul R. Lawrence, “Problems of the Matrix

Organization,”

Harvard Business Review, May–June 1978, 131–42. In this

early examination of matrix systems, the authors describe several patholo-

gies that are still relevant today.

49. Ibid., 132.

50. Dempsey, interview. Dempsey highlighted that personnel assigned to

current intelligence centers often maintain loyalty to parent organizations, a

problem exacerbated because the centers do not “own” the people assigned.

51. See Joint Publication 3-0,

Doctrine for Joint Operations, 2-8, 2-10 (see

chap. 2, n. 16).

Operational control (OPCON) is “the authority to perform

those functions of command over subordinate forces involving organizing

and employing commands and forces, assigning tasks, designating objec-

tives, and giving authoritative direction necessary to accomplish the mis-

sion.”

Administrative control (ADCON) is “the direction or exercise of author-

ity over subordinate or other organizations in respect to administration and

support including organization of Service forces, control of resources and

equipment, personnel management, unit logistics, individual and unit train-

ing, readiness, mobilization, demobilization, discipline, and other matters

not included in the operational missions of the subordinate or other organ-

izations.”

52. See Joint Publication 0-2,

Unified Action Armed For

ces

(see chap. 2,

n. 28).

53. See also, Davis, “Problems of the Matrix Organization,” 133. “A ma-

trix does not leave such matters [r

elationships between functional and pr

uct managers] in an indefinite status; it is a definite structure and not a ‘free

form’ organization.”

54. Ibid., 134.

THE PUSH AND PULL OF INTELLIGENCE REFORM

137

5. Ibid., 140. The authors use the term

ecision strangulation

o de-

scribe conditions where efforts to achieve buy-in from across functional

areas can inhibit the decision-making process.

56. The CJCS is required to provide dissenting service viewpoints to the

president in a timely manner, so there remains some incentive to maintain

a unified front. Because the service chiefs are no longer primary advisors to

the president, their willingness to challenge the chairman’s advice to the

president on all but the most important issues appears to be rather low.

57. For example, see Hackman,

Leading Teams, xi, who describes five

enabling conditions that are necessary to create the conditions for organiza-

tional success: a real team, compelling direction, enabling structure, sup-

portive context, and expert coaching.

58. Ibid., 134, describes three key elements (not exclusive) in establish-

ing a supportive context: the reward system, the information system, and

the educational system. I have added a fourth in recommending that the

policies and procedures governing the activities of the IC in adjusting the

balances is a particularly important element in the present case.

59. The sheer quantity of instructions, doctrine documents, and direc-

tives is impressive. To get some idea of the scope and number of different

documents, see http://www.dtic.mil/doctrine/index.html (accessed 13 Jan-

uary 2005). For example, there are 65 documents and instructions related

to the joint staff (see Joint Staff, Department of Defense, “Current List of

Joint Staff Documents, Instructions, Manuals, Guides, and Notices,” 14 No-

vember 2004, http://www.dtic.mil/cjcs_directives/support/js/jsirpt.pdf

(accessed 13 January 2005). This is not to argue that the number of sup-

porting documents is a key element of success; however, some form of ex-

plicit policy and procedural formulation provides a stronger contextual

baseline for the constituent organizations.

60. Daniel Wirls,

Buildup: The Politics of Defense in the Reagan Era

(Ithaca, N.Y.: Cornell University Press, 1992), 173. DOD procurement reform

occurred at the same time as the Goldwater-Nichols debate. In the case of

procurement reform, Congress passed a law requiring the DOD to establish

an Office of Operational Testing and Evaluation, the DOD failed to request

funding for the office, and Secretary of Defense Weinberger failed to nomi-

nate a director as mandated by the law, delaying implementation by 18

months.

61. MacEachin, in Godson,

U.S. Intelligence at the Crossroads, 65–69.

The author describes how incentive structures within the analysis arm of

the CIA created preferences towards specific types of analytic products that

in many cases wer

e not meeting customer requir

ements. However, because

of the reward system, analysts continued to produce reports that provided

the greatest opportunity for advancement at the expense of meeting the in-

telligence needs of the policy maker.

THE PUSH AND PULL OF INTELLIGENCE REFORM

138

Chapter 6

The Question of Balance

But this view [that organizational structure doesn’t matter]

rests on a common error, the confusion of organization with

organizational structure.

—James Q. Wilson, Bureaucracy

Reorganization will, at best, only create conditions in which

wise and imaginative leadership can flourish.

—Schlesinger Report, 1971

We are at the beginning of intelligence reform, not the end.

We are making our first tentative steps toward changing the

way we do intelligence in this country, and we are decades

away from completing it.

—Joan Dempsey

“Intelligence Reform—At Last”

The final r

eport and r

ecommendations of the 9/11 Commis

sion set of

f a flurry of government activity that began with the

issuance of new executive orders and culminated in the pas-

sage of intelligence refor

m legislation in December 2004. While

this might appear to end the debate, it may in fact r

epresent

only the first of several steps towar

ds intelligence refor

m. For

example, by February 2005 lawmakers had already signaled

that the new legislation might need to be tweaked and ac-

knowledged that ambiguity remains.

In addition, the execu-

tive branch and the leadership of the IC must now determine

the scope and nature of the internal measures needed to im-

plement reform. Finally, as the last 50-plus years have shown,

calls for further changes will undoubtedly be raised in the fu-

ture, either in response to perceived failures or as a potential

means to increase efficiency and reduce costs.

As the mandated r

estructuring pr

oceeds and future adjust-

ments are considered, these efforts may benefit from examina-

tion of three topics relevant, even central, to the subject of in-

telligence r

efor

m: (1) the 1986 Goldwater

-Nichols r

efor

m of the

139

DOD and its applicability to the IC, (2) the findings and rec-

ommendations of past efforts to modify the IC, and (3) the

competing interests inherent in the IC that have influenced

the pace and character of change, along with the potential im-

pact of recalibrating the underlying balances. Ultimately,

these balances may be readjusted via legislative mechanisms

that provide more enduring reform or via internal measures

that allow dynamic adjustments in response to evolving con-

ditions.

A Goldwater-Nichols Act

for Intelligence?

Statements by members of Congress, various reports in-

cluding the 9/11 Report, and the testimony of 9/11 Commis-

sioners and others reveal that many stakeholders explicitly

considered Goldwater-Nichols a compelling model for legisla-

tive r

eform of a large bureaucracy. A likely motivating factor

for crafting a Goldwater-Nichols Act for intelligence is that

some of the 9/11 Commission’s findings echo criticisms that

eceded the passage of Goldwater-Nichols. These included (1)

failures of joint action and the related imbalance between re-

sponsibility and authority; (2) structural barriers, including

the dual-hatting of key personnel; and (3) personnel policies

that deterred high-quality people from pursuing joint duty. A

further incentive was that in many respects the Goldwater-

Nichols legislation had its intended effect—perhaps not as

great as that desired by more ardent advocates of change, but

certainly influencing the DOD in the intended direction. Joint

action improved markedly in comparison to similar joint opera-

tions predating the act, the altered structure of the JCS im-

proved the effectiveness of the joint staff and power of the

CJCS, and the quality of joint personnel improved while nega-

tive per

ceptions associated with joint duty decr

eased.

It is important to note that Goldwater

-Nichols reform fo-

cused only on certain aspects of the DOD organization: pri-

marily the combatant commanders, the role and function of

the JCS, and the supporting joint staf

f. Thus, the military in

telligence organizations that make up such a large part of the

THE QUESTION OF BALANCE

140

IC were neither a target of Goldwater-Nichols reform nor sig-

nificantly affected by the resulting legislation. By contrast, the

intelligence organizations within the DOD are a major compo-

nent of any conceivable intelligence reform and are directly af-

fected by any ensuing changes.

A series of operational failures and subpar performances,

including the failed Iran hostage rescue in 1979, the confused

lines of authority that allowed the Marine Corps barracks

bombing in Beirut in 1983, and the invasion of Gr

enada in

that same year

, spurred the calls for DOD reform. Critics at-

tributed many of the identified problems to inadequate au-

thority and unclear command relationships that, if improved,

could better compel and enforce joint cooperation and effort.

To address this, Goldwater-Nichols vested greater authority in

the combatant commanders, primarily by increasing the com-

manders’ line authority over subordinate components and by

developing more explicit command relationships between the

services (which supply the forces) and the joint force com-

manders. Although the combatant commanders still had rela-

tively little authority over service budget processes, their new

authority allowed them and their subordinate commanders to

create new C

structures that became core elements in the

success of military operations after 1986. Importantly, the

structur

e of these organizations has continued to evolve, as

have the doctrine and training supporting the operation of

these C

structur

es.

DOD reform advocates also identified structural issues, par-

ticularly as they related to the service chiefs and the joint staff.

The ensuing debate centered on the dual-hatting of the service

chiefs and the services’ undue influence on the joint staff, which

under

mined the quality of military advice. T

o address these is-

sues, Goldwater

-Nichols changed the or

ganization of the JCS,

incr

eased the power of the CJCS by making the CJCS the pri

mary military advisor to the pr

esident, cr

eated the position of a

vice chair

man to assist the CJCS in the execution of his/her du-

ties, and reduced the services’ influence over joint staff action

by placing the staff directly under the control of the chairman.

Notably, the act did not change the dual-hatting of the service

THE QUESTION OF BALANCE

141

chiefs; instead, it relied on the CJCS to provide independent ad-

vice in the absence of undue service influence.

The debates over DOD reform also addressed the poor qual-

ity of joint personnel. The Goldwater-Nichols Act mandated

changes in the promotion practices of the services, made joint

service a requirement for service in senior grades, and de-

manded fairly detailed reporting by the services to chart

progress in meeting congressional requirements. In consider-

ing similar personnel changes within the IC or other govern-

ment agencies, several aspects of the Goldwater-Nichols Act

deserve highlighting. They include designating a large number

of senior positions as requiring joint duty experience, tracking

the promotion rates of individuals assigned to joint duty posi-

tions to ensure they continue to be promoted at equal or greater

rates than their peers, and, to avoid a revolving door in joint

assignments, prescribing the minimum tour lengths (usually

over two years) needed to receive joint credit.

Similar, Not Identical

While the DOD and the IC had similar motivations for

change, fundamental or

ganizational dif

ferences remain that

affect the options available to r

efor

mers. For example, instead

of refor

ming a single department—albeit one with four sepa-

rate services—reforming the IC requires altering a structure

comprising 15 discrete organizations that fall under multiple

executive departments and have a mixture of civilian and mili-

tary personnel. The IC also faces unique challenges, such as the

need to improve the quality of analysis and increase informa-

tion sharing, which argues against a one-on-one correspondence

and suggests the need for a more tailored approach. Further-

more, these challenges are often interrelated; hence, organiza-

tional changes designed to address one issue may have unin-

tended effects on other aspects of the IC.

The differences between the IC and the DOD reform are per-

haps best illustrated by the or

ganizational r

ealities that affect

the options available to reformers as they address the common

challenges. In the area of joint action, the DOD needed to co-

dinate the activities of the four services mor

e ef

fectively.

However

, a common organizational leader, the SecDef, allowed

THE QUESTION OF BALANCE

142

the DOD to create unified commands that gave one individ-

ual—a unified commander—both the authority and the requi-

site control over forces to plan and execute operations in a

more effective fashion.

The IC faces a more difficult problem: it is attempting to cre-

ate the equivalent of a unified command in an interagency

context. This places unique constraints on the leader of such

an organization—in this case, the new DNI. First, the leader

must mobilize the activities of personnel from across multiple

federal departments and agencies, each with its own culture

and operating principles—in many respects an even more dif-

ficult task than that faced by a combatant commander

. Sec-

ond, the interagency nature of such a structure implies that

the DNI will have far less control over personnel than a com-

batant commander does. Goldwater-Nichols codified that a

unified commander charged with planning and overseeing the

execution of operations should have authority and control over

forces permanently or temporarily assigned to that command.

Both the fundamental characteristics and underlying con

straints of interagency operations, however, suggest that new

interagency constructs may be r

equired over the long ter

m if

the IC is to replicate more closely the command relationships

and authorities currently vested in a DOD combatant com

mander.

While reform advocates identified structural impediments in

both the DOD and the IC, the natur

e of these impediments dif-

fers decidedly. Reformers of the DOD had to decide if dual-

hatting the service chiefs unnecessarily compromised their abil-

ity to provide sound advice and determine how to increase the

independence of the joint staff. Increasing the power and inde-

pendence of the CJCS and placing the joint staff under CJCS

control helped mitigate these concerns. In the IC context, con-

cerns over the dual-hatting of the DCI revolved around span

of control, perceptions of undue influence by the CIA, and the

seemingly inadequate authority of the DCI to manage the IC.

The DCI alr

eady had direct control over a community manage-

ment staff; therefore, the DCI’s insufficient authority to em-

ploy that staf

f effectively was more pertinent than the inde-

pendence of the staf

f fr

om outside influence. While separating

THE QUESTION OF BALANCE

143

the roles of the DCI is relatively easy, increasing the authority

of the leader of the IC (similar to that of a department head) is

more problematic. The interagency structure of the IC also

limits options for improving the overall effectiveness of the

community staff.

Finally, the IC faces unique constraints that affect its abil-

ity to implement uniform personnel practices that encourage

joint duty. Again, the IC’s heterogeneous composition makes it

impossible to replicate DOD joint personnel practices com-

pletely. At least initially the IC will likely have far fewer joint

positions than senior executives within the IC. Defining both

the number of joint positions and the departmental/agency

senior positions requiring joint credit is only one among sev-

eral requirements. Others include tracking promotion rates

across 15 separate organizations, determining a fair allocation

of joint positions within the IC, tracking and enforcing compli-

ance with joint duty requirements, and adequately document-

ing performance in a somewhat uniform manner across mul-

tiple personnel systems.

The objective of a Goldwater-Nichols Act for intelligence—

the successful reform of a large government bureaucracy—is

certainly a worthy one. However, the interagency natur

e and

constraints of the IC complicate any dir

ect application of the

Goldwater-Nichols legislative measures and ensuing organiza-

tional structures. An overview of previous intelligence reform

efforts, their common r

ecommendations, and associated find-

ings can provide some insight into why this is so.

Intelligence Reform: The Last 50 Years

Since the passage of the National Security Act of 1947, mul-

tiple studies and commissions have examined various aspects

and agencies of the IC. A review of the various proposals re-

veals several common themes, including the need to increase

centralization and control of the IC, provide the DCI greater

authority to govern the IC, improve the quality of analysis and

estimates, and improve congressional oversight.

As early as 1949, reformers advocated better coordination

among IC activities. The common rationales for such r

ecom-

THE QUESTION OF BALANCE

144

mendations were to avoid duplication and to centralize plan-

ning, programming, and budgetary processes. Centralized

control over the IC would allow one person, the DCI, to adju-

dicate between competing programs and IC organizations and

thereby create a more efficient and cost-effective organization.

Further, centralization offered the opportunity to develop stan-

dardized personnel practices and policies and focus limited in-

telligence resources on high-priority issues.

It soon became apparent to reform advocates that the feder-

ated nature of the IC meant that the DCI had insufficient au-

thority to compel the hard programmatic and budgetary trade-

fs necessary to improve efficiency. The first recommendations

for increasing the DCI’s authority led to creation of deputies to

relieve the DCI from administrative duties within the CIA and

later evolved into an outright call for a separation between the

DCI’s roles to, in the words of one study at the time, “pinpoint

responsibility” for management of the IC on one individual. In

addition, many studies recommended that the DCI be given in-

eased budgetary and personnel authority. In theory, these

additional powers would enable the DCI to compel cooperation

within the IC.

A third theme revolved around calls for improved intelligence

analysis and estimates. At first these calls resulted fr

om a de-

sire to avoid the conditions that preceded Pearl Harbor, where

intelligence was poorly coordinated among competing intelli-

gence or

ganizations. Initially, dissatisfaction centered on the

perceived unwillingness of departments to share intelligence

with the CIA. Over the years other charges were levied including

excessive politicization, a lack of focus on the consumer/policy

maker’s intelligence needs, and the failure to incorporate dis-

senting views into analytic products. Such recommendations

had a simple rationale: the government wanted and needed the

best possible intelligence analysis to aid policy makers. The so-

lutions proffered have remained consistent over the last 50

years, including impr

oved competitive analysis and the use of

ed teams, increased information sharing among agencies, and

the incorporation of dissenting views into products.

A fourth theme that sur

faced somewhat more sporadically

was the call for gr

eater congr

essional oversight. While such con

THE QUESTION OF BALANCE

145

cerns emerged as early as 1955, they were not taken seriously

until the investigations into intelligence activities in the 1970s,

when they resulted in the establishment of the HPSCI and the

SSCI. However, the controversy did not end there. Various stud-

ies, including the 9/11 Report, have continued to highlight ju-

risdictional difficulties, particularly with the two Armed Services

committees. The reasons for such recommendations included

the need to increase executive branch accountability, improve

congressional expertise in this relatively arcane area, and elimi-

nate fragmented oversight within Congress.

Given this history, it is not surprising that the 9/11 Com-

mission renewed these calls for reform. The more intriguing

question is why such consistent and persistent calls for reform

took over 50 years to implement. There are at least two partial

explanations. First, much actually changed over the 50 years.

The IC became more centralized, additional deputies assumed

some of the DCI’s administrative burdens, the DCI’s authori-

ties over budgets and personnel gradually increased (albeit

marginally), and an IC staff developed to assist the DCI in co-

ordinating IC activities. Congress established permanent intel-

ligence oversight committees. However, the 9/11 Commission

and resulting executive and legislative actions clearly indicate

these measur

es were consider

ed insuf

ficient.

Ultimately, the reasons for the relative stasis in the struc-

ture of the IC may have less to do with bureaucratic intransi-

gence than may at first appear. A mor

e accurate explanation

for the lack of change may rest in the underlying tensions in

herent in the IC and its structure.

A Question of Balance

By the very natur

e and structur

e of the US gover

nment,

conflicting interests will inevitably arise as the checks and bal-

ances inherent in the system come into play. These checks

and balances promote tensions between opposing sides and,

depending upon the relative power among the parties involved,

ultimately encourage varying degrees of compromise. Thus, it

is not surprising that tensions exist with regard to the role and

structur

e of the IC, and that attempts to recalibrate various

THE QUESTION OF BALANCE

146

balances often result in compromise that yields less signifi-

cant change than originally advocated.

Perhaps less evident is that the cycles of conflict and com-

promise occur at multiple levels within the government bu-

reaucracy. First, conflict sometimes arises among the

branches of government. For example, over the years the leg-

islative and executive branches have disagreed over such is-

sues as the extent and authority of congressional oversight,

the authorities of intelligence agencies to conduct intelligence

activities, and the pr

oper structure of the NSC staff to coordi-

nate intelligence activities. Second, conflict is not limited to in-

terbranch disputes. Within the executive branch, the various

departments that support the IC have at various times argued

with the DCI over the proper role and exercise of the DCI’s

power and authority. Issues that include the sharing of infor-

mation between organizations, intelligence support to depart-

mental operations, and the allocation of funding among intel-

ligence organizations have all required, and continue to

require, a degree of balancing among competing interests.

Finally, balancing takes place even within specific depart-

ments and agencies. The allocation of funds within each de-

partment/agency, managerial priorities, and the policies and

ocedures that pr

omote desired behaviors ar

e just some of

the balancing acts addr

essed at the intradepartmental level

and below. The sheer number of parties, competing interests

or tensions, and or

ganizational levels at which trade-offs occur

suggests that simple solutions are unlikely.

This study has discussed four of these tensions: (1) central-

ization versus decentralization, (2) connecting the dots versus

failing to consider alternative analysis, the so-called groupthink

oblem, (3) infor

mation sharing versus information protection,

and (4) meeting the intelligence needs of the national policy

maker versus those of a par

ent department or or

ganization. If a

fundamental tension exists, it may well be the first: that is, the

tension between centralizing and decentralizing for

ces or, put

another way, between cohesion and autonomy. Each of the

other three includes aspects of centralizing/decentralizing

for

ces. For example, infor

mation sharing implies gr

eater cen

THE QUESTION OF BALANCE

147

tralization, while the desire for departmental autonomy in ful-

filling intelligence needs implies greater decentralization.

Since its inception the IC has followed a gradual trend towards

greater centralization, with the creation of a DNI in 2004 con-

tinuing the trend. Multiple reports and studies of the IC have ex-

pressed excellent reasons for greater centralization. For example,

the fundamental functions of training, organizing, and equipping

often benefit from greater centralization. Centralized control of-

fers the opportunity to r

educe redundancy in the acquisition of

new systems, pr

ovide greater coherence in planning and budget-

ing, focus limited intelligence resources on high-priority issues,

and set and enforce uniform policies and procedures. Establish-

ing more uniform personnel systems may assist in the move-

ment of personnel among agencies in the community, ensure

that compensation and other incentives are fairly distributed

across the enterprise, and help to create a more uniform orga-

nizational culture. Finally, centralization allows the government

to assign responsibility for the conduct of organizational activi-

ties—to designate the “quarterback” who the 9/11 Report ar-

gued did not exist.

However, there are also compelling reasons for decentraliza-

tion. Individual organizations often have unique training and

equipping functions that centrally contr

olled processes cannot

fully meet. Gr

eater autonomy encourages specialization; thus,

organizations can develop systems and personnel that meet

unique departmental r

equirements for highly specialized

skills, such as cryptography or imagery analysis. Decentral-

ized analysis ensures that analysts tailor their products to the

needs of the departments and can offer alternative viewpoints

during the development of more comprehensive intelligence

estimates. Decentralization allows gr

eater integration with the

consumers of intelligence, pr

oviding mor

e tailor

ed intelligence

to assist the operations of the supported department, service,

or agency. Even r

edundancy is not always the unqualified ill

so often portrayed. In addition to meeting departmental inter

ests, redundant systems offer a surge capacity (e.g., in time of

war) and the ability to pool resources across organizations in

the face of common thr

eats.

THE QUESTION OF BALANCE

148

A second tension involves the degree to which intelligence

analysis should be centralized to promote connecting the dots

while maintaining the diversity of analytic opinion that dis-

courages groupthink. Centralizing intelligence assets around

enduring problems and improving information sharing—as in

the NCTC approach—should help to create conditions that

foster the development of a more complete analytic picture.

Advocates of centralization argue that such an approach en-

sures that the greatest amount of intelligence is gathered,

shared, and analyzed, thereby enabling more holistic analysis.

Advocates of decentralization point to recent IC failures, such

as flawed intelligence on Iraqi WMD capabilities, as evidence

of the need to maintain a diversity of opinion. Moreover, as

centralization increases the manpower support requirements

for multiagency centers could create a steadily increasing

drain on the finite analytic resources of the individual depart-

ments. This would restrict not only the ability to maintain a di-

versity of analytic opinion but also the ability to make dedi-

cated and specialized analysts available to meet specific

departmental requirements.

In examining the thir

d source of tension, the advantages of

improved information sharing seem patently obvious until

they are weighed against some of the potential consequences

of increased openness. Information sharing yields enormous

benefits, particularly as they relate to crafting more accurate

and complete analytic pictur

es. Judging by the 9/11 Report,

this balance needed to shift more towards the sharing side.

Improved sharing offers the prospect of earlier warning against

threats and the consequent saving of lives as well as more

complete and accurate intelligence to help shape the decisions

of national policy makers. However, such a shift is complicated

by the greater probability of failures to protect information ade-

quately. The risk of loss or compromise of information grows

as information is shared with a wider community. Should

such losses become too gr

eat, they could begin both to jeop

dize the quality and quantity of information received and to

imperil analysis efforts. Thus, recalibration requires a careful

assessment of both the positive and negative implications of

incr

eased infor

mation sharing.

THE QUESTION OF BALANCE

149

A fourth tension—one that threatened to jeopardize passage

of the 2004 intelligence reform legislation—involves balancing

the intelligence needs of the national policy makers against

those of the departments, in most cases the DOD. This so-

called national versus tactical debate is perpetuated by the am-

biguity of the terms and the considerable grey area between

purely national and departmental interests. At first glance,

such a dispute may look like simple bureaucratic infighting;

however, there are legitimate interests at stake, again compli-

cated by finite intelligence resources. On the one hand, the in-

telligence needs of national policy makers should dominate, be-

cause the intelligence they r

eceive shapes their policy choices

and the eventual actions of the departments. Therefore, policy

makers must exert control over community-wide resour

ces.

Conversely, the actions and activities of the departments often

have national security implications, and these actions depend

on intelligence that meets their more specialized r

equirements.

Thus, a feedback loop exists between the decisions of policy

makers and the actions of the departments, with the availabil-

ity of intelligence affecting both. Further complicating the issue

is that the departments can meet only some of their legitimate

and specific intelligence needs with purely departmental intel-

ligence resources; thus, some resources must be shared across

the IC to r

educe costs. Determining the pr

oper allocation of

shared resources to meet sometimes conflicting requirements

is thus central to the debate.

What Can Be Done?

If the prescriptive measures of Goldwater-Nichols do not

apply directly to the IC and if the system is inherently fraught

with numer

ous competing inter

ests, how can the IC be r

formed and who should do it? Fortunately, multiple levels and

different branches of government have a variety of recalibration

mechanisms or tools they can apply to adjust various aspects of

the underlying tensions. This combination of actors, tools, and

targets of reform offers varying degrees of permanence, preci-

sion, and agility as briefly described in the following table.

THE QUESTION OF BALANCE

150

In some cases, comprehensive structural or statutory changes

may of

fer the best way to adjust balances. Her

e the legislative

branch can play an important r

ole, particularly when a shock

to the system is desired. Thus, Congress used legislation such

as the National Security Act or Goldwater-Nichols as a tool to

eate conditions that allowed for fundamental r

efor

m. The

enduring quality of such legislative measur

es suggests that

THE QUESTION OF BALANCE

151

Table 1. Elements in intelligence community reform

Actors Mechanisms Targets Characteristics

Congress • Legislation • Organizational • Enduring

• Oversight structure • Relatively imprecise

• Appropriations • Top-level • Institutional focus

incentive • Large-scale

systems adjustments

• Relative • Threat of

authority of key congressional action

players encourages

executive branch

action

President • Executive • Organizational • Less enduring than

orders structure legislation, more

• Interagency • Top-level formal than internal

processes incentive efforts

• Hire/fire systems • More responsive than

authority over • Relative legislation, less than

IC leaders authority of internal measures

• Granting of key players • Limited in some

status instances by

statutory constraints

DNI, agency • Policy, rules, • Internal • Dynamic, responsive

heads, and and regulations organizational • Precision, finely tuned

senior IC • Personnel structure adjustments

leaders practices (hire, • Internal • Intra- and inter-

fire, promote) processes organizational focus

• Funding • Organizational • Difficult to implement

• Establishment culture large-scale change

and enforcement • Incentive • Greater familiarity

of norms systems with internal

• Support processes and

infrastructure constraints

• Ability to quickly

assess impact of

measures and adjust

as required

such actions should be thoughtfully considered and debated

to ensure understanding of the full implications of the re-

forms. However, legislative measures cannot adjust balances

dynamically. Thus, applying the tool of congressional over-

sight may sometimes be more appropriate and can encourage

the executive branch to utilize its own recalibration mecha-

nisms.

In instances that call for less enduring and more dynamic

adjustments, the executive branch has additional recalibra-

tion tools at its disposal. While executive orders are more eas-

ily rescinded than comparable laws, many have remained in

for

ce for decades, suggesting at least some degree of perma-

nence for broad, less inhibiting orders. Moreover, executive or-

ders can be issued quickly: in just over a month after the 9/11

Report appeared President Bush had issued three executive

orders implementing significant aspects of the report.

Further down the executive branch hierarchy are inter-

agency mechanisms to adjust balances in response to the cur-

rent international environment and consumer requirements

for intelligence. The IC uses an interagency construct to bal-

ance the competing intelligence needs of policy makers and

the departments in response to the ongoing world events.

Such a construct, while lacking some of the permanence of the

mechanisms previously discussed, offers a degree of precision

not afforded by either statutes or executive orders. Finally, in-

tradepartmental and intra-agency measur

es are also avail-

able. For example, changes in departmental or agency policies

and incentive structures can recalibrate the balances be-

tween, for example, information sharing and information pro-

tection. At this level, such measures allow sharply tailored so-

lutions, as direct line managers are often best positioned to

assess the impact of new policies and incentive structures.

The multiple tools available to recalibrate a balance can also

be applied to various aspects of a balance, such as the power

elationships among the competing parties, incentive struc

tur

es and their effect on behavior, and underlying support in-

frastructures. First, the relative power of the competing advo-

cates on each side of the balance can be adjusted. Thus, the

Goldwater

-Nichols Act gave the combatant commanders

THE QUESTION OF BALANCE

152

greater authority over assigned forces, thereby reducing the

role of the services in commanding and controlling joint oper-

ations. In a similar vein, increased budgetary, tasking, or per-

sonnel authority may assist the new DNI to assert manage-

ment control over the IC vis-à-vis the departments. At the

highest levels, of course, power relationships depend upon the

relationship of key leaders with the president.

Second, well-crafted law, policy, directives, and regulations

may also assist in recalibration efforts not by changing power

relationships but by changing the underlying incentive struc-

tures, thereby encouraging or discouraging various behaviors.

For example, Goldwater-Nichols changed the incentives for joint

duty, thereby improving the quality of personnel in joint posi-

tions. Other, less tangible, recalibration mechanisms include

the priority that leaders, managers, and even peers place on

certain behaviors. For example, if each of these groups consis-

tently stresses the importance of challenging analytic assump-

tions, a more competitive analytic environment may be more

likely to flourish.

Finally, technical changes to the underlying support infra-

structure can aid recalibration. Improvements in communica-

tion architectur

es can foster greater analytic collaboration and

also enable gr

eater information sharing. For example, im

proved interoperability, if implemented correctly, may also im-

prove information sharing without a concomitant loss in the

ability to protect infor

mation.

Summary

Reforming the IC, given the breadth of its operations and its

multidepartment span, may actually prove to be one of the

mor

e dif

ficult r

efor

m efforts ever tackled within the US govern-

ment. Unlike the National Security Act of 1947, which unified

separate cabinet-level departments under a single depart-

ment, the Intelligence Reform Act of 2004 lets the IC retain its

interagency nature. The Goldwater-Nichols Act worked to more

strongly unify the efforts of four separate military services, but

again these services functioned under the auspices of a single

department. Still, both defense and intelligence r

eform had

THE QUESTION OF BALANCE

153

common objectives: to improve joint action, establish clearer

lines of authority, clarify management structures, and enhance

the quality of joint personnel. However, the different contexts

meant that intelligence reform advocates had different options

available to them, making direct application of Goldwater-

Nichols measures impractical in certain areas.

As an example of successful reform of a large government

bureaucracy, Goldwater-Nichols remains a noteworthy model

with important implications for the IC. The lessons of the

Goldwater

-Nichols Act—clear lines of authority, more explicit

command relationships, and the importance of career incen-

tives to promote desired behavior—can help any organization

attempting to improve overall effectiveness.

The effect of Goldwater-Nichols teaches other lessons as

well. First, legislation is often only a beginning. How the orga-

nization accepts and implements reforms may have an even

larger impact on overall success. In the case of Goldwater-

Nichols, the DOD generally accepted the reforms and internal

efforts, particularly within the unified commands, and continued

to solidify the measures. Similarly, how well the IC and the

heads of its component departments cooperate and actively in-

stitutionalize reform may significantly affect the success of re-

for

m measures.

Second, deter

mining causality—that is, the degree to which

specific legislative measures had their intended effect—is dif-

ficult at best, as changes do not occur in a vacuum. Goldwa

ter-

Nichols undoubtedly assisted in transforming the DOD into a

more effective joint organization; however, concurrent develop-

ments such as on-going technological advances (e.g., precision

weaponry, stealth, and vastly improved surveillance and tar-

geting systems), impr

oved inter

operability (particularly in C

chitectur

es), and the intangibles of leadership and quality of

personnel also contributed str

ongly to the DOD’s impr

oved

per

for

mance. In fact, by 2004 the IC was not the same or

gani-

zation it had been on 11 September 2001. The or

ganization

had already reacted by making internal changes to address

many aspects of the 9/11 Report. This does not minimize the

importance of legislation in certain instances; however

, pru

dence

THE QUESTION OF BALANCE

154

suggests not giving too much or too little credit to legislative

measures alone.

Third, the fundamental tension between centralization and

decentralization has slowed the pace of intelligence reform.

While reform advocates often pointed to the advantages of in-

creased centralization, such as improved efficiency, less dupli-

cation, and greater standardization, advocates of decentraliza-

tion could justifiably argue the need for tailored intelligence

support in meeting departmental requirements. The executive

branch departments, particularly the DOD, can never conduct

operations without intelligence support and thus almost al-

ways pr

efer direct control over intelligence resources to being

placed in the role of customer. Furthermore, the interagency

nature of the IC means that only the president has authority

over both the executive branch departments and the head of

the IC. Thus, it is little wonder that intelligence reform has

proven so difficult and ponderous.

Finally, reform must do more than change organizational

structur

e. Goldwater-Nichols certainly modified the structure

of the DOD; however, changes to incentive systems and new

doctrine codifying command r

elationships and the structure of

various joint organizations, coupled with technical changes to

improve inter

operability, were also important factors in suc

cess. Again, legislation represents only part of a successful re-

form effort. In this case, adjusting incentive structures to re-

war

d desired behaviors such as cross-agency duty, developing

new doctrine and procedures to govern interagency structures

and operations, and improving cross-agency communication

infrastructures are just some of the steps likely to be required

in furthering reform efforts.

In sum, the unique organizational context of the IC in itself

suggests one reason why changes until 2004 were modest.

This context is also interwoven with multiple competing ten-

sions that further complicate reform. Governmental entities at

various levels can adjust these tensions, using several dif

fer

ent r

ecalibration mechanisms that target numerous aspects of

the organization. The recently passed intelligence reform leg-

islation lays the gr

oundwork for reform to begin in earnest;

however

, true r

efor

m may depend primarily upon the inter

nal

THE QUESTION OF BALANCE

155

efforts of the IC itself. These efforts have the advantage of

greater precision and the ability to recalibrate tensions dy-

namically as required. The degree of cooperation between

agencies and departments; changes in IC policies, practices,

and culture; alterations in the underlying supporting infra-

structure; and—perhaps most important—the effectiveness of

the community’s leaders and personnel in instituting reform

also play key roles in the reform effort.

The sheer complexity of the IC and of the balances involved

suggests that reform will not come easily and will not satisfy

the staunchest advocates on either side of the competing bal-

ances. In the end, adjusting the fulcrum to achieve the desired

balances may require more art than science; however, such an

adjustment can be informed by a broader understanding of

the competing tensions and the potential impact of recalibra-

tion—the ultimate objective of this research.

Notes

1. Walter Pincus, “National Intelligence Director Proves to Be Difficult

Post to Fill: Uncertainties over Role, Authority Are Blamed for Delays,”

Washington Post, 31 January 2005, A04.

THE QUESTION OF BALANCE

156

Appendix A

Interviews

Joan Dempsey, executive director, President’s Foreign Intelli-

gence Advisory Board

John M. Deutch, institute professor at the Massachusetts In-

stitute of Technology and former director, Central Intelli-

gence Agency

John Gannon, majority staff director, House Select Committee

on Homeland Security and former chairman, National In-

telligence Council

Lt Gen Michael V. Hayden, USAF

, for

mer director, National Se

curity Agency

Maj Gen Paul J. Lebras, USAF

, commander, Air Intelligence

Agency

Maj Gen Glen Shafer, USAF, retired, former dir

ector of Intelli-

gence, Joint Staff

R. James W

oolsey, former dir

ector

, Central Intelligence Agency

Multiple congr

essional staffers

Members of the 9/11 Commission staff

157

159

Appendix B

Chronology of Intelligence Reform Proposals Related to IC Reform

Reform Recommendations

Proposal Dates Related to Community Rationale Resulting Actions

Management

Dulles- 1949 – Improve national – Other intelligence – In 1950, DCI Bedell

Jackson- estimate process agencies not part of Smith created a Board

Correa Report – Better coordinate estimative process of National Estimates

activities of IC – Avoid lack of

coordination considered

significant cause of

Pearl Harbor

First Hoover 1949 – CIA establishes better – Improve intelligence – In 1950, DCI Bedell

Commission relations with other estimates Smith created a Board

elements of IC – Avoid duplication and of National Estimates

– Improve reduce number of

centralization of competing intelligence

intelligence analysis

estimates

Second Hoover 1955 – Create Ex Dir of CIA – Free DCI to focus more – Opposite occurred,

Commission – Establish on community affairs DCI created deputy for

commission of – Prevent abuses, establish Community Affairs

members of reasonable checks & – EO 10656 created

Congress and balances

of PBCFIA

private citizens to precursor to

oversee IC the PFIAB

PBCFIA, 1960–1961 – Separate DCI from – Unclear – None

PFIAB CIA

160

Reform Recommendations

Proposal Dates Related to Community Rationale Resulting Actions

Management

Schlesinger Report 1971 – Create a director of – Pinpoint responsibility – Intelligence Resources

National Intelligence or for management of IC, Advisory Committee to

separate roles remove duplication, advise DCI on

of DCI centralize management preparation of

– Create White House- of budget and planning consolidated budget

level coordinator of processes

– IC Staff in 1972

National Intelligence

Murphy 1975 – DCI delegates – Strengthen community – President Ford orders

Commission management of CIA management implementation of 20

to deputy and – Increase prestige of DCI of 30 recommendations,

devotes more vis-à-vis other primarily those dealing

time to intelligence w/internal CIA practices,

community organization did not agree w/public

responsibilities heads disclosure of budget

– Move DCI office – Provide independent

closer to advice to

president the president

– Strengthen PFIAB – More closely tie

– NSC-level consumer requirements

organization with producers

establishes intelligence – Improve cost-

requirements and effectiveness of

capabilities intelligence

– Develop consolidated

intelligence budget

Rockefeller 1975 – Form joint oversight – Strengthen legislative – Separate intelligence

Commission committee within oversight of the IC oversight committees

Congress – Improve transparency of CIA established (Senate

– Publicly release the – Provide better oversight and 1976, House 1977)

CIA budget assess quality of intelligence – Not implemented

– Strengthen PFIAB estimates – Unclear

161

Reform Recommendations

Proposal Dates Related to Community Rationale Resulting Actions

Management

Church 1975 – Draft omnibus legislation – Adequately define roles – Legislation proposed

Committee detailing charters of and responsibilities of to create a director of

major organizations individual agencies National Intelligence

within the IC – Improve coordination who retained control of

– Reaffirm the DCI’s and management the CIA; however,

power to coordinate effectiveness of the IC legislation failed to be

the activities of the – Improve management passed

IC by statute

efficiency and coordinate – Formulation of

– Grant statutory IC activities more national intelligence

authority to effectively budget and creation of

establish national – Too large a span of IC staff done through

intelligence control executive branch

requirements and actions—reprogramming

formulate budget to authorities provided

include reprogramming in 2004

funds and provide – Passed in 2004

DCI with an IC staff

– Consider separation of

DCI and director CIA roles

Congress 1987–1989 – Establish a separate – Better coordinate overall – Not passed

(100–101st) DNI efforts of IC

– Increasing expansion of IC

– Perceived intelligence failures

– Allegations of illegality

Congress 1991–1992 – Establish separate – Reap benefits of end of Cold War – Some provisions (charters for

(Boren-McCurdy) DNI and significantly – Enhance community management intelligence agencies

increase authority and increase efficiency and and greater

– Release top-line accountability articulation of DCI’s

budget amount – Improve quality of coordinating

– Augment analysts analysis that is “overly responsibilities) added

with outsiders fr

cautious, caveated, and to FY 1994 Intelligence

academe and private sector consensus driven”

Authorization Act

162

Reform Recommendations

Proposal Dates Related to Community Rationale Resulting Actions

Management

Aspin-Brown Sep 1994– – Strengthen authority – Enhance community – Budget figures released for

Mar 1996 of DCI via non-statutory management fiscal years 1997 and 1998

means to include – Improve quality of analysis but not thereafter

greater say in appointment – Provide greater

of agency heads, transparency of

dual-hatting some intelligence activities

agency heads under DCI,

establishing two deputy DCIs

– Improve relationships

between analysts

and policy makers,

increase use of open

source information, develop

stronger ties with

experts outside IC

– Release top-line budget figures

IC21 1995–1996 – Increase authority of – Address gap between – Congress passed P.L. 104-293

DCI via increased control DCI’s responsibilities – Established Committee

over budget, stronger and authorities on Foreign Intelligence on NSC

community management – Improve NSC supervision – Established deputy DCI

staff, greater say in of IC and deputy DCI for

appointment of agency heads – Aid DCI in management Community Management

– Establish of community affairs – Provided three

Committee on – Increase and improve assistant DCIs to aid in

Foreign Intelligence congressional oversight community management:

within the NSC Collection, Analysis &

– Establish two Production, Administration

deputy DCIs – Appointment of directors

– Establish joint intelligence of NSA, NRO, NGA

oversight committee in required DCI concurrence

Congress – Strengthened DCI

authority over the budget

William M. Leary, ed., The Central Intelligence Agency: History and Documents (Tuscaloosa, Ala.: University of Alabama Press, 1984), 137.

Richard A. Best, Proposals for Intelligence Reorganization, 1949–2004 (Washington, D.C.: CRS, 29 July 2004), 8.

Commission on Organization of the Executive Branch of the Government (also known as the Second Hoover Commission), “Intelligence Activities,” May 1955, 59. “The CIA has

been freed by the Congress from outside surveillance of its operations and its fiscal accounts. There is always a danger that such freedom from restraints should inspire lax-

ity and abuses which might prove costly to the American people. . . . Some reliable, systematic review of all the agencies and their operations should be provided by congres-

sional action as checkrein to assure both the Congress and the people that this hub of the Intelligence effort is functioning in an efficient, effective, and reasonably economi-

cal manner.”

Schlesinger Report, 26 (see chap. 3, n. 29).

Best, Proposals for Intelligence Reorganization, 17.

Church Committee report, book 1, 429 (see chap. 3, n. 40).

Alfred B. Prados, IC Leadership: Development and Debate Since 1947 (Washington, D.C.: CRS, 27 June 1989), 7.

Boren, “The IC: How Crucial?” 56 (see chap. 3, n. 78).

Best, Proposals for Intelligence Reorganization, 31.

Ibid., 32.

Ibid., 34–35.

Ibid., 35–37. The subject of ADCIs was contentious at best, with the executive branch reluctant to accept them. In fact, after the initial incumbents in both the DDCI for Commu-

nity Management and ADCI for Administration left their positions, the executive branch did not move to fill the offices.

163

165

Appendix C

Milestones in the Intelligence Community

Year Administration Studies/Commissions/ Actions

Proposals

1946 Truman - Central Intelligence Group formed by presidential directive

1947 - National Security Act (P.L. 80-253) establishes CIA and DOD

- Intelligence Advisory Committee established by NSCID 1, 12 December

1949 Jackson-Dulles-Correa Report

- Improve national estimate process

- Improve coordination of IC

First Hoover Commission

- Improve national estimates - CIA Act (P.L. 81-110)

1950 DCI Bedell Smith:

- Establishes Board of National Estimates

- Appoints deputy director of CIA

1952 National Security Agency established by presidential memo

1953 Eisenhower Deputy director of CIA authorized by P.L. 83-15

1955 Second Hoover Commission

- DCI focuses more on community affairs

Senator Mansfield submits proposal

for joint intelligence oversight committee

1956 President’s Board of Consultants on Foreign Intelligence

Activities (PBCFIA), precursor to the President’s Foreign Intelligence

Advisory Board (PFIAB), established by EO 10656, 6 February

1958 U.S. Intelligence Board formed by NSCID 1, 15 September merging

the Intelligence Advisory Committee and the United States

Communications Intelligence Board

166

Year Administration Studies/Commissions/ Actions

Proposals

1960 PBCFIA suggests the possibility of separating the DCI from the CIA

1961 National Photographic Intelligence Center established by NSCID 8,

18 January

1961 Kennedy PFIAB recommends - PFIAB replaces PBCFIA by EO 10938

separation of two roles of DCI - Defense Intelligence Agency established on

1 August by DOD Directive 5105.2

1968 Nixon - DCI Helms establishes National Intelligence Resources Board

to review all IC programs and budgets

1971 Schlesinger Report Presidential memo issued, directing:

- Creates a director of national - Enhanced leadership role in managing IC

intelligence or strengthen DCI - Establishment of National Security Council Intelligence Committee

to remove duplication and - Establishment of Intelligence Resources Advisory Committee

centralize planning and budgeting to advise DCI on preparation of consolidated budget

- Creates White House-level - Creation of Defense Mapping Agency

coordinator of national intelligence - Establishment of IC Staff

1972 - Defense Mapping Agency, later consolidated under NGA

- IC Staff established, precursor to Community Management Staff

1973 DCI Schlesinger broadens composition of IC staff to make it more

representative of the IC

1975 Ford Murphy Commission

- DCI delegates management of CIA

to deputy/devotes more time to

community responsibilities

- Moves DCI office closer to president

- Strengthens PFIAB

- NSC-level organization establishes

intelligence r

equir

ements and capabilities

- Develops consolidated intelligence budget

167

Year Administration Studies/Commissions/ Actions

Proposals

1975 Ford Rockefeller Commission

- Forms joint oversight committee

within Congress

- Publicly releases the CIA budget

- Strengthens PFIAB

1976 Church Committee

- Drafts omnibus legislation detailing

charters of major organizations

within the IC

- Reaffirms the DCI’s power to

coordinate the activities of the

IC by statute

- Grants DCI statutory authority to EO 11905 issued

establish national intelligence - Supersedes Nixon’s 1971 memo

requirements and formulate budget - Creates Committee on Foreign Intelligence (CFI) and Operations

to include reprogramming funds and Advisory Group

provide DCI with an IC staff Establishes Intelligence Oversight Board

- Considers separation of DCI and Creates Deputy DCI for the IC

Director CIA roles

Pike Committee

- Separates DCI roles Senate establishes Select Committee on Intelligence

1977 Carter House establishes Permanent Select Committee on Intelligence

1978 Legislation (S. 2525) EO 12036

proposed to: - Abolishes Ford’s CFI

- Create charters governing - Establishes Policy Review Committee and a Special Coordinating

intelligence agencies Committee on NSC

- Create DNI to head IC

1981 Reagan - Grants DCI cabinet rank

- Issues EO 12333 which

-- Abolishes Carter’s NSC structures

-- Clarifies roles and responsibilities of

organizations in IC

168

Year Administration Studies/Commissions/ Actions

Proposals

1987 National Intelligence Reorganization

Act (S. 1820) introduced to

- Establish separate DNI

1989 Bush National Intelligence Reorganization

Act (S. 175) introduced to

- Establish separate DNI Abolishes Reagan’s NSC structures

1992 Intelligence Reorganization Act of Congress passes Intelligence Authorization Act of FY 1993

1992 (S. 2198) introduced to (P.L. 102-496) which includes the Intelligence Reorganization

- Establish separate DNI and Act of 1992:

significantly increase authority - Establishes National Intelligence Council to produce national estimates

- Release top-line budget amount - Specifies duties of the DCI as head of the IC to include developing

- Consolidate analytic elements of and approving an annual national intelligence budget and

IC into single organization eliminating waste and duplication

- Augment analysts with outsiders - IC Staff renamed Community Management Staff

from academe and private sector

1993 Clinton EO 12863

- Makes Intelligence Oversight Board a standing committee

of PFIAB

1996 Aspin-Brown Congress passes PL 104-296

- Strengthens authority of DCI via - Establishes Committee on Foreign Intelligence on NSC

non-statutory means to include greater - Establishes Deputy DCI and Deputy DCI for Community Management

say in appointment of agency heads, - Establishes three Assistant DCIs

dual-hatting some agency heads - Requires DCI concurrence in appointment of directors of

under DCI, establish two NSA, NRO, NGA

deputy DCIs - Strengthens DCI authority over the budget

- Improves relationships between analysts - Congress passes PL 104-201

and policy makers, increases use of - Establishes the National Imagery and Mapping Agency

open source information, develops

stronger ties with experts outside IC

- Releases top-line budget figures

169

Year Administration Studies/Commissions/ Actions

Proposals

1996 Clinton IC 21 Study

- Increases authority of DCI via increased

control over budget, stronger community

management staff, greater say in

appointment of agency heads

- Establishes Committee on Foreign

Intelligence within the NSC

- Establishes two Deputy DCIs

- Establishes joint intelligence oversight

committee in Congress

Note: In addition to detailing information found in commission reports, studies, executive orders, and public law, this appendix consolidates information from Best’s review

of the IC (see Richard A. Best,

Proposals for Intelligence Reorganization, 1949–2004 (Washington, D.C.: CRS, 29 July 2004) and tables found in Alfred B. Prados, Intelligence

Community Leadership: Development and Debate Since 1947

(Washington, D.C.: CRS, 27 June 1989), Appendix B-2 and Appendix C, 75–79.

Senate, Select Committee to Study Governmental Operations with Respect to Intelligence Activities, Book IV: Supplementary Detailed Staff Reports on Foreign and Military In-

telligence

, 94th Cong., 2d sess., 23 April 1976, report 94-755, 8–9 (hereafter cited as Karalekas). The first 107 pages of this report were prepared by a staff member, Anne Kar-

alekas. The CIG was responsible for coordination, planning, evaluation, and dissemination of intelligence. It also was granted overt collection responsibility. The National In-

telligence Authority (NIA), a group comprised of the Secretary of State, the Secretary of War, the Secretary of the Navy, and the personal representative of the President, served

as the Director’s supervisory body. The Intelligence Advisory Board (IAB), which included the heads of the military and civilian intelligence agencies, was an advisory group for

the Director. Through budget, personnel, and oversight, the departments had assured themselves control over the Central Intelligence Group.

National Security Council Intelligence Directive No. 1, Duties and Responsibilities, 12 December 1947. See also, Karalekas, 25. The Intelligence Advisory Committee (IAC) was

created to serve as a coordinating body in establishing intelligence requirements. The IAC was chaired by the DCI and included representatives from the Department of State,

Army, Air Force, the Joint Chiefs of Staff, and the Atomic Energy Commission.

Prados, Intelligence Community Leadership, 17 (see chap. 3, n. 2). The CIA Act primarily dealt with administrative issues associated with the CIA but not the overall structure

of the IC.

Joseph S. Nye, “Estimating the Future,” in Godson, U.S. Intelligence at the Crossroads, 87 (see chap. 5, n. 22).

Prados, Intelligence Community Leadership, 76 (see chap. 3, n. 2).

Karalekas, 52.

Ibid., 62. The PBCFIA was designed to provide advice to the president on intelligence activities. The board had no authority over any element in the IC.

National Security Council Intelligence Directive No. 1, Basic Duties and Responsibilities, 15 September 1958. See also Karakelas, 62–63. The establishment of the USIB was a

recommendation of the PBCFIA to merge the IAC and the USCIB and was designed to allow the DCI to better coordinate community activities. With no significant bureaucratic

levers, the board had limited authority. See also, Aspin-Brown, A-14 (see chap. 3, n. 7). According to this history, the USIB was used unevenly by the DCIs, with the personal

preferences of the DCI playing a large role in its overall utilization.

Prados, Intelligence Community Leadership, 43 (see chap. 3, n. 2).

National Geospatial-Intelligence Agency, “NGA History,” 1, http://www.nga.mil/StaticFiles/OCR/nga_history.pdf (accessed 15 December 2004).

Aspin-Brown, 34 (see chap. 3, n. 7). Major functions of the PFIAB included advising the president on the overall quality and adequacy of intelligence as well as the manage-

ment, personnel, and organizational arrangements.

170

Prados, Intelligence Community Leadership, 76 (see chap. 3, n. 2).

Best, Proposals for Intelligence Reorganization, 17. See also, Aspin-Brown, A-14 (see chap. 3, n. 7).

“Reorganization of the IC,” Weekly Compilation of Presidential Documents, 4 November 1971, 1482.

Prados, Intelligence Community Leadership, 76 (see chap. 3, n. 2).

Aspin-Brown, A-15 (see chap. 3, n. 7).

Karalekas, 85.

Aspin-Brown, 30 (see chap. 3, n. 7). This committee was composed of the director of central intelligence, the deputy secretary of defense for intelligence, and the deputy as-

sistant to the president for national security affairs. It was charged with controlling budget preparation and resource allocation and establishing priorities for collection and

production.

Aspin-Brown, 30 (see chap. 3, n. 7). The PRC assumed most of the functions of the abolished CFI including establishment of requirements and priorities and budget review.

Prados, Intelligence Community Leadership, 49 (see chap. 3, n. 2).

Aspin-Brown, 30 (see chap. 3, n. 7).

Ibid., 31.

Intelligence Reform

A Question of Balance

Air University Press Team

Chief Editor

Emily J. Adams

Copy Editor

Carol D. Norris

Book Design and Cover Art

Steven C. Garst

Composition and

Prepress Production

Vivian O’Neal

Quality Review

Mary J. Moore

Print Preparation

Joan Hickey

Distribution

Diane Clark