Wald

WALD FINAL 5/19/2006 9:30 AM
ANALYSTS AND POLICYMAKERS: A

CONFUSION OF ROLES?
Patricia M. Wald∗
During the past year, we have witnessed two major cause celebres
involving the relations between national policymakers and intelligence analysts
in the sensitive areas of foreign relations and national security. The first
involved the issue of whether the analysts who worked on the National
Intelligence Estimate and other pre-war assessments were pressured or
intimidated into finding that Iraq had Weapons of Mass Destruction (WMD) by
high-level policymakers who wanted and needed such conclusions to further
their preordained objectives to unseat the Iraqi President, Saddam Hussein, by
military force. The second contratemps arose during the confirmation hearing
of John Bolton as United States Ambassador to the United Nations. Opponents
of the Bolton nomination asserted that on several occasions during his tenure as
Assistant Secretary of State he had tried to transfer or have removed analysts
who would not alter their findings on WMD to fit his own judgments on
Cuba’s or Syria’s activities. Though the accounts of the interface between
Bolton and intelligence analysts (and their superiors) varied considerably, the
accusations of political interference with analysts’ judgments on key foreign
policy matters gave rise to provocative dialogues in the Senate as to the proper
role of both policymakers and analysts when they join to produce critical
judgments—some that could lead to war—on the capabilities and, more
importantly, the intent of hostile governments and terrorist groups to
manufacture and use such weapons.
At first, it seems hard to fathom how our extensive national intelligence
apparatus has been operating for nearly sixty years without those relationships
having been explored publicly before. Now, however, the ever-escalating
stakes involved with the technological development of and access to new
weapons, as well as the emergence of new kinds of enemies and perhaps a
growing recognition of an accumulation of missed opportunities and past
mistakes over the past half-century, have pushed the Intelligence Community
(IC) to front and center. In the case of Iraq, two comprehensive and scathing
∗
Judge, U.S. Court of Appeals, D.C. Circuit, 1979-1999; Chief Judge, 1986-1991;
Judge, International Criminal Tribunal for the former Yugoslavia, 1999-2001; Member,
President’s Commission on Intelligence Capabilities of the United States Regarding
Weapons of Mass Destruction, 2004-2005.
241
WALD FINAL 5/19/2006 9:30 AM
242 STANFORD LAW AND POLICY REVIEW [Vol. 17:241
critiques by the Senate Select Committee on Intelligence (SSCI)1 and the

President’s Commission on the Intelligence Capabilities of the United States
Regarding Weapons of Mass Destruction (WMD Commission),2 on which I
served (2004-2005), of the tradecraft involved in analysts’ pre-Iraq War work
product did not entirely satisfy critics. Some suspected, notably, that the
analysts had succumbed to pressures by policymakers to arrive at their
conclusions, later proven “wholly wrong,” that Saddam Hussein was embarked
on an active effort to make and use WMD in 2002.3 This skepticism was
refueled by the release of a contemporaneous memorandum written in July
2002 by “C”—head of the British MI6 —reporting on “recent talks in
Washington” to the effect that the Bush Administration was actively planning
to go to war to achieve regime change in Iraq and that “the intelligence and
facts were being fixed around that policy.”4 No further specifics were offered
on who said what to whom, but critics of the war and the Administration seized
upon the memo as the “smoking gun” that was “proof positive” the intelligence
community had been unequivocally politicized and bent to the will of the
policymakers.5
In this short essay, I will briefly summarize the WMD Commission’s
findings both about alleged “politicization” of the Iraqi analysts and the causes
of their lamentable shortfall on the pre-war estimates of Iraqi WMD, findings
which closely parallel those of the Senate Intelligence Committee’s report
issued within the same year. After that I will probe some of my own views on
possible conceptual confusion about just what the role of the intelligence
analyst is, and the inherent difficulties in drawing lines between constructive
and destructive interchanges between analysts and the policymakers who
inevitably must provide the marching orders as to what information they want
or need to conduct military or diplomatic policy. Finally, despite profound
differences between the IC and other government bureaucracies, I offer some
modest suggestions largely gleaned from administrative law for handling the
delicate balance between professional integrity in fact-finding and usefulness to
elected policymakers. The most controversial of these is that the IC take
account of the likely consequences of its findings and judgments by setting
1. STAFF OF S. SELECT COMM. ON INTELLIGENCE, 108TH CONG., REPORT ON THE U.S.

INTELLIGENCE COMMUNITY’S PREWAR INTELLIGENCE ASSESSMENTS ON IRAQ (Comm. Print
2004), available at http://intelligence.senate.gov/iraqreport2.pdf [hereinafter SSCI REPORT].
2. COMM’N ON THE INTELLIGENCE CAPABILITIES OF THE UNITED STATES REGARDING
WEAPONS OF MASS DESTRUCTION, REPORT TO THE PRESIDENT (2005), available at
http://www.wmd.gov/report/wmd_report.pdf [hereinafter WMD REPORT].
3. See, e.g., Seymour M. Hersh, The Stovepipe, NEW YORKER, Oct. 27, 2003, at 77; see
also other sources cited in WMD REPORT, supra note 2, at 246 n.829.
4. Mark Danner, The Secret Way to War, N.Y. REVIEW OF BOOKS, June, 9, 2005, at 70,
70-71.
5. See, e.g., Danner, id., at 70; Frank Rich, It’s Bush-Cheney, not Rove-Libby, N.Y.
TIMES, Oct. 16, 2005, § 4, at 12; cf. Michael Kinsley, No Smoking Gun, WASH. POST, June
12, 2005, at B9.
WALD FINAL 5/19/2006 9:30 AM
2006] ANALYSTS & POLICYMAKERS 243
different thresholds of certainty and reliability for these findings dependent on

the purpose of or context in which they are raised.
I. WHAT THE WMD COMMISSION HAD TO SAY ABOUT ANALYSTS
The WMD Commission’s mandate was to “assess whether the intelligence

community is sufficiently authorized, organized, equipped, trained and
resourced to identify and warn in a timely manner of, and to support United
States Government efforts to respond to the development and transfer of
knowledge, expertise, technologies, materials and sources associated with the
proliferation of Weapons of Mass Destruction . . . and other related threats of
the 21st century . . . .”6 With regard to Iraq, we were directed to “specifically
examine the intelligence community’s intelligence prior to the initiation of
Operation Iraqi Freedom and compare it with the findings of the Iraq Survey
Group [the CIA-led post-invasion expert group that went into Iraq to look for
WMD immediately after official hostilities ended] . . . relating to the
development, . . . possession, . . . [or] potential or threatened use . . . of
Weapons of Mass Destruction . . . .”7 In addition, the WMD Commission was
specifically tasked to evaluate the challenge of obtaining information on WMD
in “closed societies,” and to compare the WMD intelligence we had on Libya
prior to its 2003 decision to open its weapons programs to inspection and on
Afghanistan prior to the removal of the Taliban with current assessments of
how far advanced those programs actually were.8
Appointed by President Bush in March 2004, the ten-person bipartisan
WMD Commission produced a 600-page report a year later, rich in analysis of
what went wrong in pre-war intelligence assessments and equally prolific in
recommendations as to how they might be improved. During the same year, the
independent 9/11 Commission scorned the intelligence community for not
“connecting the dots” on pre-9/11 information concerning terrorist plans
against the World Trade Center.9 Largely due to their public hearings and in
light of public opinion, Congress passed the first major overhaul of the
intelligence structure since 1947, creating an “intelligence czar,” the Director
of National Intelligence, to head the fifteen intelligence units scattered among a
half dozen executive agencies, including the Department of Defense,
Department of State, Department of Justice, Department of Energy, and the
CIA.10 The new “Czar’s” unenviable job is to insure integration of all fifteen
6. WMD REPORT, supra note 2, app. at 551 (Exec. Order No. 13,328, 69 Fed. Reg.
6901 (Feb. 6, 2004)).
7. Id. at 551-52.
8. Id.
9. NAT’L COMM’N ON TERRORIST ATTACKS UPON THE UNITED STATES, THE 9/11
COMMISSION REPORT (2004), available at http://www.9-11commission.gov/report/
911Report.pdf.
10. Intelligence Reform and Terrorist Prevention Act of 2004, Pub. L. No. 108-458,
118 Stat. 3638 (2004).
WALD FINAL 5/19/2006 9:30 AM
intelligence units’ functions, and to spur the much-needed innovations

necessary to keep up with our enemies’ increasingly sophisticated capabilities,
not only in acquiring the knowledge and expertise for development of their
own WMD and keeping that expertise and intentions secret from us, but in
discerning our WMD state of the art as well.
Retrospectively, the WMD Commission found that the IC’s assessments of
Saddam Hussein’s WMD capabilities were “all wrong.”11 The National
Intelligence Estimate (NIE) of October 2002 (the NIE is the most authoritative
product of the IC, representing the combined efforts of all relevant intelligence
agencies on a particular issue) said that Iraq was reconstituting its nuclear
program when in fact no such efforts were underway.12 Indeed, the failure of
the United Nations inspectors’ reports over an earlier four-year period to locate
any such activity had strongly suggested this outcome already. The NIE’s
mistaken conclusion was based on what the NIE called the “compelling
evidence” of Iraq’s attempted procurement of high-strength aluminum tubes
which a majority of intelligence agencies (spearheaded by the CIA) assessed
were intended for use in centrifuges in a uranium enrichment process leading
toward a nuclear weapon program rather than for the conventional rocket use
Iraq proclaimed and which two agencies—the State Department and the
Department of Energy—thought more likely.13 In the view of these agencies,
the particular centrifuges were not well suited for nuclear use. Ironically, the
Defense Department took what turned out to be the wrong side of this issue;
even more surprisingly, its error was based on poor technical analysis of the
tubes and a lack of up-to-date knowledge about missile weapons technology,
the alternative use for the tubes.14 Rumors and forged documents involving a
purported attempt by Iraq to purchase yellow cake (enriched uranium) from
Africa for nuclear weapons use, never proven and ultimately not credited by
the CIA or relied upon in the NIE, contributed to the debacle. There were
skeptical participants from the Department of Energy in the NIE process who
told the WMD Commission that, even though they discounted the centrifuge
theory and said so in the NIE, they still “went along” with the majority
judgment that Iraq was reconstituting its nuclear program because “DOE didn’t
want to come out before the war and say [Iraq] wasn’t reconstituting.”15 Only
the State Department filed a dissent on the basic issue of whether Iraq had
reconstituted its nuclear weapons program.
11. WMD REPORT, supra note 2, at 45.

12. Nat’l Intelligence Council, National Intelligence Estimate on Iraq’s Continuing
Program for Weapons of Mass Destruction 16 (2002), available at
http://www.fas.org/irp/cia/product/iraq-wmd.html (declassified excerpts) (last visited Nov.
11, 2005) [hereinafter “NIE”].
13. Id.
14. WMD REPORT, supra note 2, at 67-71.
15. Id. at 75.
WALD FINAL 5/19/2006 9:30 AM
In the biological warfare area, the WMD Commission’s assessment of the

IC was, if anything, more scathing. Previous IC assessments of Iraq’s
development of biological weapons in the late 1990s had credited Iraq’s
asserted destruction of its prior Gulf war supplies of biological weapons.
Despite recognizing that dual use equipment and components can serve both
legitimate pharmaceutical purposes and WMD development, these earlier
assessments found no evidence Iraq had embarked on a new program of
biological warfare. But by 2000, the notorious “Curveball” came onto the
scene—a former Iraqi engineer, now in the custody of a foreign intelligence
service which would not allow him to be interviewed by U.S. intelligence.
Curveball made over 100 reports on Iraq’s alleged development of mobile
production facilities for biological weapons which were relayed by the foreign
intelligence service to our own IC. From the beginning there was discord
within the CIA as to how credible Curveball was, with, frustratingly, no ability
to vet him for ourselves. His reports, however, had enough internal
inconsistencies and external contradictions to make some CIA officials
distinctly uneasy.16 A combination, nevertheless, of poor internal
communications within the CIA, and between the CIA and the Defense
Intelligence unit about the intensity of these doubts, allowed Curveball’s
reports of mobile biological warfare units to carry the day in the end and to
constitute the principal basis for the NIE and for Colin Powell’s eve-of-war
address to the United Nations on the existence of biological WMD production
facilities. After the war, and after U.S. intelligence had the opportunity to
interview Curveball personally, it was acknowledged that he was indeed a
fabricator. Additionally, another source used for confirmation of Curveball’s
revelations was discovered to be a fabricator as well,17 yet this second source’s
tainted reports were inadvertently kept in the circulated reports about WMD for
use by policymakers. With a dearth of human sources in Iraq to counter the
fabricators, their prevarications won the day.
The chemical warfare intelligence was equally faulty. The NIE estimated
stockpiles of up to 500 metric tons of chemical weapons, projections largely
based on accompanying deficiencies in Saddam’s reports to the U.N. inspectors
and on aerial imagery at suspected sites of a certain kind of truck traditionally
used to carry chemical weapons. Yet, what the analysts interpreted as increased
activity at these facilities later turned out to be likely, at least in part, the result
of taking the images at more frequent intervals.18
And, finally, the NIE found Iraq was developing unmanned aerial vehicles
(UAVs) basically to deliver biological strikes—possibly against the U.S. The
UAVs, in fact, were found by the postwar Iraq Survey Group (ISG) not to be
intended for biological weapons, and corroborating information that an Iraqi
agent had ordered software plans of U.S. sites for possible attacks was
16. Id. at 90-93, 94-105.

17. Id. at 84-85.
18. Id. at 117, 122-26.
WALD FINAL 5/19/2006 9:30 AM
mistaken.19 The inclusion of the software in orders of other neutral equipment

was belatedly conceded to be more probably inadvertent.20
This, then, was the intelligence analysis—prepared for the critical debate in
Congress and at the U.N. on whether to go to war with Iraq on the basis of
Saddam Hussein’s possession of WMD.
How could it have happened? One senior intelligence official called it a
“perfect storm”—the convergence of a series of idiosyncratic faults in the way
the intelligence agencies worked in this case and a healthy dose of just plain
bad luck.21 But the WMD Commission was not so sanguine; it labeled “the 15
intelligence organizations . . . a ‘[c]ommunity’ in name only’” that “rarely acts
with a unity of purpose”—“fragmented, loosely managed, and poorly
coordinated.”22 Many of the same faults that plagued the Iraq exercise could be
seen—though not in such magnitude—in examinations of intelligence
operations in Libya and Afghanistan. The WMD Commission also studied
intelligence operations in North Korea and Iran—the quintessential “hard
targets.” Though our results in these two countries were classified in their
entirety, we did pronounce: “Across the board, the Intelligence Community
knows disturbingly little about the nuclear programs of many of the world’s
most dangerous actors.”23 The truth was that we had an embarrassing shortage
of verifiable information on Iraq; we had pitifully few human spies in Iraq;
aerial intelligence can be useful in spotting changes on suspected sites but it
cannot always tell you what is going on inside buildings or what mobile trucks
are being used for; and signals intelligence can often be thwarted by the target
country’s change of communications systems, which erects technical barriers to
our surveillance technologies. Poor tradecraft failed to execute or properly
interpret the right tests for the most likely potential uses of centrifuges, and we
did not insist on verification of the reliability of Curveball before placing such
heavy reliance on his say-so. In short, our affirmative evidence of the existence
of WMD was very thin indeed. Had the NIE said just that, we might be in a
different place today. In the end, the Commission found that “one of the most
public—and most damaging—intelligence failures in recent American history
. . . was in large part the result of analytical shortcomings; intelligence analysts
were too wedded to their assumptions about Saddam’s intentions.”24
It was accepted wisdom in the IC that Saddam Hussein definitely intended
to restore his former WMD capabilities as soon as he could: he had those
capabilities in the Gulf War and for years thereafter, he used chemical weapons
against the Kurds and Iranians, and he steadfastly refused to come clean with
19. Id. at 132-46.

20. Id.
21. Id. at 3.
22. Id. at 5.
23. Id. at 4.
24. Id. at 3.
WALD FINAL 5/19/2006 9:30 AM
the U.N. inspectors in his declarations of what happened to these stockpiles and
production capacity. Was it not then reasonable to assume he would secretly
begin reconstructing his old WMD programs, as his frustration with the
interminable U.N. sanctions increased? The oil for food program allowed some
respite from utter economic devastation, and after 1998 he was free of the
intrusive U.N. inspectors. No one seriously considered or apparently even
raised to any cognizable level the alternative explanation for Saddam Hussein’s
conduct, a hypothesis that apparently turned out to be the truth—namely that he
was willing to destroy all the WMD he had as a way to get rid of existing
sanctions, but that he either failed through incompetence to document their
voluntary destruction or, more likely, he wanted to keep alive the deterrent
against his neighbors—Iran and Israel—stemming from a belief that he still
had WMD.
This assumption, largely based on history, that Saddam was secretly trying
to recreate his WMD capabilities, was so embedded in the IC that the analysts
too easily rejected any information that pointed the other way—even in the
absence of any solid information supporting their preconception. Call it “group
think,” “tunnel vision,” or whatever, they in effect switched the burden of proof
to require positive proof that Saddam did not have WMD rather than concrete
proof that he did have them.
The Commission also took to task the way in which the IC described their
findings, both in the original classified NIE for policymakers and in subsequent
declassified versions for public and media consumption. The uncertainties—the
gaps in the IC’s real knowledge—were not adequately conveyed; they did not
explicitly reference the historical assumptions that were driving their
analysis—Saddam’s past record of using WMD and his dissembling to the
inspectors. Of the forty caveats in the original NIE, fifteen were dropped
altogether in the public version, creating a still stronger air of certainty. (The
Commission similarly found the pieces on WMD in the President’s Daily
Briefing around this time “disturbingly one-sided.”) The language of consensus
in the NIE—“most agencies believe”—obscured the fact that the agencies that
dissented on particular issues (like the Department of Energy with the
centrifuges and the Air Force with the intended use of the UAVs) were the
most expert ones in their field, and a policy of “one agency, one vote” may not
have been the right paradigm for evaluating their assessments.25 In sum, the IC
had distressing failures before—including failures to warn of Pearl Harbor and
to predict the demise of the Soviet Union, the Indian nuclear tests, and the Bay
of Pigs—but this one was a whopper.
The Commission’s scope of recommendations involving the IC’s analysis
capabilities ranged far and wide. With a brand-new Director of National
Intelligence (DNI) coming in to administer a brand-new statute, this was not
the time for tinkering, but for big changes. The WMD Commission, like the
25. Id. at 181-86.

WALD FINAL 5/19/2006 9:30 AM
9/11 Commission before it, aimed for an integrated IC rather than the stovepipe
structure of the past; “Mission Managers” on the new DNI’s staff would
coordinate the work of all relevant agencies around particular priority issues.26
These managers would direct and oversee both collection and analytic efforts
throughout the whole IC on issues like China, North Korea, and WMD. (The
new Intelligence Reform Act created a National Counterterrorism Center to do
this for terrorist threats). A prime means to this end was the introduction of
genuine information sharing, a recommendation everyone in the community
agreed to in principle but seemed unable to effect in practice. Most experts said
outmoded technology was not the real problem in information sharing; rather,
the problem was a lack of common standards among the fifteen intelligence
agencies for who could access what kinds of information—each agency wanted
to keep its own sources and assets close to the vest (even to the point of
denying sufficient information on sources to the analysts who were supposed to
make credibility determinations on these sources in their reports). Moreover,
the intelligence collectors were and are often understandably afraid of safety
risks to their sources if their identities are widely spread about. A real culture
change would have to happen inside the community to make information
sharing work. And only if this change can be effected can the Mission
Managers be expected to coordinate the sum total of relevant information for
analytic purposes.27 Civil libertarians of course have valid concerns about
privacy as raw data races through cyberspace; the delicate balance between
security of sources, privacy concerns, and acquisition of all the information
necessary for sound analysis will not be an easy achievement, but it is crucial
to intelligence reform.28
The Commission made several key recommendations about the kind of
workforce needed in a reformed IC, including analysts. These included
recruitment of a more diversified core of analysts with training in languages
and technology, greater career mobility and incentives, opportunities to move
around among different parts of the community in the course of their careers,
and pay for performance. The new breed of recruits should be able to move
into the places and among the people where we need contacts and information,
not sit behind an embassy desk or circulate at embassy cocktail parties.
Arbitrary past standards, that our IC analysts must be, like Caesar’s wife,
“above suspicion” and cannot even have relatives here or abroad that might be
used to corrupt or intimidate them, would have to be rethought. Not only do we
26. Id. at 19, 25, 317-19.

27. Id. at 28-29, 320.
28. Since the Commission Report was issued, the U.S. Government has
acknowledged, after prior leaks, a four-year-long secret NSA surveillance of
communications between U.S. persons and alleged al Queda members abroad. The members
of the WMD Commission had no knowledge of the secret program. Many members of
Congress have said the program violated U.S. law and Congressional bills are pending to
address the issue.
WALD FINAL 5/19/2006 9:30 AM
need more agents of black, brown, yellow, and every other hue, but they need
to speak the languages of the people from whom they seek information; they
need new platforms from which to commingle and obtain information in hard
target countries. The new recruits will also need to understand and use
technical equipment to aid them in their job. Analysts especially need the
greater familiarity that comes from field exposure with the background and
context of the societies whose information they are evaluating.29
The Commission also recommended that the IC, including the analysts,
should reach out into the open society more for its raw material. There is a
well-documented tendency among IC insiders to overvalue secret or classified
information and to undervalue the open sources of information which often are
as or more reliable and comprehensive. In the age of the Internet, the
proliferation of specialized media, and street communications, such
undervaluation is negligent. Up to now, the IC’s use of open sources has been
embryonic and underfunded, and our analysts have not been adequately trained
or equipped to use them. Amazingly, the Commission found that some
intelligence analysts did not even have the capacity to utilize Google. For these
reasons, it recommended a new Open Sources Directorate in the CIA to
enhance the availability of open-source materials to analysts, collectors, and
users of intelligence.30 There were other ways as well in which the IC should
go public: using independent outside experts more frequently, especially as
they provide the kind of historical, cultural, political, and economic
background that often gives context to individual incidents or pieces of
information, and which was sadly lacking in the case of the Iraq WMD
analyses. Many outside experts can be tapped as well for specific observations
gleaned from their international visits and contacts. Particularly in the case of
biological warfare, the Commission was told the expertise lies primarily
outside the government and cannot be recruited inside; thus, new efforts had to
be made to secure the benefits of that expertise for IC research within the
confines of scientific ethical tenets.31
In general, as the WMD Commission moved along in its investigations, it
became increasingly clear that the analyst was at the core of both the major
problems and the most likely solutions in the IC. Given that the drafters of the
Iraq NIE had little new or recent information about WMD, it was the untoward
inferences the analysts drew from that information, the inadequacy of their
vetting of shaky sources, the failure to speedily withdraw from circulation
information known to be false, a disinclination to specify the assumptions they
were operating under in drawing their conclusions, and the overconfident
language in which they wrapped the conclusions that lay at the heart of the
debacle.
29. WMD REPORT, supra note 2, at 20, 321-26.

30. Id. at 22-23, 377-80, 395-400, 510-12.
31. Id.
WALD FINAL 5/19/2006 9:30 AM
In the Commission’s view, the world of the analyst had to change. First,
she must have more direct communication back and forth with the collector to
insure the collector obtains what the analyst needs and the analyst understands
the provenance of the raw data. In some cases, collectors and analysts might
profitably work side by side. In any case, the Mission Manager should assure
that all analysts and collectors working on a specific issue under his direction
are connected and share data. The analyst would have specialist, open-source
analysts to consult. Some analysts would be freed from the daily demands of
current requests to work on long-term strategic issues. Other analysts would be
specially assigned to conduct research and present alternative hypotheses and
arguments to counter the “group think” phenomenon so omnipresent in the Iraq
situation.
Analysts were encouraged to be more candid about what they know, what
they are uncertain about, and what inferences they are drawing from particular
data or assumptions. The “siren call of certainty” must be resisted. All-sourced
copies of earlier analyses or papers upon which a current analytical product
relies would be stored and indexed for diligent or skeptical policymakers to see
for themselves. Analysts would conduct “lessons learned” exercises on
important analyses to see what they did right and wrong. The President
promised, and Congress has already approved, a sizeable increase in both IC
collectors and analysts; however, the conclusion is inevitable from the
Commission’s report that hiring hundreds of new analysts will not accomplish
much unless the system under which they work is dramatically changed.32 For
once, fiscal conservatism and intelligence reform may go hand in hand.
What did we see as the greatest obstacles to reform? Natural ennui,
reluctance to share knowledge with other agencies, and turf protection; the spy
world is Washington bureaucracy in spades. Hopefully, a new DNI and new
leaders of several of the intelligence agencies would have a stake in seeing the
community regain its prestige in the world and would move a reform agenda
along. But the Commission did not assume the old ways would give way easily;
it concluded that only strong leadership from the top down would accomplish
lasting reform.
The FBI, which until recently was predominantly a law enforcement
agency, is a case in point. The FBI now has significant domestic intelligence-
gathering and analysis responsibilities, but the structure—the Intelligence
Directorate—it originally set up to oversee these new duties had virtually no
independent authority from the Agency’s old-line law enforcement structure.
The Commission recommended a new National Security Service within the
FBI which would report to the DNI as well as the Attorney General on
intelligence matters so that the DNI can coordinate domestic intelligence and
32. Id. at 387, 389, 407-09, 411-14, 424-25.

WALD FINAL 5/19/2006 9:30 AM
foreign intelligence for terrorism, WMD, etc.33 There has also been recurrent
conflict over who—the CIA or the FBI—has primary responsibility for
gathering foreign intelligence on U.S. soil—a dispute hopefully settled by a
recent Memorandum of Understanding.34 Still, a newspaper report a few
months after the WMD Report said that the FBI was having difficulty
recruiting and placing over 500 newly-authorized analysts. One-third of the
spots were vacant and many were doing “escort, trash and phone duty”; others
were merely giving low-level research assistance to regular agents; they were
frequently denied access to needed FBI intelligence and many quit in
frustration.35 The nagging worry remained whether a primarily law
enforcement culture used to gathering evidence for prosecution can reorient
itself—at least in part—to the different modes of piecing together bits and
pieces of information and producing intelligence estimates for policymakers
that will never pass through a courthouse door. The next year or so would tell
the tale.
II. WHAT HAS HAPPENED SINCE?
I have now described the Commission’s principal findings and

recommendations involving analysts, to all of which I subscribe. I also
indicated at the start that in the area of analyst-policymaker relationships, I
thought there was more profitable probing yet to be done by others. Before
moving on, however, I take a brief detour to survey what has happened to these
Commission recommendations in the year since the Report was issued.
The Commission noted that predecessor commissions had “identified some
of the same fundamental failings we see in the Intelligence Community, usually
to little effect.”36 The IC “has an almost perfect record of resisting external
recommendations.”37 Yet, a remarkable amount of implementing work on the
WMD Report appears to have been done. Within a month, the White House
announced that seventy of the Commission’s seventy-four recommendations
would be adopted, including the most controversial ones, such as the
reorganization of the FBI’s intelligence functions under a new National
Security Service which would be subject to the coordination and budget
authorities of the DNI as well as the FBI’s regular boss, the Attorney General.38
33. Id. at 30-31.

34 Id. at 31.
35. David Johnston, Antiterror Head Will Help Choose an F.B.I Official, N.Y. TIMES,
June 12, 2005, at A1; Melanie W. Sisson, The F.B.I.’s 2nd-Class Citizens, WASH. POST,
Dec. 31, 2005, at A19.
37. Id.
38. See, e.g., Dan Eggen & Walter Pincus, Spy Chief Gets More Authority Over FBI ,
WASH. POST, July 20, 2005, at A1; Negroponte Will Control Bureau’s Intelligence Side,
WASH. POST, July 30, 2005, at A1; Douglas Jehl, Bush to Create New Unit in F.B.I. for
Intelligence, N.Y. TIMES, June 29, 2005, at A1; Editorial, Intelligence Shuffle, WASH. POST,
WALD FINAL 5/19/2006 9:30 AM
The Washington Post called it a reform “potentially profound and warrant[ing]

public attention.”39 The Associated Press advised: “[E]ven some of the
administration’s critics are not calling this another reordering of the deck
chairs, but rather changes with potential to lead to better intelligence collection
and analysis.”40 The New York Times complimented the President on doing “all
Americans a service . . . by ordering reforms that offer the real possibility of
plugging the national security holes that allowed the 9/11 hijackers to plan,
train for and carry out mass murder without being noticed.”41 Some remained
skeptical that an FBI culture attuned to law enforcement, not intelligence
gathering, may simply be too deeply rooted to be changed, and that radical
reform, like the creation of a domestic MI5—the scourge of civil libertarians—
would prove necessary.42 Nonetheless, despite much backdoor lobbying and
doomsayer predictions, this basic reorganization was in place within months.
That was not all. In July, the new DNI Deputy Michael Hayden, former
head of the National Security Agency (NSA), told Congress that the NIE
process was being overhauled to assure that, unlike the Iraq NIE, future ones
would be based on credible information.43 And here an apparently “profound”
cultural change has taken place in the instructions for these seminal documents
handed from the top down. Hayden demanded “new, critical scrutiny to both
human and technical intelligence, including reports from agents, satellite
photographs and intercepted communications.”44 Their focus would be on
“who said what, why, and why do we think . . . [it] true.”45 Hayden described
the change as a “major breakthrough,” but not without its own internal costs.46
The sea change would inevitably result in estimates that “proved much less
definitive than in the past.”47 And senior policymakers would have to operate
with “a higher tolerance for ambiguity.”48 Analysts would be encouraged to be
July 4, 2005. at A16; Johnston, supra note 35.

39. Intelligence Shuffle, supra note 38.
40. Katherine Shrader, Q & A: How May Overhaul Affect Spy Agencies?, ASSOCIATED
PRESS, June 30, 2005, available at http://www.wtopnews.com/
index.php?nid=116&sid=537795.
41. Editorial, Reforms at the F.B.I. and Justice, N.Y. TIMES, July 1, 2005, at A18.
42. William E. Odom, Why the F.B.I. Can’t Be Reformed, WASH. POST, June 29, 2005,
at A21.
43. Douglas Jehl, Top Spy’s No. 2 Tells of Changes to Avoid Error, N.Y. TIMES, July
29, 2005, at A1; Terence Samuel et al., Everyone Knows It’s Nicer to Share, U.S. NEWS AND
WORLD REP., May 23, 2005, at 18 (stating10,000 classified websites and data repositories
and 125 databases have been made available since 9/11, amounting to 6000 bound copies of
the Encyclopedia Britannica).
44. Jehl, supra note 43.
45. Id.
46. Id.
47. Id.
48. Id.
WALD FINAL 5/19/2006 9:30 AM
“forthright about what they did not know.”49 One quite specific change Hayden
mentioned was that “the circle of senior intelligence officials required to be
given detailed knowledge about other agencies’s” sources would be widened.50
With respect to the NIE, Hayden (a participant in the Iraq NIE) pointed out that
each of the fifteen separate intelligence unit chiefs endorsed the final product
with only limited knowledge of the areas within the expertise of other units; in
Hayden’s case, that was NSA intercepts and communications. Now each
agency whose sources are a basis for part of the NIE must articulate to its
colleague agencies its specific “confidence in the source,”51 and presumably
the reasons why. (On a personal note, it is easy to see how in the Iraq NIE, the
CIA’s disclosure of its limited contacts with Curveball and the extent of its
heavy reliance on him, as well as the basis for the National Ground Intelligence
Center’s (NGIC) finding that the centrifuges would not likely be rocket-related,
might well have changed the direction, or at least the certainty, of that now
notoriously mistaken document.) According to Hayden, this new NIE rule had
already been in effect for months when he testified and had produced a quite
differently toned NIE on Iran.
Other reported changes directly related to the recommendations—
including more information sharing by CIA operatives in the field with the
analysts about their sources,52 and changes in the format and focus of the daily
Presidential briefings to include non-CIA sources—were “part of a broader
overhaul intended to add to the quality and breadth of intelligence provided to
policy makers” and to call attention to differences between agencies’ views of
critical intelligence issues as well as “to highlight the limits of our
knowledge.”53 The NGIC, singled out by the Commission for special criticism
of its analysis of the centrifuges, said it had instituted changes in training and
procedures to improve future analyses;54 the new HUMINT human resources
agency inside the CIA to instigate innovations in the recruitment, training and
usage of all community-wide human assets has been created, although the
Senate Intelligence Committee reportedly wanted to place it under the DNI’s
direct control;55 an Open Source Directorate to initiate and innovate more
widespread accessibility and use of open source information by analysts has
49. Id.
50. Id.
51. Id.
52. Walter Pincus, Intelligence Information Emphasized, WASH. POST, July 20, 2005,
at A4.
53. Douglas Jehl, Intelligence Gathering for Bush Is Overhauled, N.Y. TIMES, July 20,
2005, at A18.
54. Walter Pincus, Negroponte to Review Intelligence Changes, WASH. POST, July 2,
2005, at A18.
55. Walter Pincus, CIA Spies Get a New Home Base, WASH. POST, Oct. 14, 2005, at
A6; Walter Pincus, CIA to Remain Coordinator of Overseas Spying, WASH. POST, Oct. 13,
2005, at A4; Walter Pincus, New Office to Oversee Intelligence Abroad, WASH. POST, June
29, 2005, at A19.
WALD FINAL 5/19/2006 9:30 AM
been established;56 and a National Counter Proliferation Center will act as a

tasking and coordinating group for analyzing and collecting information on
WMD across the community.57 Mission managers have been allocated for “a
short list of top-priority challenges” including North Korea and Iran.58 A new
plan to target U.S. assets of banks and businesses conducting transactions with
listed Iranian, North Korean and Syrian companies involved in illegal weapons
trade has been announced.59 Two Commission recommendations, however,
were rejected: one classified, and the other suggesting sanctions against two of
the intelligence units whose mistakes in the Iraq NIE appeared most
inexcusable.60
The new CIA Director has also announced that he intends to send more
case officers and analysts abroad and to encourage them to engage in
“calculated risk-taking” in new covert arrangements—“We are going to be in
places people can't even imagine.”61 “Analysis is the engine that drives the
CIA,” he said, and further promised more competitive analysis in the future.62
“We are not afraid to publish opposing perspectives, if they exist. This gives
policymakers more with which to work.”63 And, finally, the DNI has
conducted, per the Commission’s suggestions, “a high-level review” of the
satellite programs to avoid “sucking resources from the rest of the intelligence
budget.”64
Of course it’s too soon to crow, and the proof will be in the pudding just
now settling in the oven. Still, it bears some reflection why this much has
gotten done so quickly. The answer, I think, lies in three factors: the speed and
relative efficiency with which the investigation got done so that its acceptance
and implementation was of political as well as substantive value to the
Administration; the unanimity and nonpartisan nature of its findings and the
fact that it worked quietly and without public fanfare; and finally, its
employment of staff who knew not only how the intelligence community
worked but where the bodies were buried. But other Commission reports,
gathering dust on archival shelves, have had all these attributes, too. In the end,
56. Scott Shane, Intelligence Center is Created for Unclassified Information, N.Y.
TIMES, Nov. 9, 2005, at A17.
57. Pincus, New Office to Oversee Intelligence Abroad, supra note 55, at A19.
58. Douglas Jehl, North Korea and Iran Win Special Notice at Spy Center, N.Y.
TIMES, Jan. 12, 2006, at A13.
59. Dafina Linzer, U.S. Plans New Tool to Halt Spread of Weapons, WASH. POST,
June 27, 2005, at A1.
60. Douglas Jehl, White House Is Said to Reject Panel’s Call for a Greater Pentagon
Role in Covert Operations, N.Y. TIMES, June 28, 2005, at A6.
61. Walter Pincus, Goss Plans To Expand CIA Spying and Analysis, WASH. POST,
Sept. 23, 2005, at A6.
62. Id.
63. Id.
64. WMD REPORT, supra note 2, at 546-47; Douglas Jehl, Review Leads to an
Upheaval in Spy Satellite Programs, N.Y. TIMES, Sept. 30, 2005, at A16.
WALD FINAL 5/19/2006 9:30 AM
I attribute much of our success in getting the report implemented to the

circumstance that at least a few Commission members had close enough
personal and professional ties with senior policymakers that they could—more
importantly they did—insist on getting who and what we needed to investigate
and on accessing those policymakers to urge quick action on our
recommendations. The Commission had two meetings—relatively lengthy and
substantive—with the President, relevant Cabinet members and national
security advisors. The co-chairs had additional numerous contacts with key
officials. The report also came out at a time when—in part due to the Iraqi
intelligence failures but in part due also to the new situation created by the
Intelligence Reform Act of 2004—top jobs were opening up in the community.
In several instances, Commission staff ended up in those jobs, bringing with
them their most recent commitments to the reforms in the report. And while
controversy raged in some quarters about whether Administration officials had
made overblown or inaccurate statements about the intelligence they did
receive in the prewar months, it was distinctly to the Administration’s
advantage to cooperate fully in disclosures that the intelligence they got was
faulty and in following through with earlier promises to fix it. All these factors
went into a warm welcome for the report and an unusually comprehensive
implementation of its reforms.65
Still, as the Commission recognized, structural changes alone are never
enough to reform institutions. Leadership is paramount; of late, the media is
starting to carry more critical stories of the IC, stating that the CIA is “bleeding
talent,” “adrift,” and morale is “crippled.”66 And the DNI has not escaped flak
for “not paying enough attention to the pressing need to pull together the 15
autonomous intelligence agencies,” instead laying too “light a hand” in
managing them, concentrating on his role as the President’s chief advisor and
on producing a national intelligence strategy. Concerns that the 2004
Intelligence Reform Act does not provide the DNI with sufficient authority to
manage the IC have resurfaced, especially with respect to the Department of
Defense, which has increased its own intelligence activities exponentially in
the last year.67 “Information sharing is an area where the progress has been
65. See, e.g., Douglas Jehl & Richard W. Stevenson, A Reminder of How Debate Over
Prewar Intelligence Continues to Shadow Bush, N.Y. TIMES, Nov. 15, 2005, at A23; Richard
W. Stevenson, Bush Contends Partisan Critics Hurt War Effort, N.Y. TIMES, Nov. 12, 2005,
at A1 (“After simmering for much of this year, the issue of how the Administration used
prewar intelligence has boiled over again in the last few months.”).
66. Dafina Linzer, A Year Later, Goss's CIA Is Still in Turmoil, WASH. POST, Oct. 19,
2005, at A1; see also Walter Pincus & Dafina Linzer, Panel Briefed by Spy Manager Who
Quit, WASH. POST, Sept. 22, 2005, at A9.
67. WMD REPORT, supra note 2, at 342 (urging “assertive leadership” by the DNI);
WMD Report at 539 (stating Commission’s most prominent theme is “stronger and more
centralized management of the Intelligence Community”); Siobhan Gorman, Broad Duties
Test Spy Chiefs, BALT. SUN, Oct. 16, 2005, at A1; Douglas Jehl, Little Authority for New
Intelligence Post, N.Y. TIMES, Oct. 14, 2005, at A16; Walter Pincus, Negroponte Unveils
Intelligence Strategy, WASH. POST, Oct. 23, 2005, at A9.
WALD FINAL 5/19/2006 9:30 AM
slower than hoped,” a member of the House Intelligence Committee

complained.68 He warned that at some early point the DNI is going to have to
“pick a battle with one of the agencies—and win.”69
III. THE ANALYST: A (WO)MAN IN THE MIDDLE
So far, I have dealt with the Commission’s critique of and

recommendations for reform of the environment in which the IC analyst is
recruited, trained, equipped, evaluated, and rewarded (or not). I now turn
toward an area in which the Commission’s inquiry was limited and in which I
believe a deeper institutional analysis is merited. That is the area of analyst-
policymaker relationships.
This was an aspect of the Commission’s report that attracted both attention
and criticism not for what it said but for what it did not say. The WMD
Commission said nothing about how the policymakers used the intelligence
that was given them by the IC—whether they hyped it, misstated it,
exaggerated it or even made some up. This omission was labeled “timidity” by
The New York Times and worse, puppetry, naiveté, or stupidity by others.70 The
response is a straightforward one—the Commission’s mandate was explicitly
limited to the capabilities of the intelligence community; there was never any
question that we were not authorized to evaluate the policymakers’ use of that
intelligence. Indeed, apart from the express words of the mandate, it is
unrealistic to think that any President would appoint a bipartisan Commission
to critique his or his top officials’ personal credibility in an election year.
Commission members—particularly those of the opposite party—might of
course be criticized for being satisfied with half a loaf when a comprehensive
look at how intelligence affected the Iraq War would require a perusal of the
actions of both analyst and policymaker. But having taken on a designated task,
I am not persuaded the Commission could ethically have plunged ahead like a
runaway jury to explore whatever it thought necessary.71 My own view is that it
was a service—albeit a limited one—to do the intelligence piece of the puzzle
while the evidence, and the personnel involved, were still fresh, rather than
waiting until the “right” Administration came along to do the complete job.
68. Gorman, supra note 67.

69. Id.
70. Editorial, A Profile in Timidity, N.Y. TIMES, Apr. 1, 2005, at A22.
71. As the debate about policymakers’ use of prewar intelligence heated up again in
2005, there were several media reports clarifying the Commission’s limited inquiry into the
intelligence sources themselves rather than policymakers’ use of them. See, e.g., Dana
Milbank & Walter Pincus, Asterisks Dot White House’s Iraq Argument, WASH. POST, Nov.
12, 2005, at A1 (“[T]he Commissions cited by officials, though concluding that the
administration did not pressure intelligence analysts to change their conclusions, were not
authorized to determine whether the administration exaggerated or distorted those
conclusions.”).
WALD FINAL 5/19/2006 9:30 AM
The SSCI, which set out to do its own study of both parts of the Iraqi fiasco but
which was proceeding extremely slowly in the second phase, was galvanized
into action by a two-hour Senate shutdown precipitated by the Democrats in
protest of its inaction.72 Fortunately, that follow-up examination is now being
done, for the issue continues to smolder in the public discourse.73 But had the
Commission attempted to do it in 2004, I suspect we would have destroyed the
consensus we were able to reach on the performance of the intelligence
community.
The sole aspect of the analyst-policymaker relationship with which the
Commission did deal, though not to everyone’s satisfaction, involved the issue
of whether the Iraq analysts were unduly pressured by the policymakers into
inflating their WMD findings. The Commission concluded that, despite the
grave errors of judgment and sloppy workmanship, the analysts’ findings that
Saddam Hussein had reconstituted his WMD program were not the result of
political pressure from the policymakers.74 This finding was received with
skepticism in many quarters, including respected think tanks and mainstream
media, as well as anti-war activists. For instance, reports that Vice President
Cheney made ten trips to Langley to talk with analysts and recurrent
complaints by analysts of persistent questioning from policymakers were cited
in newspaper reports as evidence of undue pressure.75 Yet, intensive
questioning of the Iraq analysts by the Commission and its staff, as well as
their sworn testimony before the Senate Intelligence Committee, produced a
chorus of denials by the analysts that their WMD conclusions reflected a
reaction to political pressures.76 Such denials of course do not by themselves
carry the day for all skeptics; they could reflect limitations of the questions
asked of the analysts or even self-denial of their real motivation. More
importantly, they do not necessarily resolve a number of problematic aspects of
policymakers’ and analysts’ roles in the intelligence-making process. These
issues arise not only from the unique structure of the IC itself and its sui
generis place in the Executive Branch of the government, but also from an
increasing tension throughout that Branch between the professional demands
for independent fact-finding and the imperatives of potent policymaking.
All of the Commission’s recommendations about the treatment and
performance of policymakers will be nullified if in the end they succumb to
72. Stevenson, supra note 62; Carl Hulse & David D. Kirkpatrick, Partisan Quarrel
Forces Senators to Bar the Doors, N.Y. TIMES, Nov. 2, 2005, at A1, A13-14; Walter Pincus,
Intelligence Probe Takes Shape, WASH. POST, Nov. 10, 2005, at A7.
73. But see Stevenson, supra note 65 (“[T]hat inquiry is unlikely to be completed any
time soon . . . .”).
74. WMD REPORT, supra note 2, at 8, 11, 51, 188-92; see also Eric Lichtblau, No
Evidence of Pressure on Iraq Data, N.Y. TIMES, Nov. 7, 2005, at A14.
75. Walter Pincus & Dana Priest, Some Iraq Analysts Fear Pressure from Cheney
Visits, WASH. POST, June 5, 2005, at A1; Jim VandeHei & Walter Pincus, Cheney’s Office Is
a Focus in Leak Case, WASH. POST, Oct. 18, 2005, at A1.
76. Lichtblau, supra note 74, at A14.
WALD FINAL 5/19/2006 9:30 AM
pressures by policymakers to make their analyses come out one way rather than
another. Thus I will try to explore what the contours of that relationship are and
what possible additional external restraints might be placed upon it, as well as
what internal controls the IC might adopt to better protect the analyst who faces
an intensified version of a dilemma that is—in the view of some—coming
more and more to permeate Executive Branch professional service.
A. Who Does the Analyst Work For?
The intelligence agencies are part of the Executive Branch. The new
Director of National Intelligence (DNI), like the former Director of Central
Intelligence (DCI), serves at the pleasure of the President, not for a specified
term like the FBI Director or the Comptroller General. Fourteen of the fifteen
intelligence units that make up the IC are located inside other Departments—
Defense, Homeland Security, Energy, Justice, and Treasury; the CIA stands
alone as an independent executive agency. Except for the CIA, these agencies
serve under the direction of their Cabinet secretaries as well as the DNI. The
basic statutory function of all these intelligence units is to provide intelligence
on request to Executive Branch officials and to Congress, which provides
oversight of the IC through its House and Senate Intelligence Committees.77 In
short, the IC’s “customers” are the policymakers, and so far as tactical
intelligence is concerned, the military command structure. These officials ask
the IC daily for specific pieces of information as well as periodically for more
comprehensive surveys of issues involved in foreign policy and military
operations. According to the Commission:
Analysts are the link between customers and the Intelligence Community. They
provide a conduit for providing intelligence to customers and for conveying the
needs and interests of customers to collectors. This role requires analysts to
perform a number of functions. Analysts must assess the available information
and place it in context. They must clearly and concisely communicate the
information they have, the information they need, the conclusions they draw from
the data, and their doubts about the credibility of the information or the validity of
their conclusions. They must understand the questions policymakers ask, those
they are likely to ask, and those they should ask; the information needed to answer
those questions; and the best mechanisms for finding that information. And as
analysts are gaining unprecedented and critically important access to operations
traffic, they must also become security gatekeepers, revealing enough about the
sources for policymakers to evaluate their reporting and conclusions, but not
enough to disclose tightly-held, source-identifying details.78
77. WMD Report, supra note 2, at 416 (“‘Typical’ customers include not only the
President and senior policymakers, but also members of Congress, military commanders,
desk officers in executive agencies, law enforcement officers, customs and border patrol
officials, and military units in the field.”).
78. Id. (emphasis added).
WALD FINAL 5/19/2006 9:30 AM
But obviously that is not the end of the matter: intuitively, the analyst’s job
is perceived as more than one of subserviently doing the bidding of the
policymaker. It is not the same, for instance, as the relationship between a
judge and a law clerk, or even a policymaker and her special assistant who asks
the aide to research a question and provide a recommendation, but who in the
final analysis makes her own judgment about whether to accept it and may
require the assistant to redo the assignment so as to make the best arguments
and find the most persuasive factual basis for a different position that the
assistant may not agree with. The analog to the analyst’s predicament is, after
five decades, still not clear. What is the analyst’s responsibility when he does
not agree with what he knows to be the policymaker’s desired result?
We can agree that the analyst’s “customer” is the policymaker, and it is the
policymaker whom she should serve by providing as much useful information
to him as possible. But should “useful” in this context be defined solely by the
policymaker? Indeed, discussions by intelligence insiders stress the dominance
of policymakers in the analyst-policymaker relationship. “[T]he needs of
policymakers—especially what hands-on officials seek in support of their daily
management of issues—have to be a central concern of intelligence makers.
This is no trivial point.”79 At the same time, however, they recognize that “the
national interest is not well-served by an assessment much admired by policy
officials that does not meet the standards of sound analytic tradecraft,”80 and
warn that “emphasis on close support of policy officials encourages
politicization of analysis.”81 But “[i]f effective relations between intelligence
and policy professions require the former to provide close support to the daily
business of the latter, any mission statement that seals off intelligence analysis
from politics seals it off from a significant role in the policymaking process.”82
Thus, “[If] the President wants to defy what the analysts see as the ground truth
in country X, analysts are professionally obligated both to point out the long
odds and to provide judgments and insights to shorten those odds.”83 The
marriage between tradecraft and policy relevance is an uneasy one.
In sum the analyst must maintain her professional ethic of not allowing her
own judgment to be distorted to satisfy a policymaker’s desire, but at the same
time to be effective she must maintain a good relationship by engaging in as
close a dialogue with the policymaker as is needed to perform the task.
But was that standard, or more accurately “guidance,” enough in a national
crisis like the one in which the Iraq NIE was issued? Given a general
awareness, inside as well as outside the IC, of the potential impact of the NIE
79. Jack Davis, Defining the Analyst’s Mission: Facts, Forecasts and Fortunetelling,
in Intelligence and the National Security Strategist: Enduring Issues and Challenges, 275,
295-97 (Roger Z. George & Robert D. Kline eds., 2004).
80. Id.
81. Id.
82. Id.
83. Id.
WALD FINAL 5/19/2006 9:30 AM
on impending executive and congressional action with respect to Iraq, as well

as a keen sense of how the Executive policymakers wanted their questions of
WMD in Iraq answered, is that reliance on the tradecraft tenets of objectivity in
fact-finding and analysis enough to keep analysts safe, or do they need
additional protections, including some “rules of engagement,” to restrain
policymakers?
First, let’s consider the importance of the context in which an analyst
makes a judgment. Is it not naiveté to think that analysts will block out
awareness of the real-world effects of their judgments in foreign policy
matters? The Commission found “the gathering conviction among analysts that
war with Iraq was inevitable” permeated the climate in which they worked.84
How could it be otherwise? That awareness indeed should actually accentuate
the importance of analysts giving the policymaker the benefit of their best
judgment based on what they know and do not know. Might there not be
different quality thresholds for different kinds of intelligence reports,
dependent openly upon their predictable consequences? Right now, the “one
size fits all” approach says that no matter how minor or major the inquiry, each
analytical piece deserves the same level of care and attention. Thus, an
intelligence estimate preparatory to deciding whether to put more agents in a
trouble spot or to initiate a covert or clandestine operation is conducted under
the same standards as one that is acknowledged to be a prerequisite to going to
war. Specifically, why would it not make sense to require a higher degree of
proof or evidence for a “finding” that will almost certainly lead to an
international crisis or war than one that is providing long-term strategic
planning data or tactical intelligence for the agencies’ more routine operations?
Consider the judicial model: the burden of proof necessary to stop a person on
the street for brief questioning is “reasonable suspicion” he may be about to
commit a crime; to arrest a suspect there must be “probable cause” to believe
he has committed the crime; to win a civil suit there needs to be a
“preponderance of the evidence” on the applicant’s side. An administrative
agency need only show “substantial evidence” (maybe less than 50%) to
withstand a judicial challenge, but to convict and imprison an accused there
must be proof beyond a reasonable doubt. The consequences dictate the degree
of confidence the system must have in the result of its deliberations. There are
other examples of proportionality of procedures and standards of proof fitted to
degrees of risk of error in the law. The Supreme Court has enumerated them:
“the . . . interest that will be affected by the official action . . . the risk of an
erroneous deprivation of such interest through the procedures used, and the
probable value, if any, of additional or substitute procedural safeguards.”85
Is it unnatural to think such a nuanced system could have application to

85. Mathews v. Eldridge, 424 U.S. 319, 335 (1976).
WALD FINAL 5/19/2006 9:30 AM
intelligence analysis which admittedly has to patch together pieces of disparate

information from sources of varying reliability and often has to use a pithy
database from which to infer momentous conclusions? Is it not common sense
to demand a higher level of confidence in concluding that a perceived enemy is
actively reconstituting a nuclear capability, including a delivery mechanism
that can reach the United States, than it is to send a commando raid into a
hostile country to see what their workers are actually doing inside a suspicious
plant? The WMD Commission found that awareness of the real-world context
in which analysts are plying their trade does affect their judgment. Our research
showed that the “general assumption that Saddam retained WMD and the
backdrop of impending war . . . affected the way analysts approached their task
of predicting the threat posed by Iraq’s WMD programs. . . . [T]his atmosphere
contributed to analysts’ use of a worst-case-scenario or heightened-burden-of-
proof approach to analysis . . . [and] to the too-ready willingness to accept
dubious information as supporting the conventional wisdom and to an
unwillingness even to consider the possibility that the conventional wisdom
was wrong.”86 If the expected consequences are likely to produce a heightened
degree of both external (by policymakers) and internal pressure on analysts,
why not explicitly build that awareness into the tradecraft in the form of a
higher burden of proof so as to impel analysts to be even more careful in their
fact-finding?
Yet that is not the way the process works now. Analysts—in good faith—
may believe foreign policy decisions are best made by elected officials and it is
not the analyst’s job to define the public interest insofar as foreign policy is
concerned. Thus, the DOE analysts who discounted the aluminum tube basis on
which other agencies found Iraqi nuclear reconstitution nonetheless voted to
support the majority opinion in the NIE which found that evidence compelling
in part because they “did not want to come out before the war and say [Iraq]
wasn’t reconstituting.”87 Although everyone may agree that the proper role of
the analyst is not a post-decisional one to back up the policymakers’ already-
arrived-at decisions, but rather a pre-decisional one to inform those decisions,
we know that the apparent policy implications of a finding or inference one
way or the other can subtly influence their judgments, particularly where the
evidence is ambiguous. Had the IC accurately evaluated the absence of
WMD—even just the absence of a nuclear weapons program—in pre-war Iraq
and said so, the real-world consequences in foreign policy direction might have
been quite different. Why should analysts not be challenged—even
encouraged—to use higher thresholds for findings where implications for
foreign policy are devastatingly apparent?
The way in which the Commission confronted accusations that
policymakers had pressured analysts into finding WMD reconstitution invoked

87. Id. at 190.
WALD FINAL 5/19/2006 9:30 AM
an admittedly narrow definition of “politicization”—whether analytical

judgments were actually changed as a result of policymakers’ interventions. It
did note the somewhat broader definition the CIA’s Ombudsman used in his
investigation of such complaints—whether there was “any unprofessional
manipulation of information and judgments” to please policymakers’
preferences.88 Based on interviews with those analysts, both the Commission
and the Ombudsman concluded there was not “politicization.”89
Critics who nonetheless accused the IC of “cooking their own books” do
make some valid points.90 The Commission did find some evidence that
analysts were subjected to pressure from within the agency, more than from
without. It cited examples where intermediate superiors discouraged contrary
opinions to the given wisdom that WMD existed and encouraged analysts to
fall into line: “In the case of pre-war assessments of Iraqi WMD, working-level
WINPAC analysts described an environment in which managers rewarded
judgments that fit the consensus view that Iraq had active WMD programs and
discouraged those that did not.”91
We criticized as well “a failure of management to actively foster
opposition views” and found this attitude continued even after post-war surveys
of the IC itself revealed no WMD.92 Inexplicably, at that point analysts who
pushed for explicit repudiation of the earlier pre-war assessments were
chastised and removed from the offending unit.93 In this sense, Iraq was a
dramatic but not an isolated example of the internal pressure for conformity: a
2004 survey revealed 17 percent of WINPAC (the anti-proliferation unit)
analysts said they worked “in an atmosphere in which some managers who
[held] strong views [made] it difficult to publish opposing points of view.”94
These findings suggest that even without any overt outside pressure on
analysts they can be affected by a climate of conformity and disapproval of
superiors if they deviate from accepted views. But, of course, those superiors
also are likely to be appointed by the top echelon of their agencies—those
leaders with the most frequent contacts with the highest level of policymakers
so that internal processes on some occasions may reflect a trickle-down of
policymaker pressures exerted at the highest levels. This is especially true
when top IC officials become publicly identified with a particular policy
position—a situation devoutly to be avoided. Granted, the leaders of the
intelligence community need to be perceived as having the confidence of the
political leaders from the President on down if their advice and judgment are to
88. Id. at 188 (footnote omitted).

89. Id. at 188.
90. See Rich, supra note 5.
91. WMD REPORT, supra note 2, at 191 (footnote omitted).
92. Id. at 191-93.
93. Id. at 193-94.
94. Id. at 191-92.
WALD FINAL 5/19/2006 9:30 AM
be taken seriously and if they are to access resources for their agencies. Some
Presidents have been known to cut their CIA directors out of the loop entirely.
Yet, once the IC leader voices a public opinion or predilection on an issue still
under consideration inside the IC, he can be sure it will plummet to the bowels
of the IC. Thus, no less than the analyst, like Caesar’s wife, the intelligence
chief should be above suspicion as to the impartiality and fact-based nature of
his judgments and reports to policymakers. The Commission did, however,
relate some disturbing past examples of changes in institutional judgments by
prior heads of the CIA to reflect policymakers’ preferences.95
Intelligence analysts, of course, do not operate in a free-standing
environment; they operate within bureaus and agencies and hierarchies,
subordinate to superiors who evaluate their work for competence and
reasonableness. A superior should retain the right to correct or reject an
analyst’s work on the basis of its intrinsic worth as analysts can be sloppy or
oblivious or just plain unpersuasive. There must be room for the ordinary give-
and-take between superiors and subordinates in the IC as elsewhere in the
Executive Branch.
But, the right to find or argue responsibly against the prevailing views of
an agency seems still not to be stabilized in practice. It should and hopefully
will now not only be part of the analyst’s initial training, but also a part of the
continuing on-the-job education of intermediate managers and superiors who
pass on the analyst’s work and do his evaluations. Only where the managers
support the analysts’ independence up the line can it be achieved. Thus, it was
a comfort to read in the John Bolton confirmation hearings that analysts whose
judgments allegedly were repeatedly being challenged by Mr. Bolton were
protected from alleged retaliatory actions by their superiors in the State
Department.96
The WMD Commission sought to institutionalize mechanisms of dissent,
alternative analyses, and “Red Teams” to argue a devil’s advocate position on
important issues. In that way the policymaker can at least be apprised of the
best thinking in the IC even if it runs contrary to the intelligence community’s
ultimate position. If Deputy DNI Hayden’s testimony is credited, that is
happening now with respect to the NIE which may be less “definitive” than
those of old.97 But, candidly, not every request for information or even every
NIE needs or can expect to get a systematic development of alternative
hypotheses or a devil’s advocate exercise even though palpably crucial ones
like the Iraq NIE should. Even in these less dramatic cases, however, the
internal agency position and culture may require changes in the direction of
95. Id. at 248 n.847.

96. Glenn Kessler, As Vote Nears, Focus is on Bolton’s Actions, WASH. POST, May 12,
2005, at A5.
97. See Jehl, supra note 43; see also Maura Reynolds, Intelligence Office Gives
Dissenters Their Due, L.A. TIMES, April 14, 2006 at 1 (ten top intelligence officers report
they now incorporate dissenters’ views into their analyses).
WALD FINAL 5/19/2006 9:30 AM
tolerance for dissent. For example, it may be useful to keep systematically an

internal record of serious differences or disputes between analysts and their
superiors that is available to later investigators or ombudsmen. The
Commission itself urged that “[a]nalysts . . . readily bring disagreement within
the Community to policymakers’ attention, and . . . be ready to explain the
basis for the disagreement.”98
In the end, the integrity of the analytic process depends on the analyst’s
own and the IC’s perception of her role, as well as the agency’s tolerance and
encouragement of responsible contrary views. Although egregious incidents of
policymakers exerting overt pressures to affect intelligence judgments may
occur, the Iraq experience suggests the more sinister influences are likely to be
subtle ones that permeate the internal workings of the IC itself.
B. Rules of Engagement: When Do Policymakers Cross the Line?
There is, however, no doubt that on occasion determined policymakers do

make their desired results known to the analysts. The IC promotes the ideal that
“close professional ties between analysts and policymakers usually promote
frankness, mutual respect, and even mutual dependence . . . these in turn
promote analysis to help understand how to deal with tough problems, not
analysis to please.”99 The WMD Report, as well, advised that:
Good-faith efforts by intelligence consumers to understand the bases for analytical
judgments, far from constituting “politicization,” are entirely legitimate. This is
the case even if policymakers raise questions because they do not like the
conclusions or are seeking evidence to support policy preferences. Those who
must use intelligence are entitled to insist that they be fully informed as to both the
evidence and the analysis.100
So far, so good. But this ideal posits two government servants searching in
tandem for the best intelligence available with which to make public policy.
There is of course a darker side—the policymaker hell-bent on getting facts and
judgments from intelligence analysts to back up his preferred position. The
former head of intelligence at the Department of State unequivocally
announced: “policymakers never once applied any pressure on coming up with
the ‘right’ answer on Iraq.”101 DOE analysts testified similarly. Not a single
98. WMD REPORT, supra note 2, at 419 (the Commission also urged analysts across
the IC to explain in a common parlance to policymakers varying degrees of certainty in their
work).
99. Davis, supra note 79, at 287.
100. WMD REPORT, supra note 2, at 83 (quoting CENTRAL INTELLIGENCE AGENCY,
DCI NONPROLIFERATION CENTER, NEW EVIDENCE OF IRAQI BIOLOGICAL WARFARE PROGRAM
(SIR 2000-003X) (Dec. 14, 2000)).
101. Id. at 83 (quoting CENTRAL INTELLIGENCE AGENCY, DCI CENTER FOR WEAPONS,
INTELLIGENCE, NONPROLIFERATION, AND ARMS CONTROL, IRAQ: MOBILE BIOLOGICAL
WARFARE AGENT PRODUCTION CAPABILITY (WINPAC IA 2001-050X) (2001) at 1, 7).
WALD FINAL 5/19/2006 9:30 AM
analyst interviewed said his judgment had been affected by policymaker

pressure.102 Yet, much has been made of the fact that Vice President Cheney
made ten visits to intelligence analysts while they were preparing various
reports on WMD (not the NIE).103 And one skeptic has commented to me
personally that in business, if a mid-level official working on a report gets
several visits from the CEO, he quickly gets the message of how his report
should read. (I’m not sure that’s right; the message may be only: “This is
important. Do it right.”) The question, of course, is when do insistent
questioning and repeated demands for more evidence to support the
policymakers’ conclusions become illegitimate pressure? Should it depend
primarily on the individual self-confidence or seniority of the analyst to
withstand excessive demands? Is that the gold standard for the relationship? Or
do policymakers have some obligation to adhere to restraints on the intensity
and level of their engagement with analysts?
Given the inevitable psychological advantage of a high or even mid-level
policymaker in one-on-one interactions with the analyst, there is risk that even
without any overt demands on the part of the policymaker the analyst may alter
her judgments based on what she thinks the policymaker wants to hear and how
his good graces might benefit or injure her career. One of the State Department
officials interviewed suggested that arm twisting is rarely used or needed; some
analysts may engage in self-censorship to retain good relationships with their
policymaking clients. When that happens, it is hard to discern from the outside
unless an analyst’s conclusions suddenly and sharply vary from prior
expressions on the same subject. Self-originating changes will more likely be in
tone or emphasis. Thus, the Commission report compares pre-NIE judgments
of the IC that Iraq “could” have biological weapons to the NIE estimate that
Iraq “has” biological weapons. In 2000, a report stated that “credible reporting
from a single source (Curveball) suggests” Iraq has produced biological
102. Id. at 80 (quoting NIE, supra note 12, at 5, 35).

103. See Hersh, supra note 3; VandeHei & Pincus, supra note 75. Though it was not in
the Commission’s purview of inquiry, many media reports have focused on alleged
pressures from the Vice President’s office to find a connection between Iraq and al Qaeda.
One quotes Richard Kerr, former Deputy CIA Director, to the effect that in his thirty-two
years of service he had never seen such “hammering” by an administration on finding
evidence of the link. Editorial, Decoding Mr. Bush’s Denials, N.Y. TIMES, Nov. 15, 2005, at
A28. Another cited the SSCI Report as describing “repeated unsuccessful efforts by the
White House and its allies in the Pentagon to persuade the CIA to embrace the view that Iraq
had provided support to al Qaeda,” including draft speeches rejected by Secretary Powell
and CIA Director George Tenet as exaggerated and unsubstantiated. Stevenson, supra note
62. David Ignatius describes a Cheney aide’s insistence on establishing the link by
identifying Iraq intelligence agents in a photo with Mohammed Atta in Prague despite CIA
skepticism: “The CIA would literally measure ears and noses in surveillance photos of the
alleged meeting to show the report was phony, but Cheney’s aides would tell them to go
back again, and then again”; “In a heated conversation . . . CIA Deputy Director John
McLaughlin is said to have insisted: ‘I’m not going back to the well on this. We’ve done our
work.’” David Ignatius, Op-Ed., The Real Crime: White House vs. CIA was the Wrong
Battle, WASH. POST, Oct. 30, 2005, at B7.
WALD FINAL 5/19/2006 9:30 AM
weapons but “[w]e cannot confirm” that it has.104 A year later, based on the
same source, a report pronounced “that Iraq continues to produce at least . . .
three BW agents.”105 In 2002, the CIA Director told Congress, “we know Iraq
has developed a . . . capability to produce biological warfare agents,”106 and the
2002 NIE reported with “high confidence” that Iraq not only had biological
weapons but a program, “all key aspects” of which were active and that most
elements were larger and more advanced than they were before the Gulf War.
107
Although this example does not implicate policymaker pressure, it
illustrates the importance of subtle evolutions in the inferences drawn from
ambiguous evidence on the same topic. Self-inflicted subservience can be
offset only by training in a culture and ethic of independence instilled in the
analyst from the start of her career and in a strictly enforced requirement that
the analyst state clearly all the evidence and assumptions underlying her
conclusion.
On the other hand, some policymakers are not above pursuing more
powerful kinds of pressures. As the Bolton hearings were being held, Robert
Hutching, a former Chairman of the National Intelligence Council, suggested
that pressuring technique is part of the “pattern and practice” of some
policymakers.
This is not just about the behavior of a few individuals but about a culture that
permitted them to continue trying to skew the intelligence to suit their policy
agenda -- even after it became clear that we as a government had so badly missed
the call on Iraq W.M.D. . . . When policy officials come back day after day with
the same complaint and the same instruction to dig deeper for evidence to support
their preformed conclusions, that is politicization. . . . When those officials seek to
remove from office analysts whose views they do not like, that is politicization.
The mere effort, even when it is successfully resisted, creates a climate of
intimidation . . . . 108
The Commission insisted, however:
[T]he customer should challenge the analyst’s assumptions and reasoning.
Because they are “keepers of the facts,” analysts can play a decisive role in policy
debates, a role that has temptations for analysts with strong policy views of their
104. WMD Report, supra note 2, at 83 (quoting CENTRAL INTELLIGENCE AGENCY, DCI
NONPROLIFERATION CENTER, NEW EVIDENCE OF IRAQI BIOLOGICAL WARFARE PROGRAM (SIR
2000-003X) (Dec. 14, 2000)).
105. Id. (quoting CENTRAL INTELLIGENCE AGENCY, DCI CENTER FOR WEAPONS,
INTELLIGENCE, NONPROLIFERATION, AND ARMS CONTROL, IRAQ: MOBILE BIOLOGICAL
WARFARE AGENT PRODUCTION CAPABILITY (WINPAC IA 2001-050X) (Oct. 10, 2001) at 1,
7).
106. Id. at 84 (quoting interviews with CIA Iraq WMD Review Group analysts (Aug.
3, 2004 and Sept. 20, 2004)).
107. Id. at 90 (quoting NAT’L INTELLIGENCE COUNCIL, NATIONAL INTELLIGENCE
ESTIMATE, IRAQ’S CONTINUING PROGRAMS FOR WEAPONS OF MASS DESTRUCTION 5, 35 (Oct.
2002)).
108. Douglas Jehl, Tug of War: Intelligence vs. Politics, N.Y. TIMES, May 8, 2005, at
A12.
WALD FINAL 5/19/2006 9:30 AM
own. A searching examination of the underlying evidence for the analysts’ factual
assertions is the best way to reassure policymakers that the analysts’ assertions are
well-grounded. We reject any contention that such engagement is in itself
inappropriate or that the risk of “politicizing” intelligence cannot be overcome by
clear statements to analysts as to the purpose of the dialogue.109
Vigorous—even acrimonious—exchanges are not unusual in Washington
among politicians, Congressmen and witnesses, in Cabinet meetings and
bureaucratic deliberations, even occasionally in courtrooms. But should there
be limits to how far ardent policymakers can go in trying to bring analysts into
their camp? The policymakers are usually more powerful politically than the
analysts so it is not an even playing field.
Even so, there are strategic advantages on both sides that can affect the
relationship. The important question for our purposes is whether the analyst’s
stakes need shoring up to ensure she can hold her ground without fear of
draconian consequences.
Policymakers often need intelligence to further their cause. The
policymaker has options to be sure, but they are limited. The policymaker is, at
bottom, not required to seek intelligence at all in his policymaking if he
chooses not to, or he is entitled to reject what he gets and seek other bases on
which to act. Of course the realpolitik is that his policy judgments are likely to
be more persuasive if he can explicitly assert they are backed by the IC or
implicitly create that impression (as Colin Powell did by having George Tenet,
the DCI, sit alongside him at the notorious U.N. speech on Iraq). Though
sometimes the policymaker does not refer directly to an intelligence base, he
strongly suggests it by using such phrases as “we do know that” or “there’s no
question that . . . .”110 It is also an option for policymakers who do not like or
trust the intelligence they are receiving to seek alternate sources inside the IC;
indeed, an inquiry into alternative views should ordinarily be encouraged when
reasonable doubts are raised by a policymaker about one analyst’s work. (This
is of course quite different from a search for alternative views outside of
regular IC channels which leads to stovepipe operations such as the Office of
Special Plans in the Defense Department, which reportedly provided a channel
for Iraqi defectors to reach policymakers directly without going through the
normal intelligence processes.) A clear line for policymakers to draw is that
they not create the impression of relying on the IC as the source of their facts or
conclusions when it does not support them.
Thus, there is a premium on the policymaker’s securing the imprimatur of
the IC as to the intelligence on which he bases his policy decision. To this end,
the policymaker may legitimately press the analyst within reasonable limits to
pursue all leads, or even argue with her on the inferences she draws from
available evidence. The policymaker may even look for another analyst to see
109. WMD REPORT, supra note 2, at 418 (emphasis added).

110. See Editorial, They lied, WASH. POST, Sept. 22, 2005, at A6 (quoting Rumsfeld,
Rice and Bush on WMD in 2000-02).
WALD FINAL 5/19/2006 9:30 AM
if he can get a different opinion, not in itself always a bad thing. But there
seems to be agreement that to try to have the resistant analyst herself
transferred or removed is considered out of range.
Former Reagan official Morton Abramowitz said he knew of policymakers
who sought another analyst when they were unhappy with the one they had, but
it was “somewhat unusual” to try to transfer or remove the offending analyst.111
Abramowitz added that intelligence is by its nature so uncertain it is easy for
policymakers “to pick the intelligence they like.”112 Former Deputy CIA
Director McLaughlin agreed: “it’s perfectly all right” for a policymaker to
disagree with an analyst and challenge his work vigorously, but it is something
different to ask for the analyst’s transfer because of disagreements “unless
there is malfeasance involved.”113
The current consensus thus seems to be that a policymaker crosses the line
in retaliating against an analyst by seeking her transfer because of disagreement
about what inferences or conclusions the intelligence will bear. When that
occurs, the analyst’s superior should refuse the policymaker’s request, as was
apparently done in the alleged Bolton incident. Indeed, Secretary of State
Powell reportedly visited the beleaguered analysts after one such incident to
reassure them of the Department’s support.
There is at least one other area in which policymakers cannot tread, and
that is in altering analysts’ reports themselves. That is in marked contrast to
other Executive Branch agencies where a lively dispute is ongoing as to
whether and how much political appointees may edit and revise work that
emerges from the “scientific” fact-finders. Thus, a recent controversy centered
on changes made by a White House Council on Environmental Quality official
to estimates by the scientific staff on the link between emissions and global
warming, thereby producing what The New York Times called “an air of doubt
about findings that most climate experts say are robust.”114 The White House
responded that the “changes . . . were part of the normal interagency review
that takes place on all documents related to global environmental change.”115
But others said the White House reviewers were not scientists and so were not
entitled to make those changes. The National Academy of Science warned that
“the administration’s procedures for vetting reports on climate could result in
excessive political interference with science.”116
111. Kessler, supra note 96, at A5.

112. Id.
113. Id.
114. Andrew Revkin, Bush Aide Edited Climate Reports, N.Y. TIMES, June 8, 2005, at
A1.
115. Id.
116. Id.; see also Marc Kaufman, Review of ‘Plan B’ is Faulted, WASH. POST, Nov.
15, 2005, at A1 (Government Accountability Office calls FDA rejection of over-the-counter
sale of “morning after” pill before staff conducted scientific study “unusual”); Daniel Smith,
Political Science, N.Y. TIMES, Sept. 4, 2005, § 6 (Magazine), at 37; Editorial, Strange
WALD FINAL 5/19/2006 9:30 AM
Although the IC is not a “science agency” like the National Institutes of

Health or the National Science Foundation, there appears to be a strong if not
absolute presumption that policymakers may not unilaterally change or alter the
formal conclusions or findings of the intelligence agencies to suit their
purposes. Thus, the policymakers could not take an IC report concluding there
was no WMD in Iraq and change it to a conclusion that there was. This is not
true of all agencies, where the top level of political appointees apparently can
and do make changes to inferences drawn from the facts and even as to the
sources of the “facts” themselves.
But sometimes the policymaker does not try to change the underlying
intelligence judgment itself but insists either on drawing his own inferences
from the data provided or, more egregiously, misstates what the intelligence
consensus is. A Powell aide said during the Bolton hearings that special vetting
instructions were given for the Deputy Secretary of State to review personally
all of the nominee’s speeches before they were given to assure that the IC’s
positions were not misstated.117 The Chairman of the National Intelligence
Council said: “I wouldn’t say he was making up facts. . . . Let’s say that he
took isolated facts and made much more of them to build a case than I thought
the intelligence warranted. It was a sort of cherry-picking of little factoids and
little isolated bits that were drawn out to present the starkest-possible case.”118
Bolton on the other hand argued that a policymaker had a right to state “his
own reading of the intelligence” even when it differed from that of the
intelligence agencies so long as he does not purport to speak for the
community.119 In Bolton’s case, it was asserted that he had (unsuccessfully)
tried to limit internal review of his speeches solely to ensure that no secret
sources or methods were revealed, and not to permit critiquing of his
conclusions about what the intelligence showed. He wanted analysts to be
instructed not to refute what he said but only to offer additional evidence in
support of his views.
Bolton was correct to the extent that a policymaker may express his own
conclusions about what data mean so long as he makes it clear they are his and
not the IC’s. But, if a policymaker decides to take a different or stronger stand
than the analyst’s based on the underlying data in an IC report, must he
disclose that difference in his public utterances? Obviously, ordinary listeners
to a policymaker’s pronouncements that “we know,” “we believe,” or “there is
no question that” country X has WMD will assume official intelligence
estimates back up his statements. Thus, the State Department’s intelligence
bureau reportedly complained on one occasion that Mr. Bolton’s proposed
Behavior at the FDA, N.Y. TIMES, Nov. 15, 2005, at A26.

117. Douglas Jehl, No. 2 at State Dept. Was Said to Put Restrictions on Bolton, N.Y.
TIMES, May 10, 2005, at A9.
118. Id.; see also Douglas Jehl, Report Warned Bush Team About Intelligence Doubts,
N.Y. TIMES, Nov. 6, 2005, at A1.
119. Id.
WALD FINAL 5/19/2006 9:30 AM
language in a speech on Cuba’s possession of biological weapons and terrorism

would be perceived as an official judgment even though it did not comport with
the IC’s evaluation.120 It is currently not clear if the policymaker must
announce her differences with the IC.
Is it too much of an infringement on his political speech that the
policymaker be required to spell out any differences he has with the IC in
public? This is a crucial but still unsettled tension for the IC. May he rely on a
source the IC tells him is not credible?121 Does the IC have a right or duty to
respond to a policymaker’s misstatement, express or implied, of its position? In
the Iraq WMD case, George Tenet, the CIA Director, was later asked by
Congress if he did not feel compelled after the fact to correct misstatements in
the President’s or Vice President’s public speeches about assertions that Iraq
had sought to buy yellow cake for uranium enrichment from an African nation
or that there were significant ties between Saddam Hussein and al Qaeda. Tenet
replied that he did not think the IC had a duty to monitor policymakers’
speeches for accuracy of intelligence but that in fact he had actually (albeit
months later) informed officials of their mistakes. The WMD Commission took
the view that intelligence agents who find out that policymaker actions or
statements are inconsistent with existing intelligence should inform the
policymakers but then let the officials decide what is appropriate to say in
public.122
But in the end, is there any ethical line to which policymakers should be
held in their use of intelligence products? A policymaker can certainly disagree
with the judgment an analyst makes so long as he does not explicitly or
implicitly present it publicly as the official position. But when there is such a
dispute, does he have to say so in public? He can go over the analyst’s head to
her superior in search of a different result from a different analyst. He can
argue and disagree with the analyst and even send her back to look for more
120. Douglas Jehl, Bolton Asserts Independence on Intelligence, N.Y. TIMES, May 12,
2005, at A1.
121. See Douglas Jehl, Report Warned Bush Team About Intelligence Doubts, N.Y.
TIMES, Nov. 6, 2005, at A6 (reporting that a top member of al Qaeda in American custody
was identified as likely fabricator months before his statements were used by Administration
officials to show how Iraq trained al Qaeda members in biological and chemical warfare).
122. WMD REPORT, supra note 2, at 419 (“Finally, analysts and Intelligence
Community leaders have a responsibility to take note, whenever possible, of what their
customers are doing and saying, and to tell those customers when actions or statements are
inconsistent with existing intelligence. We do not mean to suggest that analysts should spend
all of their waking hours monitoring policymakers, or that analysts should have a ‘veto’ over
policymaker statements. Rather, when aware of upcoming speeches or decisions, analysts
should make clear that they are available to vet intelligence-related matters, and analysts
should—when necessary—tell policymakers how their statements diverge from existing
intelligence. Having fulfilled this duty, analysts must then let politically-accountable
policymakers determine whether or not a statement is appropriate in light of intelligence
judgments.”).
WALD FINAL 5/19/2006 9:30 AM
evidence or to pursue additional avenues of research—up to the point, that is,

where repetition becomes intimidation or harassment. He can probably ask for
the raw material from which the analyst works and draw his own conclusions,
if he labels them such, although the interagency clearance process for speeches
may rule that out as a realistic option in most circumstances. But unless the
analyst is incompetent, he can’t ask for her removal or transfer or change the
underlying IC judgment.
Make no mistake: the analyst-policymaker relationship can be a complex
one in which analysts—particularly junior ones—start out at a disadvantage
from a power vantage point. Some kind of instruction, simulation, and role
playing during their training would seem helpful. But the deeper issue is
whether policymakers should be held to some ethical standards as well in their
engagements with analysts and made aware in their often brief tours of duty of
those standards. Right now, any standards are “cultural,” not well understood,
and remedies are distinctly ad hoc. Indiscriminate labeling of intense
interchanges between analysts and policymakers as “politicization” does not
contribute much to the dilemma, but neither is it realistic to attach the “P” label
only if intelligence is actually changed. According to the New York Times,
“the [Bush] administration’s view has been that policy makers do not cross the
line unless they force intelligence analysts to change their conclusions.”123
Similarly, the Senate Intelligence Committee reported that even though the
Administration officials had “pressed analysts to turn up evidence of a
connection between Iraq and Al Qaeda” in roughly the same timeframe as the
WMD inquiries, the analysts stood firm and so there was no breach of
conduct.124
This “outcome” test is not a satisfactory one since it depends on the
analyst’s endurance rather than the policymaker’s ethic. And, as noted, the
analyst’s endurance will often depend on the support she can expect from her
own superiors in the agency. A critical factor for analysts is that they have an
institutional assurance that when dialoguing is done, they will not be forced to
change their analysis at the risk of suffering transfer, removal, or bad
evaluations. An agency which operates completely outside public view needs
high internal levels of monitoring. In this vein, agency ombudsmen should
carefully evaluate transfers, firings, and also voluntary departures under protest
where there are allegations that departures are due to resisting pressures to
conform—inside or outside. Analysts need instruction on how to resist pressure
from outside and inside: who to raise such concerns with, how to record them,
and how to channel their dissents.
Most important, those inside and outside the IC should appreciate the
ambivalence of the analyst-policymaker relationship: a co-dependent one, to be
sure, but one in which the integrity of the intelligence gathering and analysis
123. See Jehl, supra note 108, at 12.

124. Id.
WALD FINAL 5/19/2006 9:30 AM
functions must be scrupulously respected despite its inclusion in a political

branch of the government. The limits on policymakers’ discretion in using the
intelligence process and products urgently need discussion and clarification.
IV. SHOULD THERE BE ADDITIONAL CONTROLS—EXTERNAL AND INTERNAL—

ON THE ANALYST-POLICYMAKER RELATIONSHIP?
The IC, as contrasted with most other Executive Branch agencies, operates
almost entirely out of the public eye—even its budget numbers are secret—and
its covert activities are not even disclosed to Congress at large, but only to the
President and a select few on the House and Senate Intelligence Oversight
Committees. Sanitized versions of the intelligence agencies’ major reports, like
the NIE on pre-war Iraq WMD, are sometimes made public, but these versions
invariably omit classified material which is seen only by relevant
policymakers. Moreover, the accountability mechanisms which govern other
Executive Branch agencies—the Administrative Procedure Act (APA), the
Freedom of Information Act (FOIA), the Federal Advisory Committee Act
(FACA), and even judicial review—usually exempt the IC from their
requirements. The results are no statutory standards and no transparency for the
process by which intelligence reports are made. Furthermore, Supreme Court
jurisprudence provides ballast for Executive Branch arguments that a
deliberative or Presidential privilege protects analyst-policymaker
communications from outside review even by the courts.125 Congressional
oversight committees on intelligence have generally been viewed as
frustratingly lacking the capability and/or the will for sustained or detailed
oversight of major intelligence programs, so it is highly unlikely that they
would undertake serious review of the standards for policymaker-analyst
interactions.126
In sum, it is a cold legal climate in which to propose statutory guidelines
for policymaker-analyst relations. That still does not prevent the IC itself—
especially now that it is under some centralized supervision in the DNI—from
adopting controls or ethical principles for its own people and from using its
influence to safeguard them from exploitation by policymakers.
For instance, so far as I know, no record is now kept of the process or
inputs by which an intelligence report (even one of substantial significance) is
produced. The WMD Commission recommended that a report reference all
prior reports or sources relied upon but not what outside contacts were made
with the drafters and by whom. Obviously, such a record would be an internal
document but could prove useful in discouraging untoward or excessive
overtures from the policymaker as well as irresponsible complaints from the
125. See, e.g., Cheney v. U.S. Dist. Court, 542 U.S. 367 (2004); In re Cheney, 406
F.3d 723 (D.C. Cir. 2005) (en banc).
126. WMD REPORT, supra note 2, at 337-40.
WALD FINAL 5/19/2006 9:30 AM
analyst. I repeat, as well, my prior suggestions that both analysts and newly-
appointed policymakers be educated initially about responsibility and ethical
line-drawing in their relationships, and that analysts are provided the proper
channels in which to complain if they believe that limits have been
transcended.
V. CONCLUSION
The WMD Commission recognized the centrality of the analysts’ role in

the reform of the intelligence process. It produced a great many
recommendations that should enhance the recruitment, training, and working
conditions of analysts, including rousing support for their freedom to dissent
responsibly without fear of job diminution. The Commission stressed that
analysts should be candid with policymakers about what they know, what they
don’t know, what they think, and on what evidence their opinion is based, as
well as the level of certainty with which they state facts or findings. The
recently installed NDI and his deputies have said they agree with and will
follow the Commission recommendations, providing ample opportunity for
alternative hypotheses and points of view. They accept the internal costs of
such a policy about face—that policymakers will have to operate with “a higher
tolerance for ambiguity.”127
Some may indeed speculate what the effect of such a tolerance policy will
be on decision-making, whether it will be slowed down as analysts wait for
more accurate evidence or pepper their products with hedges and ambivalent
conclusions. I believe the Commission’s view—certainly my own—is that
policymakers are constantly confronted with the need to make critical decisions
on imperfect knowledge, and they are better served if they know which
intelligence judgments are solidly grounded and which are reasonably informed
guesses. We see little gain in confident but mistaken judgments such as those
made about Iraq WMD in the NIE and other IC products. In the final analysis,
policymakers cannot delegate hard decisions that simply lack the desirable
amount of fact foundation.
The Commission dealt only briefly with the problem of “politicization” of
the intelligence process through pressure on analysts to come to a preordained
conclusion. It found no such outside pressure to control the Iraq NIE analysis
although it acknowledged that analysts were affected by a gathering climate of
impending war and by an internal atmosphere that discouraged and discounted
views contrary to the community’s assumption that because Saddam Hussein
had once had and used WMD, he would reconstitute those programs as soon as
inspectors left. Left unexplored were the deeper questions of whether there
were or should be any specific policies or controls over how much
policymakers can push their own views with analysts, when repeated demands
127. Jehl, supra note 43; Reynolds, supra note 97.

WALD FINAL 5/19/2006 9:30 AM
for more information or different modes of interpretation go over the line into
harassment, and, if that point is reached, what recourse the analyst has.
Conversely, the need for internal controls on the degree of pressure that the
policymaker can bring to the “engagement” was left unanswered. The
assertions and counter-assertions that characterized the Bolton nomination
raised these questions starkly, but there seemed to be consensus only on the
issue that a policymaker goes over the line if he seeks to transfer or remove an
analyst because he does not like the substantive judgments of the analyst.
I suggest that the analyst’s position might be made more secure by some
modest mechanisms drawn from administrative law, such as internal logs of all
contacts made with the analyst during the engagement, ombudsman focus on
the reasons for transfers, resignations and removals, an express recognition that
the context in which intelligence reports are made may mandate higher
thresholds of certainty when the predictable effect of the reports may be war or
other national crises, and, finally, training of analysts in holding their own
against politically powerful policymakers. These are modest suggestions, but
they could strengthen the analyst’s position and, in that regard, make her better
equipped to provide the policymaker with the highest level work of which she
is capable. Minimally, I believe the subject of policymaker-analyst relationship
deserves greater attention and if possible more disciplined guidance than the IC
has provided so far.
We appear to be headed into a period when national crises will be ever
more frequent and our dependence on the best intelligence we can get ever
more necessary not only to crucial decision making but also to our credibility
in the rest of the world. We do not need more “cook your own books”
accusations. Giving the analyst greater protection in this relationship will cost
little and profit us much.128
128. In March 2006, while this essay was in galley, Paul R. Pillar, now retired, who
served as CIA National Intelligence Officer for the Near East and senior analyst for the
Middle East published an article, Intelligence, Policy, and the War in Iraq, calling the
relationship between the intelligence community and the policymaking process “broken and
badly [in need of] repair.” He alleged that, in the case of Iraq, policymakers ignored the
import of intelligence reports which was “to avoid war.” “The policymakers,” Pillar said,
used the intelligence not to inform their decisions, but to justify decisions already made; they
“cherrypicked” data from reports and subtly, though not overtly, through insistent demands
for more analysis, pressured analysts into producing the policymakers’ desired conclusions.
He pointed out, as the WMD Commission had, that pro-WMD pieces were more hospitably
received by superiors than skeptical ones. While finding no single clear fix to the problem,
he recommends as desirable a forthright declaration by the government that intelligence
must be insulated from policy, oversight by a nonpartisan office like the GAO for
monitoring the relationship, and the restructuring, of the IC as a semiautonomous body
overseen by governors with fixed terms. According to SSCI Chairman Pat Roberts, Pillar did
not make similar criticisms when interviewed by the committee. See Paul R. Pillar,
Intelligence, Policy, and the War in Iraq, FOREIGN AFFAIRS, March/April 2006, at 25.
WALD FINAL 5/19/2006 9:30 AM

Wald

Uploaded by

Copyright:

Available Formats

Wald

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Wald

Uploaded by

Copyright:

Available Formats

WALD FINAL 5/19/2006 9:30 AM

ANALYSTS AND POLICYMAKERS: A

242 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

critiques by the Senate Select Committee on Intelligence (SSCI)1 and the

1. STAFF OF S. SELECT COMM. ON INTELLIGENCE, 108TH CONG., REPORT ON THE U.S.

2006] ANALYSTS & POLICYMAKERS 243

different thresholds of certainty and reliability for these findings dependent on

I. WHAT THE WMD COMMISSION HAD TO SAY ABOUT ANALYSTS

The WMD Commission’s mandate was to “assess whether the intelligence

244 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

intelligence units’ functions, and to spur the much-needed innovations

11. WMD REPORT, supra note 2, at 45.

2006] ANALYSTS & POLICYMAKERS 245

In the biological warfare area, the WMD Commission’s assessment of the

16. Id. at 90-93, 94-105.

246 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

mistaken.19 The inclusion of the software in orders of other neutral equipment

19. Id. at 132-46.

2006] ANALYSTS & POLICYMAKERS 247

25. Id. at 181-86.

248 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

26. Id. at 19, 25, 317-19.

2006] ANALYSTS & POLICYMAKERS 249

29. WMD REPORT, supra note 2, at 20, 321-26.

250 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

32. Id. at 387, 389, 407-09, 411-14, 424-25.

2006] ANALYSTS & POLICYMAKERS 251

II. WHAT HAS HAPPENED SINCE?

I have now described the Commission’s principal findings and

33. Id. at 30-31.

252 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

The Washington Post called it a reform “potentially profound and warrant[ing]

July 4, 2005. at A16; Johnston, supra note 35.

2006] ANALYSTS & POLICYMAKERS 253

254 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

been established;56 and a National Counter Proliferation Center will act as a

2006] ANALYSTS & POLICYMAKERS 255

I attribute much of our success in getting the report implemented to the

256 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

slower than hoped,” a member of the House Intelligence Committee

III. THE ANALYST: A (WO)MAN IN THE MIDDLE

So far, I have dealt with the Commission’s critique of and

68. Gorman, supra note 67.

2006] ANALYSTS & POLICYMAKERS 257

258 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

A. Who Does the Analyst Work For?

2006] ANALYSTS & POLICYMAKERS 259

260 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

on impending executive and congressional action with respect to Iraq, as well

84. WMD REPORT, supra note 2, at 190.

2006] ANALYSTS & POLICYMAKERS 261

intelligence analysis which admittedly has to patch together pieces of disparate

86. WMD REPORT, supra note 2, at 190.

262 STANFORD LAW AND POLICY REVIEW [Vol. 17:241

an admittedly narrow definition of “politicization”—whether analytical

88. Id. at 188 (footnote omitted).

2006] ANALYSTS & POLICYMAKERS 263

95. Id. at 248 n.847.

264 STANFORD LAW AND POLICY REVIEW [Vol. 17:241