Introduction
In this essay I will look at the different factors that
determine intelligence failure in general to see which of these were present in
the US intelligence’s failure to find large-scale weapons of mass destruction
(WMDs) in Iraq. To do this, intelligence in general will be described and the
intelligence cycle will be explained and utilised to analyse the different
stages of where intelligence can fail. However, using the intelligence cycle to
aid the understanding of intelligence failures is solely a method of
simplification. It does not imply that failures cannot be continuous over
several steps of the intelligence cycle. I will explore below, especially when
looking at failures in the Iraq invasion, how the practical distinction between
failures at separate the stages is rather blurry. Failures were made in
collection, analysis, dissemination and communication, to name a few. Despite
the findings of the prominent investigative reports after the US invasion of
Iraq in 2003, I believe that politicization of the intelligence community (IC)
on this matter was a prominent factor in obscuring the intelligence findings
and ultimately led to the intelligence failure of the Iraq War 2003. I will
explain why in the following paragraphs.
Intelligence and
Intelligence Failure
Intelligence is data and knowledge collected from a range of
different sources. In a modern political context, it is “the official, secret
collection and processing of information on foreign countries to aid in
formulating and implementing foreign policy, and the conduct of covert
activities abroad to facilitate the implementation of foreign policy” (Random,
1958). Intelligence is often collected to acquire knowledge of other countries’
military strength, economic power, internal political situation and levels of
domestic unrest. The primary job of intelligence is to reduce uncertainty by
identifying issues of policy relevance with policy makers, collecting and
analysing information (Gentry, 2008:267), and issuing policy makers with timely
advance warning of potential threats (Pythian, 2006:401). The intelligence is
used to estimate threats, capabilities and intentions of adversaries (Betts,
1978:68). Political leaders analyse these warnings in the relevant strategic
and domestic political context, make decisions under conflicting pressures and
manage policy-implementing agencies (Gentry, 2008:267). In this way,
intelligence is used to guide policy makers and other senior decision makers in
the national security and defence arenas (CIA, 2007).
Intelligence failure can be at blame if events “of strategic
significance” occur without forewarning (Pythian, 2006:401). Intelligence
failure can occur when a state fails in collecting or analysing information,
national leaders fail to make sound policy on the disseminated intelligence or
fails to act effectively on the information received (Gentry, 2008:249). Gentry
(2008:248) argues that reasons for intelligence failure can include
organizational, cultural, cognitive or psychological factors. The intelligence
cycle includes the direction from policy makers, the “requirement” of what
should be investigated, the collection and analysis of intelligence, the
finished product of intelligence analysis that is delivered to the
policymakers, and again the directions from policy-makers on implementation and
further targets of analysis. According to Betts (1978:63) crucial mistakes are
sometimes made by the professionals that produce the finished analysis, but
most often by the decision makers who consume the products of intelligence
services. It is impossible to account for all the possible mistakes that can be
made in attempting to take advantage of intelligence, but the following
paragraphs will give examples of how possible failures can erupt on all stages
of the intelligence cycle and how it is all linked.
Failure
Determinants at Different Stages
The directions given by the policymakers in relation to the
target of the intelligence-gathering are relevant because different targets
require different forms of intelligence. For example, human intelligence
(HUMINT) can be a beneficial way of collecting information if the collecting
country has an official presence in the country. Additionally, targets of
intelligence, usually states or non-state actors, often know that they are targets
of intelligence and may act to hinder foreign intelligence efforts through
active defences like counter-intelligence activities, deception and concealing
sensitive installations (Gentry, 2008:255), which can lead to misleading
information and wrongful analysis. One writer who focuses on the act of
deception by intelligence targets is Barton Whaley (1973:2). He claims that it
was Hitler’s deceptions, rather than mistakes made by the Soviets, that made
the German invasion of the Soviet Union in June 1941 so surprising.
Additionally, collection of intelligence is always incomplete (Dahl, 2005:37),
and different targets require different forms of collection. Analysts have
concluded that the rising global threat of terrorism, for example, poses new
and difficult challenges for the IC, which can best be overcome by more HUMINT
(Dahl, 2005:33).
Failures in the intelligence cycle can also happen at the
stage of analysis, and when information is processed by the intelligence
agencies. Analysts try to identify significant information, distinguish noise
from relevant signals, and make accurate forecasts (Gentry, 2008:252). Errors
at this level can occur in misunderstanding of information and recognition of
“noise” instead of relevant signals. Additionally, errors can happen through
cognitive or institutional bias by the analyst. There might also be a problem
of compartmentalization of information, because of restrictions on information
sharing, security concerns and fears of compromising sources that creates mistrust
between intelligence agencies (Hulnick, 2006:962). By sharing intelligence,
gaps in information can be filled, and conflicting analysis can be found. The
2004 Madrid bombings underlined how compartmentalisation of intelligence can
impede security operations. Some of the suspects in the Madrid bombing had been
known to the French and Spanish police in 2001, and had, since 2003, been on a
list of suspects issued by the Moroccan police for a series of café bombings in
Casablanca, but were still living openly in Madrid. Several countries seemed to
have a few pieces of intelligence each (Finn and Richburg, 2004), which, if
shared, could have led to a prevention of the attack.
Other major sources of intelligence failure can be the
method of dissemination; the communication to the policy-makers, and possibly
also the access to raw material outside the intelligence agencies. For
communication to be effective, analysts must present clear, accurate and
persuasive warnings. A former Defence Intelligence Officer claimed that
analysts cannot only give reports stating that a bombing might happen, they
must tell the policy-makers what this means and what is really likely to happen
(Dahl, 2005:47). Both insufficient information and an overload of information
can lead to unclear messages being communicated. Where there is ambiguous
information and limited time for thorough assessment of sources, intelligence
analysis can be driven by intuition and conclusions can be led by instinct. As
stated by Betts (1978:71): the greater the ambiguity, the greater the impact of
preconceptions. Another source of challenge in the intersection between
analysts and policy-makers in the US, is that raw reporting from the collection
process usually reaches both at the same time (Hulnick, 2006:961). Some of this
raw intelligence may be incomplete, contradictory or wrong, and policy
officials sometimes take the reporting as having been judged or evaluated
(Hulnick, 2006:962). This can create serious problems on both sides.
In the final stage of the intelligence cycle, intelligence
is communicated to policymakers, who must then react appropriately and
implement policies accordingly. The task of the policy-maker is to place the
intelligence warnings in political contexts and make decisions based on their
political circumstances and limited tools for implementation. Factors such as
psychology, experience and interests can increase cognitive bias of a
policy-maker in interpreting warnings and may lead decisions to become
defective policies (Genry, 2008:254). It was noted in relation to the Beirut
bombings in 1983 that failures in collection and processing of information were
much less significant than the attitudes and lack of action by the consumers of
intelligence (Dahl, 2005:46). Directors of intelligence agencies can tend to
dismiss critical intelligence, and cling to data that supports continued
commitment to established policies (Betts, 1978:65). Additionally,
policy-makers might not have trust in the intelligence personnel, and may
therefore refrain from acting on certain warnings. According to Gentry
(2008:256) presidents Truman, Johnson, Nixon and Bill Clinton sometimes
displayed suspicion of the CIA and at times ignored intelligence. One Director
of Central Intelligence (DCI) reportedly quashed a CIA report warning of the
dangers of invading Cambodia in 1971, because President Nixon had already
decided to invade (Ibid, 2008:252).
Additionally, policy-makers’ bias can occur as a source of
intelligence failure if politicians create an environment of presumed facts,
and indirectly force intelligence agencies to operate in concert with those
presumed facts. Or failure can arise if intelligence professionals attempt to
increase personal or institutional favour by providing specific intelligence messages
that they imagine are wanted or necessary for the leaders (Gentry, 2008:252).
It is also said that “the principal cause of surprise is not the failure of
intelligence, but the unwillingness of political leaders to believe
intelligence or to react to it with sufficient dispatch” (
Wohlsetter, 1962:227)
. Desires to prevent recrimination may drive intelligence
agencies to withhold warnings until uncertainties recede, and therefore keep
raw information to themselves (Gentry, 2008:253). The allocation of time and
resources for intelligence professionals provide additional constraints (Betts,
1978:68). Additionally, agencies responsible for the implementation of policies
are subject to certain limitations that impact their ability to respond to
intelligence warnings. This way, failures that appear on the implementation
stage may superficially look like intelligence failures, but actually reflect
structural issues created through previous policy decisions (Gentry, 2008:256).
The Intelligence
Failures of the 2003 Iraq Invasion
In 2003 the US invaded Iraq to eliminate the perceived
threat posed by Saddam Hussein’s possession of weapons of mass destruction
(WMDs) (Pythian, 2006:400). The invasion has later been named ‘the perfect
intelligence failure’, as failure occurred on all stages of the intelligence
cycle, and no WMDs were found (Hulnick, 2006:967). Some see it as the worst
intelligence failure since the founding of the modern intelligence community
(IC). The
National Intelligence Estimate
(
NIE
) that ‘justified’ the
intervention in Iraq was based on reporting from unreliable sources and biased
preconceptions grounded in the previous experiences of WMD programs in Iraq
(Hulnick, 2006:967). Subsequent investigations into the failed intervention
claimed that the failure was in collection and analysis, poor management and
organizational weaknesses (Pythian, 2006:401).
Since the US did not have an official presence in Iraq,
collection of intelligence relied on a few HUMINT sources that were dependent
on “defectors and foreign government services” (US Congress, 2004:24). One of
these was ‘Curveball’, a chemical engineer from Baghdad (Betts, 2007:602), who
provided the bases of around 112 separate reports (US Congress, 2004:149), and
subsequently was deemed to be a fabricator (Ibid, 2004:462). Analysts further
failed to investigate dual-use equipment bought through illicit channels (Ibid,
2004:14), and assumed that these were for the development of a WMD program, and
not actually for the legal tactical rockets (Pythian, 2006:408). The
Silberman-Robb commission (2005:52) claimed that this constituted “errors in
technical and factual analysis”. A compartmentalization of information made the
DCI unaware of dissenting opinions within the IC (US Congress, 2005:28). Additionally,
the CIA tended to deny information to more specialist agencies that could have
provided input that challenged existing assumptions (Ibid, 2004:28). These,
along with similar findings, led the Silberman-Robb Commission (2005:5) to
describe the IC as “fragmented, loosely managed and poorly coordinated”.
Politicization
on All Levels
Even though the subsequent investigations rejected
politicization, and absolved the Bush administration of all charges (Pythian,
2006:401), some reports gave hints of an environment that was not conducive to
questioning the dominant assumption on Iraq, which strengthens Betts’
hypothesis that intelligence failure usually lies with the consumers of
intelligence (Pythian, 2006:418). Senator Ron Widen stated that the Bush administration
had “repeatedly and independently made the case for war not by relying on US
intelligence but by ignoring and directly contradicting the same” (US Congress,
2004:489-490). Senior Bush administration officials had made forceful public
statements for war (Silberman-Robb, 2005:189) and prevalence of repetitive
tasking of intelligence personnel was found, including questions of judgments
on a particular issue over and over again by “senior customers” (US Congress,
2004:456). According to Hulnick (2006:967) some also speculate that that the
NIE was drawn up in order to meet the political needs of the White House.
Richard Clarke (2004:264) claimed that the Bush administration entered office
“with Iraq on its agenda”, and a leaked minute from a meeting record shows the
head of MI6, Sir Richard Dearlove, reporting to the prime minister that
“military action was now seen as inevitable” and that “intelligence and facts
were being fixed around the policy” (Danner, 2005).
Deception practices employed by Iraq in the past, and the
failure to account satisfactorily for WMDs during the 1991 war, gave US
policymakers and intelligence professionals logical reasons for why there was
evidence denying the existence of WMDs (Pythian, 2006:408). In a context of
seemingly obvious guilt, an environment of pressure to find WMDs and a
political desire to intervene in Iraq, analysts shifted the burden of proof
from requiring evidence of WMDs, to require evidence showing that Iraq did not possess
WMDs, a theory that could not be disproved (Silberman-Robb, 2005:168). The NIE
was written with the assumption that the US was going to war (US Congress,
2004:505), and intervention could only take place if the IC concluded that Iraq
still had illegal WMD programs. This justification was the only way to secure
public support for the Bush administration to initiate a war (Betts, 2007:598).
This sentiment caused findings of evidence that denied WMDs to be downplayed
and ignored. Even when Saddam Hussein’s son-in-law, Hussein Kamel, told his
debriefers that old stocks of WMD had been destroyed, this was not believed
(Jervis, 2006:40). Thin amounts of evidence gave room for preconceptions and
the ability to shape verifications according to these. Influenced by decisive
policymakers, judgments on all levels were driven by circumstantial evidence
(Betts, 2007:602) and the necessity of finding WMDs to justify invasion of
Iraq.
Cirincione et al (2004:50) confirm my opinion, and found it
unlikely that the behaviour at policymaker level did not create an environment
of pressure to reach a conclusion confirming WMDs in Iraq. Along with the
increased publicity of the immediate threat of Iraq’s WMD, the intelligence
judgments became more absolute and supportive of the administration’s case
(Pythian, 2006:417). IC managers failed to utilize mechanisms in order to
challenge the prevailing conclusions (US Congress, 2004:23), and concerns
regarding credibility of sources were not conveyed to policy makers. The Senate
Select Committee explained this behaviour as “groupthink” (US Congress,
2004:18), referring to a desire for unanimity overriding a realistic appraisal
of alternative courses of action (Janis, 1983:9), recognized in the selective
use of information and collective rationalization of WMD intelligence in Iraq
(US Congress, 2004:18). The conclusion reached by the analysts went beyond
anything that could safely be reached on the basis of available intelligence
(US Congress, 2004:14), and the Senate Select Committee found that the
NIE
“did not accurately portray the
uncertainty of the information”, but formulated assumptions and theories as if
they were facts (US Congress, 2004:17). According to Betts (2007:605) the Key
Judgments of the summary conveyed a message that the conclusions “derived from
observed activities as much as preconceptions” and assumed intentions. This
shows how politicisation possibly influenced the IC to such an extent as to
form the intelligence according to preconceptions and desires of the Bush
administration.
Conclusion
In many ways, the failure to find WMDs in Iraq could be
described as the perfect intelligence failure, as failures can be found to have
occurred at all stages of the intelligence cycle. It ranged from collection of
intelligence, analysis, dissemination, within management, in the structure of
the IC, and at the level of policymakers (Pythian, 2006:419). However, as has
been shown in the previous paragraphs, I believe that politicisation ultimately
led to obscure intelligence findings, which consequently led to the
intelligence failure of the Iraq War. As the collection of intelligence is
carried out by humans, ensuring neutrality is of wide importance, and therefore
strategies of ensuring this should be employed. In the case of the 2003 Iraq
invasion, neutrality was far from ensured. Policymakers and senior officials
created a bias towards the existence of WMDs in Iraq, and successfully
implemented this bias at all stages of the intelligence cycle, in order to
enable a conclusion justifying an intervention in Iraq. Given the secretive
nature of intelligence, and the minimal political constituency of intelligence
agencies, presidents can blame intelligence for their own errors in policy and
its execution, and I believe this is in fact what happened in the 2003 Iraq
invasion. Intelligence was twisted by politicisation and pressure from
government officials, which ultimately culminated in a failed invasion and the
onset of war.