Over the course of the last century, failures in intelligence have happened regularly, sometimes at significant political loss, and occasionally, resulting in devastating loss of life such as the terrorist attacks of 11 September 2001 (9/11).  Intelligence failures have been well documented throughout the history of the United States Intelligence Community (IC), and in the case of an earlier intelligence failure at Pearl Harbor, the failure was even the catalyst that brought about the changes that formed the modern IC.   What has not been as well documented or even understood is why these intelligence failures persist after reforms and changes within the IC, almost as if after each failure, reforms are made and then new weaknesses emerge, like playing whack-a-mole.  While the history of these failures is well documented, there is little in the way of a systematic theory to explain why these intelligence failures persist.

Richard K. Betts worked toward developing a theory of intelligence failure in his article “Analysis, War, and Decision: Why Intelligence Failures are Inevitable”, which established the paradoxes within intelligence that prevent the elimination of failure.[1]  Betts argues that failures in intelligence are not only inevitable, but even natural because the failure is primarily a result of politics and psychology, rather than analysis and organization.[2]  Betts’ article was written in the wake of the intelligence failures of the Vietnam War, and the reforms of the IC after the Church Committee inquiries.  While Betts developed a theory of the causes of failure, the areas of error, and the imperfect solutions, the discussion of his theory is situated in a particular time and space.  His theory is sound and should withstand application to the challenges of the modern IC in the 21^st century.

The aim of this essay is to utilize Betts’ theory of intelligence failure to look at the failure on the part of the IC that resulted in the tragic attacks of 9/11.  First it will examine the core principles of Betts’ theory of intelligence failure.  Part two will look at the IC in the period prior to 9/11 and pinpoint where the system was broken, which allowed the failure to occur.  Part three will look at the reforms enacted as a result of the findings of the 9/11 Commission Report and the Intelligence Reform and Terrorism Prevention Act of 2004 (IRTPA) to see if the changes are in accordance with Betts’ theory and recommendations.  The essay will finally examine the state of the current IC and determine if intelligence failures are still inevitable as Betts’ theory predicts they will continue to be.

Betts’ Theory of Intelligence Failure

Betts places intelligence failure squarely on the shoulder of the consumer, arguing that, “The most crucial mistakes have seldom been made by collectors of raw information, occasionally by professionals who produce finished analyses, but most often by the decision makers who consume the products of intelligence services.”[3]  While he does allow room for improvement on the part of the intelligence institutions, Betts argues that failure is political and psychological, rather than the fault of flawed organization and institutions.  Improvements can be made in the analytic process, but as Betts believes the flaws are political and psychological, it is in these areas he is most critical, and offers a framework for understanding where failures occur and solutions to address them, though he readily admits they are imperfect solutions.  Betts recognizes that an empirical study of intelligence failures is well developed, but because lessons are not always learned from mistakes, he tries to create a normative theory to understand why failure occurs and how it can be prevented.[4]

Intelligence fails in three ways, and of these three causes of failure, Betts describes failure in perspective as the most reassuring.  Whereas the nature of success is preventing a quantifiable disaster, a failure is acute and measurable.  Because failure is distinct and tangible, we perceive what we can quantify.  Our failure in perspective is our inability to measure the ratio of intelligence success against failure.  Betts describes this cause of failure as reassuring because there is an implicit ignorance of the picture as a whole; we lose sight of the many daily successes and instead focus on the lone failure.[5]  Betts identifies pathologies in communication as his second cause of failure, which he also considers most common.  Due to gaps in the procedures of transmitting intelligence analysis to policy makers, as well as the struggle to convince policy makers of the importance of the particular intelligence at hand, the structure of the IC is the most often attributed source of failure.  Because these challenges in institutional communication are the most identifiable cause of intelligence failure, the organization and management of the IC is the most targeted area of reform after an intelligence failure.[6]  The third and final cause of intelligence failure according to Betts is the paradox of perception.  Betts regards this source of failure as the most crucial, and most difficult to address.  The paradox of perception is that preconceived notions held by either analysts or policy makers impede them from accurately assessing a situation.  Even if the intelligence process can be perfected, the consumption of the intelligence will never be perfected due to psychological limitations and influencing political factors.[7]

Betts believes it is important to locate the broken links between analysis and decision makers.  He identifies three areas where the paradoxes and pathologies of intelligence are most pervasive: in attack warning, operational evaluation and defense planning.  The challenge in providing attack warning is the inherent challenge of discerning and predicting enemy intentions in a timely manner.  In the major examples of surprise attack, such as Pearl Harbor, the evidence of the intended attack existed, but it did not pass quickly enough through the levels of management to reach policy makers in time to act to prevent the disastrous attack. An additional characteristic of the available intelligence prior to a major attack is the inability of policy makers to accept the validity and urgency of the available intelligence because it was inconsistent with the prevailing estimates or assumptions.  Betts identifies additional impediments in the disbelief of policy makers in the provided intelligence as well as hesitancy in communication within the IC as contributing conditions leading to unsuccessful attack warnings.[8]  In examining the cause of disconnect between policy makers and analysts in utilizing intelligence for operational evaluation, Betts draws much of his insight from practices during the Vietnam War.  He found that when evaluating and assessing the success of tactical operations while they were occurring, operational intelligence agencies, such as the military agencies took a more optimistic view of the available intelligence to support the organization’s activities during the conflict.  He found that non-operational intelligence agencies, such as the CIA, were more pessimistic in their assessment of the available intelligence.  Additionally, when provided differing assessments of a situation, policy makers are inclined to seek optimistic intelligence that supports the policy that they promote in the given situation.[9]  The final area where intelligence breaks down and failure occurs is the arena of defense planning.  Like attack warning, the intelligence regarding the capabilities and intentions of the enemy are nebulous and difficult to estimate or predict.  The use of intelligence in the capacity of defense planning can allow decision makers to use the intelligence to justify favored policies or preferred future operations or missions.  Budgetary priorities can complicate organizational assessments of future threats, and the relationship between intelligence and policy can become murky and without clear distinction.[10]

Betts finds the greatest threat to successful utilization of intelligence in “[W]ishful thinking, cavalier disregard of professional analysts, and, above all, the premises and preconceptions of policy makers.”[11]  The improper use of intelligence by decision makers, generally stemming from their own preconceptions and lack of objectivity are in his opinion the cause of intelligence failure.  Yet Betts recognizes that it is difficult to distinguish between actual intelligence failure and policy failure.  In order to do so, he draws attention to the tangible barriers that prevent analytic accuracy.

The three obstacles to successful analysis identified by Betts are the ambiguity of evidence, ambivalence of judgment, and the atrophy of reforms.  Ambiguity of evidence can include excessive, or conflicting as well as limited or inadequate intelligence.  The danger for analysts can be in oversimplifying the ambiguities or minimizing their significance, which can in turn allow the policy maker to misunderstand the uncertainty within the intelligence at hand.  When the ambiguities are so pervasive, the natural tendency of both the analyst and policy maker is reliance on instinct and intuition to compensate for what is unknown or unclear.[12]  The second obstacle in analysis is the tendency to err on the side of caution, which Betts expresses as ambivalence of judgment.  When presented with inconsistency or ambiguity, analysts can be in danger of providing assessments that are too cautious, and in not making a definitive assessment, nothing new is said and the value of the intelligence is diminished.  By not making a definitive assessment, policy makers are again given the opportunity to use an ambivalent assessment or portions of a conflicting or vague assessment to support a particular policy as it suits them.[13]  The final barrier to analytic accuracy identified by Betts is the atrophy of reforms implemented in the wake of an intelligence failure.  Many reforms are costly, both in time, effort, and financial expense, and unless they provide practical, tangible benefit to the intelligence consumer, changes in intelligence analysis and consumption will quickly erode.[14]

Betts recognizes that the solutions to these obstacles in accurate analysis are inadequate and that in most cases, the solution to one obstacle creates new vulnerabilities in other areas, creating the inescapable dilemma of the whack-a-mole approach to eliminating intelligence failure.  One of the most common solutions in the wake of intelligence failure is to assume the worst.  By treating every threat as serious and genuine, all threats become equal and sensitivity to actual threats decreases.  This solution is generally not appropriate and is counterproductive and expensive.[15]  Another solution that is common after an intelligence failure is the desire to consolidate and streamline the IC to simplify and coordinate analysis.  Betts recognizes that interagency competition and bureaucracy have a nagging tendency to reoccur, and that decreasing or consolidating personnel and resources creates new problems.[16]  Other reforms that are often implemented include multiple advocacy and devil’s advocacy.  The former is the intentional utilization of multiple or dissenting perspectives and the later is the intentional use of opposing perspectives in analysis.  While each method can provide new insight or valid opinions, it can also increase ambiguity, allowing decision makers to choose the perspective or opinion that suits them or their policies best.[17]  The purposeful creation of dissenting opinions can also lend validity to perspectives that are not serious and should not be considered, or conversely, are dismissed because the opposing argument was created out of requirement rather than legitimate concern, and it can cause decrease in sensitivity to alternative arguments with actual merit.[18]

Perhaps the most interesting of Betts’ solutions is his attention to the psychologies of policy makers.  Since he considers most failures to occur as a result of decision makers, he argues that these decision makers need to understand their own psychologies, and be aware of their own attitudes and modes of perception as well as the differing methods and perceptions of the intelligence producers in order to understand the biases in the intelligence products that they recieve.  Betts recognizes that cognition is impractical to measure or legislate, and that preconceptions and paradigms are difficult to alter, and most policy makers simply do not have the time or interest for serious reflection.  Even though these policy makers are more often than not the cause of intelligence failure it is impossible to expect the consumer to be aware of and correct for their central role in the cause of intelligence failures.  Betts suggests that the most ideal solution would be a Director of Central Intelligence (DCI) that was both an intellectual and professional analyst as well as a talented politician who could thoroughly understand the product provided to the president, and be able to skillfully advocate for unusual or different analysis, as well as illustrate why different intelligence agencies may reach different conclusions.  Naturally this scenario would require a president and DCI with an interest in this more time-consuming form of intelligence briefing, as well as the personal relationship required for this unorthodox approach to intelligence consumption.[19]

In light of the challenges to analytic accuracy and the very human flaws of each new policy maker using intelligence, the imperfect solutions Betts provides offer opportunity for minor improvements that allow for the reduction, but not elimination of intelligence failures.  Betts admits that mistakes will still occur and that, “Although marginal reforms may reduce the probability of error, the unresolvable paradoxes and barriers to analytic and decisional accuracy will make some incidence of failure inevitable.”[20]  Betts concludes that tolerance for disaster is all that we are left with since intelligence failures are not only inevitable, but also even natural.[21]

Pinpointing Responsibility for the Failure of 9/11

The terrorist attacks of 11 September 2001 in New York, Washington, and Pennsylvania were the most devastating attacks in living memory.  The surprise nature of the attacks and the seeming lack of warning have led most Americans to believe it was the greatest intelligence failure of our time.  While many in the public may hold this opinion, it is not entirely accurate to describe the attacks as a failure of intelligence, but more appropriately a failure in policy that, unfortunately, was probably inevitable.

Betts identified three causes for intelligence failure: failure in perspective, pathologies in communication and the paradox of perception.  Failure in perspective is the case for many who felt the IC was asleep at the wheel and that the attacks of 9/11 came out of nowhere.  In fact, the IC has vigilantly protected the US from terrorist attacks after the end of the Cold War, thwarting threats such as the Manila air plot and millennium terrorist plots.[22]  It was because of the devastating scale of 9/11 that we saw it as an intelligence failure distinct from the many daily success of the IC both before and since the attack.  The 9/11 Commission Report identifies communication as perhaps the most easily remediable failure of 9/11.[23]  The compartmentalization within the CIA, competition and rivalry between the CIA and the FBI, and legal barriers to shared knowledge did inhibit effective communication between agencies, and better coordination and communication within the IC is a legitimate area for reform.  However, it was not poor communication that caused the failure that day, since the 9/11 Commission Report states emphatically in multiple instances that the both Presidents Clinton and Bush, as well as their top policy advisors, were very much aware of the threat posed by bin Ladin.[24]  Because the top decision makers of both administrations “got the picture”, this implies that the true failure occurred because of paradoxes of perception.

The relationship between intelligence analysis and policy making in intelligence failures was explored in the interesting recent article by Stephen Marrin, “The 9/11 Terrorist Attacks: A Failure of Policy Not Strategic Intelligence Analysis”.[25]  The article seeks to debunk the assumption that intelligence analysis influences decision-making and serves as the foundation for policy making.  Marrin argues that, “The reason decision makers frequently ignore intelligence analysis is because intelligence analysis is never the sole foundation for decisions, but rather duplicates the analysis and assessment conducted by the decision-makers themselves who are usually well aware of both the details and strategic importance of the issues under consideration.”[26]  The statements of DCI George Tenet corroborate the 9/11 Commission report that emphasized that both Bush and Clinton and their top national security advisors understood the gravity of the threat posed by bin Ladin, yet Marrin believes that the 9/11 Commission Report absolves these leaders of their responsibility by suggesting that because of their meager and slow moving policy in response to the threat, they could not possibly have understood the nature of the threat posed.[27]  The 9/11 Commission report lists a series of classified reports issued between 1998 and 2001, many of which were included in the Presidential Daily Brief (PDB), which clearly stated the known intentions and ambitions of bin Ladin, including evidence of possible attacks targeting New York, Washington and airports, as well as the use of airlines, hijackings and bombs.[28]  The intelligence analysis provided to decision makers between 1998 and 2001 quite clearly stated that the US faced a serious threat, and potentially imminent attack, but the CIA struggled to secure additional resources from the White House and Congress to counter the threat.[29]  The Clinton White House was described by author Steve Coll as the, “[M]ost sensitive to the political consequences of both terrorism and failed covert action, rang loud alarm bells about the threats but also proved cautious about operations that might go bad.”[30]  In light of the knowledge of the enormity of the threat, new policy towards eliminating the threat of al Qaeda and bin Ladin was slow moving and tenuous, and policy changes to remove the source of Islamist terrorism and terrorist grievances, such as US policy in the Middle East, particularly Saudi Arabia and Israel, or providing economic development or education, do not appear to even have been considered.[31]  Both Marrin and the 9/11 Commission Report agree that the embassy bombings of 1998 were the opportunity for policy change, and Marin describes the missed opportunity as a policy failure, and the 9/11 Commission Report states that the, “[M]ajor policy agencies of the government did not meet the threat.”[32] By their statements to the 9/11 Commission Report, top policy makers were aware of the magnitude of the threat, and recognized that their own policy was not effective to guard against the attack.  Richard Clarke in particular, as the “[N]ational security decision-maker most responsible for coordinating counterterrorist policy did not believe that current policies were adequate as a way of addressing the threat.”[33]  Clinton’s administration struggled to confront the threat from bin Ladin in an impossible political climate of impeachment proceedings, partisan warfare, bombing Iraq and the memory of disaster in Somalia, and instead opted for covert action that had limited prospects for success.

In his theory of intelligence failure, Betts’ highlighted the three areas where the system breaks down between intelligence analysis and policy makers as attack warning, operational evaluation and defense planning.  In the case of 9/11, there were barriers between what the CIA knew and tried to express to policy makers and what these decision makers understood and were able to create policy in response.  Warning of impending attack was expressed clearly to the incoming Bush administration, which was initially slow to understand the threat; the now public PDB of 6 August 2001 indicates that the IC was very forthcoming regarding the likelihood of attack.[34]  As previously mentioned, Betts’ suggestion that even in the face of an attack warning, policy makers tend to discount the warning if it contradicts prevailing assumptions or norms, or discount the urgency of the threat.  This is most certainly the case, as it was as late as 4 September 2001 that the Bush Cabinet met to assess a draft of a new policy toward al Qaeda and Afghanistan; if the administration had understood the urgency of the impending threat presented to them, surely stronger and swifter action would have been taken.[35]    Operational evaluation and defense planning as areas of broken links between analysis and decision makers can be examined together, as the CIA did provide pessimistic assessment of covert action targeting bin Ladin in Afghanistan during the Clinton administration, as well as use this threat to secure more funding for the agency to implement defense against the growing capabilities of al Qaeda, just as Betts predicted could happen.  President Clinton himself chose to pursue possible responses to the terrorist threat that were offered to him that would have the least impact on the domestic politics he faced, such as covert action and the expensive but unmanned weaponized Predator drones to target bin Ladin and al Qaeda.[36]  Given the knowledge that bin Ladin was determined to attack in the US and that hijackings and airline bombings were potential methods for attack, increasing both border defense, airline security, and cooperation and awareness among domestic agencies could have been appropriate ways to increase homeland security against the al Qaeda threat.  These kinds of defensive policies were implemented successfully prior to the new millennium, but as the 9/11 Commission Report recognized, this kind of government-wide vigilance and awareness were not implemented again in response to the acute threat of terrorist attack during the summer of 2001.[37]

Betts also highlighted three areas where analysis fails, and the case study of 9/11 demonstrates some analysis failure, but as Betts also suggested, the limited failures of analysis point toward a greater failure in policy.  Betts indicated that too much ambiguity in evidence, from either too much or too little information, could leave room for policy makers to misunderstand the certainty, or uncertainty of the analysis provided.  While there were many ambiguities of evidence leading up to the attack, the PDB of 6 August 2001 demonstrated that while the precise attack was unknown, likely targets and methods were known.  As mentioned previously, Betts described ambivalence of judgment as a barrier to analytical accuracy, suggesting that when analysts err on the side of caution in light of the ambiguity of evidence, policy makers are given a greater degree of freedom and less liability in choosing policy without a definitive assessment from analysts.  There is strong evidence that the IC did not fail to provide decisive judgment, and that in fact, “From George Tenet to the lowest-paid linguist in the Counterterrorist Center, the system was biased toward sounding the alarm.”[38]

The final area where analysis can flounder is from the atrophy of reforms, which was unfortunately the case within the IC prior to 9/11.  After the embassy attacks in 1998, DCI Tenet declared war on bin Ladin and al Qaeda in a memo distributed within the IC.  Yet the lack of actual authority over priorities and budgeting of the majority of the agencies within the IC left the DCI in a weak position to carry out his declared war.  This war was poorly understood within the larger IC, and little was done to implement new priorities at the other agencies in the IC aside from the CIA.[39]  DCI Tenet also implemented new leadership, priorities and operations for the CIA’s Counterterrorist Center in early 1999 despite his inability to gain additional funding for fighting the bin Ladin threat.[40]  While these changes within the CIA made progress, the new enthusiasm within the Counterterrorist Center did not produce fast enough results, and the inherent weakness of the DCI as leader and chief analyst of the IC did not lead to unified action within the IC.

Reform of the IC and Betts’ Imperfect Solutions

Perhaps the most prominent reform passed as part of the Intelligence Reform and Terrorism Prevention Act of 2004 (IRTPA), which enacted many of the recommendations of the 9/11 Commission Report, was the creation of a director of national intelligence (DNI), which replaced the director of central intelligence (DCI).  As previously stated, Betts predicted that organizational flaws were the most identifiable contributions to failure, and the most likely target for reform.  The new DNI is still the president’s primary intelligence advisor, but differs from the previous DCI position through oversight of national intelligence, which includes foreign and domestic intelligence as well as homeland security.[41]  Betts was skeptical of consolidation within the IC, believing that the differing approaches and biases of each organization within the IC contributed unique perspectives that should not be lost, and that agency competition was practically impossible to eliminate.[42]  While agency competition is unlikely to ever cease, the creation of the National Counterterrorism Center (NCTC), which is staffed by personnel from multiple agencies, should enable government-wide cooperation and united effort that the US lacked, particularly in the domestic arena before 9/11.[43]

This essay has already established the drawbacks of multiple advocacy and devil’s advocacy practices within intelligence analysis, yet there are specific instances where such practices offer benefit.  The 9/11 Commission Report cited a failure of imagination within the IC, particularly the failure to imagine aircraft as weapons, and suggests the institutionalization of imagination as necessary.[44]  Betts cautions that the formal use of outlandish possibilities becomes superficial and regarded as crying wolf, but the 9/11 Commission Report draws attention to a systematic approach to preventing specific surprise attacks developed during the Cold War that were disregarded in the 1990s that may offer a legitimate approach to planning for the unimaginable.[45]

The most important solution Betts suggested was the need to change the cognition of the policy maker.  He emphasized that what is really needed is reflection and impartial, non-political approaches to assessing and acting on intelligence on the part of the decision maker.  President Bush was seen by many as lacking the depth of knowledge and serious thought necessary to act on grave matters of national security, and his administration was accused of misusing intelligence for political purposes, exhibiting the exact traits that Betts cautioned against.  President Obama is in many senses the opposite, embodying the reflective academic that struggles to act decisively.  While Obama may be just the conscious and careful president that Betts would admire, taking too long to consider all of the options and their implications is exactly the trap that allowed the previous two presidents the room to not take any action at all, until it was too late.  Just as Betts predicted, by solving one area of potential failure, another hole in the system emerges, continuing the inevitable game of whack-a-mole.

Concluding Thoughts

Intelligence failures are inevitable due to the political and psychological limitations of decision makers that inhibit them from utilizing intelligence to develop effective policy responses to the threat that the intelligence warns of.  Betts developed a theory of intelligence failure to understand why failure continues to occur and where error generally occurs within the intelligence process.  The case of 9/11 illustrates that intelligence failure continues to occur within the modern IC just as Betts predicted.  As Betts’ theory predicted, the failures of 9/11 were the result of some shortcomings of analysis and organizational structure, with much of the failure occurring on the part of the decision makers.  Although intelligence failures can never be eliminated, the IC has learned from its mistakes by developing what Peter Gill and Mark Phythian describe as “intelligence learning”, which has reduced failures at the level of collection and analysis.[46]  The reforms of IRTPA in 2004 have made great strides towards eliminating failure at the organizational level, but “intelligence learning” has not yet occurred at the level of the policy maker.  While Obama may be more cognizant of the dangers of using intelligence to serve political goals, the president and his administration change, and each new group is fairly blind to the lessons learned by the previous administration.  “Intelligence learning” has not yet been institutionalized for the office of the president, and as a result, failures in policy will be attributed to failures in intelligence, and they will continue to inevitably occur.  More attention needs to be drawn to the responsibility of the policy maker in intelligence failure, and Betts, Marrin, Gill and Phythian are all direct and forthcoming at placing the blame squarely on the shoulders of the decision makers.  Unfortunately, the most widely recognized analysis of intelligence failure was the 9/11 Commission Report, which did identify a failure in policy, but fell far short of directly of placing blame on any individual or administration.  Until a greater recognition of the role of the policy maker in intelligence consumption is realized, policy failures will continue to cause what the public perceives as intelligence failures, and theses failures will inevitably continue.

Further Reading on E-International Relations

• On the Role of Intelligence in the Pacific Theater of the Second World War
• What’s Stopping Us? The Failures Behind Famine Prevention in the 21st Century
• The Gouzenko Affair and the Development of Canadian Intelligence
• Cuban Intelligence after the Cold War: A Case Study in Adaptation and Influence
• The Convention on the Elimination of All Forms of Discrimination Against Women’s Failures in Fiji
• Can the Use of Torture in Intelligence Gathering Be Justified?