Intelligence in Public Media

Intelligence and Surprise Attack—Failure and Successes from Pearl Harbor to 9/11 and Beyond
Erik J. Dahl (Georgetown University Press, 2013) 277 pp., endnotes, bibliography, index.
Constructing Cassandra—Reframing Intelligence Failure at the CIA, 1947–2001
Milo Jones and Philippe Silberzahn (Stanford University Press, 2013) 375 pp., endnotes, bibliography, index.

Reviewed by Randy P. Burkett

On publication in 2013, these two books were part of a growing cottage industry of works designed to dissect intelligence failures and help the Intelligence Community—and CIA in particular—do its job better. Both draw from, and seek to improve on and update, classic works like Roberta Wohlstetter’s Pearl Harbor: Warning and Decision (Stanford University Press, 1962) and Richard Betts’s Surprise Attack: Lessons for Defense Planning (Brookings Institution Press, 1982).

Each book starts by providing a framework to understand the complexities involved in preventing a future intelligence failure and then uses this framework to revisit historic case studies of infamous failures. Erik Dahl’s book goes two steps beyond and discusses intelligence successes as well as failures and some unique challenges of dealing with terrorist attacks. Jones and Silberzahn dismiss the “dark matter” of what they assume are “CIA’s many intelligence successes” because these successes become “nonevents.” (14) Strangely, even though Jones and Silberzahn assert that CIA suffers from multiple defects that spawn intelligence failures, they make no effort to examine why these supposed flaws did not affect the “many intelligence successes” they chose to ignore.

Dahl is a former naval intelligence officer and is currently a professor teaching homeland security at the Naval Postgraduate School. His insider status shows through, as he examines the complexity of preventing an intelligence failure. Dahl looks at four steps that need to be accomplished to prevent a failure: intelligence collection, analysis, dissemination, and the policymakers’ willingness to accept the intelligence provided. This fourth step is a unique contribution of his work. Dahl expects a lot from intelligence producers. To avoid being blamed with a failure, CIA must not only collect sufficient, specific information to provide for both strategic and tactical warning, it must also shepherd this information through an analytic production process that results in timely dissemination to the policymakers who need it, and convince them of the accuracy of the intelligence—even though the message is likely not one the policymakers will want to hear. If policymakers refuse to accept the intelligence, then Dahl lays the blame back on CIA, insisting the rejection must be due to a lack of clarity or specificity.

Dahl recognizes this dilemma, writing, “The argument can be made that I am setting the bar too high; it may seem unreasonable to hold the Intelligence Community at fault if leaders do not listen. And at the same time, it may appear that I am letting decisionmakers off too easily, if a failure on their part is defined as an ‘intelligence’ failure. My answer is that we need to take a broader view, and realize that the surprise can be just as great, and the harmful effects just as serious, when a surprise attack is successful, whether or not intelligence professionals ‘called it right.’” (20) Given the fact that policymakers are unlikely to ever admit that intelligence “called it right” but they just did not believe the news, the result will likely be more blame in the future.

Despite this drawback, Dahl’s work provides a valuable new look at this subject through its review of intelligence successes, his insights into the unique challenges of warning of terrorist plans, and his comprehensive examination of 9/11. Dahl is not afraid to take on conventional wisdom, such as when he asserts that Wohlstetter was wrong in Pearl Harbor; the issue was not “picking out the signals from the noise,” but rather the failure to collect sufficient specific intelligence to provide tactical warning.  He points out the same problem happened on 9/11—the failure was not a lack of strategic warning that something was coming but that actionable intelligence was missing to allow policymakers to respond in time. The book’s appendix, listing 227 “Unsuccessful Plots and Attacks against American Targets, 1987–2012,” is a particularly valuable contribution to the literature and provides future researchers with leads to look at what went right in these cases.

In contrast to Dahl’s balanced and insightful work, the authors of Constructing Cassandra appear to have little understanding of how intelligence works and rely on a wide variety of secondary sources to prop up their cultural and organizational arguments. The book, which Jones has described as essentially his doctoral dissertation, begins with a list of what CIA has gotten wrong—including the usual cases of the Cuban missile crisis, the fall of the Shah of Iran, “missing” the collapse of the Soviet empire, and 9/11. Jones and Silberzahn explain that their book is not meant to be an attack on CIA but rather “to understand why the agency fails so often.”

The authors attribute this failure to CIA’s “identity and culture” and assert that these factors drive the analysts’ perceptions and questions, resulting in their missing pertinent information or discarding it because it fails to conform to their perception of reality. The authors, both instructors in European business schools, try to make the case that identity and cultural issues (which they further define as a lack of diversity in the workforce, an overreliance on “scientism” in analysis, a fixation on secrets, a demand for consensus in finished analysis, and a servile attitude to policymakers as “customers”) are the roots of CIA’s failures. (50)

While CIA has readily acknowledged that it continues to build a more diverse workforce, the authors’ first assertion, that the agency, and particularly the analytic workforce, is composed of white, Anglo-Saxon, Protestant men who came from Ivy League schools and never traveled or lived overseas is bizarre. Dr. Mark Lowenthal, former senior CIA analyst and manager, addressed this claim in a joint appearance with Milo Jones at the Spy Museum on 3 September 2014.[1] According to Lowenthal, the workforce Jones describes looks nothing like what he was part of and, if it existed at all, it would have been the workforce of the 1960s, not that of the 21st century.

Lowenthal also took strong exception to the authors’ description of “scientism” as a basis for analysis. While on one hand the authors decry the large number of political science majors among the analyst ranks, they also recommend bringing in more anthropologists and sociologists—as if one form of social science would balance another.

The authors get little right when it comes to analysis. They assume CIA analysts pay little or no attention to open source products and only rely on “secrets.” They advocate “educating” analysts rather than “training” them—ignoring the fact that analysts come in as well-educated individuals who are often selected because of their proven academic expertise in an area, and are then “trained” in the writing style and the range of classified products that they can now use, in addition to the open source products they relied on as students.

The two business school professors appear to be confused by the use of the term “customer” in intelligence and demonstrate no understanding of how tasking originates in the intelligence cycle. They assume that analysts adhere to a “the customer is always right” business-style attitude and tailor their products inappropriately to tell customers what they want to hear, that analysts will not produce anything that goes against the consensus of everyone they work for, and that they provide false levels of predictive certainty (again, based on an overreliance on “scientism” in describing human actions). The authors incorrectly assume that CIA managers create their own taskings and, as such, they advocate changing the relationship with consumers to teach them that CIA “is not in the answer-fetching business and arguably should give the consumers more questions to consider rather than answers.” (15)

If the authors had remained with the title’s premise, to identify Cassandras who were correct but unheeded in specific intelligence failures and then examine why they were right and everyone else was wrong, the book could have been useful. However, the Cassandras cited do not live up to the title. For example, in the case of the Cuban missile crisis, DCI John McCone is offered as an example and while it is true that McCone believed the Soviets would take a risk in Cuba at a time when CIA experts disagreed with that assessment and omitted it from a Special National Intelligence Estimate one month before the nuclear missiles were found, McCone admitted that his view was based on his “judgment” that had little value until hard evidence was obtained. (150) The best the authors can suggest is that the missiles might have been identified more quickly if McCone had been believed—not that the crisis could have been averted. None of the four intelligence failures they cite offers a clear, well documented example of a person (or persons) who clearly predicted exactly what happened in advance and had views that could have changed the outcome.

Jones and Silberzahn are right on one point however: intelligence successes do tend to result in “nonevents” or at least in the public’s not hearing about what occurred until 25 to 30 years have passed. As a result, intelligence failures will continue to seize the public’s attention and generate more books intended to fix intelligence so that the failure will never happen again.

In the interim, intelligence professionals will have to continue to live with accusations of failure, real and imagined, and continue to provide US policymakers with the answers they need to keep the country safe. Dahl’s book is a welcome addition to an intelligence officer’s bookshelf and will likely be a valued text for intelligence classes. Jones and Silberzahn need to study the realities of CIA and the intelligence process more closely before offering ideas for improvements.

[Top of page]


[1]Spycast, Constructing Cassandra, talk and interview with Milo Jones, 3 September 2014, accessed 22 November 2016,

[Top of page]

All statements of fact, opinion, or analysis expressed in this journal are those of the authors. Nothing in any of the articles should be construed as asserting or implying US government endorsement of their factual statements and interpretations. Articles by non-US government employees are copyrighted.

Posted: Dec 22, 2017 12:10 PM
Last Updated: Dec 22, 2017 12:10 PM