WarningThis information has been archived for reference or research purposes.

Archived Content

Information identified as archived on the Web is for reference, research or recordkeeping purposes. It has not been altered or updated after the date of archiving. Web pages that are archived on the Web are not subject to the Government of Canada Web Standards. As per the Communications Policy of the Government of Canada, you can request alternate formats on the "Contact Us" page.

Book Reviews



Reviewed by Douglas Agnew

Print PDF

For more information on accessing this file, please visit our help page.

by Robert Jervis
Ithaca, NY: Cornell University Press, 2010
248 pages, $US27.95 (cloth)
ISBN: 978-0-8014-4785-3

Reviewed by Douglas Agnew


Robert Jervis’ latest book is a bit cheeky. His examination of why the CIA was slow to realize the Shah might fall is centered upon a recently de-classified report he wrote on that subject for the Agency in the spring of 1979. That said, this is actually an effective device for comparing the reasons for intelligence failures separated by more than two decades. Jervis also casts his net wider than most post mortems, bringing to bear his understanding of analysts and of social science methods.

Jervis challenges the assumption that fixing the ‘machinery’ of intelligence will solve the problem of intelligence failures. The reasons for a failure may lie as much in analysts’ minds as in the structure and processes of their agencies. One of Jervis’ main themes is that intelligence analysts may rely upon intuition rather than upon social science methods. As a result, they may leave crucial assumptions unexamined and unexplained, or, they may fail to develop testable hypotheses. Jervis also makes the points that better analysis may make judgements less certain, rather than lead to wholly different judgements, that it may not lead to more effective policies, and that specific intelligence failures may be avoidable, but the overall phenomenon is not. Anyone seeking to excuse a policy failure or to assign blame for an intelligence failure will find these conclusions disturbing. Equally disturbing for intelligence agencies is Jervis’ conclusion that they seldom have the advantage in understanding revolution and political developments. As he puts it: “[The] CIA and its counterparts are in the business of stealing secrets, but secrets are rarely at the heart of revolutions.”

In the case of the fall of the Shah, there was little contact between political and economic analysts in the CIA. Little attention was paid to the political consequences of economic change, and scant thought was given to how Iranians might use political means to effect economic change. In Jervis’ view, analysts were more like journalists than social scientists, seeking to tell a story, but rarely looking for alternative explanations. They did not realize that a limited human intelligence capability was giving them the wrong impression of the revolutionary movement, and so, they over-estimated the role of the middle class and under-estimated that of religion. Analysts believed the Shah was strong and decisive, and would crack down on dissent if required to do so, a belief which was never actually brought to fruition. They did not realize this belief would prove to be wrong only when it was too late. They did not read the CIA’s own assessments of the Shah from the 1950s and 1960s, and did not realize how much their perception of the Shah had changed.

Jervis identifies a number of failures by analysts in the lead-up to the war against Iraq, some of which may not be apparent at first blush. The American and British intelligence communities did attempt to see the situation through Saddam’s eyes but, they did not realize how very isolated he was. Neither did they understand his regimes’ efforts to balance the need to secure relief from sanctions with the need to deter Iran, nor that Saddam was engaged in a self-defeating bluff. As Jervis writes: “When the situation is this bizarre, it is not likely to be understood.” The inferences drawn about Iraq and weapons of mass destruction were very plausible; however, they were expressed with too much certainty. This was due in part to analysts’ failure to consider alternative explanations. That said, the alternatives may not have been viewed as credible, and, arguably, even what is now believed to have been the case did not fit the evidence any better than did the prevailing assessment of the situation.

Jervis also examines the environment in which the US intelligence community had to operate. Given the justification for the war, that is, to remove a tyrant seeking weapons of mass destruction, it can be argued that there was only a limited role for intelligence. A corollary to this argument is that intelligence errors were then irrelevant. The American intelligence community consistently warned that the aftermath of the invasion of Iraq would be difficult, and denied that there was credible evidence of Saddam’s involvement in 9/11, all in the face of repeated challenges from Vice-President Cheney. Then again, policy makers are right to scrutinize what they believe to be incorrect judgements, and to require a higher level of proof from intelligence agencies when the available evidence does not support their preferred policies. The Bush Administration created “… an atmosphere that was not conducive to critical analysis,” which thereby exacerbated analytical shortcomings. Once the intelligence community accepted that the United States and the United Kingdom were committed to war, it stopped examining evidence with critical eyes.

The author spends little time examining possible structural changes to intelligence organizations, or the American intelligence community writ large. Rather, he considers possibilities such as the practice of peer review typical of the scholarly world. One of his most interesting and unusual suggestions is that one role of intelligence may be to raise questions, to identify fundamental issues to policy makers, and to highlight the judgements upon which decisions should be predicated. Such a role would, no doubt, cause discomfort for those policy makers who expect intelligence to provide only answers, even more so for those who expect intelligence to provide support to policy preferences.

One of the very few and minor flaws in Jervis’ book occurs in the very last chapter, where he uses historical examples to illustrate his points. They are certainly relevant, but Jervis pre-supposes a fairly broad knowledge of intelligence history, and he does not explain them in much detail. Nevertheless, Why Intelligence Fails offers many insights, many more than those mentioned herein. Some students of history or practitioners of intelligence may take issue with the factors Jervis considers, or with the weight he assigns to each in the two intelligence failures. Nevertheless, all will find their copies of Why Intelligence Fails a ‘mass of yellow highlighter’ by the time they have finished reading it.

CMJ Logo

Major Douglas Agnew, an Intelligence Officer, is currently serving as the Canadian Defence Intelligence Liaison Officer to the Defense Intelligence Operations Coordination Center, US Defense Intelligence Agency, Washington, DC. He holds a Master’s degree in War Studies from the Royal Military College of Canada.

Top of Page