Intelligence failures preceding September 11

The intelligence failures preceding September 11 revealed a central weakness in the U.S. intelligence system: agencies held specialized information, but no effective mechanism brought their separate judgments together. The book neither treats the attacks as plainly predictable nor dismisses the failures as visible only in hindsight. It distinguishes the value of decentralized expertise from the need to make dispersed knowledge collectively usable—and argues that simply concentrating authority at the top would not solve the problem.

A Fragmented Intelligence System

In the postwar period, the Central Intelligence Agency was initially the key player in American intelligence. Over time, however, the community became more fragmented, with overlapping responsibilities distributed among the Central Intelligence Agency, the National Security Agency, the National Imagery and Mapping Agency, the National Reconnaissance Office, the Defense Intelligence Agency, the intelligence arms of the three major military services, and the Federal Bureau of Investigation. Although the director of the Central Intelligence Agency was nominally in charge of the intelligence community, the book says that this authority brought little practical supervision; much of the intelligence budget came from the Department of Defense, while the Federal Bureau of Investigation operated almost entirely outside the community.

The arrangement was decentralized in a way that permitted agencies to pursue a shared broad purpose—protecting the United States from attack—through different missions and methods. But separate agencies could not reliably turn their findings into a shared assessment. The book’s distinction is important: decentralization could support specialization, independence, and local knowledge, yet it could also leave valuable information stranded in one part of the system rather than incorporated into a collective judgment.

Missed Warnings and the Problem of Hindsight

Before September 11, the system had failed to anticipate the 1993 bombing of the World Trade Center, the 1998 bombings of the U.S. embassy in Kenya, and the attack on the USS Cole in Yemen. The attacks of September 11, 2001, made the shortcomings of intelligence gathering seem undeniable. A congressional inquiry concluded that the intelligence community had failed to capitalize on the individual and collective significance of information that appeared relevant to the attacks, and that agencies had missed opportunities to disrupt the plot.

The book immediately qualifies that judgment. An inquiry conducted after an attack can review a large body of information with knowledge of what happened, making previously scattered material appear more obviously connected than it was at the time. In her account of the intelligence failure at Pearl Harbor, Roberta Wohlstetter described numerous signals of an impending Japanese attack but argued that it was unreasonable to expect people to select the decisive ones from the surrounding confusion. Strategic surprise is intrinsically difficult: if a large-scale naval attack was hard to foresee, the book asks, how much harder might it have been to predict an attack carried out by nineteen men?

That caution does not settle the matter. The intelligence community had failed to anticipate four major terrorist attacks from 1993 through 2001. The book therefore leaves open a narrower possibility than foreknowledge of the precise attacks: a different organization might have made it more likely that analysts would recognize the collective significance of the information already available and produce a concrete estimate of the risk.

Aggregation, Not Simply Centralization

The proposed remedy is not to put every agency under a single intelligence authority. Richard Shelby’s critique emphasized the need for a truly central intelligence agency, an arrangement that could make information easier to gather under one roof. The book argues that such centralization would also sacrifice some advantages of diverse, independent agencies with specialized knowledge. A small group of decision-makers, however intelligent, is not necessarily better positioned than the larger community to discern the significance of all its information.

The missing element was aggregation: a way to combine not only information but judgments. The agencies contained distinct kinds of expertise, but there was no dependable process for drawing on the collective assessment of National Security Agency analysts, Central Intelligence Agency officers, and Federal Bureau of Investigation agents. In the book’s framework, decentralization without aggregation yields a collection of groups, not an intelligence organization capable of learning from its own distributed knowledge.

Some possible improvements were strikingly practical. Linking agency computer databases could have helped information travel while allowing the agencies to retain their autonomy. Yet two years after September 11, the government still lacked a single unified watch list drawing on data from across the intelligence community. The contrast captures the book’s central claim: intelligence need not be centralized in one decision-maker for separate findings to become more accessible and collectively useful.

The Limits of the Lesson

The case does not establish that better information sharing would have prevented September 11, or that the attacks could have been predicted with certainty. It instead separates two questions often collapsed in accounts of failure: whether the precise event was foreseeable, and whether a system might have improved its chances of detecting a serious threat in the evidence it possessed. The first may remain unanswerable; the second turns attention to organizational design.

The episode thus illustrates both sides of decentralization. Agencies need room to specialize and develop judgments from their own information, but their independence is insufficient if findings cannot be connected and assessed across institutional boundaries. For the book, the failure was not proof that all intelligence should be directed from the center. It was evidence that distributing knowledge without a way to aggregate judgment can leave a system less capable than the expertise it contains.