United States National Security Agency

The National Security Agency (NSA) is one of the major intelligence agencies in the decentralized U.S. intelligence community described in the book. Its significance lies less in any specific operation than in the organizational problem it represents: specialized agencies can gather distinct knowledge, but without effective ways to connect their information and judgments, the wider system may fail to recognize what its parts collectively know.

Place in the Intelligence Community

After the 1947 National Security Act created the Central Intelligence Agency, the intelligence system did not become the centralized operation envisioned by William J. Donovan. Over time it developed into an overlapping collection of agencies, including the NSA, the Central Intelligence Agency, the National Imagery and Mapping Agency, the National Reconnaissance Office, and the Defense Intelligence Agency, as well as intelligence arms of the three major military services. In theory, the director of the Central Intelligence Agency oversaw the community; in practice, that supervision was limited, and much of the funding for intelligence operations came from the Department of Defense.

The Federal Bureau of Investigation operated almost entirely outside this intelligence community, despite the obvious relevance of information about foreign terrorists inside the United States to foreign intelligence agencies. Rather than Donovan’s envisioned central repository for information and analysis, the country had virtually autonomous groups pursuing the broad shared goal of protection from attack through different organizational channels. The NSA belonged to that fragmented structure, not to a unified chain of command capable of ensuring that information reached every agency that needed it.

Fragmentation and the September 11 Inquiry

Before September 11, 2001, the weaknesses of the system were often overlooked. Intelligence agencies had failed to anticipate the 1993 World Trade Center bombing, the 1998 U.S. embassy bombings in Kenya, and the bombing of the USS Cole in Yemen. After the September 11 attacks, the Congressional Joint Inquiry concluded that the intelligence community had failed to capitalize on the individual and collective significance of information relevant to the plot. Agencies had missed opportunities to disrupt it, and information that might have helped uncover and prevent the attacks had passed unnoticed.

Congress identified insufficient information sharing as a central failure. Instead of combining their holdings into a coherent picture of threats, agencies produced localized snapshots. Senator Richard Shelby criticized the Federal Bureau of Investigation’s decentralized structure for leaving its information holdings fragmented into largely independent fiefdoms, and argued that an agency standing above the disputatious bureaucracies should take charge of U.S. intelligence. His proposed remedy revived Donovan’s call for centralization, but the book distinguishes that response from the underlying diagnosis: agencies needed to share information, yet putting a small group at the center to sift it would sacrifice the diversity, independence, and local knowledge that decentralized specialists can provide.

Decentralization Without Aggregation

The book’s argument is that the intelligence community’s failure was not decentralization by itself but decentralization without aggregation. The division of work among agencies could support specialization and more finely informed analysis. A perspective that seemed irrelevant or noisy to one set of analysts might reveal a crucial clue or relationship to another. The system’s missing capacity was a way to aggregate not only data but also the different judgments of the people examining it. There was no effective mechanism to draw on the collective knowledge of NSA analysts, Central Intelligence Agency personnel, and Federal Bureau of Investigation agents.

This distinction makes the NSA part of a larger organizational question rather than a simple example of a bureaucracy that ought to be centralized. A central authority might make information easier to collect in one place, but it would not necessarily be wiser than the larger community, and it could discard the benefits of specialized perspectives. The book’s alternative is to preserve agencies’ autonomy while making their knowledge collectively useful. Linking agency databases, for example, could ease information flow without requiring agencies to surrender their independence. Even two years after September 11, the government lacked a single unified watch list drawing on data from across the intelligence community—a practical gap that simpler information-sharing systems might have addressed.

Local Knowledge and Organizational Design

This tension reflects a broader principle of decentralization: people close to a problem may possess experience-specific knowledge that is difficult to summarize or transmit. Decentralized arrangements can encourage independence and specialization while allowing people to coordinate and address difficult problems. Their weakness is that valuable information found in one part of a system may never reach the rest. The organizational challenge is therefore to retain local knowledge while creating reliable means to assemble it into a larger picture.

The same problem appears in the book’s discussion of corporations. A rigid hierarchy can obstruct information as it moves between people doing work and executives making decisions; one company chart showed 223 different routes a new product idea might have to travel through the bureaucracy. In contrast, the book argues that decisions about local problems should, where possible, be made by people with direct knowledge of them. The intelligence agencies’ experience illustrates why decentralization alone is insufficient: distributing expertise creates potential value, but only communication and aggregation allow that value to inform collective decisions.