Strategic Intelligence for American World Policy (1949) by Sherman Kent is the founding text of intelligence analysis as a professional discipline. Written after Kent’s wartime service in the OSS’s Research and Analysis Branch and before his appointment as chairman of the CIA’s Board of National Estimates (1952), the book argues that intelligence analysis is an intellectual discipline with its own standards, methods, and professional ethics — not a bureaucratic function subordinate to military command or political preference.
Core argument
Kent’s central claim is that the intelligence analyst occupies a position analogous to the academic researcher: the analyst’s obligation is to the truth of the assessment, not to the preferences of the consumer. The policymaker needs to know what is happening and what is likely to happen; the analyst provides this; the policymaker decides what to do. This separation of knowing from deciding — which Kent understood as essential to both intellectual integrity and policy effectiveness — grounds the discipline’s resistance to politicization and defines the analyst-policymaker relationship as the profession’s central problem.
Kent identified three components of strategic intelligence: intelligence as knowledge (the body of assessed information about the world), intelligence as activity (the process of collection, analysis, and dissemination), and intelligence as organization (the institutions that perform the activity). Each component has its own requirements and pathologies; the discipline’s health depends on getting all three right.
Influence
The book’s influence is institutional: the CIA’s Directorate of Intelligence, the Board of National Estimates, the National Intelligence Estimate as a product format, the concept of estimative intelligence, and the professional identity of the intelligence analyst all derive from Kent’s framework. His insistence on analytic independence — however imperfectly realized — provides the standard against which departures are measured.
Limitations
Kent’s framework assumes that the policymaker needs and will use the analyst’s assessment. Paul Pillar’s work demonstrates that this assumption is frequently wrong: policymakers often make decisions on criteria the intelligence assessment does not address, rendering the assessment structurally irrelevant rather than politically contested. The 2026 Iran war presents this pattern: the intelligence community’s assessment was not rejected — it was bypassed.
Related concepts
- Analyst-policymaker relationship — the relationship Kent defined
- Estimative intelligence — the product type he formalized
- Politicization — the corruption his framework was designed to prevent