Unlike other intelligence reports, which focus on current intelligence, National Intelligence Estimates are forward-looking assessments.  Such Estimates, from the earliest days of the modern U.S. intelligence system—the product of the National Security Act of 1947—have been considered to be the best analysis of specific issues of national importance or of national crisis situations that could be brought to bear by the Director of Central Intelligence (DCI), with the concurrence of the other intelligence organizations of the United States Government.  As DCI Walter Bedell Smith put it in a 1950 meeting of the Intelligence Advisory Council,

Accordingly, the responsibility for drafting Estimates, after briefly being assigned to CIA’s Office of Research and Estimates (ORE), was located in CIA’s Office of National Estimates (ONE) as of November 1950.  ONE performed its estimative task fully, preparing more than 1,500 of them until the office was disestablished in November 1973. [2]   ONE was a small organization, consisting of a Board of National Estimates of between five and twelve senior experts, a professional staff of 25-30 regional and functional specialists, and a support staff. [3]

Estimates could be requested (tasked) by the President, members of the National Security Council, any member of the United States Intelligence Board (USIB—predecessor of the National Foreign Intelligence Board, discussed below), or by the leadership of ONE itself.  Upon completion by ONE—a process that averaged about 6-8 weeks, Estimates were forwarded to the DCI, who presented them to the weekly USIB meeting for final concurrence.  At this point, if individual bureaucracies had specific objections to judgments made in the Estimate, they would be discussed, registered, and entered into the final draft.  Final copies of Estimates were disseminated by ONE to 100-300 individuals or offices within the U.S. Government, depending upon classification levels, subject and relevance.  After publication, many Estimates also were subjected to a formal review of “intelligence gaps” or shortfalls of information it was hoped could be addressed by intelligence collectors. [4]

To improve responsiveness to intelligence needs and to better engage the Intelligence Community members [5] in the drafting of estimative intelligence, the ONE was succeeded in 1973 by National Intelligence Officers.  This group of substantive experts became the National Intelligence Council in 1979. [6]   The final approval for NIEs currently is the responsibility of the National Foreign Intelligence Board, which is chaired by the DCI or Deputy DCI, and consists of the heads of the principal intelligence collection and analytic services in the US Government. [7]

To this day, Estimates remain controversial.  Again, to paraphrase Sherman Kent, estimating is what you do when you do not know something with exactitude or confidence.  In discussing large or complex topics, National Intelligence Estimates necessarily have to delve into a realm of speculation, a dense process of trying to separate out the probable from the possible from the impossible, and of providing answers to difficult but important questions with an appropriate degree of uncertainty about incomplete information. 

Estimates written at the specific request of a policy principal, or focused on an ongoing crisis, are likely to be read avidly and be an important factor in crisis management and decisionmaking.  If they are highly technical and involve weapons of mass destruction, they will be read carefully and be factored into long-range planning processes, particularly by military consumers.  If they are more general overviews of internal politics, economic development, or even foreign policy, they are less likely to be read by key policymakers, but they may be highly useful in educating middle-level officials and other members of the Intelligence Community on general policy issues and potential problems just over the (invariably short) horizon of the policy players. 

In any case, Kent’s advice to those charged with preparing Estimates remains sound.  An Estimate,

Equally important, NIEs are documents of record, contributions to institutional, and perhaps national history.  Current intelligence analysis disappears quickly and even more thoroughly than yesterday’s newspaper.  Mid-range analysis is usually remembered only if it’s wrong.  But Estimates put the big judgments on the record, they represent the collective knowledge of hundreds of intelligence analysts, and they are intended to stand a test of time—in most cases, two to five years.  So in a sense, they are written for historians as well as policymakers.