Intelligence Analysis and Assessment

David Charters, Stuart Farson, Glenn P. Hastedt

• Publisher: Routledge
• Year: 2013
• Print ISBN: 9780714647098
• eBook ISBN: 9781135212452
• Topic: History
• Subtopic: Military & Maritime History
• Index: History

These essays cover: assessment systems now in place in Britain, the USA, Germany and Australia; the bureaucratic dynamics of analysis and assessment; the changing ground in intelligence; and the impact of new technologies and modes of communication on intelligence gathering and analysis.

II. NATIONAL ASSESSMENT SYSTEMS

Assessment Machinery: British
and American Models

MICHAEL HERMAN

‘Assessments’ (or to the Americans ‘estimates’) are where intelligence bears most closely on policy. There is an intelligence ideal of objective and authoritative assessment for the top level of government. Disagreements over intelligence interpretations should be disengaged from interdepartmental policy differences and solved prior to decision-taking. The principle is the same as in the British Government Statistical Service’s objective ‘to make sure that the Cabinet need never argue about statistics’.¹

There has been considerable academic writing about intelligence-policy relationships at this top level and the nuances of presenting intelligence’s alternative and minority interpretations there. The actual machinery for assessment has received rather less attention. In this, countries have choices broadly between British and American organizational models. The first is the system of the Joint Intelligence Committee (JIC); the second is the American framework for producing National Intelligence Estimates (NIEs). These arrangements and their resemblances and differences are the subject here.

TOP LEVEL DECISION-TAKING AND DESIDERATA

The top level decision-takers considered here are Presidents and Prime Ministers; members of the White House staff, the National Security Council (NSC), the Overseas Policy and Defence Committee of the British Cabinet and other formal and informal groups; and other senior government members, officials and service officers operating collectively. Intelligence assessment for these groups is the business of presenting conclusions and supporting arguments and evidence in forms easy for them to use; it is distinct from the more detailed desk-level analysis and research that precedes it. Assessment of this kind must be capable of viewing foreign states as a whole and relating them to others with a broad focus, as national assessments. These need the ability to tap government sources across departmental boundaries, to bring all relevant knowledge to the table through processes of interdepartmental integration or community assessment. These arrangements need to promote objective analysis, with the discipline of scrutiny and argument to eliminate departmental biases. The product must command interdepartmental agreement to give it credibility, so that the decision process is not obstructed by intelligence wrangles.

Assessment of this kind has its purely intellectual difficulties. The subject-matter is what Sherman Kent called ‘speculative knowledge’, and ‘the speculative-evaluative element’ of analysis; the prerequisite is ‘not the important but gross substance which can be called recorded fact; it is that subtle form of knowledge which comes from a set of well-stocked and well-ordered brain cells’.² Reflecting on assessments (estimates) Ray Cline wrote that:

Estimates are careful descriptions of the likelihood that certain things will exist or occur in the future. National Intelligence Estimates are papers setting forth probable situations or occurrences that would make a major difference to our national security or our foreign policy. When the answers are clear to questions about the future, there is no need for an estimate. The easy questions are never asked. An estimate tries to reduce the inevitable degree of uncertainty to a minimum in making calculations about future situations. ... It is not easy to make estimates that are well-calculated in terms of evidence and logic, carefully set forth with great objectivity, and plainly relevant to decision-making.³

But they are also team efforts, coming from the orchestration of groups. The original military intelligence model was of the intelligence officer producing ‘his’ assessments for ‘his’ commander, and much intelligence is still organized in this way, departmentally, with primary allegiance to a single department. It evolved in the second half of the nineteenth century as part of war ministries and admiralties, and comparable non-departmental agencies emerged only slowly in this century. Defence intelligence, the biggest block of analysis, is still subordinate to Defence departments. Departmental subordination also applies to the American State Department’s Intelligence and Research organization (INR) and the similar intelligence divisions of Treasury, Energy and Commerce, and to their British equivalents. Serving collective top-level divisions means that intelligence has to come together in some supra-departmental way beyond purely departmental analysis. Doing this poses the organizational problem.

There is nothing unique about it. Intelligence emerged as part of organized knowledge within government: ‘Between Burke and Balfour, the “expert” – a protean image of authority and rational knowledge -became a key factor in the “technique of government” that accompanied the nineteenth century revolution in government.’⁴ In this process there has been continual debate on how to organize this expertise. British government statistics for example have slowly evolved from departmental towards supradepartmental structures, in an evolution not dissimilar from intelligence’s .⁵ A.J.P. Taylor’s description of these statistics in the 1930s – ‘four separate departments collected industrial statistics; five classified employers of labour; two produced rival and conflicting figures concerning overseas investment’ – could well have applied to the departmental intelligence of the time.⁶

There are no tidy solutions. Most nations are not interested in community assessment anyway. But for those that seek it there is a broad choice between emphasizing interdepartmental arrangements that enable departments to cooperate collegially, and forms of central intelligence that supplement or supplant the departmental system. The former is characteristic of the British JIC; the latter is epitomized in America by the position of the Director of Central Intelligence (DCI) and the Intelligence Directorate of the Central Intelligence Agency (CIA); but in fact both national systems have elements of both collegiality and centralism, as emerges from an account of their evolution.

BRITISH AND AMERICAN EXPERIENCE UP TO 1945

National assessment of states in the round evolved out of twentieth century warfare; total war needed total assessment. World War I was fought with unsophisticated central direction, and one symptom was the independence of military, naval and (for Britain) blockade intelligence. Yet the war brought home the need for central defence planning, and hence for viewing foreign military power in terms of national industrial capacity, demography and cohesion as well as the traditional military yardsticks. Britain in the interwar years developed some integrated military planning around the Chiefs of Staffs Committee, and this formed the basis for Grand Strategy in World War II. It also produced the need for similarly integrated intelligence inputs.

This was met in Britain through two routes. One was to establish non-departmental, ‘national’, central units for analysis that fell outside departmental boundaries. The Security Service emerged between 1921 and 1931 as the non-departmental agency for internal security.⁷ The real prototype of non-departmental analysis for top level decisions was however the tiny Industrial Intelligence Centre (IIC), established in 1931 to study the German economy on behalf of all government users.⁸ World War II produced more non-departmental, inter-service units. Most of them were for dealing with particular collection sources – for Sigint, photographic interpretation, and strategic prisoner-of-war interrogation- or with covert action and political warfare. But some were for data-handling and analysis. Tri-service warfare, especially the amphibious invasions of Europe, could not be planned on a single-service basis and needed similarly integrated intelligence support; the biggest example of this kind was the Inter-Service Topographical Department, responsible inter alia for collecting and collating the beach intelligence needed for the Normandy landings.⁹

But non-departmental models still encountered departmental objections, so the other line of British development was to retain departmental organizations but superimpose interdepartmental coordinating machinery upon them. The JIC was formed in 1936 with a membership of the three service intelligence departments and (later) the IIC and the Foreign Office, and became effective from 1939 onwards when Foreign Office membership (and chairmanship) brought the integration of political, military and economic assessment.¹⁰ It was also stimulated in that year by the need for some warning machinery to evaluate the contradictory reports of short-term German intentions. A separate Situation Report Centre established for this purpose was soon incorporated within it. This gave the JIC the general responsibility which it bore throughout the war, in support of the Joint Planning Staff, the Chiefs of Staff, Churchill as Minister of Defence, and the War Cabinet, for ‘the assessment and coordination of intelligence received from abroad with the object of ensuring that any Government action which might have to be taken should be based on the most suitable and carefully coordinated information available.¹¹ A crucial addition to it was the creation in 1941 of the Joint Intelligence Staff (JIS), made up of departmental representatives, originally just to act as the drafting sub-committee but in the event with a far greater effect on the output. All this was joined with the Committee’s other responsibility – not dealt with here – for the management of the intelligence system as a whole.

The American system before 1941 of separate Army and Navy intelligence plus the FBI was even more departmental than its British counterpart, and after Pearl Harbor came under similar pressures for supradepartmental structures. Many of the wartime British bodies became combined Anglo-American entities or were duplicated in American versions. In addition, the Research and Analysis Division (R and A) of the American Office of Strategic Services (OSS) developed as a large, allpurpose analysis agency, supporting and competing with the services’ organizations.¹² Under British pressure the interdepartmental JIC model was also adopted, and interlocking committees in Washington and London produced agreed intelligence inputs to the two countries’ Combined Chiefs of Staffs Committee.¹³

World War II thus ended with the lesson that purely departmental intelligence was not enough. Pearl Harbor burnt the lesson into American memory that some central point was needed for warning intelligence, and the war’s higher strategy demonstrated the comparable need for integrated longer-term assessment. Comparison of Allied wartime arrangements with the uncoordinated and unsuccessful systems of the Axis seemed to clinch the point. Private intelligence services had proliferated under Hitler, with no machinery below Head of State for synthesising the results; Japanese intelligence was inefficient and unsystematic. Yet in Britain and America there remained an unstated – perhaps hardly understood – conflict between the two supradepartmental variants: central organizations on the one hand and collegial or community machinery on the other.

POST-1945 MODELS

In the post-war reconstruction the British moved towards central, ‘national’ collection agencies, notably in establishing Government Communications Headquarters (GCHQ) as the signals intelligence organization. In analysis there were moves in this same direction when economic, topographical and some scientific intelligence were centralized in a new Joint Intelligence Bureau (JIB).¹⁴ But the main weight was placed on continued departmentalism with community machinery added to it. Service intelligence remained single-service; the JIB’s role was limited to gap-filling on subjects of common interest. The JIC was confirmed as the organ of community assessment, though initially with a military slant.¹⁵ Subordinate JICs were established overseas, and the joint intelligence structure developed in Malaya became standard for counter-insurgency campaigns.¹⁶ ‘Jointery’ – cooperation between the three services and other departments through committees – became the standard British solution in intelligence as elsewhere. Yet to some this seemed a concession to single-service prejudices against the logic of central analysis. Professor R.V. Jones’s verdict on the postwar structure established for scientific intelligence was that ‘in six hours the experience of six years was jettisoned.’¹⁷

An official snapshot of the post-war JIC system was first given in the report of the Falklands Committee chaired by Lord Franks in 1983.¹⁸ The ‘Joint Intelligence Organisation’ is based on the Cabinet Office. Apart from its managerial responsibilities the JIC makes ‘assessments for ministers and officials of a wide range of external situations and developments’. These are prepared by geographically-based Current Intelligence Groups (CIGs), made up of ‘those in the relevant Departments with special knowledge of the area.’ The groups are serviced and chaired from the Cabinet Office by members of the ‘Assessments Staff’, made up mainly of seconded Foreign and Commonwealth Office (FCO) and Ministry of Defence civil servants and service officers. CIGs’ assessments are normally (but not necessarily) considered in draft by the JIC itself, made up of the heads of the security and intelligence agencies and representatives of the Ministry of Defence and Treasury and ‘other Departments, including the Home Office, as appropriate.’¹⁹ Prior to the Franks recommendations the committee was chaired by a senior FCO official as it had been since 1939.

This is essentially the World War II system, with a wider remit and a transfer to Cabinet Office control in 1957. The Assessments Staff is a more powerful successor of the wartime JIS. The CIGs are an expanded and formalized version of wartime inter-service discussions. On Franks’ recommendation, the JIC chairman became a full-time Cabinet Office official appointed by the Prime Minister. This was a novelty, though the idea had been urged in print by a retired deputy chairman some years earlier, and completely ignored.²⁰ Subsequently, the post has reverted to being the part-time responsibility of a busy FCO official, at present the head of the Cabinet Office’s Overseas Secretariat. The structure retains its heavily collegial character as a set of committees of equals.

Central intelligence made more progress in America with the evolution of the OSS’s R and A to become the basis of CIA’s Intelligence Directorate. The National Security Act of 1947 established the CIA ‘to correlate and evaluate intelligence relating to the national security, and provide for the dissemination of such intelligence within the Government using where appropriate existing agencies and facilities. ’²¹ The post of DCI, also the CIA’s Director, was established as intelligence adviser to the President and the NSC.

However the DCI and CIA were never intended to become sole players. The 1947 Act charged the DCI and CIA with coordinating departmental intelligence for national assessment, and with services of common concern. (The DCI was also given the origins of his modern managerial responsibilities for the community, but like those of the JIC these are ignored here.) The subsequent emergence of CIA as a comprehensive central analysis agency was a matter of bureaucratic growth and accident – including the services’ refusal to release sensitive material to it. From 1950 onwards NIEs were presented by the DCI as community products, produced through a JlC-like process.²² What emerged was a mixed system; partly collegial, but with the DCI and CIA as potentially powerful central elements. For convenience ‘CIA’ is used here to refer to its Intelligence Directorate only, not its Operations Directorate for Collection and covert action.

There were many subsequent adjustments to the American machinery.

Up to 1973, estimates were drafted on the DCI’s behalf by the Office of National Estimates (ONE), which coordinated them with the community before final approval by the Board of National Estimates. In 1973 ONE was abolished and replaced by the rather looser system of National Intelligence Officers (NIOs), with changes in the forms of approval.²³ Other minor changes have been made periodically, but the basic framework of national assessment by NIEs and the like (including SNIEs, or Special NIEs for matters of urgency) has remained unaltered.

One earlier change, common to Britain and America, is also relevant: the creation of central service intelligence organizations as part of the coordination of defence. America created the Defense Intelligence Agency (DIA) in 1961, though it left powerful single-service agencies in place. The DIA’s effects have been muted by the single-service agencies, which have tended to make it seem just one service voice among many. Britain amalgamated the three service elements and the JIB into the Defence Intelligence Staff (DIS) in the defence reorganization of 1964. One British result was that inter-service debates ceased to be conducted in the JIC forum, with some consequent decline in the Foreign Office’s original 1939 role of inter-service mediator. However its influence was strengthened by the consolidation within it shortly afterwards (with the change of name to the Foreig...

Table of contents

• Cover
• Title page
• Copyright page
• Contents
• I. INTRODUCTION
• II. NATIONAL ASSESSMENT SYSTEMS
• III. THE PRODUCER/USER INTERFACE
• IV. NEW ANALYTICAL PRIORITIES
• V. THE OPEN SOURCE REVOLUTION
• Notes on Contributors