HLSS3

ARTICLE

The Politics of Intelligence and the Politicization of Intelligence: The

American Experience

GLENN HASTEDT*

ABSTRACT The relationship between intelligence analysis and policy decisions is a contentious one with both policymakers and intelligence analysts frequently expressing frustration over its underlying dynamics and with each faulting the behavior of the other. This article examines one aspect of this relationship, the manner in which intelligence analysis can become politicized. Rather than view politicization as an aberration it is treated here as a normal feature of intelligence analysis. A typology of politicization organized around the concepts of hard and soft politicization is presented and illustrated with historical examples from the American experience with intelligence analysis.

The relationship of intelligence analysis to policy decisions is a continuing topic of concern and controversy. Here we analyze the American experience with an eye toward coming to a clearer understanding of the politics of intelligence and the extent towhich has become politicized. To this end a typology of intelligence politicization is presented and a series of historical cases are used to help us distinguish between different types of politicization. In doing so we are able to come to a more complete understanding of the dynamics of the intelligence analytic process and the behavior of analysts and policymakers. In no small measure public and policymaker perceptions treat intelligence

analysis as connecting dots. It is a mechanical exercise. The politicization of intelligence is an aberration, one that can be remedied by organizational changes and changing policymakers. The view from inside the intelligence community is not as sanguine with examples of politicization being readily produced across administrations and in varied organizational settings. In no small part the gap between these two perspectives is so large because of differing notions of how to define and think about politicization. For those whominimize its occurrence politicization involves the overt manipulation of

q 2013 Taylor & Francis

*Email: hastedgp@jmu.edu

Intelligence and National Security, 2013 Vol. 28, No. 1, 5–31, http://dx.doi.org/10.1080/02684527.2012.749062

intelligence and the intelligence process. For those who see it as a more pervasive phenomenon it can also take more subtle forms. The argument presented here is aligned with those who see politicization as

a reoccurring if not normal feature in intelligence analysis; so much so that we need to develop a framework for distinguishing different forms of politicization that goes beyond the mechanism used and focuses on fundamental characteristics.1 The proposal put forward here is that the concepts of soft and hard politicization drawn from the imagery of hard and soft power along with the context in which this politicization takes place provide the basis for such a framework.

Intelligence Estimating

Three distinct focal points are routinely used for organizing discussions of intelligence analysis. The first is the concept of an intelligence cycle.2 In it intelligence analysis is broken down into a series of functional stages (requirement setting, information collection, evaluation, and reporting) each leading logically to the next. The second is the debate over whether intelligence analysis is a craft or science.3 Is it a skill honed through an apprenticeship and life-long study or a body of knowledge or universal truths arrived at through a structured investigation under controlled circumstances? The third framing device used to organize discussions of intelligence analysis is the long-standing debate over the proper relationship between analysts who produce intelligence and consumers who request it and use it.4 One position holds that the two must be kept separate so that analysis does not become corrupted by the intrusion of viewpoints and considerations that do not bear directly on the task of understanding and assessing a situation or problem. The opposing position holds that rather than promote objectivity an arm’s length relationship between analyst and consumer only encourages irrelevance. Analysts must know the concerns, priorities and blind spots of consumers if they are to produce intelligence that is germane to the decisions being made and courses of action contemplated. Individually or as a group these approaches do not provide an adequate

foundation for understanding the dynamics of intelligence analysis.Missing is any systematic attention to the variety of ways in which the exercise of power takes place in intelligence analysis. This is especially true for the first two which black-box intelligence analysis and treat it as a self-contained unit of analysis whose operation can be examined with little or no reference to

1For another formulation of a politicization framework see Gregory Treverton, ‘Intelligence Analysis: Between “Politicization” and Irrelevance’ in Roger George and James Bruce (eds.) Analyzing Intelligence (Washington, DC: Georgetown University Press 2008) pp. 91–106. 2For a discussion of the intelligence cycle see Arthur Hulnick, ‘What’s Wrong with the Intelligence Cycle’, Intelligence and National Security 21/6 (2006) pp.959–79. 3Allen Dulles, The Craft of Intelligence (New York: Harper & Row 1963). 4The classic account of this starting position is Sherman Kent, Strategic Intelligence for American World Policy (Princeton, NJ: Princeton University Press 1949).

Intelligence and National Security6

external influences. In doing so they hearken back to the early literature on civil–military relations which posited separate areas of responsibility and expertise for military professionals and civilian policymakers and engaged in lengthy discussions about the meaning and boundaries of military professionalism as well as the classical literature on diplomacy which treats it as a neutral area of expertise whose practitioners stand apart from society. The third focal point recognizes that intelligence analysis operates within a

political context but still falls short of providing a framework for understanding the dynamic interaction between the two. The arm’s length perspective is in many respects similar to the first two approaches in that it seeks to wall off intelligence analysis from the broader political system. The co-mingling of the two spheres results in the politicization of intelligence which undermines the analytic process. While this may be the result it need not be. As a strategy for bringing about change politicization is value neutral and may be used by reformers as well as obstructionists. By itself the argument for establishing a close relationship between analysts and consumers is an inadequate starting point because it fails to effectively incorporate into its reasoning the many motivationsaconsumermayhave inestablishingaclose relationship. Improving thequality and relevanceof intelligence is one goal but presidents have also been motivated to establish closer relations with intelligence by paranoia, distrust, anger and personal political and policy agendas they seek to advance.

Intelligence Analysis and Politics

A more direct and systematic attention to the interaction between politics and intelligence analysis is anecessary supplement to the existing frameworks for at least four reasons. The first involves the changing nature of the policymaking environment in which intelligence analysis operates. Two trends, one long developing and the othermore recent, have brought intelligence and politics into closer and more continuous contact. The long-developing trend is the end of bipartisanship as a sustaining and guiding concept in the conduct of national security policy.5 Calls for ‘rallying around the flag’ increasingly go unheeded as foreign and defense policy problems are now debated and argued over with the same vehemence and duration as are domestic policy problems. The more recent trend is the onset of what some refer to as ‘the crisis after

the crisis’. Central to this period are the arrival of blame games. The crisis after the crisis is the period when the original set of events that created a crisis has passed but its political and policy repercussions continue.6 For the US intelligence community this crisis was 9/11. From a foreign policy perspective 9/11 created a security crisis in which the nature of the threats facing the United States appeared to change dramatically. From a political perspective 9/11

5Harry Howe Ransom, ‘The Politicization of Intelligence’ in Stephen Cimbala (ed.) Intelligence and Intelligence Policy in a Democratic Society (Dobbs Ferry, NY: Transnational Publishers 1987) pp.25–46. 6Arjen Boin, Paul ‘t Hart, Eric Stern and Bengt Sundelius, The Politics of Crisis Management (Cambridge: Cambridge University Press 2005) p.100.

The Politics of Intelligence and the Politicization of Intelligence 7

called into question the capabilities and policies not only of the Bush administration but the US national security establishment as a whole. Crises are signals that the modes of thought and action that we have been relying upon to control the world and organize our daily lives have failed.7 Crisis management from this second perspective is not about rationally responding to external threats but dealing with fear, perceptions of risk, and reestablishing core beliefs within the political system. Assigning blame becomes a central tool of crisis management. According to Mark Bovens and Paul ‘t Hart, ‘political scapegoating . . . is a way of reestablishing these deeply held convictions (myths) about politics as the realm of rational action, control, and policy choice’.8 In this view laying blame and scapegoating are rational responses to a crisis because they have a potentially cathartic and healing effect.9

Intelligence is particularly vulnerable to the assignment of blame on several fronts.10 First, as we noted above, the popular conception of intelligence analysis sees it as little more than connecting the dots. The second handicap confronting intelligence in the blame game over intelligence failures is that evidence that policymakers were alerted to a problem is easily finessed by policymakers with the rebuttal that while the intelligence on the subject may have been presented it was not convincing. Any counterargument on the part of intelligence can easily be undermined by pointing to areas of doubt and disagreement among analysts inside and outside of the intelligence community over themeaning and significance of data, a problem inadvertently compounded by interest in competitive analysis. A second overarching reason for a greater systematic attention to the

interaction of politics and intelligence analysis lies in the subjective nature of professional knowledge. Viewed from the outside professional standards are seen as objective and neutral but as Thomas Kuhn and others have argued a different picture lies beneath the surface. Professional bodies of knowledge are supported by a complex system in which those whowould enter the system be it science, law, medicine, or the military learn from and are mentored by those already in the profession and in the process develop a commitment to its fundamental tenets. These guiding assumptions and definitions of best practices emerge from a period of competition with other schools of thought which gradually fall by the wayside. Once accepted as normal, the triumphant perspective then serves as the basis for determiningwhat questions to ask, how to assess the validity and reliability of information and how to draw inferences

7Barry Turner and Nick Pidgeon, Man-Made Disasters (London: Butterworth Heinemann 1997) pp.4–5. 8Mark Bovens and Paul ‘t Hart, Understanding Policy Fiascos (New Brunswick, NJ: Transaction 1998) pp.138–9. 9Arjen Boin, Allan McConnell and Paul ‘t Hart, ‘Governing After Crisis’ in Arjen Boin, Allan McConnell, and Paul ‘t Hart (eds.)Governing After Crisis (Cambridge: Cambridge University Press 2008) p.9; and Boin, ‘t Hart, Stern and Sundelius, The Politics of Crisis Management, p.18. 10See Paul Pillar, ‘Great Expectations’, Harvard International Review 27/4 (2006) pp.16–21.

Intelligence and National Security8

from it. In doing so some questions, assertions, and conclusions lie beyond the scope of this normal pattern of inquiry and go unexamined. Change when it occurs is not linear but the product of periodic challenges

to the dominant professional ethos that emerge when a critical mass in the number of anomalies occur for which no satisfactory accounting can be given by it. These challenges are met with resistance ultimately yielding either the reassertion of the dominant paradigms position, compromise or polarization and the replacement of one paradigm with another. For Kuhn, this act of replacement did not necessarily mark an increase in knowledge or insight. It was only a different way of looking at a phenomenon.11

While rejecting the view that intelligence analysis is at base an engineering exercise in which information is collected and can be brought together as if one were connecting dots, there is often not a sufficient appreciation for the extent to which the organizational structure in which it is presented and the professional knowledge and standards used in intelligence analysis fit comfortably within this framework. Once adopted, the method of analysis, the definition of the situation or a judgment about an intelligence problem can quickly become enshrined in organizational routines and patterns of thinking that constrain analysis and produce blinders much as does a connect the dots approach to analysis. Director of Central Intelligence (DCI) Admiral Roscoe Hillenkoetter established the format for interagency coordination of National Intelligence Estimates in 1948 after the Central Intelligence Agency (CIA) had been in existence for only one year.12 One scientist commented that after a 1945 estimate that the Soviet Union could get a bomb in five years, ‘every year that went by you kept saying five years’.13 More broadly a similar type of thinking came to characterize US thinking on Iran. As Gary Sick observes for all too many US officials ‘the Shah was Iran and Iran was the Shah’.14

Third, politics and intelligence analysis are more closely linked than standard accounts would suggest due to the great variability in the nature of intelligence questions and the uses of intelligence. Not all intelligence questions are created equal. Some appear in a regular schedule. Others emerge only periodically. Still others are new. Some intelligence questions permit time for reflection and the gathering of new information. Others are presented on a short timeframe or in a crisis where the analytic process occurs at an accelerated speed. On occasion questions may be so routine and non- controversial that they entail little more than exercises in connecting the dots.

11Thomas Kuhn, The Structure of Scientific Revolutions (Chicago, IL: University of Chicago Press 1962) and Stephen Cole,Making Science: Between Nature and Society (Cambridge, MA: Harvard University Press 1992). 12See John Ranelagh, The Agency: The Rise and Decline of the CIA (New York: Simon & Schuster 1986) p.174. 13Lawrence Freedman, U.S. Intelligence and the Soviet Strategic Threat (Princeton, NJ: Princeton University Press 1977) p.64. 14Gary Sick, All Fall Down: America’s Fateful Encounter with Iran (New York: Random House 1985) p.31.

The Politics of Intelligence and the Politicization of Intelligence 9

Other times they are highly controversial and touch directly on highly visible policy decisions. Similar variability exists in the uses of intelligence analysis. Intelligence

may alert or sensitize a policymaker to a potential problem; it may provide insight into how to evaluate a situation; it may provide assistance in selecting a policy option; or it may be useful to a policymaker in his or her effort to sell a policy to others. On other occasions when policymakers have made up their minds on what to do it may be of little influence. Intelligence is not unique in this regard. BobWoodward notes in The Commanders that the Pentagon was the center of decision-making in the months before the George H.W. Bush administration’s decision to invade Panama, a time where the adminis- tration’s attention was elsewhere. In the case of the Persian Gulf War, the White House paid attention to little else leading Woodward to note: ‘when the President and his advisers are engaged, they run the show’.15

Finally, it needs to be noted that politicizing an issue is not by definition corrupting it. It is value neutral strategy. It says nothing about the agenda of those employing it. Those seeking change may politicize an issue as a means of bringing a fresh look at a problem, breaking down ingrained ways of thinking, and stimulating the consideration of alternative explanations and contingencies. Defenders of the status quo may politicize an issue as a means of protecting existing policies or values from being undermined by new policy initiatives or opposing groups.

Varieties of Politicization

At its core politicization is a strategy used by participants in the policy process to advance their goals and thwart the efforts of others to achieve competing goals. And, as with all political strategies it is a matter of mobilizing power resources. Borrowing from the literature on US foreign policy and world politics we can identify two different politicization strategies: hard politicization and soft politicization.16

As described in this literature soft power involves the use of non-coercive means to achieve one’s objectives. Foremost among them are the value base on which decisions are made, the rules by which they are made and the institutional system wherein the decision is made. Accordingly, soft politicization of intelligence analysis involves deliberate attempts to alter the assumptions underlying an analysis, the decision rules by which an analysis moves forward, and the institutional setting within which these deliberations occur.Hard power involves the use of coercion to eliminate options and if need be to impose an outcome. From this vantage point the hard politicization of intelligence analysis involves deliberate attempts to coerce analysts into adopting a certain set of assumptions or conclusion or in the extreme overruling analysts and imposing a conclusion on the analysis.

15Bob Woodward, The Commanders (New York: Simon & Schuster 1991) p.33. 16Joseph Nye, Soft Power: The Means to Success in World Politics (New York: Public Affairs Press 2004).

Intelligence and National Security10

Attention to whether the politicization of intelligence analysis takes a hard or soft form is one key dimension to understanding the politics of intelligence. A second important factor is the setting within which the politicization of intelligence takes place. The larger the number of participants involved in making the decision the greater the potential number of views present and the greater the number of alliances possible. The more open the decision-making process the greater the number of strategies available to participants. Following this logic and consistent with the broader literature on policymaking within the United States we can identify three principal decision-making settings at the national level. The first is a bureaucratic setting. Although no bureaucracy is totally immune from the influence of outside forces some are more impenetrable than others. The intelligence community was long among the most closed decision-making systems and still is in many respects. The more closed the decision-making system the less visible are efforts at politicizing intelligence analysis. A second setting might be best described as an elite system. Here, participation in intelligence analysis extends beyond the intelligence community bureaucracy to includekey congressional figures alongwith representatives from themedia, think tanks and advocacy groups. The final setting finds the mass public actively involved in the politicizationof intelligence analysis.Where in the elite system the mass public is present more as a spectator than active participant, they are now actively involved as ‘referees’ whose support must be gained for the analysis to be accepted as truthful and accurate. Politicization here becomes increasingly public thereby adding another dimension, the publicization of intelligence, to the dynamics of intelligence politicizing.17

A Framework for Studying the Politicization of Intelligence Analysis

Taken together the these two dimensions, the nature of politicization and the context within which it occurs, provides us with a framework for better understanding the politics of intelligence.

Closed System

(1) Consensus politics: This is the normal context within which intelligence analysis operates. At times a single definition or view of the analytic task exists; at other times disagreement exists within or between intelligence organizations as to the validity of the information before them and its meaning. In either case the conflicts are limited and occur within the bounds of ‘best practices’. Established practices exist and are used for registering disagreements. The analytic process occurs within a bureaucratic setting that is largely shut off from external forces other

17GlennHastedt, ‘The NewContext of Intelligence Estimating: Politicization or Publicizing’ in Stephen Cimbala (ed.) Intelligence and Intelligence Policy in a Democratic Society (Dobbs Ferry, NY: Transnational Publishers 1987) pp.47–68.

The Politics of Intelligence and the Politicization of Intelligence 11

than the president and his key advisors who act as customers setting the intelligence requirements and receiving the intelligence products.

(2) Soft politicization: At this point consensus politics gives way to efforts to control the terms of the analysis either through controlling guiding assumptions, changing the participants through exclusion or inclusion, alteringdecision rules, or changing the institutional settingwithinwhich the analysis is carried out and approved. Disagreements now incorporate the added presence of competing bureaucratic interests and policy priorities. It is now possible to speak of winning and losing in producing analysis.

(3) Hard politicization: Efforts to control the analytic process and outcome now move beyond changing participants, rules or institutions to undertaking more direct actions to assure the desired outcome. Intimidation is one strategy used when that fails participants may arbitrarily reject or ignore an analysis, substitute their own for one they receive, or pick among competing analyses. The key here is that this rejection, substitution or selection is not based somuch on the accumulated evidence or professional expertise as it is on considerations external to the analytic process. Evidence that this is taking place can be found in the rejection or disregard for intelligence information and analysis without explanation or using arguments known to be without merit and/or references to such factors as public opinion, personal prestige, image, party politics, and intuitional factors as key reasons for moving forward in a particular direction.

Elite System

(4) Consensus politics: Intelligence analysis is no longer the sole province of professionals within the intelligence community nor is intelligence purely an executive function. Increasingly members of Congress and representatives from think tanks and the media become important figures in the production, dissemination and evaluation of intelligence through holding hearings, making public statements and framing news stories. At this stage one can still speak of bipartisanship existing among elites so that these external groups serve more to affirm existing views and analytical positions than challenge them.

(5) Soft politicization: Here again atmosphere of consensus has given way to zero sum contest in which winners and losers are clearly identifiable. As with all cases of soft politicization victory is achieved not by imposing an outcome but by shaping the ideas, voices, rules, and settings within which the analysis is taking place. In making disputes more visible the elite system setting of the intelligence process also heightens the stakes by more closely identifying intelligence with policy preferences creating a second zero sum game where policymakers, interest groups, and political parties win and lose.

(6) Hard politicization: Intelligence analysis here is a struggle of wills in which one side seeks to impose an outcome on others. Unlike in a system

Intelligence and National Security12

of consensus politics in which the analytic contest is largely among carried out among equals with only one consumer of intelligence, the involvement of decision-makers from outside the Executive Branch and the heightened visibility attached to the analytic process makes imposing an outcome a competitive process that plays out over time rather than a singular decision-making act.

Public System

(7) Consensus politics: Such a period tends to occur in the aftermath of a foreign policy crisis which finds the American public having ‘rallied- around-the-flag’ in support of a given world view or solution to a problem that is also advanced by intelligence professionals and policymakers. It is generally a brief period because the three groups involved (intelligence professionals, elites and mass public) tend not to share the same depth of commitment or understanding of a problem, making it difficult to maintain this consensus. This is likely to be especially true for the mass public which historically has been minimally knowledgeable about the specifics of national security issues and tends to gravitate to and away from policy problems with great frequency.

(8) Soft politicization: Efforts to steer analysis by the adoption or rejection of certain assumption, expand or contract the number of participants in the process or alter the rules and institutional setting within which the analysis will be carried out are now played out in a quite visible and public fashion. Competing policies more than competing intelligence analysis are likely to drive the contest with the outcome itself still in the future. In this contest the mass public moves from being a passive audience over intelligence analysis to an active participant in determining how it will proceed.

(9) Hard politicization: The involvement of the mass public in judging the adequacy and appropriateness of intelligence analysis moves the public from being a participant in the analytic process to being the judge of the merits of a line of analysis. Policies and the intelligence on which they are based are increasingly inseparable. Selling a policy to a skeptical public thus also means selling the analysis on which it is based. The strategies adopted in this public relations/campaign atmosphere are designed to create intimidating pressure to bear on participants and decision-makers to adopt a given position.

Case Studies in the Politicization of Intelligence Analysis

In this section a series of intelligence analytic efforts are examined that are frequently identified as cases where the politicization of intelligence occurred. The purpose is not to address the accuracy or inaccuracy of those analytical efforts but to highlight the variety of ways in which politicization occurred, a

The Politics of Intelligence and the Politicization of Intelligence 13

variety that is obscured by overly generic references to politicization and one that hinders efforts to respond to it.

Bomber Gap

Long-range bombers were the original delivery system of the nuclear age. The primary Soviet heavy bomber for this purpose was the Bison which made its debut at the 1954MayDay Parade and in the process set inmotion the ‘bomber gap’. This case demonstrates consensus politics in a closed setting as the intelligence estimating process produced disagreements which were dealt with in the established pattern of inserting dissents into intelligence products. Calculating backward using theUS production strategy as its reference point,

US Air Force intelligence pegged 1952 as the starting point for work on a prototype and identified 1956 as the earliest point at which full production would take place and 1960 when a bomber force would come into existence. The Bison’s reappearance in greater than anticipated numbers in 1955 led to a more pessimistic conclusion on the part of US Air Force intelligence with the Pentagonputtingout apress release asserting that ‘these observations establish a new basis for our estimate of Soviet production’.18 Production was now estimated to have begun in 1954 and a force of 60 Bison was held to exist. The projected growth in Soviet bomber strength now was judged to be superior to the anticipatedability of the StrategicAirCommand todefend theUnitedStates. Yet anotherflyover aspart of anAviationDayParadeon13July 1955 reinforced the sense that in thewords of DCIAllen Dulles ‘every indication pointed to [the Soviets] having adopted [the Bison] as a major element of their offense strength and to an intention to produce these planes more or less as fast as they could’.19

Produced in secret, the predictions of a rapid and strategically significant increase in the number of Soviet bombers were leaked to the press. Briefings to Congress soon followed as did a February 1956 decision by the Senate Armed Services Committee under the chairmanship of Stuart Symington who had been Secretary of the US Air Force in the Truman administration to hold hearings on the matter. President Dwight Eisenhower sought but failed to get this hearings cancelled by sending a supplemental budget request to increase production of the US B-52 bomber to Congress. Not all intelligence agencies shared in the assessment that a large and

threatening bomber force was being created but the presentations made paintedadirepicture comparing the anticipated trend inUSandSoviet bomber production, an analysis which the Strategic Air Command (SAC) commander GeneralCurtis LeMay stated theCIAconcurredwith. Inhis testimonyDefense

18John Prados, The Soviet Estimate (Princeton, NJ: Princeton University Press) 1982, p.42. In addition to the sources that follow additional information on the bomber gap can be found at William Lee,Understanding the Soviet Military Threat, National Strategy Information Center, Agenda paper #6 (1977) and Raymond Garthoff, ‘Estimating Soviet Military Intentions and Capabilities’ in Gerald Haines and Robert Leggett (eds.) Watching the Bear (Langley, MD: Center for the Study of Intelligence 2003). 19Prados, The Soviet Estimate, p.43

Intelligence and National Security14

Secretary Charles Wilson disagreed with the Air Force analysis leading Symington to accuse Wilson of ‘unconstitutionally contradicting’ the US Air Force and CIA positions.20 Army and navy officials also presented less pessimistic analysis of the situation and Admiral Arthur Radford, Chair of the Joint Chiefs of Staff, noted in his testimony that ‘we normally overestimate communist capabilities’. An August National Intelligence Estimate (NIE) on the US–Soviet military

balance continued to adopt the Air Force position of a rapidly emerging national security threat of major proportions. Dissent from the army and navy grew and the CIA undertook a new review of the economic implications and foundations of the presumed Soviet buildup. This analysis questioned whether a build up of the Air Force’s estimated proportions could be sustained by the Soviet industrial base. The additionofU-2photographs in1956addedanewelement to the debate.

Its photographsdocumented the existence of fewer bombers than theAir Force hadestimated toexist.The reducedestimateswerepresented toCongress in1957 during budget hearings. A disbelieving Symingtonwrote toDulles protesting the downward revision in Soviet bomber strength now being presented. With NIE 11-4-57 the bomber gap began to fade from public view. It

concluded that while technical difficulties may have been involved the more likely explanation was that ‘Soviet planners have deliberately decided on a relatively modest bomber program’.21 Disagreements within the intelligence community continued with the air force now dissenting through a footnote claiming a higher figure was possible. The Army, Navy and Joint Chiefs of Staff intelligence also inserted dissenting footnotes arguing for lower estimates. In the intelligence analytic process over the bomber gap we see soft elite

politicization as intelligence is leaked to the press and congressional hearings are held in which conflicting views are presented. The public is largely a passive participant in this debate which brings together intelligence and policy.

Missile Gap

In short order the bomber gap was followed by the missile gap. In this intelligence controversy the politicization of intelligence moves from soft – attempting to expand the number of participants in themissile debate – to hard with its involvement in the 1960 presidential race and efforts to sway voters. John Prados observes that from 1954 to 1956 the Air Force considered

estimates of Soviet ballistic missile capabilities to be the province of military intelligence.22 This did not mean that the CIA had ignored the question. An

20Ibid., p.44. 21Ibid., p.48. 22Ibid., p.60. In addition to the sources that follow additional information on the missile gap can be found at Edgar Bottom, The Missile Gap: A Study in the Formation of Military and Political Policy (Rutherford, NJ: Farleigh Dickenson Press 1971); Roy Licklider, ‘The Missile Gap Controversy’, Political Science Quarterly 85/4 (1970) pp.600–15; and Peter Roman, Eisenhower and the Missile Gap (Ithaca, NY: Cornell University Press 1995).

The Politics of Intelligence and the Politicization of Intelligence 15

NIE on Soviet satellite capabilities was produced in 1954 that accurately predicted the 4 October 1957 Sputnik launch and the CIA participated in providing analysis for national estimates on Soviet capabilities. DCI Dulles did not, however, press the issue of greater CIA involvement in these estimates until 1955 after the Air Force had come under criticism from the second Hoover Commission and Congress. The context of intelligence estimating on Soviet missile strength began to

change in other ways. Even before Sputnik newspaper accounts began to refer to intelligence estimates that wrote in alarmist terms about the rapidly growing threat Soviet missiles posed to US security. In February 1956 Joseph Alsop and Steward Alsop wrote that US estimates had consistently underestimated Soviet developments for several years and cited an estimate which predicted a 1958 intercontinental ballistic missile (ICBM) test. In July 1957 Stewart Alsop asserted that ‘convincing evidence’ now existed that the Soviet Union had tested their first experimental ICBM. Prados comments that before Sputnik’s launch few ‘attentive observers’

such as Senator Symington were paying attention to this publicization of intelligence.23 Its launch ‘electrified America’. Stewart Alsop wrote in December that ‘the prospect which immediately confronts us is that the Soviet will achieve this replacement [of manned bombers] before we do. There will be a gap – in the Pentagon it is known simply as The Gap’. Only a few days after Sputnik, Senator Lyndon Johnson, who harbored presidential ambitions, called hearings before his Preparedness Investigating Subcommit- tee of the Armed Services Committee. Among those who testified were scientists Edward Teller and Vannevar Bush who, in turn, stated that the Soviets were close to having an operational ICBM capability and that the United States was behind in the missile race. Retired general James Gavin who had headed the Army’s research and development unit warned that the Soviet Union was now in a position to force the United States to ‘live and conduct . . . international diplomacy under a canopy of fear’. Senator John Kennedy accused the Eisenhower administration of pursuing a policy of economic security at the cost of sacrificing military security.24

In April 1957 the National Security Council (NCS) established an advisory panel, the Security Resources Panel under the chairmanship of H. Rowan Gaither to study the adequacy of US civil defense plans. In the midst of Johnson’s hearings the Gaither Report was presented to Eisenhower and its findings leaked to the press. Prados notes that the Gaither Committee functioned more as a consumer of existing information than a generator of new analysis and that it did not have unrestricted access to intelligence.25 The CIA appears not to have provided access to U-2 photographs while the Joint Chiefs of Staff withheld information on Air Force capabilities.

23Prados, The Soviet Estimate, p.63. 24Gerald Clarfield and William Wiecek, Nuclear America (New York: Harper & Row 1984) pp.165–6. 25Prados, The Soviet Estimate, p.68

Intelligence and National Security16

In its report the Gaither Committee went beyond its initial narrow charge reporting that:

by 1959 the USSR may be able to launch an attack with ICBMs . . . against which SAC will be almost completely vulnerable under present programs . . . the next two years seem critical to us. If will fail to act at once [to improve US missile defense capabilities], the risk, in our opinion, will be considerable.26

Journalist Chalmers Johnson wrote a detailed account of the Gaither Report in The Washington Post in December 1957 prompting Senator Johnson to ask Eisenhower to release the full report. Eisenhower refused, prompting several members of the Gaither Committee to begin publicly lobbying in support of its recommendations. Controversy over the Soviet ICBM estimates continued to be carried out

both in public and within the bureaucracy. Lawrence Freedman describes NIEs on this subject as ‘containing footnote upon footnote’.27 The CIA and the Air Force were the lead participants in the debate although the Army and Navy also were active in it. The Office of the Air Force Chief of Staff for Intelligence was staffed by those who held a missile gap to exist. One CIA official is described as saying ‘to the Air Force every flyspeck on film was a missile’.28 Emblematic of the analytic conflict was disagreement over the meaning of a lull in Soviet missile testing between April 1958 and March 1959. The CIA interpreted it as indicative of performance problems. The Air Force concluded the opposite. Testing had stopped because technological problems had been solved and the Soviets were preparing to go into production. In public Eisenhower and senior figures in the administration took a

generally cautious note while some in the press such as Stewart Alsop and some in the administration such as Air Force Secretary James Douglas and SAC Commander General Thomas Power expressed more alarmist positions. Symington, who like Johnson was interested in the Democratic presidential nomination, also endorsed the notion of a missile gap telling Eisenhower he had ‘other’ intelligence on Russian military capabilities. Later Symington would assert that ‘the intelligence books have been juggled so that the budget books can be balanced’.29

The Eisenhower administration spoke out publicly in defense of its policies but to too little effect since Eisenhower never released the intelligence on which his administration’s urging of restraint and caution were based. Eisenhower addressed the nation in speeches; Dulles explained at a conference how intelligence estimates were arrive at; newspaper reporters were briefed; and officials were sent to testify before Congress. Eisenhower

26Ibid., p.71 27Freedman, U.S. Intelligence and the Soviet Strategic Threat, p.75. 28Ibid., pp.71–2. 29Prados, The Soviet Estimate, p.90.

The Politics of Intelligence and the Politicization of Intelligence 17

also found it necessary to increase defense spending in the face of these pressures. By one count military expenditures had increased by 20 per cent from where they were in 1955–56. John Kennedy was elected president in November 1960. As had his

challengers for the Democratic nomination, he campaigned in part on the theme that a missile gap existed. Upon taking office his administration abandoned this position embracing the conclusion that followed logically from NIE 11-60 which asserted that no operational ICBM sites exited. Secretary of Defense Robert McNamara maintained that the missile gap continued to exist only because of conflicts of opinion within the intelligence community. His observation contained an element of truth. The Air Force and SAC

remained the primary dissenter to increasingly lower estimates of Soviet ICBM strength. The Air Force briefed McNamara using its own calculations rather than those in the NIE and found itself isolated and in opposition to the shared positions of the CIA, State Department, and Army in NIE 11-8/1-61 that posited Soviet ICBM strength at less than 35 as well as the position held by the Navy that less than ten existed. The Air Force made one final presentation to Kennedy that 600–800

ICBMs existed but failed to sway him. There followed a directive from McNamara to the military forbidding dissenting services from making direct appeals to Congress should there budgetary requests be denied.30

Intelligence analysis during the missile gap presents us with a variety of different forms of politicization. Consensus driven politics in a closed setting was evidenced in the frequent use of footnotes in intelligence estimates. We see soft politicization in a closed system as the Air Force responds to criticism from the Hoover Commission by agreeing to the establishment of new intelligence unit. An attempt at soft politicization in an elite setting prior is evident prior to Sputnik in the leakage of intelligence to the press. The limited attention it received at that time is replaced by widespread attention and concern after Sputnik.

Vietnam Estimates

As was the case with the bomber gap, the politics of intelligence on Vietnam presents evidence of both the presence of hard and soft politicization. In many respects the question that the Kennedy administration posed to US national security officials early in its term in office was not could Vietnam be won, but when could US forces be removed? Victory seemed within reach. Secretary of Defense Robert McNamara commented to the press that every quantitative measure showed they were winning. Similarly, President Kennedy spoke confidently in his 1963 State of the Union address that ‘the spear point of aggression has been blunted’. The inherent optimism of the Kennedy administration often captured in

the catch phrase the ‘best and brightest’ was reinforced by the tenor of official

30Ibid., p.119.

Intelligence and National Security18

reporting from Washington through official channels. Harold Ford notes in evaluating this reporting that:

the historical record is replete with instances where supervisors and field commanders, the men charged with demonstrating operational progress in the programs assigned to them, overrode their subordinates’ negative facts and judgments . . . Dissenting junior officers were urged to ‘get on the team’ and on occasion were frozen out or moved out by their superiors.31

Dissent was further stifled by a McNamara decision in early 1963 that only finished intelligence reports should be sent to Washington thereby reducing the potential for contradictory information or weak supporting information to be seen. A key intelligence estimate produced in this overall atmosphere of

optimism and growing Kennedy administration expectation of victory in Vietnam was NIE 53-63. It concluded that: ‘we believe that Communist progress has been blunted and that the situation is improving . . . the Viet Cong can be contained militarily’. Presented in April 1963 this assessment was far different from a draft submitted to the US Intelligence Board in February of that year. It had concluded:

the struggle in South Vietnam at best will be protracted and costly[,] very great weaknesses remain and will be difficult to surmount. Among them are lack of aggressive and firm leadership at all levels of command . . . lack of trust between peasant and soldier . . . a very inadequate intelligence system and obvious Communist penetration of the South Vietnamese military organization.32

DCI JohnMcCone toldMcNamara in the 1962 that ‘he was not optimistic about the success of the whole United States program’. Yet when presented with this pessimistic assessment he sharply criticized those responsible for it for having prepared a document that was at variance with ‘the people who knew Vietnam best’ and directed that their views be incorporated into a revised estimate. When contacted, Military Assistance Command Vietnam’s [MACV] General Paul Harkins wanted a statement included attesting to

31Harold Ford, CIA and Vietnam Policymakers: Three Episodes 1962–1968 (Washington, DC: Center for the Study of Intelligence, Central Intelligence Agency 1998) p.9. In addition to the sources that follow additional information on Vietnam estimates can be found at Estimative Products on Vietnam, 1948–1975 (Washington, DC: National Intelligence Council 2005); Sam Adams, ‘Vietnam Cover-up: Playing with Numbers’,Harpers, May 1975; T.L. Cubbage II, ‘Westemoreland vs. CBS: Was Intelligence Corrupted by Policy Demands?’ Intelligence and National Security 3/3 (1988) pp.118–80; and James Wirtz, Intelligence Failure in War: The American Military and the Tet Offensive (Ithaca, NY: Cornell University Press 1991). 32Ford, CIA and Vietnam Policymakers, p.1.

The Politics of Intelligence and the Politicization of Intelligence 19

South Vietnam’s ‘steady and notable progress in gaining support of the people’. Major General Victor Krulak, McNamara’s Special Assistant for Counterinsurgency and Special Activities, stated South Vietnam’s military capabilities had ‘increased markedly’ and Viet Cong initiated activity was ‘well below 1962 levels’. Army Chief of Staff General Earle Wheeler challenged the assertion of peasant–soldier conflict and wanted the statement regarding poor leadership to be ‘heavily qualified’.33

The NIE’s drafters held fast in their conclusions but were overruled. In April the Board of National Estimates (BNE) released the optimistic assessment that McCone had engineered. In explaining McCone’s abrupt shift from doubter to supporter of Kennedy’s Vietnam policy Ford puts forward a series of factors external to the merits of the estimate under consideration.34 First, an NIE estimate released just prior to the Cuban Missile Crisis that failed to anticipate the placement of Soviet missiles in Cuba.McCone was criticized by the President’s Foreign Intelligence Advisory Board for this estimate. Ford speculates that McCone ‘almost certainly’ had lost some confidence in the BNE. Second,McCone had strong personal ties to Krulak who argued against the analysis and former President Eisenhower who warned of the dangers of a defeat in South Vietnam. Both hard and soft politicization in a closed setting emerged here. Soft

politicization took the form of pressures related to the exhortation to ‘get onboard’ while hard politicization took the form of the requirement to include the Army’s perspective on the estimate and then to make it the basis for the final NIE. President Lyndon Johnsonwas no less committed to victory inVietnam than

was Kennedy and in 1964 his administration began planning for newmilitary operations against the communists. One such initiative came from Walt Rostow who headed the State Department’s Policy Planning Staff. In 1963 he ordered a study done of the potential for series of sanctions of increasing magnitude to affectNorthVietnam’swillingness to continue thewar. A special task force was charged with carrying out the analysis. Members of the task force were told that examining the option of a US withdrawal from Vietnam was not part of their charge. After a brief but intensive study an interagency subcommittee of the Task Force concluded that ‘it is not likely that North Vietnam would (if it could) call off the war in the South even though U.S. actions would in time have serious economic and political impact’.35

This conclusion echoed that held by the CIA’s Office of National Estimates (ONE). Theywere, however, at variancewith administration thinking. Johnson had earlier told McCone that planning for placing added pressure on North Vietnamshouldbe ‘speededup’.Evenbefore thecommittee reported its negative assessment Rostow had told Secretary of State Dean Rusk its finding supported the use of military and other sanctions against North Vietnam.36 This episode

33Ibid., p.14 34Ibid., pp.18–19. 35Ibid., p.50. 36Ibid., p.51

Intelligence and National Security20

illustrates hard politicization in a closed context as the intelligence analysis was overruled because it did not support administration policy. The CIA was again involved in analyzing the potential effectiveness of

bombing the north in April 1964. Rather than a formal intelligence analysis the question was addressed through a war game. The CIA was well represented in the exercise. The fundamental conclusion reached was that the United States was facing a no-win situation with the limited options of expanding the war effort and risking a Chinese Korea-like intervention or de- escalating. Ford, who participated in this exercise, noted that it had no impact on US policy planning.37 In fact, as with the earlier analysis, its only high-level recipient was George Ball who asked why the United States would proceed with planning for bombing of the north in the face of this finding. The November 1964 Viet Cong attack on Bien Hoa air base led to another

assessment of the impact of US air strikes against the north. This one was carried out by a NSC working group led by Assistant Secretary of State William Bundy. The intelligence analysts tasked to this project concluded consistent with the findings of existing NIEs but in opposition to the majority opinion that bombing the north would probably not lead to its abandonment of the Viet Cong. The final report, however, failed to reflect their thinking even as a dissent. Bombing the north did not have the desired effect and the war continued.

The analytic focus now shifted to center on the strength of the Viet Cong forces. MACV had been in charge of producing this estimate. CIA intelligence units including the ONE Estimates, the Office of Research and Reports and the Office of Current Intelligence had only occasionally carried out their own analysis of Viet Cong troop strength. When they did they found MACV estimates to be off the mark and considerably too low. In September 1966 MACV figures placed Viet Cong irregular troop strength at 100,000– 120,000. In January 1967 an ONE estimate placed it at 250,000–300,000 at the end of 1965. A May 1967 MACV Order of Battle estimate placed total irregular enemy strength in South Vietnam at 292,000. The CIA estimate was ‘in the 500,000 range and may even be higher’.38

By 1966 McNamara had come to have serious reservations about MACV estimates and asked the CIA’s Directorate of Intelligence to produce a private assessment for him on the future course of the war. He was presented in August with the negative assessment that ‘planned US measures were not likely to deter the North Vietnamese’. McNamara thereupon in April 1967 asked for periodic CIA estimates of enemy troop strength, the effectiveness of the pacification program, and US operations against North Vietnam.39

Steadily the dispute over MACV’s troop estimates grew as efforts to reconcile the conflicting positions failed to produce an agreement. Finally, in May 1967 DCI Richard Helms tasked the CIA and military intelligence analysts with coming up with an agreed upon Order of Battle for inclusion in

37Ibid., pp.57–8. 38Ibid., p.89. 39Ibid., p.88.

The Politics of Intelligence and the Politicization of Intelligence 21

NIE 14-3-67. With the Defense Intelligence Agency (DIA) supporting MACV’s estimate of less than 300,000 no agreement was reached and Helms dispatched the analysts to South Vietnam for further discussions. There the impasse continued and in many ways deepened with George Carver, Helms’ Special Assistant for Vietnam, cabling to Washington that the undertaking was ‘frustratingly unproductive since MACV stonewalling, obviously under orders’. One CIA analyst summed up the experience of producing Vietnam estimates by noting that ‘some of the issues were so divisive that it was almost senseless to try for an Estimate’.40 The truth of his observation was confirmed later by several participants. Colonel Gaines Hawkins, chief of MACV’s Order of Battle section, stated: ‘our hands are tied’. ‘We were under orders not to yield on their “commander’s estimate” of fewer than 300,000 men’. Colonel Daniel Graham later stated ‘of course’ he felt MACV’s figures were wrong but defended them because it was ‘the commander’s position’.41

The impasse was broken when Carver accepted MACV’s estimate.42 In his analysis of these negotiations Ford puts forward several hypotheses to explain both MACV intransigence and Carver’s sudden embrace of its position. General William Westmoreland and MACV were trapped by its own past estimates, administration policy, and public expectations. Having long maintained that the Viet Cong were suffering great losses any sudden increase in numbers would be politically ruinous for the administration and the careers of many in and out of the military. Carver’s about-face is attributed both to some degree to the existence disagreements within the CIA about the validity of the high range estimates of 500,000 but mostly due to the political reality expressed by Helms: ‘we had to come up with agreed figures . . . we had to get this O/B question off the board and that it didn’t mean a damn what particular figures were agreed upon’.43 As Ford summarizes it, ‘intelligence was outgunned’.44

Yet, the political logic of MACV’s calculations suggested the numbers did matter and the final estimate reflected this point. NIE 14-3-67 placed Viet Cong regular troop strength at 118,000 and the number of its guerrilla forces at 70,000–90,000 producing at total strength of 208,000. This figure was below the 249,000 agreed upon by Carver. Moreover, it was almost 100,000 below the estimate from the previous year. The logical conclusion to draw from these figures was the United States was winning the war. NIE 14-2-67 was published in November 1967. The Tet offensive began in January 1968. Here again we see hard and soft politicization in a closed setting taking

place. Soft politicization occurred in the form of McNamara’s request to the CIA to provide him with a private assessment of the situation, a request that effectively altered the structure of the decision-making situation by providing an additional outlet for the CIA’s views to enter the analytic process. Hard

40Quoted in Ranelagh, The Agency, p.463. 41Ford, CIA and Vietnam Policymakers, p.94. 42Ibid. 43Ibid., p.99. 44Ibid., p.95.

Intelligence and National Security22

politicization took the form of the Army’s refusal to move from the 300,000 troop strength number and then to actually have the Viet Cong’s troop strength reduced in the final estimate.

B Team

The end of the missile gap did not bring an end to the controversy over the US–Soviet strategic balance. It was soon followed by the anti-ballistic missile (ABM) gap and a more general controversy over what some saw as long- standing CIA underestimation of Soviet nuclear capabilities. The B Team episode shows several forms of politicization to have been present. The very creation of the B Team was an act of soft politicization in a closed setting as it tried to change the composition of those involved in the analysis and in the process bring a different set of values to bear on the problem. Hard politicization in a closed setting is an apt characterization of Prados’ description of B Team–CIA meetings on the Soviet military threat. The ‘October surprise’ leaking of its findings is an example of a failed hard politicization strategy in a public setting. This charge that the CIA had underestimated Soviet nuclear capabilities

was endorsed by conservative members of the President’s Foreign Intelligence Advisory Board (PFIAB) who included William Casey, Clare Booth Luce and Edward Teller. In 1975 they suggested that a second competing analysis be undertaken of Soviet missile strength. DCI William Colby deflected their requests but their proposal was accepted by DCI George H.W. Bush in May 1976, some five months after assuming that position and approved by President Gerald Ford.45

Ford was in the midst of a hard fought campaign with Ronald Reagan for the Republican Party nomination for president. Conservatives had been emboldened by Reagan’s strong showing in Republican primaries and began to press Ford on national security matters. A key dimension to this pressure was the establishment of the Committee on the Present Danger. Its media campaign included articles, speeches, and giving testimony before Congress on the growing Soviet threat. The net effect of this pressure was to move Ford from a supporter of the Nixon–Kissinger foreign policy of détente to one who no longer wished to be associated with it. With input from the PFIAB three B teams were established to look at

various aspects of the strategic balance. Key members were Richard Pipes, William R. Van Cleve, Daniel Graham and Thomas W. Wolfe. While their backgrounds were varied (historian, political scientist, head of the Defense Intelligence Agency, and Air Force intelligence officer respectively) they all

45Anne Hessing Cahn, ‘Team B: Trillion Dollar Experiment’, The Bulletin of the Atomic Scientists, April 1993, pp.24–5. In addition to the sources that follow additional information on Team B can be found at Lawrence Freedman, ‘The CIA and the Soviet Threat: The Politicization of Estimates, 1966–1977’, Intelligence and National Security 12/1 (1997) pp.122–42 and Robert Reich, ‘Re-examining the Team A-Team B Exercise’, International Journal of Intelligence and Counterintelligence 3/3 (1989) pp.387–403.

The Politics of Intelligence and the Politicization of Intelligence 23

shared a conservative outlook on world affairs which had an aggressive and militarily superior Soviet Union as its centerpiece. Others who shared this view served as advisors to the B Team. Foremost among themwas Paul Nitze. The B Teammet with and exchanged views with CIA officials in October, a

meeting that was almost immediately leaked to the press with the B Team’s views prominently being mentioned in what some described as an attempt at an ‘October surprise’ to thwart Jimmy Carter’s election as president.46 Prados describes meetings between CIA analysts and the B Team as ‘bloody’ and a ‘back alley cat fight’ with the B Team taking the exercise as a ‘mandate to deconstruct CIA analyses, even criticizing estimates cast decades earlier’.47

Among the conclusions reached by the B Team were that ‘past NIEs were filled with unsupported and questionable judgments about what it is the Soviet government wants and intends’.48 It judged Soviet political and military doctrine to be offensive in nature and rooted in a science of conquest. Looking to specific weapons systems it held that the Backfire bomber will ‘probably’ be produced in substantial numbers and the ‘Soviets have mounted ABM efforts . . . of a magnitude that is difficult to overestimate’. The B Team report also took exception to the established view that economic difficulties in the Soviet Union would constrain weapons production. Nine days before the election the Committee on the Present Danger

released a report arguing the Soviet Union’s ‘unparalleled military buildup’ presented ‘the principal threat to our nation, world peace and the cause of human freedom’. Pipes pushed for publication of the B Team’s report in early December but this bid was blocked by the CIA. A little less than one month before Carter’s inauguration, on 26 December 1976, The New York Times carried an extensive article on the B Team. Three congressional committees quickly proposed hearings.49

Prados finds that in spite of the media frenzy created by the B Team’s assertions the ‘report reveals nothing stunning at all’, concluding that the methodological criticisms presented in terms of data analysis and conclusion drawing were concerns commonly voiced in the process of producing estimates.50 Similarly the substantive issues raised were long-standing ones which analysis had often split analysts. Still, the B Team’s analysis did have its desired effect. The 1976 NIE, while not adopting the B Team’s analysis in its entirety, did present a sufficiently more stringent view of Soviet behavior and capabilities that the State Department and CIA both inserted dissenting footnotes. Cast more broadly, the B Team exercise is seen as having created serious impediments for Carter’s arms control policies and laid the foundation for the Reagan administration’s foreign policy.

46Cahn, ‘Team B: Trillion Dollar Experiment’, p.24 47John Prados, ‘Team B: Trillion Dollar Experiment’, The Bulletin of the Atomic Scientists, April 1993, p.28. 48Ibid. 49Cahn, ‘Team B: Trillion Dollar Experiment’, p.25 50Prados, ‘Team B: Trillion Dollar Experiment’, pp.28–31.

Intelligence and National Security24

Central America

The orchestrated release of intelligence on El Salvador by the Reagan administration in an effort to garner broad support from the American people for its policy along with the counter leaking of contradictory information is an example of hard politicization in a public context. What mattered in this contest was not which intelligence analysis was correct but which intelligence analysis and policy the public would support. During his presidential campaign and upon taking office Ronald Reagan

sought to highlight the extent to which US national security was still threatened by Soviet aggression both directly through its military arsenal which created a window of vulnerability or indirectly through the actions of surrogates. El Salvador was cited as a textbook case of indirect aggression by communist powers. Going public with intelligence was a key element in his strategy of convincing those who doubted his assertions and redirecting the thrust of American foreign policy. Central to this effort was a State Department White Paper ‘Communist Interference in El Salvador’.51

References to the White Paper began to appear in the press in early February 1981 when it was reported that documents considered authoritative indicated that the Soviet Union and Cuba had agreed in late 1980 to deliver tons of US weapons left behind in Vietnam to Marxist guerrillas in El Salvador.52 Later that month, the State Department made public a summary of its report. This was soon followed by an article in The New York Times reporting the existence of a 29-page dissent drawn up by the NSC, State Department, Defense Department and CIA. In June another series of articles appeared criticizing the White Paper for questionable translations and misidentifying the authorship of various documents. Reagan administration officials acknowledged that the report had

‘overreached’ the evidence in blaming the Soviet Union for giving support to the guerrillas but maintained that intelligence gathered from electronic and human sources supported its main conclusion: Cuba and Nicaragua were aiding the guerrillas. This rebuttal was weakened when a State Department official charged that Secretary of State Alexander Haig had pushed the Latin American bureau to produce the report. With the public remaining unconvinced a second intelligence offensive

began in which the administration released classified intelligence, a move

51‘Communist Interference in El Salvador U.S. Department of State, Bureau of Public Affairs, Office of Communication’, Special Report #80, 21 February 1981. In addition to the sources that follow additional information on the Reagan administration’s policy to Central America can be found at Laurence Whitehead, ‘Explaining Washington’s Central American Policies’, Journal of Latin American Studies 15 (1983) pp.321–63; James Scott, ‘Interbranch Rivalry and the Reagan Doctrine in Latin America’, Political Science Quarterly 112/2 (1997) pp.237– 60; and Anne Karalekas, ‘Intelligence Oversight: Has Anything Changed?’ Washington Quarterly 6/3 (1983) pp.22–31. 52This case study is adapted from Glenn Hastedt, ‘The New Context of Intelligence Estimating: Politicization or Publicizing? in Stephen Cimbala (ed.) Intelligence and Intelligence Policy in a Democratic Society (Dobbs Ferry, NY: Transaction Publishers 1987) pp.47–68.

The Politics of Intelligence and the Politicization of Intelligence 25

which Reagan acknowledged presented dangers to US intelligence sources. It began on 2March when Haig told the House Foreign Affairs Committee that evidence of foreign control over the El Salvadoran rebels was ‘overwhelming and unrefutable’ although he declined to identify the sources of this intelligence.53 The same day Senator Barry Goldwater spoke out in support of Haig’s contention citing a briefing by DCI William Casey to the Senate Intelligence Committee documenting Nicaraguan actions that aided the rebels. The administration followed by releasing intelligence photos showing among other things new military base construction following the well- established Cuban model, Soviet tanks and personnel carriers, and four airfields. Haig then announced that a Nicaraguan rebel leader had been captured in El Salvador further establishing the validity of the Nicaraguan connection. A second rebel leader was then identified. Plans to release electronic intercepts to further bolster their argument were used in a closed- door briefing for key officials but not made public due to CIA objections. Again the desired effect was not achieved. The photos of a Nicaraguan

military buildup could not prove intent allowing them to maintain the moves were defensive and in some cases had been recommended by the United States when Somoza was in power. The first captured Nicaraguan turned out to be a student on his way home from Mexico and the second recanted his story.54

Iraq War

Central to the George W. Bush administration’s strategy for building public support for the Iraq War was the public use of intelligence. The American public needed persuading and intelligence was the key resource at its disposal. It thus constituted a case of hard politicization in a public setting. The politicization of intelligence was also evident in the closed bureaucratic setting. Soft politicization took the form of creating a new intelligence unit in the Pentagon and the absence of an identifiable point in the decision process for dissent to be heard and registered. Hard politicization took the form of cherry-picking intelligence that was then presented in public where it became part of a broader strategy of politicizing intelligence. The George W. Bush administration first went public with intelligence in

the lead-up to the Iraq War in late summer 2002 with a speech by Vice President Dick Cheney.55 Speaking to the National Convention of the Veterans of Foreign Wars on 26 August he asserted:

53Hastedt, ‘The New Context of Intelligence Estimating’, p.62. 54Ibid., p.63. 55This case study is adapter from Glenn Hastedt, ‘Public Intelligence: Leaks as Policy Instruments’, Intelligence and National Security 20/3 (2005) pp.419–39. In addition to the sources that follow additional information on the Iraq case can be found at Robert Jervis,Why Intelligence Fails (Ithaca, NY: Cornell University Press 2010); Richard Betts, Enemies of Intelligence (New York: Columbia University Press 2007); and Chaim Kaufmann, ‘Threat Inflation and the Failure of the Marketplace of Ideas: The Selling of the Iraq War’, International Security 29/1 (2004) pp.5–48.

Intelligence and National Security26

there is no doubt that SaddamHussein has weapons of mass destruction . . . we know Saddam has resumed his effort to acquire nuclear weapons. Among other sources we’ve gotten this first hand testimony from defectors, including Saddam’s own son-in-law who was subsequently murdered at Saddam’s direction.56

In actuality, Iraq’s uranium enrichment program had not been restarted after the 1991 Persian Gulf War. Also that month Andrew Card, Bush’s White House Chief of Staff, set up

the White House Iraq Group to coordinate the public pronouncements made by the White House on Iraq. Among its initiatives was bringing tougher intelligence reports and press clippings on Iraq for a White Paper which asserted Iraq was seeking to obtain uranium oxide from Africa, satellite photos showed signs of an accelerating nuclear program, and Saddam had launched a crash drive for nuclear weapons. The White Paper was never published and the above assertions proved inaccurate but they formed the core of later public pronouncements made by Bush in September and by a second White Paper released by the administration in early October. It contained maps and photographs purporting to show facilities under construction dedicated to Iraq’s nuclear program. Absent in the public version of the White Paper were doubts raised by the intelligence community about the quality of the intelligence at hand and true nature of the Iraqi nuclear program.57 A 2004 post-invasion study led by Charles Duelfer concluded that a concern for Iran and not the United States was the primary motivating factor in Saddam Hussein’s pursuit of nuclear weapons.58 Where the administration’s public comments stressed their possible use in international conflicts others suggested that Iraq’s pursuit of nuclear weapons might be better interpreted in the context of Saddam Hussein’s ongoing efforts to security his power within Iraq. Some of these themes were central to Bush’s State of the Union Address of 23 January 2003. In it he stated: ‘the British government has learned that Saddam Hussein recently sought significant quantities of uranium from Africa’.59

Reports to this effect had been received by US intelligence agencies but had not received widespread acceptance. They were, however, stovepiped to high-ranking administration officials including Cheney who engaged in a ‘year-long tug-of-war’ with the agency over the validity of the intelligence. Secretary of State Colin Powell voiced his public support for the

56‘Vice President Speaks at VFW 103rd National Convention’, http://georgewbush- whitehouse.archives.gov/news/releases/2002/08/20020826.html (accessed 6 December 2012). 57See National Security Archive, ‘CIAWhites Out Controversial Estimate on Iraq Weapons’. In particular see the Senate Intelligence Committee Report on Prewar Intelligence, Section X. 58Central Intelligence Agency, ‘Comprehensive Report of the Special Advisor to the DCI on Iraq’s WMD’, ,https://www.cia.gov/library/reports/general-reports-1/iraq_wmd_2004/ index.html. (accessed 8 December 2010) 59‘State of the Union Address’, 2003 ,http://whitehouse.georgewbush.org/news/2003/012803- SOTU.asp. (accessed 5 December 2010).

The Politics of Intelligence and the Politicization of Intelligence 27

interpretation that Iraq was seeking nuclear material. Speaking to the House International Affairs Committee he stated: ‘with respect to the nuclear program, there is no doubt that the Iraqis are pursuing it’.60 The CIA took a more cautious position in a declassified report sent to Congress, observing: ‘Baghdad may be attempting to acquire materials that could aid in reconstructing its nuclear program’.61 DCI George Tenet had managed to get a reference to Iraqi efforts to acquire nuclear material from Africa removed from a speech Bush gave in Cincinnati in September 2002. The CIA objected again but to no avail. In his Cincinnati speech Bush made his point about Iraqi intentions using other intelligence, stating: ‘satellite photographs reveal that Iraq is rebuilding facilities at sites that have been part of its nuclear program in the past’.62

Public intelligence was at the heart of Powell’s address to the United Nations Security Council on 5 February 2003. Accompanied for effect by Tenet, Powell assured his listeners that: ‘every statement I make today is backed up by sources, solid sources . . . These are not assertions’. Powell argued that ‘we know from sources that a missile brigade outside Baghdad was dispersing rocket launchers and warheads containing biological agents’. He continued that satellite photos showed ‘banned materials have recently been moved from a number of Iraqi weapons of mass destruction facilities’ and that the United States had ‘first hand descriptions of biological weapons factories on wheels and rails’. Finally, he asserted that the United States had ‘detected one of Iraq’s newest UAV [unmanned aerial vehicles] in a test flight . . . [that could be used] to deliver biological weapons against its neighbors or if transported to other countries including the United States’.63

In the days before the invasion Bush also referenced US intelligence. In a radio address to the nation he said: ‘we have sources that tell us that Saddam Hussein recently authorized Iraqi field commanders to use chemical weapons’.64 Bush also referenced intelligence in his final ultimatum to Iraq on 17 March observing that ‘intelligence gathered by this and other governments leaves no doubt that the Iraqi regime continues to possess and cancel some of the most lethal weapons ever devised’.65

60Seymour Hersh, ‘The Stovepipe’, The New Yorker, 27 October 2003 ,http://archive. truthout.org/article/seymour-hersh-the-stovepipe. (accessed 6 December 2012). 61Ibid. 62See Walter Pincus, ‘Bush Team Kept Airing Iraq Allegation’, The Washington Post, 8 August 2003, p.A10; and Walter Pincus and Mike Allen, ‘CIA Got Uranium Mention Cut’, The Washington Post, 13 July 2003, p.A1. 63Transcript of Powell remarks at UN, ,http://abcnews.go.com/US/story? id=90856&page=1. (accessed 6 December 2010). 64President’s Radio Address, 8 February 2003 ,http://www.gpo.gov/fdsys/pkg/PPP-2003- book1/pdf/PPP-2003-book1-doc-pg140.pdf. (accessed 6 December 2012). 65Dana Milbank, ‘Bush Remarks Confirm Shift in Justification for War’, The Washington Post, 1 June 2003, p.A18.

Intelligence and National Security28

The intelligence community was roundly criticized for the mistakes it made in analyzing the data that led to these assertions. In the words of David Kay: ‘we were almost all wrong’.66 A second conclusion has been presented by Paul Pillar who served as National Intelligence Officer for the Near East and South Asia from 2000 to 2005. He observes that intelligence on Iraqi weapons program did not drive the US decision to go to war. Intelligence played a relatively small role in the decision-making process. Going to war was a political decision. Most of the intelligence analysis available to policymakers suggested caution if not avoidance of war forecasting the difficulties of spreading democracy through the region and ruling over post- war Iraq. Controversy surrounds why the intelligence used by policymakers was off

target. Charges of politicization were levied at the administration. Both the Robb-Silberman Commission and the Senate Select Committee on Intelligence absolved the Bush administration of charges of having pressured analysts into adopting a view supportive of its policy. Beneath this judgment, however, lies a more complex picture of efforts by the administration to steer and direct intelligence in the desired direction. Administration officials, Pillar notes, cherry-picked intelligence, presenting

in public evidence and analysis at variance with the position of analysts leaving them ‘to register varying degrees of private protest’.67 Further complicating matters was the absence of any identifiable point in the formal decision-making process that brought analysts and policymakers together to address the issues involved in going to war. Pillar also comments on other ways in which the actions of policymakers commutatively affected the content and flow of intelligence analysis. Reports containing favorable conclusions had an easier time navigating through the approval process; questions were asked on some subjects but not others; critical findings were repeatedly reviewed; and some questions were repeatedly asked.68 Most overtly, a new intelligence unit, the Policy Counterterrorism Evaluation Group, was set up in the Pentagon to analyze data and critique the consensus view of the intelligence community on the extent of the link between Saddam Hussein and Al Qaeda.

Conclusion

The politicization of intelligence analysis is not confined to these examples from the past. The Carter administration, for example, provides evidence of hard and soft politicization. National security advisor Zbigniew Brzezinski tried to limit DCI Stanfield Turner’s access to the president and establish himself as his Carter’s intelligence advisor but Turner ‘ventured into the arena of making net assessments of Soviet and US forces

66‘Kay: “We Were Almost All Wrong’”, CBS, 5 December 2007 ,http://www.cbsnews. com/2100-500257_162-596595.html. (accessed 6 December 2012). 67Paul Pillar, ‘Intelligence, Policy, and the War in Iraq’, Foreign Affairs 85/2 (2006) p.19. 68Ibid., pp.21–3.

The Politics of Intelligence and the Politicization of Intelligence 29

(which had been off limits to DCIs since the 1940s)’.69 Carter released a sanitized CIA estimate on the global energy situation in an effort to change the direction of US energy policy. Carter was also the target of intelligence politicization. Senator Frank Church and others dramatized and made public intelligence on the presence of a Soviet brigade in Cuba as a means of furthering their own re-election chances and in the process brought the ratification of Strategic Arms Limitation Treaty (SALT) II to a halt. Moving forward in time, the popular uprisings in the Middle East in spring

2011 brought forward congressional questioning about the adequacy of the intelligence presented to the White House before the crisis began and comments from ‘current and former administration officials’ that President Barack Obama was critical of their performance.70 Only months before, in December 2010, a generally positive military assessment of the Afghan War brought forward references to a NIE which presented a more pessimistic view.71

DCI Robert Gates, whose first nomination to that position was objected to by career intelligence officers on the grounds of having politicized intelligence analysis on the Soviet Union, observed in a March 1992 speech to CIA officials that ‘the problem of politicization is as old as the intelligence business’. He went on to comment on the multiple shapes and forms it can take as well as the disagreement over how to define it.72

This essay has sought to affirm both aspects of Gates’ observation: the politicization of intelligence is enduring and it takes a variety of forms. It has sought to move beyond this recognition to providing a framework in which the politicization of intelligence both within an administration and across administrations can be studied and compared. Doing so provides us with a more complete understanding of the dynamics of the intelligence analytic process and a basis for more clearly evaluating the behavior of analysts and policymakers in intelligence analysis.

Acknowledgements

This is based on a paper presented at the International Studies Association Meeting, Montreal, Canada on 19 March 2011.

69Douglas Garthoff, Directors of Central Intelligence as Leaders of the U.S. Intelligence Community 1946–2005 (Washington, DC: Center for the Study of Intelligence 2005) p.142, footnote 19. 70Mark Mazzetti, ‘Obama Faults Spy Agencies’ Performance in Gauging Mideast Unrest, Officials Say’, New York Times, 4 February 2011. 71Elisabeth Bumiller, ‘Intelligence Reports Offer Dim Views of Afghan War’, The New York Times, 14 December 2010, p.1. 72Robert Gates, ‘Guarding Against Politicization’, Studies in Intelligence 36/5 (1992) pp.5–14.

Intelligence and National Security30

Notes on Contributor

Glenn Hastedt is professor of political science and justice studies at James Madison University where he serves as chair of the Justice Studies Department. He is the author of American Foreign Policy: Past, Present, Future (Longman 2012) and editor of the American Foreign Policy Annual Edition (McGraw-Hill). His writings on intelligence have appeared in Intelligence and National Security, The International Journal of Intelligence and Counterintelligence and various edited volumes.

The Politics of Intelligence and the Politicization of Intelligence 31

Copyright of Intelligence & National Security is the property of Routledge and its content may not be copied or

emailed to multiple sites or posted to a listserv without the copyright holder's express written permission.

However, users may print, download, or email articles for individual use.