Answer questions in paragraphs

Policy Preferences and Bureaucratic Position: The Case of the American Hostage Rescue Mission Author(s): Steve Smith Source: International Affairs (Royal Institute of International Affairs 1944-), Vol. 61, No. 1 (Winter, 1984-1985), pp. 9-25 Published by: Oxford University Press on behalf of the Royal Institute of International Affairs Stable URL: http://www.jstor.org/stable/2619777 Accessed: 19-02-2018 14:38 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

http://about.jstor.org/terms

Royal Institute of International Affairs, Oxford University Press are collaborating with JSTOR to digitize, preserve and extend access to International Affairs (Royal Institute of International Affairs 1944-)

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Policy preferences and bureaucratic position: the case of the American hostage rescue mission

STEVE SMITH*

Within two days of the seizure by student revolutionaries of the American embassy in Tehran on 4 November 1979, planning began on a possible rescue mission. Initial estimates of the probability of success were 'zero', given the severe logistic problems involved in getting to the embassy in Iran and back out of the country without losing a large number of the hostages as casualties. Nevertheless, as negotiations dragged on with very little promise of success, and as the 1980 American presidential election campaign approached, the decision was made to undertake a very bold rescue mission. Photographs of the charred remains of the burnt-out helicopters in the Dasht-e-Kavir desert provide the most vivid image of the failure of that mission.

The decisions about the mission were taken at three meetings on 22 March, 11 April and 15 April 1980 by a very small group of people (on average, there were nine participants). Since 1980, the hostage rescue mission has received considerable coverage in the press and in the memoirs of the participants in that decision-making process. As such, it is an excellent case-study for one of the most widely cited but rarely tested theories of foreign policy behaviour: the bureaucratic politics approach.

The theoretical background

The dominant theories of why states act as they do derive from the basic assumption of rationality. Most theories of foreign policy are based on the premise that states act in a more or less monolithic way: foreign policy is, accordingly, behaviour that is goal-directed and intentional. Of course, many practitioners and academics quickly move away from the monolith assumption, but they can rarely command the kind of detailed information that would enable them to assess precisely what the factions are and how the balance of views lies in any decision-making group. It is, therefore, very common to talk of states as entities and to analyse 'their' foreign policies according to some notion of a linkage between the means 'they' choose and the ends these must be directed towards. Since practitioners and academics do not literally 'know' why state X undertook action Y, it becomes necessary to impute intentions to the behaviour of states. The rationality linkage makes this task much easier; hence the popularity of the idea of the national interest, which incorporates very clear and powerful views on what the ends of governments are in international society, and, therefore, on hpw the behaviour can be linked to intentions. The most important attack on this viewpoint has been the 'bureaucratic politics approach', most extensively outlined by Graham Allison in his Essence of decision.' According to this approach, foreign policy is the result of pulling and hauling between the various components of the decision-making

: Steve Smith is Lecturer in Politics in the School of Economic and Social Studies at the University of East Anglia, Norwich.

1. Graham Allison, Essence of decision (Boston: Little, Brown, 1971).

0020-5850/85/1/0009-117$3.00 (? 1985 International Affairs

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

10 POLICY PREFERENCES AND BUREAUCRATIC POSITION

process. Foreign policy may, therefore, be better explained as the outcome of bureaucratic bargaining than as a conscious choice by a decision-making group. As Allison puts it, the outcome of the decision-making process is not really a result but 'a resultant-a mixture of conflicting preferences and unequal power of various individuals-distinct from what any person or group intended'.2 The critical point is that these conflicting preferences are determined, above all, by bureaucratic position. Foreign policy, according to this perspective, is therefore to be explained by analysing the bureaucratic battleground of policy-making, rather than imputing to something called the state a set of motives and interests. On the bureaucratic battleground, the preferences of the participants are governed by the aphorism first coined by Don Price, 'where you stand depends on where you sit'.'

Since the publication of Essence of decision in 1971, Allison's claims for the explanatory power of the bureaucratic politics approach have been challenged by several writers.4 While the logical structure of his models and their applicability to countries other than the United States have come under attack, the most damaging criticism has concerned the extent to which Allison's theory of decision-making is able to explain the events of the crisis his book focuses on-the Cuban missile crisis. Allison's bureaucratic politics model has been criticized for not being able to explain the policy preferences of those who made US policy during the crisis: his aphorism 'where you stand depends on where you sit' does not fit the evidence. Although Allison goes into a detailed examination of who proposed what policy,5 Desmond Ball has written, 'in the case of the Cuban Missile Crisis, Allison's bureaucratic politics approach would generally have been unable to have predicted from the basis of a person's position in the bureaucracy, what his position on the question of the missiles in Cuba would be'.6 Robert Art, in turn, argues that the aphorism has to be qualified 'with so many amendments before it begins to work that when it does we may not be left with a bureaucratic paradigm, but may in reality be using another'.7

The decision of the United States government to attempt a rescue of the 53 American hostages held in Iran offers an excellent opportunity for a case-study to act as a further test of Allison's claims about bureaucratic position and policy preference. Whereas it has become broadly accepted that bureaucratic position is likely to have some impact on the more routine areas of foreign and defence policy- making, doubts about the applicability of the approach to crisis decision-making have constituted a serious weakness in its explanatory power. The rescue mission offers a rare opportunity to examine the extent to which the approach can explain crisis decision-making: it was a decision taken over just three meetings; there were very few participants; there is ample evidence on the positions adopted by the participants; and, above all, it was a very controversial and bold decision, one which raised fundamental questions about the acceptability of the use of force.

2. Allison, Essence of decision, p. 145. 3. See Allison, Essence of decision, p. 176. 4. These criticisms are summarized in Steve Smith, 'Allison and the Cuban missile crisis: a

review of the bureaucratic politics model of foreign policy decision-making', Millennium: Journal of International Studies, Spring 1980, Vol. 9, No. 1, pp. 21-40.

5. Allison, Essence of decision, pp. 193-210. 6. Desmond Ball, 'The blind men and the elephant', Australian Outlook, 1974, Vol. 28, p. 77. 7. Robert Art, 'Bureaucratic politics and American foreign policy', Policy Sciences, 1973, Vol. 4, p.

473.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 11

The planning process for the rescue mission began on 6 November 1979, just two days after the hostages were seized in Tehran.8 During the winter and spring the planning continued, focusing on the composition and training of the rescue force, on the precise location of the hostages and the nature and location of their captors, and on the enormously complex logistic problems involved in mounting the mission. These preparations continued in secret alongside an equally complex process of negotiation for the release of the hostages with the various elements of the Iranian government (including a secret contact in Paris). Bargaining was also under way with the United States's allies, in an attempt to persuade them to impose sanctions on Iran. As noted above, there were three key meetings at which the rescue plan was discussed (on 22 March, 11 and 15 April 1980), although the actual decision to proceed, taken on 11 April and confirmed on 15 April, was in many ways only the formal ratification of what had by then become the dominant mode of thinking among President Carter's most senior advisers. There were two schools of thought in the initial reaction to the seizure of the hostages: first, that the United States should impose economic sanctions on Iran; secondly, that it should make use of international public opinion and international law to force the Iranian government to release the hostages. As these measures appeared less and less likely to succeed, the US government became involved in attempts to persuade its allies to join in economic sanctions-a move that succeeded just two days before the rescue mission.

President Carter's initial reaction to the seizure was to stress the importance of putting the lives of the hostages first. He declared on 7 December 1979, 'I am not going to take any military action that would cause bloodshed or cause the unstable captors of our hostages to attack or punish them'.9 Yet leaks from the White House indicated that military plans were being considered. By late March 1980, President Carter and his advisers were becoming convinced that negotiations were not going to be successful, a view confirmed by the secret source in the Iranian government. At a meeting held on 22 March at Camp David, the President agreed to a reconnaissance flight into Iran to find an initial landing site for the rescue force (Desert One). The plan called for eight RH-53 helicopters from the aircraft carrier Nimitz to fly nearly 600 miles, at a very low altitude and with radio blackout, from the Arabian Sea to Desert One. There, they would meet the rescue force of 97 men (codenamed 'Delta Force') who would have arrived from Egypt via Oman on four C-130 transport aircraft. The helicopters would refuel from the C-130s and then take Delta Force to a second location (Desert Two) some 50 miles south- east of Tehran, where Central Intelligence Agency (CIA) agents would meet them and hide the rescue force at a 'mountain hideout'. Delta Force would remain hidden during the day before being picked up by CIA operatives early the next night and driven to a location known as 'the warehouse' just inside Tehran. From there they would attack the embassy and the Foreign Ministry where three of the hostages were held, rescue the hostages, and take them to a nearby soccer stadium, where the helicopters would meet them and transfer them to a further airstrip at Monzariyeh, to be taken to Egypt by the C-130s. The planning process had meant that very definite deadlines had emerged: by 1 May there would only be 16 minutes

8. For an account of the rescue mission see Steve Smith, 'The hostage rescue mission', in Steve Smith and Michael Clarke, eds., Foreign policy implementation (London: Allen & Unwin, forthcoming, 1984).

9. Robert D. McFadden, Joseph B. Treaster and Maurice Carroll, No hiding place (New York: New York Times Books, 1981), p. 197.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

12 POLICY PREFERENCES AND BUREAUCRATIC POSITION

of darkness more than required for the mission; by 10 May, the temperature would be so high that it would seriously hamper helicopter performance. 1 May appeared to be the latest feasible date for the mission, and by late March the planners were recommending 24 April for the mission (primarily because a very low level of moonlight was expected that night). But the rescue mission failed. It never got beyond Desert One. Of the eight helicopters assigned to the mission, one got lost in a duststorm and returned to the Nimitz, and two suffered mechanical break- downs. This left only five helicopters in working order at -Desert One, whereas the plan had called for six to move on to Desert Two. The mission was subsequently aborted, and, in the process of manoeuvring to vacate Desert One, one of the helicopters hit a C-130, causing the death of eight men.10

It is critical, in any discussion of the applicability of the bureaucratic politics approach, to focus on the actual decisions that led to this mission, and to review the positions adopted by the participants. We have so much information available that this is not a very difficult task, although very important issues are raised by any attempt to ascribe policy preferences to bureaucratic actors (a point which will be taken up in my conclusion). Nevertheless, we know that the three meetings of 22 March, 11 and 15 April were the decisive ones, and we know who took part and what they said. The key meeting in terms of the actual decision was on 11 April, when the 'go-ahead' was given. The meeting on 22 March was important because at it President Carter gave permission for aircraft to verify the site for Desert One. The meeting of 15 April was important because Cyrus Vance, the Secretary of State, presented his reservations about the decision. As Zbigniew Brzezinski, President Carter's National Security Adviser, pointed out: 'In a way, the decision [on 11 April] had been foreshadowed by the discussion initiated at the March 22 briefing at Camp David. From that date on, the rescue mission became the obvious option if negotiations failed-and on that point there was almost unanimous consent within the top echelons of the Administration."' A virtually identical set of people were present at those meetings. On 22 March, there attended President Carter, Walter Mondale (the Vice-President) Cyrus Vance (the

Secretary of State), Harold Brown (the Secretary of Defense), David Jones (the Chairman of the Joint Chiefs of Staff), Stansfield Turner (the Director of the CIA), Zbigniew Brzezinski (the National Security Adviser), Jody Powell (the Press Secretary), and David Aaron (the Deputy National Security Adviser). On 11 April, the same participants convened, except that Warren Christopher, the Deputy Secretary of State, replaced Cyrus Vance, and Carter's aide Hamilton Jordan replaced Aaron. The final meeting on 15 April was attended by the same people who attended on 11 April, except that Vance replaced Christopher.

In order to outline the positions adopted by the participants in this decision- making group, the participants can be divided into four sub-groups: President Carter, 'hawks', 'doves', and 'presidential supporters'. (These terms are only intended as analytical shorthand.) For reasons that will be discussed in more detail in the conclusion, the terms actually raise questions that are quite fundamental to the internal logic of the bureaucratic politics approach. Put simply, this concerns the issue, 'why are hawks hawks and not doves?' What this case-study shows is

10. The details of the planning for and the subsequent failure of the mission are discussed in Smith, 'The hostage rescue mission'.

11. Zbigniew Brzezinski, Power and principle (London: Weidenfeld & Nicolson, 1983), p. 493.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 13

that this is indeed a critical theoretical concern which has been rather neglected in the study of bureaucratic politics. Before we discuss that issue in detail, it is sufficient to say that this case-study supports the contention that bureaucratic position explains policy preference. The evidence indicates that these four sub- groups of people adopted consistent positions, positions which a priori seem predictable from their bureaucratic.base. Although there is a risk of fitting evidence to a preconception, the conclusion to be drawn from the detailed press discussions and subsequent academic and personal accounts is that these groups acted in accordance with what the bureaucratic politics approach would suggest: namely, that the National Security Adviser, the Secretary of Defense, the Chairman of the Joint Chiefs of Staff and the Director of the CIA would support military action (the position of the National Security Adviser will be returned to in my conclusion); the Secretary of State, and in his absence his deputy, would oppose it; those individuals who were bureaucratically tied to the President (the Vice-President, the Press Secretary and the Political Adviser) would be fundamentally concerned with what was best for the Carter presidency; and President Carter, although clearly more than just another bureaucratic actor, would act in a way that reflected bureaucratically-derived as well as personal influences.

President Carter

The key to understanding President Carter's position lies in the interaction between his desire to avoid the blatant use of American military power and the great pressure on him to satisfy his public and 'do something'. From the earliest days of the crisis, he was attacked in the press and by the Republican Party for failing to act decisively. 1980 was, of course, presidential election year, the President's public opinion rating was poor, and he was being challenged strongly for the Democratic Party's nomination. His promise not to campaign for the election so long as the hostages were in Iran made his situation worse. He was advised by his campaign staff that decisive action was needed (especially after the fiasco of the morning of the Wisconsin primary, on 1 April, when the President announced that the hostages were about to be released)12. That inaccurate assessment was seen by many as a reflection of his lack of control over events; it was also portrayed as manipulating the issue for his own political ends.

Another factor which added to the president's frustration was the desire to make the allies go ahead with sanctions against Iran.'3 It later turned out that the allies' belief that the US administration was planning military action was their main incentive to join in the sanctions, in the hope of forestalling it. But the critical moment came when the President felt that the only alternative to military action was to wait until, possibly, the end of the year for the release of the hostages by negotiation. That was the impression he gained in the early days of April: information coming out of Tehran indicated that the release of the hostages would be delayed for months by the parliamentary elections due to be held in Iran on 16 May. Indeed, by the time the rescue mission was undertaken, the favourite estimate of how long the new government in Iran would take to negotiate was five or six months.14 So, as a result of fear that the hostages might be held until the end of

12. See Hamilton Jordan, Crisis: the last year of the Carter presidency (New York: G. P. Putnam's Sons, 1982), pp. 248-9.

13. Brzezinski, Power and principle, p. 487. 14. Jimmy Carter, Keeping faith (London: Collins, 1982), p. 512.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

14 POLICY PREFERENCES AND BUREAUCRATIC POSITION

1980, President Carter determined on a change in policy: 'We could no longer afford to depend on diplomacy. I decided to act.'15 In fact, the President threatened military action on 12 and 17 April,16 unless the allies undertook economic sanctions. This action (which, he said, had not been decided on yet) would involve the interruption of trade with Iran. (This was widely interpreted as meaning a naval blockade or the mining of Iranian harbours.)17 Of course, this was a deliberate smokescreen: accordingly, when on 23 April the European countries agreed to the imposition of sanctions on Iran, the White House let it be known that this would delay any military action until the summer!18

Yet the desire of the President for drastic action is only part of the story. It is evident that he was also extremely concerned to limit the size of the operation, in order to avoid unnecessary loss of life. At the briefing with the mission commander, Colonel Beckwith, on 16 April, Carter said: 'It will be easy and tempting for your men to become engaged in gunfire with others and to try and settle some scores for our nation. That will interfere with your objective of getting our people out safely. In the eyes of the world, it is important that the scope of this mission be seen as simply removing our people.'19 W. Safire has argued that the reason why the mission was unsuccessful was precisely because Carter wanted the rescue to be a humanitarian rather than a combat mission, and stipulated only a small force with very limited back-up.20 Hence, in explaining President Carter's position on the rescue mission, two factors seem dominant: a personal concern to ensure that the mission was not to be seen as a punitive military action, and a role-governed perception that American national honour was at stake. Although, in early March, the President had been unwilling to give the go-ahead even to a reconnaissance mission, when negotiations with the Iranian leadership suddenly broke down in the first few days of April (leading to his embarrassment over the Wisconsin primary statement) he decided to go ahead with the rescue mission. At the meeting of 22 March, the participants were informed that 1 May was the last deadline for a mission, and that the night of 24-5 April was the best time, owing to the low projected levels of moonlight. The need to get the rescue force into the area therefore meant that a decision was necessary by mid-April. At this juncture, the breakdown of talks, the fiasco of the Wisconsin primary announcement, the predictions of, at best, a five or six-month delay in the release of the hostages, and the feeling among his key advisers that he had to act resulted in Carter's decision to proceed.

Carter's actions were, of course, a response to a number of factors. The bureaucratic politics approach draws our attention to certain of these: specifically, his desire for re-election, and his perception of his responsibility as the individual charged with protecting American national honour. Clearly, Carter's personality was an important factor (and one can easily imagine other Presidents handling the situation in slightly different ways), but the bureaucratic politics approach seems much more useful in identifying the kinds of considerations that would be important to Carter than concentrating on notions of what would be most rational for the

15. Carter, Keeping faith, p. 506. 16. New York Times, 17 Apr. 1980, p. A.1; 18 Apr. 1980, p. A.1. 17. Boston Globe, 16 Apr. 1980, p. 1. 18. New York Times, 23 Apr. 1980, p. A.1. 19. Jordan, Crisis, p. 263. 20. W. Safire, International Herald Tribune, 29 Apr. 1980, p. 5.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 15

American nation. This is not to imply that bureaucratic factors are the only important ones in explaining what Carter did; but it is to claim that a bureaucratic perspective paints a far more accurate picture of what caused Carter to act as he did than any of the rival theories of foreign policy-making.

The hawks

The leading political proponents of military action throughout the crisis were Brzezinski and Brown. Drew Middleton wrote, 'for months, a hard-nosed Pentagon view had held that the seizure of the hostages itself was an act of war and that the United States was, therefore, justified in adopting a military response'.21 Indeed, just two days after the hostages were taken, Brzezinski, Brown and Jones began discussing the possibilities of a rescue mission.22 Their discussions led to the conclusion that an immediate mission was impossible, but Brzezinski felt that 'one needed such a contingency scheme in the event ... that some of the hostages either were put on trial and then sentenced to death or were murdered.... Accordingly, in such circumstances, we would have to undertake a rescue mission out of a moral as well as a political obligation, both to keep faith with our people imprisoned in Iran and to safeguard American national honor.' In fact, Brzezinski felt a rescue mission was not enough: 'It would better if the United States were to engage in a generalized retaliatory strike, which could be publicly described as a punitive action and which would be accompanied by the rescue attempt. If the rescue succeeded, that would be all to the good; if it failed, the U.S. government could announce that it had executed a punitive mission against Iran.'23 This punitive action, he thought, could take the form of a military blockade along with airstrikes. In the earliest days of the crisis, Brzezinski, Turner, Jones and Brown began to meet regularly in private and discuss military options; Brzezinski alone took (handwrit- ten) notes. It was this group which directed the planning for the mission (which used military and CIA personnel) and gave the eventual plan its most detailed review.24 Similarly, it was Brzezinski who pressed for the reconnaissance flight into Iran, agreed on 22 March, and the same group of four who proposed the rescue plan at the 11 April meeting, led by Brown and Jones. But it is clear from the available evidence that Brzezinski was the political force behind military action.

As early as February, Brzezinski felt increasing pressure from the public and from Congress for direct action to be taken against Iran. Brzezinski thought there were three choices: to continue negotiations, to undertake a large military operation, or to mount a small rescue mission. What swung him away from his earlier first choice, a punitive military operation, was the consideration that, after the Soviet intervention in Afghanistan in December 1979, any military action might give the Soviet Union additional opportunities for influence in the Persian Gulf and Indian Ocean: 'It now seemed to me more important to forge an anti-Soviet Islamic coalition. It was in this context that the rescue mission started to look more attractive to me.'25 As negotiations failed, Brzezinski sent a memorandum to Carter on 10 April in which he argued that a choice must be made between a punitive military action or a rescue mission. Given his fears about the spread of Soviet

21. Drew Middleton, 'Going the military route', New York Times Magazine, 17 May 1981, p. 103. 22. Jordan, Crisis, pp. 258-9. 23. Brzezinski, Power and principle, pp. 487-8. 24. Brzezinski, Power and principle, pp. 488-9. 25. Brzezinski, Power and principle, p. 489.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

16 POLICY PREFERENCES AND BUREAUCRATIC POSITION

influence, Brzezinski recommended the latter option, concluding, 'we have to think beyond the fate of the 50 Americans and consider the deleterious effects of a protracted stalemate, growing public frustration, and international humiliation of the US.'26 At both the 11 April and 15 April meetings, Brzezinski spoke forcefully in favour of the mission.27

Brown and Jones were the main advocates of the actual rescue plan (Brzezinski still wanting more than the others in the way of a slightly wider retaliatory strike). These two men presented the plan to the 11 April meeting, and conducted the detailed private briefing with Carter on 16 April; it was Harold Brown who gave the detailed account, and defence, of the mission to the press after its failure. It was also Brown who spoke against the Christopher/Vance position at the 11 April and 15 April meetings.28 Finally, both Brown and Brzezinski spoke very strongly in justification of the mission after its failure, stating that it had been morally right and politically justified.29 Brzezinski was said to be 'downright cocky about it [the mission] in private and insisting that military action might be necessary in future'.30 He also warned America's opponents: 'Do not scoff at Arrmerica's power. Do not scoff at American reach.'31

Turner, the Director of the CIA, was also very much in favour of the mission, so much so that it appears that he did not voice the very serious doubts about the mission which had been expressed in a report by a special CIA review group, prepared for him on 16 March 1980. According to this report, the rescue plan would probably result in the loss of 60 per cent of the hostages during the mission: 'The estimate of a loss rate of 60 per cent for the AmEmbassy hostages represents the best estimate'.32 The report also estimated that the mission was as likely to prove a complete failure as a complete success. Yet it was exactly at this time that the review of the plan was undertaken by Brzezinski's small group. To quote Brzezinski again: 'a very comprehensive review of the rescue plan undertaken by Brown, Jones, and me in mid-March led me to the conclusion that the rescue mission had a reasonably good chance of success though there probably would be some casualties. There was no certain way of estimating how large they might be [emphasis added]'.33 Turner was involved in the detailed briefings of the President; at the meeting of 11 April he even said, 'The conditions inside and around the compound are good'.34 The evidence does not suggest that he made his agency's doubts public at any of these meetings, either in the small group or in the group of nine.

To sum up: the positions adopted by those classified here as 'hawks' could have been predicted in advance. What is striking about the evidence is the consistency with which these four men-Brown, Brzezinski, Jones and Turner-proposed policies that reflected their position in the bureaucratic network. (I must, however, point out one problem raised by this analysis concerning the issue of whether bureaucratic position per se causes these people to be hawks. On the one hand,

26. Brzezinski, Power and principle, p. 492. 27. Jordan, Crisis, p. 251; Brzezinski, Power and principle, pp. 493-4. 28. Brzezinski, Power and principle, pp. 492-3, 494. 29. Keesings Contemporary Archives 1980, p. 30532. 30. The Times, 1 May 1980, p. 16. 31. International Herald Tribune, 28 Apr. 1980, p. 1. 32. Pierre Salinger, America held hostage (New York: Doubleday, 1981), p. 238. 33. Brzezinski, Power and principle, pp. 489-90. 34. Jordan, Crisis, p. 251.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 17

their position puts them in contact with certain ways of thinking, and they come to share ways of seeing the world; yet bureaucratic position alone might not necessarily determine whether a person is a hawk or a dove. This problem, 'why are hawks hawks?', will be discussed again in my conclusion.) Nevertheless, the evidence indicates strongly that bureaucratically-determined thinking on the part of Brown, Brzezinski, Jones and Turner was a major input to the decision-making process. To the extent that the bureaucratic politics approach explains the policies adopted by these individuals, it illustrates the weaknesses of rationality-based theories of US foreign policy.

Presidential supporters

The next group to consider are those who do not fit into the traditional 'hawks- doves' characterization of US government. These are individuals whose primary loyalty is to the President, and who would therefore be expected to adopt positions that promised to bolster the President's domestic standing. Unlike those groups discussed so far, the first concern of this group is not the nature of US relations with other states, but, rather, the domestic position of the President. Mondale, Powell and Jordan seem to have been neither 'hawks' nor 'doves' in their views of the Iranian action; rather, their policy proposals show that their concern was first and foremost with the effect of the crisis on the Carter presidency. This can be seen very clearly in Jordan's memoirs,35 which reveal both a loyalty to Carter and an evaluation of the rescue mission in terms of how it helped Carter out of a domestic political problem. 'I knew our hard-line approach would not bring the hostages home any sooner, but I hoped that maybe it would buy us a little more time and patience from the public.' The rescue mission was 'the best of a lousy set of options'.36 Throughout his memoirs, at every juncture of the mission's planning, failure and consequences, Jordan's position is consistently one in which he advocates what he believed would benefit the President. This determined his reaction to Vance's objections (Vance was failing to support the President when he needed it, thereby putting Carter in an uncomfortable position), to the failure of the mission (Congress's reaction would be to concentrate on the lack of consultation and it might accuse Carter of violating the War Powers Resolution), and to Vance's resignation and his replacement by Ed Muskie37 (the former created a problem for Carter, the latter was a vote of confidence in Carter's political future).

The evidence also unambiguously supports the contention that Mondale and Powell were motivated above all by an awareness of the President's domestic standing and their perceptions of how it might be improved. Brzezinski notes that Powell, Mondale and Jordan 'were feeling increasingly frustrated and concerned about rising public pressures for more direct action against Iran'.38 All of them seemed to think that direct action was needed to stem this public pressure, especially after the Wisconsin primary announcement on 1 April. As Powell put it on 1 April: 'We are about to have an enormous credibility problem. The combination of not campaigning and that early-morning announcement has made skeptics out

35. Jordan Crisis, pp. 248-89. 36. Jordan, Crisis, pp. 248-9. 37. Jordan, Crisis, pp. 264, 275, 283 & 285, respectively. 38. Brzezinski, Power and principle, p. 490.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

18 POLICY PREFERENCES AND BUREAUCRATIC POSITION

of even our friends in the press.'39 Salinger argues that Carter's 'campaign for re- election registered the frustrations of the American public. While his political fortunes had risen after the taking of the hostages, he was beginning to slip in the polls and had lost a key primary in New York to Senator Kennedy. Jimmy Carter was now in the midst of a fight for his political life, and it looked as if he was losing. A military operation that freed the hostages would dramatically alter the odds.'40 The position of the 'presidential supporters' was summed up in Mondale's contribution to the 11 April meeting, when he said, 'the rescue offered us the best way out of a situation which was becoming intolerably humiliating'.41 (Interestingly, humiliation was also a concern of the hawks.) Before agreeing to the mission at that meeting, Carter informed those present that he had discussed the matter fully beforehand with Mondale, Powell and Jordan and that they all felt strong action was required.

The 'presidential supporters', then, proposed policies which reflected their own bureaucratic position. Mondale, Powell and Jordan had no vast bureaucratic interests to represent, nor was their chief concern the relationship between US foreign policy and other states. Each of them owed their influence to their position vis a' vis President Carter (as, of course, did Brzezinski), and their concern was to act so as to aid his presidency, above all his domestic political fortunes. In contemporary press reports, it was these three men who voiced concern about the President's relations with Congress and his chances of re-election. This was in contrast to both the 'hawks' and the 'doves' who were far more concerned with Carter's relations with Iran, the Soviet Union and US allies. As in the case of the 'hawks', the policy preferences of the 'presidential supporters' seem to have been predominantly determined by their bureaucratic role.

The doves

The evidence that bureaucratic role determines policy stance is strongest of all in the case of the 'doves': Cyrus Vance, the Secretary of State, and Warren Christopher, the Deputy Secretary of State. Not only did the two men take virtually identical stands on the subject of the rescue mission, but, as will be discussed below, Christopher did not know what Vance's position was when he attended the 11 April meeting.

From the earliest days of the crisis Vance had advised against the use of military force.42 At the meeting on 22 March, Vance agreed that a reconnaissance flight should go ahead in case a rescue mission should prove necessary (in the case of a threat to the hostages' lives), but argued against 'the use of any military force, including a blockade or mining, as long as the hostages were unharmed and in no imminent danger. In addition to risking the lives of the hostages, I believed military action could jeopardize our interests in the Persian Gulf ... Our only realistic course was to keep up the pressure on Iran while we waited for Khomeini to determine that ... the hostages were of no further value. As painful as it would be, our national interests and the need to protect the lives of our fellow Americans

39. Jordan, Crisis, p. 248. 40. Salinger, America held hostage, p. 235. See also Newsweek, 5 May 1980, pp. 24-6, for a

discussion of the domestic context. 41. Brzezinski, Power and principle, p. 493. 42. Cyrus Vance, Hard choices (New York: Simon & Schuster, 1983), p. 377.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 19

dictated that we continued to exercise restraint.'43 After this meeting, Vance felt there was no indication that a decision on the use of military force was imminent, and on 10 April he left for a long weekend's rest in Florida. But on the very next day the meeting was held that made the decision to go ahead with the rescue mission. Jody Powell explained to the press later that Cyrus Vance was on a well earned vacation and that 'Vance was not called back because it would have attracted too much attention when the operation had to remain secret'.44 There is no evidence as to why the meeting was called in his absence, but it is clear that Vance did not know that the mission was being so seriously considered, and that everyone else involved knew that Vance would disagree. Tom Wicker argues that Vance was deliberately shunted aside from the critical meeting in order to weaken his (and the State Department's) ability to prevent the mission from proceeding.45 All the Carter, Brzezinski and Jordan memoirs say is that Vance was on 'a brief and much needed vacation' (Carter), 'on vacation' (Brzezinski), and 'in Florida on a long overdue vacation' (Jordan).46 In many ways the exclusion of Vance can be interpreted as a symptom of what Irving Janis calls 'groupthink'; other symptoms can also be determined in this case-study of the phenomenon, which refers to the tendency for groups to maintain amiability and cohesiveness at the cost of critical thinking about decisions.47

The President opened the meeting of 11 April by saying that he was seriously considering undertaking a rescue mission, and he invited Brown and Jones to brief those present on the planned mission. At this point, Jordan turned to Christopher and said: "'What do you think?" "I'm not sure. Does Cy know about this?" "The contingency rescue plan? Of course." "No, no-does he realize how far along the President is in his thinking about this?" "I don't know ... I assume they've talked about it. ' 48 When the briefing finished, Christopher was first to speak. He outlined a number of alternatives to a rescue mission: a return to the UN for more discussions, the blacklisting of Iranian ships and aircraft, the possibility of getting European support for sanctions against Iran. Brown immediately dismissed these as 'not impressive', and he was supp-orted by Brzezinski, Jones, Turner, Powell and Jordan, all of whom wanted to go ahead. Christopher was alone in his opposition to the plan. He declined to take up a formal position on the rescue mission since he had not been told about it in advance by Vance; he therefore felt that Vance had either accepted the plan or had felt that the State Department could not really prevent its going ahead. According to one press report after the failure of the mission, Christopher 'was led to believe that the Secretary [Vance] already knew that the President had decided to undertake the mission'.49 His impression

was reinforced when Carter informed the meeting that Vance 'prior to leaving for his vacation in Florida, had told the President that he opposed any military action but if a choice had to be made between a rescue and a wider blockade, he preferred

43. Vance, Hard choices, p. 408. 44. International Herald Tribune, 29 Apr. 1980, p. 1. 45. Tom Wicker, 'A tale of two silences', New York Times, 4 May 1980, p. E.23. 46. Carter, Keeping faith, p. 506; Brzezinski, Power and principle, p. 492; Jordan, Crisis, p. 250. 47. See Irving Janis, Groupthink, 2nd edn. (Boston: Houghton Mifflin, 1982). For a discussion of

the extent to which groupthink was present, see Steve Smith, 'Groupthink and the hostage rescue mission', British Journal of Political Science, forthcoming, 1985.

48. Jordan, Crisis, p. 251. 49. Wicker, 'A tale of two silences', p. E.23.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

20 POLICY PREFERENCES AND BUREAUCRATIC POSITION

the rescue'.50 Christopher knew that Vance had opposed the use of military force, but it is logical to assume that he felt all he could do was to offer non-belligerent alternatives (they were, after all, State Department people being held hostage) to any use of military force, but remain silent on the actual mission; particularly as it had been strongly suggested that Vance had already agreed to it. In support of this conclusion, it is interesting to note that Christopher did not contact Vance on holiday to tell him what had happened. When Vance returned on 14 April and was told what had happened at the meeting, Christopher explained that he had not telephoned him on holiday: 'Under the impression that Vance, who was taking his first time off in months, had given his tacit, if reluctant, approval to the plan, Christopher decided not to disturb him with word of the decision to go ahead'.51

Vance's reaction to the news was 'that he was dismayed and mortified'.52 Vance writes: 'Stunned and angry that such a momentous decision had been made in my absence, I went to see the President'.53 At this meeting Vance listed his objections to the mission, and Carter offered him the opportunity to present his views to the group which had made the original decision in the meeting to be held on 15 April. Vance's statement at that meeting focused on issues almost entirely dictated by his bureaucratic position.54 He said, first, that to undertake the mission when the United States had been trying to get the Europeans to support sanctions on the explicit promise that this would rule out military action, would look like deliberate deception; secondly, the hostages, who were State Department employees, were in no immediate physical danger; thirdly, there were apparently moves in Iran to form a functioning government with which the United States could negotiate; fourthly, that even if it succeeded, the mission might simply lead to the taking of more American (or allied) hostages by the Iranians; fifthly, it might force the Iranians into the arms of the Soviet Union; and, finally, there would almost certainly be heavy casualties (he cited the figure of 15 out of the 53 hostages and 30 out of the rescue force as a likely death-toll)55.

After Vance's comments, Brown turned to him and asked him when he expected the hostages to be released; Vance replied that he did not know.56 No one supported Vance: his objections were met by 'a deafening silence'.57 Although Vance said later that, after the meeting, a number of participants told him that he had indeed raised serious objections, no one mentioned them at the time58an example of 'groupthink'? Carter noted that Vance 'was alone in his opposition to the rescue mission among all my advisers, and he knew it'.59 In their memoirs, Carter and Brzezinski put Vance's subsequent resignation down to tiredness: 'He looked worn out, his temper would flare up, his eyes were puffy, and he projected unhappiness

50. Brzezinski, Power and principle, p. 493. 51. McFadden et al., No hiding place, p. 220. 52. Brzezinski, Power and principle, p. 493. 53. Vance, Hard choices, p. 409. 54. The following summary of Vance's views is based upon Vance, Hard choices, pp. 409-10; Jordan,

Crisis, pp. 252-4; Brzezinski, Power and principle, pp. 494-5; Carter, Keeping faith, p. 507; The Times, 29 Apr. 1980; International Herald Tribune, 29 Apr. 1980; New York Times, 28 & 29 Apr. 1980; Keesings Contemporary Archives 1980, p. 30532; Facts on File, Vol. 40, No. 2060, 2 May 1980.

55. Doyle McManus, Free at last! (New York: New American Library/Los Angeles Times, 1981), p. 149.

56. Brzezinski, Power and principle, p. 494. 57. Wicker, 'A tale of two silences', p. E.23. 58. Wicker, 'A tale of two silences', p. E.23. 59. Carter, Keeping faith, p. 513.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 21

... Cy seemed to be burned out and determined to quit' (Brzezinski); 'Vance has been extremely despondent lately ... for the third or fourth time, he indicated that he might resign ... but after he goes through a phase of uncertainty and disapproval, then he joins in with adequate support for me' (Carter)60 Even worries expressed by Vance about the details of the plan at the 16 April briefing were dismissed on the grounds that they reflected his opposition to the raid in principle.61 On 21 April, Vance offered his resignation to Carter; it was accepted, with the agreement that it would not be made public until after the rescue mission, whatever the outcome. Vance duly resigned on 28 April. The press reports about his resignation suggested that opposition to the mission was only the last incident in a long line, and that Vance's resignation stemmed from his battle with Brzezinski over the direction of US foreign policy. As a White House aide said, it had been 'clear for some time that Mr Vance was no longer part of the foreign policy mainstream in the Carter Administration'.62

That Vance and Christopher opposed the rescue mission is not, in itself, proof of the applicability of the bureaucratic politics approach. What is critical is that their opposition was generated not simply from their personal views, but more as a result of their bureaucratic position (although there is a problem in weighting these). Three factors warrant this conclusion. First, Christopher, without knowing Vance's position on the rescue mission, and having been told (erroneously) that Vance supported it, still outlined alternatives. In fact, his opposition to the mission was on the same grounds as Vance's, even though he was led to believe that his superior had given the go-ahead. Secondly, Vance's statement at the 15 April meeting very clearly reflected State Department concerns. The response of Brown and Brzezinski did not address the problems Vance had outlined (for example, the position of the allies), but stressed issues such as national honour and security. These are role-governed policy prescriptions. Thirdly, Vance was not opposed to a rescue mission as such, but only to one at a time when negotiation was still possible; his objection did not simply reflect a personal attitude towards violence.

Bureaucratic politics and implementation

This article has tried to analyse the relationship between policy preference and bureaucratic position at three meetings. However, it is important to note that the bureaucratic politics approach subsequently developed by Allison and Halperin63 is also significantly useful in explaining how the decision was actually implemented. Three features of their approach seem especially worth noting. First, their theory explains the composition of the actual rescue force. As Halperin and Halperin have argued, the 1948 Key West agreement on the division of responsibilities in the US armed forces resulted in the forces used for the rescue mission being made up of contingents from the navy, army and air force.64 Because the navy had responsibility for any operations from ships, the helicopters used on the mission had to be navy

60. Brzezinski, Power and principle, p. 496; Carter, Keeping faith, pp. 510-11. 61. Jordan, Crisis, pp. 263-4. 62. International Herald Tribune, 29 Apr. 1980, p. 1. 63. Graham Allison and Morton Halperin, 'Bureaucratic politics: a paradigm and some policy

implications', in Raymond Tanter and Richard Ullman, eds.: Theory and policy in international relations (Princeton: Princeton University Press, 1972), pp. 40-79.

64. The 1948 Key West Agreement was a compromise between the three branches of the US armed services over which service would perform which missions and would be responsible for procuring certain kinds of equipment.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

22 POLICY PREFERENCES AND BUREAUCRATIC POSITION

ones-yet their pilots and equipment were simply not ideal for the job. None of the pilots concerned had had experience of such a mission; and the helicopter chosen was one normally used to sweep mines. Halperin and Halperin conclude: 'Because of the Key West boundaries, then, the presence on ships of the best- qualified fliers would tip off enemy intelligence that business was not as usual ... Pentagon planners should not have had to chose a mine-sweeper helicopter and between pilots who knew the equipment and pilots who knew the mission'.65 Secondly, bureaucratic politics operated in determining the nature of the command structure of the rescue force. The most surprising aspect of this was that there was no single overall commander. To quote Richard Gabriel: 'In typical "systems" fashion the operation was conceived and assembled in components, each with its own commander. Thus, at Desert One, there were no less than four commanders: the rescue force commander, the air group commander, the on-site commander, and the helicopter force commander. In addition, the Joint Task Force Commander was not even on the ground with his staff; instead he was located aboard ship in the Persian Gulf.'66 The result was an inability to improvise when things went wrong. Each commander recommended aborting the mission, since the lack of a sixth helicopter at Desert One challenged each unit's planning assumptions. No commander could assume control outside his specific area. This problem of organization had recurred throughout the planning, with each service wanting a piece of the action. The eventual rescue force was arrived at not by logic but by bureaucratic bargaining. This was one of the conclusions of the Joint Chiefs of Staff's own review of the reasons for the failure of the mission.6' Thirdly, there is strong evidence of a serious bureaucratically-determined clash between the President and leading Congressional figures. After he had decided to go ahead with the mission, President Carter, with conflicting advice from Brzezinski and Vance, decided not to inform Congress of the plans. He did call in Robert Byrd, the Senate Majority leader, and inform him that there were plans for a rescue mission, but he did not tell him that the mission was actually under way.68 When the failure of the mission was announced, a large number of Congressional leaders attacked the administration for breaking the terms of the War Powers Resolution.69 Indeed, Senators Frank Church and Jacob Javits had written to Cyrus Vance the day before the rescue rhission took place, stressing that the President was required to consult Congress before sending US troops into a potentially hostile situation.'? While executive-legislative struggles are built into the wider political setting in which decision-making occurs, these struggles were bureaucratically-determined to the extent that any President would have had to face these pressures simply by being President. The role of President, by being located in a network of bureaucratic relationships with, for example, the Congress, axiomatically involved these strug- gles. Bureaucratic battles thus detract from any conception of a US national interest.

65. Morton H. Halperin and David Halperin, 'The Key West key', Foreign Policy, Winter 1983-84, No. 53, pp. 124-5.

66. Richard Gabriel, 'A commando operation that was wrong from the start', Canadian Defence Quarterly, Vol. 10, No. 3, 1980-1, p. 9.

67. Joint Chiefs of Staff, Special Operations Review Group, Rescue mission report (Washington, DC: Joint Chiefs of Staff, 1980), p. 50.

68. Congressional Research Service, Iran: consequences of the abortive attempt to rescue the hostages (Washington, DC: Congressional Research Service, 2 May 1980), p. 26.

69. New York Times, 28 Apr. 1980; The Times, 29 Apr. 1980. 70. Keesings Contemporary Archives 1980, p. 30534.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 23

The nature of this dispute is a classic example of the way in which bureaucratic battles cut across notions of a 'national interest'.7

Conclusion

In the three key meetings that led to the decision to undertake the hostage rescue mission, the evidence presented here suggests that the participants adopted positions that reflected their location in the bureaucratic structure. The influence of bureau- cratic structure makes it possible to explain the change in policy that occurred between'the 22 March meeting and that of 11 April., In each case, the same group proposed a rescue mission, and the same group (Vance o.n 22 March, Christopher on 11 April) opposed it. The change came about because the 'presidential supporters' and President Carter himself felt that the situation had altered significantly. While this alteration was due in part to external events (the breakdown of negotiations), the evidence presented above suggests that an even stronger reason was the extent of domestic criticism of Carter's inaction (especially after the Wisconsin primary fiasco). The 'presidential supporters' felt it was 'time to act'. For similar reasons, Cyrus Vance's inability to change the rescue decision at the 15 April meeting is also explicable from a bureaucratic political standpoint. In the event, of course, his doubts were only too clearly vindicated. What this case-study shows, therefore, is the limitations of an attempt to explain foreign policy decision-making as if the state were monolithic and as if 'it' had interests. Such an approach makes policy- making appear rational, and this is a major reason for the popularity of such a perspective; but the case of the hostage rescue mission amply demonstrates the limitations of such conceptions of rationality, in that the key decisions are more powerfully explained by the bureaucratic politics perspective.

However, this conclusion requires some qualification since it raises fundamental problems about the precise claims advanced by proponents of the bureaucratic politics approach. As was noted earlier, the question that must be addressed is whether bureaucratic position alone leads to the adoption of certain policy pos- itions.72 As it stands, the bureaucratic politics approach is rather mechanical and static; it commits one to the rather simplistic notion that individuals will propose policy alternatives because of their bureaucratic position. Two problems emerge when this is applied to a case-study such as this one. The first is that the bureaucratic politics approach lacks a causal mechanism; it cannot simply be true that occupying a role in a bureaucratic structure leads the occupant to hold certain views. The second relates to the wider issue of belief systems, in that certain individuals are 'hawkish' irrespective of their precise position in a bureaucracy. The latter problem is most clearly illustrated by the case of Brzezinski, since it is arguable that whatever position he had occupied in Carter's administration, he would have adopted roughly similar views. Together, these problems force us to focus on one issue, namely, the exact meaning of the notion of role in the context of the bureaucratic politics approach.

This issue has been dealt with in the literature in the work of Alexander George and of Glenn Snyder and Paul Diesing.73 George is concerned with the ways in

71. Allison's other approach, the 'organizational process model', clearly has implications for this case. See Smith, 'The hostage rescue mission.

72. I am much indebted to Phil Williams for making me aware of the problems in my original formulation of this argument.

73. Alexander George, Presidential decision-making in foreign policy: the effective use of information and advice (Boulder, Colo.: Westview, 1980); and Glenn Snyder and Paul Diesing, Conflict among nations (Princeton, NJ: Princeton University Press, 1977).

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

24 POLICY PREFERENCES AND BUREAUCRATIC POSITION

which US decision-makers use (and abuse) information and advice in the policy process. He examines in some depth the ways in which individuals and bureaucracies will select information to assist their rather parochial goals. In other words, through his study of the use of information, George arrives at precisely the same kind of concern that this study has led to, namely, the relationship between individuals and their policy advocacy. More saliently, in their comprehensive survey of crisis

decision-making, Snyder and Diesing discuss the psychological make-up of those groups of individuals named in their study (as in this) 'hawks' and 'doves'. They believe that 'hard and soft attitudes are more a function of personality than of governmental roles',74 and they offer a very useful summary of what the world- views of hard and soft-liners are. As such, the works of George and of Snyder and Diesing are the best available discussions of the impact of role on belief and of belief on information -processing. Yet neither of these two pieces of work provides the kind of analysis that would be required in order to solve the problems

this case-study has indicated in the bureaucratic politics approach.

While it is clear that it is simplistic to assume that bureaucratic position per se causes policy preference, it is equally clear that bureaucratic position has some impact. Role, in and of itself, cannot explain the positions adopted by individuals;75 after all, the very notion of role implies a certain latitude over how to play the role. Further, a role does not involve a single goal, and there is therefore significant room for manoeuvre and judgement in trading off various goals against each other.

Thus, for example, it is not a sufficient explanation of Vance's position just to say that he was Secretary of State. There was a complex interplay between his role, his personality, the decision under consideration, and other personal and bureaucratic goals. Yet role occupiers do become predisposed to think in certain, bureaucratic, ways, and for a variety of psychological reasons they tend to adopt mind-sets compatible with those of their closest colleagues. In addition, individuals are often chosen for a specific post because they have certain kinds of world-views. So for reasons of selection, training, and the need to get on with colleagues, it is not surprising that individuals in certain jobs have certain world-views. Neither George nor Snyder and Diesing unravel the problem this leads to-how to distinguish between personal beliefs and bureaucratically instilled beliefs-indeed, the evidence of this case-study indicates that this is too crude a way of thinking about the problem. To answer the question, 'why are hawks hawks?' it is not sufficient to say either, that is the individual's personal world-view (for no account is given of the origins of that personal world-view) or, it is because of their bureaucratic position (as this passes over the crucial issue of the individual's ability to interpret their role). Thus, while it is clearly the case that Brzezinski was a hawk, it is neither accurate to say that this was because he was National Security Adviser (since this would not in and of itself cause hawkishness), nor to say that his views were simply personal (since it is surely the case that, had he been Secretary of State, he would have had to argue for courses of action other than those he did argue for-given the State Department's concern with getting the allies to agree on sanctions).

74. Snyder and Diesing, Conflict among nations, p. 297. 75. I am grateful to Martin Hollis for suggesting these problems of over-mechanistic role conceptions

to me.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms

Steve Smith 25

This case-study therefore leaves us with some critical questions unanswered. On the one hand, the empirical findings are important in that they illustrate the weaknesses of the rational actor approach as an explanation of foreign policy behaviour. States are not monoliths, and we might impute very misleading intentions to them if we assume that decisions are rational in this anthropomorphic way. The evidence indicates that the bureaucratic politics approach is very useful in explaining the decision to make an attempt to rescue the hostages. The linkage between the policy preferences of those individuals who made the decision and their bureaucratic position is a more powerful explanation of that decision than any of the alternatives. But, as in the case of any theory, it is tempting to try to portray the outcome as something completely explicable by it. It is not: the bureaucratic politics approach overemphasizes certain factors and underemphasizes others. On the other hand, the theoretical implications of this case-study force us to consider the issue of the sources of the beliefs of decision-makers. The 'hawks-doves' dichotomy is brought out very strongly in this case-study; and yet the bureaucratic politics approach as it stands is not capable of supporting a convincing mechanism for linking position and world-view. Therefore, the empirical and theoretical implications impel us to consider precisely that mechanism. What is needed is to link the concept of individual rationality with the structural influence of bureaucratic position. Neither Allison's work nor the subsequent work of George and Snyder and Diesing does this. This article, therefore, points both to the utility of the bureaucratic politics approach and to its theoretical weaknesses. The very fact that bureaucratic position was so important in determining policy preference over the decision to attempt to rescue the hostages makes the clarification of the nature of bureaucratic role all the more important. The nexus between individual rational action and bureaucratic structure appears to be one of the most promising, but also one of the most complex, avenues for foreign policy analysis to explore.

This content downloaded from 130.166.3.5 on Mon, 19 Feb 2018 14:38:32 UTC All use subject to http://about.jstor.org/terms