Pragmatism over principle: US intervention and burden shifting in Somalia, 1992–1993

ABSTRACT The conventional wisdom about the 1992 US intervention in Somalia is that it was a quintessentially humanitarian mission pushed by President George H. W. Bush. This article challenges that interpretation, drawing on newly declassified documents. The Somalia intervention, I argue, was largely a pragmatic response to concerns held by the US military. In late 1992, as the small UN mission in Somalia was collapsing, senior American generals worried about being drawn into the resulting vacuum. Hence they reluctantly recommended a robust US intervention, in the expectation that this would allow the UN to assemble a larger peacekeeping force that would take over within months. The intervention ultimately failed, but the military learned useful lessons from this experience on how to achieve smoother UN handoffs in the future and thus effectively shift longer-term stabilisation burdens to the international community.

New evidence presented in this article raises serious questions about all three of these interpretations. First, declassified National Security Council (NSC) documents indicate that the deployment of US troops to Somalia was not driven primarily by humanitarian considerations. Instead, high-ranking Bush administration officials were motivated by pragmatic concerns: senior American generals, in particular, worried that because of prior US actions, if the small United Nations Operation in Somalia (known as UNOSOM) imploded under attacks from hostile militias, US troops might have to be brought in to extract the UN peacekeepers. The United States might then be left with primary responsibility for dealing with the famine and civil war, without a clear exit strategy. The generals viewed an immediate US intervention aimed at restoring a modicum of stability, followed by a handoff to a strengthened UN mission, as a preferable alternative. Second, the Somalia intervention was not pushed by President Bush; instead, it was recommended and indeed conceived by the US military under the leadership of General Colin Powell, chairman of the Joint Chiefs of Staff (JCS). Third, the documents indicate that Bush initially envisioned Restore Hope as a shortterm operation of only a few weeks. But such a temporally limited operation would have done little do diminish pressure for US intervention in Bosniahence, the Somalia-for-Bosnia argument is problematic. 4 In November 1992, UNOSOM was under siege, while tens of thousands of Somalis remained at risk of starvation. General Powell thus recommended a massive but temporally limited US intervention under national command to facilitate the delivery of relief supplies and prevent a wholesale collapse of the UN mission. The JCS expected that this would (a) allow the United States to retain the initiative, minimising the risk to American forces, and (b) offer the UN the necessary breathing space to assemble a larger and more capable peacekeeping force that would take over as soon as possible. Powell persuaded Bush that this would be a low-risk course of action. Consequently, Bush launched a large-scale US intervention that reflected the military's preferences in December 1992.
In the short run, Operation Restore Hope was able to somewhat stabilise the situation and drastically reduce civilian deaths. But the transfer to UN troops, formally completed in May 1993, was implemented too hastily: it was driven by a political timetable in Washington more than by progress on the ground, and US forces withdrew before the follow-on UN mission (UNOSOM II) was fully staffed. To complicate matters, UNOSOM II's ambitious new enforcement mandate drew the ire of Somali warlords, resulting in frequent attacks on the peacekeepers and increasingly turning the UN into a belligerent in the country's civil war. During the summer of 1993, US combat troops went back into Somalia to support the UN in its pursuit of hostile militia leaders. The mission ended in failure after 13 American soldiers were killed during the infamous 'Black Hawk down' incident in October, which led first US and then also UN authorities to mandate a scaled withdrawal of all international forces.
A quarter century after Operation Restore Hope, the American military continues to view this experience as a source of useful lessons. 5 The failure in Somalia increased the US military's reluctance to deploy American combat troops on humanitarian missions abroad. But it also appears to have prompted significant learning by the military as an organisation on how to more effectively hand off longer-term stabilisation to UN forces, should the president order similar interventions in the future.
Scholars of individual and organisational learning have long pointed out that experiences of failure, by challenging the status quo, produce a 'learning readiness that is difficult to produce without a felt need for corrective action'. 6 In line with this, after Somalia, the military's attention was drawn to previously overlooked challenges in implementing UN handoffs successfully, prompting a search for solutions. In particular, the Somalia experience drove home the importance of an integrated political-military effort based on clear benchmarks for measuring progress ahead of the transfer to UN forces.
As a result of learning from Somalia and from subsequent more successful experiences elsewhere, the idea that handing off longer-term stabilisation to UN missions is a central component of the US exit strategy for humanitarian interventions has become embedded in US military doctrine. Current US Army doctrine, for instance, states that officers planning humanitarian interventions should aim to 'accomplish transfer of authority [to follow-on multilateral missions] as early as possible. The timing of the transfer [ought to be] part of the initial negotiations '. 7 This article draws on recently declassified documents and interviews with high-level policymakers to elucidate how the notion of UN handoffs as central to the US exit strategy for humanitarian interventions first emerged during civil-military deliberations about Somalia in late 1992. The article also clarifies the motives behind this intervention by tracing the decision-making process and analysing the positions of key actors. The first section of the article reviews US and UN policy on Somalia during the first part of 1992, highlighting the US military's initial opposition to armed intervention. The second section zeroes in on the Washington policy debate in November 1992: I show that military leaders, in a stunning about-face, abandoned their prior opposition to intervention and essentially recommended a large-scale deployment of US combat troops. The third section clarifies the influence of this recommendation on Bush's decision to intervene and elucidates the reasons behind the military's change of heart. The fourth section discusses key factors that resulted in ultimate mission failure, before the conclusion returns to lessons learned.

Dealing with a complex humanitarian emergency
Somalia descended into a state of violent chaos in early 1991, after Cold War superpower funding for longtime strongman Siad Barre dried up, his regime collapsed, and various factions began to vie for control of the country's politics. Factional warfare destroyed the country's agricultural production and caused large-scale population displacement, yielding a humanitarian crisis of massive proportions. 8 By early 1992, more than 300,000 Somalis had died from malnutrition; an additional 1.5 million faced an immediate risk of starvation, and approximately 3000 were dying daily. 9 The summer 1992 US food airlift and the first 'safe area' proposals Plenty of food aid had been stocked in Somalia by international relief organisations. However, as aid convoys increasingly came under attack from armed militias, delivering that food to hungry civilians became exceedingly difficult. In April 1992, UN Secretary-General Boutros Boutros-Ghali recommended that 500 armed peacekeepers be deployed to Somalia to facilitate relief supplies in and around the capital, Mogadishu. The proposal was welcomed by US officials at the Department of State (DOS), especially in the Bureau for African Affairs. 10 However, senior policymakers at the White House and the Department of Defense (DOD) were reluctantnot least because under the UN's assessed contribution system, the United States would be responsible for 30 per cent of the mission's financial cost. Because of these hesitations in Washington, in the spring of 1992, only 50 UN observers were dispatched to Somalia. 11 The pressure on President Bush to support more serious international action on Somalia grew over the summer. After the United Nations Security Council (UNSC) sent armed peacekeepers to Bosnia in June, Boutros-Ghali bluntly criticised the West for caring only about the 'rich man's war' in the Balkansdrawing significant media attention. 12 Furthermore, in July, both houses of the US Congress unanimously adopted a resolution calling for the deployment of armed UN peacekeepers to Somalia to protect aid deliveries. 13 US presidential elections were scheduled for November 1992, and persistent portrayals of the Bush administration as callous in the face of massive suffering in Somalia could damage the president politically. It was thus probably not a coincidence that on the eve of the 1992 Republican National Convention, Bush announced several initiatives on Somalia. 14 First, in late July, the administration offered to support a UNSC resolution authorising the deployment of 500 peacekeepers to Mogadishu, as recommended by Boutros-Ghali. Shortly thereafter, Bush let it be known that US aircraft would transport the peacekeepers to Somalia. 15 The most important US policy change came on 14 August, when the White House declared that 'starvation in Somalia is a major human tragedy' and announced that 'the United States will take a leading role' by initiating a 145,000-ton emergency food airlift to the country. 16 The Pentagon's concerns that the offer to transport the UN peacekeepers, combined with the food airlift, might put the administration on a slippery slope toward US military intervention were brushed aside on this occasion. 17 The US food airlift, known as Operation Provide Relief, delivered tens of thousands of tons of food to Somalia. However, it soon became apparent that much of the food could not reach its intended beneficiaries because of inadequate security along major highways and looting by armed gangs. 18 Pro-intervention officials at the State Department consequently began to float the idea of establishing safe areas, or 'zones of tranquility', in rural southern Somalia, to be guarded by small numbers of American ground troops. The goal, they argued, would be to protect critical facilities such as airports, feeding centres and key roads, keeping the militias away so that food could be handed out to vulnerable populations. 19 Yet military planners on the Joint Staff pushed back vigorously against the idea of deploying small numbers of US troops to establish safe areas in Somalia, insisting 'that it was just complete nonsense, that it wouldn't work without muscular military power behind it'. 20

UN peacekeepers under fire
As the security situation in Somalia gave no sign of improvement, on 28 August 1992, the UNSC authorised 3000 additional troops for UNOSOM. 21 Several countries, including Belgium, Canada, Egypt, and Nigeria, offered troops. 22 But Somalia's most powerful warlord, Mohammed Aideed, had not been consulted on this expansion of the UN force, and he reacted furiously. First, he placed significant restrictions on the deployment of the initial contingent of 500 Pakistani peacekeepers authorised by the UNSC in July, to which he had previously given his consent. Second, he warned that any further, forcible UNOSOM deployment would be met by violence. 23 By October, the movement of relief workers in Somalia's south had become extremely difficult because of constant harassment by armed militias, and US officials worried that 'at most only half of the population threatened with starvation [was] being reached' by the airlift. 24 On 12 November, Aideed upped the ante, demanding that the Pakistani battalion that had managed to set up base at Mogadishu airport withdraw completely. That same day, after UN authorities declared that they had no intention of complying, the peacekeepers came under fire. 25 It now appeared unlikely that a mere 3000 additional peacekeepers, without a robust mandate and the capabilities to enforce it, would be able to make much of a difference on the groundindeed, they might not even be able to deploy. A secret CIA report warned US policymakers that the 12 November attack signalled a significant escalation of Aideed's anti-UN campaign. 26 The report predicted that a decision 'to try to move ahead with the [UN] deployment… will prompt Aideed to escalate'; specifically, it 'would probably spur him to bombard the airport and Pakistani bivouac area in a bid to force an evacuation'. 27 Deteriorating security conditions and concerns that an evacuation of UN peacekeepers under fire might have to be 'done by US forces', 28 in a high-risk operation, set the stage for a rethink of US policy.

US intervention seriously considered, November 1992
To fully understand the political importance of UNOSOM's looming collapse in late 1992, the broader context needs to be taken into account. The US food airlift was increasingly ineffective; it was also very expensive and could not be sustained indefinitely. Earlier in the fall, under pressure from DOD, the Bush administration had agreed to wind down the airlift and discontinue it by the end of January 1993. 29 That risked leaving the United States with a policy vacuum that could become politically unsustainable if the humanitarian crisis persistedpotentially saddling American forces with a long-term stabilisation burden after the UN peacekeepers withdrew. As Frank Wisner, the undersecretary of state for international security affairs, reminded his colleagues during an interagency meeting, America's actions up to that pointthe food airlift and the transportation of UN troops to Somaliamade it impossible for the United States to simply extricate itself: 'One reason for why this is our problem is because of what we have done'. 30 Senior military leaders had for months been resisting greater US involvement, seeking to disabuse their more interventionist civilian colleagues in the administration of the notion that American power could be used at little cost to defuse or resolve the humanitarian crisis. Admiral Frank Bowman, who was deeply involved in interagency deliberations on Somalia as the director of political-military affairs on the Joint Staff, recalls that 'There was a whole lot of pushback from the combatant commanders, [ Bush's own recollection is that he decided to explore options for more vigorous action on Somalia after losing the presidential election to Bill Clinton on 3 November 1992. As Bush and his wife were watching television around mid-November and saw images of 'those starving kids', the president summoned General Powell and Secretary Cheney to the White House and told them, 'Iwecan't watch this anymore. You've got to do something'. 33 However, Bush did not tell Powell and Cheney that he wanted to see a US military intervention. Powell says he 'vividly' remembers the mid-November White House meeting, at which Bush made it clear he 'wanted to do something'; but Powell insists that neither the president nor Scowcroft appeared 'anxious to send American soldiers to Somalia'; in fact, 'they were reluctant to get into it'. 34 Admiral David Jeremiah, Powell's deputy at the time, confirms that Bush was not driving policy on the issue of US intervention. 35 With the exception of the 14 August decision on the food airlift, which had been taken by the president with little prior consultation, the development of US policy options on Somalia was left to officials below cabinet level in the NSC's Deputies Committee. The deputies had been meeting regularly on Somalia through 1992, but without much urgency. 36 In late November, the discussion among the deputies at last became more focused and crystallised around two options: first, a significantly enlarged UN peacekeeping operation involving at least 15,000 troops, with US logistical support but without active US participation; second, a US-led coalition operation under national command. Because of the presumption of continued JCS opposition to any deployment of US troops, the first option initially appeared more likely to garner the administration's support. 37 The DOS proposal for a strengthened UN force US intelligence analysts had concluded by 10 November that the existing plan to insert 3000 additional peacekeepers was 'seriously flawed'. 38 Even if US aircraft were able to transport the peacekeepers 'into areas in which local agreement has been obtained', there was 'a high likelihood that at least one or another of the groups inserted by the United States… would [subsequently] require evacuation from a potentially hostile situation'. 39 The intelligence community concluded that 'a force far larger than the 3500 [500 + 3000] currently envisaged… would be needed'. 40 On 12 November, this same conclusion was adopted by the Policy Coordination Committee for Africa, an interagency group of mid-level officials tasked with preparing the agenda for the deputies. 41 Over the next several days, Undersecretary of State Wisner secured Eagleburger's blessing for a 'much stronger US leadership role in pressing the UN to mount a credible security force in Somalia'. 42 Wisner and Boutros-Ghali, who were personally acquainted from the former's time as US ambassador to Egypt, then began to discuss various options for a significantly enlarged UN peacekeeping operation of between 15,000 and 20,000 troops, with robust rules of engagement. 43 In preparation for a deputies meeting scheduled for 20 November, Wisner and his team at DOS prepared a detailed policy proposal. Their recommendation was that the administration 'offer to work with [Boutros-Ghali] to put together a coalition of nations prepared to participate in… a much larger force'. 44 The DOS proposal noted that there was 'no need at this point to commit US ground troops to such an operation' (the United States would only provide logistical support)a qualification that seemed intended to make the proposal acceptable to DOD and especially the JCS. The discussion initially focused on the DOS proposal. Wisner began by reminding his colleagues that 'the number dying is increasing dramatically'. 48 He then related that 'in discussions with the UN, I have been struck that they see the need to… [deploy] a force up to 15,000--20,000, perhaps inserted without permission'. 49 After Wisner had presented his case, Howe, the deputy national security adviser, asked, 'If the UN gets in trouble and needs reinforcement, are we willing to do that, even at some risk to our people'? 50 Wolfowitz seized on this question to voice his doubts about Wisner's proposal, which, he feared, could result in American forces being drawn in later, under more challenging circumstances. 'I am worried that this is [a] band aid', he affirmed. 'If we get involved with big resource implications, and there is "no US role for ground forces", that might shift if it doesn't work'. 51 The DOS proposal, he continued, 'will be costly and [is] likely to suck us in on the ground'. 52 Instead of considering such risky half measures, Wolfowitz suggested, 'we need to look at a serious use of force if we are serious'. 53 McCaffrey spoke next. 'You must be willing to destroy the forces that oppose food deliveries'. Therefore, 'there needs to be a transfer to a coalition…, and the ROE [rules of engagement] need to change'. 55 Then came the bombshell statement: 'it is resolvable by US forces, including combat troops', McCaffrey declared. 'It is a doable mission'. 56 Under this scenario, he explained, US and coalition troops would have a 'whack at it for three or four months, and then set up a police force'. 57 This represented a significant about-face, after months of insistence by the JCS that deploying US combat troops to Somalia would be too risky and was out of the question.
Howe seemed surprised and asked the representative of the US military for clarification: 'Are you saying that… [the United States should] take it on frontally, the US taking the lead, sorting it out and then leaving a UN police force?' McCaffrey replied by emphasising that the uniformed leadership had little faith in the DOS proposal: 'There is real skepticism that the UN will come up with the numbers or the will to do the job… This is not even an amateur effort… The US has the capability'. 58 By speaking out strongly against the DOS proposal and underscoring the preferability and feasibility of a US-led coalition operation followed by a UN handoff after a few months, McCaffrey had for all practical purposes come close to recommending such a two-step approach. This was all the more remarkable because, as noted, policymakers had until then hardly considered the possibility of a significant US troop deployment.
No decision was taken at the 20 November meeting, but Wolfowitz's and McCaffrey's statements had dramatically shifted the terms of the debate. In a follow-up small group meeting on 23 November, the deputies laid out three options for the president. Option one was an expanded UN mission: the United States would offer logistical support for a 'significantly larger UN force that would be prepared to deploy and operate effectively even without the agreement of local factions'. 59 Option two was a 'coalition of nations [acting] under the authority of a new UN resolution', under national rather than UN command, but without direct participation of US troops. 60 Option three was a US-led coalition: the United States would seek 'a new UNSC resolution that permits the US to create, lead, and contribute the bulk of a coalition force that would operate under UN authority but not UN command'. This last option would likely involve a 'six month presence prior to a hand-off to an adequately sized and equipped UN peacekeeping force'. 61 55 Ibid. 56 Ibid. 57 Ibid. 58 Ibid., 6-7. 59 'Options Paper: Next Steps in Somalia', small group meeting of DC (23 November 1992), 1. 60 Ibid., 2. 61 Ibid., 3.
The deputies indicated their support for option three, noting that it 'stands the best chance of rapid success'. 62

The military's influence on Bush's decision to intervene
On 25 November, Bush met with his senior foreign policy team to discuss the Somalia options. After Powell presented the military's plans for each of the three options, the president approved option three in principle, involving a large US intervention force of more than 20,000 troops. 63 The decision to intervene was made contingent on (1) a UNSC mandate for the use of force, (2) participation from other countries, and (3) the prospect of a relatively quick handoff of to a strengthened UN peacekeeping mission. 64 On 3 December, the UNSC unanimously approved Resolution 794, which authorised the US-led coalition 'to use all necessary means to establish… a secure environment for humanitarian relief operations in Somalia'. That same day, Bush convened an NSC meeting with his principal advisers and informed them that the decision to intervene had become final. 65 On 4 December, Bush announced in a televised address to the nation that US troops would intervene to 'create a secure environment' for food deliveries. He emphasised that this would be a limited operation with a clear end point: 'once we have created that secure environment, we will withdraw our troops, handing the security mission back to a regular UN peacekeeping force'. 66 How Powell convinced the president Previous accounts have claimed that 'the decision [to deploy US troops] was the president's, especially since no recommendations accompanied the options paper' developed by the deputies. 67 In a technical sense, the decision was of course the president's. But the view that the options paper developed by the deputies on 23 November included no 62 Ibid. 63  recommendation is incorrect. The way in which the deputies had framed the problem and identified possible policy answers narrowed down the president's choice significantly. The deputies, taking on board the Pentagon's concerns about immediately deploying an enlarged UN force, had recommended against going forward with this option (option one) in their 23 November paper. 68 Furthermore, as noted, the deputies acknowledged in the same paper that a coalition effort with US ground troops in a lead role had the best chance of succeeding.
An annex to the 23 November options paper mentioned that 'Option 3 action line awaits military in-put', making it clear that the JCS needed to provide further details regarding this option's feasibility. 69 Consequently, whether the president would go for option two (a coalition without US troops) or some version of option three (a US-led coalition) depended, to a significant degree, on what the JCS was going to recommend. A 24 November paper authored by the Joint Staff acknowledged that option two was probably unfeasible because other countries would not contribute significant forces unless the United States did. 70 This could be read as a cautious endorsement of option three by the military.
The key recommendation to move ahead with option three was subsequently made by General Powell during the 25 November White House meeting. By highlighting the pros and cons of the various options, Powell effectively, if reluctantly, came to recommend a full-scale US intervention. 71 Powell himself remembers his advice to the president that day as follows: I said, 'we have a plan on how to do it'. And the plan reflected my own view that if the UN is going to be able to do this, then we first had to send in something big enough to scare the militias. It was for this reason that I recommended to the president that we send in a large US force. 72 Powell's change of heart, backed by Secretary Cheney, appears to have played a central role in convincing President Bush of the wisdom of immediate action by American forces. As Wolfowitz sums it up, 'once the president was persuaded to do this, Powell and the military loyally implemented what we all understood to be the president's intention. But Powell had a lot to do with shaping the president's view of it'. 73 It seems that on this occasion, the 68  military did not flout the civilian principals' wishes and/or impose its own views by political maneuvering. 74 Instead, Powell and other military leaders genuinely seem to have persuaded President Bush that a large-scale US intervention, followed by a handoff to UN peacekeepers, would be the best course of action. 75 Hypotheses on the military's change of heart What explains the senior officers' recommendation to intervene, after they had insisted for months that deploying US combat troops to Somalia was out of the question? It appears to have been influenced by concerns that since the United States had staked its credibility on a successful relief effort, if the UN peacekeepers came under sustained attack, American forces might be drawn in to support them or facilitate an evacuation. In September, the Bush administration had stationed 2100 Marines off the coast of Somalia, announcing that the goal was to 'provide cover' for the UN peacekeepers. 76 Thus, apart from the administration's political commitment as a result of the airlift, the credibility of the US military was arguably at stake. As Wolfowitz, the Pentagon's third-highest-ranking civilian at the time, remembers the internal DOD conversations, 'the idea was, if you give the UN peacekeepers a mandate and they can't fulfil the mandate, then the pressure comes on you. There's a credibility issue'. 77 The CIA predicted that if small numbers of US forces were sent in to support the UN operation, Aideed 'would probably respond by engaging them'. 78 In a worst-case scenario, if the UN peacekeepers withdrew entirely, American forces might be left with primary responsibility for stabilising Somalia. Research shows that policymakers are more likely to opt for quick offensive action when they believe there is a high chance of armed confrontations involving their country's armed forces in the near future. 79 The military, in particular, tends to favour offensive action in such circumstances because it views it as a way of 'exercising the initiative [and] fighting the war that it [has] planned'. 80 Consistent with these findings, in the Somalia case, the JCS appears to have concluded that it would be better to move to the offensive by deploying a large US force to establish safe distribution points for humanitarian relief, instead of intervening reactively to defend or extract UN peacekeepers under attack.
The US military and Chairman Powell in particular believed that the risk to American troops could be minimised by deploying overwhelming force for a well-defined purpose. This 'decisive force' approach had been vindicated by the 1991 Persian Gulf War victory, and it was subsequently enshrined in the 1992 US National Military Strategy, drafted under Powell's guidance. 81 When it came to planning for the Somalia intervention, Bush followed Powell's advice to deploy a massive force: 38,000 coalition troops were ultimately deployed as part of Operation Restore Hope, with 21,000 coming from the United States. 82 Powell was confident that such a large force would intimidate the Somali militias; consequently, expectations of US casualties were low, and in early December, he assured Bush that US and coalition forces were unlikely to encounter major resistance. 83 It is possible that Powell thought the Somalia intervention would further vindicate the decisive force concept and demonstrate its applicability to non-traditional missions. 84 Another, related factor that appears to have made the senior generals and admirals more supportive of US intervention was their ability to define a limited mission for American forces. In early November, Powell and Wolfowitz met with Fred Cuny, a humanitarian relief expert who had been involved in Operation Provide Comfort, the US-led humanitarian intervention in northern Iraq in 1991, and who was now working as a consultant on Somalia. Cuny made the case that 'if you stay away from what's politically decisive, people won't interfere with you too much'. 85 This argument struck a chord with the DOD leadership. Once senior Pentagon officials began to think of the mission as simply ensuring the safe delivery of relief supplies, it appeared more feasible. Powell was turned around, and civilian policymakers from other agencies began to hear that, provided US military objectives were narrowly defined, this was a 'doable mission'. The key factor, Wolfowitz insists, is that 'the mission changed'from helping the UN stabilise Somalia to opening up the supply routes for humanitarian relief. 86 To minimise the risk of serious clashes with Somali militias, the uniformed leaders were also adamant that American forces do not have, as their mission, the disarmament of any of the Somali groups. 87 President Bush again accepted the military's request. Consequently, when Boutros-Ghali told Bush during a telephone conversation on 8 December 1992 that 'disarming the gangs, and the heavy weapons, will be important', Bush replied: 'I agree it is important. But we have not made it part of our mission statement'. 88 (The US military's staunch resistance to contemplating any form of disarmament, including confiscation of heavy weapons, would subsequently complicate the handoff to UN forces).
Finally, as noted in the introduction, some authors have suggested that Powell and other senior generals came to support a US intervention in Somalia because they expected that this would reduce the likelihood of US intervention in Bosnia, which the military dreaded far more. 89 But this argument is problematic. Powell was certainly opposed to US intervention in Bosnia, and he made his opposition public in September 1992. 90 However, Operation Restore Hope, the US intervention in Somalia, was intended as a short-term operation. As Bush declared during an NSC meeting on 3 December, 'I expect that within 40 days troops can start coming out'. 91 Even under a four-to six-month time frame, which had been the working assumption during earlier discussions among the deputies, the plan was for US troops to be out of Somalia by the spring of 1993. 92 A US withdrawal from Somalia by the spring of 1993 would have left the incoming Clinton administration with ample room to intervene in Bosnia. This lends credibility to Powell's claim that he 'wasn't thinking about that, or worrying about that' (a US intervention in Bosnia) when planning the Somalia intervention. 93

A UN handoff as central to the US exit strategy
After the end of the Cold War, the Bush administration had come to see UN peacekeeping operations as a relatively low-cost means of managing governance crises in peripheral regions of the world while limiting US liability. 94 The administration, including DOD, had invested heavily in developing the UN's capabilities for robust peacekeeping; however, the UNOSOM experience over the course of 1992 had revealed that the world organisation continued to face serious challenges in implementing this type of operations. 95  in Somalia as a test case 'to demonstrate that the UN can act decisively in the post-Cold War world'. 96 At the same time, as noted, the American military worried that simply beefing up UNOSOM in late 1992 would be difficult and might draw in US troops under unfavourable circumstances. The JCS thus put forward the idea of a US 'bridging operation'. The expectation was that a large-scale US operation, limited in time, could offer the UN the necessary breathing space to assemble a larger and more robust peacekeeping force that would take over within months. This expectation seems to have played a crucial role in persuading the military, and the administration more generally, to support a US intervention. 97 As Scowcroft, the former national security adviser, explains, 'we finally hit upon this notion, well if we reopen the food supply lines, then we can turn it over to the UN forces, and they will be able to keep them open even though they don't have the force to open them'. 98 The groundwork for the handoff to a strengthened UN force was laid by military planners at US Central Command (CENTCOM) and on the Joint Staff, in cooperation with Robert Gallucci's office for political-military affairs at the State Department. 99 CENTCOM planners defined the desired end state for the US operation as 'the creation of an environment where UN and relief organizations can assume responsibility'. 100 The contingency plans that CENTCOM had developed since early November 1992 foresaw a US operation in four phases, with phase IV involving a handoff to a strengthened UN peacekeeping force (UNOSOM II) after a period ranging from between three to eight months. 101 As Admiral Bowman recalls, from the military's perspective 'the UN handoff was really a requirement. It was present in all the CENTCOM plans and certainly all the Joint Staff recommendations during the interagency discussions'. 102 The US military in fact sought to entrench the principle of a UN handoff in the UNSC resolution authorising the US-led intervention. At a 30 November Deputies Committee meeting, Admiral Jeremiah presented an amendment to the draft UNSC resolution then under discussion in New York that would have committed 'UNOSOM [to] assume the role of the coalition forces upon their departure'. 103 Ideally, the JCS would also have wanted to keep UNOSOM functioning in some form while the US-led coalition intervened, to facilitate and speed up the subsequent transition. Admiral Bowman was dispatched to New York to lobby the UN secretariat on these points, but Boutros-Ghali turned down the US military's requests. 104 UNSC Resolution 794, which authorised the US-led intervention, simply mentioned that the Security Council would 'make the necessary decision' for a transition back to UN peacekeeping based on a recommendation from the secretary-general, once the latter had ascertained that conditions were sufficiently improved. 105 In other words, there would be no automatic handoffinstead, a further UNSC resolution would be needed.

Incomplete handoff and mission failure
The first contingent of US troops from the I Marine Expeditionary Force under the command of Lieutenant General Robert B. Johnston landed on the shores of Mogadishu on 9 December 1992. 106 In the short run, Operation Restore Hope was a success: by threatening to use overwhelming force, the US-led coalition intimidated the Somali militias as expected. The improved security environment allowed the UN and private relief organisations to massively step up food deliveries around Mogadishu and in southern Somalia, so that 'deaths from starvation and disease fell sharply'. 107 At the same time, because risk-averse coalition forces declined to disarm local militias and steered clear of interference in Somali politics, they failed to tackle the underlying sources of the humanitarian crisis. 108 On 29 January 1993, General Johnston, under pressure from authorities in Washington, declared that US troops had accomplished their mission and recommended a transition to UNOSOM IIsignificantly earlier than the original CENTCOM plans had foreseen. 109 Boutros-Ghali, however, repeatedly delayed the transfer to UNOSOM II, insisting that the conditions were not yet right. In a 3 March report to the Security Council, the UN secretary-general pointed out that the security situation remained volatile; for that reason, he recommended a robust mandate for the follow-on mission that ought to 'include disarmament'. 110 Serious planning for the transfer could begin only after the Security Council accepted this recommendation in Resolution 814 of 26 March. 111 The new US administration under President Clinton also had to compromise on another issue: the agreement negotiated between the Bush administration and the UN secretariat in November 1992 had committed the United States to leaving a small 2500-man quick reaction force (QRF) on ships off the coast of Somalia after the handoff, to be called upon for special rescue missions; but in March 1993, the Clinton administration had to offer a beefed-up QRF and an additional 4000 US troops for logistical support, before Boutros-Ghali agreed to the transfer. 112 The transfer finally occurred on 4 May, when UNOSOM II took over and Operation Restore Hope was terminated. Yet the deployment of the strengthened UN peacekeeping force was still far from complete: at the time of the transfer, UNOSOM II comprised only about 16,000 troopssignificantly short of its authorised strength of 28,000. 113 Furthermore, the mandate for UNOSOM II, enshrined in Resolution 814, included disarming the Somali factions, as requested by Boutros-Ghali; but this put the UN troops on a path of collision with Aideed, the most recalcitrant of Somalia's warlords.
After 24 Pakistani peacekeepers were killed in an ambush by Aideed's militia in June 1993, UNOSOM II began to request frequent support from the QRF, and US forces became engaged in a manhunt for Aideed. 114 In August, lobbying by Admiral Howe, who had been appointed as the UN's new special envoy for Somalia, persuaded the Clinton administration to deploy US Army Rangers and Delta Force commandos to the country, resulting in a further escalation of the violence. 115 In short, far from achieving a smooth handoff, American troops were increasingly drawn back in, to support a mission that was now largely defined at UN headquarters in New York.
Things came to a head on 3 October 1993, when Somali militias shot down a US Black Hawk helicopter, Rangers sent on a rescue mission were ambushed, and 18 US soldiers died in the ensuing firefight. 116 A few days after the 'Black Hawk down' incident, President Clinton, under strong pressure from Congress, announced that he would pull all US troops out of Somalia over a five-month period. The hunt for Aideed was effectively over, and the UNSC formally suspended all efforts to apprehend Somali militia leaders. 117 Without American support, the UN mission was no longer viable, and Boutros-Ghali concluded that all UN forces should be withdrawn by the spring of 1995. 118 military implementation plans for humanitarian interventions including 'demonstrable milestones and measures of success'. 123 Second, another lesson learned was that, to enable a smooth handoff to UN peacekeepers, the initial US intervention needs to fit into a broader, comprehensive approach aimed at tackling the underlying political sources of the crisis. Because coalition forces in Operation Restore Hope declined to take on politically sensitive tasks, by the time of the handoff in May 1993, many of the underlying sources of the conflict continued to persist. 'The US military's view was, we'll accomplish our limited military objectives and be done', explains Colonel William Flavin, who was involved in planning for Restore Hope. By contrast, today, partially as a result of the experience in Somalia, 'the question you ask is, how are those military objectives nested inside the overall political-military objective?' 124 For Somalia, US policymakers probably placed too much hope in UN-led political reconciliation talks scheduled for early 1993. Talks held in Addis Ababa, Ethiopia, in March 1993 yielded an agreement among Somali faction leaders that committed them in principle to establishing representative political councils. This led US authorities to expect that the relatively benign security environment established by Restore Hope would continue. 125 Yet this optimism was unwarranted. As Nora Bensahel has noted, compliance with the Addis Ababa agreement was 'not at all in Aideed's interests', because although Aideed was the militarily dominant local player, he 'simply did not have the popular support necessary to guarantee a majority' in the representative councils. 126 For the Addis Ababa agreement to have a serious chance of success, international forces needed to marginalise Aideed militarilyas US defence leaders belatedly recognised in the summer of 1993.
Finally, the US military learned from Somalia that the United States should obtain clearer commitments from relevant multilateral bodies that there will indeed be a rapid handoff to adequately equipped follow-on missions. Disagreement between UN and US authorities on the objectives of Operation Restore Hope (notably, on the subject of disarming the militias) in late 1992 made Boutros-Ghali and his staff reluctant to firmly commit to a UN handoff, and it subsequently hampered planning for the transition.
On the day before American forces landed in Somalia, Boutros-Ghali had insisted in a letter to President Bush that without disarmament of the factions, he did 'not believe that it will be possible… to create conditions in which… the task of protecting humanitarian activities can safely be transferred' to UN peacekeepers. 127 After Bush declined to make disarmament part of the US-led mission, the UN secretary general repeatedly delayed the transferuntil the Americans accepted in UNSC Resolution 814 that disarming the militias would be part of the mandate for UNOSOM II, with a beefed-up US QRF to support it. 128 As a result of the Somalia experience, when it came to planning for the 1994 Haiti intervention, senior US generals and admirals pushed back against the notion that one could expect 'the UN eventually to come in', without having obtained a clear, agreed upon mandate up front with guarantees concerning the handoff. 129 By the time President Clinton sent US troops to Haiti in September 1994, planning for the follow-on force had already reached an advanced stage, and in the spring of 1995, the mission was smoothly handed off to UN forces as planned. 130 Subsequent UNSC resolutions authorising US interventions have become increasingly specific in codifying the details of UN handoffs. For instance, Resolution 1497, authorising a US-led operation in Liberia in 2003, explicitly mandated the establishment of a 'follow-on United Nations stabilization force' under regional leadership within a maximum of two months. A similar template was adopted for the Balkans (albeit with NATO at the forefront), and for several non-combat relief missions over the last 25 years. 131 The notion that UN handoffs are a central component of the US exit strategy for humanitarian interventions has taken hold within the armed services, to the point that 'it is now practically a doctrinal assumption', as a former senior defence official explains. 132 Reflecting this, the UN handoff idea is enshrined in the current US Army Field Manual and the Joint Doctrine document on multinational operations. 133 It nevertheless appears that the military 'still has issues in describing this [idea] in sufficient detail'. 134 One reason for the lack of more detailed doctrinal development in this field may be that multilateral cooperation is ultimately a matter for civilian policymaking and, as such, lies outside the military's purview. If we have not witnessed a larger number of multilateral handoffs, however, this probably has little to do with the lack of detail in current doctrine: the main reason appears to be that there have not been many US humanitarian interventions involving 'boots on the ground' to begin with. When such interventions do occur, we are likely to see the US military working hard to achieve a smooth handoff to multilateral missionsfollowing the lessons of Operation Restore Hope.