This paper has benefitted significantly from fieldwork and the insights of both survivors of the conflict and humanitarian, human rights (HR) and other personnel who were working in or on Sri Lank as the crisis peaked.
1In 2009, tens of thousands of civilians perished in the final months of the brutal Sri Lankan war that began in 1983. They died as a direct result of armed conflict and the political and military strategies of the warring parties, compounded by the international community’s collective prioritization of geo-political considerations over the action needed to curb unrestricted warfare. The United Nations (UN), as well as various other political, human rights, and humanitarian actors, failed to effectively challenge the inhumanity that led to the death of 40-70,000 civilians (Keenan, 2012; Internal Review Panel on United Nations Action in Sri Lanka (IRP), 2012: 14). In the immediate aftermath of the war, humanitarians were reluctant to confront the full scale of the protection problems inherent in the internment of some 300,000 of the war’s survivors in militarised camps (Nash, 2012: 2).
2This article focuses on humanitarian decision-making during the end phase of the war. It does not address the underlying causes and dynamics of the armed conflict, or the failure of development and human rights initiatives to address the structural fault-lines at the heart of the crisis in Sri Lanka. The first section provides a snapshot of the historical and political context that shaped the operating environment in which relief programming was implemented. The following section analyses events as the war came to its bloody end, examining the poor definition of humanitarian life-saving, the critical role of data as death and deprivation took its toll, and the consequences of a decision-making paradigm that relied on a false dichotomy between getting access to people in need and their need for protection from all-out war. The subsequent section on internment reviews the way in which the relief system grew to be implicated in assisting closed camps where many inmates were abused, even as material conditions improved over time. The final section provides an analysis of the significance of the Sri Lankan experience for humanitarian action in other crisis settings.
3Since its creation in 1948, Sri Lanka’s history has been littered with numerous pogroms that left thousands dead. It also includes two Sinhala insurrections (1971 and the late 1980s) that were brutally suppressed with great loss of life. Twenty-six years of intermittent civil war and failed peace processes produced a death toll that reached the tens of thousand by the final months of the armed conflict in 2009. At the same time, however, Sri Lanka also has a long history of harmonious relations between various indigenous groups and others who, for the majority of the island’s history, have peacefully co-existed and mutually contributed to a vibrant and multi-cultural society.
4Sri Lanka’s 450 years of colonisation altered the island’s social, economic and political systems in a manner that still reverberates today. This includes the pivotal role ethnicity has come to play in Sri Lankan society, as well as the development of a chauvinistic interpretation of Sinhala nationalism that has been championed by some Buddhist clergy since the days of the anti-colonial struggle. Racism, and the notion that ethnic minorities, such as Tamils, represent a threat to Sinhala identity is a central and corrosive feature of contemporary Sri Lankan politics.
5The state-sponsored marginalisation of the Tamil community included the disenfranchisement of Tamil plantation workers, originally imported from India by the British, shortly after independence in 1948. This pattern of systemic discrimination includes the 1956 Sinhala Only Act, restrictions on the university admission of Tamils, and a new Constitution (1972) that affirmed Buddhism as the state’s premier religion, which excluded the predominantly Hindu, Muslim, and Christian Tamils. Similarly, the state implemented several development models, such as state-sponsored irrigation, and colonisation schemes, which re-arranged the ethnic balance in some areas. This added to the Tamils’ sense of grievance and a growing mood of militancy in the north of the country.
6The Liberation Tigers of Tamil Eelam (LTTE) was launched in 1976, under the leadership of Velupillai Prabhakaran, in order to seek greater autonomy for Tamils. Prabhakaran was ruthless in his insistence that the LTTE was the sole representative of the Tamil people. The Tigers’ use of suicide attacks, their indiscriminate violence in the south of the country, and the expulsion of all Muslims from the Northern province in 1990 are but a few examples of Prabhakaran’s totalitarian approach. Prabhakaran’s strategy to end the deeply-resented anti-Tamil ethnic discrimination lacked the support and legitimacy needed to advance the quest for greater Tamil autonomy.
7The zero-sum politics espoused by Prabhakaran was mirrored in the nationalist rhetoric of the two main Sinhala political parties, the Sri Lanka Freedom Party (SLFP) and the United National Party (UNP). Both parties frequently allied themselves with extreme Sinhala groups to undermine conflict resolution processes and win elections.
8Exclusivist identity politics has thus long fuelled opposing narratives of nation and victimhood, weakening support for a pluralist, inclusive and democratic Sri Lanka. These factors significantly complicated the dynamics of the war and efforts to address its humanitarian consequences. Changes in global power relations, spurred in particular by the GWOT, coupled with the geo-political agendas of regional and other actors, also played a critical role in the end phase of the war.
9The LTTE had a history of overcoming setbacks, but Prabhakaran failed to understand the implications of 9/11 for both global politics and the resource mobilisation tactics of the Tigers (Anderson, 2011). As a result, in 2004, the LTTE suffered serious setbacks when a breakaway faction sided with Colombo (Anderson, 2011). The devastating 2004 Indian Ocean Tsunami further hindered the organization, as it brought a massive influx of aid that divided the community and became a deciding factor in the 2005 presidential elections, which at Prabhakaran’s urging many Tamils boycotted (Feinstein, 2007). This boycott, in combination with the aid, benefitted Mahinda Rajapakse, who won with a slim majority on an anti-Western, ‘war for peace’ platform (Nalapat, 2011).
10Rajapakse also benefitted from the increasing demonization of the LTTE, which was banned by numerous countries that chose to overlook Colombo’s reciprocal use of terror. The Sri Lankan government had an abysmal human rights record, characterized by the erosion of the independence of the judiciary, significant limitations upon the freedom of the media and a system of governance that exhibited all the hallmarks of an elected tyranny (Isaac, 2010). The Rajapakse regime was adept in its use of GWOT narratives, effectively labelling the government’s participation in an enduring civil war as a counter-terror campaign (Macae, 2012). Rajapakse and various pro-government media outlets were vociferous in their articulation of an anti-Western rhetoric, which in turn was well received by hard-line nationalists and segments of the Buddhist clergy (Walton, 2012). This polemic depicted human rights as a tool of Western hegemony, discredited by double standards, and a threat to national sovereignty. Rajapakse was anxious to deflect calls for a ceasefire and avoid a Responsibility to Protect (R2P) military intervention that, he argued, would likely result in civilian casualties (Thottam, 2009).
11The Rajapakse regime may be likened to a seasoned poker player, extremely sensitive to shifts in global and regional power dynamics (Bhadrakumar, 2009). Interlocking and competing Beijing-Delhi agendas meant that the government benefitted enormously when, in the 1990s, India switched its allegiance and provision of military and political support from the Tigers to Colombo (Senanayake, 2009). Other important suppliers of arms and military technology included Israel, Iran, Pakistan, Russia, the United Kingdom (UK) and other European Union (EU) states, while the United States provided counter-insurgency training (Harneis, 2010). Similarly, China emerged as a key player in Sri Lankan politics in the final years of the war, when it proved itself a steadfast ally of Colombo in the UN Security Council (SC) and other fora (Bhadrakumar, 2009; Harneis, 2010). At the same time, in light of Colombo’s portrayal of the US and its allies as pro-LTTE, the Western powers lost much of their Cold War era influence in Sri Lanka.
12Sri Lanka’s troubled history of state formation history, ethnicity and racism-driven politics, deep-seated grievances and narratives of victimhood, played out in a context of the GWOT and competing geo-political agendas, all contributed to the operating environment of the relief system during the final months of the war. As the war intensified in the Vanni, the last stretch of territory held by the LTTE in northern Sri Lanka, Colombo’s anti-Western rhetoric and vilification of humanitarian actors worked to significantly complicate the task of saving lives.
13The inhumanity of the war in the Vanni – the steady slaughter of a besieged population held against its will in a diminishing strip of coastal shrub-land – was shocking. The targeted and wilful killing of civilians is itself, unfortunately, not uncommon in conflict situations, but the relief community’s failure to leverage its knowledge of the atrocities to elicit meaningful action was. A few relief workers, including national staff trapped in the Vanni, strove to call attention to the bloodshed, but the humanitarian system lacked the muscle to challenge the broad consensus amongst key UN member states ‘that war should be given a chance’ (Good Humanitarian Donorship [GHD], 2011). The groundbreaking report of the UN Secretary-General’s Internal Review Panel (IRP) eventually concluded that the events in Sri Lanka marked a grave and systemic failure of crisis management, human rights, and humanitarian action (IRP, 2012: 28).
14The aid community had significant experience in Sri Lanka, but was unprepared for the rapid escalation that marked the end of the war. Relief actors had worked with both sets of warring parties. They knew the 2004 LTTE split had been detrimental to the Tigers. They also knew that Rajapakse’s strengthened government forces had successfully pushed Prabhakaran’s fighters out of the Eastern Province between mid-2006 and mid-2007. Rajapakse had promised his Sinhala national alliance that he would crush the Tigers. In doing so, he formally opted out of the 2002 ceasefire agreement in January 2008, from which time Colombo shifted its focus to the Northern Province and the remaining LTTE-controlled territory in the Vanni (International Crisis Group [ICG], 2011). The Sri Lankan Army (SLA) captured Kilinochchi, the de facto capital of the LTTE, in January 2009.
15Notwithstanding this change in circumstances, the Common Humanitarian Action Plan (CHAP), which articulated the overall objectives, strategy and approach of the relief system, reflected a business-as-usual approach for 2008. It focused on those displaced as a result of the war in the east, and made no reference to either increasing preparedness or potential changes in the Vanni humanitarian caseload (CHAP, 2008).
16Interviews conducted by this author with a cross-section of field, headquarter relief, diaspora and other personnel in Sri Lanka and elsewhere in 2012 indicate that poor analysis of the changing situation was compounded by a dysfunctional relief coordination structure, a lack of unity within the humanitarian system, and a reactive approach to the unfolding crisis. Agencies with protection-specific responsibilities – namely, a duty to oppose policies and practices that impair the safety, well-being, and dignity of crisis-affected groups – such as the UN High Commissioner for Refugees (UNHCR), were seen as ‘in bed with the government’ and reluctant to challenge egregious instances of harm (Keen, 2009: 81). The World Food Programme (WFP), one of the most high profile agencies in Sri Lanka, kept its distance from non-food issues generally ‘and from protection issues in particular’ (Keen, 2009: 93). UNICEF and other UN agencies were also perceived as inadequately committed to addressing contentious issues.
17The government’s Consultative Committee on Humanitarian Assistance (CCHA) was widely seen as an Orwellian structure that served as an instrument of Colombo’s propaganda in a sophisticated campaign to intimidate and emasculate the relief community. Colombo used visa and travel permits, the threat of expulsion from the country, character assassination and allegations of pro-Tiger sympathies to coerce and control the relief system. Such practices of intimidation included the presence of ‘uniformed and armed soldiers’ in field coordination meetings (Keen, 2009: 52). Threats to staff safety, particularly national staff, were a valid concern. In August 2006, 17 Action Contre la Faim (ACF) local staff were murdered in an execution-style killing in their agency compound shortly after the SLA retook Muttur in the Eastern province. The aid community protested the killing of their colleagues, but did not seize the potential watershed moment to re-negotiate the space and parameters of humanitarian action. The relief system also failed to devise a collaborative strategy for contesting Colombo’s dangerous discursive distortion of humanitarian values that imperilled both those in need of aid and attempting to provide it. As a result, as the fighting intensified, the manipulation of humanitarian action by both warring parties for their own strategic goals, on and off the battlefield, was a key factor in the intensification of suffering and death.
18In the Vanni, the Tigers held a monopoly on the use of force, while the government controlled access to several essential services, including health care. This arrangement meant that the warring parties were keen to dominate and exploit humanitarian programming as part of a wider struggle for leverage and legitimacy. In the final months of the armed conflict, the besieged population was used for strategic advantage by both sides. In the process, each party also used aid convoys to shelter or camouflage their military movements, while vehemently denying doing so (Weiss, 2011: 105).
19The Tigers used the population, including children, as a ready recruitment and conscription pool. As the war moved into its final stage, the LTTE leadership calculated, erroneously, that mass civilian deaths would oblige the international community to intervene and, in the process, provide some respite to its fighters. This explains, in part, the LTTE’s determination to hold onto frightened civilians who, in turn, were used to slow the advance of the SLA (ICG, 2010: 24-26). For all intents and purposes, this strategy was accepted by the diaspora, an important source of Tiger funding and support.
20The Rajapakse regime’s manipulation of the relief system, to the detriment of life-saving action, played out in different ways. In order to limit the availability of critical supplies, Colombo greatly restricted relief personnel’s access to people in need. This allowed it to manipulate population movements and provoke an exodus from the Vanni (Keen, 2009: 55). The government greatly restricted the supply of essential medicines, but as it was eager to avoid provoking international concern, it allowed a small amount of food to get through (Keen, 2009: 63-70). Colombo thus disingenuously engaged in endless discussions on access to humanitarian aid in order to ‘demonstrate its good intentions’, enabling it to distract the relief system while simultaneously presenting its military campaign as a ‘humanitarian hostage rescue’ exercise (Keen, 2009: 55)
21The general reluctance of the relief system to challenge the instrumentalisation of humanitarian action, the Machiavellian tactics of the government, including its co-option of the relief narrative for military objectives, and the disruption of live-saving programmes all set the scene for the end stages of the war. Perennially on the defensive, the relief system further compromised its credibility, as well as its practical capacity to deliver on its responsibilities, when it failed to prioritise the safety and protection of those exposed to unrestricted warfare.
22Colombo advised the UN on 3 September 2008 that it could not guarantee the safety of UN staff after several artillery shells hit Kilinochchi, a town the warring parties had agreed to respect as a conflict-free zone. Many relief staff were upset when the UN subsequently ‘fell into line without a fight’, despite clear evidence that the crisis was deepening and needs were increasing (University Teacher of Human Rights Jaffna [UTHR-J], 2008). The departure of humanitarian agencies, with the exception of the International Committee of the Red Cross (ICRC) and Caritas, served only to compound the vulnerability of those in the Vanni. The UN was also aware that the government had ‘been bombing and shelling its civilians for over 20 years’ and that Colombo’s pledges ‘to ensure zero civilian casualties’ had no credibility (UTHR-J, 2008). Nonetheless, the office of the UN Secretary-General (UNSG), Ban Ki Moon, issued a statement that failed to challenge Colombo’s ‘request for the relocation of UN humanitarian staff’, made no mention of the SLA shelling of Kilinochchi, and voiced no specific concerns regarding the safety of national staff and other civilians unable to leave the Vanni. Instead, the statement merely reminded the warring parties of their obligations under international humanitarian law (United Nations Secretary General Report [UNSG], 2008).
23The passive approach of the relief system to this particular crisis highlights Sri Lanka’s status as a critical and persistent failure of the humanitarian community. Although, since the end of the Cold War, both individual organisations and the relief system have invested heavily in their capacity to address the protection issues in crisis settings, there was little evidence of this in Sri Lanka. At the time, the relief system largely failed to acknowledge that the tactics and practices of unrestricted warfare designed to increase and exploit the suffering of those in the Vanni were matters that fell under the mandate of humanitarian agencies (IRP, 2012: 28). Issues concerning the physical safety of civilians were, nearly without exception, seen as too complicated to address. Such situations were ‘described as political’, and therefore beyond the remit of the relief system, despite the fact that, in any given context, humanitarian actions invariably have political implications (IRP, 2012: 19).
24As the war entered its final stage, the biggest threat to the people in the Vanni was the military strategy of the warring parties, characterised by contempt for the suffering of civilians. The relief system was aware of the political dynamics shaping the life-threatening crisis, but stuck to a response model with limited relevance to the policies and practices driving up the death toll.
25The bulk of the relief system withdrew from the Vanni in September without protest, becoming almost immediately consumed with protracted negotiations to regain access for food delivery convoys. There was, of course, a food shortage in the Vanni, but there was a much greater need to counter the unrestrained warfare and policies that restricted the availability of essentials such as anaesthetics, antibiotics and paediatric medicines (HRW, 2008: 32). The relief system’s preoccupation with convoys and the organisation of camps for Internally Displaced Persons (IDPs) resulted in an ‘excessive proportion’ of time and energy invested in addressing material, ‘compared to protection’, needs (Keen, 2009: 62).
26Whatever the claims of individual agencies, the prioritisation of material over non-material needs is the norm, rather than the exception, in most crisis settings. The combination of policies, tactics, and narratives used to justify no-holds-barred warfare, however, coupled with the reluctance of the relief system to take a principled and assertive stance on the inhumanity of the war, makes Sri Lankan case unique. Although the humanitarian community devised mechanisms to identify overall contextual objectives, and measures to evaluate the results, it did not similarly develop a unified protection strategy. The 2009 CHAP made no reference to the protection problems inherent to the deepening crisis in the Vanni, the practice of indiscriminate warfare, the rising death toll, or the deliberate blockage of life-saving medicines.
27Opinion among those interviewed for this article varied as to the appropriateness and utility of the overall relief approach. One interlocutor noted that there was an unending preoccupation ‘with tents and tarpaulins’, while some argued that, despite the fact that the convoys had ceased by January 2009, attention to protection would jeopardise access. Others indicated that a narrow focus on maintaining a post-war presence and the provision of support for planned IDP camps deflected attention from protection issues. In addition, as the conflict escalated, there was significant confusion as to what, precisely, protection might mean in the context of humanitarian action. The relief system’s review of its 2008 programme, for example, made no reference to war-related death and injuries, but did highlight the value of child-friendly spaces and children’s clubs (CHAP, 2009: 7).
28There was broad consensus among both field and headquarters staff that humanitarian workers were scapegoated for the failure of others, including the mutually-reinforcing relationship between development models and societal divisions in Sri Lanka and the poor performance of inter-governmental human rights and conflict resolution mechanisms. A number of respondents stressed the fact that Western and other states, despite routine reiterations of humanitarian and human rights standards, were aligned with Colombo’s GWOT positioning and keen to see the end, however bloody, of the LTTE. Beijing, which habitually claims to oppose interference in the internal affairs of other states, used its Security Council seat to block UN action, despite the fact that this policy clearly favoured one side over the other. New Delhi, similarly, played a particularly cynical role, making ‘pro-forma admonitions’ of Colombo’s violence while simultaneously encouraging and facilitating the rapid demise of the LTTE (Nalapat, 2011; Gokhale, 2009).
29Against this background, most relief actors argued that, access and security permitting, it made sense to focus on practical issues such as camp construction or convoys. They emphasised that the lack of support for the humanitarian imperative – the prioritisation of lifesaving over all other considerations – meant that there was limited value in challenging the inhumanity that drove the unfoldingcatastrophe. Similarly, pushing for recognition of the protected status of civilians under international humanitarian law was not perceived as a practical or tangible contribution to the material measures needed to help the besieged survive the war.
30The general reluctance of the relief system, with the exception of the ICRC, to assertively address the practice of indiscriminate warfare was compounded by the inability of humanitarian actors, following their withdrawal from the Vanni, to determine the size and nature of the humanitarian caseload. By mid-2008, nearly the entire population of the Vanni was dependent upon the humanitarian system for their survival. In early 2009, a UN official estimated that only 50 per centof the relief goods required had been provided during the previous year (Keen, 2009: 58).
31An accurate assessment as to the number of vulnerable persons and the nature of their non-material and material needs is fundamental to effective humanitarian planning and impartial programming. In Sri Lanka’s contested operating environment, given Colombo’s objective to remove people from the Vanni, such figures also had strategic significance, (Weiss, 2011:113). The UN used an array of population estimates that ranged from 400,000 in November 2008 to 250,000 in February 2009 (Weiss, 2011:178; IRP, 2012: 39). The Catholic Diocese of Mannar, using local government data, concluded that there were 429,059 people in the Vanni in October 2008 (IRP, 2012: 38). Colombo, conversely, claimed that there were ‘no more than 50,000 civilians’ inside the war zone (Weiss, 2011: 178). When the conflict and the rate of killing intensified in 2009, Colombo used its self-determined population numbers ‘to rebut allegations of civilian deaths’ (Report of the Secretary-General’s Panel of Experts on Accountability [POE], 2011: 39). Many interviewees argued that the use of artificially-reduced population estimates was yet another signal of the government’s belief that the relief system was malleable and easily quarantined.
32A critical feature of the end phase of the hostilities was the use of the besieged population by both parties to advance their war aims. Many civilians in the Vanni were fearful of crossing into government-held territory, while, at the same time, the Tigers used intimidation and coercion to hold onto civilians even as the LTTE-controlled areas shrunk to a narrow coastal spit of sand (Weiss, 2011: 181). The government’s unilateral declaration of so-called No-Fire Zones (NFZ), used to concentrate civilians into designated areas that were subsequently bombarded, was brutal in its deception and execution (POE, 2011: ii). Colombo announced the first NFZ as part of its purported zero-civilian casualty policy in January 2009, shortly after the Tigers began retreating eastward from Kilinochchi. Frightened civilians crowded into the NFZ, where they had been assured they would find ‘sanctuary’ (Weiss, 2009: 112). As the SLA advanced and frontlines changed, Colombo announced a second NFZ in February and a third in early May. It was quickly apparent, however, that the Zones were Orwellian constructs that ‘had little to do with protecting lives’, as their primary purpose was to advance the SLA campaign (Weiss, 2009: 113). It soon became conventional wisdom ‘that government-marked safe zones were the least safe’ areas (UTHR-J, 2009: 34).
33Survivors of the NFZs told this author of devastating shelling, and the decision of one family to take an injured child out of hospital upon the realisation that such structures were particularly prone to artillery strikes. A UN Panel of Experts concluded that Colombo ‘systematically shelled hospitals’, to the extent that such structures were repeatedly ‘hit by mortars and artillery…despite the fact that their locations were well-known to the Government’ (POE, 2011: ii). In February, the SLA repeatedly shelled the second NFZ ‘from all directions, including land, air and sea’, even though it was packed with civilians (2011: 28). Though the ICRC and doctors in the Vanni reported on the deteriorating situation, Colombo invariably challenged the validity of such reports, denying access to independent journalists and framing calls to curb the fighting as evidence of pro-Tiger and pro-terror sympathies (Perera, 2009).
34The UN and the diplomatic community in Colombo were aware of the catastrophic situation in the Vanni. They had access to satellite imagery, a constant stream of information from medical personnel and other individuals in the warzone, and the eyewitness account of a combat-experienced UN international security staff member, Harun Khan. He led the last UN convoy into the Vanni in mid-January, when he had to seek shelter from heavy shelling at a UN-designated area in the first NFZ. He made numerous calls to SLA headquarters to demand a halt to the barrage, but to little effect (Weiss, 2009: 106-120). Following M. Khan’s return to Colombo, UN security staff established a dedicated monitoring team to track and analyse events in the Vanni. It soon concluded that the majority of casualties were the result of government fire (IRP, 2012: 72). By the end of January, the humanitarian, human rights, and politics arms of UN were engaged in an increasingly vocal behind-the-scenes campaign against the attacks on civilians. The UN was more comfortable highlighting the problems posed by the LTTE, however, than challenging the government’s narrative or its determined disregard for the laws of war and humanitarian norms (IRP, 2012: 72).
35The primary task of the humanitarian system in Sri Lanka was to help those in immediate danger to survive the crisis, but few relief actors were willing to risk the fallout likely to stem from a confrontation with the government. There also appears to have been a poor understanding of the essence of humanitarian action, particularly in relation to non-material needs. According to the UN’s internal report, staff did not adequately appreciate their responsibility to pre-empt the killing of civilians. It found that the overall approach to the government ‘amounted to a failure by the UN to act within the scope of institutional mandates to meet protection responsibilities’ (IRP, 2012: 27). Louise Arbour, former UN High Commissioner for Human Rights, concluded that the UN’s muted approach ‘verged on complicity’ (Weiss, 2009: 140). The Inter-Agency Standing Committee, the primary mechanism for relief system coordination, did little to challenge the total-war agenda that effectively made pawns of those in the Vanni. The open practice of unrestricted warfare also made a mockery of the UNSC’s decade-old initiative to address the protection needs of civilians in conflict settings.
36The UN system, including the Secretary-General, various heads of agencies and the Emergency Relief Coordinator, had irrefutable evidence that the so-called safe zones were subject to routine and indiscriminate shelling, in clear violation of humanitarian law. The UN was aware that time was running out for those trapped in the NFZ, but persisted with its strategy of quiet diplomacy and periodic public announcements calling for an end to the use of heavy weapons and indiscriminate attacks. The UN also knew that, even as it took extreme measures to intimidate or silence the media and others critical of the Rajapakse regime, the government was sensitive to attacks upon its reputation. At a meeting of the UNSG’s Policy Committee in March, however, and in subsequent correspondence between senior UN leadership, several senior officials questioned the legitimacy of what the UN’s intelligence on the Vanni, expressing concern about upsetting Colombo authorities by placing the LTTE and the government ‘on the same footing’ (IRP, 2012: 66-68). As a result, the UNSG and his senior team allowed debate regarding the actual number of people killed to prevail, diverting attention from the action needed to curb the mounting death toll. When the UN downplayed its knowledge of the impact of the war on civilians, it effectively joined the ‘diplomatic dance’ of the Security Council members, expressing concern about the killings in the Vanni while simultaneously backing Rajapakse’s war agenda (Macrae, 2010).
37The UN humanitarian chief and Under-Secretary General for Humanitarian Affairs John Holmes did not mention specific countries when he referenced the effects of their diplomatic footwork but, presumably, given their general ambivalence about the killings in the Vanni as they awaited the ‘inevitable defeat of the LTTE’, he had the permanent five members of the UNSC in mind (Macrae, 2010). The French and British Foreign Ministers visited Colombo in April 2009, as it became clear the war was ending, and called for a ceasefire. President Obama waited until a few days before the end of hostilities to call on the LTTE to surrender and ask Colombo to cease the ‘indiscriminate shelling’ (Whatley, 2009). The muted advocacy and public handwringing of various high profile personalities included that of India’s Foreign Affairs Minister, Pranab Mukherjee. He met President Rajapakse in January 2009 and, at that meeting, was apparently reassured that Colombo ‘would respect the safe zones and minimise the effects of conflict on Tamil citizens’ (CNN, 2009).
38Just as truth is often the first casualty of war, humanitarian actors who failed to be upfront about the inhumanity of unrestricted warfare were complicit in the widespread indifference to the death and suffering of those trapped in the Vanni. In this way, the experience of the war set the tone for its aftermath.
39Though the battle for the Vanni came to a bloody end on 18 May 2009, the death and suffering did not stop. The war’s survivors weredetained and confined to closed, militarised camps that were constructed and maintained with the help of the humanitarian community.
40The relief system had ample time to draft its post-conflict strategy for action outside the Vanni, and spent months planning for an anticipated mass exodus that, in the immediate aftermath of the war, eventually reached 300,000 (Nash, 2012: 2). Relief actors knew that a few hundred of those who fled the Vanni in 2008 were held against their will in closed camps. Agencies thus were rightly wary of legitimising unlawful detention, and provided only limited services to such camps.
41When Colombo announced that it intended to clear large tracts of land for camps in an area known as Manik Farm, the UNHCR developed a useful Aide-Mémoire to share with the government. This documentclarified the basis for humanitarian engagement and acknowledged the necessity of developing a formal screening process to separate combatants from civilians. It emphasised that camps should be civilian in character, and that both residents and relief actors should enjoy unrestricted movement.
42When, beginning in April 2009, those who survived the war sought succour in large numbers, it soon became apparent that Colombo was not responding to the situation as a crisis facing vulnerable, displaced people. Government authorities insisted on separating family members and pursuing a non-transparent screening process. Some of those who were screened were executed.More than 70,000 people remained unaccounted for at the end of 2012 (IRP, 2012: 14). There were numerous reports of sexual violence, including rape, as endemic to the camps. Many of those interrogated were dispatched to so-called surrender camps – another, given the lack of evidence of individuals volunteering to surrender, of Colombo’s Orwellian terms (POE, 2011: 44).
43Relief actors knew that the camps were little more than an oppressive tool that, while meeting some of the material needs of their captive population, flouted basic humanitarian and human rights norms. Nonetheless, even when two UN national staff members were abducted, tortured and heldincommunicado until they were accidentally discovered in a prison in Colombo,relief agencies pursued a business-as-usual approach(Lee, 2009). The humanitarian system undermined its own ability to give effect to its principled objectives when, once again, it prioritised its working relationship with the government over meaningfully challenging the abuse and confinement of those who survived the war.
44Relief actors persisted in relating to survivors as IDPs, rather than tackling the protection issues posed by their effective internment. Humanitarians abandoned their joint action strategy in favour of unilateral programming. Though there was some productive lobbying aimed at limiting the duration of internment, for the most part, agencies were preoccupied with material supplies. This, as before, undermined the ability of the humanitarian agencies, to the great detriment of the internees, to effectively tackle the system of terror that ruled the camps.
45The experience of the humanitarian community in Sri Lanka represents a collective and systemic failure. Its actions or, more accurately, lack of action, cannot be separated from the loss of thousands of lives. A small number of agencies and individuals were steadfast in challenging the inhumanity of the conflict in the Vanni, but the overall humanitarian platform was characterised by ethical confusion, the unaccountable de-prioritisation of protection problems, and pusillanimity. The failure of the humanitarian community to conduct a collective and dedicated reflection upon the results of its intervention in Sri Lanka does not bode well for the future of principled and effective humanitarian action.
46Much of the death and suffering of the Sri Lankan conflict can be attributed to the warring parties and their supporters, but people also died as a result of the relief system’s failure to acknowledge its core responsibility of saving those in immediate danger. The relief system ignored the lessons learned from previous mass atrocities, becoming a tool in the execution of agendas antagonistic to the values of humanitarianism. Specific instances of such failures included the relief system’s uncontested departure from the Vanni, the withholding of important evidence regarding the dead and the dying, and the provision of predominantly uncritical support for internment.
47Sri Lanka shows what happens when UN humanitarian entities and relief system mechanisms – the humanitarian country team, global protection cluster, and IASC – abdicate their protection responsibilities and sideline principled and protective humanitarian action. The relief system’s anaemic approach to protective humanitarian action reflects, in part, the corporate concerns of agencies that are in danger of becoming too-big-for-principles, particularly in instances where protective interventions might put long-term programming at risk. In settings where humanitarians ought to be indignant and inspirational in the face of inhumane practices, decision-making needs to be shaped by considerations that prioritise the safety and survival of endangered groups
48The relief system will continue to stumble unless it devises a more effective strategy for coping with narratives and agendas that rationalise or misrepresent the drivers, dynamics and human costs of contemporary crises. It also needs to acknowledge that organised humanitarianism is strongly informed by its Western origins. The space for humanitarian action is thus often proscribed by the perception that humanitarian programming plays an integral role in Western-led interventions associated with regime change and hegemonic agendas. Unless the relief community insulates itself from geopolitical agendas and impartially engages with all parties to a conflict, it will remain a victim of the shifting shoals of purported anti-terror campaigns and anti-Western discourses.
49Ignoring the lessons of Sri Lanka will likely mean the same mistakes are repeated elsewhere.
50Having worked with Norah Niland when I was the UN Emergency Relief Coordinator from 2007 to 2010, I have always admired her commitment to human rights and principled humanitarian action. I also believe that we need to closely examine case studies such as Sri Lanka to see what lessons we can learn. There was certainly a failure to prevent a terrible tragedy for many thousands of Tamil civilians in the first half of 2009. However, I cannot allow Norah’s piece ‘Sri Lanka: Unrestricted Warfare and Limited Protective Humanitarian Action’to pass unchallenged. While her paper makes many good points, I believe that overall it is in danger of contributing to a mythology surrounding the 2009 events in Sri Lanka. This does not only scant justice to the efforts of many humanitarian actors at the time, but more importantly it risks distorting future action by giving undue primacy to advocacy and a particular interpretation of human rights and protection principles above all other concerns. In my view, the internal UN report on Sri Lanka by Charles Petrie (UN Secretary General’s Internal Review Panel) fell into the same trap. Norah draws very heavily on this report, as if it is the only truth, as well as on the highly critical views of a few individuals from inside the humanitarian system who have dominated the recent public record. This is unbalanced and - if the wrong lessons are learned - I worry about what it might mean for future action by the UN when faced with similar intractable dilemmas.
51I was the Emergency Relief Coordinator in the period Norah describes. My colleagues and I spent more time agonising over what was happening in Sri Lanka and what we should do in response than on any other issue in my three and a half years in the job. I visited the country five times, and had many hours of discussions with the country’s leadership and the humanitarian community there, ably led, in the most difficult of circumstances, by Neil Buhne, the UN Resident/Humanitarian Coordinator at the time. I have been heavily criticised by both Sri Lankan parties (the government and the LTTE), at different times, for supposedly being partisan, but of course was as careful as I possibly could be to avoid taking a political side.
52I take full responsibility for whatever the humanitarian community did and did not do at the time. I continue to think that we did a much better job than the conventional wisdom now seems to have concluded. My aim is not to say that we got everything right – we surely made many mistakes, particularly with the benefit of hindsight. Rather, I want to argue that the debate needs to avoid sweeping and often unsupported statements about what happened (including incidentally, sometimes, those regarding the actions of the Sri Lankan government themselves). I want to take a much closer and hard-headed look at why we took the approach we did, and whether we could have done things differently. Above all, I wish to look at whether it would have had a more positive impact if we had behaved differently, before reaching far-reaching conclusions for the future.
53Norah tries harder than many to avoid the trap of actually blaming the humanitarian community for what happened, rather than the protagonists in the military confrontation on both sides, or the wider international community for its failure to get engaged or intervene in what happened to the Tamil civilian population. She talks at one point of the widespread view among those involved that humanitarian workers were scapegoated for the failures of others. But, while accepting at some points that neither the government nor the international community was listening, she nevertheless contributes to the scapegoating: she suggests that this tragedy could have been averted or significantly mitigated if the humanitarian community had only given greater primacy to protection of the civilian population, and asserted more strongly and publicly the overwhelming need to stop the disaster that overtook so many of the Tamil civilians effectively taken hostage by the LTTE. For example, she states at one point that humanitarian actors who failed to be upfront about the inhumanity of unrestricted warfare were complicit in the widespread indifference to the death and suffering of those trapped in the Vanni.
54This is the nub of the argument. The reality is that the international media and the international community did in fact know exactly what was going on in the Vanni in the later stages of the war – mostly because we in the humanitarian community were constantly telling them. Lack of action was not due to lack of information. A look back at the press reports of the time will amply confirm this.
55A particular example of the debate is the long-running dispute about whether we could and should have released or highlighted more casualty figures at a late stage in the war. There may be a legitimate argument here about whether we were right to stop issuing our own figures. But much of the polemic simply ignores the fact that us shouting more about the figures, whether or not accurate, would not and could not have made a real difference, for two reasons. Firstly, similar figures were being released by others anyway. Secondly, and more importantly, no-one in a position to do anything about it really wanted to know any more than they did already. Everyone knew that civilians were being killed in large numbers, through a combination of the LTTE’s cynical use of them as human shields and of the government indifference (at best) to their fate. In the ongoing Syrian conflict, exact casualty figures have never been lacking but have not forced action, despite the obvious horrors happening there.
56It is obviously impossible to say with certainty what would have happened if the humanitarian reaction had been different. But I would argue that there is little to no evidence to support the view that the Sri Lankan government would have listened and changed its approach if we had been even more forceful/shrill in public and taken a more exclusively protection-focussed approach, or that the international community would have put greater pressure on them to do so. Plenty of evidence points to the contrary. My own extensive dealings with the Sri Lankan government gave me the conviction that - while we could and did achieve some improvements at the margins by engaging with them over their tactics in various areas - they were in no way prepared to deviate from their goal of destroying the LTTE militarily, whatever the civilian cost, now that they had the chance to do so. Neither we nor the international community more widely had the influence or the arguments to stop them (in fact most of the international community had no problem with this LTTE-related goal, whether or not they worried about the methods being used). Not being able to help everyone was not a reason for not helping those we could.
57Nevertheless, the question clearly arises as to whether this pragmatic argument is enough. Did we not have the responsibility to speak out at maximum volume about the protection concerns and the horrors of the war, whether or not we thought this would make any difference, regardless of the consequences for the rest of the humanitarian operation? Were we not guilty of compromising our principles just to stay there and help where we could, and therefore, as some have suggested, guilty of an unacceptable trade-off? Let me address this in two ways.
58First, we did not in fact keep quiet about the facts simply to protect the operation. We did tell the world regularly what we knew, as I have already suggested. We did not keep quiet about the casualty statistics, when we stopped publicising them, for fear of an adverse government reaction, as some have since alleged, but because we could not stand behind them with confidence. There was no simple calculated trade-off or unprincipled yielding to undoubted government attempts to intimidate us.
59What did drive our decisions was the considered view that, while we could almost certainly not halt the war or change the brutal tactics of either the government or the LTTE (though we never stopped trying to achieve both), we could influence some of their behaviour towards the Tamil civilians, and force them to take more account of international humanitarian norms. We also believed that by remaining engaged we could provide real, practical assistance to the Tamils ourselves, at a time when they desperately needed it. The alternative of leaving them entirely to the mercy of a hostile or uncaring government seemed, and seems to me still, unacceptable in principle as well as practice.
60I am sure that in practice we did have a useful effect on the government in a number of important areas, such as:
61their reluctant permission for at least some aid to reach the war zone through the ICRC, and the evacuation of wounded civilians;
62their preparations for the eventual exodus of the Tamil civilians, imperfect though these remained;
63their reluctant willingness to allow international NGOs to work in the camps and in the north of the country after the end of the war;
64their gradual relaxation of the conditions in the camps, and commitment to the return of those held there;
65their reluctant willingness to move with a degree of speed towards normalisation of the north after the war.
66Secondly, the kind of account given by Norah and others is in my view unduly dismissive of the importance of the aid we were able to provide ourselves, particularly after the war. I do not think we were obsessed by tents and tarpaulins, as someone apparently claimed, but the Tamils who emerged at the end of the conflict were in terrible shape and badly needed help in all kinds of practical ways – health, nutrition, shelter, emergency education for the children, and so on. If we had not been there, working with the government, they would still have received some aid, but it would have been much poorer in every way than that they received from us, and with us on the spot. This should not simply be dismissed as of little or no value compared to the protection concerns, not least because the fact that we were there, engaging practically with the authorities on a daily basis, enabled us also to improve protection of those concerned from abuses in a myriad of ways, including the facilitation of contact with their relatives. If we had not been able to help the dead, at least we were able to help those who had survived in significant ways.
67Norah’s account of what happened with respect to the camps in particular does not do justice to what the humanitarian community achieved. When the authorities first began to build a camp in anticipation of the end of the war, there is no doubt that the kind of shelters they were building looked as though they were intended to house people for a very long time. Stories that these were really internment camps, presaging an attempt by the government to repopulate large parts of northern Sri Lanka with Sinhalese, were rife. I do not know whether the government, or parts of it, ever really had that intention, but the strong protests of the humanitarian community against any such policy did seem to have the effect of inhibiting any such intentions. The next stages of the camps, built with the assistance of the humanitarian community, were much more obviously temporary.
68At the end of the war, when the trapped Tamil civilians had finally poured out, there were still many claims that the government would keep them locked up for years. The conditions of the camps were certainly unacceptable in principle and practice, with razor wire around them, the military surrounding and patrolling them, no freedom of movement, little or no communication with relatives, and informers being taken round regularly to identify LTTE cadres supposedly hiding among the camp populations. We debated several times whether we should have nothing to do with the camps in such circumstances. But we always came back to the view that we could do more good by staying engaged, and demanding improvements. In addition, we were also advised by the UN Secretary-General’s then-adviser on IDP rights, Walter Kaelin, that what the government was doing in terms of security precautions around the camps in the immediate aftermath of the fighting was not against international law, although the longer it went on, the less true that would be.
69Our response was therefore to hang on in the camps, while pressing the government at every opportunity to take away the wire and the soldiers, allow freedom of movement, and begin to make preparations for returns as soon as possible. We made it clear that we would not stay to help if they did not show significant, time-bound progress in these areas. Our efforts were gradually rewarded, both in gradually improving terms and conditions for those in the camps, and later the emptying of the camps much more quickly than most had dared to hope – indeed, at one point, we had to ask the government to move people out less rapidly because the required mine clearance had not been adequately done, and the conditions for acceptable returns had not been met. We consistently refused, incidentally, to help the government with camps for the ‘surrendees’ because we did not believe that conditions for us to do so in those cases were met, for example as long as the government would not allow the ICRC to have access to them.
70Again, the crucial question is whether we could have provided this help and if we could have had as much influence on the government in such areas if we had focussed more exclusively on the stopping of the war and the protection concerns of those caught up in it, in the way it has been suggested we should. I do not believe so. To repeat, this government was (and is) highly resistant to public attempts to pressure them, and inclined to ignore entirely those who do this to the exclusion of efforts to work with them in practical ways, and those who focus more on making a public splash than on trying to get their arguments accepted in private. That was the balance we were attempting to strike throughout that period – not disregarding or keeping quiet about terrible things that were happening, but at the same time working in a multitude of ways, including behind closed doors, to improve the situation. This included incidentally using those inside the government, ministers as well as officials, who sympathised with our agenda and views. It is wrong to portray the government of the time as monolithic.
71It seems to me that this is the responsibility of the humanitarian community in situations like this. While essentially motivated by human rights concerns, we are not in the same position as the human rights community who can speak out publicly in countries where they have no effective presence on the ground, without jeopardising anything practical they are engaged in. We had other practical responsibilities we needed to take seriously, which themselves contributed to protection, as well as our fundamental protection concerns.
72Does this mean there are no red lines for the humanitarian community, which if crossed, would provoke a refusal to work with the government at all? I think that such lines must exist, but that in the Sri Lankan case, they were not crossed, although we were very close to it at times. In other words I strongly believe that we did more good by staying and helping the Tamil community wherever we could than we could ever have done through hopeless and likely counterproductive, but principled, shouting from the rooftops, and refusing to engage with an unprincipled but strong and determined government.
73All this is a debate that will no doubt continue. I hope it does, not least because it will help keep up the pressure for an independent investigation of what really happened during the last stages of this brutal war, and help all those who still do not know what happened to their loved ones. But I also hope that there will continue to be a fair representation of what the humanitarian community were trying to do, and indeed achieving, in the most difficult of circumstances, rather than vilification based on the assumption that we somehow did not care about what was happening to the Tamils stuck on the beaches of north-east Sri Lanka. We cared more than anyone else. We also saved a significant number of lives, relieved a lot of suffering, and gave hundreds of thousands a chance to rebuild their lives. It is not at all clear to me that a different approach, where the UN would have in effect behaved like Human Rights Watch, could have achieved more for the people, even if we might have been applauded more by some for taking a stand.
74In her article ‘Sri Lanka: Unrestricted Warfare and Limited Protective Humanitarian Action’, Norah Niland offers a damning critique of humanitarian action—or, more accurately, inaction—in the Sri Lankan civil war in 2009. Of course, she is not the first to do so. The humanitarian response during both the final stages of the war and the post-war era has been widely criticised (see, for example, Goodhand, 2010; Humanitarian Policy Group, 2010; Internal Review Panel, 2012; Keen, 2009). There is broad agreement that a different course of action, involving greater public advocacy and, in particular, the dissemination of casualty figures, should have been taken.
75On my reading of her piece, the following points represent Niland’s central arguments. In Sri Lanka in 2009, UN and non-governmental humanitarian agencies neglected protection work—and specifically various forms of advocacy—in favour of efforts to secure or maintain the access required to provide material assistance such as food and shelter. According to Niland, this was mistaken for two main reasons. First, she argues that it demonstrates a mis-conceptualisation of the humanitarian imperative and a corresponding mis-prioritisation of relief assistance over protection activities. Second, the logic of the approach is presented as flawed, as abstention from engaging in the more ‘political’ activities associated with protection did not, in fact, facilitate access for material assistance. The government and LTTE’s attempts to manipulate humanitarian action, in combination with the humanitarian community’s susceptibility to such manipulation, exacerbated both dimensions of the problem. A similar scenario played out in response to the post-conflict internment of survivors. Moreover, Niland asserts that there is a great danger these mistakes will be repeated elsewhere.
76With respect to policy, this last point is key, and I fear Niland is absolutely right. Despite a raft of criticism, it is not at all clear that the necessary changes have been made to prevent a repeat of the same type of mistakes made in Sri Lanka. Perhaps part of the reason is that we do not know what such changes would look like, because most critiques of the international humanitarian community’s actions in Sri Lanka—including Niland’s—do not take us far enough. Simply critiquing the response in Sri Lanka, and identifying what should have been done differently in that particular context, is not enough to tell us what should be done in the future, in contexts which will undoubtedly be quite different. From my own recent interviews with staff from humanitarian agencies in Myanmar, for example, it is clear that the Sri Lankan experience is being borne in mind in Rakhine, but that it does not provide many answers. On the one hand, the protection working group produces advocacy notes regarding protection concerns, which are then passed on to the diplomatic community in Yangon, to inform high-level dialogue with the Myanmar government. On the other hand, ‘red lines’ are discussed but not been agreed upon, because it is unclear where the lines should be drawn, and what course of action should be pursued if the lines are crossed.
77Keeping in mind the importance of drawing lessons from the Sri Lankan experience that might be applied to other contexts, I wish to highlight three issues to which I believe Niland pays insufficient attention. I hope that by further examining her arguments, and situating them within broader debates regarding the role of protection in humanitarian action, this comment will thus help to move the discussion forward.
78The first point I want to emphasise is that we do not have a definitive understanding of cause and effect regarding what happened in Sri Lanka, and must thus exercise caution in drawing conclusions. Implicit to Niland’s contribution is the assumption that more public advocacy would have increased protection, but framing this as a straightforward, linear relationship is problematic. There were almost certainly several interacting factors at play here, including a two-directional relationship between every possible dyad of the following components: humanitarian advocacy, humanitarian access, government behaviour, LTTE behaviour and lifesaving interventions. Advocacy, moreover, is not simply a quantitative variable, but also a qualitative variable, and lifesaving interventions may be both material (e.g. food and shelter) and non-material (e.g. dialogue and denunciation). Perhaps the question should not be how much advocacy is required, but what kind (DuBois, 2007). What kind of advocacy would have been most effective in changing government and LTTE behaviour for the better? Would public naming and shaming by the UN and other humanitarian agencies have been most effective? Would it have been more effective to pass information to the diplomatic community in Colombo, leaving it to them to put pressure on the Sri Lankan government? Would simply discussing the casualty figures have been enough? Or would humanitarian actors also have had to advocate for specific responses from the government, the LTTE and international political actors? What kind of advocacy would have maximised the trade off between advocacy and access?
79Assessing and accurately attributing causality is immensely complicated in the complex and dynamic contexts of armed conflict and humanitarian action. Although we know what happened in the face of inaction, we do not know that this course of events could have been avoided or mitigated had UN agencies and NGOs been far more vocal (Humanitarian Policy Group, 2010, 4). We cannot know the counterfactual outcome with any certainty, and this limits the lessons we can derive from the Sri Lankan experience. It may be that whatever course of action taken by the international humanitarian community, no more lives could have been saved. Strong public advocacy would almost certainly have eased the conscience of the international humanitarian community, but it would not necessarily have saved the lives of Sri Lankan civilians. There is a lack of consensus among International Relations scholars regarding the extent to which ‘naming and shaming’ improves governments’ respect for human rights (Franklin, 2008; Hafner-Burton and Ron, 2009). What is undisputed is that public criticism does not always lead to the desired behaviour change. There is a danger that humanitarians delude themselves, and others, with promises of protection alongside relief assistance, when they are unable to deliver meaningful protection (DuBois, 2009). Greater acknowledgement of the limits of humanitarian action is essential both to ensure that the humanitarian community does not make (potentially life-threatening) false promises to civilians about the kind of protection it can offer, and to minimise the potential for the presence and work of humanitarian agencies to be used by others (particularly the UN Security Council and powerful donor states) to conceal political inaction.
80If attributing causality is difficult after the event, gaps in information relating to cause and effect are even more acute during an emergency. Niland rightly identifies poor decision-making as a key part of the problem in Sri Lanka, but I think she underestimates the difficulty of making good decisions in emergency contexts. For example, she highlights the failure of the international humanitarian community both to engage effectively with the Sri Lankan authorities, and to speak out publicly against the abuses they perpetrated. This was at least in part due to the mistaken expectation that abstaining from public advocacy would enable their continued access to the civilian population. We now know that access was ultimately denied regardless, but in 2008 and early 2009, humanitarian agencies could not have known with certainty what the result of their near-silence would be. Some may have suspected, and even predicted, the subsequent outcome, while others—as Niland’s interview data indicates—believed they would save more lives by keeping quiet (Niland, 2014, 7). This is not unique to the Sri Lankan context. Humanitarian actors will always be making decisions under conditions of uncertainty. Sophisticated political analysis is required in order to make the best possible decisions—and international humanitarian agencies do not always possess the expertise necessary for such analysis (Darcy, O'Callaghan and Bonwick, 2007, para. 17). It may be that better decisions—and thus more effective protection—are beyond the capacity of the international humanitarian community as it stands.
81Single case research allows for the kind of in-depth analysis not usually possible in comparative approaches, but this comes at the price of generalizability. If an aim of evaluating and critiquing humanitarian action is to learn lessons for future events and crises of an uncertain nature in vastly differing contexts, this poses a particular problem. There is a related tension, evident in Niland’s piece, between treating the Sri Lankan experience as an exceptional situation, and viewing it as reflecting humanitarian crises more generally. She claims, for instance, that the mistakes and problems highlighted by the Sri Lanka case are the norm rather than the exception, and are thus likely to be repeated elsewhere (Niland, 2014, 6 and 11). At the same time, however, she suggests that ‘the relief community’s failure to leverage its knowledge of the atrocities to elicit meaningful action’—one of her central criticisms—was unusual (Niland, 2014, 4). The fact is that in other contexts, events will not necessarily play out as they did in Sri Lanka. How, then, can we draw more general conclusions and lessons from this case?
82In a context like Sri Lanka, assuming that access would be maintained or granted if controversial advocacy was avoided may have been naive. The international community’s failure to engage in advocacy once the hoped-for access was denied, leaving little or nothing to lose, may be impossible to justify. Such a situation, however, is one end of a spectrum, and in many emergencies the situation will be much less clear-cut. This means the Sri Lankan experience alone will offer insufficient guidance, and we need to think more generally about the relationships between protection, advocacy and access. Similarly, with respect to the kind of advocacy that should be undertaken, and the issues such advocacy should deal with, the picture in Sri Lanka may have been more straightforward than elsewhere. Efforts at private advocacy with the government had got nowhere, and credible casualty figures existed. Moreover, UN officials directly experienced artillery fire, and thus had first-hand knowledge of civilian targeting, including—indeed primarily—by the government (Internal Review Panel, 2012, para. 17). Instead of being publicly disseminated, reliable data was deliberately concealed and misrepresented. However, acknowledging that this was the wrong decision does not offer us much guidance for future contexts, where information may be less credible, numbers less certain, and private dialogue with the relevant authorities yield some limited success. In such situations, should data be shared nonetheless? How reliable must the data be? And with whom should it be shared? In order to answer these questions, we need to move beyond the single case of Sri Lanka to try and identify the conditions and situations in which particular advocacy strategies are appropriate.
83Despite the difficulties in attributing causality in any given case, and the problems of generalising from individual cases, critiques of humanitarian action tend to focus on individual cases. It is thus perhaps unsurprising that knowledge of what strategies are effective in protecting civilians, under what conditions, remains limited (Reichhold and Binder, 2013). At this stage then, it may be more productive to move beyond the many critiques of humanitarian action in Sri Lanka, and to undertake a systematic, comparative analysis of existing evaluations of a range of humanitarian responses found to have differing levels of success with respect to civilian protection. When seeking to improve future policy, identifying what has worked in the past is as important as identifying what has failed. In the meantime, as long as we are examining a single case, the best we can do is to identify the conceptual, institutional, economic and political factors that contributed to poor decision-making, and focus on these factors, rather than the decisions themselves. In other words, we need to understand why poor decisions were made.
84Niland does address this issue, albeit often only implicitly. For example, she identifies confusion over the meaning of ‘protection’, agencies such as UNHCR being ‘in bed with the government’, the political manipulation of humanitarian actors, a lack of independence, and insufficient focus on impartiality as all contributing to the mistakes of the humanitarian community in Sri Lanka. A more systematic analysis of the factors behind poor decision-making, however, would be more useful than this somewhat anecdotal approach. In addition, an important factor, omitted by Niland, is the lack of sufficient political expertise among humanitarian actors. The ability to analyse and assess the likely effects of different strategies in a given context is essential to making good choices, and sophisticated political analysis is all the more important in the face of manipulative efforts by armed parties to the conflict and other actors. The reality is that those who seek to manipulate humanitarian action for their own benefit or advantage are often more skilful political operators than humanitarians themselves.
85Moreover, with respect to some of the issues highlighted by Niland, I think we need to peel back another layer and ask why these issues arise in the first place. Research from other contexts may help shed some light on these questions. For example, with regard to the apparent confusion as to the meaning of protection—and the prioritisation of relatively easy and non-confrontational activities— part of the problem may be conceptual (Bradley, 2012, 43-54). It may also be the result of political pressure and institutional incentives (Hart and Lo Forte, 2013). Identifying the causes of confusion and mis-prioritisation is the first step to addressing them. Similarly, we need to ask why the UNHCR was ‘in bed with the government’, why humanitarian actors were so easily manipulated and why independence and impartiality were lacking. It is only through answering these questions that we can identify—and subsequently address—the roots of poor decision-making. A comparison of the strategies, action and inaction of the different humanitarian agencies operational in Sri Lanka might help to clarify some of these issues, but for the most part Niland discusses them collectively rather than individually. Further examination of Niland’s own interview data with these questions in mind thus has the potential to help us better understand more about why poor decisions were made in Sri Lanka.
86Undoubtedly humanitarian action was limited in Sri Lanka, both in terms of ambition, focusing on material assistance rather than protection, and in terms of the range of activities pursued, largely avoiding ‘political’ and confrontational activities. Yet this observation prompts as many questions as it answers. First, as I have argued above, understanding why the humanitarian response was so limited in Sri Lanka—and how better decisions may be made in the future—requires deeper analysis of the Sri Lankan case and a systematic comparison with other experiences. There exists a wide array of scholarly knowledge, tools and methods which could be utilised in efforts to understand why decision making was so poor in Sri Lanka, and to identify the conditions under which different kinds of advocacy—or other humanitarian activities—will be most effective in saving lives.
87Second, a more fundamental question exists regarding the extent to which humanitarian action should, in fact, be limited. An undercurrent to both Niland’s piece and my response is the question of how far humanitarian action—particularly in pursuit of protection objectives—should engage with politics. I have suggested that greater political expertise is necessary to make better decisions about how to protect civilians. Niland argues for greater advocacy, and implicitly suggests that as ‘humanitarian actions invariably have political implications’, humanitarians should not shy away from engaging in politics (Niland, 2014, 6). In going down this road, however, there is a danger that humanitarians will be ‘drawn into discussions in which they have limited competence or expertise’, and thus do more harm than good (Humanitarian Policy Group, 2007, 1). Should humanitarians aim to develop greater political expertise, or should they draw a line on advocacy and political engagement at the limits of their existing expertise? What, in other words, are the parameters of the humanitarian project?
88Finally, the Sri Lanka case study raises questions as to the broader limits of humanitarian action. Humanitarians cannot physically stop the bullets and the bombs, and in a context of unrestricted warfare, this imposes significant limits on what they can reasonably achieve, however wide those parameters are set. Honesty and humility about the limits of humanitarian action may be as important as efforts to expand the scope of humanitarian ambition and activities, and to enhance capacity for good decision-making.