ABOUT THE AUTHOR
Marcos Ferreiro is an aerospace engineer and humanitarian practitioner, with field experience in Somalia, DR Congo, Sudan, Nigeria, Kenya, Mauritania, Haiti, Ecuador, Brazil and India. He is a Barrié de la Maza fellow and Harvard graduate in Public Administration in International Development.
In May 2012, a Pakistani doctor was sentenced by his government for carrying on a fake vaccination campaign that helped the US government identify and kill Osama bin Laden.1 A few weeks later, gunmen opened fire on a UN vehicle in Karachi (Pakistan), wounding a Ghanaian doctor who was part of a polio campaign. Previously, some Taliban leaders had announced that they would not allow vaccinators to treat children there “unless the US halted its campaign of drone strikes against militants in the area,” and claimed that the vaccination campaign was “a cover for American espionage.” Meanwhile, in Somalia – another battlefield of the Global War on Terror – the militant group Al-Shabaab is insistently banning vaccination campaigns. Indeed, in the midst of a famine declaration, in 2011 this Islamist militia kicked out several humanitarian agencies from the areas under its control, accusing them, amongst other charges, of “collecting data under the guise of demographic surveys, vaccinations reports, demining surveys, nutrition analyses and population censuses, using field data and analyses for dishonest policies and programmes.”
Militant movements have good reasons to be suspicious of Western humanitarian agencies, even if they may be wrong on their allegations. USAID administrator Rajiv Shah has recently reaffirmed the strong links between aid and national security: “[Former Defense] Secretary Gates was right when he said development is a lot cheaper than sending soldiers,” and official reports to the US Congress bluntly include USAID expenses in Iraq and Afghanistan on the costs of the Global War on Terror. The question is, what is the price of linking aid efforts with military goals?
This is not a new concern, nor are Shah’s declarations especially striking. Since the wave of ‘humanitarian interventions’ in the 90s and ever more so since 9/11, roles between humanitarian workers and other actors have become increasingly blurred. In the last decade, we have heard NATO Secretary General Anders Fogh Rasmussen suggesting that NGOs should be the “soft power” component to military strategy in Afghanistan, and describing this country as “the prototype” for engagement between NATO and NGOs; we have also heard then-U.S. Special Representative for Afghanistan, Richard Holbrooke, remarking the importance of aid organizations on intelligence gathering; and, at the beginning of the Afghan campaign, then-US Secretary of State Colin Powell referring to INGOs as “force multipliers” and “an important part of our combat team.” In sum, the military seems to be increasingly aiming to co-opt the NGO community under a discourse based on the need to break cultural barriers and misunderstandings, and the assumption of sharing common goals. Military relief operations within counter-insurgency (COIN) strategies and the mushrooming of militarized contractors have contributed to greater confusion.
Yet, some warn that the humanitarians should not only be concerned by the blurring with the military but also and foremost, with the trends in political engagement by donor states. Indeed, many point out that the blurring of humanitarian and political-military agendas is provoking an increase of violence against aid workers and the loss of access in some complex emergencies. However, despite all the criticisms to the blurring of lines, its consequences are hard to measure. How much of the current violence trends against aid workers in Somalia, Sudan or Afghanistan is due to the deployment of foreign troops, the implementation of stabilization approaches or UN integration policies? With so many confounding factors in play, any claim of causality can be easily challenged. This haziness hinders the ability of decision-makers to evaluate the severity of the problem, and provides arguments to champions of comprehensive approaches, who often put the burden of proof on the shoulders of their critics.
A Divided Humanitarian Community
Can the military and the humanitarians pursue shared objectives? On the one hand, many humanitarian agencies and scholars strongly reject that there are such common goals between the military and the humanitarians, and maintain that what separates military from humanitarian actors is not just cultural barriers and misunderstandings; rather, there are deeper, intrinsic differences between core aims and principles, especially with regards to the impartial delivery of aid by the latter vs. the tactical and conditional aid by the former. Some say that the line separating humanitarian and military action is one that by definition of the essence of humanitarian action “cannot be bridged”, and maintain that the aim of humanitarians is “not to end wars, bring peace, build states, or promote democracy, but to limit the devastation of war by providing relief aid to people in plight” at both sides of the frontline.
On the other hand, a large portion of the aid community believes that assistance should transcend the basic humanitarian imperative of saving lives and be directed towards the root problems. That is, that it should go beyond alleviating the symptoms and seek a transformation of the society. Thus, some multi-mandate organizations find it appropriate aligning their goals with the donor states’ stabilization agendas, and are willing to sacrifice neutrality to include in their portfolios nation-building, liberal peace, human rights and economic development. The issue becomes even more divisive when confronting operational independence vs. financial reliance on donors involved in the conflict. Here, some humanitarian agencies are taking very controversial decisions:
“Yes, we do accept funds from the US Government for our operations in Afghanistan, but not from the Department of Defense, so as to preserve our independence and non-partisanship.” (Former president of a top American humanitarian agency)
Besides, some relief agencies are progressively seeking more armed protection from a large array of international forces and private security firms, creating a growing dilemma between the humanitarian agencies’ values and their actual security needs:
“Aid agencies cannot have it both ways: asking for armed escorts to reach populations in need one day and criticizing those same military forces for blurring the lines the next cannot be a solution.” (P. Krahenbuhl, ICRC Director of Operations)
Eventually, humanitarians know that it is not just external factors but also the choices they all make what affects their security, and this realization is creating tensions within the humanitarian community.
Drivers of the Blurring of Lines: A Snapshot
UN integration; military involvement in relief aid; whole-of-government stabilization efforts; counter-terrorism legislation; and ‘humanitarian’ interventions invoking the Responsibility to Protect norm. Though this list does not intend to be comprehensive, these are some of the main exogenous variables that feed the blurring of international military, political and humanitarian agendas. All these issues are strongly interrelated, and share a willingness to provide coherent solutions to complex emergencies while addressing security threats. Although sound on paper, their implementation has created significant tensions yet to be addressed. Hereafter is a brief review of the challenges that each of them is facing.
Military Aid Embedded in Counter-Insurgency
Hearts and minds COIN strategies are not a new phenomenon. However, since the end of the Cold War there has been an increasing engagement of the military in relief operations, carried on as a component of their combat tactics but presented under the label of humanitarian action. In Afghanistan and Iraq, many voices have denounced the instrumentalization of aid, pointing out that the military pushes the population to take sides, putting the beneficiaries at risk and blurring the lines with the genuine aid workers. The Provincial Reconstruction Teams (PRTs), conformed by joint civil-military teams, are perhaps a paradigm of this blur. While they implement ‘quick impact projects’ on relief, rehabilitation and stabilization, their final goal is to attract the population esteem, gather intelligence, undermine local support to the insurgents and gain force protection.
From a humanitarian and development perspective, aid projects carried on by the military in conflict zones are hardly ever effective because they lack the expertise, the required human resources, the necessary time span and the proper needs-based approach. Besides, some analysts warn that COIN strategies endanger civilians, as they are forced to take sides and risk retaliation from the opposition. Their failure to have a transformative effect on the communities concerned is ultimately their failure to win hearts and minds and therefore a failure to stabilize the area and protect their own troops. From the military point of view, COIN operations such as the ones carried on by PRTs play an important role in getting international publicity, maintaining support from home constituencies and improving the troops’ morale. Nevertheless, some highlight that a fatal flaw of the COIN in scenarios such as Afghanistan is that local leaders “never seemed to actually share or buy into this legitimization strategy of winning the population away from the insurgents through services,” which drastically affects the outcomes. Similar complaints can be heard from other contexts:
“We were highly inefficient with our services. I spent 10 million dollars on my own in cash, on projects that I am sure that had a very low return on security and overall effectiveness.” (US-Force Tank Commander & Infantry Platoon Leader deployed in Iraq on 2004-2005) 
Western governments have lately called for ‘coherent,’ ‘comprehensive,’ ‘3D,’ or ‘whole of the government’ approaches to stabilize fragile states, in an effort driven mainly by geostrategic interests, national security agendas and the War on Terror. State failure is perceived as a consequence of the breakdown of public service, so stabilization efforts are increasingly oriented towards the delivery of basic services (through massive investment in aid) and the promotion of the rule of law, liberal democracy and a western-style economic model based on free markets.
Despite the efforts, stabilization is failing to deliver what it promised in contexts such as Somalia, Iraq and Afghanistan. In these countries, the entire aid architecture is apparently doing more to destabilize rather than stabilize the country. Some analysts suggest two key reasons to explain this failure: (1) addressing the wrong drivers of insecurity, and (2) spending too much, too quickly. The fact is that stabilization theory focuses on socio-economic problems such as poverty, illiteracy, and lack of social services. However, if the drivers of conflict are mainly political (such as grievances due to unfair power sharing and wealth sharing), they cannot be effectively addressed through primarily socio-economic activities. Indeed, aid projects may just feed conflict by distributing resources in a way that exacerbates these grievances. Besides, donors’ institutional priorities and incentive structures result in attempts to spend hastily, including donors often putting pressure on their field representatives to allocate all the budget within brief and rigid timeframes, with scarce human resources and limited to the zones where their troops are deployed:
“Some donor agencies are spending hundreds of millions of dollars in Afghanistan relying on hasty analyses made by students who come for short internships. Due to the pressure they have to spend the entire budget, whoever comes with the most stupid idea will get funds for it. The Afghans know it and take advantage.” (Former delegate of a Western donor agency in Afghanistan)
Unsurprisingly, some observers warn that stabilization as implemented right now fuels corruption and undermines both security, relief and development objectives, becoming a “self-defeating” strategy: a perverse system that will only perpetuate the conflict. In particular, in environments where there is very little implementation capacity, and even less capacity to provide oversight over the implementers, some analysts suggest that spending amounts of aid that exceed both the implementation and absorption capacities of aid agencies, the local government, and local communities, virtually guarantees failure, and conclude that “less can be more.” The bottom line is that the relationship between relief, development and security is more complex than anticipated by supporters of hearts & minds strategies: it is not aid but security and governance what local communities mostly demand in conflict contexts.
Western donor states and local governments are increasingly promoting counter-terrorism legislation that criminalizes the transfer of resources to terrorist groups or individuals. Applied literally, this amounts to the criminalization of any relief provided to populations located in areas controlled by armed groups identified as ‘terrorists’ because setting up aid operations “necessarily involves negotiating directly with the de facto authorities and providing indirect material support to their political economy.”
Thus, counter-terrorism laws are progressively used to legitimize the reluctance of governments to abide by international humanitarian law (IHL). Tensions between the two sets of laws have become more apparent in the cases of Somalia, Afghanistan, Pakistan and Palestine, with humanitarian organizations facing mounting restrictions and donors conditioning funding to growing security and political requirements, such as no-contact policies, which hinder their efforts to be impartial actors. Remarkably, these new regulations go against the humanitarian principles that the same donor governments have subscribed through intergovernmental fora such as the Good Humanitarian Donorship initiative.
On the other hand, even though humanitarian action may not aim to support any political or military organization, it does to some extent help to strengthen them – only marginally in most cases, but sufficiently to concern other stakeholders. The future of humanitarian assistance in regions controlled by armed groups designated as terrorist organizations will therefore require common efforts “to delineate a new balance between humanitarian imperative and national security priorities.” 
The Responsibility to Protect (R2P)
The protection of civilians is another important factor in the blurring of humanitarian, political and military spheres, especially when R2P is invoked. Certainly, the non-coercive actions that humanitarians can take may have a limited impact when civilians are targeted. On the other hand, some observers alert that protecting populations through military interventions “does not mean restoring law and order in the same way that police forces do in a peaceful country, but creating a new political order through violence.” In this line, others have recently warned that stopping genocide has become “a political project” that operates with wrong tools, under wrong assumptions, aiming at the wrong goals:
“[R2P architects] have composed a story, based on ethics rather than evidence, that incorrectly assumes all perpetrators of mass political violence are insatiable killers and that dictates who should respond (Western nations), how (with military intervention) and why (for justice and democracy). It is a morality tale that undermines the best ways to deal with the worst crimes.” 
The fact is that claims of politicization of the R2P norm seem well founded, as it is called upon when there seems to be an interest by any of the permanent members of the Security Council (e.g. Libya 2011) and conveniently ignored when the affected country is not worthwhile and out of the radars of the Western media (e.g. Sri Lanka, 2009; Yemen, 2009). Besides, there is a legitimate concern on the impact that coercive force may have on the ability of humanitarian agencies to carry on their operations, as they may be perceived as party to the conflict.
More importantly, some observers warn that military interventions may do more harm than good to the population they reportedly intend to protect. A recent study carried on in Burma, Sudan, South Sudan and Zimbabwe found that local communities considered their own actions to protect themselves as more important than anything done by peacekeepers, which was perceived as “rather modest, non-existent, or even counter-productive.”
“For the time being, much more needs to be done in building a consensus on the means and methods of implementing R2P. In particular, much work needs to be invested in researching the military objectives and tactics of such operations, the role and responsibilities of humanitarian organizations, in establishing clear protocols to rebuild broken societies in line with human rights values and norms, and in prompting processes of reconciliation as a premise for sustainable peace.” 
Integration intends to confront old problems, such as poor inter-agency coordination and funding shortages. Moreover, in line with the R2P norm, it aims to revert the apparent passivity of the international community under past massacres of the 90s where humanitarian action was perceived as a meager substitute of political action. But it also creates new challenges, especially in situations of ongoing conflict where the UN pursues a policy of structural integration. This is currently the case in 11 of the 18 integrated presences where there is a peacekeeping or political mission. In such scenarios, UN humanitarian agencies have been dragged into a schizophrenic situation, in the obligation to deliver aid in a neutral, impartial and independent manner (and more importantly, with the need to be perceived as such) while being embedded in a broader political agenda.
The UN Secretary-General’s Decision 2008/24 recognized the internal contradictions that integration generates for aid agencies, and tried to address the humanitarians’ concerns by remarking on the need to “take full account of recognized humanitarian principles, allow for the protection of humanitarian space, and facilitate effective humanitarian coordination with all humanitarian actors.” In order to do so, it provided flexibility in the country level arrangements, which can take “different structural forms,” and highlighted the importance of “strategic partnership” upon structural integration.
Despite these arrangements, integration remains controversial. Many humanitarian actors accept the need for greater coherence within the UN system, while simultaneously fear the “subjugation of humanitarian priorities to the UN’s political objectives.” Some affirm that, in practice, with integration UN humanitarian agencies have sacrificed operational freedom and independence. Specifically, some observers highlight that OCHA has become “cornered almost to irrelevance,” “abolished altogether, left under-funded or relegated to a diminished role of general coordination and interfacing with NGOs, without any significant influence to further humanitarian action within the UN system.” The concern on the increased politicization and partisanship of humanitarianism has pushed InterAction to demand a ‘humanitarian exception:’
“Humanitarian actors can no longer afford to ignore the contradictions inherent in the UN’s policy on integration. A humanitarian exception to the integration rule is required.”
Meanwhile, UN political and peacekeeping actors argue that the concept of integration has been misunderstood. Some proponents of UN integration even insinuate that hostility towards these arrangements among some humanitarian actors “stems from resistance to greater scrutiny over aid diversion and other sensitive issues.” A study recently commissioned by the UN Integration Steering Group suggests that the frictions come mainly from a general lack of understanding and awareness, as well as lack of ownership, transparency and accountability by the parties involved. This study does not question nor challenge the concept of integration, as this point was out of its terms of reference. It just identifies good and bad practices, and recommends a number of technical arrangements to fine-tune its completion. However, what if tensions do not arise just from details on the dissemination and implementation of the policy, but from deeper conceptual contradictions due to incompatible mandates? Decision 2008/24 underscores that the main purpose of integration is “to maximize the individual and collective impact of the UN’s response, concentrating on those activities required to consolidate peace.” The humanitarian agencies’ overarching goal is to save lives, which is often at odds with the pursuit of peace in an armed conflict. Whether there is structural integration or not, UN humanitarian agencies are confronting a collective adaptive challenge: a conflict between the values they stand for and their practices, with multiple goals that often become antagonist. They need to bridge this gap, and there is no simple fix.
Ultimately, integration is here to stay. At this point, the debate whether it is good or bad for humanitarian action seems sterile. Instead, UN humanitarian agencies should focus on how to adapt their values and actions to this new reality.
“Integration is a UN-mandated policy. Withdrawing from (it) is not an option. We believe in its intended aim of greater strategic coherence […] At the same time, we cannot allow integration to impede the effective provision of humanitarian assistance to people in need.” (Valerie Amos, UN Under-Secretary-General for Humanitarian Affairs and Emergency Relief Coordinator)
Impact on Access: A Shrinking Humanitarian Space
There is a strong perception within the humanitarian community that, from the perspective of aid worker security, “humanitarian space is indeed diminishing” and that the last decade has seen “an increasing disregard for international humanitarian law.” This view is widely shared by donors and UN agencies:
“For the past few years, civilians have increasingly become victims of conflicts […] International humanitarian law is less and less respected […] There also seems to be a growing tendency to close the door to humanitarians, preventing them from helping the victims.” (ECHO)
“There has been an increasing tendency for International Law, including International Humanitarian Law, Human Rights Law and Refugee Law, to be ignored or blatantly violated.” (EU, European Consensus on Humanitarian Aid)
“Increasingly today acquiescence for humanitarian access is lacking.” (John Holmes, then-UN Under-Secretary General for Humanitarian Affairs and Emergency Relief coordinator)
“In recent years, bureaucratic constraints, politically and economically motivated attacks on humanitarian personnel and active fighting have increasingly contributed to limited access to conﬂict-affected populations.” (OCHA)
Most claims about the shrinking of humanitarian space point out to the blurring of lines with political and military agendas, which have made access to vulnerable groups more difficult. On the other hand, there are growing voices that challenge the dominant narrative. This way, some analysts and agencies consider it a “myth” and claim that there was “no past golden age for humanitarian action:”
“Evoking a ‘golden age’ in which aid actors were able to realize their ambitions unfettered is to underplay the very real difficulties encountered during the forced displacements in Ethiopia in the 80s, for example, or, in the 90s, the massacres in ex-Yugoslavia and the genocide in Rwanda.” (Marie-Pierre Allié, President of MSF-France)
Opponents of the shrinking narrative point out that many of today’s challenges are indeed the result of “an expanding humanitarian system that has extended its reach and ambitions into types of conflict and crisis previously off-limits.” The fact is, in global terms, there continues to be growth in humanitarian operations, and some observers affirm that, firstly, in most cases “access to vulnerable populations is better now than in previous periods,” and secondly, that contrary to widespread opinion, there are “probably fewer violations [of IHL] today than in the past.” These scholars underscore that the recent increase in attacks on humanitarian workers is “confined to a few high-risk conflicts,” concluding that there is “no conclusive evidence that humanitarian space is declining over time.”
Reframing the Debate
Many are questioning the appropriateness of debating the ‘humanitarian space.’ The concept seems to have different meanings to different organizations, which drives to apparently sterile discussions. Indeed, the ICRC has stopped using the term because “it would imply the acceptance of a non-humanitarian space, i.e., a space where IHL does not need to be respected,” and also, in practical terms, because “there is simply no such thing as a pre-established, protected humanitarian space.” MSF, who coined the term in the 90s, offers a similar point of view, highlighting the agencies’ responsibility:
“Contrary to the “shrinking space” theory—which frees aid actors from any responsibility for conquering and defending their own sphere of activity—there are no legitimate perimeters to humanitarian action, valid at all times and in all situations, which become clearly visible once the mists of “military-humanitarian confusion” have lifted and humanitarians are protected from any political fallout. There is, however, a space for negotiation, power games and interest-seeking between aid actors and authorities. MSF’s freedom of action is not rooted in a legal and moral “space of sovereignty” that simply needs to be proclaimed in order to be automatically acknowledged and respected. It is the product of repeated transactions with local and international political and military forces. Its scope depends largely on the organisation’s ambitions, the diplomatic and political support it can rely on and the interest taken in its action by those in power.”
In any case, regardless of the label, the real concern is whether the access of agencies to populations in need and/or the access of the communities to relief aid are being increasingly reduced. If so, the subsequent questions are whether this is just a localized and/or temporary phenomenon circumscribed to certain contexts, what the root causes are, to what extent the blurring of lines may explain this trend, and what can be done to revert it. This is apparently the case in places where donor states’ counter-terrorism legislations are currently applied (See Box 1).
Context and Agency Choices
Some observers point out that, contrary to global trends, the humanitarian footprint has indeed shrunk in contexts such as Somalia, Afghanistan, Pakistan, Chad, Darfur and Iraq. Some suggest that the contraction of aid activity/access – measured by the surge in suspensions, withdrawals and relocations following violent incidents – is a consequence of “the obstacles and conditions created by militaries, governments and non state actors.”  Somalia may be one of the most clear-cut examples (see Box 2).
In Afghanistan, some observers claim that UN integration and coherence agendas hinder access to vulnerable groups  and force agencies to conduct much of their work through third party organizations. For most humanitarian organizations, “access and operational space are almost nonexistent” in many parts of the country, reportedly because aid agencies “are seen as adjuncts to the US-led coalition intervention.” Similar scenarios are described in Iraq, Pakistan, Somalia and Darfur, where the UN (and the international community as a whole) is perceived as aligned with one of the sides, alienating the other side.
On the other hand, in countries such as DRC, CAR and Liberia, some scholars affirm that UN integration arrangements have supported increased access for humanitarian actors “by facilitating the use of mission logistical assets, the provision of area security by UN peacekeeping forces and the use of UN military escorts,” and describe the integrated mission in DRC as a “success story.” Finally, in Chad, Sudan and Afghanistan, many observers complain about a ‘bunkerization’ of the UN agencies that, allegedly, limits the movement of staff as well their capabilities to negotiate with local stakeholders. Apparently, UN integration favors a more risk-averse approach in operational security management policies and practices, which in turn impacts upon access.
In sum, there does not seem to be a general trend in access that fits all contexts. The impact of the blurring of lines on the ability of humanitarian agencies to access their beneficiaries is very much context-specific and, perhaps most importantly, agency-specific. That is, it is not just driven by external factors beyond the agencies’ control, but also by the steps and choices taken by the humanitarian organizations in a given scenario, as well as by their own mandate. A good example is the ICRC in Afghanistan: in the period 2005-2011, while most humanitarian organizations were forced to reduce their activities and the presence of expatriates, ICRC managed to go against the grain, sustaining and later expanding its operations in the country without a significant increase in security incidents. (see Table 1)
Table 1 – The ICRC in Afghanistan (2005-2011) 
Impact on Staff Security
Violence against humanitarian workers has skyrocketed post-9/11, from an average of 32 severe incidents and 38 killed per year in the period 1998-2001 to an average of 148 severe incidents and 97 killed per year just one decade later, in the interval 2008-2011.
The aid industry has experienced impressive growth in the last decade, almost tripling its budget in the interval 2000-2010. Globally, this may be the main reason for the surge in violence. However, not only the absolute figures of violence but also the attack rates have significantly increased in the three most violent settings – Afghanistan, Somalia and Sudan – which together account for half of the incidents and killings accumulated worldwide in the last 15 years. Thus, in these three countries, the growth of aid alone does not justify the rise in incidents. To explain this surge, many point at a growing politicization of attacks due to the ‘blurring of agendas,’ with aid workers “increasingly being targeted by militants for their association with Western military and political campaigns.”
On the other hand, some analysts suggest that it is not that conflict situations have become more violent in recent years, but, rather, that “international agencies have become more exposed to risks in conflict zones.” In fact, many agencies have moved from the periphery to the center of conflicts, operating in more dangerous places and, subsequently, affecting the interests of belligerents who, in the past, used to see humanitarian action as non-threatening. This increased exposure to risks has much to do with the ‘blurring of mandates’ – with most development agencies now routinely claiming to act on humanitarian grounds – and the “competitive pressure for operational outreach in areas of high insecurity.”
Some aid agencies have been very vocal on the new risks created by UN integration, “including that of associating humanitarian workers with one side of a conflict, and the consequent risks of attacks on humanitarian workers and the people they are assisting.” (See Box 3). The fact is that integrated missions help local governments to “reshape the social, political, and economic structures of countries in conflict,” often carrying on their transformative agendas against the interests of previously prevailing armed groups that may remain militarily active. Insurgents are subsequently tempted to undermine their political foes by targeting aid workers, typically the weaker links of these integrated missions.
ICRC recently warned that when humanitarian action becomes part of military strategies, “the risks for aid agencies in the field grow exponentially.” The co-option of NGOs for the implementation of joint operations, the abuse of humanitarian rhetoric by the armed forces, the misuse of humanitarian identification material (e.g. the use of white cars and civilian attire by US Special Forces) and, in sum, the “instrumentalization of aid”, augments the current confusion of roles and agendas, and humanitarian workers end up becoming “guilty by association.” In these environments, the universality of the values promoted by the U.N. “no longer guarantees the security.”
“When donors are occupiers and combatants, security and access will be compromised. Counter-insurgency operations do militarize and politicize humanitarian operations, and security and access will be reduced. Whether this is part of a longer-term trend however will depend on the likelihood of future US/NATO combat interventions”. 
Even donor agencies such as ECHO admit that being engaged in nation-building “inhibit NGOs’ ability to adhere to humanitarian principles,” with repercussions on aid workers’ security, and the European Union Consensus on Humanitarian Aid recognize the dangers of mixing up the military and the humanitarians: “In complex emergencies recourse to civil protection assets […] can result in exposing relief workers as well as the affected population to attacks from warring parties […] In order to avoid a blurring of lines between military operations and humanitarian aid, it is essential that military assets and capabilities are used only in very limited circumstances in support of humanitarian relief operations as a last resort.” But, so far, incumbent states have failed to translate these concerns into effective policies. Existing civil-military guidelines have also failed to be effectively translated into military doctrine.
Measuring the Impact
Criminal/rent-seeking motivations; political violence linked to local dynamics; anti-westernization backlash; the agency’s mandate, scope and behavior; increased exposure to riskier environments; and spillover effects of a heavier humanitarian footprint, are some of the drivers of violence against aid workers in any given context. It is difficult to disentangle the impact of the blurring of lines (a blurred concept itself) from these other factors, and even more so to analyze separately the impact of the different intervening variables. A recent study on UN integration concluded that available data “provide no clear evidence of a direct link between UN integration arrangements and attacks on humanitarian workers.” The question is whether any causal effect can ever be estimated with statistical validity, taking into account the many confounding variables, the small sample size, and the absence of a counterfactual.
Interestingly, in Afghanistan, the number of major violent incidents against aid workers over time is poorly correlated with the weight of the humanitarian response, but strongly linked with other variables that can be considered proxies to the international political and military intervention, such as the expenditure of the US State Department, the official development assistance (ODA) and the cost of USA military operations in this country (see figures 1 to 8). 
Barely 12% of the variation of the number of attacks against aid workers in Afghanistan can be explained by the variation of humanitarian aid expenditure. However, roughly 90% can be explained by the variation of the other three variables, which might be used as proxies for overall stabilization, development and military expenditures. This study does not claim to prove causality, as it does not control for omitted variable bias nor for reverse causality. Nevertheless, the remarkable predictive power of these three factors suggests that, beyond mere humanitarian aid, there may be something in the international response in Afghanistan which influences vigorously on the security of humanitarian workers.
In the cases of Somalia and Sudan there is no significant divergence between humanitarian action and ODA’s correlation with the violence against aid workers (see figures 9 to 16). This may be mainly because more than half of the official development assistance is spent in humanitarian action, which implies a strong correlation between both variables.
In Sudan, approximately 90% of the variation of the number of aid workers’ security incidents can be explained by the variation of either ODA or humanitarian expenditure. This strong correlation suggests that, in this country, the size of the humanitarian enterprise may be the main explanatory variable. The predictive power of both variables drops to roughly 60% in Somalia. Here, violence is apparently more complex to disentangle, and other confounding factors with significant impact seem to be in play. In sum, the blurring of lines might be a key driver of violence against aid workers in Afghanistan, but, apparently, it does not play such a relevant role in explaining attack trends in Sudan and Somalia.
Targets by Association vs. Direct Targets
Political violence against aid workers cannot be blamed just on misperceptions due to the blurring of lines. Attacks are not only ‘by association’ but also because of who humanitarians are and what they do. John Holmes, then-UN USG/ERC, alerted in 2009 that aid workers were increasingly directly targeted: “(In Somalia and Afghanistan) we are too often being attacked for who we are […] Too often, UN and NGO flags and emblems are no longer protections but provocations.”
Thus, humanitarians are confronting two distinct threats. On the one hand, ‘associative targeting’ may be due to an apparent breach of the principles of neutrality and independence, which may lead a warring party to perceive that the humanitarian agency is backing its foe. This way, in Afghanistan, some observers warn that the Taliban are attacking aid workers “for their perceived support of alien political agendas,” which reportedly converts them in legitimate targets in the eyes of the insurgents. On the other hand, ‘direct targeting’ is intrinsic to the nature of humanitarian work, and is due to the direct influence of relief agencies in the conflict, or linked to their cultural and ethnic identity. This way, when civilians are deliberately targeted, assisting civilians may also be perceived as taking sides; thus, aid agencies can become targets as well. The provision of aid itself can justify the attack, “in that it represents an obstacle to the objectives of belligerent groups trying to gain or maintain control of the local area or to undermine central authority.” In these cases, humanitarian principles per se cannot protect aid workers, who become more useful dead than alive for those pursuing destabilization strategies.
Afghanistan might be a model case of direct targeting. As shown on figures 17 and 18, attacks against aid workers are significantly associated with civilian casualties caused by anti-government forces. However, there is no correlation at all between the former and civilian casualties caused by pro-government forces (see figures 19 and 20). This might suggest that, as conflict expands, armed opposition groups increasingly target both civilians and the agencies that aid them. However, there is a general view held by NGO staff and supported by the Afghan NGO Safety Office that attacks on humanitarian agencies are “not the official policy of organized armed opposition groups, rather they are as a result of criminality, mistaken identity or as a result of decisions taken locally.” The question is whether this perception is well founded or just wishful thinking from a humanitarian community that takes more comfort in the idea of being targets by association than in becoming direct targets, the former threat being more palatable than the latter.
The last two decades have witnessed the increasing blurring of lines between humanitarian actors, foreign troops engaged in counter-insurgency and counter-terrorism, and donor states pursuing national security and stabilization. In several complex emergencies with foreign troops on the ground there has also been a sharp increase of violence against humanitarian workers, not only in absolute numbers but also in rates. Thus, this escalation cannot be justified solely by the growth of the aid system. Moreover, while in global terms the humanitarian footprint is expanding, there are quite a few contexts (South and Central Somalia, Afghanistan, Pakistan, Chad, Darfur and Iraq) where the trend is reversed, mainly due to withdrawals and suspension of activities following security incidents, threats or bans from the de facto ruling authorities.
How much is to be blamed on the blurring of lines? It depends. Available data suggests that in Afghanistan, the blur of mandates and agendas is a key driver of both attacks against aid workers and loss of access; however, in this country aid workers are not only being targets by association, but also becoming direct targets due to their involvement in local conflict dynamics, rent-seeking behaviors and other causes that have nothing to do with misunderstandings about their identity and mission. In Somalia, the blurring of agendas and widespread suspicion around humanitarian agencies’ missions is hindering access to most of the central and southern regions – areas controlled by rebel movements. However, the current political and military support of the international community to the local government does not seem to have much explanatory power when analyzing causes of violence against aid workers (perhaps because the baseline was already extremely high). The third country analyzed, Sudan, also suffers from access restrictions, in this case from the national government, which perceives humanitarian agencies as a part of a larger international political agenda. However, looking at the available data, it is hard to claim that the blurring of lines is a significant driver of violence in this country. It is rather the size of the humanitarian enterprise in Sudan what mostly determines the variation of violence hitting aid workers, with war-profiteers and rent-seekers apparently behind most of the incidents.
Finally, as the ICRC has shown in Afghanistan, violence is not only context-specific but, more importantly, agency-specific. That is, humanitarian actors cannot only blame external factors, but should also examine themselves, their actions, decisions and behaviors. Indeed, while the humanitarian community has shown its concern with the increasing involvement of the military in relief assistance, it should be consisten with the trends in the international political engagement in these contexts. The fact is that many NGOs are financially dependent on donor states that are one of the warring parties, and end up walking on a razor’s edge, seeking a clear distinction from the military while following belligerent donors’ political agendas. The challenge is even bigger for UN humanitarian agencies embedded in integrated missions and/or working under political mandates from the Security Council. The reality is that humanitarians are doomed to coexist with foreign troops in scenarios where donor governments pursue political agendas. Humanitarian agencies must be aware of the cost of this cohabitation, and learn to deal with it.
“Humanitarian organizations must debate the consequences of their choices in a more self-critical and honest fashion and genuinely decide how they wish to operate.” (Krahenbuhl, ICRC Director of Operations)
 Walsh, Declan, ‘Gunmen Attack U.N. Vehicle in Pakistan, Wounding Polio Doctor’, The New York Times, July 17, 2012.
 Al-Shabaab al-Mujahideen, ‘OSAFA Fact-finding Committee Conducts Organization Performance Appraisal’, Press release, November 28, 2011, available at http://ansar1.info/showthread.php?t=37132 (last visited July 11, 2012).
 Loewenberg, Samuel, ‘Interview: A Business-Like Approach to Foreign Aid’, Foreign Policy, May 3, 2012.
 Belasco, Amy, ‘The Cost of Iraq, Afghanistan, and Other Global War on Terror Operations Since 9/11’, CRS Report for Congress, Congressional Research Service, March 29, 2011.
 NATO, ‘Speech by NATO Secretary General Anders Fogh Rasmussen at the Strategic Concept Seminar in Helsinki’, March 4, 2010, available at http://www.nato.int/cps/en/natolive/opinions_61891.htm (last visited July 11, 2012).
 Burns, Robert, ‘Envoy laments weak US knowledge about Taliban’. Associated Press, April 07, 2009.
 Powell, Colin, ‘Remarks to the National Foreign Policy Conference for Leaders of Nongovernmental Organizations at Yale Law School: September 11, 2001: Attack on America’, October 26, 2001, available at http://www.yale.edu/lawweb/avalon/sept_11/powell_brief31.htm (last visited July 11, 2012).
 Collinson, Sarah, Samir Elhawary and Robert Muggah, ‘States of fragility: stabilization and its implications for humanitarian action’, HPG Working Paper, ODI, May 2010.
 This paper uses ‘aid workers’ and ‘humanitarian workers’ indistinctively, meaning ‘the personnel of UN humanitarian agencies, Red Cross/Red Crescent movement, and NGOs with programmes in humanitarian contexts’. This definition is the same as the one used by Humanitarian Outcomes on the Aid Worker Security database (source of this paper for the data on attacks against humanitarians).
 Barry, Jane and Anna Jeferys, ‘A Bridge too far: Humanitarian Agencies and the Military in Humanitarian Response’, HPN Network Paper, ODI, January 2002.
 Hofman, Michiel and Sophie Delaunay, ‘Afghanistan: A Return to Humanitarian Action’, MSF, March 11, 2010, available at: http://www.doctorswithoutborders.org/publications/article.cfm?id=4311&cat=special-report (last visited July 11, 2012).
 Personal communication, February 21, 2011.
 Hofman and Delaunay (2010).
 Krahenbuhl, Pierre, ‘The militarization of aid is perilous’. Stars and Stripes, January 15, 2011.
 Hofman and Delaunay (2010).
 See e.g. Runge, Peter, ‘The Provincial Reconstruction Teams in Afghanistan: Role model for civil-military relations?’ Association of German Development NGOs (VENRO), October 2009, available at http://www.bicc.de/publications/publicationpage/publication/the-provincial-reconstruction-teams-in-afghanistan-role-model-for-civil-military-relations-268/ (last visited July 11, 2012).
 See e.g. ECHO, ‘Humanitarian Space Under Pressure’, European Community Humanitarian Office, page updated on August 18, 2009, available at http://ec.europa.eu/echo/policies/humanitarian-space_en.htm (last visited July 11, 2012); Steering Committee on Humanitarian Response (SCHR), ‘SCHR Position Paper on Humanitarian-Military Relations’, January 2010, available at http://ochanet.unocha.org/p/Documents/Steering%20Committee%20for%20Humanitarian%20Response-%20SCHR%20position%20paper%20on%20humanitarian-military%20relations%20(2010).pdf (last visited July 11, 2012); and Hofman, Michiel, ‘Dangerous Aid in Afghanistan’, Foreign Policy, January 12, 2011.
 OXFAM International, ‘Iraq: Humanitarian-Military Relations’, Oxfam Briefing Paper, March 2003, available at http://ochanet.unocha.org/p/Documents/OXFAM%20INTERNATIONAL%20Iraq%20Humanitarian-Military%20relations%20(2003).pdf (last visited July 11, 2012).
 Fishstein, Paul and Andrew Wilder, ‘Winning Hearts and Minds? Examining the Relationship between Aid and Security in Afghanistan’, Feinstein International Center, January 2012.
 Personal communication, February 24, 2011.
 See Fishstein and Wilder (2012); and Collinson et al. (2010).
 See Fishstein and Wilder (2012); Aikins, Matthieu, ‘Last Stand in Kandahar. Can the military’s massive counterinsurgency gamble salvage the Afghan war?’, The Walrus, March 8, 2011; and UN Monitoring Group, ‘Report of the Monitoring Group on Somalia pursuant to Security Council resolution 1853 (2008)’, UN Security Council. S/2010/91, March 10, 2010.
 Fishstein and Wilder (2012).
 Personal communication, March 6, 2011.
 Aikins (2011).
 Fishstein and Wilder (2012).
 See Collinson et al. (2010); and SCHR (2010).
 Weissman, Fabrice, ‘Criminalising the enemy and its impact on humanitarian action’, The Journal of Humanitarian Assistance, June 28, 2011.
 Bruderlein, Claude, ‘International Islamic charities facing increased restrictions under counter- terrorism regulations in Israel, UK, and US for their activities in Gaza’, Program on Humanitarian Policy and Conflict Research, Harvard University, Blog posted on January 30, 2012, available at http://hpcrresearch.org/blog/claude-bruderlein/2012-01-30/international-islamic-charities-facing-increased-restrictions-unde (last visited July 11, 2012).
 Egeland, Jan, Adele Harmer and Abby Stoddard, ‘To Stay and Deliver. Good practice for humanitarians in complex security environments’, Office for the Coordination of Humanitarian Affairs (OCHA), Policy Development and Studies Branch, February 2011, available at http://reliefweb.int/report/world/stay-and-deliver-good-practice-humanitarians-complex-security-environments-enar (last visited July 11, 2012).
 Weissman (2011).
 Bruderlein (2012), ‘International Islamic charities…’
 See e.g. Bruderlein, Claude, ‘Syria: Humanitarian corridors will depend on international commitment to protect civilians’, Program on Humanitarian Policy and Conflict Research, Harvard University, Blog posted on February 24, 2012, available at http://hpcrresearch.org/blog/claude-bruderlein/2012-02-24/syria-humanitarian-corridors-will-depend-international-commitment- (last visited July 13, 2012). See also Metcalfe, Victoria, Simone Haysom and Stuart Gordon, ‘Trends and challenges in humanitarian civil-military coordination. A review of the literature’, HPG Working Paper, ODI, May 2012.
 Weissman, Fabrice, ‘’Not in Our Name’: Why Médecins Sans Frontières Does Not Support the ‘Responsibility to Protect’’, Criminal Justice Ethics, Vol. 29, No. 2, August 2010.
 De Waal, Alex, ‘How to End Mass Atrocities’, The New York Times, March 9, 2012.
 Metcalfe et al. (2012).
 See e.g. Weissman (2010); and De Waal (2012).
 IRIN, ‘Security: New report on R2P challenges humanitarians’, IRIN News, February 10, 2012.
 Bruderlein, Claude, ‘Syria: Time to Reflect on Responsibility to Protect’. Program on Humanitarian Policy and Conflict Research, Harvard University, Blog posted on February 13, 2012, available at http://hpcrresearch.org/blog/claude-bruderlein/2012-02-13/syria-time-reflect-responsibility-protect (last visited July 13, 2012).
 See e.g. Metcalfe, Victoria, Alison Griffen and Samir Elhawary (2011), ‘UN Integration and Humanitarian Space’. An Independent Study Commissioned by the UN Integration Steering Group, HPG / Stimson, December 2011. See also Tan, Vivian, ‘Ogata calls for stronger political will to solve refugee crises’, UNHCR, May 27, 2005, available at http://reliefweb.int/report/afghanistan/ogata-calls-stronger-political-will-solve-refugee-crises (last visited July 13, 2012).
 OCHA, ‘Humanitarian Issues: Integrating peacekeeping and humanitarian work – how to make it work?’, January 10, 2012, available at http://www.unocha.org/top-stories/all-stories/humanitarian-issues-integrating-peacekeeping-and-humanitarian-work-–-how-mak (last visited July 13, 2012).
 Elhawary, Samir, ‘The impact of UN integration on aid worker security in Somalia’, Humanitarian Practice Network (HPN), Humanitarian Exchange Magazine, Issue 53, March 2012.
 Derderian, Katharine, Eric Stobbaerts, Iesha Singh, Simone Rocha and David Melody, ‘UN Humanitarian Reforms: A View from the Field’. Humanitarian Practice Network (HPN), Humanitarian Exchange Magazine, Issue 39, June 2007.
 Former NGO Country Director in Afghanistan and Iraq, personal communication, March 6, 2011. Indeed, there is no OCHA office in Iraq, while the Afghanistan OCHA office was dismantled when the integrated mission was launched in 2002, being reestablished as a separate body of the integrated mission seven years later and just after strong pressure from international NGOs.
 Derderian et al. (2007). This report describes experiences in Darfur, South Sudan, the DRC, Haiti, Liberia, Côte d’Ivoire, Somalia and Uganda.
 InterAction, ‘A Humanitarian Exception to the Integration Rule’, InterAction Policy Statement, December 2011, available at http://www.interaction.org/document/interaction-statemetn-un-integration (last visited July 13, 2012)
 Elhawary (2012).
 Metcalfe et al. (2011).
 UN Secretary-General’s Decision N° 2008/24 (2008), ‘Integration’, June 26, 2008.
 OCHA (2012).
 Tennant, Vicky, Bernie Doyle and Raouf Mazou, ‘Safeguarding humanitarian space: a review of key challenges for UNHCR’, Policy Development and Evaluation Service, UNHCR, February 2010, available at http://www.unhcr.org/4b68042d9.html (last visited July 13, 2012).
 ECHO (2009).
 European Union (EU), ‘The European Consensus on Humanitarian Aid’, Official Journal of the European Union, C25/1. January 30, 2008.
 Holmes, John, ‘Attacks on aid workers harm humanitarian efforts’, The Sydney Morning Herald, August 19, 2009.
 OCHA, ‘OCHA on Message: Humanitarian Access’, Version I, April 2010, available at http://ochanet.unocha.org/p/Documents/OOM_HumAccess_English.pdf (last visited July 13, 2012).
 See Hubert, Don and Cynthia Brassard-Boudreau, ‘Shrinking Humanitarian Space? Trends and Prospects on Security and Access’, The Journal of Humanitarian Assistance, November 24, 2010. See also Donini, Antonio, ‘Between a rock and a hard place: integration or independence of humanitarian action?’, International review of the Red Cross, Volume 93, number 881, March 2011.
 Collinson, Sarah and Samir Elhawary, ‘Humanitarian Space: a review of trends and issues’, HPG Report 32, ODI, April 2012.
 IRIN, ‘Aid policy: The myth and mystique of humanitarian space’, IRIN News, May 2, 2012.
 Allié, Marie-Pierre, ‘Humanitarian Negotiations Revealed. The MSF Experience’, Claire Magone, Michael Neuman and Fabrice Weissman (eds), Médecins Sans Frontières, 2011.
 Collinson and Elhawary (2012).
 GHA database, Global Humanitarian Assistance, available at http://www.globalhumanitarianassistance.org/datastore (last visited May 30, 2012).
 Hubert and Brassard-Boudreau (2010).
 See e.g. Collinson et al. (2010); Hubert and Brassard-Boudreau (2010).
 See Collinson and Elhawary (2012).
 Maurizio Geremia, ICRC Head of Unit for Relations with Arms Carriers, personal communication, March 15, 2011.
 Krahenbuhl (2011).
 Allié (2011).
 See e.g. Egeland et al. (2011); and Pantuliano, Sara, Kate Mackintosh and Samir Elhawary with Victoria Metcalfe, ‘Counter-terrorism and humanitarian action. Tensions, impact and ways forward’, HPG Policy Brief 43, ODI, October 2011. See also Featherstone, Andy, ‘Afghanistan: a case study. Strengthening principled humanitarian response capacities’, Humanitarian Policy Group (HPG), January 2012; Bruderlein (2012), ‘International Islamic Charities…’; and Donini, Antonio, ‘Afghanistan: Humanitarianism under Threat’, Feinstein International Center, Briefing Paper, March 2009, available at http://sites.tufts.edu/feinstein/2009/afghanistan-humanitarianism-under-threat (last visited July 13, 2012).
 Pantuliano et al. (2011).
 Pantuliano, Sara and Victoria Metcalfe, ‘Neutrality undermined: the impact of counter-terrorism legislation on humanitarian action in Somalia’, Humanitarian Practice Network (HPN), Humanitarian Exchange Magazine, Issue 53, March 2012.
 See Egeland et al. (2011); and Stoddard, Abby, Adele Harmer and Katherine Haver, ‘Aid Worker Security Report 2011’, Humanitarian Outcomes, August 2011. See also Stoddard, Abby, Adele Harmer and Victoria DiDomenico, ‘Providing aid in insecure environments: 2009 update’, HPG Policy Brief 34, ODI, April 2009.
 Egeland et al. (2011).
 See e.g. Donini (2011); and Hofman (2011).
 Featherstone (2012).
 Donini (2009).
 See e.g. Elhawary (2012); and Donini, Antonio, ‘Humanitarian Agenda 2015: Final Report. The State of the Humanitarian Enterprise’, Feinstein International Center, March 2008, available at http://sites.tufts.edu/feinstein/2008/humanitarian-agenda-2015-the-state-of-the-humanitarian-enterprise (last visited July, 13, 2012). See also Donini (2009); and Stoddard et al. (2011).
 Metcalfe et al. (2012).
 Donini (2008).
 See e.g. Metcalfe et al. (2012).
 Al-Shabaab al-Mujahideen, ‘Regarding Various Non-Governmental Organizations and Foreign Committees Working in Somalia’, Press Release, July 20, 2009, available at http://nefafoundation.org//file/nefa_shabaab0709.pdf (last visited July 11, 2012), and also available (with a slightly different translation) at http://patronusanalytical.com/files/Al_Shabaabs_NGO_liaison_office_announces_closure_of_UN_offices_in_Somalia.php (last visited July 11, 2012)
 Al-Shabaab al-Mujahideen, ‘OSAFA Fact-finding Committee Conducts Organization Performance Appraisal’, Press release, 28 November 2011, available at http://ansar1.info/showthread.php?t=37132 (last visited July 11, 2012).
Blacklisted organizations: UNHCR, WHO, UNICEF, UNFPA, UNOPS, FSNAU, Danish Refugee Council (DRC), Norwegian Refugee Council (NRC), Concern, Norwegian Church Aid (NCA), Cooperazione Internazionale (COOPI), Swedish African Welfare Alliance (SAWA), the German Agency For Technical Cooperation (GTZ), Action Contre la Faim (ACF), Solidarity and Saacid.
 Al Shabaab Al-Mujahideen (2011).
 OCHA (2012).
 Elhawary (2012).
 ICRC, ‘Somalia: ICRC temporarily suspends distributions of food and seed’, ICRC News Release 01/04, January 12, 2012, available at http://www.icrc.org/eng/resources/documents/news-release/2012/somalia-news-2011-01-12.htm (last visited July 14, 2012)
 Médecins Sans Frontières (MSF), ‘MSF Closes Two Large Medical Centers in Mogadishu after Killings of Staff ‘, MSF press release, January 19, 2012, available at http://www.doctorswithoutborders.org/press/release.cfm?id=5733&cat=press-release (last visited July 14, 2012)
 Source for staff and sites: ICRC Annual Reports, 2005-2011, available at http://www.icrc.org/eng/resources/annual-report/index.jsp (last visited July 14, 2012). Source for security incidents: Humanitarian Outcomes (2012), Aid Worker Security Database, available at https://aidworkersecurity.org/ (last visited: May 27, 2012).
 Source: Humanitarian Outcomes (2012). The year 2008 marked a high point in the absolute number of attacks against civilian humanitarian operations. The subsequent decline in attacks is interpreted by some analysts as mainly due to the shrinking presence of international aid agencies in two of the most violent settings (Somalia and Darfur), rather than improving security conditions (see Stoddard et al., 2011).
 GHA database (2012)
 See Stoddard et al. (2009); and Humanitarian Outcomes (2012).
 Stoddard et al. (2011).
 Bruderlein, Claude and Pierre Gassmann, ‘Managing Security Risks in Hazardous Missions: The Challenges of Securing United Nations Access to Vulnerable Groups’, Harvard Human Rights Journal, Vol. 19, 2006.
 See Donini (2008); and Collinson and Elhawary (2012).
 Bruderlein and Gassmann (2006).
 OXFAM International, ‘OI Policy Compendium Note on United Nations Integrated Missions and Humanitarian Assistance’, OXFAM, January 2008, available at http://policy-practice.oxfam.org.uk/publications/united-nations-integrated-missions-and-humanitarian-assistance-115058 (last visited July 14, 2008).
 Bruderlein and Gassmann (2006).
 UN Security Council Resolution S/RES/2041 (2012). March 22, 2012.
 Featherstone (2012).
 Donini (2011).
 Hofman and Delaunay (2010).
 Featherstone (2012).
 Krahenbuhl (2011).
 Donini (2009).
 Bruderlein and Gassmann (2006).
 Hubert and Brassard-Boudreau (2010).
 ECHO, ‘Report on Security of Humanitarian Personnel: Standards and Practices for the Security of Humanitarian Personnel and Advocacy for Humanitarian Space’ European Commission, Humanitarian Aid Department, January 2004, available at http://reliefweb.int/report/world/report-security-humanitarian-personnel-standards-and-practices-security-and-advocacy (last visited July 14, 2012).
 European Union (2008).
 See Metcalfe et al. (2012). See also Whiting, Michael, ‘Military and humanitarian cooperation in air operations in Haiti’, Humanitarian Practice Network (HPN), Humanitarian Exchange Magazine, Issue 53, March 2012.
 Metcalfe et al. (2011).
 Period analyzed: 1998-2010. Violence data extracted from Humanitarian Outcomes (2012). Humanitarian response measured in total expenditure (constant prices 2009), extracted from GHA database (2012).
 ODA: period 1998-2009; source: GHA database (2012). US State Department and US military operations data: period 1998-2011. Sources for US Gov. data: USAID, ‘Budget Justification Fiscal Year 2001’, US Agency for International Development, 2001 (period 1998-2000); and Belasco (2011), period 2001-2011.
 Violence data extracted from Humanitarian Outcomes (2012). Humanitarian and ODA expenditures extracted from GHA database (2012).
 Holmes (2009).
 Donini (2009).
 Bruderlein and Gassmann (2006).
 Stoddard et al. (2009).
 Source: The Guardian, based on UNAMA’s reports, available at http://www.guardian.co.uk/news/datablog/2010/aug/10/afghanistan-civilian-casualties-statistics#data (last visited July 19, 2012).
 Another interpretation would be that aid workers and/or civilians were unintended collateral damage caused by the insurgents in their attacks. While it would show how little insurgents cared about IHL, there would be no direct targeting involved. However, given the typology of the attacks (kidnappings, ambushes, carjacking…), this hypothesis seems unlikely.
 Featherstone (2012).
 See Egeland et al. (2011); Stoddard et al. (2009); and Stoddard et al. (2011).
 Collinson et al. (2010).
 Krahenbuhl (2011).