Introduction

Pope Benedict XVI, Russian President Vladimir Putin and United Nations Secretary-General António Guterres have at least one thing in common: they each, at different times and in reference to different contexts, called for or ordered the opening of so-called ‘humanitarian corridors’. Whether it was to evacuate wounded civilians in South Ossetia in 2008, to implement a daily ceasefire in Syria’s Eastern Ghouta in 2018, or to assist populations in Ethiopia’s Tigray region in 2021, respectively, the notion is now so frequently invoked that it goes unnoticed in mainstream public discourse despite having no legal basis or strictly agreed upon definition.

Supposed to allow the unobstructed deployment of humanitarian aid and/or the evacuation of civilians, humanitarian corridors have most frequently been used in contexts of armed conflicts to secure passage through disputed territory. Their existence is grounded in what constitutes a core objective of the law of war: protecting civilian populations from military operations. But the fact that they are by definition temporary and limited in geographical scope undermines existing obligations of all parties under international humanitarian law to allow civilians in areas affected by fighting to leave in search of safety and impartial aid to reach those in need. It is on that basis that the concept is not one which the International Committee of the Red Cross (ICRC) wishes to promote, even though it acknowledges its possible necessity for other humanitarian organisations (ICRC, 2003) and has resorted to it on occasions.¹

Numerous examples of failed corridors and safe zones becoming the target of attacks during the post-Cold War era also suggest a disconnect between the apparent straightforwardness of the concept and its practical enforcement. Whether political forces engaged in armed conflict – state or non-state actors – will consent to the opening of a humanitarian corridor depends on how concerned they are about their population, on the importance they assign to their international image and how these two factors hang in the balance compared to conflicting military interests. Yet arguably, it is also the vagueness of the concept itself which allows for diverging interpretations and, therefore, convenient ambiguity.

This article is a historically informed review of humanitarian corridors both as a modus operandi and as a terminology. It draws on an internal reflection from Médecins Sans Frontières (Doctors Without Borders) around the applicability of the notion to contemporary fields of intervention. A potential bias of this analysis is that it examines in priority instances gone wrong, thus omitting a number of contexts where the outcome was possibly more favourable. The purpose of this summary is therefore not to rule dogmatically for or against humanitarian corridors in general, but to explore its grey areas and to analyse in which cases, to what extent, under which conditions and at the cost of what possible manipulations the use of such corridors can be beneficial for the people concerned.

Origins of a Notion

Borrowed from the Italian corridore, meaning ‘narrow passage’ or ‘place where one runs’, the word corridor was used in the seventeenth century in the military sector to refer to the path that went around the fortifications and allowed rapid communication with troops. The term then underwent numerous changes in meaning in architectural semantics before being picked up by other fields such as biology and geography, where it was used in a more or less metaphorical sense to designate a passage, most often long and narrow, connecting two spaces. It is difficult to date the exact time when the image was applied to humanitarian action but worth highlighting that the idea to supply or evacuate a besieged city is not specific to humanitarianism.

Even when such operations made it possible to save human lives, they were first implemented by States themselves in the form of airlifts supporting politico-military ambitions. Well-known examples are the 1948–49 airlift by the American army into the city of Berlin, which served to provision both the garrisons of the Allied Forces and the civilian population during the blockade, and Operation Poomalai by the Indian Air Force over the besieged town of Jaffna in June 1987, intended to spare civilian casualties as well as to support the Tamil separatist movement somewhat symbolically. Organising mass evacuation of civilians has also long been the prerogative of States, for humanitarian, political or military motives. Famous examples from the Second World War include Operation Kindertransport in which Jewish children were evacuated from areas under Nazi rule to the UK, and Operation Pied Piper in which British civilians, mostly children, were relocated from high- to low-risk areas to protect them from aerial bombings.

The airlift technique was applied to humanitarian action as early as 1968 in Biafra, to circumvent the blockade imposed on the enclave by the Nigerian army. Considered the world’s largest civilian airlift, with 5,300 flights delivering 60,000 tons of humanitarian supplies over a period of about two years, the Biafran airlift resulted from the efforts of an ad hoc coalition of international non-governmental organisations and the ICRC, frustrated by the failed negotiations with the Nigerian government. But it wasn’t until the late 1980s and early 1990s, at a time marked by the resurgence of nationalist tensions following the disappearance of East–West antagonism and when the media coverage of the humanitarian consequences of multiplying intra-State conflicts favoured international intervention, that the humanitarian corridor terminology became part of the interventionist diplomacy lexicon.

Under the aegis of the United Nations, States tried to contain conflicts within their borders by creating neutral zones linked to humanitarian corridors supposed to encourage populations to stay or return to their country of origin by guaranteeing them assistance. The nature and modalities of assistance delivery thus radically changed to meet the needs of internally displaced populations, with humanitarian action increasingly deployed within zones of disputed sovereignty and no longer on their periphery. Consequently, humanitarianism at the heart of conflict dynamics and containment policies combined led relief operations to depend on the existence of humanitarian corridors in Sudan in 1989, Iraqi Kurdistan in 1991, former Yugoslavia in 1992 and Rwanda in 1994 (Jean, 1997).

The Absence of a Normative Framework or Consistent Definition

The legal concept of ‘relief corridors’ appears in resolution 45/100 of the UN General Assembly from 14 December 1990. It alludes to the principle of free access to victims enshrined in the Geneva Conventions and reaffirmed in numerous resolutions of the UN General Assembly and Security Council since 1988 but it is nowhere defined (Nyabeyeu Tchoukeu, 2018). In fact, in spite of the rise of the ‘right to interfere’ and ‘responsibility to protect’ doctrines and the Security Council urging States to allow ‘effective and unimpeded’ delivery of humanitarian aid under increasingly binding terms, to the point of sometimes authorising the use of force to secure assistance operations, no subsequent resolution explicitly refers to corridors as a concrete way to deliver humanitarian aid and/or to organise the evacuation of civilians. Perhaps a notable exception is resolution 2165 of 14 July 2014 on Syria, which authorises the opening of border crossings meant to allow the delivery of aid to opposition-held areas – an idea very similar to that of humanitarian corridors (Gillard, 2013; Hall, 2021).

Whether it is by land, sea, river or air, a humanitarian corridor is a strip of territory that is supposed to allow the unobstructed deployment of humanitarian aid and/or movement of civilians. This basic definition notwithstanding, the notion is ambiguous on both the intended objectives and the exact nature of the access constraints the corridor is supposed to circumvent. The two functions of a corridor, whether it be to transit aid or civilians, can co-exist at the same time. In the latter case, the corridor may be used both for the repatriation of a population (as was the case in Turkey-Iraqi Kurdistan in 1991–96 and in Zaire-Rwanda in 1996) and for its evacuation, which remains the most common case. The evacuation can be general and massive or selective, depending on who is deemed to be under threat and vulnerable. For example, in Syria, the term ‘humanitarian corridor’ has been used widely to refer to the ‘sheltering’ of all civilians (in Aleppo in 2016, in Eastern Ghouta in 2018 and in Rubkan in 2019), whereas in reference to Gaza, the term referred to evacuation routes for the wounded only, set up as part of humanitarian pauses lasting a few hours. More recently, the term took on yet another meaning to designate a specific mechanism of relocating refugees from Lebanon, Ethiopia and Libya to Italy and France, under memoranda of understanding signed between public authorities and civil society organisations.

The intended objectives of a corridor are largely contingent on the context in which it is implemented. Most of the time, it provides access to an area whose sovereignty is challenged, during an armed conflict, as was the case in Bosnia in 1992–95, in Chechnya in 2000, in Darfur in 2003–06, and more recently in Ukraine, for the evacuation of the cities of Donetsk, Luhansk and Horlivka in 2014.² But the opening of a humanitarian corridor can also be a response to border closure in peacetime. In 2014, for example, during the Ebola epidemic, Côte d’Ivoire and Senegal kept their borders closed with neighbouring Guinea and Liberia but opened humanitarian corridors to help expedite aid to these two countries heavily affected by the outbreak. Sometimes, corridors have as well been envisaged by third parties in an attempt to circumvent bureaucratic constraints – in Myanmar after cyclone Nargis in 2008, for example.

Feasibility and Protection Dilemmas

Paradoxically, humanitarian corridors can only exist in hostile but relatively stable environments, to connect two well-defined territories, each under the fairly permanent control of clearly identifiable authorities. Archetypal yet exceptional, such situations are those of enclaves such as in Biafra in 1968, in Bosnia in 1992–95 or in Liberation Tigers of Tamil Eelam-held Vanni in Sri Lanka until 2009. But insofar as they impact the overall dynamic of conflicts by distorting issues of territorial control, contradicting siege strategies and disrupting military operations, the consent of both parties to the opening of corridors is almost always extremely difficult to obtain and fragile, which explains why corridors are frequently attacked soon after they are opened (Hoffmann, 2020). In Chechnya in 1999, for example, corridors intended to allow civilians to flee bombed villages were themselves the target of bombings, so much so that some referred to them as ‘death corridors’. Syria is another case in point, with reports of major escalations in ground offensives in Eastern Ghouta just days after resolution 2401 calling for the deployment of humanitarian convoys was unanimously passed in February 2018 (Price, 2020).

Corridors enforced by military means are equally vulnerable to attack and even the presence of an international force mandated by a resolution of the UN Security Council has proven ineffective, if not counterproductive. In former Yugoslavia and Rwanda, the corridors connecting safe zones to the outside world were targeted as much as the zones themselves, by both parties. Between July 1992 and January 1996, for instance, the Sarajevo airlift was the target of over 270 incidents caused by both the Bosnian Serbs to prevent humanitarian aid and the Bosnian Muslims to provoke Western politicians into military intervention.

Humanitarian corridors are also likely to have the negative side-effect of diverting international attention from atrocities happening elsewhere and can be a deliberate stunt by belligerents wishing to regroup, build up arsenals or redeploy their forces. Such accusations have for instance been made against all parties during so-called ‘humanitarian pauses’ in Syria. Therefore, contrary to the image conveyed, the ethical dilemmas posed by humanitarian corridors in terms of protection concern both what happens within the corridor’s perimeter and its potential wider impact on hostilities.

Better Some than No Access?

Despite the difficulty to secure them, humanitarian corridors are often seen as a necessary compromise. The matter of their efficiency in terms of aid delivery is rarely addressed, both because they are usually deployed where there is no alternative access and because no one can know in retrospect what kind of assistance would have been possible had the corridor not existed. But their restrictive nature merits a few comments. Regardless of their specificities, whether they cross borders, frontlines, or both, all humanitarian corridors have two essential characteristics: they last only during a humanitarian pause or a resolution and rarely throughout a conflict, and they cover only a specific strip of territory. Yet, referring to an area (or moment in time) as ‘special’ or ‘humanitarian’ entails portraying everything around it as ‘not humanitarian’. Therefore, questioning the impact of humanitarian corridors requires looking at what happens both within the delimited space and outside it, not only in terms of protection as pointed to above but also in terms of needed assistance.

Inside the special zone supplied by dedicated corridors, one could argue that the aid channelled rarely meets the needs. Syria’s Bab al-Hawa border crossing from Turkey, for example, was depicted as a ‘crucial lifeline’ supposed to allow aid to reach the three to four million inhabitants of north-west Syria. But corridors often only give access to a small area, while the population in need of assistance is generally scattered.³ Additionally, the appreciation of who needs assistance and who doesn’t is not immune to political interpretation and manoeuvring, as illustrated by the competing narratives put forward during Security Council discussions around the successive extensions of Syria’s cross-border aid mechanism.

Corridors are also not excluded from more direct forms of diversion or confiscation of aid through looting or blocking of convoys and control of beneficiaries. In this respect, the factors that led to the famine in Bahr el-Ghazal in 1998 during Operation Lifeline Sudan are well documented, including the way Khartoum and the Sudan People’s Liberation Army controlled aid distribution points and denied foreign agencies direct access to beneficiaries, despite the opening of ‘corridors of tranquillity’.

More generally, just as the setting up of ‘safe zones’ may come at the price of abuses committed outside them (Keen, 2017), setting up humanitarian corridors usually implies renouncing broader access and potentially leaving other needs unaddressed. In that respect, humanitarian notification systems also known as ‘deconfliction arrangements’⁴ may be corridors’ most contemporary and, arguably, subtle incarnation. A response to the rise of aerial strikes on civilian sites particularly in Syria and in Yemen, such systems are supposed to reduce hazard for humanitarian personnel through the sharing of geographic coordinates of aid operations and movements with warring parties. But further to the fact that the arrangement has failed to prevent all incidents, it has the potential to restrict access by putting humanitarian organisations in the position of having to wait for tacit clearance or risk being denied protection.

A Fig Leaf of Respectability for Highly Political Intentions

Corridors are inevitably part of a wider context of instrumentalisation of humanitarian rhetoric, best illustrated by the Ethiopian government’s June 2021 announcement of a so-called ‘humanitarian ceasefire’ in Tigray, all the while continuing to hamper aid delivery. Apart from the intended or unintended negative side-effects presented so far, there is a number of ways in which humanitarian corridors have been blatantly manipulated in history. Occasionally, supply routes opened on humanitarian grounds have been abused by armed groups to smuggle weapons and munitions or recruit and repatriate troops. More often, belligerents have agreed to them for no other purpose than to keep up appearances on the international stage, soften their image and increase their credibility – for example when Russia announced the opening of humanitarian corridors in Grozny in 1999 a few days after opening fire on civilians fleeing the city.

By establishing mechanisms for the exchange and transfer of populations which upset the balance of power, evacuation corridors are particularly prone to being used for war strategies. Parties may use them to force a population to move from a rebel zone to a government zone, as was the case in Syria in February 2019 when Russia unilaterally opened two humanitarian corridors to help the displaced population living in Rubkan return to government-held areas, after blockading the camp and obstructing assistance for years. Corridors intended to evacuate a besieged city may also serve war objectives by rendering it a legitimate target (even if this means negotiating the exit of rebel leaders before the final assault, as was the case in Mosul in 2016) or, in extreme cases, endorsing a policy of ethnic cleansing by forcibly emptying a region of its inhabitants (a clear objective of Serb nationalists in the Bosnian war, for example). Conversely, civilians have been used as human shields by rebel groups keeping them hostage in an attempt to secure military positions or to force governmental forces to commit war crimes. This was, for example, the strategy of the LTTE when the Sri Lankan government troops advanced on the Vanni region in early 2009.

Ultimately, what matters is the reason underlying the decision to open or to call for a humanitarian corridor. Corridors can be misappropriated by a third party or the international community to conceal a military intervention or, on the contrary, to create a diversion, reassure public opinion and hide political impotence in the face of the real problem. Several authors have analysed, for example, the policy of mass humanitarian aid provision in Bosnia in 1992 as a diversionary strategy due to the lack of politico-military initiatives to end the conflict (Jean, 2004). As pointed out earlier, there are also situations where calls for humanitarian corridors or cross-border mechanisms are considered ways to circumvent onerous administrative requirements and to force affected states into consenting to in-country presence. A well-known example of such political manoeuvring is in Myanmar in May 2008, where advocates of the right to intervene in France, the UK and the United States spearheaded calls for the opening of air and sea corridors following Cyclone Nargis.

Conclusion: Translating Historical Analysis into Practice

Far from the harmless right of passage they are made to embody, humanitarian corridors resemble more a timid assertion of the principle of free access to victims. Their normalisation in public discourse should therefore prompt caution. Corridors may be a useful tool to implement temporary emergency interventions in hard-to-reach areas while continuing to advocate for more permanent and broader access, but the cost of building a separation between a humanitarian and a non-humanitarian space shouldn’t be underestimated. In that sense, one could argue that so-called ‘deconfliction systems’ alter the principle of free access to victims into a series of negotiated exceptions and ultimately hinder the enforcement of international humanitarian law.

Without overlooking or underestimating contextual specificities, practitioners might consider taking away three practical highlights from this short historical review. First, they might prefer avoiding the terminology of humanitarian corridors because of its profound ambiguities and vulnerability to political exploitation as well as the false sense of safety the notion projects. Instead, it may be more useful to call an evacuation corridor exactly that, or to claim temporary and restricted access to a specific area. Second, the option of negotiating a restricted ‘right of passage’ might be better used as a last resort in contexts of open fighting and certainly not made the ‘new normal’ of humanitarian action, or practitioners risk mistaking ‘passage’ for ‘access’ in the face of bureaucratic impediments. Third, the political consequences of participating in the evacuation of civilians should be carefully weighed, both in terms of their prospects for return and in terms of what such an evacuation implies for those left behind.