George was followed home. As he reached for his keys to his apartment in Tel Aviv, he was startled by a voice from behind.
‘Pack your belongings’, a policeman ordered. George was told he had a week to return to South Sudan or be detained indefinitely in Israel.
George had originally fled South Sudan for Egypt during the Second Sudanese Civil War in the 1980s. He failed to find secure protection in EgyptFootnote 1 and so crossed the Sinai Dessert in 2008, entering Israeli territory with the help of smugglers. Between 2007 and 2012, roughly 60,000 other East African asylum seekers crossed into Israel in a similar manner. Upon reaching the country, they could not apply for refugee status. In 2012 all South Sudanese nationals, numbering approximately 1,200, were told they must return or face detention.Footnote 2
As the policeman drove away, George called Operation Blessing International (OBI), a humanitarian organization with offices in Jerusalem. He asked for help returning to South Sudan, and was given a free flight home and travel documentation. By 2012, nearly all South Sudanese in Israel had repatriated via similar means.
It is against international law to indefinitely detain asylum seekers without first establishing if they are refugees.Footnote 3 What is less obvious is whether humanitarian organizations should help individuals return to avoid such detention.
The UN claims it should.Footnote 4 Over the last decade, it has helped 7.2 million refugees repatriate, many from detention.Footnote 5 They help because, even if governments detain refugees,Footnote 6 the UN is using no coercion itself, and is helping refugees obtain freedom through repatriation.Footnote 7 It is analogous, one could claim, to civil servants clandestinely helping individuals flee persecuting regimes. During the Rwandan genocide and the Holocaust, such civil servants were celebrated as helping individuals escape injustices.Footnote 8 Of course those who fled were coerced; that is why it was commendable to help them.
Yet unlike fleeing danger to safety, refugees who return home may be trading one injustice for another. In this case, ‘repatriation facilitators’, as I refer to them here, cannot normally justify their actions by appealing to the outcomes of return. In this eventuality, NGOs have justified their assistance by referring to refugees’ consent.Footnote 9 But it is unclear if there is consent, given the presence of coercion.
Most of today’s debates on immigration focus on the actions of states and their agents towards migrants and refugees.Footnote 10 There have been fewer empirical studies on NGO and UN roles in immigration control, and even fewer discussions on the ethics of such agencies. In this article, I consider what dilemmas repatriation facilitators face, and how they may better address them.
In the following section I will elaborate on my methodology, which combines empirical political science and normative political philosophy. I will use this methodology to address three ethical dilemmas that emerged from my empirical research. I will first address the coercion dilemma, which occurs when facilitators help with coerced returns without causally contributing to the coercion. I will then address causation dilemmas, in which facilitating return does causally contribute to coercion. Finally, there is a dilemma that cuts across both types of dilemmas: in risk dilemmas, NGOs and refugees lack sufficient data to calculate the risks of return. In such cases refugees cannot come to an informed decision, even if coercion can be mitigated.
Before proceeding, it is necessary to precisely state the aims and clarify the assumptions of this article to avoid misunderstanding about the highly contentious questions addressed.
I shall consider the ethicsFootnote 11 of facilitators’ actions, rather than their legality, focusing on whether these are consistent with NGO and UN aims of ensuring only safe and voluntary returns.Footnote 12
The refugees under consideration are primarily those who the UN claims should not be forcibly returned, but instead given asylum or the opportunity to apply for refugee status. These are individuals whose lives will likely be at risk from persecution if they return.Footnote 13 Using the UN definition permits discussion of the UN’s facilitation dilemmas according to its own standards. In a similar vein, I use the definition of coercion provided by the International Organization of Migration (IOM), a major global repatriation facilitator. According to the IOM, coercion occurs when one is repatriating to avoid detention, but also when one lacks basic necessities if they stay, such as food and shelter.Footnote 14
Though I mostly focus on refugees fleeing persecution, I will at times discuss individuals fleeing food insecurity, lack of medical care and general violence.Footnote 15 I assume that coercing such ‘survival migrants’Footnote 16 to leave is unethical if the state has the capacity to accept such individuals, and if accepting these migrants is the only way to ensure that they obtain basic human rights. This claim is supported not only by philosophers who believe in open borders, such as Joseph Carens,Footnote 17 but also by those who defend states’ right to exclude immigrants, such as David Miller, Matthew Gibney and even some states themselves.Footnote 18 As such, it serves as a ‘minimal ethical standard’Footnote 19 determining when the state should not deport, while still leaving open the question of who repatriation facilitators should help return. For simplicity, I will refer to individuals as ‘refugees’ even if their return is unsafe for reasons related to general violence or food insecurity, rather than persecution.
Though I make the above assumptions throughout the article, one may accept my general conclusions while still disagreeing on who deserves asylum. My goal is not to settle the debate about who states should protect, but to resolve the puzzle of who should be helped to return by the aforementioned organizations if governments are coercing individuals to leave.
METHODOLOGY
The data I present in this article serves a similar purpose as court judgements in jurisprudence, medical cases in bioethics and thought experiments in moral philosophy. The data serve as the dilemmas we start with, which have yet to be addressed in contemporary debates in political theory or public policy.
Political theorists raising ethical dilemmas typically rely on cases found in the existing empirical literature.Footnote 20 I will also refer to the existing literature, citing cases of return from a diverse range of countries. However, there are very few in-depth studies of repatriation; the few that exist tend to describe the aggregate experience of repatriating populations, and often only before return.Footnote 21 With some exceptions,Footnote 22 studies rarely explore how individuals may have been subject to different injustices both before and after repatriation. I aim to capture this range of dilemmas for NGOs and the UN with my own in-depth empirical research on repatriation from Israel to South Sudan.
A diverse range of examples is useful for normative theorizing. For example, in fictional trolley examples used in moral philosophy, a person must always decide if it is just to kill one person to save five, but the details of each example vary, with one case involving pushing a man to his death and another involving pulling a switch.Footnote 23 The variation between cases highlights whether our intuitions change in response to new variables, helping us determine if these variables are normatively significant. I will employ a similar approach when formulating a general ethical guideline for repatriation.
I choose to focus on the case of NGOs in Israel because there is evidence that they had managed to facilitate return that was voluntary and safe in the midst of widespread government detention. I seek to determine if they succeeded in this endeavour even though the United Nations High Commissioner for Refugees (UNHCR) struggled to do so.Footnote 24 Unlike repatriation facilitated by UNHCR, OBI in Israel was helping a relatively small number repatriate, and so had significant resources to interview each refugee individually, to ensure they were not coerced into returning. Importantly, they took no government funds, relying on private donors, to avoid acting as an arm of the government’s immigration goals.
To determine if return was voluntary and safe, I interviewed refugees after return to learn about their reasons for repatriating and their conditions after return. I first travelled to Juba, Aweil and Wau in South Sudan in March and April 2012, interviewing twenty-seven individuals. In the summer directly following this trip, the Israeli Government announced that all were required to repatriate.Footnote 25 Almost all 1,200 South Sudanese nationals in Israel returned at this time, mostly via NGOs. I therefore travelled to East Africa again to interview these new returnees. Because many migrated to UgandaFootnote 26 shortly after returning to South Sudan, I conducted fieldwork in Kampala and Entebbe in 2013, interviewing twenty-nine returnees.
In December 2013 I reached South Sudan a second time, arriving two days before the South Sudanese Civil War broke out. I stayed for six weeks, interviewing sixty-one individuals in Juba who returned from Israel, roughly half of whom were Nuer and forced by Dinka militias to flee their homes to the UN’s internally displaced persons (IDP) camp. I also interviewed nine returnees who had fled to Ethiopia in June 2014, visiting them in Gambella, situated along the border with South Sudan.
When I arrived in each country, I called two to five contacts provided to me by repatriation facilitators and friends in Israel. I then used a snowball methodology to interview their acquaintances, their acquaintances’ acquaintances and so forth, until all links were exhausted. After transcribing each qualitative interview, I coded responses for subjects’ reasons for return, including detention or threats to deportation in Israel, and their conditions after return, including food access, income, medical care, education, shelter, displacement, and deaths from medical complications, cross-fire and ethnic-based violence.
My approach contrasts with survey-based studies that ask closed-ended questions, which ask respondents to rate, on a scale, their access to food, their level of security and so forth.Footnote 27 While such surveys are invaluable for recording important variables, they have the drawback of limiting the full range of refugees’ responses, which I tried to capture fully through in-depth qualitative interviews.
The interviews I conducted were not with a randomly selected sample. This was partly because I could not obtain a full list of phone numbers of those who returned, or access those in especially insecure areas. Nonetheless, I strived to obtain a sample that was as diverse as possible, and sought to counteract survivor bias whenever possible, travelling extensively within each town and surrounding rural areas, interviewing returnees without cell phones, secure housing or access to close tarmac roads. During the war, I also conducted interviews in both UN IDP camps in Juba, including one in the Jebel neighbourhood, where ethnic cleansing was especially widespread. Overall, there was most certainly still bias towards those who were better off and alive, but this was mitigated to the best extent possible, given the constraints.
One might suppose that we cannot rely on the responses of those who returned. They may misrepresent how much they were coerced to return, and how difficult their conditions are today, especially if they were not satisfied with their choice to repatriate. My method of sampling was designed to mitigate this possibility. Because I interviewed individuals living in a diverse range of countries and regions, a significant portion was very satisfied with their return, but still recalled being coerced into returning, and later fleeing their homes after their return. If even these individuals recall similar challenges to those who regretted repatriating, this provides stronger evidence of the accuracy of such testimonials. I also witnessed conditions described by respondents, such as overcrowding, unhygienic latrines, food scarcities and soldiers firing into IDP camps. As such, I could corroborate the responses of many interviewees regarding these conditions.
This original data from Israel and East Africa are central to this article. However, I situate it within the broader range of repatriation cases. The case of Israel is not unique because of the dilemma NGOs faced. What was unique was the NGOs’ greater financial investment to avoid these dilemmas. If such extraordinary measures fail to succeed, this highlights the depth of the problem and the need for an ethical analysis.
COERCION
Coercion dilemmas occur when NGOs and UN agencies are faced with the choice of either helping with return or watching refugees face confinement in camps, detention or an inability to access basic necessities. I will first describe this dilemma, and then consider how we might resolve it.
The Dilemma
In 1991 two million Kurdish refugees fled Iraq, most hoping to reach Turkey. They reached a mountainous area separating the two countries, but Turkish officials refused to grant them entrance. While current theorists focus on the Turkish policy,Footnote 28 there was also an ethical dilemma for NGOs: they could do nothing, forcing refugees to stay in the mountains, or help them return to Iraq, and risk being killed.
Within four days, 1,500 died from exposure, and the rest were uncertain what would happen if they stayed. Like in Israel, no NGOs claimed that the Turkish Government’s response was ethical. But helping with return seemed preferable, because the Turkish Government refused to change its policy regardless.Footnote 29
One might suppose that coercion dilemmas are not relevant when claims for asylum are heard in wealthier countries where, one might hope, genuine refugees would be given residency rights and freedoms. Yet even when claims are heard, strict evaluation criteria mean many refugees are denied refugee status.Footnote 30 They are then detained and wish to repatriate. Some do, with the help of NGOs, and end up displaced again after their return.Footnote 31 Even if one believes that states have acted legally according to a strict definition of international law, it seems unlikely they are acting ethically, and so it remains unclear whether NGOs should assist with such returns.
This dilemma may be more pronounced when states lack the capacity to accept refugees. In such cases, states may deny refugees the right to work as well as lack the means to provide them aid to survive. This was the case between 1982 and 1984 when Djibouti both denied refugees work visas and reduced their rations, compelling many to return to Ethiopia.Footnote 32 More recently, Burundian refugees faced a choice between living in camps in Tanzania or returning to Burundi without access to basic necessities.Footnote 33 Similarly, the Ugandan Government recently revoked land from South Sudanese refugees, and refugees in both Uganda and Kenya are often confined to camps, limiting their freedom.Footnote 34 In such cases, we may blame wealthier states for failing to provide aid to poorer host states to increase their capacity to accept refugees,Footnote 35 but if refugees feel compelled to repatriate without this aid, it remains unclear whether NGOs and the UN should help with return.
As noted, the current academic discussions focus almost entirely on state injustices,Footnote 36 but the few scholars who do discuss the ethics of UNHCR repatriation tend to assume that a coerced return is, by definition, unprincipled.Footnote 37 Their position is that UNHCR uses a distorted definition of ‘voluntariness’, in which a refugee in detention is considered sufficiently free to consent to return,Footnote 38 and that UNHCR has a ‘repatriation culture’.Footnote 39 This critique is incomplete. Though the definition of voluntariness is skewed and the culture of repatriation problematic, UNHCR may be helping with involuntary returns because doing nothing is far worse.
When OBI began its repatriation programme in 2010, it too was facing a coercion dilemma. Though South Sudanese were not prevented from working,Footnote 40 they were denied legal residency, and some were detained. They could not apply for refugee statusFootnote 41 and, even if they could, their claims would likely be denied, as Israel provides refugee status to only 0.25 per cent of applicants.Footnote 42
Though conditions were difficult in Israel, returning to South Sudan entailed significant risks. South Sudan had only recently emerged from decades of war, mainly fought between southern Sudanese opposition forces and the ruling northern Sudanese forces.Footnote 43 From 1991, the southern Sudanese forces split into two opposing groups, one mainly from the Dinka ethnic group, and the other mainly of the Nuer ethnic group.Footnote 44 When South Sudan eventually achieved independence from northern Sudan in 2011, a coalition government was formed in Juba, comprised of both Nuer and Dinka ethnic groups, but the president stifled dissenting voices,Footnote 45 and inter-ethnic violence continued into 2012, with thousands of civilians killed that year alone.Footnote 46 As a result of the instability, the country lacked basic services, including food security and healthcare.Footnote 47
Given that Israel let South Sudanese work in 2011, and given the conditions in South Sudan at the time, many refugees stayed in Israel. Consider, for example, Alek, who explains why she had initially left South Sudan, and why she did not return in 2011:
I am from Unity State, and we fled the war to […] Khartoum when I was a young girl. Later, I married there, and had four kids, and crossed into Israel, via Egypt, in 2007. I was in prison for half a year, but then released, so decided to stay. It was good. I worked, at first, in the Renaissance hotel in Tel Aviv. The kids went to school.Footnote 48
But others wished to return, such as Joseph:
My state is Lega State […] I was born in Khartoum in 1982, but came back to South Sudan from 1995 until 2000, so I was familiar with Juba. I went to Egypt in 2000, and in 8 August 2005 I went to Israel […] I went to prison for one year, and after one year they released us. I worked in a hotel, but could not get an ID, or legally start a business. So I saved $20,000. I was in touch with my family in Juba, and so asked for help returning.Footnote 49
Joseph was one of the first refugees to return. At the time, many human rights organizations opposed OBI’s assistance, claiming Joseph and others had few rights, and so their return was involuntary.Footnote 50 In response, OBI hired the Hebrew Immigrant Aid Society (HIAS) to interview each refugee, asking them ‘Why do you want to return?’. If an individual said they were only returning to avoid detention, their return was viewed as involuntary and not supported.
In total, OBI and HIAS helped roughly 900 individuals return between 2009 and 2012. Once an asylum seeker left Israel they could not re-enter Israeli territory.Footnote 51 But OBI and HIAS were convinced that this choice, though irreversible, was entirely voluntary.
OBI’s intentions seemed genuinely humanitarian. It was a Christian humanitarian organization with a strong history of providing food, shelter and medical assistance to all denominations in developing countries.Footnote 52 It had never, until 2010, been involved in repatriation. Nor had HIAS, a humanitarian organization founded in 1881 to assist Jews fleeing pogroms in Russia and Eastern Europe, but which later focused on helping non-Jewish refugees, resettling 3,600 refugees from Vietnam, Cambodia and Laos in the United States.Footnote 53 HIAS said it opposed repatriation in other contexts, refusing to assist with repatriation from Kenya due to the risks involved.Footnote 54 In Israel it made an exception, as it could conduct individual interviews to ensure there was no coercion.
In total, of the 126 subjects I interviewed, sixty-seven returned because they thought life was better in South Sudan and not only to avoid difficult conditions in Israel. However, there was a marked distinction between those who returned prior to and after 2012.
That year, thousands of Israeli citizens marched through the streets of Tel Aviv, calling for the expulsion of African asylum seekers, described by the prime minister as ‘flooding the country’Footnote 55 and by one politician as a ‘cancer to the body’.Footnote 56 Legislation was passed to detain asylum seekers,Footnote 57 and all South Sudanese were told they must returnFootnote 58 or face imprisonment,Footnote 59 with the exception of those who had medical problems. Alek describes life during this period, and why she changed her mind about staying in Israel:
Every day started with a mess. You go outside and they tell you, ‘Go back to your country! Why are you here? Your country has money! Go home!’
In June they took my husband’s visa and said, ‘We will not give you a new visa’. We were left without work for two months. I said ‘What? What will I do […]?’ So I thought, ‘I will say thank you to God that we are healthy and go back’.Footnote 60
Alek called OBI, which eventually agreed to help her return.Footnote 61 Hundreds of others soon followed. Of the 126 subjects I interviewed in East Africa, thirty-seven had returned to avoid detention, and thirty-six returned partially or wholly because they could no longer work, fearing they would lack access to basic necessities if they stayed. Fourteen left because they feared deportation.
It is not immediately clear whether OBI’s first policy of refusing coerced returns was better than its second policy of supporting such returns. Neither was more principled than the other. It may seem ethical to only help with voluntary returns, but this would force refugees to stay in detention. The case demonstrates that the dilemmas of repatriation cannot be avoided even when working independently from the government, and even with the best of intentions and resources.
Resolving the Dilemma
To resolve this dilemma relating to coercion, we must address a pressing question: whether refugees can truly give their consent when faced with coercion. In many cases outside the sphere of repatriation, consent may very well be valid even if there are only injurious alternatives. A patient is perfectly capable of giving consent to life-saving surgery, even though the alternative to surgery is death. As such, some philosophers argue that cases of ‘third-party coercion’ are also cases of valid consent.Footnote 62 For example, imagine that Abbey threatens to shoot Babu if he does not buy Cathy’s watch. Cathy sells Babu her watch because she does not want him shot by Abbey. Babu’s consent seems valid for Cathy, even if not for Abbey. One could similarly argue that refugees’ consent is valid for repatriation facilitators, even if it is not valid consent for the government.
However, according to a number of ethicists, consent would be invalid for Cathy if she could easily persuade Abbey to put her gun down.Footnote 63 Cathy should do this, instead of selling her watch. In other words, Cathy’s duty is to get Abbey to stop threatening Babu, and therefore Babu’s consent is not valid for Cathy. This approach is consistent with the Good Samaritan principle, which holds that agents should help those in great need if they easily can.Footnote 64 If there is nothing that Cathy can do, then Babu’s consent is perfectly valid for her, but not if she can easily help stop Abbey’s violent threat.
With repatriation to dangerous countries, we may ask if a facilitator can easily raise money for basic necessities and legal aid to avoid detention. If instead it raises money for repatriation, then it fails to honour the Good Samaritan principle. Of course, basic necessities may be an ongoing cost, while repatriation is a one-off cost. But if a refugee lacks necessities after they have returned, it is unclear if the repatriation facilitator can simply ignore their needs. If they owed them this aid before return, an action absolving them of this duty without alleviating the need seems unethical.Footnote 65
In addition to Good Samaritan duties, organizations may have costlier duties. Humanitarian organizations in particular, which were created precisely to protect vulnerable populations, should be held to a higher standard in protecting these populations. This may translate to special duties, such as lobbying for policy changes, providing legal aid and raising money for necessities. Demanding costly duties from Cathy, by contrast, could infringe upon her right to a personal life. While organizational staff also have a private life, they have voluntarily agreed to allocate an insulated portion of their lives to the goals of the organization, so their personal lives are not infringed upon.
Some organizations may also have costly duties because they have significant power.Footnote 66 When an agent has power, they have a greater ability to help others, and so may have a greater duty to help.Footnote 67 For example, a doctor on a flight may have a duty to save a life because she can more easily do so, even if this is difficult for her. Similarly, Medicins Sans Frontiers may have a duty to widely publicize a famineFootnote 68 because they are more able to do so. If repatriation facilitators have a greater ability to publicize the plight of refugees and lobby for the end of coercive conditions,Footnote 69 they should take these actions, even if they are more difficult than only helping with repatriation.
There are situations in which repatriation facilitators do work hard to end coercive conditions, but fail to create any change. In such cases, assisting with return may be legitimate. For example, when Kurdish refugees were trapped between Iraq and Turkey, NGOs tried and failed to persuade the Turkish Government to provide them asylum. More refugees were likely to die from exposure, and so NGOs acted ethically when helping with their return. Similarly, had OBI and HIAS worked hard to end detention, but failed, perhaps helping with return would have been legitimate, so long as South Sudanese nationals were aware of the risks.
This conclusion is predicated on the assumption that repatriation does not itself causally contribute to coercion. If there is such a causative link, then further considerations become relevant, which I will now address.
CAUSATION
Causation dilemmas encompass three categories of causal scenario. In all three, helping refugees to repatriate causally contributes to coercive government policies. As such, return should generally not be facilitated, with some exceptions.
Simple Counterfactual Causation
In cases of simple counterfactual causation, an agent causes an event if, had the agent not acted as she had, the event would not have occurred.Footnote 70
If the government is detaining refugees to encourage return, and an organization makes return possible, this can motivate the government to detain more refugees than it otherwise would. The IOM is an example of an organization that may have such an impact. It visits survival migrants in detention around the world, taking down their details and trying to secure their passports so they can repatriate, when they otherwise would not be able to.Footnote 71 If governments are only detaining refugees so that they repatriate, and refugees are only repatriating because of the IOM, then the IOM may be causally contributing to detention.
UNHCR may contribute to coercive policies in a similar manner. In 1994 and 1995 it began facilitating the repatriation of the Rohingyan refugees from Bangladesh to Burma. Soon after, the Bangladeshi Government significantly increased its pressure on refugees to return, seeing that their return was now possible, as it was funded by UNHCR.Footnote 72 Similarly, in 2012, one Israeli Knesset report states that an NGO had established that repatriation for South Sudanese was possible, and so the government could endorse a more aggressive return policy for those who had not yet returned.Footnote 73
Facilitating return may also increase a government’s capacity to use coercive measures. When OBI helped a refugee return from detention, the government quickly filled his cell with a new refugee, who had previously not been detained, maintaining the government’s policy of filling the detention centre to its maximum capacity.Footnote 74 Thus in this scenario, repatriation efforts directly determine the rate of detention at a given time.
The case of Israel raises an additional complication, overlooked in the examples above. OBI and HIAS were not the only agents facilitating return. The government also began its own repatriation programme in 2012,Footnote 75 eventually returning thousands of asylum seekers.Footnote 76 In other countries, UN agencies, multiple private charities and refugees themselves pay for transport home.
In such a scenario, any single NGO helping with return may seem to have no impact on the level of coercion, nor may it have any impact if any single means of repatriation fails. If existing bodies have the capacity to repatriate all refugees, a single NGO may well not causally contribute to coercion. For, were it to discontinue its repatriation services, refugees would still be able to repatriate at the same rate, via a different facilitator. However, if the other facilitators are incapable of facilitating all refugees, then each facilitator directly contributes to the rate of detention. The more agencies that are available for repatriation, the more refugees can repatriate, freeing up cells for further detention.
When NGOs’ actions are necessary for coercive policies, coercion is not a mere background condition, but is dependent on repatriation. This leads to a simple argument for NGOs discontinuing repatriation services, related to the Good Samaritan principle: refraining from helping with return is costless. If this costless act of omission helps refugees avoid detention and coercive conditions, then, as organizations created to help others, they should exploit this omission to efficiently achieve their goals.
We might argue that, in some cases, causally contributing to coercion does not harm refugees. In my sample, some refugees did not particularly mind that the government threatened to detain them or revoke their visas, because they would have returned regardless, for reasons unrelated to coercion. Some missed their families, or wished to contribute to the development of their country.
Even for these cases, it may be wrong for NGOs and the UN to help with return, because it is wrong to causally contribute to coercive policies, even if those subject to coercion do not feel subjectively worse off. For example, imagine again that Abbey puts a gun to Babu’s head, telling him to buy Cathy’s watch, but Babu secretly wanted to buy the watch regardless. When Cathy sells her watch, she may be making Babu’s life better in some ways, but she is also causally contributing to Abbey’s act of raising a gun to another person’s head. In such cases, Cathy should refuse to sell Babu her watch if she knows that this refusal will make Abbey put down her gun. She should wait until Abbey does this, and only then sell Babu her watch.
In a similar way, NGOs and the UN should avoid encouraging governments to detain refugees, as the act of detention is especially unjust, even if many refugees would have returned regardless. Repatriation facilitators should wait until the government ends detention, and only then agree to help with return.
Causation as Influence
There are instances in which repatriation is not necessary for coercion, and so does not cause coercion in the counterfactual sense. Facilitating repatriation may still be wrong according to other ethical criteria. Sometimes a person wrongly causes an event by influencing it, even if their actions were not necessary for the general event to occur.Footnote 77
For an example of such causation, imagine there is an assassin, and she pulls her trigger, leading the bullet to shoot out of her barrel into the heart of a victim, unjustly killing him on the spot. She also has a hundred backup assassins, who would have killed the victim had she not killed him first.Footnote 78 As such, she was not necessary for his death, or even almost necessary for it.Footnote 79 She still causally contributed to his death if she influenced the particular way the death transpired. This would be the case if, in a world without her, the bullet would have flown in a slightly different direction, piercing the victim’s heart in a different place, while in a world without other assassins, her bullet would have still flown in the same direction, piercing the victim’s heart in the same way.
In such cases, even if the assassin influences the event, and so causally contributed to it, we might still claim that she did not influence it in a way that harmed the victim; he would have been killed regardless. Nonetheless, as noted above, we have duties to avoid causally contributing to injustices, even if the victims are made no worse off from the injustice. The assassin may be acting wrongly by causally contributing to the injustice that occurred, regardless of whether the victim is worse off compared to a world in which the assassin does not pull her trigger. In a similar sense, a single NGO may be wrongly causally contributing to an injustice even if, had they not provided return, the general injustice would still have occurred.
In cases where we causally contribute to injustice by influencing the event, such causal influence may still be justified if the influence is significantly helpful for the victim. The assassin, for example, may know she can more accurately shoot the victim directly at the centre mass of his body, leading to a quicker death, compared to the backup assassins. If the assassin is in no way responsible for the presence of other assassins, and is shooting the victim only to reduce suffering, pulling the trigger may be morally justified. In a similar manner, an NGO can justifiably help with repatriation in cases where, though the help causally contributes to unjust coercion, it can also ensure a much safer return than would otherwise take place. However, unless the NGO is quite certain that its actions significantly help with return, it should avoid helping with repatriation, to avoid causally contributing to injustice.
Uncertainty
In some cases, a given NGO has essentially no influence. Its actions are not necessary for coercion, and it does not influence coercion or the safety of return. This may be the case if there are multiple NGOs, each one providing equally safe repatriation, such that if one pulled out, the level of coercion and safety of return would be the same. Similarly, there may be only one NGO, but the government is detaining refugees both to encourage return and to placate protesters, or to deter new refugees from arriving in the country. We might suppose that an NGO assisting with return here in no way causally contributes to coercive policies. For, had it not been for repatriation, there would still be other decisive reasons for the government to detain refugees. In such cases, an NGO may still have a strong reason to avoid helping with return.
An agent has a reason to avoid an act if she subjectively suspects that it may increase the probability of a harmful event occurring,Footnote 80 even if she is not ultimately necessary for the outcome and does not influence it. Imagine two assassins pull their triggers at the same time, both bullets flying out of their barrels simultaneously, piercing the victim’s heart in the same location at the same moment, such that neither assassin influenced his death.Footnote 81 One reason that each assassin acted wrongly is that, at the time she pulled her trigger, she could never be 100 per cent certain the other would pull her trigger. In choosing to pull her own trigger, she increased the probability, in her mind, of the death occurring.
When there are multiple facilitators helping with return, then neither can ever be 100 per cent certain that the others will make return possible. In choosing to help with repatriation, they risk possibly increasing the chances of repatriation occurring, and thus the chances of coercion occurring. Similarly, when the government has multiple reasons for using coercive policies, the NGO can never know for certain that the government will still detain refugees in the event that repatriation is no longer a possibility. As such, repatriation should be discontinued, so that NGOs are certain that they are not causally contributing to injustice.
Nonetheless, an exception may be made if the government has a large number of reasons for detaining refugees, such that detention would almost certainly continue even if repatriation ceased. Helping with such coerced returns is not ideal, but may be ethical, as the causal impact on coercion is unlikely, and the benefits significant if refugees truly wish to return.
We have, as such, reached a general conclusion: coerced repatriation should only be facilitated if it does not significantly contribute to the coercive policies, and if all efforts have been made to first stop the coercive policies. Such repatriation is ethical on balance, assuming refugees are aware of the risks.
RISKS
Refugees cannot always be aware of the risks of returning, as there is often insufficient data on the country they are returning to. Current discussions of repatriation tend to overlook this point, focusing on the importance of a safe return.Footnote 82 But the most unsafe areas in the world are the most difficult places to conduct research in, and so the risks are most difficult to calculate. When the safety of return is impossible to establish, there is a risk dilemma.
Such was the case for Ugandan refugees living in Zaire and Southern Sudan in the 1980s,Footnote 83 Rohingya refugees living in Bangladesh in the 1990sFootnote 84 and South Sudanese refugees living in Israel in the 2000s. Some had last lived in their countries as young children, or not at all, their parents having fled before they were born.Footnote 85 The problem is not only that refugees often do not know what will happen if they go back; there is often insufficient data to even predict the likelihood that they will be harmed.Footnote 86
Consider how this impacted South Sudanese refugees in Israel. Seven subjects, including Alek and George, returned to Unity State, where 140,000 had been displaced in 2012,Footnote 87 but where even an estimated death toll has never been publicized.Footnote 88 Twenty-three subjects returned to Upper Nile, three returned to Abyei and one returned to Warap State, where tens of thousands were displaced in all three areas, at least hundreds killed, but the precise number of displaced and killed is unknown.Footnote 89 Ten returning were from Jonglei. Here, data are slightly more complete, but still sparse. One estimate states that 200,000 were displacedFootnote 90 and at least 2,700 civilians killed in 2011–12,Footnote 91 but the precise number of deaths was never confirmed, and the total population of Jonglei had never been accurately counted in a reliable census.Footnote 92 Seven returnees were from the town of Akobo in Jonglei, where between 250 and 1,000 civilians were killed between 2011 and 2012, but the precise number was never confirmed, and the total population of the town never counted.
In discussions of informed consent, scholars emphasize that individuals should be told about possible side effects and the chances of improvement from an intervention.Footnote 93 Such informed consent is rarely possible for refugees. This is because there is rarely detailed data available on atrocities as they are occurring. Indeed, even after an atrocity, data are rarely available, except in rare cases, such as the Holocaust and the Rwandan genocide.Footnote 94 In most cases, no organized logs are available for refugees to know the risks of returning.
I asked subjects why they returned, given the unknown risks. Most responded that it was precisely because they did not know the risks that they returned. Alek used a comparative judgement:
I was in prison for six months in Israel. I didn’t like it. If I don’t know what it’s like in South Sudan, but I know I hated prison in Israel, I would prefer to go to South Sudan […] it might have been worse, but it might have been better.Footnote 95
Alek is from the Dinka tribe, but grew up among the Nuer, and speaks the languages of both tribes fluently. Two years after her return, Dinka militias came to her home, believing she was Nuer. She fled, returning two days later to find her furniture and clothes stolen. ‘When we come home’, she explains, ‘people on the street look at us. They don’t ask questions. They don’t know what tribe I’m from’.
She does not regret her choice to repatriate from Israel,Footnote 96 but others did. Of the fifty-eight adults I interviewed at the end of 2012–2014, thirty-two wished they had never returned at all. Samuel, who returned from Israel to avoid detention, felt this way:
I was at home in Juba when soldiers started shooting at 10:00pm. We thought we were not included in the problem. But soldiers shot at my friends who stayed with me. Two were killed. I put on a pair of shoes – they weren’t mine – and ran. I left all my things. I arrived in the UN IDP camp on the December 17th. I cannot leave. So I am in a prison. I think, ‘In Israel prison would have been better because my enemy is not outside of the prison doors’.Footnote 97
OBI did not ignore the possibility that many would regret returning because of the lack of information on the conditions in South Sudan. To address this problem, the NGO called returnees and visited them after repatriation, to find out if each individual was safe. If they were, OBI took this as evidence that other refugees could safely return to these areas. However, when I visited families in secondary towns and rural villages, and those in IDP camps, they explained that they never received a call or visit from OBI. This was partly because they had no electricity, and so their cell phones rarely worked. As such, it appears OBI’s attempts to visit returnees suffered from a type of survivor bias. Those most difficult to contact were the least able to access regular food, medicine or housing. Once an individual fled South Sudan, they no longer held the same cell phone numbers, and so those I interviewed in Ethiopia or Uganda were never in touch with OBI.
As such, despite pouring significant resources into flying regularly to South Sudan, the organization largely failed to learn about individuals after return. In 2012, towards the end of OBI’s project, aid organizations that had assisted families while they were in Israel made efforts to contact them after their return. They found that at least twenty-two of 500 returning children died within their first year post-repatriation.Footnote 98 More children likely died during this period, as most were never contacted by any aid organization. Out of the twenty-seven adults and children I interviewed in South Sudan in 2012, only five had access to any income or food security. All twenty-nine I interviewed in Uganda in 2013 lacked a basic income after return.
During the war, I learned about the conditions of an additional 110 returnees, based on the sixty-one interviews I conducted in 2013 and 2014. Of the thirty-two returnees who were of the Nuer ethnic group, all had fled Dinka militias. Twenty-four of these individuals had no income or family support before fleeing to IDP camps. They lived off less than one meal per day, mostly consisting of corn flour, and so failed to obtain the basic nutrients necessary for survival according to World Health Organization standards.Footnote 99 Thirty-seven of the 110 returnees were living outside of the IDP camp but inside South Sudan, nineteen of whom had no income and also lacked food security. Twenty-five had left South Sudan, and only two of these individuals had an income. The remaining returnees were without basic medical care or food security. In total, I learned of one subject killed in crossfire during the war, and four killed because of their ethnicity, including two children shot at gunpoint, aged three and five. There were most likely more I never heard about, due to survivor bias in my sample.
Given the uncertain risks, should NGOs and the UN assist with return?
When risks are certain, then refugees should perhaps be able to decide for themselves whether they would prefer to stay in detention or return. This is because the trade-off between freedom and security is highly subjective. People have very different answers to the question, ‘would you rather stay indefinitely in a cell, or have a one in ten chance of being killed over the next twenty years?’. When risks are unknown, we cannot ask a refugee this question. As such, when facilitating return, NGOs cannot be certain they are providing a service that is consistent with refugees’ preferences. Yet one may also feel that refugees have the right to consent to a future with unknown risks, rather than stay in detention.
To address this dilemma, repatriation facilitators should gather more information on past returnees, act to counter the effects of survivor bias, and establish a more complete picture of the environment refugees will be returning to. When possible, the facilitators should venture into rural areas, rather than relying on cell phones alone.
Often, even when information can be gathered, it soon becomes irrelevant as political conditions change in a country of origin. For example, Juba was relatively safe in 2012, but unsafe for Nuer citizens in 2013. Because conditions can change, NGOs and the UN should lobby for re-admittance agreements, allowing refugees who returned to re-enter the host country. In the 1990s, the governments of Sweden, France and the United Kingdom allowed Bosnian refugees to either briefly travel to Bosnia before making a final decision or to re-enter after returning.Footnote 100 On a more limited scale, UNHCR organized ‘go-and-see’ visits for Burundian refugees in Tanzania.Footnote 101 More generally, refugees may have a ‘right to regret’Footnote 102 their decision, and return to the host country, given the risks involved in returning. Importantly, if a given NGO is the only body making return possible, they might refuse to help with return until the government agrees to allow such readmission to refugees.
Ultimately, repatriation facilitators may fail to persuade the government to institute such a policy, and research may still be impossible to conduct. In such cases, facilitators should at least be explicit to refugees about their inability to disclose the risks, and recognize they may not be providing a safe return, but merely a flight home to an unknown future.
CONCLUSION
When a refugee is detained, her choices are far from voluntary. Given that this is the case, humanitarian agents have two options, neither ideal. They can help with an unsafe return, and free refugees from detention, or refuse to help, forcing them to stay. In reality, this dilemma comes in three forms, requiring three distinct policies.
In the first form, the government will arrest refugees, force them into detention or deny them visas regardless of whether they return. NGOs and the UN should lobby for an end to such policies, and appeal to donors to provide food security and shelter. If they fail, it may be ethical to facilitate return, so long as refugees are aware of the risks.
In the second form of the dilemma, repatriation causes coercion. Facilitators are not mere third parties, as their actions impact government policies, whether intentionally or not. The more refugees are able to repatriate from detention, the more spaces become available in detention centres. This not only allows the government to detain more refugees, it gives them a reason to, seeing that past detainees were persuaded to return. In such cases, NGOs and the UN should not help with return unless their assistance has only a small impact on coercion, and ensures a much safer return than would otherwise take place.
In all cases where assistance is ethical, the risks of return should still be clearly communicated to refugees. When risks are not known to facilitators, and refugees know little about the country they are returning to, it remains unclear if return should be provided. If facilitators clearly communicate to refugees that the risks are unknown, perhaps refugees should still have the choice to repatriate. But in helping with such return, refugees may be returning to a country where, had they known the risks, they would not have returned. As such, repatriation does not truly assist refugees, and may be contrary to the very goals of the NGO.
The general lessons I raise may be applicable to areas outside of refugee repatriation. NGOs often provide assistance to vulnerable populations who may feel forced to accept the assistance because their options are constrained by coercive government policies. For example, NGOs may provide housing to those who have been displaced, food to those whose land has been illegally revokedFootnote 103 or medical assistance to those who have been tortured.Footnote 104 Though these cases may be different from repatriation, it might still be useful to consider if NGOs have the ability to end coercive policies, if their assistance causally contributes to further coercive policies, and if recipients are aware of the risks of accepting various forms of aid.
Further research is necessary to determine how generalizable my conclusions are. For now, we can at least consider how NGOs and the UN, when helping with repatriation, might change their policies and practice.
Today, repatriation facilitators spend little of their budget on lobbying for the end of coercive conditions, and more on flights, stipends and coordinating return. This is partly because NGOs and the UN often rely on government grants, at times competing with other NGOs to repatriate refugees at the lowest possible cost, at the fastest possible rate.Footnote 105 But even NGOs that raise their own funds, such as OBI, continue allocating their entire budget to repatriation, feeling pressure from refugees who want to return as quickly as possible, to avoid detention. Though refugees have good reasons to return quickly, NGOs have good reasons to slow down return, freeing up resources for lobbying, and possibly dissuading governments from detaining quite so many refugees. Such a policy shift for NGOs may mean fewer refugees can return, but fewer may want to if conditions improve in the host country.
Even when conditions will not improve, because governments refuse to change their policies, NGOs and the UN can still ensure a more informed return. They can conduct more rigorous post-return evaluations, interviewing former refugees in rural and insecure areas, to determine displacement and mortality after return. If researchers cannot access unsafe areas, they may still be able to call relatives of those who returned, to find out their conditions. More importantly, they can still travel to safer surrounding countries and interview those who have fled after repatriating.
Today, the UN explicitly states that it lacks the capacity to conduct such research.Footnote 106 This may be because the current budget is earmarked for repatriation itself, paying for the transport of hundreds of thousands of refugees annually.Footnote 107 Funds should be shifted from maximizing the number who return towards establishing a minimum quality of information to provide refugees before return.
When George called OBI in 2012, it might have implemented a different policy, in light of these conclusions. It is unclear that George’s detention was inevitable. The NGO might have done more to persuade the government to provide George residency, or to provide greater residency rights for South Sudanese nationals in general. OBI should also have waited to facilitate this return, to see whether the government would eventually free him, seeing that he had no way of going back. Even if continued detention was likely, George was not making an informed choice. He was forced to fly to a country he knew little about. When he landed, he failed to find permanent shelter, employment or security.Footnote 108 In some ways, this was not OBI’s fault. The NGO lacked access to information on the risks, and there was a limit to how much information they could find. But if information is not available, perhaps a ticket should not be available either.
For George, and millions of others, immigration control involves not just force, but assistance. How organizations provide assistance can impact how governments respond, and how refugees react. If we are to have a fuller picture of what an ethical refugee policy would entail, we must shift our focus away from the policeman who followed George home, and onto NGOs who sit in small offices, answering calls from refugees who feel they need help returning, and quickly. While the urgency is clear, the best policy is not.
