Australian Year Book of International Law
As far as I am aware Ivan Shearer is not known for people trafficking. But he has been active for many years in international humanitarian law as well as in refugee law and that accounts for his indirect responsibility for this article. It was his lectures on refugee law at the Institute of International Law and International Relations in Thessaloniki in 1982, which I attended, that inspired my interest in refugee law, and this essay on people trafficking is an attempt to explore how refugee law may be used to address the international protection needs of victims of trafficking in human beings.
People trafficking is not a new phenomenon. For centuries, across the globe, people have been sold into servitude: compelled to work for somebody else for little or no reward. Nowadays trafficking is most notorious in the sex trade and has become perhaps the greatest humanitarian tragedy to arise out of the collapse of communism in east-central Europe, although it is a worldwide phenomenon. It is, in most cases, a form of slavery. This article considers some ways in which international law may be used to protect the victims of trafficking.
There is on the face of it plenty of international law dealing specifically with trafficking. In addition to older anti-trafficking law, such as the United Nations (UN) Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others 1949, recourse may also be made to the UN Convention on the Elimination of All Forms of Discrimination against Women 1979, the UN Convention on the Rights of the Child 1989, its Optional Protocol on the Sale of Children, Child Prostitution and Child Pornography 2000, international laws against slavery, and various International Labour Organization (ILO) conventions. Furthermore, general human rights law, as well as laws on the protection of migrant workers, may play a role. In addition to these instruments there is developing a specialised regime in the European Union context, the Council of Europe has sought to address the issue, as has the UN High Commissioner for Human Rights and even the International Maritime Organisation has adopted anti-trafficking measures. Furthermore, particularly in the European context there have been adopted a plethora of soft-law measures under the auspices of the International Organisation for Migration, the Organisation for Security and Cooperation in Europe and the Stability Pact for South Eastern Europe. Despite this not insubstantial activity, there was a real need for an internationally acceptable binding instrument. Accordingly, when the UN Convention against Transnational Organised Crime (UNCTOC) was opened for signature in Palermo in December 2000, simultaneously adopted was the Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (Trafficking in Persons Protocol, or TIPP).
TIPP acknowledges in its Preamble that there was, despite the measures outlined above, ‘no universal instrument that addresses all aspects of trafficking in persons’. TIPP is intended to fill the gap. However, there are good reasons to consider alternative legal means of tackling the crime of people trafficking and assisting its victims. First, TIPP is not universal. It required only 40 ratifications and accessions to enter into force. That means that there may be many states in which trafficking in human beings occurs that remain unbound. Second, the effectiveness of the Protocol will depend on the will and ability of states to enforce it individually and to cooperate in that enforcement. In the meantime, trafficking is a major growth industry and is responsible for some of the most egregious breaches of human rights taking place in the world today. Third, TIPP deals with trafficking as a transnational offence only, albeit with a rather generous definition of ‘transnational’. It does not purport to deal with trafficking that takes place entirely within one country. Thus a woman who has been trafficked in her own country but succeeds in fleeing to another would not be a victim of trafficking for the purposes of TIPP. Nor would a woman who has been trafficked within one foreign country, who in any case would, in the normal scheme of things, be expected to rely upon the protection of her home state, should she escape those controlling her, rather than that of the country within which she has been trafficked. It must be stressed that trafficking in human beings does not have to be transnational; rather TIPP only addresses transnational trafficking in human beings. Other treaties, like the South Asian Association for Regional Cooperation (SAARC) Convention on Preventing and Combating Trafficking in Women and Children for Prostitution 2002, apply also to trafficking in human beings that is wholly within national frontiers.
Of course trafficking can have many purposes. Most notoriously this entails moving women and children to other countries for their exploitation in the sex trade, but there are other reasons: adoption of children, forced labour, use of children in criminal activity, begging, military uses, sports, arranged marriage and trafficking in organs.
This paper considers how the Convention Relating to the Status of Refugees 1951 (Refugee Convention), as amended by the Protocol Relating to the Status of Refugees 1967, may be used to assist victims of trafficking by speculating on the extent to which such persons might be considered as members of a particular social group. In particular, it focuses on women who have been trafficked in the sex trade, generally the most public and egregious example of the practice. However, the conclusions drawn may have a wider application.
Perhaps the crucial novelty of TIPP is that it treats those who have been trafficked as victims rather than criminals who have breached the destination state’s laws, particularly with regard to immigration and employment. TIPP requires parties to adopt measures to assist the victims of trafficking, including, possibly, even permitting them to remain in their territory on a permanent basis in appropriate cases. This provision recognises that there may be no option for the victim to return to their state of origin, precisely the basis upon which asylum may be granted under the Refugee Convention.
A key element of the 2001 Global Consultations on International Protection of the UN High Commissioner for Refugees (UNHCR) involved the promotion, strengthening and consolidation of appropriate procedures by receiving states for the welfare of asylum seekers. Such procedures may well play an important role in assisting victims of trafficking. However, they are concerned more with the day-to-day treatment of asylum seekers in the receiving state. This is exceptionally important: the regime may have a major influence on the immediate future of victims. But the aim here is a different one: to assess how the plight of the victims may be addressed at the fundamental level by inclusion in the definition of refugees contained in the Refugee Convention itself. If they indeed may qualify as refugees then their position and status in the states in which they find themselves will be that bit healthier because they will have access to the facilities available to other refugees, in particular shelter and health care, as well as protection against refoulement.
The issue with regard to victims of trafficking is whether the persecution to which they have been subjected is such that it may be covered by the Refugee Convention. Article 1A(2) of the Convention provides, in part, that a refugee is someone who:
owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it.
To qualify as refugees in this sense victims of trafficking will have to satisfy the following criteria:
• they must have a well-founded fear of persecution;
• that fear is the reason why they are outside their own country; and
• they are unable or unwilling to return to their country because of that fear.
Several issues arise here. First, most victims of trafficking, if they are in another country, will have a fear of the authorities in the country where they find themselves, rather than their home state, because they may have transgressed immigration and employment laws and are liable to sanction for doing this. TIPP attempts to address this by treating those who have been trafficked as victims rather than offenders. Therefore, it will be necessary for a victim of trafficking to demonstrate that her fear is because of persecution in her home state. This is not improbable, since most are recruited in their home state and may reasonably believe (and are frequently led to believe by their captors) that should they escape to their home states then they may easily be found again and punished or victimised for having fled.
The position is different where the trafficking has taken place within one country (the victim’s home state) and the victim escapes to another. In such a situation it will be easier to demonstrate that there is a fear of persecution with regard to her state of nationality.
The Refugee Convention requires that the asylum seeker be outside her country of nationality by reason of having a well-founded fear of persecution. Victims of trafficking will not normally be outside their country for that reason. Rather, they may have been effectively kidnapped or duped into believing that they were travelling abroad for some other kind of employment. They have not left because they feared persecution. However, being outside their country it is possible for them to acquire a well-founded fear of persecution should they return, precisely because of their involvement in the sex trade: either at the hands of the traffickers or in the community at large. This is recognised in the UNHCR Handbook, which provides that the refugee does not have to have left her country on account of a well-founded fear, if circumstances have changed while they were out of the country. It is specifically acknowledged that a person may become a refugee as a result of their own actions while abroad. If this is so for those who voluntarily take action that creates for them a well-founded fear of persecution, it must surely also be so, but even more legitimately, for those who are placed in such circumstances quite involuntarily, because of their exploitation by others.
Even if such a fear be established, another problem arises. In some countries, only persecution by the state has been recognised as legitimate to qualify someone as a refugee for the purposes of the Convention; persecution by non-state actors would not then be regarded as a qualifying factor. With trafficking the persecution is most likely to come from non-state actors such as organised criminal gangs. It is true that some states may, through incompetence, lack of resources, corruption of their officials or even connivance, encourage the practice but no state will ever admit that it engages in trafficking. Therefore, there might be an issue, in some countries, in qualifying trafficking as persecution for the purposes of the Convention. Nevertheless, most countries nowadays do accept that persecution may be by non-state actors where the state is unable or unwilling to offer protection, and trafficking by criminal gangs could qualify. This aspect of the test would require careful consideration of the facts with regard to the state of origin of the victim. In particular it would require an assessment of the conditions in which traffickers operate, as well as the ability, and will, of the national authorities to act against traffickers. Where it appears that traffickers are able to operate without interference from the authorities, perhaps even with their informal co-operation, it will be easier to establish that there is no effective protection available in the victim’s home state. This will be relevant to assessing the validity of the fear of persecution.
Inconsistencies amongst European Union (EU) member states on this matter have proved problematic. However, as part of the development of a common immigration and asylum policy for the EU, the acceptance of persecution by non-state actors has become entrenched. A proposal from the European Commission for a Council Directive on qualification for international protection (refugee status or subsidiary protection) (12 September 2001) sought to harmonise EU state regimes on the provenance of the persecution by proposing that it may also originate from non-state agents or actors where a state is unable or unwilling to provide effective protection. The finalised version confirms this.
A victim of trafficking may be persecuted if she returns to her home country in that she is ostracised by her family or community because she has been employed in the sex trade. This of course also entails persecution by non-state actors but, if such persecution is accepted, such ostracism could well amount to persecution for the purposes of the Refugee Convention. This raises the question, how bad must the feared persecution be for it to qualify? It is unlikely that the decision by family members to reject a woman because she had been trafficked would, of itself, be enough. Persecution is not defined in closed terms; the notion is flexible but clearly requires that the treatment feared must be a serious interference with basic rights such as a ‘threat to life or freedom on account of race, religion, nationality, political opinion or membership of a particular social group’. Other serious violations of human rights, if for the same reasons, would also qualify. Furthermore, where the individual has been subjected to various measures not in themselves necessarily amounting to persecution, such as discrimination, these various factors taken together might amount to persecution. Where a woman has been trafficked and this is known, her fear of rejection or ostracism may well amount to a fear of persecution, depending upon the society from which she originates. While she will not have been persecuted in this sense before being trafficked, it is what has happened in the process of trafficking that raises the risk. The questions will then be twofold but related: assessment of the well-foundedness of that fear on the part of the woman; and the gravity of the persecution facing her because of what has happened to her in the meantime. In some societies, it may be enough for her family to reject a woman to make her whole existence in her original community untenable. In others, while regrettable at the personal level, it will not prevent the woman being able to re-establish herself in her home state.
Assuming that the victim can demonstrate persecution for the purposes of the Refugee Convention, as well as a well-founded fear of returning to her home state, there remains one more hurdle: the persecution must be based on the victim’s race, religion, nationality, membership of a particular social group or political opinion. Of course, a woman may well be trafficked because she has a particular nationality or religion or political opinion; this is after all one way of victimising a particular group of people who are at the mercy of others as a consequence of an armed conflict, but it is more likely that she will be trafficked simply because she is a vulnerable woman.
In this scenario the most likely way to meet the criteria is to establish that the woman is being persecuted because she is a member of a particular social group. This has been one of the most keenly debated elements of the Refugee Convention, in part at least because it has been seen as a way to expand the scope of the Convention in line with evolving notions of who may constitute discrete groups in society.
The notion of what exactly constitutes a particular social group was extensively discussed in the Second Track of the UNHCR’s Global Consultations meeting in San Remo in September 2001. The summary conclusions from that meeting, which are intended to reflect broadly the understandings emerging from the discussion, include the following:
• There is no requirement that a group be cohesive in order to be recognised as a particular social group within the meaning of the Convention; that is, there need be no showing that all members of a group know each other or associate together.
• A particular social group is a group of persons who share a common characteristic other than their risk of being persecuted, and which sets them apart. The characteristic will ordinarily be one that is innate, unchangeable, or that is otherwise fundamental to human dignity.
It is well accepted that the fact of persecution alone is insufficient to establish a particular social group. There must be some other characteristic that the group shares in common, for instance that they have an innate or immutable characteristic, such as sexual orientation, that exposes them to persecution, or else one that is so important that they should not be required to change it, such as being human rights activists.  In other words, the persecution must be driven by one of the grounds recognised in the Refugee Convention; a group cannot generally be said to exist where the only thing those affected have in common is the fact of their persecution.
The issue then is whether victims of trafficking may be members of a particular social group. To establish this we cannot rely on the fact of persecution alone; there must be some other factor that they have in common, either innate or so fundamental that they should not be required to change it. In the case of victims of trafficking (for our purposes considering only women in the sex trade), what do they have in common other than that they have been persecuted? The European Commission’s Draft Directive pointed out that sexual violence may be perpetrated against women because of their nationality, religion or political opinion.
Such women would clearly fall within the definition of refugee because the persecution is against them specifically as women (sexual violence) but driven by their nationality, religion or political opinion.
In such cases it is not problematic to qualify the women as members of a particular social group. However, trafficked women are not usually trafficked because of their nationality or opinions, they are trafficked principally because they are women who are vulnerable and can be commercially exploited. The commentary to article 7 suggests that sexual violence can also be inflicted purely for the reason of gender, and that, in such situations, the persecution ground, member of a particular social group, could apply. Following this logic, sexual enslavement of women may also be said to be inflicted purely due to their gender and they may qualify as a particular social group: women who have been subjected to trafficking for the purpose of sexual enslavement.
Another possibility is that certain victims of trafficking, but not all, may qualify as a member of a particular social group if they come from societies that are likely to ostracise them (should they ever go home) should the fact that they have worked in the sex trade become known, although this would exclude those from societies or cultures that might be more tolerant and perceive the women as victims.
The fact that the victim is a woman is clearly unchangeable. The fact that she works in the sex trade is changeable if she is able to get out. But the fact that she has worked in the sex trade is unchangeable, and may expose her to persecution from her own society should she return there. It is surely also relevant here that the woman, if trafficked, by definition has not chosen to undertake this forced work. In this sense she has had the characteristic of being a former involuntary worker in the sex trade forced upon her. This is not fundamental in that she chooses to be a member of such a group: no such choice has been made, but it may be said to be innate in that it cannot be changed.
This discussion comes very close to relying on the fact of persecution alone to define the group, something that is clearly not accepted as sufficient to establish a particular social group. But the persecution feared in the future is not based on past persecution alone. Rather, it is based on their status as former workers in the sex trade. In particular, they are former sex-trade workers who have been forced into the work, but it is not the element of force, or persecution, alone that distinguishes them. This goes beyond characterising the group by persecution alone. In this context, the focus is not so much on the victimisation as such of these women (which might be said to inform much of TIPP) but rather on their victimisation because of who they are.
One thing is clear at this stage in the development of the international protection regime for victims of trafficking in human beings: there is a willingness by some tribunals to find ways to treat them as refugees where they are unwilling to be repatriated. Nevertheless, it is difficult to make out a clear case that all victims of trafficking in human beings as such may be regarded as a particular social group.
Perhaps the Convention definition of a refugee is not as restrictive as the above analysis indicates. For sure, the persecution must be feared with regard to the woman’s own state of nationality or residence, and in many, if not most, cases the victim has been trafficked to another country, thus suggesting, on the face of it, that protection is available by returning to the state of nationality.
In such cases the Refugee Convention is clearly redundant. But frequently the reality is significantly more complex. If we look at trafficking as a process involving many stages and participants in two or more countries (including source, transit and destination states) the picture changes. This is the approach adopted by TIPP. It deals with transnational trafficking but acknowledges that only relatively few players may be involved in the actual trafficking across frontiers. However, others may still bear responsibility because of their role in recruitment, harbouring and transporting of women, guarding them and perhaps subjecting them to physical, psychological and sexual abuse to break their will. In such cases the persecution may well be said to take place in the woman’s home state as well because of the role of each participant in the chain.
From this perspective, it becomes possible to discern a gap in the apparent restrictions of the Refugee Convention; a way to make it work to assist the victims of trafficking. Although the victim is physically present in the country to which she has been trafficked (and therefore at risk in that country), the original source of the risk is her home state. That is the state from which she was recruited, kidnapped or tricked into the power of the traffickers. That is the state, in other words, in which she first encountered those whose actions resulted in her being trafficked into another country. And that is, furthermore, the state where, should she be returned, she may be at risk, if not of victimisation by her own community or the local police, then of retribution at the hands of the traffickers who first took control of her. In that case the woman is at risk in at least two countries: her state of origin and the state to which she has been trafficked. From the perspective of refugee law, this might strengthen her case not to be returned to her state of origin, so long as the state where she has been trafficked (the destination state, where she has requested asylum) is able to offer effective protection. This is by no means fanciful. The destination state, if it is a western one, may well have the infrastructure and the resources (if not the political will) to offer effective protection.
A third alternative is that, should the victim succeed in escaping to a third country where she has not been trafficked, she may well be able to succeed in a claim to asylum vis-à-vis her own country.
Given the difficulties in using the Refugee Convention to secure international protection for victims of trafficking in human beings, it is worthwhile considering other options. The Convention should certainly not be regarded as the principal means of protection.
It has already been stressed that, even if a particular social group (victims of trafficking) exists, its significance will be limited because we are dealing, in the case of women who have been trafficked across international frontiers, with a risk of persecution in their home state if they are to be covered by the Convention. It will not operate to protect those in a foreign country – subject to the arguments set out above – unless it can be demonstrated that their state of nationality is unable or unwilling to help. The point argued here is that nationality, or citizenship, has to be effective nationality. Since the Refugee Convention aims to provide protection to those who cannot get it from their own country, it would defeat the whole purpose of the agreement to rely on a refugee’s nationality when that is in reality a pretence, that is, where the state of formal nationality is not offering the protection necessary.
The notion that nationality must be effective is well-recognised and has been analysed in the context of dual, or multiple, citizenship. While that is not the issue here, the test should be the same: is the state of nationality offering effective protection or not? The Refugee Convention requires that asylum seekers must have a well-founded fear of persecution in each country of their nationality, reflecting the ethos of the Convention, that protection should first be sought from the state of nationality before imposing a burden on other countries. If an asylum seeker has more than one citizenship, it means, in the normal scheme of things, that each of these countries must admit that person and thus provide protection. Only if each fails to do so, may international protection be granted.
This issue was central to a series of cases in the 1990s in Australia, concerning asylum seekers from East Timor, at the time under Indonesian control. The Australian government argued that they were not entitled to asylum in Australia. The issue was not one of persecution at the hands of the Indonesian government. Rather, it was argued that, because they appeared to have Portuguese citizenship, they should seek entry to that country rather than Australia. The issue was whether or not the Portuguese nationality was the effective nationality for the purposes of the convention. This entailed, at the least, admitting the individual to the national territory and providing a haven from persecution.
These decisions were without doubt correct in so far as they required that protection be genuine, real or effective. In other words, one could only be expected to seek protection from the state of nationality if it would meet these criteria. While the cases concerned dual nationality, the principle applies equally in cases of single nationality: the protection must be genuine, real or effective. Where the state of nationality is in fact unwilling or unable to provide protection, it is effectively rendering that person stateless, as far as international protection is concerned. In other words, lack of effective nationality for the purposes of refugee law is tantamount to statelessness for the purposes of refugee law. This is so irrespective of how many citizenships an individual possesses.
This may suggest another way to apply the Refugee Convention to victims of trafficking. Lack of state protection is de facto statelessness, which may amount to persecution. The Executive Committee of the UNHCR, on 5 October 2001, recognised the problem of statelessness that many victims suffer:
Strongly condemning the trafficking of persons, especially women and children, which represents a grave violation of their human rights; expressing concern that many victims of trafficking are rendered effectively stateless due to an inability to establish their identity and nationality status; [ExCom] calls upon States to cooperate in the establishment of identity and nationality status of victims of trafficking so as to facilitate appropriate resolutions of their situations, respecting the internationally recognized human rights of the victims.
One of the main difficulties faced by trafficked women is that their passports or identity papers are often confiscated by those who control them. Should they escape they may be unable to prove their identity or citizenship. If they are in a foreign country then they should be able to turn to the diplomatic or consular authorities of their home state for assistance. However, if that state declines to assist they may be rendered effectively stateless. A practical problem is that, frequently, the victim’s state may have no diplomatic or consular representation of its own, or the victim may be so far away that she is unable to get access to it. Such practical problems do not, and cannot, mean that the victim has lost her citizenship: in the absence of legitimate renunciation or some other act that has the effect of causing the loss of citizenship, it will subsist, and should the victim return to her state of nationality it will be obliged to admit her. However, that is not the situation here. Rather, we are confronted by a variety of circumstances that may result in the victim being unable, through no fault of her own, to seek or obtain the protection that her citizenship should give her: loss of passport or identity documents; confiscation of these documents by the traffickers; inability or unwillingness of the state of nationality to assist even when aware of the problem; lack of diplomatic or consular presence in the destination state; physical remoteness of diplomatic or consular presence from the victim.
Where such circumstances exist, the victim is unable to receive protection from her state of citizenship. She is not prima facie a refugee, unless she can demonstrate a fear of persecution with regard to her own state. However, the deliberate failure to assist by the state of citizenship may amount to persecution. In other words, the persecution here is by the state rather than the traffickers. The Preamble to the Convention Relating to the Status of Stateless Persons 1954 makes clear that merely being stateless does not make you a refugee. However, this restriction clearly refers to the status of the individual rather than the persecution. In this case, the issue is not their status; victims will still have to be a member of a particular social group. Rather, the crucial point is that, by making the victims de facto stateless, the state of citizenship may effectively be persecuting them. In this case, it would be legitimate for the victim to seek asylum in the destination state, assuming the other criteria for refugee status have been satisfied.
Deliberate and manifest disregard by the state for the fate of its citizens in another country is perhaps rather rare, or at least difficult to prove. More likely to arise is the situation where the victim’s state of nationality is unable to assist, meaning that there is no effective protection. In such cases it is rather difficult to speak of persecution by that country. But where there is effectively a failure to protect, what are the consequences of such failure? It may render the victim de facto stateless, in which case the destination state may have obligations towards her.
De facto statelessness arises because the individual is in a foreign country, is treated by the authorities as an alien, yet is unable to obtain the protection of her own state. Victims of trafficking are even more disadvantaged: not only have they been physically or sexually exploited and remain at risk of further exploitation, if they cannot get effective assistance their chances of escape and return to a normal life are further reduced. This has been recognised by the UNHCR.
At the very least, one may say that states should do all they can to reduce the incidence of statelessness. This is the ethos behind the Convention on the Reduction of Statelessness 1961. That treaty addresses deliberate acts by states that have the effect of causing statelessness, whereas here we are dealing more with statelessness by default (although it may be that some states act deliberately to cause statelessness in that, knowing of the problem, they nevertheless do nothing to address it). Article 7(3) of the convention provides, in part: ‘a national of a Contracting State shall not lose his nationality, so as to become stateless, on the ground of departure, residence abroad, failure to register or any similar ground’. This clearly seeks to prevent states legislating for deprivation of nationality in this way. But the mischief is the same, whether deliberate or not: if states acknowledge that they should not do this deliberately, it seems reasonable to query the legitimacy of a situation whereby statelessness happens through neglect, benign or otherwise.
The denunciation by ExCom of trafficking and its concern that it may lead to effective statelessness emphasises the serious consequences this may have for the victims. It is arguable that de facto statelessness, at least where the victim’s state knows about the problem but does not try to prevent it, may itself amount to persecution. Clearly, the prevention of statelessness would normally be in the domain of the government itself. The potential harm, in the sense of the rights that may be affected, that may be caused to a person who is rendered stateless is certainly serious enough to constitute persecution.
The analysis so far suggests that there is limited scope to classify victims of trafficking as refugees. In particular, establishing that victims are members of a particular social group will be a significant hurdle. But modern refugee law is not confined to the Refugee Convention. That instrument seems to suggest an all-or-nothing approach: you either are a refugee or you are not (although the stringency of the distinction is mitigated by the principle of non-refoulement). However, the reality is different. While most countries are exceptionally cautious about granting refugee status, many nevertheless permit aliens to remain on their territory even though they are not recognised as Convention refugees, because they accept that these people may nevertheless have legitimate fears about returning to their home states. The Refugee Convention, concerned only with those seeking asylum owing to a well-founded fear of persecution for one of the five recognised reasons, simply does not address all the causes that make people flee their home states or fear returning to them. But their fears are nonetheless legitimate. Once one adds to this fear the obligation of non-refoulement, it is readily evident why some aliens, not qualifying as Convention refugees, may nevertheless be permitted to remain.
Asylum seekers may have abandoned their own countries, or be afraid to return there, for a wide variety of legitimate reasons. These may include natural disaster, civil upheaval, armed conflicts, or indeed persecution by the state or non-state actors that breaches basic or fundamental rights not covered by the Refugee Convention. Victims of trafficking in human beings may belong to this category. Any woman who has been trafficked then rescued from her situation may have very good reason to wish to remain in the country to which she was trafficked rather than be repatriated. In particular, she may have legitimate fears of retribution by the traffickers, ostracism by her family or community, or harassment by police or other law-enforcement officials. Any of these factors may make it safer for the woman to remain in the state to which she was trafficked (or perhaps in some third state). Indeed, TIPP recognises that this may be a preferable option for some women and encourages states to assist them wherever possible, while offering little in the way of substantive obligation. The issue therefore is, to what extent the legal regime of subsidiary protection may operate to offer protection.
Destination states have adopted various methods to regulate the status and presence of such non-Convention refugees. The practice has become so common that attempts are now being made to formalise this non-Convention refugee status and define its content. In UNHCR parlance this is complementary protection; the European Commission calls it subsidiary protection: those who do not qualify as Convention refugees but are nevertheless in need of international protection.
The precise ambit of subsidiary protection ratione personae is uncertain. The UNHCR has stressed the need to reach some consensus on what this actually means, but has not itself come up with concrete proposals. However, the European Commission has, in September 2001, in the form of a proposal for a Council Directive. The principle of subsidiary protection is accepted. However, there remains disagreement as to the precise content of protection, and who should be entitled to it. In fact, while the extent of rights accompanying subsidiary protection status may be unclear, it is secondary to the core issue here, viz, whether victims of trafficking in human beings may qualify for the status in the first place.
The most recent EU practice defines as entitled to subsidiary protection persons who are not refugees but who nevertheless face ‘a real risk of serious harm’ if returned to their own country. ‘Serious harm’ includes, as one of three possibilities, ‘torture or inhuman or degrading treatment or punishment of an applicant in his or her country of origin …’.
The other possibilities are death penalty or execution, plus serious and individual threat to a civilian’s life or person by reason of indiscriminate violence in situations of armed conflict. The original proposal was even wider: it would have included violations of human rights ‘sufficiently severe to engage Member States’ obligations’ as well as threats to life, safety or freedom arising out of ‘systematic or generalised violations’ of human rights.
To what extent does this reflect a pre-existing international law standard? It certainly represented part of the practice of some EU member states, as the Commission made clear:
No specific EU acquis directly related to subsidiary or complementary protection exists but the ECHR and the case law of the European Court on Human Rights provides for a legally binding framework, which informed the choice of categories of beneficiaries in this Proposal. The categories and definitions of persons listed in this Chapter do not create completely new classes of persons that Member States are obliged to protect but represent a clarification and codification of existing practice. The three categories … are drawn very much from the disparate Member State practices and are believed to encompass the best ones.
In other words, the original proposal was considered in 2001 to reflect the existing practice of EU member states. However, even if this were so, the question remains to what extent that practice was based on duties under international law, and to what extent it represented merely non-binding compassionate or humanitarian will.
Victims of trafficking are most likely to argue that they have been subjected to, or are at serious risk of, torture or inhuman degrading treatment or punishment, and that is really the only one of the qualification criteria that we need to assess in this context. There is no doubt that the real risk of such treatment in future creates an obligation, at least on parties to the European Convention on Human Rights and probably at general international law, not to return a person to a territory where that risk exists. The risk here, of course, is not normally that the country where the victim is present will be the one where the torture or inhuman and degrading treatment is feared. Rather, the threat emanates from the state to which she might be removed. However, responsibility lies not only with the state where the woman is subjected to such treatment but also with the state that exposes her to a real risk of such treatment by expelling her.
If we consider the type of treatment to which victims of trafficking are exposed, there is little doubt that it may amount to torture, certainly to inhuman and degrading treatment. Some are raped by those who have control over them. Each occasion on which they are forced to engage in sexual acts with customers may also be rape if they have not consented. Even if this does not amount to torture it may certainly be inhuman or degrading treatment. Some trafficked women have been subjected to extreme physical maltreatment by their captors; this may also amount to torture. Being held in sexual slavery, being sold (perhaps at auction), being drugged by their captors, held in captivity with insufficient food and medical care, at risk of violence from their captors and customers: all of these may amount to inhuman and degrading treatment.
To benefit from subsidiary protection, a woman who has been trafficked will have to show that she fears being subjected to torture or inhuman and degrading treatment at the hands of persons in her own country, that is, the one from which she was trafficked. Exactly the same arguments apply as with regard to refugee status: the fear must be with regard to the home state, so she will have to show that she should be allowed to remain in the state to which she has been trafficked. Clearly, in this situation there are at least two potential sources of threat: from traffickers in source and destination states. If there is no threat in her home state then it will be appropriate for the woman to return there.
The threat of inhuman or degrading treatment may also arise due to the woman being ostracised by her own community should it become known that she has worked in the sex trade. In such cases the destination state may also have to offer subsidiary protection.
Finally, there are two other situations where a trafficked woman may be able to seek subsidiary protection. The first is where she has been trafficked within her own country but has escaped to another, and fears serious breaches of her rights should she return. The second is where the woman has been trafficked to another country but manages to escape to yet another. If it appears that the woman has a well-founded fear of being subjected to torture or inhuman or degrading treatment in both the source and in the destination state, then that state to which she has escaped may have to offer subsidiary protection.
In all of these cases, it will be necessary to demonstrate that the state with regard to which the trafficked woman has a well-founded fear is either unable or unwilling to offer protection. Otherwise, there is no need for international protection.
The subsidiary protection regime offers perhaps more hope to trafficked women than the Refugee Convention. While the latter may well serve to protect some victims of trafficking it seems unlikely to be available to all of them. The principal hurdles are establishing that the woman has a well-founded fear of persecution with regard to her home state (even though she has been trafficked to another) and that she is a member of a particular social group. These hurdles are not insurmountable but they may be problematic. With subsidiary protection, on the other hand, qualification rests on the existence of a well-founded fear of being subjected to serious and unjustified harm in the form of torture or inhuman and degrading treatment or punishment. It is an easier road than the long and winding one to refugee status and may be the one to absorb the traffic.
[∗] Department of Law, University of Wales, Aberystwyth.
 On people trafficking generally, see the websites of Anti-Slavery International <http://www.antislavery.org/> and the International Organisation for Migration (IOM) <http://www.iom.int/> . The latter contains special reports as well as the quarterly bulletin Trafficking in Migrants. See also United Nations Office for Drug Control and Crime Prevention, Global Programme against Trafficking in Human Beings, February 1999 <http://www.uncjin.org/CICP/traff_e.pdf> . See also K Knaus, A Kartusch and G Reiter, Combat of Trafficking in Women for the Purpose of Forced Prostitution. International Standards (2000), and related country reports published by the Ludwig Boltzmann Institute of Human Rights, Vienna, outlining the situation with regard to trafficking in individual countries. On trafficking of children see IOM, Trafficking in Unaccompanied Minors for Sexual Exploitation in the European Union (2001); and UN Commission on Human Rights, Report of the Special Rapporteur on the sale of children, child prostitution and child pornography, Ms Ofelia Calcetas-Santos, E/CN.4/1999/71. A recent overview of the regime is provided in R Piotrowicz, ‘Irregular Migration Networks: the Challenge Posed by People Traffickers to States and Human Rights’ in B Bogusz, R Cholewinski, A Cygan and E Szyszczak (eds), Irregular Migration and Human Rights: Theoretical, European and International Perspectives (2004) 137.
 For a concise discussion of slavery see R Jennings and A Watts, Oppenheim’s International Law (9th ed, 1992) 978-83.
 Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others (21 March 1950), 96 UNTS 271.
 Convention on the Elimination of All Forms of Discrimination against Women (18 December 1979), 1249 UNTS 13 esp art 6.
 Convention on the Rights of the Child (20 November 1989), 28 ILM 1448, esp arts 32, 34 and 35.
 Optional Protocol on the Sale of Children, Child Prostitution and Child Pornography (25 May 2000), 39 ILM 1290, esp art 3.
 Slavery Convention (25 September  LNTSer 19; 1926), 60 LNTS 253; Supplementary Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery (30 April 1956), 266 UNTS 40; ATS 1958, No 3.
 Forced Labour Convention (28 June 1930) (ILO Convention No 29); Abolition of Forced Labour Convention (25 June 1957) (ILO Convention No 105). More recently, the ILO adopted the Worst Forms of Child Labour Convention (17 June 1999) (ILO Convention No 182), supplemented by the Worst Forms of Child Labour Recommendation (17 June 1999) (Rec 190).
 See, for instance, the Universal Declaration of Human Rights, GA Res 217A (1948) art 4: ‘No one shall be held in slavery or servitude; slavery and the slave trade shall be prohibited in all their forms’.
 Convention on the Protection of the Rights of All Migrant Workers and Members of Their Families 1990 (18 December 1990), 30 ILM 1521. This entered into force on 1 July 2003 after achieving 20 ratifications, all of which are countries of emigration. Support from destination countries has been underwhelming.
 See, eg Council of the European Union, Council Framework Decision of 19 July 2002 on Combating Trafficking in Human Beings (2002/629/JHA).
 Rec No R (2000) 11 of the Committee of Ministers to member states on action against trafficking in human beings for the purpose of sexual exploitation (adopted 19 May 2000), and the Rec and Res referred to therein. See also the Commitment and Plan of Action adopted at the Budapest Conference of 20-21 November 2001 in preparation for the Second World Congress against Commercial Sexual Exploitation of Children. Most recently, in 2004 the Council was working on a Draft Convention against Trafficking in Human Beings.
 Office of the High Commissioner for Human Rights, Recommended Principles and Guidelines on Human Rights and Human Trafficking (2002).
 Interim Measures for Combating Unsafe Practices Associated with the Trafficking or Transport of Migrants by Sea (16 December 1998) MSC/Cird 896.
 See, eg IOM and EU, Brussels Declaration on Preventing and Combating Trafficking in Human Beings (18 September 2002) <http://europa.eu.int/comm/
justice_home/news/forum_crimen/2002/workshop/brusels_decl_en.htm>; OSCE Declaration on Trafficking in Human Beings (7 December 2002) MC.DEC/2/03; Decision No.557 – OSCE Action Plan to Combat Trafficking In Human Beings, (24 July 2003) PC.DEC/577; Stability Pact for South Eastern Europe ,Task Force on Trafficking in Human Beings, Statement on Commitments, 11 December 2002 <http://www.osce.org/attf/pdf/sofia/sofia_commstatement.pdf> .
 Convention against Transnational Organised Crime, A/AC 254/4/Rev 9, reprinted in (2001) 40 ILM 353.
 Protocol to Prevent, Suppress and Punish Trafficking in Persons, especially Women and Children (15 November 2000), 40 ILM 377. On TIPP, see A Gallagher, ‘Human Rights and the New UN Protocols on Trafficking and Migrant Smuggling: A Preliminary Analysis’ (2001) 23 Human Rights Quarterly 975; R Piotrowicz, ‘Traffic in People Gets the Red Light’ (2001) 75 Australian Law Journal 35.
 Art 17. It came into force on 25 December 2003. UNCTOC, the parent convention, has been in force since 29 September 2003. Destination states that have ratified TIPP include Belgium, Canada, France, Norway, Spain and Sweden. As of September 2004 Australia had signed but not ratified.
 TIPP has to be read in conjunction with UNCTOC, the framework convention, which it supplements (art 1(1)). Art 3(2) of UNCTOC defines a transnational offence as any of the following: an offence committed in more than one state; an offence committed in one state but a substantial part of its preparation, planning, direction or control occurs in another state; an offence committed in one state but involving an organised criminal group, which is active in at least two states; an offence committed in one state but having substantial effects in another. The definition of trafficking in art 3(a) of TIPP allows substantial scope to states in tackling trafficking in human beings. It includes ‘recruitment, transportation, transfer, harbouring or receipt of persons, by means of the threat or use of force or other forms of coercion, of abduction, of fraud, of deception, of the abuse of power or of a position of vulnerability or of the giving or receiving of payments or benefits to achieve the consent of a person having control over another person, for the purpose of exploitation’. There is no reference here to the transnational aspect of the process. The definition covers those involved at all stages of the trafficking process. The offence of transnational trafficking is perpetrated not only by those who actually cause the victim to cross the frontier, but also those involved in the process of getting the victim from where she first came under the control of the traffickers to the destination where she will be exploited.
 TIPP, art 3(a). See, on reasons for trafficking of children, the Report of the UN Special Rapporteur, above n 1, 12-18; Trafficking in Human Beings: Implications for the OSCE, ODIHR Background Paper 1999/3, sects 2.1, 2.2, 2.3.
 Convention Relating to the Status of Refugees (28 July 1951), 189 UNTS 150.
 Protocol Relating to the Status of Refugees (4 October 1967), 606 UNTS 267.
 See above n 17, art 7(1).
 The Refugee Convention starts from the premise that protection is most appropriately provided by the state of nationality. It is only where such protection is unavailable that other states are obliged to consider claims for asylum. Hence art 1A(2) defines as refugees those who, inter alia, are outside the country of their nationality and unable or unwilling to return there owing to a well-founded fear of persecution: in other words, they are unable or with good reason unwilling to rely on the protection of their home state. This is confirmed by the UNHCR, Handbook on Procedures and Criteria for Determining Refugee Status (1992)  (hereinafter UNHCR Handbook), that provides in part: ‘Wherever available, national protection takes precedence over international protection.’
 UNHCR, Global Consultations on International Protection, 4 September 2001, Reception of Asylum Seekers, Including Standards of Treatment, in the Context of Individual Asylum Seekers (EC/GC/01/17). See especially para g), dot point 2, of the Recommended Measures for the Reception of Asylum Seekers in Individual Asylum Systems: ‘Single women with special security needs should be provided with separate and safe accommodation.’ Also para g), dot point 8: ‘Special attention should be paid to the risk of child trafficking, in particular separated and unaccompanied female asylum seekers. Special accommodation arrangements, counselling and protection arrangements are necessary for them.’ Similar recommendations are made in UNHCR, Reception Standards for Asylum Seekers in the European Union (2000) 17-19.
 TIPP explicitly criminalises the activities of those who are involved in trafficking: arts 3(a) and 5. Furthermore Part II of TIPP, entitled ‘Protection of victims of trafficking in persons’, and the provisions therein make it clear that such persons are not to be treated as criminals but rather as victims of a serious offence who are entitled to substantial assistance.
 Such treatment could be at the hands of the traffickers, who may have contacts in the areas from which women have been trafficked and who report their return. Women in some countries are also vulnerable to harassment by the local police, especially where the police are corrupt, because the police know that the women have been working in the sex trade, perhaps in breach of the law of the country where they were trafficked but, more importantly, because they can threaten to reveal to her local community that the woman has been working in the sex trade. This may expose the women to ostracism or other forms of victimisation. See IOM, Deceived Migrants From Tajikistan (2001) 20, 22; A O’Neill Richard, International Trafficking in Women to the United States: A Contemporary Manifestation of Slavery and Organized Crime (2000) 40; Trafficking in Human Beings: Implications for the OSCE, above n 20, sect 5.2 and fn 60.
 Above n 24, 94.
 Ibid 95.
 Ibid 96.
 The Refugee Convention does not actually say that the persecution must be at the hands of the state but this has been the interpretation of some countries, including France and Germany. On this generally, see G Goodwin-Gill, The Refugee in International Law (2nd ed, 1996) 70-74; the UNHCR Handbook, above n 24 65, states: ‘Persecution is normally related to action by the authorities of a country. It may also emanate from sections of the population that do not respect the standards established by the laws of the country concerned … Where serious discriminatory or other offensive acts are committed by the local populace, they can be considered as persecution if they are knowingly tolerated by the authorities, or if the authorities refuse, or prove unable, to offer effective protection.’ Australian courts accept that persecution may be at the hands of non-state actors: see R Germov and F Motta, Refugee Law in Australia (2003) 216-20 and cases cited therein. The United Kingdom generally accepts persecution by non-state actors as sufficient: R v Secretary of State for the Home Department; Ex parte Adan and Aitsegur  UKHL 67;  1 All ER 593. For discussion of the extent of the UK’s acceptance of persecution by non-state actors, see I MacDonald and F Webber, MacDonald’s Immigration Law and Practice (5th ed, 2001) 12.51-12.57.
 Proposal for a Council Directive laying down minimum standards for the qualification and status of third country nationals and stateless persons as refugees, in accordance with the 1951 Convention Relating to the Status of Refugees and the 1967 Protocol, or as persons who otherwise need international protection, COM (12 September 2001) 510 final. See further R Piotrowicz and C van Eck, ‘Subsidiary Protection and Primary Rights’ (2004) 53 International & Comparative Law Quarterly 107.
 Arts 9(1c) and 11(2). In its explanation of art 9, the Commission argues that the inclusion of persecution by non-state actors where the state is unwilling or effectively unable to provide protection actually reflects the practice of the vast majority of member states. This is a significant move away from the (non-binding) Joint Position on the harmonised application of the refugee definition adopted by the Council of the European Union in 1996, which was much more restrictive concerning the origins of persecution: Joint Position 96/196/JHA [5.2]. G Goodwin-Gill, ‘The Individual Refugee, the 1951 Convention and the Treaty of Amsterdam’ in E Guild and C Harlow (eds), Implementing Amsterdam. Immigration and Asylum Rights in EC Law (2001) 140, 153-54.
 Council Directive 2004/83/EC of 29 April 2004 on minimum standards for the qualification and status of third country nationals or stateless persons as refugees or as persons who otherwise need international protection and the content of the protection granted (30 September 2004) OJ L 304/12.
 Above n 27.
 UNHCR Handbook, above n 24, 51.
 Ibid 53.
 For a detailed analysis see A Aleinikoff, ‘Membership in a Particular Social Group: Analysis and Proposed Conclusions’ Background Paper for Track Two of the Global Consultations of the UNHCR (2001), available at the UNHCR website: <http://www.unhcr.ch> , at Global Consultations, Documents, Second Track Meetings. Amended version: A Aleinikoff, ‘Protected Characteristics and Social Perceptions: an Analysis of the Meaning of “Membership of a Particular Social Group”‘ in E Feller, V Türk and F Nicholson (eds), Refugee Protection in International Law (2003) 263.
 Conclusions 4 and 5, available at the UNHCR website: <http://www.unhcr.ch> , at Global Consultations, Documents, Second Track Meetings. Also at Feller, Türk and Nicholson, above n 38, 312. A very similar definition has been adopted by the UNHCR in two instruments: Guidelines on International Protection: Gender-Related Persecution within the context of art 1A(2) of the 1951 Convention and/or its 1967 Protocol relating to the Status of Refugees (7 May 2002) HCR/GIP/02/01, 29 and Guidelines on International Protection: ‘Membership of a particular social group’ within the context of art 1A(2) of the 1951 Convention and/or its 1967 Protocol relating to the Status of Refugees, (7 May 2002) HCR/GIP/02/02 1. The definition adopted there is: ‘a particular social group is a group of persons who share a common characteristic other than their risk of being persecuted, or who are perceived as a group by society. The characteristic will often be one which is innate, unchangeable, or which is otherwise fundamental to identity, conscience or the exercise of one’s human rights.’
 This is the position of courts in several jurisdictions that deal with significant numbers of applications for asylum. See eg Applicant A & Anor v MIEA & Anor (1997) 190 CLR 225, 242, 263 (Australia); R v Immigration Appeal Tribunal and another; Ex parte Shah; Islam and others v Secretary of State for the Home Department  UKHL 20;  2 All ER 545, 552 (UK).
 Canada (Attorney-General) v Ward  2 SCR 689.
 Draft Directive, above n 32, commentary to art 7(4) provides, in part: ‘where the applicant for international protection is a woman, account shall be taken of the fact that persecution, within the meaning of the Geneva Convention, may be effected through sexual violence or other gender-specific means. Where the form of persecution is gender-specific this should not obscure the reason why the persecutory act occurred. For example, sexual violence can be inflicted on refugee women because, for instance, of their religion, political opinion or nationality. In these cases sexual violence is purely a form of persecution and any of the Convention grounds elaborated in … may be applicable.’
 ‘[S]exual violence to refugee women, such as female genital mutilation can also be inflicted for the one and only reason of their gender. In such situations, the persecution ground “membership of a particular social group” could apply.’ Ibid. There is certainly some willingness to show flexibility in establishing particular social groups. See the decision of the House of Lords in Shah and Islam, above n 40. Similarly, the Immigration and Refugee Board (Refugee Division) of Canada has been prepared, in a contentious decision, to treat former sex trade workers as a particular social group: Case T98-06186, 2 November 1999. Since then British courts and tribunals have recognised that some victims of trafficking could be a member of a particular social group in particular circumstances, without going so far as to assert that all victims would qualify: Secretary of State for the Home Department v Dzhygun (2000) Immigration Appeal Tribunal Decision 00TH00728, (particular social group consisting of women from Ukraine forced into prostitution against their will); Miss AB v Secretary of State for the Home Department (2003) Immigration Appellate Authority Decision CC/64057/2002 (particular social group consisting of girls trafficked from West Africa); Appellant v Secretary of State for the Home Department  UKIAT 00023 (particular social group consisting of women from the north east of Albania); Petition of Olga Shimkova, Outer House, Court of Session, 23 December 2003 (particular social group consisting of women trafficked from Ukraine).
 See above, n 19.
 Art 1A(2) 2.
 See, in particular, Jong Kim Koe v Minister for Immigration and Multicultural Affairs (1997) 143 ALR 695 and Lay Kon Tji v Minister for Immigration and Ethnic Affairs  1380 FCA. For detailed discussion of these issues, see R Piotrowicz, ‘Refugee Status and Multiple Nationality in the Indonesian Archipelago: Is there a Timor Gap?’ (1996) 8 International Journal of Refugee Law 319 and ‘Lay Kon Tji v Minister for Immigration & Ethnic Affairs: The Function and Meaning of Effective Nationality in the Assessment of Applications for Asylum’ (1999) 11 International Journal of Refugee Law 544.
 Conclusion on International Protection(s). Executive Committee Conclusions No 90 (LII) – 2001.
 Convention Relating to the Status of Stateless Persons (28 September 1954), 360 UNTS 117.
 See ExCom, Note on International Protection, 13 September 2001 68: ‘The Convention does not specifically refer to any particular groups of people who may face problems when forcibly displaced. Yet there are persons whose protection needs might require a tailored response with additional safeguards. Among them are … survivors of sexual and gender-based violence’; 70: ‘Women and girls, who are victims of trafficking and have been stranded en route, have given rise to acute protection needs …’ A/AC.96.951 96/951. The problems confronting trafficked women in this regard are acknowledged even more forcefully in ExCom’s Conclusion on International Protection of 5 October 2001, above n 47.
 The Universal Declaration of Human Rights, art 15(2), provides: ‘No one shall be arbitrarily deprived of his nationality …’. It is not necessarily unlawful to deprive a citizen of their nationality but the recognised grounds all entail some act that serves to alienate the citizen from the state: Jennings and Watts, above n 2, 878-80. This is not the case with victims of trafficking. Furthermore, deprivation is generally permissible only where the person affected will not be rendered stateless as a consequence.
 Convention on the Reduction of Statelessness (13 August 1961), 989 UNTS 175.
 The UNHCR expressly raised this in a commentary on migration control in the EU: Office of the UN High Commissioner for Refugees, Reconciling Migration Control and Refugee Protection in the European Union: A UNHCR Perspective, October 2000, 40: ‘the Convention definition of a refugee does not cover the protection needs of all persons. Those who may not necessarily come within the ambit of the Convention refugee definition as formulated in 1951 but who nevertheless need international protection are commonly referred to as refugees falling under UNHCR’s wider competence. States have long supported protection and assistance activities undertaken for those categories of refugees by UNHCR, which include, for example, persons fleeing the indiscriminate effects of armed conflict or serious public disorder, albeit with no specific element of persecution or link to one of the five grounds enumerated in the Convention’.
 Art 6(3) requires states to ‘consider implementing measures’ for the physical, psychological and social recovery of victims. This is to include appropriate housing, counselling and information in a language that they can understand (especially with regard to their legal rights); medical, psychological and material assistance; employment, educational and training opportunities. This is an extensive list. It entails potentially substantial outlay of resources by the state and clearly envisages that the victim might remain for some substantial period of time in the national territory – hence the non-obligatory language. Art 7(1) takes this further by obliging states to ‘consider’ the adoption of legislative or other measures that enable victims to remain, temporarily or permanently.
 For a discussion of some of the uncertainties and possibilities see G Noll, ‘Fixed Definitions or Framework Legislation? The Delimitation of Subsidiary Protection Ratione Personae’, paper for the seminar ‘International Protection Within One Single Asylum Procedure’, 23-24 April 2001, Norrkőping, reproduced in: Subsidiary Protection: Improving or Degrading the Right of Asylum in Europe, conference documents, Academic Network for Legal Studies on Immigration and Asylum in Europe, Brussels, 16-17 November 2001. See also Piotrowicz and van Eck, above n 32.
 Global Consultations on International Protection (4 September 2001) Complementary Forms of Protection, EC/GC/01/18.
 Above n 32.
 See the Proposal of 12 September 2001 as amended, art 15(b). 7944/04, Asile 21, 31 March 2004.
 Above n 32, 27-28.
 On the generally binding status of the principle, see Goodwin-Gill, above n 31, 167-71. On its normative character for parties to the ECHR, see Soering v United Kingdom  ECHR 14; (1989) 11 EHRR 439, 88 91; HLR v France  ECHR 23; (1997) 26 EHRR 29, 40 (dealing specifically with the threat from non-state actors).
 Rape has been recognised as torture by the European Court of Human Rights (Aydin v Turkey  ECHR 75; (1998) 25 EHRR 251) and by the International Criminal Tribunal for Rwanda (Prosecutor v Akayesu Case No ICTR-96-4-T, ICTR T ch I, 1998, 597). Not all rape is torture, but in these cases the rape was carried out by someone acting in a position of authority.
 Cases dealing with inhuman and degrading treatment or punishment invariably have involved alleged breaches of human rights by the state. However, given that persecution by non-state actors is recognised for the purposes of refugee status, it is logical to extend that to subsidiary protection. HLR v France, above n 59, supports this. As to the type of treatment that may be inhuman or degrading, see D Harris M O’Boyle and C Warbrick, Law of the European Convention on Human Rights (1995) 61-73, 80-84.