Abuse, Exploitation, and Floating Jurisdiction: Protecting Workers at Sea

IN 2013, the Bangladeshi media reported that at least 40 fishermen had been bound hand and foot and tossed into the sea to drown. Despite video evidence, no prosecutions followed. In a survey of Cambodian men and boys sold to Thai fishing boats the following year, many reported having witnessed officers killing workers. These abuses are extreme examples of a much broader pattern. Some sectors of the fishing industry remain heavily dependent on forced labour, defined in international law as labour carried out by workers who have not offered themselves voluntarily for work, and who continue to work under the threat of penalties. They also regularly expose workers to violence and mistreatment, as well as to dangerous and unsanitary working conditions. Their catches eventually find their way on to the plates of consumers in wealthy economies. The Gulf of Thailand, for instance—the epicentre of forced labour in the fishing industry—is also the biggest source of shrimp for the US market, in addition to being a major supplier of raw materials for the pet food, dietary supplement, and aquaculture industries. The abuse of workers on the margins of the global economy is, of course, depressingly widespread. Back on dry land, ‘Export Processing Zones’ often boast of their relaxed regulatory environments—meaning, in practice, that many of a country’s safety and environmental laws may be suspended, and union

provisions, with the result that a significant proportion of fishers working today are not covered by its rules. 13 A third factor which makes the fulfilment of fishers' rights much less straightforward is their distinctive legal position. Within 'territorial seas', which extend for 12 nautical miles from a country's coast, the coastal state can, if it chooses, apply to seafarers the full gamut of legal protections ordinarily offered to local workers. 14 But outside of those narrow strips of water, fishers are subject (and only subject) to the jurisdiction of the country under which their vessel is 'flagged'. So-called 'flag states' enjoy a kind of quasi-territorial jurisdiction over life on board ship: 15 each vessel is a minute extension of a state's legal territory, and it is to this flag state that workers whose rights have been violated must therefore appeal for redress. They, if anyone, must enforce the rights stipulated in international agreements, assuming they are parties to them in the first place. 16 To date, however, only 18 countries have ratified the Work in Fishing Convention. 17 This presents distinctive difficulties for seafarers, over and above those faced by other categories of migrant workers. Most migrant workers are at least present upon the fixed territory of a single state-a state in which, if they are lucky, lawyers, unions, activists, or ordinary citizens might be prepared to help defend their rights in extremis. The fisher might be on a specific state's legal territory, but it is a moving slice, and they will often be far from possible sources of assistance. Moreover, the legal ground beneath the fisher's feet is subject to sudden shifts. They could find themselves under the jurisdiction of two or more different states on the same day, without leaving their boat, since vessels will often 'flag out' to another jurisdiction in order to avoid difficulties with the law. Changing flag is both quick and relatively inexpensive for owners. For individual fishers, the consequences can be bewildering. It is not only the case that contacting lawyers or citizens who might help is made difficult by seafarers' geographical isolation. Fishers may not even know to which state they must appeal, even assuming that the authorities in the actual flag state would be interested in helping if they were able to do so.
In practice, of course, flag states are often chosen by vessel owners precisely because they will not offer effective recourse for employees with grievances. States which are prepared to rent their flags to vessel owners without insisting on even basic respect for workers' rights-so-called 'Flag of Convenience' (FOC) countries-operate as mobile exploitation havens, sustaining a race to the bottom in effective legal protection. Panama's Law Decree 8/98 of 2005, for 13 Papanicolopulu 2018, p. 39. The scope for exclusion is described in Articles 3.1 and 4.1 of the Work in Fishing Convention.
14 But note that some countries-including South Africa, for example-explicitly exempt 'persons employed on ships at sea' from basic employment protections; Couper 2005, p. 8. 15 Churchill 2005, p. 135. 16 Chen and Shan 2017, p. 2. 17 <https://www.ilo.org/dyn/norml ex/en/f?p=1000:11300 :0::NO:11300 :P11300_INSTR UMENT_ID:312333>. instance, rescinded previous protections of the right to strike, the minimum wage, and compensation for injuries; and the country is not a signatory to any treaty limiting working hours. 18 Alongside the famous case of Panama, Liberia, Honduras, and Equatorial Guinea are key FOC states operating what are called 'open registries'. These states charge a one-off fee for the registration of a ship, but thereafter agree not to levy taxes on the income they generate. Perhaps more importantly, they explicitly advertise their services on the basis that they will minimize the risk of exposure to any legal liability. 19 Seafarers themselves are not drawn from FOC countries in any significant numbers, and very few ships are owned there. In fact, vessels will typically never have any reason to visit the ports of their FOC states, making it very difficult for a seafarer with a grievance to initiate legal proceedings, even if flag states were serious about upholding their rights. 20 This distinctive legal situation places fishers in an unenviable position on the frontier of the contemporary capitalist economy, with jurisdiction-and therefore notional responsibility for the protection of their rights-potentially subject to rapid shifts, and effective legal protection often far out of reach. Many fishers are migrants, in the sense that they are living and working in a foreign legal jurisdiction. But the vulnerabilities of seafarers as a distinctive category of migrants have received little or no attention within political philosophy to date-even within the extensive literature on migration justice. This article aims to remedy that neglect.
The next section describes some of the historical and structural conditions which have increased the vulnerabilities of fishers, and gives an indication of the scale of abuse. Section II assesses who can be held morally responsible for instances of abuse: I will argue that, alongside traffickers and unscrupulous ship owners or skippers, both flag states and coastal states can bear significant moral responsibility as enablers of abuse. In the final section I propose a number of means of better protecting the rights of fishers-including rights notionally guaranteed under international law, such as the right to be paid for work, to receive an employment contract, and to receive adequate nutrition and sanitation. The reforms that I will explore range from the relatively modest through to the much more radical. Moves to increase awareness of abuse are worthwhile, as are efforts to induce flag states or coastal states to take the protection of workers more seriously. But if a major underlying cause of the exploitation and abuse of fishers is, as I have argued, the distinctive nature of legal jurisdiction at sea, then a fully adequate response is likely to involve reform to longstanding assumptions and practices concerning the power to make and enforce laws on the open ocean.

I. PATTERNS OF ABUSE AT SEA
Abuses at sea are not a new phenomenon. Seafarers have long complained of short rations, long hours, and inhospitable working conditions. But there is no doubt that the scale of these problems grew along with the emergence of industrial-scale shipping and fishing, which also deepened ethnic, religious, and gender-based subcurrents to abuses at sea. The emergence of steam ships in the 19th century, in particular, enabled both a great wave of trade globalization, and a major internationalization of the ocean-going labour force. Whereas training a crew for a sailing ship had required several years, labourers recruited to the new steamships could be turned to stoking furnaces below decks immediately. This innovation facilitated a depression of wages across the shipping and fishing industries, and the emergence of often miserable working conditions. Traditional maritime countries were able to recruit crews from Asia and pay them far less than their locally born counterparts, and to give them inferior accommodation; as a consequence, Asian crews were far more susceptible to diseases such as tuberculosis and pneumonia. 21 The internationalization of crews was given further impetus by the oil crisis of the 1970s. The owners of fishing fleets were faced not only with rising fuel costs, but also with the introduction of new fishing technologies which drove operating costs up and sent many small-scale operators out of business. Savings were made on the maintenance of boats and equipment, with sometimes tragic consequences. 22 But labour costs were the primary site for cost-cutting. The oil crisis also significantly accelerated the process of 'flagging out'. 23 By 2001, vessels registered with Flag of Convenience states accounted for 53 per cent of the gross tonnage of vessels at sea. 24 As far as we can tell, the ultimate ownership of vessels remained largely unchanged during this period. 25 At the extreme, the contemporary fishing industry has now become a major global site for forced labour. 26 In the Indian Ocean and the South China Sea, fishing and aquaculture industries rely extensively on trafficked workers. 27 But the Gulf of Thailand is probably the centre of abuse within the fishing industry. 28 There are now an estimated 200,000 migrant fishers working in the seas around Thailand, mostly recruited from Cambodia, Myanmar, Laos, Indonesia, and the Philippines. 29 A survey by the International Labor Organization found that at least one-sixth of workers on long-haul boats had been either coerced or deceived into working against their will. 30 Most of the victims are men, and fishing as a 21 Hyslop 2014, p. 216. 22 Couper et al. 2015, p. 33. 23 Borovnik 2004, p. 39. 24 Frawley 2005 89. These figures include vessels engaged in shipping as well as fishing. 25 Couper 2005, p. 26. 26 ILO 2013a. 27 Chantavanich et al. 2016. 28 Marschke and Vandergeest 2016, p. 40. 29 Couper et al. 2015, p. 125. 30 ILO 2013b whole remains a deeply male-dominated activity. But many women are exploited on the peripheries of the industry, either in fish processing or by being forced into prostitution in ports. 31 Even in cases where workers have actively sought out employment on fishing boats, their position can be characterized by deep vulnerability. Many fishers pawn family assets in order to pay fees to the agents who assign them to vessels. These assets may then be seized if the fisher breaks contract. 32 Unexpected costs are often added to the 'debts' of fishers, resulting in a form of bonded labour. 33 But while contracts may therefore offer considerable leverage to employers, they often do little or nothing to protect the interests of workers. In his study of fishing workers in Thailand, Philip Robertson reported that he 'was unable to identify a single fisher who had ever received a written contract'. 34 If workers are paid, it will typically be as a share of profits at the end of a period of 12 to 30 months at sea. 35 But what is paid may be far from what was expected. As noted earlier, the International Labor Organization's Conventions outlaw deductions from pay to cover food and accommodation, and insist that travel to and from a vessel should be paid for by employers, rather than employees. But workers are frequently unaware of their legal rights, and in practice deductions are very commonplace. Fishers who defend their rights face the threat of abandonment in foreign ports. 36 Given that fishers are often asked to surrender their identification documents upon boarding a vessel, 37 the consequence of abandonment is often arrest for illegal entry into the port state. 38 For others, the consequences of protest may be still worse.
It might be expected that workers plying the waters off the coasts of wealthy liberal democracies would achieve more consistent protection of their rights. Aside from their territorial seas, many states possess 200-nautical mile Exclusive Economic Zones (EEZs) within which they alone can determine who is able to fish. Although EEZs do not bring in tow jurisdiction over persons (other than that necessary to enforce restrictions on the taking of fish), I argue below that coastal states could, if they chose, leverage their resource sovereignty in order to better protect workers' rights. But to date they have shown very little interest in doing so. Conditions of modern slavery have been observed in the fisheries of wealthy countries such as New Zealand, 39 and an estimated two-fifths of the squid shipped from New Zealand has been caught using coerced labour. 40 The 31 Nakamura et al. 2018, p. 3. 32 Couper et al. 2015Derks 2010, p. 922. 33 Marschke and Vandergeest 2016, p. 40. 34 Robertson 2011. 35 Derks 2010, p. 918. 36 Between 1996 and 2001 alone, 3,500 seafarers contacted the International Transport Workers Federation after being abandoned in foreign ports ;Frawley 2005, p. 95. 37 Derks 2010, p. 926. 38 Chen and Shan 2017, pp. 5-6. 39 Macfarlane 2017; Stringer et al 2014. 40 Skinner 2012.
1,000 Filipino fishers who work off the coast of Scotland, meanwhile, are employed by British companies, but earn considerably less than British fishers would, and often face inferior working conditions. 41 As a result, they are three times more likely to die at sea than their British counterparts. 42 The scallop fisheries off the British coast have also been found to be the site of forced labour. 43 The contemporary fishing industry, then, demonstrates extensive vulnerability and a pattern of systematic abuse in some sectors. Although some stretches of the ocean undoubtedly offer better prospects than others for workers, even the marine territories of developed countries have witnessed abuse and exploitation. Which actor or actors, though, bear moral responsibility for sustaining these abuses?

II. ABUSE, CONTRIBUTION, AND COMPLICITY
The contemporary global economy involves complex supply chains which draw many actors into some kind of contact with wrongdoing. The most proximate culprits of wrongdoing are, to be sure, those who employ, and abuse, workers. In the case of the fishing industry, this means the owners and skippers of individual vessels. Using forced labour, making people work for long hours or in dangerous conditions under the threat of violence, and cheating workers out of wages to which they are entitled are clear-cut cases of abuse. They also count as instances of exploitation, since they involve actors taking unfair advantage of others' vulnerability in such a way as to deprive them of a fair return on their labour. In most cases owners and skippers will not plausibly be able to claim ignorance about this mistreatment. Although owners and skippers are therefore the principal wrongdoers-because they are directly causally responsible for exploitation and abuse-people traffickers who supply fishing vessels with unwilling workers fall into the same category. They, too, knowingly violate the rights of vulnerable workers in order to gain material rewards. The same can potentially be said of agents as well, even if they are not involved in people trafficking per se. Unlike traffickers, agents find openings for people who have volunteered to work. But if agents supply workers in the full knowledge that they will be abused, or if they mislead workers about what they can expect, this is enough for them to be properly considered co-principals in abuse, even if it is others who will go on to commit these abuses. 44 All of these actors, then, can be considered to be 'outcome' responsible for abuses, in the sense that they make clear causal contributions to sustaining those abuses. 45 41 Howard 2012, p. 337. 42 Couper et al. 2015, p. 37. 43 Carrell 2018, p. 1. 44 On the idea of being a co-principal in wrongdoing, see Goodin 2019. 45 For the distinction between outcome responsibility (which identifies agents who have contributed to a wrong) and remedial responsibility (which identifies agents who have a duty to put that wrong right), see Miller 2007, ch. 4. This much is perhaps straightforward. But a number of other actors can potentially be considered complicit in wrongdoing within the fishing industry. Drawing attention to their complicity is not intended, to be clear, to detract from the outcome responsibility borne by those who directly abuse or exploit fishers. But it is important to have a full moral picture, not least since this picture will help us understand better who ought to bear the burdens of eradicating injustice within the fishing industry. I will return to that question in the next section, but the answer might plausibly be that quite a diverse set of actors possess some remedial or reparative responsibilities. 46 Accounts of complicity differ in their views about what kind of action or intention qualifies one as complicit in an outcome. On some accounts of complicity, actors who knowingly participate in shared practices which eventuate in abuse can be considered complicit in that abuse, even if their participation is not individually causally necessary for abuse to occur. What matters, on such views, is a shared participatory intention. 47 But while that kind of shared intention may be reprehensible, it is not obvious why it qualifies as complicity in cases where one's actions are not causally necessary to the outcome in question. On the other hand, it does seem appropriate to charge people with complicity in an outcome, regardless of whether they share a participatory intention to create that outcome, so long as they are aware that their actions are causally necessary contributions to its emergence, and they voluntarily make those contributions anyway. That is the conception of complicity that I will adopt in this section. 48 The fishing industry provides many plausible examples of complicity construed in those terms. One example would be corporations further down the supply chain. Firms which buy the fish sold by fishing outfits-to sell it on to consumers or convert it into pet food or fodder for fish farms-can sometimes be considered complicit in wrongdoing even if they do not commit it themselves. For that judgement to be sound, it need not be the case that these corporations intend for abuses to happen. It is sufficient for the judgement of complicity to carry through that they are aware of abuses, that they are likewise aware of their own role in sustaining them, and that they could act in such a way as to avoid sustaining them. 49 We might consider consumers too to be complicit in a similar way in abuses. Individual consumers' outcome responsibility for wrongdoing is typically 46 In this section, I focus on actors who possess outcome responsibility because of their role in sustaining or enabling abuse. They may well bear remedial responsibility as a result (I discuss remedial responsibilities in the following section). But a full moral picture will also acknowledge that actors can possess remedial responsibilities on account of their ability to make a difference. For an excellent discussion of how responsibilities can be grounded on contribution and capacity, see Barry and Øverland 2016. 47 Kutz 2000, p. 81. 48 Lepora and Goodin 2013, pp. 80-3. 49 Goodin 2019. Some states require large corporations to take active steps to ensure that abuses do not occur further down the supply chain. The French 'Duty of Vigilance Law' is a leading example; Loi relative au devoir de vigilance des sociétés mères et des entreprises donneuses d'ordre, 2017. minimal, in the sense that abuse would still go on if any particular consumer stopped eating fish products, 50 and in that sense it makes little sense to accuse them of complicity-at least on the definition I am assuming here. This likely marks out a difference from corporations, not least since the global fish business is dominated by just a handful of corporations which exercise a powerful influence on the fishing economy, 51 and could presumably exert considerable pressure for reform if they chose to. But individual consumers nevertheless partially constitute the market for fish sourced in unethical ways. Even if their causal contribution to injustice is negligible, it might still be argued that they should take actions to avoid being a constitutive part of a bad practice. 52 This judgement, though, must be sensitive to the availability of other options. There are many people in the developing world who possess very few sources of protein and essential micronutrients besides fish, and who can therefore, regrettably, scarcely afford to exit the market for cheap fish. If ceasing to eat fish sourced in unethical ways is very costly, this does not mean that I am no longer playing a role in a bad practice, but it may well mean that my doing so is not blameworthy. 53 People in the developed world, by contrast, typically have plenty of alternative, and non-costly, sources of protein and micronutrients. For them, avoiding unethically sourced fish is likely morally required, and the failure to do so potentially blameworthy.
Consumers considered collectively, by contrast, are a cause of the mistreatment of workers just like corporations are, in the sense that their willingness to buy goods sourced in unethical ways encourages the further production of those goods. 54 But consumers considered as a collective are probably not the right kind of entity to bear outcome responsibility, or to attract the charge of complicity: consumers do not come together to coordinate their actions, and presumably do not either intend or foresee the outcomes which result from their purchases. 55 In that sense, consumers are rather different from large corporations with fixed decision-making structures and audit trails. Things look different from the perspective of the states in which consumers live, however. Individual states will plausibly bear outcome responsibility if they fail to prevent their citizens (avoidably) consuming fish, wherever the predictable outcome is that exploitation and abuse are perpetuated. If they refuse to do so, they can be considered complicit in those abuses. In the next section I will outline some ways in which consumers' states might promote greater adherence to human rights standards in the market for fish. 50 The same may not be true in some sectors of the economy, in which one purchase can be enough to trigger an order from suppliers for more products. But that does not appear to be the case in the fishing industry. 51  Thus far I have mainly dealt with highly familiar categories of employers, agents, corporations, and consumers. But I want to argue in the rest of this section that in the fishing case, there are two further and less familiar sets of actors which can bear significant outcome responsibility for abuses in the fishing industry. First, I will argue that those who determine policy in FOC states are crucial enablers of wrongdoing, and complicit in it. Second, I will argue that those who determine policy in coastal states can be considered to enable, and to be complicit in, wrongdoing. I will not seek to resolve here the empirical question of who actually determines state policy in these cases. My claim is simply that those who determine state policy, whoever they are, can bear a significant degree of moral responsibility for labour injustices at sea.
First, then, policy-makers in flag of convenience states are crucial enablers of injustice. By selling their flags without making good-faith efforts to ensure that minimal labour standards are adhered to, flag states as a whole are complicit in any abuses that result. The existence of FOC states is a necessary causal contribution to wrongdoing by others, and the governments of FOC states are well aware of the role that they play. Without the existence of flags of convenience, it would be impossible for owners and skippers to get away with abuse (at least on anything like its present scale). Since FOC countries by definition make no significant effort to ensure compliance with even minimal labour standards, these states operate as mobile exploitation havens, facilitating a race to the bottom in labour standards. FOC states make it much easier for vessel owners to evade legal scrutiny, by putting in place policies that thwart efforts to trace the beneficial ownership of vessels. They allow each vessel on their books to be registered under a different company name, for instance, making it more difficult for victims of injustice, or those who would defend them, to establish ownership, and eradicating legal cross-liability. 56 It might be, however, that an individual FOC country could plausibly claim not to be complicit in labour abuse. I have been assuming thus far that complicity involves making a necessary causal contribution to a bad outcome. 57 If FOC countries collectively reformed their practices, insisting on effective protection of workers' rights before a vessel could be registered, it would no longer be possible for abusive owners and skippers to get away with that abuse. It is very doubtful, however, that the same outcome would follow if any individual FOC country reformed its practices: unscrupulous owners would presumably swap flag to other FOC states instead. In response to this objection, we have two options.
The first is to accept that this would enable a state to escape from the charge of complicity, and note that another moral charge may still be pressed. It does appear that the individual FOC state in these circumstances knowingly continues to act as a constitutive part of an appalling practice. The likelihood that other 56 Christy 2019, pp. 51-5. 57 Lepora and Goodin 2013, ch. 6. unscrupulous registries would take up the slack does not excuse it from blame for acting in this way, provided it has alternatives. Even if preventing abuse is not a feasible option, ending one's own role as a facilitator of abuse may be. This may not be true for very poor states, for which registration fees are a significant portion of national income. In those cases, their role may be less blameworthy. But when alternatives are available, flag states must make good-faith efforts to ensure that they are not providing cover for abuse and exploitation. To refuse to do so is to wrongly act as a constitutive part of those practices.
The second, and in my view more satisfying, response is to untie the charge of complicity from the requirement of a uniquely necessary causal contribution. Imagine the following case. A factory owner declares that his employees are not working hard enough, and that he is going to buy a gun to force them to do soeven if they drop dead in the process. There are 20 people who could sell him a gun. Fifteen of them will refuse to do so. Five of them-including me-will agree, even if they fully understand the consequences. Assume now that the factory owner randomly knocks on my door and I sell him the gun with which he goes on to abuse his workers. Here it seems to me that the fact that my causal contribution is not uniquely necessary-because eventually one of the other four unscrupulous people would have sold him a gun instead-does not absolve me from the charge of complicity. I have facilitated his wrongdoing by supplying a means necessary to his ends. The fact that I merely belong to a set of agents who would have acted similarly does not establish, it seems to me, that I am not complicit in his wrongdoing. Just as the presence of other abusive employers in a marketplace does not excuse any given employer from blame for abusing workers, 58 the presence of other potential facilitators of abuse does not excuse any particular facilitator from the charge of complicity when they knowingly and avoidably agree to be the one who makes the necessary causal contribution. If so, it does appear that individual FOC states will find it hard to escape the charge of complicity in abuse within the fishing industry.
Coastal states too can bear significant moral responsibility. There are two ways in which coastal states can be complicit in, and therefore share moral responsibility for, exploitation and abuse. The first is in their capacity as port states, where fish catches are unloaded and sold. Port states can, in principle, play a vital role in defending workers' rights, by boarding ships and assessing the conditions in which fishers live and work. Coastal states possess the legal right to board vessels which have docked in their ports. While in port, anyone guilty of abuse can be punished under domestic law, regardless of their nationality or the flag state of their vessel. Those prerogatives are not always exercised, however. Unfortunately, so-called 'port of convenience' countries compete according to the laxness of their inspections. Their desire to attract more trade outweighs any interest in protecting the rights of workers. 59 If port of convenience states turn a 58 Ibid., pp. 41-2. 59 Campling et al. 2012, p. 183. blind eye to abuse, they can be considered to be complicit in injustice, in the sense that they play a necessary causal role in enabling wrongdoing to take place. If vessels engaged in abuse could not unload, their incentive to engage in abuse would be radically curtailed. Again, for the judgement of complicity to be valid, it is not necessary for the authorities in port states to intend the wrongdoing in question. It is sufficient that they know that it is likely to take place, and understand that they are facilitating it. Assuming they possess alternatives, their complicity will be blameworthy.
Second, and probably more significantly, coastal states which allow fishing vessels to ply their Exclusive Economic Zones can, I argue, share in moral responsibility for exploitation and abuse on those vessels. In a sense, the seafarer working within a country's EEZ is like the ultimate circular migrant: contributing to the economy by catching the fish which recoup the costs of fishing access, but not enjoying any form of membership, or indeed any of the labour or safety or environmental protections which even a visiting worker would normally expect. If coastal states are aware of the abuse of workers in their EEZs, but fail to take available steps to prevent it, they are complicit in that abuse.
This claim, though, faces an immediate objection. As I noted earlier, coastal states enjoy the right to exercise full jurisdiction over persons within their ports and territorial seas, whether they choose to exercise it or not. 60 But the same does not apply to EEZs, in which more than 90 per cent of the world's fish are caught. 61 Here, coastal states enjoy a form of partial or quasi-jurisdiction which extends only as far as resource rights; actions taken against persons within an EEZ will therefore be actions taken to enforce rules around fishing, rather than to enforce labour rights as such. In general, vessels can only be stopped at sea by law enforcement or military vessels flying the same flag. (One exception is that warships are entitled, under the Convention on the Law of the Sea, to board ships which are suspected of being engaged in the slave trade, 62 and other specified crimes, although this power is rarely exercised). 63 Either way, the coastal state's powers certainly do not extend to the routine protection of labour rights. 64 How, then, can they be considered morally responsible for abuse?
Coastal states are, in fact, far from powerless when it comes to preventing exploitation and abuse within their EEZs. Legally, states can determine who can fish in their EEZs and who cannot. If a vessel is fishing legally in Canada's EEZ, it is because Canada has allowed it to do so. States can therefore leverage their resource sovereignty in order to require fishing vessels to extend basic protections 60 Churchill 2005, p. 135. 61 Zacharias and Ardron 2020, p. 205. 62 UN Convention on the Law of the Sea, 1982, Art. 110(2). 63 Garcia-Llave et al. 2015. 64 The 1995 UN Fish Stocks Agreement does also give members of regional fisheries organizations the right to board vessels in the relevant areas. But while inspectors can board and gather information, they cannot make arrests or institute judicial proceedings. Flag states alone must follow up any reported violations. Harrison 2017, p. 192. to foreign fishers working within their fishing zones, and they can refuse to sell access to vessels sailing under flags of convenience. There have, to be sure, been very few cases in which coastal states have shown any willingness to make this move. New Zealand is an honourable exception. In 2014, a coalition of NGOs, lawyers, and organized labour brought about a significant change in government policy. Whereas New Zealand's fisheries were previously the site of forced labour, fishing permits are now only issued to ships which agree to implement the very same protections which fishers would enjoy in New Zealand's territorial seas. 65 This represents real progress. 66 It shows that effective coastal-state control over labour practices is possible within EEZs, and this fact serves as a condemnation of other states which possess, but refuse to deploy that potential control. Things may be different, to be sure, in cases where coastal states simply lack the practical (as opposed to legal) capacity to determine who fishes in their waters and who does not. Off the coast of West Africa, for instance, an estimated 40 per cent of fish are caught illegally. 67 If a state lacks the capacity to determine who fishes in its EEZ and who does not, then it does not make a causal contribution to, and hence is not complicit in, exploitation and abuse. But although some developing countries suffer from rampant illegal fishing which they undoubtedly lack the capacity to control, this is not typically the case in the developed world. When the capacity to monitor abuse exists, it would be wrong for coastal states to wash their hands of the problems faced by fishers in their EEZs. If they refuse to exercise this capacity, they are complicit in exploitation.

III. PROTECTING WORKERS
I argued in the previous section that global fishing supply chains draw many actors into contact with abuse and exploitation, and that quite a diverse set of actors can bear moral responsibility grounded on their contribution to or complicity in these widespread problems. Progress in easing abuse and exploitation may well involve all of these actors changing their practices. In addition, actors can bear remedial responsibilities on the basis of their capacity to make a difference: if innocent others can ease patterns of exploitation and abuse without making undue sacrifices, it is often the case that they ought to do so. 68 In this section I will assess four sets of reforms which offer the promise of reducing fishers' vulnerability to abuse and exploitation. They range from the relatively modest to the more radical and transformative. 65 Couper et al. 2015, p. 167. 66 In fact, New Zealand has gone further than this, and now requires vessels fishing in its EEZ to sail under a New Zealand flag. This is one way of ensuring control over onboard labour practices, and perhaps a particularly effective one. But this further step is not strictly necessary: the crucial step is to refuse to sell fishing rights to vessels which do not agree to be bound by core labour standards. 67 Agnew et al. 2009. 68 Much will depend, of course, on how we define 'undue' sacrifices. For an argument that we are required to take on 'moderate' costs in order to remedy severe suffering, see Barry and Øverland 2016. A. Transparency A variety of transparency-oriented reforms could help to diminish patterns of abuse and exploitation within the fishing industry. First, establishing a global registry of licensed fishing vessels, including information on beneficial or ultimate ownership, would improve the ability of seafarers to pursue legal remedy against abusive employers-though they would still require the right to sue under national law, as well as access to legal counsel. 69 It would assist port states in protecting workers' rights, moreover, by clarifying who is responsible for their care, and making their abandonment by unscrupulous employers less likely.
A second category of reforms would provide greater information to consumers and the general public about labour practices within the fishing industry. Developed-country consumers are often unaware of the conditions under which the fish they eat have been sourced. This forms a stark contrast to public awareness about the sustainability of fish stocks: ordinary citizens are increasingly conscious of the depletion of some fish species, and many of them therefore choose to buy fish with 'sustainable' certification. Until now certification schemes have not helped in the case at hand, because they have concentrated on stock sustainability rather than labour rights. The leading vehicle for certification within the fishing industry is the Marine Stewardship Council (MSC). It has only quite recently taken its first steps towards connecting certification to respect for labour rights, by refusing certification to companies which have been prosecuted for using forced labour in the previous two years. This sets a fairly low benchmark for accreditation, but it is a positive step. So too are recent moves to extend MSC accreditation to pet food produced with fish as a raw material. Nevertheless, respect for labour rights is not yet one of the core standards which the MSC employs in order to define a well-managed fishery. 70 That must change, if consumers are to be empowered to make informed decisions.
In some cases, fish are used as raw materials in the manufacture of complex consumer products, and here it is not obvious that explicit consumer-oriented certification is going to be the most effective policy. If so, it makes more sense to place the onus on the corporations which process fish products to ensure that they are not using supplies harvested by way of exploitative or abusive labour conditions. The United Nations' Guiding Principles on Business and Human Rights make clear that corporations have a responsibility to 'Avoid causing or contributing to adverse human rights impacts through their own activities', and to 'Seek to prevent or mitigate adverse human rights impacts that are directly linked to their operations, products or services by their business relationships, even if they have not contributed to those impacts'. 71 69 Anderson and Mensah 2005, p. 548. 70 <https://www.msc.org/en-us/stand ards-and-certi ficat ion/fishe ries-standard>. 71 UN Guiding Principles on Business and Human Rights, New York, 2011, Pillar II, Principle 13.
A third information-related reform, which would oblige corporations to act on this responsibility, would require corporations to ensure that their supply chains are free from abuse, and to report their findings publicly. At present, few companies systematically track the origins of the fish they use in their products. But they do often possess the technical capacity to do so. 72 The challenge is to oblige them to use it. In 2012, California took an important step by passing the Business Transparency Law, which imposed reporting requirements on large corporations, requiring them to document exactly how they can be sure that their imports are trafficking-free. 73 The only national-level reporting law to date applies in the UK, where the Modern Slavery Act of 2015 obliges all large companies to report on the risk of slavery and forced labour within their supply chains. This example should be reproduced by all countries which possess the capacity to do so. For those which do not, support for building such capacity should be a priority of the international community.

B. Leveraging Coastal-State Resource Sovereignty
I argued in the last section that coastal states can potentially leverage their resource sovereignty in order to promote labour justice. The possession of EEZs gives states jurisdiction over natural resources, but does not explicitly grant them direct jurisdiction over persons (other than what it takes to enforce resource rights). It is not obvious, therefore, that coastal states can directly promote workers' rights within these marine territories. 74 As a result, many states have washed their hands of the problem of abuse at sea, even if it occurs within their own EEZs. But if coastal states have the capacity to leverage their resource sovereignty in such a way as to prevent abuses, this is morally indefensible. Assuming that this does not prevent them from meeting their own citizens' basic rights, states should only sell access to fish stocks to companies willing to guarantee the protection of workers' rights, and to agree to periodic checks on their working conditions. New Zealand has shown that such an approach is possible, and that it can reap dividends even within the context of a regime of exclusive flag-state jurisdiction.
Indeed, many states appear to have a legal duty to prevent some forms of abuse. Article 4 of the European Convention on Human Rights, for instance, clearly prohibits slavery and forced or compulsory labour, 75 and a case could therefore be made that signatory states have a duty not to permit those practices 72 Nakamura et al. 2018, pp. 3, 7. 73 Bang 2014 It is possible, however, that a coastal state might invoke Article 62(4) of UNCLOS, which states that foreign fishers should comply not only with the coastal state's conservation measures, but also with 'other terms and conditions established in the laws and regulations of the coastal State'. The examples given in 62(4) do not include labour rights, but they are part of an 'inter alia' list of laws and regulations. If this does allow coastal states to directly impose labour standards, there may be no need to turn to the argument about leveraging resource sovereignty. 75 Convention for the Protection of Human Rights and Fundamental Freedoms, Rome, 1950, Article 4, 'Prohibition of Slavery and Forced Labour'. within their EEZs. More widely, Pillar I of the UN Guiding Principles on Business and Human Rights declares that states will be in breach of their international human rights law obligations 'where they fail to take appropriate steps to prevent, investigate, punish and redress private actors' abuse' within their jurisdiction. 76 Concerned citizens should lobby their representatives to act on these principlesfor instance by generalizing some variant of New Zealand's policy-so as to make access to fish stocks conditional on observing minimal labour standards. This would represent a considerable boost to working conditions, especially in developed countries which possess the requisite monitoring capacity. When this option is open, taking it up is required in order to avoid moral complicity with exploitation and abuse.
Such a reform would not require the creation of new international institutions or laws. It would simply require states to deploy their existing resource sovereignty in a more ethical way. If coastal states were to take advantage of the resource sovereignty granted to them under international law, the payoff for the human rights of workers at sea could be enormous. Since the vast majority of the world's fish catch occurs within the EEZ of one state or another, if owners were excluded from even a significant proportion of these zones unless they agreed to improve labour standards (and to accede to monitoring of their compliance with international agreements), we could expect significant improvements to occur.
While coastal states leveraging their resource sovereignty would shift industry incentives from the 'supply' end, rendering access to stocks conditional on the observance of labour rights, port states could also exert a powerful role at the 'demand' end of the chain. If fishers could not unload fish from ships crewed by forced or otherwise abused workers, this would also significantly reduce the incentive to engage in exploitation and abuse. For this approach to make a substantial impact, however, port states would need to show a consistent determination to board and inspect fishing vessels. To date, the most significant innovation has been the Port State Measures Agreement, which entered into force in 2016. Signatory states agree to robustly monitor and inspect fishing vessels in their ports in order to drive out illegal fishing. While the focus of this Agreement is not on labour rights per se, a more robust and consistent approach to inspecting vessels is expected to also pay dividends for workers on those vessels who may have suffered from abuse. To date, only a third of states have become parties to the Agreement, 77 however, and encouraging wider uptake should be a priority.

C. The 'Genuine Link'
In many ways, the practice of flag-state jurisdiction lies at the heart of the difficulties faced by fishers. Against a backdrop of frequent non-enforcement of even the most basic international norms, how might flag states be encouraged to 76  show a more consistent concern for workers sailing under their flags? It has been argued that changing the way in which vessels come to sail under a country's flag in the first place is a promising place to start. Many international legal instruments, including the Law of the Sea Convention, already declare that there must be a 'genuine link' between a state and any vessel sailing under its flag. Unfortunately, the Convention never defines what would constitute such a 'genuine' link, and there has been no legal consensus on this question to date. 78 The 1986 UN Convention on Conditions for Registration of Ships declared that a meaningful genuine link would exist if, and only if, a significant equity stake was owned within the flag state, if the identify of beneficial owners was readily discernible, and if the flag state provided a significant portion of a vessel's crew. 79 Reforming the notion of the genuine link in such a way would make it easier for victims of abuse to achieve effective legal recourse against their employers. It might also become harder politically for flag states to tolerate the mistreatment of workers in the industry if many of those workers were in fact their own citizens. Unfortunately, the 1986 Convention has not attracted sufficient ratifications to come into legal force. 80 Until it does, the only 'link' which is legally required between a vessel and a flag state consists in the very act of registering a ship. 81 That situation continues to smooth the path for those who would abuse and exploit workers.

D. Ending Exclusive Flag-State Jurisdiction
The reforms discussed so far could make a significant positive difference to the prospects of people working in the global fishing industry. But they would not suffice to eradicate bad practices. Take the proposal that coastal states should exercise the leverage they possess over foreign fishing fleets plying their EEZs in order to demand compliance with basic labour standards. One reason this could only represent a partial solution to the problems of exploitation and abuse is that some fishing takes place on the high seas, where there is no coastal-state resource sovereignty capable of being leveraged. On the high seas, fishers are wholly dependent upon the scruples of their employers, and the flag states with which their vessels are registered. Furthermore, even within EEZs, many poor coastal 78 Churchill 2005, p. 133. 79 UN Convention on Conditions for Registration of Ships, Adopted by the United Nations Conference on Conditions for Registration of Ships, 7 Feb. 1986. 80 It has also been suggested that the Convention's recommendations are drafted in such an openended way that they would be unlikely to significantly constrain states' rights to confer their flags even were they to come into force; Barnes 2015, p. 307. 81 Frawley 2005, p. 87. In an important case before the International Tribunal on the Law of the Sea, Guinea argued that it ought to be authorized to board and exert jurisdiction over a vessel engaged in illegal fishing, because there was no 'genuine link' between the vessel and its flag state. The Tribunal, however, ruled that there was no justification for refusing to accept the authority of a vessel's existing flag, even if another state possesses evidence that there is no genuine link between flag state and vessel. Thus it reaffirmed the predominant role of flag states even in cases of weak linkage and apparently illegal behaviour. Harrison 2017, p. 174. states appear to lack the practical capacity to monitor and enforce compliance with basic labour rights.
I have also suggested that greater sharing of information can help to build political, legal, and consumer pressure on owners to curtail their participation in instances of exploitation and abuse, and that beefing up the 'genuine link' clause would make it easier to hold FOC countries to account for their inaction in protecting labour rights at sea. But these reforms, though important, would also be limited in their impact. The ocean economy possesses many grey zones, and practices such as 'trans-shipment'-in which fish caught illegally, using slave labour, or under flags associated with malpractice, are transferred, before landing, to boats otherwise operating legally-represent a major problem for anyone concerned to drive out exploitation and abuse. The movement of fish across vessels seriously clouds the legal picture and makes it much harder to drive abusive owners out of business.
Against a background of massive non-compliance, weak enforcement capacity on the part of many coastal states, and pervasive informational problems, it is difficult to escape the conclusion that fully protecting the most basic rights of workers will involve direct action by a wider range of non-flag-state authorities which encounter vessels on which abuses take place. Only once owners and skippers know that third parties are realistically likely to board vessels, to inspect working conditions, and to punish wrongdoers, will such wrongdoers encounter a genuine deterrent.
Such proposals set a clear collision course with the practice of exclusive flag-state jurisdiction (EFSJ). But such a collision is unavoidable if the rights of some of the most vulnerable are to be protected. The practice of EFSJ may once have made moral sense. For the bulk of maritime history, switching flag was a relatively rare event. The nationality of seafarers, owners, and flag states typically overlapped, meaning that crew could, in theory, form a stable expectation that the country in which they, and their vessels, possessed nationality would protect their interests in extremis. The globalization of the fishing industry has changed that background fundamentally. Most seafarers-including fishers-are now sailing on vessels registered in countries other than their own, which are owned in yet other countries. In this context, EFSJ has acted as a lubricant to a race to the bottom in labour standards, and has seriously attenuated the protection afforded to workers. If flag states were reliable enforcers of labour rights, EFSJ might not represent such an obstacle to justice. But in a situation where flag states explicitly compete according to their laxness in enforcing core standards, it serves to shelter unscrupulous employers from legal redress.
There are two chief ways in which EFSJ might be transcended as a practice. The ideal solution would be to establish an international body legally entitled to police the observance of labour rights at sea. It would exercise a form of 'universal jurisdiction', insofar as it could prosecute abuses anywhere in international waters, regardless of the nationality of a vessel. The principle of universal jurisdiction already applies in the case of piracy, which can be punished by vessels belonging to other flag states. 82 But this proposal would extend it to wider cases of exploitation and abuse. The relevant international body would be globally funded, according to a formula factoring in both capacity and contribution to the problem of abuse and exploitation at sea. Its agents would possess the legal right to board vessels suspected of infringing the rights of workers. Skippers found guilty of abuse would be liable to prosecution under the auspices of an appropriate international court, which would need to be empowered to pursue claims for damages against owners, and, in serious cases, to seek their extradition. Such a body could operate both on the high seas and within EEZs, where the majority of fishing takes place. Its activities would not infringe upon the sovereignty of coastal states, because outside their 12-mile territorial seas they possess no general jurisdiction over persons in the first place.
A second-best option would be to license concerned third parties-which might include, but need not be limited to, states in whose EEZs a vessel was operating-to enforce minimal labour standards. Such a proposal would no doubt from time to time attract charges of vigilantism, and could spur tit-for-tat action. Cases where the justification for boarding a vessel was legally uncertain could, however, be referred to an international tribunal for adjudication.
On either option, we would see a substantial development of existing constraints upon the practice of EFSJ. The 1926 Slavery Convention declares that states have a duty to 'prevent and suppress the slave trade', within all territory under its 'sovereignty [or] jurisdiction'. 83 But on the high seas, the ability of states to prevent slavery on vessels sailing under other nations' flags is distinctly limited. The Law of the Sea Convention declares that military vessels operating anywhere in the world possess the legal right to board any vessel suspected of engagement in slavery, regardless of the flag it is sailing under. But it does not grant them the right to seize those vessels or to liberate those suspected of being held in slavery. 84 Suspicions concerning 'forced' or 'compulsory' labour, by contrast, generate no right of boarding at sea, and remain a matter for the flag state alone. The 1956 Supplementary Convention on Slavery, likewise, declares that states have a duty to 'progressively' abolish debt bondage and other slavery-like practices, but it does not extend the right to board vessels to those cases. 85 Here we can discern two problems. The first is one of normative scope: though grievous in its own right, it is not clear why slavery should be seen as unique in 82 See Song (2015) for a valuable argument in favour of extending rights of universal jurisdiction beyond cases of piracy. Song sees universal jurisdiction as an important means of enforcing human rights principles in cases where the responsible countries cannot enforce them themselves. I want to claim that we must (also) turn to universal jurisdiction in cases where the flag state is not willing to enforce human rights principles. 83 Slavery Convention, Geneva, 1926, Art. 2. 84 UN Convention on the Law of the Sea, 1982, Art. 110(2). The scope of the right is restricted to boarding a vessel and inspecting documents to verify a vessel's right to fly its flag. 85 Supplementary Convention on the Abolition of Slavery, the Slave Trade, and Institutions and Practices Similar to Slavery, Geneva, 1956, Art. 1. its power to justify constraints on the purported freedom of vessels to escape third-party scrutiny and enforcement of international legal norms. Although slavery is a gross affront to individual autonomy, it is part of a broader continuum of practices (including bonded labour, the refusal to issue written contracts, and the use of physical threats, the confiscation of passports, and/or the threat of abandonment to ensure compliance with unreasonable demands) which infringe upon the rights of many vulnerable migrant workers. Indeed, many legal instruments declare that forced labour and other slavery-like practices are equally unlawful. 86 That recognition ought to be reflected in an expanded remit for boarding foreign vessels at sea.
The second problem concerns the content of the rights granted to those seeking to interdict vessels engaged in slavery or other forms of abuse. The two Conventions discussed above grant foreign vessels the right to board and inspect any vessel suspected of engagement in the slave trade, and to report infractions to the flag state. But they do not grant any power to make arrests, or even to free those found to be held in conditions of slavery. 87 States can already tow vessels into port if they are suspected of illegal fishing; this right should be extended to cases where vessels appear to be engaged in slavery and other grievous abuses. 88 The granting to third parties of the power to arrest culprits and liberate victims also ought to be a priority, as should their extension to cases of bonded labour and debt servitude, as well as classic forms of slavery.
There is no doubt that such proposals will be controversial. The unfinished human rights revolution has increasingly seen sovereignty-and rights against foreign intervention-qualified in light of at least basic commitments to human rights. At the same time, states have often fiercely resisted those qualifications by pointing to their purported interest in sustaining political and territorial independence. Whatever the weight of such objections in archetypical cases of foreign intervention, it is hard to see how they resonate anything like as strongly in the case of fishing vessels, notwithstanding the fact that they are indeed floating slices of a state's sovereign territory. The claim to exclusive jurisdiction over vessels at sea is ripe for revolution in light of the basic interest of workers in freedom from coercion, exploitation, and physical abuse.

IV. CONCLUSIONS
The evolving Law of the Sea has displayed a consistent blind spot on the topic of the human rights of those who live and work at sea, and not least those workers who decide-or are pressed-to join one of the most dangerous of all professions. That neglect has been mirrored within political philosophy, which has shown little interest to date in the distinctive position of one of the most vulnerable 86 Both slavery and forced labour are explicitly forbidden, for instance, by Article 8 of the 1966 Covenant on Civil and Political Rights. 87  categories of migrant workers. I have begun to remedy that neglect, by detailing the abuses which fishers are often exposed to. I have also attempted to identify the various actors who bear moral responsibility for exploitation and abuse within the fishing industry. One of the most striking conclusions has been that actors in coastal states are complicit-in ways which ordinary citizens, as well as political philosophers, may never have had occasion to reflect upon-in the global problem of the mistreatment of fishers. Finally, I have surveyed a number of proposals for reform which promise to improve the prospects for workers at sea. While a number of them are accessible without any legal or institutional reform, I have argued that consistent protection of fishers' rights will ultimately require us to rethink the nature of legal authority at sea. While many nation-states may be reluctant to abandon the practice of exclusive flag-state jurisdiction, in the situation of massive non-compliance in which we find ourselves it represents a serious and unjustified obstacle to the protection of the basic rights of many of the most vulnerable.