1 Introduction

States increasingly rely on extraterritorial border controls to prevent migrants from entering their territories. The purpose of these policies is to apply border controls “before borders are reached” (Moreno-Lax 2017). As migrants are unable to cross the national border, they cannot apply for asylum and do not enjoy protection from “refoulement” – the core rights of the international refugee regime. States are thus acting contrary to obligations of respect for human rights by extending border controls to locations beyond the national territory.Footnote 1 This is no accident, of course, as the very purpose of external border controls is for states to “skirt their obligations without formally withdrawing them” (Shachar 2020; p. 12).

Border polices that deny migrants’ access to the state are controversial not just in terms of legal obligations, however. Scholars also dispute the legitimacy of current border practices. Legitimacy determines the state’s right to rule: the state only has the right to enact policies and laws that are legitimate. Hence, the charge that a border policy of the state is illegitimate means to suggest that the state has no right to pursue it.

The requirements that determine the legitimacy of border controls are of course widely debated. Some argue that respect for human rights is necessary but also sufficient for legitimate border policy (Miller 2013). Others insist that state legitimacy is more demanding. Migrants ruled by the border policies of the state are subject to legitimate rule only if they either are not subject to domination (Sager 2017) or are afforded a democratic say in designing the policies that potentially exclude them (Niño Arnaiz 2024; Abizadeh 2008).

Yet, normative questions about what makes border policies legitimate are secondary to what it means for the state to rule (Tasioulas 2010; p. 98). Legitimacy standards apply only to policies and laws that rule people. Hence, the claim that border controls are not legitimate depend on reasons to believe that migrants are in fact ruled by them.

In so far as the state is ruling the population present on the territory it controls, conceptual issues about the meaning of “rule” are seldom relevant. Whatever rule is taken to mean, the resident population is invariably ruled by the state and standards of legitimate rule clearly applicable. Sometimes, this is taken to be true also for the state’s attempt to rule migrants by border controls. Following Miller (2021) migrants are expected to – and should – comply with laws of immigration that are legitimate.

But it is not obvious that laws and policies governing state borders do rule actual and potential migrants excluded by them. Whether migrants are ruled by extraterritorial border controls depends on what is meant by “ruling” in theories of state legitimacy. If a policy or law is either legitimate or illegitimate only with respect to people ruled by it, extra-territorial border controls can be illegitimate only if they rule migrants in normatively relevant ways.Footnote 2

The emerging literature on the legitimacy of extraterritorial border controls proposes two distinct conceptions of “rule”. The first view is that migrants are ruled by extraterritorial border policies only if they subject migrants to coercion. States that engage in “remote control” are acting illegitimately when migrants are subject to morally impermissible coercion (Bloom and Risse 2014) or force-wielding institutions that fail to robustly respect their human rights (Schmid 2022). The basic idea is that the state rules migrants only through coercion.

The second view is that the legitimacy of extraterritorial border controls depends on the extent to which the state exercises justified authority. If border practices that extend beyond territorial borders are illegitimate, this is because they are unjustified exercises of authority over migrants (Sandven 2024; Scherz and Sandven 2022; Shachar 2020). The idea is that the state rules migrants only if they are subject to the authority claimed by the state.

“Coercion” and “authority” are distinct conceptions of state rule. Of course, states often subject people to coercion and authority at the same time. In the domestic context, a resident citizen that is arrested for a crime is both subject to the coercive powers of the state and to the state’s claimed authority to govern her conduct. But coercion and authority are nevertheless conceptually distinct. The legitimate authority to decide does not entail the right to coerce. Likewise, the right to use coercion does not entail the legitimate authority to decide (Raz 1986; p. 25; Roughan 2013; p. 24).

Given these distinct accounts of what it means for the state to “rule” the primary question is in what sense (if any) migrants are ruled by extraterritorial border controls. This question is critical since a policy can be illegitimate with respect to people only if they are ruled by them. Hence, extraterritorial border controls can be illegitimate with respect to migrants only if migrants are rule by these policies.

However, the label “extraterritorial border controls” is often used to describe various policies that involve different modes of governance and rule.Footnote 3 The following analysis focuses exclusively on what is known as “carrier-sanctions”: the policy of preventing migrants from travelling to the state by enforcing requirements of prior approval to entry on anyone embarking on “carriers” (airlines, ferries) in foreign ports and airports. Carrier-sanctions are closely related to visa-policies that require prospective migrants to gain prior permission of entry (Baird 2017) and create what FitzGerald (2020) describes as an impenetrable “dome” over the territory of the state.Footnote 4 Indeed, carrier-sanctions is arguably the “primary tool” by which states deny migrants entry by applying border controls before borders are reached (Gammeltoft-Hansen 2012).

For present purposes, the relevant feature of carrier-sanctions is that they are enforced in territories controlled by other states. This policy may seem to extend the ruling activities of the state, not just beyond its own borders, but also into the borders of other states. In this regard, carrier-sanctions differ from, for example, interceptions of migrants in the high seas. Although maritime interceptions are applied extraterritorially and aim to prevent migrants from reaching the territory of the state, they are not attempts to extend the rule of the state into the territory of another state – the high seas are not sovereign state territories (Milanovic 2016, p. 239). In fact, state borders are constitutive elements of the international system of states as indicated by norms of territorial state jurisdiction and the principle of non-intervention.Footnote 5 There is a strong presumption in international law against the state enforcing its laws on territories controlled by other states (Ryngaert 2008; p. 24).

The claim defended is not that these features of the international system of states excludes the possibility of extraterritorial rule of migrants by definition. The more specific argument developed is that carrier-sanctions do not involve the exercise of either state coercion or state authority over migrants in the territories of other states and that standards of legitimate state rule do not, therefore, apply to these policies.

In the first section of the paper the argument is made that while carrier-sanctions are often coercive this is not the relevant metric of state “rule”. In the second section, the argument is that the state does not claim the authority to rule migrants present in foreign states by carrier-sanctions. Hence, neither the view that legitimate rule depends on permissible coercion nor the view that legitimate rule depends on legitimate authority applies to the ruling activities associated with carrier-sanctions. The conclusion is not, however, that migrants prevented from travelling due to carrier-sanctions are not ruled by any state. In the third and final section two claims are defended. The first is that legitimacy standards apply to the state with de facto authority over the territory where carrier-sanctions are enforced. The second is that considerations of justice apply to states that introduce carrier-sanction legislation.

2 Legitimacy and Coercion

The capacity for coercion is a defining feature of the state: states are organizations that can enforce decisions against potentially unwilling subjects. By the use of sanctions and punishments the subject population is coerced into doing what they would not otherwise have done.Footnote 6 But bending the will of others is often morally problematic. Following an influential strand of thinking, the justification of coercion is key to the legitimacy of the state.Footnote 7 The state is legitimate only when the subject population is ruled by morally permissible coercion.

2.1 How Carrier-sanctions Coerce Migrants

Legitimacy so conceived is an attribute only of coercive state rule. The claim that carrier-sanctions and other extra-territorial border controls are not legitimate is therefore premised on the fact that these policies subject migrants to coercive rule that is not permissible. But it is not obvious that carrier-sanctions entail coercion against migrants. Consider, for example, the account of state coercion influentially proposed by Hans Kelsen. On his view, the law is a “technology” designed to regulate the coercive activities of the state. State coercion is regulated by law, meaning that the law also provides for coercion: legal norms are instructions that define the conditions for when public officials are permitted or required to impose punishment or sanctions against subjects (Kelsen 1991; p. 54).

The crux is that carrier-sanctions are legal norms that provide instructions for when the state should perform coercive acts against carriers – but not against migrants. Migrants prevented from embarking on an aircraft carrier are apprehended by privately employed airline staff rather than by public officials. Carrier-sanction legislation is not designed to authorize state coercion against migrants but only against carriers. As argued by Blake: “the carrier sanction isn’t coercion against the one seeking to migrate, but against the carrier providing him with the means needed to migrate” (Blake 2019; p. 115).

The fact that carrier-sanctions do not authorize coercive acts against migrants has not prevented critics from concluding that this policy constitutes illegitimate coercion of migrants. According to Bloom and Risse (2014; p. 76), carrier-sanctions amount to “indirect” coercion. The indirectness consists in the fact that coercion is “channeled” through carriers, which equals what Ataner (2004) describes as “indirect interdiction”. A plausible reading of this argument is that migrants are subject to state coercion because the actions of the state explains why they are subject to coercion by carriers. In fact, this indirect way of exercising coercion is making carrier-sanctions worse in terms of legitimacy according to Bloom and Risse. In addition to subjecting migrants to coercion that is impermissible, carrier-sanctions obstruct both democratic and judicial control and thereby undermine the procedures that normally serve to legitimize public decisions.Footnote 8

Yet, the assertion that “indirect” coercion challenges the legitimacy of the state requires justification. It remains to be explained how it could be that a coercive act that is not performed by the state is nevertheless compromising the legitimacy of the state. After all, states that pursue carrier-sanctions do not partake in coercive acts that apply to migrants.

Following Blake, the legitimacy of carrier-sanctions would be an issue if coercion was transitive. If coercion is transitive, a coercive act undertaken by the staff of the carrier is equivalent to a coercive act undertaken by public officials. Yet Blake denies that coercion is transitive in this scenario. Migrants coerced by carrier-sanctions are not subject to state coercion since a coercive act is attributable only to the agent that performs the act (Blake 2019; p. 103). No matter why carrier employees refuse migrants to travel, it is they, not the state, that refuse migrants to embark.

There is reason to doubt, however, that a coercive act can be attributed only to the agent that performs it. A person that hires a hitman is not absolved from murder just because she did not pull the trigger. Similarly, an officer that orders a private to execute a prisoner of war remains responsible for a war crime even if the officer did not participate in the execution. Criminal law regularly attributes responsibility for coercion to instigators of coercion even if they do not directly participate in the coercive act.

One possible way to account for this is by following Gerver et al. (2023; p. 957) in arguing that coercion is transitive whenever it can be explained by the coercive actions of others. The fact that migrants subject to coercion by carriers is explained by the policy of carrier-sanctions shows that the state “causally contributes” to the coercion of migrants. Since the state causally contributes to migrants being coerced, there is reason to conclude that the state is subjecting migrants to coercion.

But it is unclear that causal explanation is sufficient for coercion to be transitive. Consider a parent that drives his car and is stopped and fined by the police for driving too fast. As a result of the hefty fine coercively imposed on him, the parent gets furious and starts beating his children that are also present in the car. Provided that the parent would not otherwise have acted thus it seems that the coercive act performed by the police is part of the explanation for why the children are beaten. Yet, it would be absurd to conclude that the police has therefore subjected children to coercion. The fact that one agent causally contributes to another agent making a coercive act against a third agent is not sufficient to conclude that the first agent subjects the third agent to coercion.

For coercion to be transitive in the normative relevant way, the outcome must be explained by actions that are intended to result in another agent performing a coercive act. The person that hires a hitman is considered responsible for murder only because his actions causally contributed to and intended to cause the murder. By analogy, we would say that the coercive effects of carrier-sanctions are attributable to the state only if they intended to make migrants subject to coercion. Although the state is not the agent that coerces migrants, the state is – just like the person hiring a hitman – responsible for the fact that migrants are coerced. Having concluded that coercive acts against migrants are attributable to the state even if only the carrier is performing the coercive act, there is reason to believe that carrier-sanctions rule migrants in ways that makes the state liable to standards of legitimacy.

2.2 Two Objections Considered

However, there are further potential objections against the claim that carrier-sanctions are illegitimate because they subject migrants to impermissible coercion. The first is the now familiar argument by Miller (2010) that preventing is distinct from coercing. Migrants denied entry to the state are not made to do what they would not otherwise have done. Instead, they are prevented from doing what they would otherwise have done. The point is that the concept of coercion only applies to actions that exert pressure on others to make them do something. This observation is relevant to carrier-sanctions provided that the policy is only preventing migrants from doing what they would otherwise have done. Carrier-sanctions are preventive and therefore not coercive on this understanding.Footnote 9

The distinction between preventive and coercive acts is nevertheless suspicious. Consider the paradigmatic case of coercion: the robber that pressures the victim to do what she would not otherwise have done (presumably, handing over her valuables to the robber). The robber clearly performs a coercive act on Miller’s account. Yet, this account would not consider as coercive the robber exerting pressure on the victim not to call for help. Making the victim not call for help is not making the victim do anything – it is a preventive, not coercive, act. And yet, it seems obnoxious to conclude that the victim is not thereby subject to coercion. There is consequently reason to recognize that preventive actions can be coercive: carrier-sanctions are no less coercive just because they prevent migrants from doing what they would otherwise have done.

The second objection is that it is unclear that coercion is necessarily impermissible and that policies that involve subjection to coercion are therefore illegitimate. In fact, coercion is a pervasive feature of both the state and regular human interaction. We are recurrently subject to pressure in the attempt to make us do (or not do) what we would not otherwise have done (or not done) (McCloskey 1980; p. 342).

Yet, the predominant view is that coercion is impermissible at least when it violates the autonomy of the person (Cf. Taylor 2022). Provided that it is pro tanto impermissible to compromise a person’s autonomy, it follows that coercive actions by the state that violate the autonomy of the person are presumptively illegitimate. If, as already argued, carrier-sanctions do subject migrants to coercion, there is a case for the claim that carrier-sanctions are illegitimate.Footnote 10

This reasoning must be qualified, however. My autonomy is not invaded just because the concert I wanted to attend is sold out, the house I wanted to buy is too expensive, or the person I fancy is in love with another. The autonomy of a person is not compromised as soon as any option is foreclosed; it is compromised only when the person is denied an adequate range of options (Raz 1986). The reason my autonomy is not compromised when concert tickets are sold out is that a sufficient range of valuable options are likely to remain available. The relevant question about coercive carrier-sanctions is not, then, whether they prevent migrants from doing what they want but whether they deny migrants an adequate range of valuable options (Pevnick 2011; p. 85).

The answer to that question is that carrier-sanction sometimes deprive migrants of an adequate range of options. When that is the case, the autonomy of the migrant is compromised by coercive carrier-sanctions. This is particularly true when carrier-sanctions are coordinated among wealthy and democratic states and deprive migrants from poor and autocratic countries access to valuable options. The same applies to refugees that are subject to persecution or severe deprivation in their home countries and where carrier-sanctions are instrumental in denying them access to options that provide safety from persecution and severe deprivation.

Of course, the permissibility of carrier-sanctions still depends on circumstances. They are likely to be permissible if enforced only by a few states as this is consistent with migrants having access to other options of equal or comparable quality. Furthermore, carrier-sanctions are permissible with respect to migrants that already have access to an adequate range of valuable options where they live (Stilz 2020; cf. Hildago 2014).

2.3 Coercion as Currency of State Legitimacy

The notion that the legitimacy of border controls depend on the moral permissibility of coercion is premised on the idea that state coercion is the “fundamental problem of governance” (Mansbridge 2014; p. 9; Tomasi 2012; p. 65). The state, so conceived, is primarily a coercive organization with the capacity mobilize physical power, force, and violence.

The consequent image of the state is akin to the “gunman writ large”, to use H.L.A. Hart’s (1962) memorable phrase. According to this image, the state differs from the gunman only with respect to the scale of its coercive power. This view has been criticized, by Hart and others, for obscuring the fact that states wield normative as well as coercive powers. The state is an ‘institutional normative order’ that claims the authority to rule and that rules the subject population also by non-coercive means (MacCormick 1997; Raz 1979; p. 150).

Non-coercive state action include the creation of legal norms that impose duties on subjects that are not attached to coercive sanctions. The conduct of law-makers in parliaments and other elected bodies is a case in point. Their conduct is regulated by constitutional and other legal norms that impose duties that regulate the process of law-making (Raz 1999; p. 158). Norms regulating legislative procedures of national parliaments typically include duties to prepare legislative proposals by standing committees, for example. Yet, the duties that regulate the conduct of law-makers are not enforced by the threat of sanctions or punishment for non-compliance. Consequently, the state imposes duties on subjects that are not coercive.

Non-coercive state action is present also in the creation of legal norms that are not duty-imposing but that regulate the legal powers of subjects. Norms that regulate legal powers include, for example, the right to vote, the right to marriage and the power to make contracts. A legal norm that confers a legal power is not coercively enforced because it is not a norm that requires compliance. The point is that standards of legitimacy should apply also to the ruling activities of states that are not coercive. But this possibility is foreclosed if legitimacy standards apply only to rule that is coercive.

A further objection against the view that states rule only by coercive means is that it misconstrues the relationship between the law and state coercion. The state is a legal entity in the sense that it is legally empowered to create, revise, and abolish legal norms (MacCormick 1997; Raz 1999; p. 150). The legal powers of the state include the authority to introduce coercive sanctions and to undertake other forms of coercive acts. Whenever the state engages in coercion, it performs an act that it claims the right to do in virtue of its claimed authority: state coercion is “legitimate” only if authorized. Coercive acts that are not authorized are illegitimate because they cannot be traced back to the authority claimed by the state. Hence, the morally permissibility of a coercive act is not sufficient to its legitimacy. A coercive act that is not authorized by the state is illegitimate even if morally permissible.

Every state claims the authority to regulate the conduct of subjects (Raz 1999). The authority to regulate conduct is the normative bedrock also for the coercive powers of the state (Green 2016; p. 167; Tasioulas 2010; Lamond 2001; p. 55). The implication is that the relevant question about legitimacy is not if coercion is morally permissible but whether the state has the authority to engage in coercion. In case it does not, the state lacks morally justified authority to use coercion. The legitimacy of state coercion can in other words be fully accounted for by the view that the legitimacy of the state depends on the extent to which it wields morally justified authority.

In contrast to gunmen, the state does not merely wield coercive power. The state also claims the authority to regulate conduct by norms and the authority to enforce them when necessary. Coercion is a secondary feature of the state since the permissibility of coercion depends on the authority of the state. The coercive actions of the state are legitimate only if lawful and the laws and directives of the state are legitimate if only if they amount to morally justified authority.

3 Legitimacy and Authority

Writers that endorse the authority-based conception of state legitimacy argue that extraterritorial border controls regularly fail to be legitimate. Following Sandven and Scherz (2022 p. 678) the border regime of the European Union represents a “genuine instance of a claim to authority by states over individuals who reside outside of their jurisdictions”. The authority implied is illegitimate because it fails to provide “accountable human rights protection” for migrants. These ills are aggravated by the fact that border controls are delegated to FRONTEX that is largely shielded from accountability to both democratic publics and judicial mechanisms of accountability for human right abuse.

A similar argument is made by Sandven (2024) who specifically addresses the problems infecting the exercise of state authority in extraterritorial border controls. The argument is that the “externalized and coordinated nature” of border controls that are enforced beyond the territory of the state is not regulated by mechanisms that robustly protect the human rights of migrants. States that externalize border controls are making decisions that foreseeably compromise the human rights of migrants.

Both critics defend the claim that extraterritorial border controls are illegitimate exercises of authority over migrants either because they directly violate the human rights of migrants, or because they fail to provide robust or accountable mechanisms for their protection. It should be noted that none of them are specifically concerned with the legitimacy of carrier-sanctions. However, both mention carrier-sanctions (Sandven and Scherz 2022; p. 682; Sandven 2024; p. 550). And both argue that all forms of extraterritorial border controls – presumably including carrier-sanctions – are illegitimate. There is consequently reason to ascribe to them the view that carrier-sanctions are illegitimate because they are exercises of authority that undermine the human rights of migrants and that fail to provide mechanisms for robustly protecting them.

These critics are explicitly committed to the notion that the authority claimed by the state is legitimate only if the state provides the subject with reasons that “makes her comply with reasons in a given domain” (Sandven and Scherz 2022; p. 688). Authority so conceived is the right to provide reasons for action that others should act on. This formula coheres with the widely accepted understanding of authority as the capacity to provide “content-independent reasons” (cf. Valentini 2018). Legitimate authority provides subjects with reasons that they should act on independently of the content of the reasons provided. The thought is that authority is the normative power to make others do what they have moral reason to do. Hence, the legitimacy of the authority claimed by the state depends on the laws and directives of the state providing subjects with morally correct reasons for action.

An implication of authority so understood is that it is always concerned with the relationship between the agent claiming authority and the subject to whom that claim is addressed. The legitimacy of border controls is therefore concerned with the authority claimed by the state enforcing the border over actual or potential migrants. The claim that extraterritorial border controls are illegitimate is thus a claim about the authority claimed by the state over migrants beyond the borders of the state. Applied to carrier-sanctions, they are illegitimate only if they involve morally unjustified claims to authority over migrants.

Despite claiming that extraterritorial border controls involve exercises of authority over migrants outside of the jurisdiction of the state, the critics cited above do not provide much detail on how this authority is exercised. More often, the claim is that these policies are mere exercises of “unaccountable power” (Scherz and Sandven 2022; p. 689) that involve “state power over vulnerable persons” (Sandven 2024; p. 555). But power and authority are not the same.Footnote 11 Power is the general capacity to effectuate changes in the world or to get others to do what they would not otherwise do (Abizadeh 2023). By contrast, authority is the capacity to create reasons for action by normative power (Wenar 2005; Raz 1979). From the fact that the state exercises “power” over a person it consequently does not follow that the state exercises authority over that person. Given the lack of detail about the extent to which border controls are exercises of authority over migrants, there is reason to examine the issue further.

As should be clear by now, standards of legitimate authority apply to carrier-sanctions only if the state that introduce them claim the authority to regulate migrants. In the following I raise doubts about the cogency of this claim by exploring two distinct accounts of what it means for the state to exercise authority. On the first, duty-based account, the state exercises authority by claiming the right to compliance. Carrier-sanctions are exercises of authority over migrants only if the intention is that migrants should comply with them. On the second, power-based account, the state exercises authority by claiming the right to regulate the legal powers of the subject. The point is that the authority claimed by the state need not, on this view, involve a claim for compliance. A sufficient condition for carrier-sanctions exercising authority over migrants is that they regulate the legal powers of migrants.

3.1 Duty-based Authority

One influential view is that the state exercises authority by claiming the right to impose duties on subjects. The idea is that a legitimate state possesses “the right to rule over subjects who owe obedience to the authority” (Christiano 2009; Chap. 6). The authority claimed by the state is legitimate only if subjects have a moral duty to comply with its directives and laws (Edmundson 2010; p. 180). Following this account, carrier-sanctions are legitimate only if migrants have moral duties to comply with them. These duties exist only if the state has the moral authority to regulate the conduct of migrants.

However, the notion that carrier-sanctions are illegitimate exercises of authority, in this sense, presupposes that carrier-sanctions constitute an attempt by the state to impose duties on migrants. I argue below that carrier-sanctions do not seek to impose duties on actual or potential migrants and that carrier-sanctions are not, in that sense, exercises of authority over migrants.

As already observed, the primary subject of carrier-sanction legislation is carriers not migrants. For example, the UK Immigration and Asylum Act (1999) only mentions carriers as subjects of duties. Carriers are legally required to make payments to the Secretary of State for any passenger brought to the UK without proper documentation. Carrier-sanction legislation in other countries include the legal duty of the carrier to control travel documents, to refuse travel to travellers without the required documentation, to provide data about passengers to migration authorities, and to re-transport migrants refused entry on arrival (Scholten 2014; p. 17).

Hence, carrier-sanction legislation is not designed to generate legal duties for migrants. This is significant given that authority is understood as the capacity to impose duties on the subject. If a person is subject to the authority claimed by the state if and only if she is under a legal duty to comply with the directives of the state, it follows that the state does not subject migrants to its claimed authority through carrier-sanction legislation.

On the other hand, carrier-sanctions are enforced by carriers requesting valid travel documents from migrants. Hence, we might argue that migrants have duties because carriers require compliance with their requests. A request is an act where the intention is that the request is taken as a reason to perform the act requested. Since migrants are subjected to requests by carriers, they are presumably subject to a claim for compliance.

Yet, the observation that carriers makes requests on migrants does not vindicate the claim that migrants have duties towards carriers. Requests and duties are different. This is illustrated by the fact that requests may be rejected without disregarding them. Not so with duties. Rejecting a duty is to disregard the duty to perform the act (Raz 1979; Chap. 1). Hence, the fact that carriers request documentation from migrants is not sufficient to conclude that migrants have duties towards carriers.

On the other hand, it is not obvious that carriers are merely making request. Perhaps carriers issue an order to the effect that the migrant must provide the required documentation. Following Raz (ibid.) an order is intended to be a reason for action that the recipient should comply with regardless of any other reasons she might have. Whilst a request is a reason for action that should be considered, an order is a reason for action that should be obeyed. Indeed, the notion that migrants are subject to orders issued by carriers, and therefore under duties to them, seems right given the instruction provided by carriers that migrants “must provide” the necessary documentation. There is consequently reason to submit that carriers claim the authority to impose duties on migrants.

But duties towards carriers and duties towards the state are distinct. The critique against the legitimacy of carrier-sanctions is directed against the laws and directives of the state. The laws and directives of the state are illegitimate exercises of authority only with respect to subjects required to comply with the state. Duties towards carriers are duties towards the state only if they are extensions of the state’s claim that migrants comply with the law.

They are not, however. The claim that migrants are under a duty towards the carrier to provide the required documentation does not entail that migrants are under a duty to comply with the state. The migrant that refuses to show her passport or VISA to the carrier does not violate any legal precept of the state. Again, carrier-sanction legislation regulates carriers, not migrants. Any duty of passengers towards carriers depend solely on the authority claimed by the carrier, not on the authority claimed by the state.

Against this reasoning, the objection might be raised that every passenger has a duty to provide documentation according to international law. This observation lends support from the fact that most countries are signatories to international legal agreements that regulate the terms of inter-state travel. Specifically, the 1944 Chicago convention on international air travels includes the provision that passengers must comply with entry regulations and other laws imposed by states. Article 13 of the convention stipulates that the “laws and regulations of a contracting State as to the admission or departure from the territory of passengers, crew or cargo of aircraft, such as regulations relating to entry, clearance, immigration, passports, customs and quarantine shall be complied with by or on behalf of such passengers, crew or cargo upon entrance ...”. Accordingly, passengers have duties to comply with the laws regulating admission to the territory of the state. Since carrier-sanctions are laws regulating admission to the territory of the state and migrants are prospective passengers it follows that migrants have legal duties to comply with carrier-sanctions.

However, duties owed to the international legal system are not duties owed to the state. Even if, under international law, migrants have duties to comply with state regulations, this is a duty owed to international law, not to the state. While this may not make much of a difference to the agent that is duty-bound, it is quite relevant to the argument under scrutiny.

The argument under consideration is that carrier-sanctions are illegitimate because they represent illegitimate claims to authority by the state. But from the fact that migrants are under duties of international law to comply with carrier-sanctions it does not follow that migrants are subjected to the authority claimed by the state. What follows is instead that migrants are subjected to the putative authority of international law. In case international law imposes duties on migrants that are contrary to respect for human rights, and if authority is legitimate only if consistent with respect for human rights, it follows that the legitimacy of international law is questionable. But this conclusion does not entail that the legitimacy of the authority claimed by the state is questionable. The conclusion is that no grounds have been identified for the claim that migrants are subjected to the authority claimed by the state, given that by “authority” is understood the capacity to create duties in others.

3.2 Power-based Authority

On a distinct view, authority is the capacity to “change the rights and duties of others” (Brinkman 2020) and, more generally, the moral power to determine the subject’s normative positions (Applbaum 2019).Footnote 12 This account allows for the ascription of authority to acts that do not claim compliance: the state is exercising authority also when it seeks to regulate the “second-order” normative positions of the subject population. Though the second-order normative positions include both immunities and powers (Wenar 2005) only the powers are relevant in the present context. The hypothesis is that a state introducing carrier-sanctions claims the right to regulate the legal powers of migrants and is thereby subjecting them to authority: “remote” border controls exercise authority by regulating what migrants are legally empowered to do (Galloway 1991).

The legal powers of the subject are regulated whenever the state regulates the legal capacity of the individual to change her legal status by voluntary action. A clear-cut example of legal power enjoyed by the individual in any jurisdiction is the capacity to enter contractual relationships. Through the legal power to make and agree to contracts, the individual has the ability to change her legal claims and duties in various respects. For instance, by the right to take a mortgage on property, the individual acquires a duty to repay the debt against the creditor while she is at the same time provided the means to acquire claim-rights in the property purchased.

The state’s authority to determine the legal powers of subjects extends beyond the territorial borders of the state. Citizenship status is, for example, a legal status that includes the legal power of the citizen to enter her state and, often though not always, the legal power of the citizen to participate in national elections. The legal powers entailed by citizenship are by no means invalidated just because the individual citizen is present abroad. In fact, many of the rights associated with citizenship are specifically designed to apply to citizens abroad, e.g. rights to assistance from the state, rights to return to the state and expatriate voting rights (Owen 2010). The point is that the authority to regulate the legal powers of subjects is transcending the territorial borders of the state.

This observation is relevant to border controls as it indicates that the state may wield extra-territorial authority to regulate the legal powers of migrants. The question though is if carrier-sanctions do in fact regulate their legal powers? The argument in favour of this claim would run as follows: in the absence of carrier-sanctions and other extra-territorial border controls, migrants have the capacity to change their normative positions by voluntarily travelling to the state. Anyone entering the territory of another state changes her normative position since she acquires a new set of rights and duties. Given that the capacity to perform an action that changes your normative position equals the exercise of legal power, it follows that the migrant is exercising a legal power when she is travelling to another country. But these legal powers are effectively denied by states that introduce carrier-sanctions. Carrier-sanctions deny migrants the legal power to voluntarily change their normative positions by travelling to another country. Hence, the argument is that carrier-sanctions amount to claims to authority over migrants by regulating their legal powers.

It is uncertain if this argument is sound, however. The argument is premised on the understanding that legal powers are invoked whenever people perform actions with consequences for their normative positions. But this notion is often considered problematic in the literature as it seemingly implies that a person is exercising legal power when she breaks the law. By committing a crime, the agent changes her normative position by making her liable to punishment. Yet, it is odd to assert that a person is exercising a legal power when she engages in for example theft or murder (Raz 1984). The more plausible view is that legal powers are exercised only through actions that are authorized by power-conferring norms. Legal powers are necessarily “acts-in-the-law” (Kurki 2017). On this understanding, the decision to break the law is not an exercise of a legal power since there is no law that confers on the subject the power to break the law.

Legal power so conceived is not at play when the migrant decides to travel to the territory of the state. Although her decision is an attempt to change her normative position, the decision is not an exercise of a power conferred to her by the law. The decision to travel to the state is not an exercise of a legal power because the legal system does not confer on migrants the legal power to travel to the state. Of course, this is not to say that travelling to the state is necessarily illegal. Performing an act that is not recognized as an exercise of a legal power does not violate the law – it is just a nullity.

Hence, carrier-sanction legislation does not regulate the legal powers of migrants abroad. Of course, carrier-sanctions still prevent migrants from achieving the status as persons with the legal power to apply for asylum. But preventing a person from becoming a person with legal power is not the same thing as regulating the legal powers of the person. I may have the physical or social power to prevent others from exercising their legal powers. I could, for example, prevent my friend from exercising her legal power to marry if I refuse to drive him to the magistrate and there is no other means of transportation available to him. But my decision is not an attempt to exercise power-based authority over my friend as it does not change the fact that he is legally empowered to marry.

The conclusion is that a migrant located in the territory of a foreign state is not subject to the authority of the state introducing carrier-sanctions on either the duty-based or power-based conception of authority. Clearly, if carrier-sanctions do not subject migrants to the authority claimed by the state, these policies cannot be illegitimate exercises of authority over migrants.

4 Reconsidering Legitimacy and carrier-sanctions

It is now time to consider the implications of our analysis for the claim that carrier-sanctions are illegitimate. As argued in the first section, carrier-sanctions are coercive and often have morally problematic consequences for migrants. Yet, I have also raised doubts about the view that state legitimacy is primarily concerned with the extent to which subjects are exposed to coercion. Even non-coercive decisions by states can be illegitimate: a decision that does not result in coercion is illegitimate if the state does not have the legitimate authority to make it. Moreover, decisions that are coercive can be illegitimate even if they do not result in impermissible coercion. The argument is that authority – not coercion – is the relevant metric of state rule. A necessary precondition for the illegitimacy of carrier-sanctions and other extraterritorial border controls is that they rule migrants by claiming authority over them.

The second section showed that carrier-sanctions do not represent an attempt by the state to exercise authority over migrants. Irrespective of whether authority is conceived of as a claim for compliance or as the power to regulate normative positions, migrants located in foreign territories are not subjected to the authority of the state introducing carrier-sanction legislation. Hence, the notion that carrier-sanctions are illegitimate because they represent illegitimate exercises of authority over migrants abroad does not hold up. As far as we are concerned with carrier-sanctions, migrants are not “norm subjects” of the state introducing them (cf. Schotel 2011).

Now, grounds for concern with the legitimacy of carrier-sanctions do not disappear just because they are not illegitimate claims to authority over migrants. It is still the case that the human rights of migrants are jeopardized when states engage in practices that deny them recourse to asylum and rights against non-refoulement. Below I argue that a concern with illegitimate border controls apply to the state with de facto authority over migrants – that is, to the state in control of the territory where carrier-sanctions are enforced. Lastly, I conclude by the reminder that considerations of justice apply to all state action and that states presumably have duties of justice not to introduce border policies that undermine the human rights of migrants.

4.1 Carrier-sanctions and de Facto Authority

Carriers that deny migrants the right to embark are enforcing their decisions in the territory of some state. While carriers tend to act in conformity with carrier-sanction legislation introduced by one state, their actions are undertaken on territories controlled by another state. The fact that some state controls the territory where carrier-sanctions are enforced is relevant given that the authority claimed by the state is an exercise of authority only if it enjoys de facto authority. If claims to legitimate authority depend on de facto authority, it should be clear that concerns with legitimacy apply to the state that controls the territory where carrier-sanctions are enforced.

There is a compelling case to be made for the view that legitimate authority presupposes de facto authority. An agent without influence over conduct does not have authority even if she claims authority and even if that claim is morally justified. Only claims to authority that are effective in guiding the conduct of subjects can be legitimate: “no political authority can be legitimate unless it is also a de facto authority” (Raz 1986; p. 46).Footnote 13

De facto authority should not be confused with the capacity to secure conformity with rules or commands. Conformity can be achieved by any means and is, for example, secured by robbers, pirates, and gunmen that are sufficiently powerful. Only an agent that claims legitimate authority can achieve de facto authority, however. The subjects of de facto authority do not conform for prudential reasons but because they perceive themselves as duty-bound to comply (Dagger 2018; p. 5).

As already observed, the state necessarily claims legitimate authority. But the state achieves de facto authority only if subjects believe that the authority claimed by the state is legitimate such that they should comply (Raz 1994; p. 215). Generalized belief in the legitimate authority of the state is a necessary precondition for the state achieving de facto authority (Føllesdal 2018).

It follows from this understanding of de facto authority that a state introducing carrier-sanction legislation does not enjoy de facto authority in the location where it is enforced – i.e. in foreign ports and airports. The state that introduce carrier-sanctions does not claim legitimate authority over carriers and migrants abroad and such a claim is necessary for de facto authority. This is not to deny that migrants prevented to travel by carriers are subject to the de facto authority of some state. Absent failed states, the state normally wields de facto authority over all activities within its borders. It does so provided that it claims the right to regulate all conduct in its territory and that this claim is generally recognized as valid by a “sufficiently large” proportion of the subject population (Bicchieri 2006; p. 11). There is thus reason to conclude that de facto authority belongs to the state in control of the territory where carrier-sanctions are enforced: the migrant that is refused travel because of a decision made by a carrier is subject to the state that claims the authority to regulate conduct in the territory where this decision is made.

This observation challenges the claim made by Bloom (2013) that carrier-sanctions legislation delegates authority to carriers abroad and Rodenhäuzer’s (2014, 231) claim that carriers are “empowered by law” to refuse migrants’ travel. Though true that carrier-sanction legislation are attempts to induce private corporations to act in conformity with the aims of the law, it is not true that this legislation exercises authority over carriers operating in other countries. Only a state with de facto authority is exercising authority and only the state that controls the territory where carriers operate enjoys de facto authority. The implication is that we should re-consider the target state of legitimacy-claims. The critique that carrier-sanctions are illegitimate should be targeted against the state that allow carriers to enforce the carrier-sanctions of other states in their territories.

It may be objected that a state allowing carrier-sanctions to be enforced on its territory has not authorized carriers to enforce border controls introduced by other states. But do to nothing is to permit others to do things. Carriers enforce carrier-sanctions because they are not duty-bound by the state in control of the territory to refrain from so doing. By permitting conduct, the state also exercise authority (Raz 1999; p. 151). Hence, the concern with legitimacy should be re-directed towards the state that permits carriers to enforce carrier-sanctions introduced by other states. If the permissive norm authorizes activities that are not morally justified, the conclusion must be that the authority claimed by the state that permits the activity is not legitimate.

4.2 Carrier-sanctions and Justice

The second point to make is that although states introducing carrier-sanction legislation do not exercise illegitimate authority over migrants abroad, their actions may still be unjust. The charge that a decision by the state is illegitimate is not the only possible charge that can be made. A different but no less serious charge is that the state is violating precepts of justice. While state legitimacy depends on justified authority, the outcomes produced by the decisions made by states are always subject to considerations of justice. Justice and legitimacy are distinct normative categories (Carens 2014). It makes perfect sense to argue that a policy is unjust but legitimate and vice versa. A policy that is unjust but legitimate is consistent with the legitimate authority claimed by the state but contrary to considerations of justice. It is consequently conceivable that states introducing carrier-sanction legislation are acting contrary to the demands of justice even if they are not acting contrary to standards of legitimacy. Reasons to believe that these policies are unjust include, for example, the fact that they undermine the human rights of migrants and induce them to seek other, more dangerous, routes of travel that result in the risk of death (Baird and Spijkerboer 2019). The charge that carrier-sanctions are unjust is not the argument of this paper, of course. It is important to note, however, that such an argument is consistent with the thesis of this paper, which is that carrier-sanctions are not illegitimate exercises of authority over migrants abroad.