1  This entry concerns the concept of ‘deference’ and examines how it functions under international law. In ordinary usage, the term deference designates: ‘[s]ubmission to the acknowledged superior claims, skill, judgment, or other qualities, of another … [c]ourteous regard such as is rendered to a superior, or to one to whom respect is due [and] the manifestation of a disposition to yield to the claims or wishes of another’ (Oxford English Dictionary, 2020). This definition indicates that there may be a range of possible focal points, conditions, and justifications for deference. In essence, however, and as explained in further detail below, the concept of deference under international law can be defined to describe a response by one actor to the recognition of another actor’s decision-making authority (Shirlow, 2021). In such situations, deference means that the first (deferring) actor places some weight or relevance on the decision of the second (deferred to) actor in order to reach a decision on a given matter. Deference is therefore closely connected with the concept of ‘authority’, which describes ‘the ability of one actor to use institutional and discursive resources to induce deference from others’ (Barnett and Finnemore, 2004, 5; see similarly Raz, 2009, 136).

2  A requirement for some approach to deference will arise wherever a matter is determinable by more than one potential decision-maker (Shirlow, 2021). As explained in further detail below, deference assists decision-makers in situations in which there are overlapping claims to decision-making authority to settle disagreements as to where final decision-making authority lies. ‘Deference’ therefore describes approaches and reasoning devices that assist to structure the relationships between international and domestic actors, or between international actors operating in the same or different subfields of international law. The practical effect of recognizing another actor’s decision-making authority may prompt the deferring actor to adopt one of several different approaches to deference. This explains why a diverse range of analytical devices have been treated as reflective of deference in international law. Indeed, the concept of deference is closely related to other concepts which describe particular manifestations of—or approaches to—deference (Forum Non Conveniens; Comity; Domaine réservé; Margin of Appreciation; Standard of Review; Standard of Review: Investment Arbitration; Standard of Review: African Court on Human and Peoples’ Rights (ACtHPR); Standard of Review: Court of Justice of the European Union (CJEU); Standard of Review: Dispute Settlement System of the World Trade Organization (WTO); Standard of Review: European Court of Human Rights (ECtHR); Standard of Review in Appeal Proceedings; Standard of Review: Inter-American Court of Human Rights (IACtHR)) (for a detailed conceptual approach to understanding how these various approaches to deference interrelate see Shirlow, 2021).

3  Deference manages conflicts between claims to authority in one of two ways. First, deference might cause an (international or domestic) actor to take into account, give weight to, or accept as dispositive, specific, identifiable, ‘decisions’ of another (domestic or international) actor on matters of relevance to its decision-making process. In so doing, the first actor recognizes, and so defers, to the authority of the second actor. Deference in this sense describes: ‘a situation where a decision-making body … decides a case … not on the basis of its own independent assessment, but by treating the decision of some other decision-making body … as, in some degree, authoritative for its own … decision’ (Foley, 2008, 4). A focus on deference to decisions only, however, risks obscuring other important instances of deference. This is because ‘[d]eference may also refer to the allocation of authority’ (Daly, 2012, 8). Secondly, then, the term deference can be used to describe a situation in which one actor exercises ‘restraint’ when exercising its own authority, in order to give another actor greater scope to make decisions that lie outside the first actor’s scrutiny or control, even though the first actor would otherwise have (formal) authority to make decisions about those matters (Judicial Self-Restraint). Deference might thus arise not only by reference to specific decisions, but as a technique to make it more likely that one actor will have some room to manoeuvre or authority to make a decision on a given topic free from scrutiny or control by another actor.

4  To exhibit deference, the deferring actor must itself have some claim to the authority to determine a given matter. Regardless of whether deference attaches to a decision or to decision-making authority in the abstract, then, it must be distinguished from a lack of authority to decide a given matter. An actor does not defer only if it finds that it lacks formal authority (eg jurisdictional competence) to determine a matter, even if such a finding in practical effect results in another actor’s decision being treated as authoritative or final on that matter. Where an actor holds that it lacks formal authority to determine a matter, it is not deferring but is instead merely recognizing the limits of its own jurisdiction to decide. Deference, instead, arises where an actor has formal authority to determine a matter but nonetheless decides—for the epistemic or pragmatic considerations detailed below—not to exercise that jurisdiction out of deference to another actor’s authority. In such circumstances, the deferring actor ‘has some continuing authority to act, and does act; only its independent judgment is displaced, not its actual authority’ (Horwitz, 2008, 1077).

5  Deference can also be distinguished from decision-making on the basis of coercion (see generally: Weber, 1947, 53; Dick, 1978, 376; Schauer, 2004, 1567; Arendt, 2006, 93). Coercion produces obedience rather than deference. As Horwitz notes, ‘deference implies that [the deferring actor] has some power of independent decision-making, but chooses to displace its own judgment with that of [another]; obedience implies that [the deferring actor] follows [another’s] judgment because it has no choice but to do so’ [emphasis in original] (Horwitz, 2008, 1076). Deference can further be distinguished from decision-making on the basis of persuasion or agreement. Deference does not arise only because one actor accedes to ‘the views of others because they have explicitly or implicitly persuaded [that actor] that those views are correct’ (Schauer, 2004, 1571; see also: Arendt, 2006, 93; Dick, 1978, 377; Horwitz, 2008, 1074). Such a situation instead indicates decision-making on the basis of persuasion. Authority, by contrast, elicits deference to another’s views ‘even in the absence of substantive agreement’ (Venzke, 2013, 383).

6  The distinction between persuasion and authority illuminates the structure of the concept of deference: whereas persuasion describes reasoning on the basis of what are called ‘first-order’ reasons, authority prompts reasoning by reference to ‘second-order’ reasons (see, especially: Legg, 2012, 37; Raz, 1999). First-order reasons are reasons related to the merits of deciding (or not deciding) in a particular way (Raz, 1999, 41). For international law, such first-order reasons might include, for example, an assessment of the facts of a given dispute against applicable legal rules. Where an actor decides on the basis of persuasion, it undertakes its own appraisal of the first-order reasons for decision (the facts of the dispute) to reach a finding that may or may not accord with another actor’s own appraisal of those facts. Deference only arises if the actor allows its appraisal of those first-order reasons to be influenced by second-order reasons. Second-order reasons are ‘reasons to act or refrain from acting on one’s own assessment of the first-order balance of reasons’ (Legg, 2012, 18; see also: Raz, 1999, 41). An actor might for instance recognize another actor’s greater expertise for assessing whether certain conduct was justified by a public purpose. The actor might exhibit deference by referring to this second-order reason (expertise) to give some weight in its decision to a more expert authority’s appraisal of the first-order reasons (the facts) for ascertaining the existence of a public purpose.

7  Deference may stem from a treaty provision, an international actor’s procedural or inherent powers, and/or rules or principles of customary or general international law. Such sources will always form the starting point for the analysis of deference (see further Henckels, 2013, 2–3; Schill, 2012, 8; Allen, 2005, 416). Treaties in particular might direct international actors to adopt—or refrain from adopting—particular approaches to deference. An example of a treaty provision to this effect is Article 17.6 Agreement on Implementation of Article VI of the General Agreement on Tariffs and Trade (1994) (‘WTO Anti-Dumping Agreement’), which provides that (in examining a matter under the Agreement):

8  Even in the application of such directions, however, international actors will typically retain discretion to formulate an approach to deference, including to structure deference in a categorical or non-categorical way and to stipulate any conditions that might attach to it (see, for example, the interpretations of what Article 17.6 WTO Anti-Dumping Agreement requires in: Morocco – Hot-Rolled Steel (Turkey), Panel Report, 2020, para 7.2; US – Hot-Rolled Steel, Panel Report, 2001, para 7.6).

9  Where a treaty or applicable international rule is silent on the question of deference, this does not automatically mean that no deference is appropriate (compare Antaris Solar GmbH and Göde v Czech Republic, 2018, para 50). Instead, an international actor could refer to second-order reasons for deference to interpret the obligation more leniently or to give some weight to another actor’s interpretation of that obligation. Of course, certain approaches to deference might be implicitly or explicitly foreclosed by the interpretation of the rule (including any silence on deference). The Vienna Convention on the Law of Treaties (1969), for instance, is widely accepted to have displaced restrictive interpretations of treaties in favour of States (see further Orakhelashvili, 2003; Crema, 2010). Yet just because a particular approach to deference might be foreclosed under a given rule of international law does not meant that all deference per se is inapplicable. Instead, the ‘appropriateness of deference instead depends upon a close analysis of the treaty … to determine what exercises of authority might be required in its interpretation and application’ (Shirlow, 2021). Deference will be particularly likely to arise wherever an international obligation applies to empirically or normatively contestable matters. In such circumstances, to apply the obligation, the international actor will rely upon fact-intensive or normative (including moral or political) judgments that require particular types of empirical or normative epistemic authority. A need for some approach to deference will arise wherever a domestic or other international actor also holds authority to make such decisions. The international actor will then confront a question of deference: they could refuse to recognize any role for the authority of other actors; adopt the decisions of other actors as dispositive due to their superior authority; or defer the making of a decision pending the outcome of another actor’s decision-making process. The scope for deference will thus turn upon whether the international actor considers there to be a conflict (or overlap) between its own and another actor’s authority. The extent of the overlap will determine how decisive deference will be to the outcome in a given case (Shirlow, 2021).

10  The connection between authority and deference means that theories of authority can usefully explain why one actor might defer to the decisions of another actor in circumstances in which they are not persuaded or coerced to do so. Different theories of authority place emphasis variously on ‘instrumental’, ‘prudential’, and ‘non-instrumental’ second-order reasons for deference. The bulk of existing literature on deference in international law has focused on instrumental second-order reasons for deference (see especially: Henckels, 2014, 34–43; Cheyne, 2014, 40–41; Gruszczynski and Vadi, 2014, 166; Vadi, 2018, 195–201; Legg, 2012, 25–26, 145–74). According to this standard account, deference arises due to recognition by one actor that another actor has the best claim between the two to either the expertise or legitimacy needed to decide certain questions. These instrumental bases for deference are introduced in section C.1 below. The two other categories of second-order reasons that can prompt deference (prudential and non-instrumental second-order reasons) are introduced and considered in more detail in the subsequent subsections.

11  Instrumental deference arises under the ‘service’ and ‘leader’ conceptions of authority. In these theories of authority, deference follows the recognition by one actor of another actor’s superior expertise (service account) or legitimacy (leader account) to decide certain questions. These accounts of authority are ‘instrumental’ because they conceptualize deference to the authority of another actor as a means for the deferring actor to improve its own decision on a given matter. Thus, for empirically unclear or normatively contested questions, an actor might defer to another more expert or legitimate actor in order to increase its chances of making decisions that are ‘substantively good’ and/or ‘normatively legitimate’ (Elliott, 2013, 11; Horwitz, 2008, 1085; Daly, 2012, 73; Enoch, 2014, 234–35). Deference in these situations functions by reference to comparative authority: the deferring actor compares its claim to expertise or legitimacy to that of another actor in order to determine whether (and to what extent) to defer to the decisions of the latter actor.

12  In the service account, deference follows the recognition of another actor’s superior expertise (empirical epistemic authority) vis-à-vis the deferring actor. In such a situation, the deferring actor (D1) decides that another actor (D2) has greater knowledge and expertise about the issue that is before D1 for decision. The superior expertise of D2 might prompt D1 to defer to D2’s decisions on that issue, on the basis that D2 is more likely than D1 to make a ‘correct’ decision because of its superior expertise. Such superior expertise may derive from various features of D2, including its knowledge, decision-making process, insulation from potential bias, proximity to a given situation, or better resources (see further: Hershovitz, 2003, 212; McGoldrick, 2016, 35; Shany, 2005, 918; Nollkaemper, 2011, 29). The actor might be considered to be an expert because they themselves have the requisite expertise to make a particular decision. They might otherwise be considered an expert due to their institutional features or structure. The decision-making procedures of courts, for example, may make them less ‘expert’ when deciding polycentric or policy questions compared to legislatures or administrators (Carolan, 2009, 154; Harris, 2003, 641; Galligan, 1986, 243–44). Similarly, courts might be considered less expert than administrative actors at making technical determinations or socioeconomic policy decisions (Daly, 2012, 91–94; Legg, 2012, 153; Puthuppally, 2014, 88). In such circumstances deference may assist a court as deferring actor to benefit from the superior knowledge or decision-making processes of another (legislative or administrative) actor vis-à-vis these questions. Deference to the latter, more expert, decision-makers offers the court an improved chance of obtaining a correct answer than it could reach by relying only on its own decision-making authority (Horwitz, 2008, 1065–66; Henckels, 2014, 123; Van Harten, 2013, 80–81; Fahner, 2016, 62). Deference can thus provide a means by which the strengths and weaknesses inherent in institutional design can be appraised and accommodated on a case-by-case basis (Shirlow, 2021).

13  This is referred to as a service conception of authority because deferring to actors with expertise (empirical epistemic authority) serves the deferring actor by assisting them to reach more correct decisions. As Young explains, where another actor has better capacity to decide an issue, a failure to defer to that actor: ‘would, at worst, run the risk of replacing the correct answer … with [the deferring actor’s] own incorrect assessment. At best, it would give rise to inefficiency, with the [deferring actor] replicating the reasoning of the [other actor], reaching the same conclusion’ (Young, 2009, 563–64). Deference in these circumstances reduces the ‘error costs’ associated with the deferring actor’s exercise of decision-making authority (Horwitz, 2008, 1090). Deference will follow in a service account of authority wherever an actor perceives that: a question might produce several answers—some of which are more likely to be correct than others—and another actor is better able to identify those correct answers (Bodansky, 1999, 620). Deference to an expert authority in such circumstances is ‘no different from relying on a calculator or computer; deference is justified in exactly the way that one justifies using any tool as the best means of achieving one’s predetermined ends’ (Soper, 2002, 35–36).

14  Under the leader conception of authority, deference follows from the recognition that another actor has superior legitimacy (normative epistemic authority) to make a given decision compared to the deferring actor. There might be a range of reasons to conclude that another actor is a more legitimate decision-maker. A source-based view of legitimacy, for instance, would treat as legitimate any actor with a formal positional claim to authority (see, for example: Thomas, 2014, 735; Tyler, 2006, 376; Buchanan and Keohane, 2006, 405; Hurd, 1999, 381). Under such an approach, the deferring actor might treat as particularly authoritative the decisions of actors that have been vested with a ‘right to rule’ according to a religious theory, particular customs or traditions, or legal (eg constitutional) powers. Alternatively, an actor might be treated as a legitimate authority due to the procedural qualities of its decision-making process. Under a procedural view of legitimacy, decisions might thus be treated as legitimate (and so deferred to) if made via a participatory procedure or by a decision-maker that abides by norms of due process or fairness (Franck, 1988, 706; Hurd, 1999, 67–73; Shany, 2005, 4; Bodansky, 2012, 13–14). Finally, legitimate authority might derive from a more substantive theory of legitimacy that focuses on the normative features of decisions (see further: Franck, 1988, 17; Bodansky, 2012, 12; Thomas, 2013, 751). Using this approach to legitimacy, the deferring actor might defer to another actor’s decision provided it exhibits particular substantive values such as equality, fairness, or respect for human rights.

15  This is called a leader conception of authority because it offers a reason for deferring to actors in respect of issues to which no ‘right’ answer can be given (Soper, 2002, 48–49). Where there is no ‘expert’ that can settle such questions, the leader conception of authority recognizes that an actor might be given a ‘right to decide’ and so have the authority to lead others by deciding these contested issues (Soper, 2002, 43–45). As Hershovitz illustrates:

Considerations of safety and efficiency demand that people drive on the same side of the road, but it does not matter much if it is the left or the right. An entity that has the ability to enforce a solution may be a legitimate authority regarding which side of the road people drive on, despite its lack of expertise or any claim to better judgment (2003, 207–8).

16  Prudential second-order reasons might also prompt deference (Bickel, 1962). Deference may, for instance, follow from a prudential assessment that acknowledging another actor’s decision-making authority will lead to stability, prestige, or enhanced legitimacy for the deferring actor or the regime in which they operate. Deference in a prudential account does not operate to improve decision-making but is instead used strategically, including for example to signify ‘acceptance of one’s position in the hierarchy’, including as a tool of ‘appeasement’ (Fragale and others, 2012, 374). Communication studies in both social and professional settings, for instance, indicate that ‘deference behaviors affect how others perceive a communicator’ (Fragale and others, 2012, 374). The language of deference might thus be deployed as ‘no more than a rhetorical device used to “soften” the impact of adverse findings or to give … some sense of protection against possible criticism for “judicial activism”’ (Arai-Takahashi, 2002, 232; see, also: Calvert and Hayes, 2012, 18; Leonhardsen, 2014, 138; Meagher, 2011, 130; Judicial Activism). Such strategic uses of deference are likely to be particularly difficult to discern, though may underlie instances in which an international actor uses the language of deference in circumstances in which it ‘does not seem to serve any useful purpose’ (Kratochvil, 2011, 335; see, also: Gerards, 2011, 106; Meagher, 2011, 130; Candia, 2014, 9).

17  Non-instrumental second-order reasons might also prompt deference to another actor’s decision-making authority. Where it is prompted by non-instrumental second-order reasons, deference does not arise because the deferring actor recognizes another actor’s capacity to make a better decision, or because deference is recognized to be prudentially useful to the deferring actor. Deference in a non-instrumental account is instead used to express respect for certain inherent values, such as empathy, fairness, autonomy, subsidiarity, respect, trust, gratitude, or identification with a ‘community’ (Soper, 2002, 157–58, 170, 181–82). As Dan-Cohen explains: ‘the respect conveyed by deference can be a matter of trust [and to convey] … one’s belief that the authority will not lead one astray’ (Dan-Cohen, 1994, 36). Deference might, for instance, operate not on the basis of one actor’s recognition of another’s superior expertise or legitimacy, but rather by reference to the constraints faced by that actor in a given situation (see, for example Hoexter, 2000, 501–2). This might explain why sometimes deference is accorded to actors operating in a crisis scenario or under other practical constraints. An international actor might, for instance, defer to domestic actors responding to an economic crisis by assessing their conduct by reference to what was known at the time of the crisis (and the relevant acts), as opposed to what was known at the time of the international actor’s review of those acts (Brems and Lavrysen, 2015, 152; Paine, 2019). Deference in such circumstances does not offer a means of improving decision-making. The international actor is less constrained than the original decision-maker, and so presumably more able to assess the first-order reasons to produce an empirically or normatively ‘good’ decision. The international actor might nonetheless opt to defer by reference to non-instrumental second-order reasons, including out of respect or fairness. This could include, for example, because they recognize that they ‘would perform analogous acts’ if placed in the same position as the actor to whom they are deferring (Soper, 2002, 170), it being unfair to judge the State’s decisions with the benefit of hindsight.

18  The possible justifications for deference may overlap in practice. An actor might elect to defer to another on the basis, for example, of both empirical and normative epistemic authority justifications. A substantively legitimate authority might, for example, be particularly likely to also be an expert authority, insofar as expertise itself might be a source of legitimacy (Horwitz, 2008, 1093; see, also: Dyson, 2006, 103). For this reason, some theorists reject the leader conception of authority. As Raz argues, for example: ‘[i]f the primary point of authority … is to improve conformity with reason, it is implausible to think that someone who contributes not at all in that respect [ie as a non-expert] … can have legitimate authority’ (2009, 157). Thus, depending on an actor’s approach to conceptualizing the second-order reasons that can prompt deference, the above justifications for deference could operate in parallel, to differing degrees, and in differing combinations in any one situation. Scholars have also disputed which justifications should guide decisions about deference in international law. Some authors favour considerations of expertise, whereas others favour legitimate authority-based justifications (compare, for example: Makela, 2013, 347; Dyson, 2006, 107; Steyn, 2005, 352; Raine, 2013; Clayton, 2004; Gerards, 2004, 182–83; Young, 2010, 822). As these justifications have typically been identified based on analyses of deference in domestic systems, any analysis of international approaches to deference must also consider whether these justifications should be adapted as a basis for deference in international law or in specific international regimes.

19  The connection between deference and authority explains why deference can be treated as categorical (defer/not defer) or more flexible (more/less deference). The reason for these different structures of deference is that the second-order reasons underlying authority can be treated as exclusionary or non-exclusionary of the first-order reasons for decision. That is, deference to another actor’s authority might result in the available first-order reasons for decision cancelling out, being balanced alongside, or being displaced by second-order reasons prompting deference (see, further Shirlow, 2021). The distinction between exclusionary and non-exclusionary second-order reasons can be analogized to the difference between an order and a request. Where deference is treated as categorical, the recognition of another actor’s authority (eg expertise) ‘replace[s] or cut[s] off the [first actor’s] independent assessment of the merits of the case’ (Westlund, 2013, 458). Authority in this sense becomes akin to an ‘order’ to decide a matter in a particular way (defer/not defer). Deference in this structure causes one actor to disregard ‘at least some reasons which are … relevant and would in the absence of the exclusionary reason have sufficed to justify proceeding in some other way’ (Finnis, 2011, 234). Where deference is treated as more flexible, the recognition of another actor’s authority (eg legitimacy) causes the deferring actor’s judgment not to be ‘pre-empted, but only balanced against the reasons for deferring to the views of another’ (Soper, 2002, 45). Authority in this sense becomes akin to a ‘request’ to decide a matter in a particular way (more/less deference): ‘[c]ountervailing reasons may outweigh the request’ (Hershovitz, 2003, 204).

20  The structure of authority might interact with the bases for recognizing authority considered above. Some theorists, for example, consider that empirical epistemic authority (expertise) will always produce categorical assessments (an actor is either an expert or not), with normative epistemic authority (legitimacy) being viewed as less binary (an actor may be more or less legitimate) (Soper, 2002, 35–36). Other theorists consider authority to necessarily produce only categorical approaches (see especially: Raz, 1979, 22–25), whereas others adopt more flexible analyses for both types of instrumental authority (see for example Daly, 2012, 199). International actors themselves adopt different understandings of the structure of authority which influence their approaches to deference. This means that ‘second-order reasons prompting deference may exclude all or only some of the relevant first-order reasons or may instead operate in a non-exclusionary way alongside first-order reasons’ (Shirlow, 2021).

21  Deference can be used to settle authority overlaps between various types of actors in different structural relationships. This section explores the incidence of deference between international actors within the international legal system, between international and domestic actors, and in several other circumstances.

22  Deference may arise between international actors, including those operating within the same institution or field of international law, or across different areas of international law. International regimes have variously been characterized as part of a unified ‘system’, as a ‘set of self-contained regimes’, or even as components in a ‘loose agglomeration or anarchy’ (see for example Han, 2006, 102; Sands, 2002, 202). Regardless of the view adopted, ‘[t]he international legal terrain definitely is becoming more congested’ and ‘[w]ith congestion comes collision, and often friction’ (Trachtman, 2011, 1). As such, international actors are increasingly likely to confront situations in which their authority overlaps with that of another international actor in the same, or a different, institution or subject-field. In such situations, deference may be invoked as a means of managing the interactions between international actors. Deference, for instance, provides a means for achieving consistency and/or informal hierarchy between different actors within the one institution or subject-field of international law. This includes deference between appellate and first instance adjudicators in the same field or regime of international law (see, for example Oesch, 2003). Deference may be used by international actors to manage overlap and fragmentation across different institutions and subject-fields. Deference in such circumstances creates either vertical or horizontal relationships of acknowledgement or respect for decision-making authority between international actors (Fischer-Lescano and Teubner, 2004, 1039).

23  Deference might, for instance, inform how an international adjudicator uses the decisions of other international adjudicators (Judicial Precedent). Some international courts and tribunals, for example, have treated the decisions of other international courts or tribunals—particularly those of the International Court of Justice (‘ICJ’) on matters of general international law—with deference. The European Court of Human Rights (‘ECtHR’) in Jones, for example, noted that ‘[i]n recent years, both prior to and following the House of Lords judgment in the present case, a number of national jurisdictions have considered whether there is now a jus cogens exception to State immunity in civil claims against the State’ (Jones v United Kingdom, 2014 (‘Jones’), para 197). The court held, however, that: ‘it is not necessary for the Court to examine all of these developments in detail since the recent judgment of the ICJ in Germany v Italy … – which must be considered by this Court as authoritative as regards the content of customary international law – clearly establishes that, by February 2012, no jus cogens exception to State immunity had yet crystallised’ (Jones, 2014, para 198). Yet not all international actors elect to defer to others in this way. The investment tribunal in Electrabel, for example, noted that certain issues presented to it for decision were ‘arguably similar to those addressed’ in the decisions of other investment treaty tribunals (Electrabel SA v Hungary, 2012 (‘Electrabel’), para 4.14). The tribunal nonetheless noted that it had decided to approach the issues ‘anew, without any deferential regard to these earlier cases’ (Electrabel, 2012, para 4.15). Deference could alternatively prompt such an international adjudicator to adopt a ‘rebuttable presumption that the decisions of international regime courts do have the character of precedent for one another’ (Fischer-Lescano and Teubner, 2004, 1044).

24  Deference can also guide the resolution of jurisdictional overlap (see especially Shany, 2003). Such jurisdictional overlaps can occur between actors in different regimes, between actors of institutionally different types, and between actors within the same regime or institution. Jurisdictional overlaps between international courts and tribunals are, for instance, becoming increasingly common as a result of a proliferation of international dispute settlement mechanisms. In MOX Plant (MOX Plant Arbitration and Cases), for example, an Annex VII Arbitral Tribunal declined to exercise jurisdiction over a dispute between Ireland and the United Kingdom concerning various provisions of the United Nations Convention on the Law of the Sea (1982) (MOX Plant, Ireland v United Kingdom, 2003 (‘Mox Plant’), para 14). The tribunal there noted that ‘there is a real possibility that the European Court of Justice may be seised of the question whether the provisions of the Convention on which Ireland relies are matters in relation to which competence has been transferred to the European Community and, indeed, whether the exclusive jurisdiction of the European Court of Justice, with regard to Ireland and the United Kingdom as Member States of the European Community, extends to the interpretation and application of the Convention as such and in its entirety’ (MOX Plant, para 21). The tribunal noted that these matters ‘essentially concern the internal operation of a separate legal order (namely the legal order of the European Communities)’, and acknowledged a ‘risk’ that, ‘until these matters are definitively resolved, there remain substantial doubts whether the jurisdiction of the Tribunal can be firmly established in respect of all or any of the claims in dispute’ (MOX Plant, paras 24 and 25). As such, ‘and bearing in mind considerations of mutual respect and comity which should prevail between judicial institutions both of which may be called upon to determine rights and obligations as between two States’, the tribunal deemed it ‘inappropriate for it to proceed further … in the absence of a resolution of the problems referred to’ by the European Court of Justice (MOX Plant, para 28). The types of relationships that produce these overlaps will be important to determining whether deference will apply as they impact the relative assessment of authority detailed above. This might produce different deference relationships between actors of different institutional types. In Eastern Sugar, for example, the respondent—the Czech Republic—argued that the nature of the European Union’s legal order meant that ‘the Arbitral Tribunal should defer or refer to the opinion of EU authorities when deciding about its own jurisdiction’ (Eastern Sugar BV (Netherlands) v Czech Republic, 2007 (‘Eastern Sugar’), para 118). While the tribunal noted that a different result might be reached had EU judicial authorities considered the matter, it concluded that ‘the views of the European Commission … are not binding on this Arbitral Tribunal but, if clear, which they are not, would at best have persuasive force’ [emphasis in original] (Eastern Sugar, para 125).

25  Deference between international actors can also occur outside of adjudicative contexts (see, especially Pratt, 2018). Such deference where offered can serve to indicate one international actor’s acceptance of another international actor’s authority. As Pratt notes: ‘[d]eference is a common form of coordination among [international organizations]; its goal is to minimize areas of overlapping jurisdiction by assigning authority to a single organization’ (Pratt, 2018, 35).

26  International actors also frequently defer to actors within domestic systems. This includes via deference to domestic interpretations of law and domestic regulatory and policy decisions. As noted above, such deference may be achieved through several reasoning devices, including deferential standards of review (see for example Vadi, 2018; Henckels, 2014; Oesch, 2003) and margin of appreciation analysis (see for example Legg, 2012; Gerards, 2011; Shany 2005). Such terms describe particular approaches to deference that produce ‘a limitation in … [an international actor’s] level of scrutiny concerning decisions taken or determinations made by a [State] because the [international actor] respects the reasons for a state’s decision or conduct even if its own assessment was different’ (Schill, 2012, 582). The types of domestic decisions that might be deferred to by international actors are diverse: they may concern different subjects (eg legal, policy, or factual issues), have been made at different times (eg before, during, or after a matter being placed before an international actor for decision), or intersect in different ways with the matters before an international actor for decision (eg as context for a dispute, the cause of a dispute, or as an attempt to resolve a dispute). Several categories of domestic decision to which an international actor might defer are introduced and discussed in the paragraphs that follow.

27  International actors are often asked to defer to domestic interpretations and applications of domestic law, and/or domestic findings of fact. To the extent that an international actor is required to assess a matter of domestic law, for instance, it may decide that a domestic actor is better placed to interpret or apply that law due inter alia to its closer institutional position or better resourcing and expertise. In such circumstances, the international actor could afford deference to more authoritative domestic actors by treating as dispositive their findings of fact or interpretations of domestic law (see further on applicable law and deference Hepburn, 2017, 109; Lourie, 2015, 516; Watts, 1996, 425). An international actor might otherwise stay its own proceedings to have the benefit of a domestic determination of a given matter, or accept such determinations provided they are made through particular (expert or legitimate) procedures, or appear reasonable or not arbitrary (as opposed to correct). States frequently request international actors, for instance, to defer to their domestic judicial decisions or proceedings on both legal and factual matters. Such requests typically arise as part of international adjudicative proceedings in which domestic judicial decisions are either challenged directly as inconsistent with an international obligation of the State or are otherwise relevant in some way to an international claim.

28  Deference can arise when a domestic judicial decision is itself challenged through an international proceeding. Domestic judicial decisions and proceedings may be directly challenged before international actors assessing claims under customary international law (see especially Paulsson, 2007) or international treaties—including human rights (see for example Art 6 (right to a fair trial) European Convention on Human Rights and Fundamental Freedoms (1950)) or investment treaties (see especially Demirkol, 2018). Whether a domestic judicial decision can constitute a breach of international law is a matter of applicable law, the interpretation of which will supply the permissible first-order reasons for the international actor’s analysis. An international actor might nevertheless interpret an international obligation less stringently in order to offer deference to domestic decision-making authority. Deference might similarly affect an international actor’s analysis of whether a domestic judicial decision is consistent with international law. Often, this will result in statements from international adjudicators to the effect that they cannot substitute their own interpretation of domestic law for that of domestic courts, or that they should not otherwise act as appellate courts for matters of domestic law. The ITLOS Tribunal in Monte Confurco, for example, emphasized that in assessing the reasonableness of a bond required by a national court for the release of a vessel and its crew, ‘it is not an appellate forum against a decision of a national court’ (Monte Confurco, Seychelles v France, 2000, para 72). Frequently this follows from the international actor’s view that the domestic actor is the more legitimate or expert authority in relation to the interpretation or application of domestic law. As the Unglaube investment tribunal stated:

29  Deference might otherwise inform an international actor’s analysis of the relevance of a domestic judicial decision to one of the matters before it for decision. A domestic court might, for instance, already have determined matters of law relevant to the international claim. In Azinian, for example, an investment tribunal held that it had to defer to (and accept as dispositive) a domestic judicial decision that a contract was void under domestic law because there was no proof that the domestic judicial decisions were themselves inconsistent with international law (Azinian and others v Mexico, 1999, paras 99–100). The GAMI investment tribunal similarly undertook to ‘give respectful consideration’ to domestic judicial decisions ‘insofar as [the domestic court] applies norms congruent with those of [North American Free Trade Agreement]’ (GAMI Investments v Mexico, 2004, para 41). Such deference might follow from instrumental, prudential or non-instrumental second-order reasons. An international actor might, for instance, consider a domestic court a more expert and legitimate interpreter of domestic law. Applying such justifications, international actors have frequently deferred to domestic judicial interpretations of domestic law. The ECtHR in Former King of Greece, for example, observed that ‘it is in the first place for the domestic authorities, notably the courts, to interpret and apply the domestic law and to decide on issues of constitutionality’ (Former King of Greece and others v Greece, 2000, para 82; see similarly Sovtransavto Holding v Ukraine, 2002, para 95). To similar effect, the investment tribunal in Grand River declined to offer its own analysis of ‘delicate and complex questions of US constitutional and Indian law’ on the basis that ‘issues of national law belong in national courts, not in an international tribunal’ (Grand River Enterprises Six Nations Ltd and others v United States, 2011, para 234).

30  International actors might also offer deference to domestic findings of fact. As the ECtHR held in Tunc, for example:

The international investment tribunal in Rompetrol, adopted similar reasoning when it was called upon to assess an alleged interference with an investment through wiretapping. That tribunal noted that it found it ‘unnecessary to go into these questions in greater detail, as it is now in possession of an authoritative determination by the Romanian High Court of Cassation and Justice’ (Rompetrol Group NV v Romania, 2013 (‘Rompetrol’), para 256). It thereafter treated as conclusive the decisions of that court, including its findings that ‘national security was not the real motive for the interception but only a convenient device’, and that the claimant ‘had not substantiated its claim of interference with its own business operations’ (Rompetrol, paras 260–61). An international actor might use one or several of the above justifications to defer to domestic judicial decisions on matters of fact. Deference might follow, for instance, from a recognition by the international actor that domestic actors have the better claim to the authority needed to interpret or apply domestic law (see further Douglas, 2003, 273; Nollkaemper, 2011, 318; Jenks, 1938, 591). Deference might otherwise follow an assessment that a domestic actor has better proximity or resources to evaluate a given matter and so is more ‘expert’ than the international actor due to its institutional position or resources (Nollkaemper, 2011, 254). A domestic actor might alternatively be considered to have greater legitimacy to determine certain matters compared to an international actor. Such superior legitimacy might follow from the procedural aspects associated with domestic as compared to international hearings, including for instance because domestic proceedings are ‘ordinarily open to the public’ (Trakman, 2014, 168).

31  Where an international actor accepts that some deference is due to the decisions of domestic actors, the structure of authority underlying such deference will influence how deference manifests. As noted above, where deference is afforded to a domestic decision the international actor might treat it either as dispositive of a certain question or highly influential on that matter. The Emmis investment tribunal, for example, adopted the view that any application of domestic law by an international actor requires that actor to ‘seek to determine … the manner in which the law would be understood and applied by the municipal courts’ (Emmis International Holding and others v Hungary, 2014 (‘Emmis’), para 175). While it clarified that domestic court decisions would not be ‘necessarily dispositive of the question’, they were ‘likely to be of great help’ (Emmis, para 176). Deference can also inform the setting of presumptions and argumentative burdens. In Al Warraq, for example, Indonesia argued that ‘[t]he principle of comity alone requires this Tribunal to assume that the Indonesian Court has acted properly unless the Claimant proves that there has been a glaring disregard of due process by the Court’ (Al Warraq v Indonesia, 2014, para 405). Arbitrator Ali Khan in Deutsche Bank similarly adopted the perspective that domestic courts are entitled to a ‘high measure of deference’ such that ‘there should be a presumption that it has acted in a proper manner’ (Deutsche Bank AG v Sri Lanka, 2012, para 106).

32  Questions of deference might alternatively arise where domestic judicial decisions provide possible alternative avenues of redress for the matter brought before an international actor for decision. Certain international rules expressly address these potential overlaps between international and domestic judicial decision-making authorities. This includes rules entailing a requirement to exhaust or have recourse to local remedies or imposing fork-in-the-road requirements (see further Haesler, 1968; Uzun Marinkovic and Kamber, 2016). Some international rules specifically privilege the exercise of international jurisdiction in such circumstances. The Statute of the International Criminal Tribunal for the Former Yugoslavia (‘ICTY’), for instance, provided that:

The application of such rules does not as such indicate deferential reasoning; their application may simply reflect the adjudicator’s respect for the limits of its jurisdictional mandate (and as such, the absence of formal authority to determine a dispute). However, such rules may provide international adjudicators with discretion to incorporate elements of deference into their analysis. Moreover, in the absence of such rules, international actors have considerable discretion to use principles of deference to inform the sequencing of their proceedings vis-à-vis a domestic court’s. Each of the authority-based justifications for deference introduced above can guide such assessments. Waiting for a domestic court to issue a decision on a matter might, for instance allow an international actor to benefit from the expertise of the domestic court on the matters at issue. The Permanent Court of International Justice in Prince von Pless noted to this effect that:

33  Analysis of domestic public policy and regulatory decisions may also elicit deference from international actors. Such decisions can be relevant to the jurisdiction of an international actor or admissibility of a claim before them, to the substance of an international obligation, or to the application of justifications, carve-outs, exceptions, and excuses (see, for more detailed discussion: Shirlow, 2021). To the extent that an international actor is empowered to analyse a State’s regulatory or policy choices, they may offer different degrees of deference to State’s determinations on those matters. This might follow, for instance, from an assessment that a domestic—rather than international—actor has the better claim to the expertise or legitimacy necessary to identify or appraise public policy and regulatory choices.

34  An international actor might use deference to inform their identification of the policy justifications that will be relevant to their ultimate determination. Where an obligation permits a state to undertake conduct that pursues a ‘public purpose’, for instance, the international actor may exhibit deference in applying that obligation by recognizing the State’s superior authority vis-à-vis the identification of such purposes. The international actor might give a State either complete or at least some discretion to identify whether particular conduct has been motivated by a relevant public purpose (Stone Sweet and Grisel 2017, 247). The investment tribunal in Paushok, for instance, declined to second-guess a state’s claim to have acted for a public interest, on the basis that: ‘[t]he definition of public interest is one that varies considerably from one State to another and it is a subject of significant public debate within each State … [t]his is more a subject for political debate than arbitral decisions’ (Paushok v Mongolia, 2011, paras 328, 337). An international actor could otherwise exhibit deference to a State’s claim to have acted for a public purpose provided that such a claim does not appear to be manifestly unreasonable, abusive, or arbitrary. The Russia – Traffic in Transit panel, for example, recognized a broad discretion for WTO Member States to identify what interests constitute ‘essential security interests’ within the meaning of Article XXI General Agreement on Tariffs and Trade (1947) (‘GATT’), such that it would be ‘for every Member to define what it considers to be its essential security interests’, save that in doing so the Member must act in good faith (Russia – Traffic in Transit, Panel Report, 2019, paras 7.131-7.132). Concretely, this interpretation entails deference to a WTO Member State’s characterization of their essential security interests provided that the Member State ‘articulate[s] the essential security interests said to arise from the emergency in international relations sufficiently enough to demonstrate their veracity’ (Russia – Traffic in Transit, Panel Report, para 7.134).

35  In addition to deference in the identification of domestic regulatory and policy goals, international actors might also exhibit deference when setting and assessing the nexus required between those goals and the State’s measure (see further Mitchell and others, 2018, 17; Young, 2014, 378; Schauer, 1977, 26; Brems and Lavrysen, 2015, 140). An international actor could, for instance, exhibit deference by adopting an approach whereby the ‘[s]uitability or fitness of a measure will largely be taken for granted’ (Cottier and others, 2017, 666–67). The international actor could otherwise review the connection between the State’s conduct and its claimed public purposes more leniently. Deference might thus prompt an international actor not to assess the validity or correctness of a State’s claim to have acted for a public purpose, but instead to give the State some deference on this question and so only assess the reasonableness of plausibility of that claim. The Russia – Traffic in Transit panel, for instance, held that WTO Member States needed to ‘meet a minimum requirement of plausibility’ in demonstrating that a particular measure had been adopted to protect an essential security interest, such as to prove that any measure purportedly adopted under Article XXI GATT are ‘not implausible as measures protective of these interests’ (Russia – Traffic in Transit, Panel Report, para 7.138). Similarly, the Glamis investment tribunal, noted that:

36  Such deference can again follow each of the various justifications introduced above. It might follow, for example, from the international actor’s recognition that the State holds greater legitimacy than does the international actor to identify public purpose or craft responses to them. As the foregoing discussion indicates, deference will frequently intersect with the first-order reasons for decision, including any balancing tests applied by an international actor to assess a State’s regulatory or policy decisions (see further Henckels, 2018; Keller and Walther, 2018). Such tests indicate to the international actor that the State’s policy decisions may be relevant to their enquiry but in themselves they do not resolve the issue of deference. Deference can nevertheless influence the application of a balancing test, including to inform the selection of the matters to be balanced, or the appraisal of those matters in relation to other reasons for decision.

37  Deference is also invoked in international law in several other situations beyond those considered above. This includes deference to actors other than those considered above. International actors, for instance, might defer to private parties in international claims. This might include deference to a company’s business judgments (Micula and others v Romania, 2013, para 1138) or to a disputing party’s framing of its case or arguments (see for example the request by a claimant for such deference in Garanti Koza LLP v Turkmenistan, 2013, para 85). International actors might also offer deference to adjudicators in commercial or transnational settings including, for instance, to contractual arbitration tribunals (see for example SGS v Pakistan, 2003, para 73). International actors might also invoke principles of deference during treaty interpretation, including to defer to the treaty text (see for example Tokios Tokeles v Ukraine, 2004, para 25; Siag and Vecchi v Egypt, 2007, para 127), to note deference (flexibility) within the treaty text (see for example Aguas del Tunari v Bolivia, 2005, para 285; Caratube International Oil Company LLP v Kazakhstan, 2012, para 336), or to give a State some deference when interpreting the obligations contained in a treaty in order to leave the State greater regulatory space (see for example PNG Sustainable Development Program Ltd v Papua New Guinea, 2015, paras 177, 256–57).

Esmé Shirlow