Jump to Content Jump to Main Navigation
The Oxford Guide to Treaties edited by Hollis, Duncan B (1st July 2012)

s.III Treaty Application, 12 The Territorial Application of Treaties

Syméon Karagiannis

From: The Oxford Guide to Treaties

Edited By: Duncan B. Hollis

From: Oxford Public International Law (http://opil.ouplaw.com). (c) Oxford University Press, 2015. All Rights Reserved. Subscriber: null; date: 17 January 2020

Treaties, application — Treaties, scope (temporal and territorial) — State practice — Universal international organizations

(p. 304) (p. 305) 12  The Territorial Application of Treaties


Legal texts are traditionally (and conveniently) applied with the aid of some good old Latin expressions: ratione personae (personal application), ratione materiae (material application), ratione temporis (temporal application), and ratione loci (spatial application).1 Although treaties may raise all four issues, the law of treaties does not address them equally. Neither the 1969 Vienna Convention on the Law of Treaties (VCLT) nor its 1986 companion takes an express position on the ratione materiae or the ratione personae issues. What subjects a treaty covers instead becomes a matter of the interpretation method put down by the Vienna conventions. The ratione personae issue is seemingly more ambiguous. Asking which persons or entities within a State party to a treaty are bound by that treaty raises difficult questions about how municipal law incorporates treaties. The Vienna conventions cautiously avoid dictating any answers in light of the different approaches taken by national legal systems. In contrast, the VCLTs contain a large number of articles on a treaty’s ratione temporis application (entry into force, non-retroactivity, termination, suspension, and so forth). But for the ratione loci application of treaties, there is but one article. Article 29 of the VCLT (‘Territorial scope of treaties’) provides that ‘unless a different intention appears from the treaty or is otherwise established, a treaty is binding upon each party in respect of its entire territory’.2

Article 29 may not be the VCLT’s shortest provision, but it is close. The main principle (‘a treaty is binding upon each party in respect of its entire territory’) consists of only thirteen words. It is a principle, moreover, rarely encountered in treaty texts, corresponding as it so often does to a text’s ‘ordinary meaning’. Indeed, a State party that refuses to apply treaty provisions to the whole of its territory would usually not be deemed to act in ‘good faith’. The logic of Article 29 thus (p. 306) parallels Article 31’s rule of interpretation, which relies heavily on both the good faith and ‘ordinary meaning’ principles.3

What saves Article 29 from the most boring banality is its pairing of the main principle favouring integral territorial application with a derogation (‘unless a different intention appears from the treaty or is otherwise established’). Combining a legal principle with its opposite cannot but prove thrilling for scholars, puzzling for diplomats, and dramatic for judges. More positively, a provision that permits—almost—everything cannot but attain a high degree of consensus among States. Thus, Article 29 reflects in a certain way customary law in so far as customary law loves above all consensual and flexible solutions. Yet, this approach may also generate confusion. The Greek delegation to the International Law Commission (ILC) characterized it as ‘a refutable legal presumption’ and queried ‘whether the inclusion of such a provision is useful in a formal text’.4 The ILC, however, took the position that ‘a State’s territory plays such an essential role in the scope of the application of treaties that it is desirable to formulate a general rule on the matter’.5

This chapter examines the legal questions associated with the territorial application of treaties. It begins by reviewing Article 29’s travaux préparatoires and the sources of exceptions to its main principle of integral territorial application. State practice is then explored in two respects. First, the chapter explores how variations in how States organize themselves—whether through the presence of colonies or a federal form of government—led to particular clauses in treaties delimiting how they would apply to some or all of a State’s territory. Second, it examines the question of territorial boundaries as a precursor to a topic left unaddressed by the VCLT, namely the extraterritorial application of treaties, particularly in the human rights context. Finally, this chapter examines the difficulties associated with territorial application in the context of treaties involving international organizations (IOs).

I.  Article 29: The Curious Relationship between a Principle and a Derogation

Article 29’s actual text differs only slightly from the wording of the earlier 1966 ILC draft.6 The article was easily approved at the Vienna Conference, with no votes against and no abstentions. Such history might suggest that no particular difficulties (p. 307) arise under a ratione loci application of treaties. This is unfortunately not the case. Although not related here in detail, Article 29’s drafting history was actually lengthy and complicated, undoubtedly due to the political context of decolonization at the end of the 1950s.

The travaux préparatoires were further complicated due to the (British) nationality and legal experiences of the ILC Special Rapporteur on the law of treaties, Sir Gerald Fitzmaurice. In 1959, Fitzmaurice drafted a highly sophisticated system for the territorial scope of treaties. Most of his provisions—in four articles—favoured a strict dichotomy between the application of treaties to a State’s metropolitan territories and to its so-called ‘dependent territories’. Given its colonies, the United Kingdom was, along with some other European powers, most concerned by this discussion. Most of the other States (and non-western European members of the ILC) were less enthused, to the point that no territorial scope draft article was submitted to the UN General Assembly in these years. But even as the issue became blurred by a rather heavy political discussion on the metropolis/colony relationship, Fitzmaurice’s early effort hinted at the customary international law on this topic. His draft Article 25 provided that: ‘Unless a treaty otherwise provides, it applies automatically to the whole of the metropolitan territory (or to all territories forming part of the metropolitan territory) of each contracting party.’7

The next ILC Special Rapporteur, Waldock, simplified Fitzmaurice’s approach considerably, paving the way to the ILC’s adoption of its first draft article on this point.8 The ILC’s 1964 draft article submitted to the General Assembly stated that ‘the scope of application of a treaty extends to the entire territory of each party, unless the contrary appears from the treaty’.9 That language thus continued Fitzmaurice’s idea that territorial application should consist of both a principle and a corresponding exception. As already noted, such duality assured the provision the general consensus among States and its easy adoption at the Vienna Conference. Yet, this very duality (or ambiguity) might also undermine the provision’s credibility and effectiveness. For starters, there is the question of what comes first. Is it the principle (applying the treaty to the State’s whole territory) or the exception (a partial application of the treaty on the territory)?

Normally, exceptions are interpreted stricto sensu, which suggests that the principle of integral application should not be easily defeated. Yet, Article 29’s own drafting history reveals a continuing retreat from the principle applying a treaty to the whole of the State’s territory. Fitzmaurice’s draft had required the treaty itself to express an exception. Waldock’s draft extended exceptions to the more vague form of a ‘contrary intention’ expressed: (i) in the treaty, (ii) ‘from the circumstances of its conclusion or the statements of the parties’, or (iii) contained in a reservation.10 The more sources for the exception, the easier it is to establish. The 1964 ILC draft was even more general (‘unless the contrary appears from the treaty’).11 The final (p. 308) 1966 ILC draft, as well as the 1969 text, go even further.12 Not only do they double the material sources for the exception (whether derived from the treaty or ‘otherwise established’), they also place (symbolically?) the exception at the head of the sentence.13

How do Article 29’s exceptions to integral territorial application apply? The first source (‘appears from the treaty’) simply requires reference to the usual methods of treaty interpretation detailed in Articles 31 and 32 of the VCLT. The alternative source (‘otherwise established’) poses more difficulties. The term ‘otherwise’14 appears to permit methods for establishing a ‘different intention’ quite different from those envisaged in Articles 31 and 32. But it is by no means clear what these alternative methods are. There is little doubt that the possibilities for identifying a ‘different intention’ are numerous. Yet, we should not minimize the importance of the verb ‘establish’.15 Albeit ‘otherwise’, the ‘different intention’ still has to be clearly shown. Indeed, if this were not the case, the distinction between the principle and its exception would be thoroughly erased, which hardly seems a correct interpretation of Article 29’s phrasing.

Both sources of the exception, moreover, require establishing a ‘different intention’ from the default rule. Thus, as Waldock recognized in his 1964 commentary: ‘The territorial application of a treaty is essentially a question of the intention of the parties.’16 Admittedly, lawyers may not be particularly fond of the word ‘intention’. There is always a question of whose intention matters. It is difficult to prove that a collective ‘intention’ exists and, assuming this is impossible, there are suggestions that the ‘intention’ of the State party to the territory of which the treaty would apply should prevail over the ‘intention’ of other States parties to the same treaty.

Unfortunately, the case law and practice are not coherent. The United Kingdom has a practice of declaring in writing to the depositary to which, if any, of its overseas territories a treaty applies, which it regards as settling any territorial application question.17 In Application No 8873/80 (X v United Kingdom), the former European Commission of Human Rights based its solution on the mere ‘intention’ of the respondent UK Government not to consider the island of Guernsey as part of UK territory as far as some electoral rights were concerned.18 The same Commission had previously found, however, in Wiggins v United (p. 309) Kingdom, that there were not ‘any significant social or cultural differences between Guernsey and the United Kingdom’ which could have proved relevant for a differentiation between the ‘metropolis’ and this Channel island, at least as far as housing regulations on this island were concerned,19 which probably means the mere ‘intention’ of the UK Government did not prevail on this point. On the other hand, in French case law, the commissaire du gouvernement of the Conseil d’Etat stated, with respect to the application of a French-Australian extradition treaty to New Caledonia, that the Convention being silent on the point, Australia had the right to think that the Convention applied on the whole of the territory of an indivisible Republic such as France.20 The UN Secretary-General, as a depositary, has taken to circulating unilateral declarations on exclusions of territorial application, a practice it describes as not ‘inconsistent’ with Article 29 of the VCLT ‘since it may be considered that the constant practice of certain States (which still comprise “non-metropolitan” Territories) in respect of territorial application and the general absence of objections to such practices have “established a different intention” within the meaning of article 29’.21

II.  Territorial Application in State Practice

How a State organizes its government with respect to the territory under its control may, of course, vary widely. Two practices in particular, however, have had a lasting impact on questions of ratione loci—(i) the presence of colonial territory distinct from that of the metropolitan territory; and (ii) the establishment of federal systems of government, which may place constitutional constraints on the federal government’s ability to make and apply a treaty on matters within the competence of one or more sub-federal territorial governments. In both contexts, States have devised particular treaty clauses to delineate whether and how a treaty will apply. Although no longer used with as much frequency, these clauses remain relevant both for applying existing treaties and the continuing occasions where States seek a treaty text to overcome the default rule of integral territorial application.

A.  Colonial clauses

States endowed with colonies (now referred to as ‘overseas territories’) have not adopted a uniform internal attitude on whether to apply treaties to such territories. One can generally speak of a British22 (and similar Dutch) system and a distinct (p. 310) French one.23 The first model clearly distinguishes between the metropolis and colony with connotations that are both negative (the colony may not benefit from an advantageous treaty) and positive (the colony preserves autonomy vis-à-vis its metropolis). The French model, with some exceptions,24 favours a kind of creeping assimilation between metropolis and colony; the constitutional principle of égalité pushing quite naturally to such a solution.

Prior to decolonization, States regularly dealt with colonial territorial application issues. Occasionally, they expressly endorsed the general principle of integral territorial application to all territories for which a State was ‘internationally responsible’.25 Exceptions were allowed via what has come to be known as the ‘colonial clause’, which took various forms. Some treaties afforded a State with colonial territory the option to extend the treaty to those territories by notification at the time it signed or consented to be bound.26 In others, the treaty provided the colonial State the option to exclude colonial territories by similar notification.27 In a few instances, moreover, a treaty would specify by name covered (or excluded) territories.28

With decolonization and the accompanying intense diplomatic and even military fighting against colonial systems, controversy enveloped the idea of allowing the remaining colonial powers the discretion to dictate whether international treaties applied to one or more of their colonies. Trying not to be politically incorrect, the ILC only stated that it preferred to avoid referring to expressions such as ‘territories for which the parties are internationally responsible’ (previously found in the Fitzmaurice and Waldock drafts) to which treaties could be extended.29 Although not necessarily warranted, the prevailing sentiment viewed use of colonial clauses as an endorsement of the colonial system itself.30

(p. 311) Of course, just because the VCLT did not address the colonial clause does not mean it is absent from international treaty-making. For starters, many existing treaties (particularly regional European treaties) still contain colonial clauses. Thus, most (if not all) Council of Europe (COE) treaties do so. Article 56(1) of the European Convention on Human Rights (ECHR) is a paradigmatic example, providing that ‘any State may at the time of its ratification or at any time thereafter declare by notification addressed to the Secretary General of the Council of Europe that the present Convention shall … extend to all or any of the territories for whose international relations it is responsible’.31 In Tyrer v United Kingdom, the European Court of Human Rights (ECtHR) emphasized that Article 56 ‘was primarily designed to meet the fact that, when the Convention was drafted, there were still certain colonial territories whose state of civilization did not, it was thought, permit the full application of the Convention’.32 But even after the 1960s collapse of European colonial empires, the clause still garners attention. In the post-colonial era, those territories for which a State party is responsible are more likely to have historical, geographic, and cultural ties to the State, a point emphasized by the Court in its Tyrer judgment which analyses the ECHR’s application to the Isle of Man. Nevertheless, efforts to delete Article 56 seem to have met strong resistance from some European States.

Today, the colonial clause remains controversial. For some, it marks an unfortunate mechanism for precluding the universal application of treaty rights (especially human rights). Thus, modern UN human rights treaties do not contain a colonial clause or otherwise allow the protection of fundamental rights to be different depending on the individual’s place of residence (whether in a metropolis or overseas territory). On the other hand, some colonial clauses continue to facilitate respect for regional and local differentiation. For example, where a State extends the ECHR to an overseas territory, Article 56(3) provides that ‘the provisions of this Convention shall be applied in such territories with due regard, however, to local requirements’. This ‘local requirements’ (‘necessités locales’) clause allows further discrimination of the Convention’s application overseas. To apply, however, the ‘local requirements’ have to be evidenced beyond doubt. That has proven neither an easy nor frequent occurrence before the ECtHR; the Tyrer case, for example, declined to find it applicable to the Isle of Man’s governmental use of corporal punishment. In Py v France, the Court did make an adjustment to permit a New Caledonian restriction on French citizens who had not resided there for a sufficiently lengthy time to elect the members of the local legislature; a restriction designed to appease strained relations between competing political and ethnic (p. 312) groups.33 In this sense, ‘colonial clauses’ do not stand unavoidably against overseas inhabitants and their human rights; they may be seen as vehicles for preserving local traditions and minority cultures for overseas societies, albeit perhaps at the expense of individual human rights.34

The sometimes heavy heritage of the colonial clause is also found in the two treaties on the European Union. Article 355 of the Treaty on the Functioning of the European Union (TFEU) divides territories under the sovereignty of the Union’s member States into a surprisingly large number of categories. Of course, to most of them, situated in Europe itself, the provisions of the Treaty apply completely. But there are some others to which the Treaty has never applied, for example the Faeroe Islands, under Danish sovereignty.35 To a large number of overseas territories, globally called ‘overseas countries and territories’, and listed in Annex II, applies a complex system of special arrangements amounting to a kind of association of these territories to the Union.36 Article 198 details that the ‘association shall serve primarily to further the interests and prosperity of the inhabitants of these countries and territories in order to lead them to the economic, social and cultural development to which they aspire’.37 Member States, moreover, cannot unilaterally or in groups change the rules and procedures regarding this association; unanimity of the Council of Ministers is required by Article 198. At least officially, therefore, this kind of colonial clause does not purport to discriminate negatively against ‘overseas countries and territories’. On the contrary, this status differentiation could be seen as a kind of ‘affirmative action’.

A similar approach is taken with respect to French overseas departments in the Caribbean region as well as to the Island of the Réunion in the Indian Ocean. Originally, the treaties of the European Communities as a whole were supposed to eventually apply to these departments. Local authorities, however, resisted. The (p. 313) Court of Justice of the European Union condemned France for not complying with its Community obligations because of, inter alia, the non-abolition of some local taxes on all goods not originating from these departments upon their entry.38 Eventually, however, rather than implementing EU law, local resistance in the French overseas departments brought about a substantial modification of the application of EU law to these departments. Article 349 TFEU, to which Article 355 refers, enables the EU Council of Ministers to adopt specific measures laying down conditions for the application of the EU Treaties to overseas departments, including common policies, in areas including customs, fishery, and agricultural policy. In doing so, the Council must take into account the overseas departments’ ‘structural social and economic situation … which is compounded by their remoteness, insularity, small size, difficult topography and climate, economic dependence on a few products, the permanence and combination of which severely restrain their development’. At the same time, this adaptation to local situations is not to result in an ‘undermining [of] the integrity and [of] the coherence of the Union legal order, including the internal market and common policies’.39 Today, the four traditional French overseas departments (Réunion, Guadeloupe, Martinique, and Guiana) have been joined in Article 349 by additional territories.40

Thus, in these more recent clauses, what one could still call a ‘colonial clause’ has become a vehicle for adjusting whether and how treaties apply to the people concerned and their elected officials. Although integral territorial application remains the default rule, treaties may still authorize a State to extend (or exclude) the treaty’s application to its overseas territories. Today, however, such an extension (or exclusion) is more often done with the involvement of the peoples concerned.41 Thus, the clause’s inclusion is less controversial, although tensions remain with its relationship with treaties that purport to declare universal rights or responsibilities.

B.  Federal clauses

Besides States endowed with overseas territories, federal States also present difficulties as far as the territorial application of treaties. The crux of the problem is that international law (with some exceptions discussed in Chapter 5) enables only federal States to enter into international treaties. But many federal States’ (p. 314) constitutions entitle their sub-federal governmental units (states, Länder, regions, provinces, or cantons) to implement—either exclusively or concurrently—the measures necessary for the federal State to comply with a treaty. In important areas, such as criminal law, human rights law, environmental law, and civil law, the federal State may simply lack domestic legal authority to apply treaty provisions in the absence of implementation by the sub-federal unit that the federal State may have no authority to require.42 Federal States thus face difficult choices in deciding whether to consent to treaties in these areas. One alternative would be for the federal State to abstain from entering into the treaty entirely. Such a ‘solution’ is highly impractical due to the fact that some of the most important States in the world have a federal form of government, while others increasingly favour it nowadays to end internal strife.43 A second alternative is for the federal State and its sub-federal units to resolve the situation internally through agreed procedures, the most notable example of which is undoubtedly the 1957 Lindau Agreement concluded by the Länder and the German Federal Government, providing that in the case of treaties affecting the Länder, the latter must give their consent (and not merely their opinion) before the Federation can validly enter into such a treaty.

A third option lies in adjusting the treaty obligations for federal States.44 Expressing a non-federal or even anti-federal view, Fitzmaurice’s 1959 report on treaties had suggested that:

the constituent states, provinces or parts of a federal union or federation, notwithstanding such local autonomy as they may possess under the constitution of the union or federation, are considered to be part of its metropolitan territory for treaty and other international purposes.45

Unlike the more general question of territorial application, the VCLT never endorsed any provision on federal States. In practice, however, treaty-makers (p. 315) have devised a ‘federal clause’ to avoid the strains created by the confrontation between international and federal national law.46

The principal example of a ‘federal clause’ remains the Constitution of the International Labour Organization. Article 19(7)(b) provides that the federal State’s obligations shall be the same as those of non-federal States for matters appropriate for federal action under the federal State’s Constitution. For matters ‘appropriate under its constitutional system, in whole or in part, for action by the constituent’ units of the federal State, the federal State obligations are limited to referring them to the appropriate sub-federal units ‘for the enactment of legislation or other action’.47 The clause, with slight adaptations, has been repeated in a number of other major multilateral agreements, including the New York Convention on Arbitration and the 1951 Convention Relating to the Status of Refugees.48 A more recent example of a federal clause appears in the COE’s Convention on Cybercrime, authorizing a federal State to avoid the treaty’s criminalization obligations ‘consistent with its fundamental principles governing the relationship between its central government and constituent States’ provided it does not ‘exclude or substantially diminish’ those criminalization obligations.49

A variation of the federal clause—known as a ‘territorial clause’—appears in a variety of commercial and so-called private international law treaties. Thus, Article 93(1) of the 1980 UN Convention on Contracts for the International Sale of Goods (CISG) provides that:

If a Contracting State has two or more territorial units in which, according to its constitution, different systems of law are applicable in relation to matters dealt with in this Convention, it may, at the time of signature, ratification, acceptance, approval or accession, declare that this Convention is to extend to all its territorial units or only to one or more of them, and may amend its declaration by submitting another declaration at any time.50

(p. 316) As written, the clause could apply to political sub-divisions of a metropolitan territory, although in practice the clause is most often invoked by federal States such as Canada.51

In recent years, there seems to be less enthusiasm for federal clauses, especially regarding treaties on human rights. The objections centre on allowing federal States to assume different (and fewer) obligations than non-federal States, especially where human rights treaties are designed to establish universal minimum standards. Some treaty texts now affirmatively oppose differential treatment for federal States. A typical ‘anti-federal clause’ is found in Article 50 of the International Covenant on Civil and Political Rights (ICCPR): ‘The provisions of the present Covenant shall extend to all parts of federal States without any limitations or exceptions.’52 Article 28(2) of the American Convention on Human Rights is less brutal, but also more ambiguous:

With respect to the provisions over whose subject matter the constituent units of the federal State have jurisdiction, the national government shall immediately take suitable measures, in accordance with its constitution and its laws, to the end that the competent authorities of the constituent units may adopt appropriate provisions for the fulfillment of this Convention.53

In one of the few judgments of the Inter-American Court of Human Rights on this provision, the Court described it as a federal clause, before asserting that, as a matter of estoppel, Argentina had always behaved ‘as if the federal State had jurisdiction over human rights matters’ although the matter concerned exclusively the attitude of the local police in the Argentinian province of Mendoza.54 Indeed, quite often, federal tribunals have developed rather a bold (and much discussed) case law in order to bridge the gap between federal and international law.55

(p. 317) III.  The Concept of ‘Territory’ in the Territorial Application of Treaties

A.  Treaty application to a State’s ‘territory’

The foregoing analysis focused on territorial application in juridical terms: deciding if a treaty applies to all of a State’s territorial units or only to some of them. But even where a treaty clearly applies to a designated territory, two questions still remain. First of all, it is important to differentiate territorial application from the material application of treaty provisions that happen to relate to territory. Waldock, in particular, emphasized this distinction:

[S]ometimes the provisions of a treaty expressly relate to a particular territory or area, e.g. the Antarctic Treaty; and in that event the territory or area in question is undoubtedly the object to which the treaty applies. But this is not what the territorial application of a treaty really signifies, nor in such a case is the application of the treaty confined to the particular territory or area. The ‘territorial application’ of a treaty signifies the territories which the parties have purported to bind by the treaty and which, therefore, are the territories affected by the rights and obligations set up by the treaty. Thus, although the enjoyment of the rights and the performance of the obligations contained in a treaty may be localized in a particular territory or area, as in the case of Antarctica, it is the territories with respect to which each party contracted in entering into the treaty which determine its territorial scope.56

Second, understanding a treaty’s territorial application requires understanding what the concept of ‘territory’ means with respect to a State’s consent, regardless of whether the State is consenting to apply the treaty to ‘its entire territory’ or only to some specific parts of it.

For his part, Waldock simply relied on authorities stating that the territorial scope of a treaty coincides with territory under sovereignty of States parties to the treaty.57 But the contemporary concept of ‘territory’ is a more composite one. In its 1966 commentary to what became Article 29, the ILC defined ‘the entire territory of each party’ to include ‘all the land and appurtenant territorial waters and air space which constitute the territory of the State’.58 This definition, however, does not resolve the issue. Neither the terms ‘territorial waters’ nor ‘appurtenant’ were used in previous (or later) treaties on the law of the sea. Moreover, ‘appurtenant’ seems also likely to instill confusion as it can be interpreted to give States the right to determine which waters belong to them. Looking to the law of the sea, however, it is possible to equate ‘appurtenant’ to the idea of waters belonging to or coming under a State’s sovereignty, namely internal waters, territorial seas, as well as the archipelagic waters of archipelagic States.59

(p. 318) Whether the additional maritime zones beyond the territorial sea are ‘appurtenant’ to coastal States so as to comprise part of a State’s territory is a more difficult question. The exclusive economic zone (EEZ), which post-dated the VCLT, is neither a part of the high seas nor a zone under coastal State sovereignty. Its elusive legal nature is underscored by Article 59 of the 1982 UN Convention on the Law of the Sea (UNCLOS), which suggests conflicts over the EEZ not specifically regulated by the Convention ‘should be resolved on the basis of equity and in the light of all the relevant circumstances, taking into account the respective importance of the interests involved to the parties as well as to the international community as a whole’.60 The most that one can say at present is that, depending on the circumstances, the EEZ may or may not be ‘territory’ under Article 29 of the VCLT.

In contrast to the EEZ, a coastal State’s rights in its contiguous zone (police surveillance powers) and continental shelf (economic rights) are rather devoid of ambiguity. With respect to the continental shelf, the Netherlands proposed that the ILC revise the territorial application provision to include ‘the entire territory of each party, and beyond it as far as the jurisdiction of the State extends under international law, unless the contrary appears from the treaty’.61 Waldock accepted this suggestion by adding a second paragraph to his draft article: ‘a treaty may apply also in areas outside the territories of any of the parties in relation to matters which are within their competence with respect to those areas if it appears from the treaty that such application is intended’.62 Yet, despite additional US support,63 the Commission rejected this adjunction. The final word on this subject came from Senjin Tsuruoka, who suggested that the clause ‘unless the contrary appears from the treaty’ contained in the ILC’s draft ‘should be interpreted fairly broadly, in the positive as well as the negative sense, so that it was understood that the treaty—if its object so required or the intention was clear—was applicable outside the territory of the parties’.64

B.  Extraterritorial application of treaties

Tsuruoka’s observation suggests a certain value in applying some treaties beyond their States parties’ territories. Still, formally, Article 29 abstains from addressing the question of extraterritorial application of treaties. Both its history and text focus on choosing between applying the treaty to the entire territory of a State party or to a part of it. As worded, the provision does not easily allow going beyond the ‘entire territory’ concept; it reads as if one may do less but not more. For its part, the (p. 319) ILC—without rejecting the idea of extraterritoriality—made clear that its draft article which became Article 29 avoided that topic:

[A]rticle [29] was intended by the Commission to deal only with the limited topic of the application of a treaty to the territory of the respective parties … In its view, the law regarding the extra-territorial application of treaties could not be stated simply in terms of the intention of the parties or of a presumption as to their intention; and it considered that to attempt to deal with all the delicate problems of extra-territorial competence in the present article would be inappropriate and inadvisable.65

Quite clearly, this commentary does not exclude applying a treaty in some extraterritorial way. It merely excludes Article 29 from governing such an application, leaving the issue to the law of treaties that lies beyond the VCLT.

But what does it mean for a treaty to apply extraterritorially? Some treaties appear to apply irrespective of territorial boundaries (such as the UN Charter’s prohibitions on the unlawful use of force). For our purposes, the concept may actually mean either: (i) that a treaty will apply to territory, ‘terrestrial’ or not, lying outside the sovereignty of any State or (ii) that it will apply to territory lying under the sovereignty of another State.

The first possibility involves applying treaties to maritime zones outside the limits of the territorial sea.66 We will not deal again with the contiguous zone, the continental shelf, or the EEZ. The big issue consists in the application of treaties on the high seas. Here, international customary law (and not Article 29 of the VCLT) allows (and more and more often obliges) States to apply treaties on ships flying their flags. As a normative matter, this is an important way to ensure the ‘rule of law’ on the high seas. But as a descriptive matter, it is no longer acceptable to describe ships as ‘floating territory’. By applying treaty provisions to a ship on the high seas, the flag State applies them to its ship and not to the high seas themselves. But is it ever possible to forget that, in this case, the ship is really located on the high seas?67 We see therefore that a treaty may geographically apply to a territory even if the latter is not under the territorial sovereignty of the State party to it. The case is quite similar to the application of treaties to territories commonly thought not to appertain to any State. The best example for this is Antarctica, even though seven States continue to raise territorial claims to some portions of the frozen continent.

The second understanding of extraterritoriality raises undoubtedly more ‘delicate’ problems. Applying a treaty to a portion (or even worse, to the whole) of another State’s territory runs counter to a fundamental principle of sovereignty and the territorial integrity of States.68 And States are rarely eager to accept (and (p. 320) even less so to claim openly) that a treaty binding on them applies on territory outside of their own national territory. Ultimately, the issue may be one of treaty interpretation.

In recent years, interpretative disputes have arisen in particular with respect to the extraterritorial application of human rights treaties.69 The ICCPR, for example, provides that ‘each State Party to the present Covenant undertakes to respect and to ensure to all individuals within its territory and subject to its jurisdiction the rights recognized in the present Covenant’.70 In General Comment 31, the ICCPR’s Human Rights Committee (HRC) interpreted this provision as a ‘disjunctive conjunction’, allowing the ICCPR to bind a State party not just for acts within its territory but also in other areas subject to its jurisdiction, which the Committee defined to cover ‘anyone within the power or effective control of that State Party’.71 Others, notably the United States and the Netherlands, however, continue to insist that the text of Article 2 and the accompanying travaux préparatoires limit the ICCPR’s application to the territory of a State party.72

Resolving such interpretative disputes (let alone the issues of State responsibility that they may generate) is difficult absent the consent of the concerned State(s) to judicial review. Several human rights treaties, however, include such consent. The most abundant (and sometimes spectacular) case law on extraterritorial application is that of the ECtHR. The Court (and before it the European Commission of Human Rights) have repeatedly said that a State party to the ECHR is accountable for acts contrary to its conventional commitments even if materially these acts take place outside the territory of that State. The Court bases this interpretation on Article 1 of the ECHR: ‘The High Contracting Parties shall secure to everyone within their jurisdiction the rights and freedoms defined in Section I of this (p. 321) Convention.’ Unlike the ICCPR, there is no territorial limitation in Article 1, meaning not only that the ECHR might extend to acts committed outside the State’s territory but, conversely, acts committed inside the said territory might not trigger the State’s ECHR obligations if it has no ‘jurisdiction’ in a given case.73

This rather surprising aspect of the ‘jurisdiction’ theory is fully admissible in case, for instance, the State has no (more) authority on the territory in which the act contrary to the Convention takes place (provided of course that no agent of the State is involved). This would apply most naturally to instances of military occupation.74 For example, during the allied occupation of Berlin, it has been argued that West Germany was not responsible for what was happening there.75 The ECtHR’s own case law, however, has not fully endorsed this view. In Ilaşcu and Others v Moldova and Russia,76 the Court held Moldova accountable for acts in Transdniestria during a period in which its government did not control this secessionist territory (and which the Court found was almost occupied by Russia). The Court acknowledged Moldova’s lack of de facto authority, but held that the jurisdictional requirement of Article 1 required it to ‘endeavour, with all the legal and diplomatic means available to it vis-à-vis foreign States and international organisations, to continue to guarantee the enjoyment of the rights and freedoms defined in the Convention’.77 That finding may explain why States subsequently sought to reserve to any ECHR obligations for territory under foreign occupation or controlled by secessionist forces. Upon ratifying additional Protocol No 7, for example, Azerbaijan declared that ‘it is unable to guarantee the application of the provisions of the Protocol in the territories occupied by the Republic of Armenia until these territories are liberated from that occupation’.78

(p. 322) Still, the genuine extraterritorial effect of the ‘jurisdiction’ theory arises vis-à-vis the occupying or controlling power.79 The classic statement of this is found in the ECtHR’s Loizidou v Turkey judgment (Preliminary Objections): ‘although Article 1 sets limits on the reach of the Convention, the concept of “jurisdiction” under this provision is not restricted to the national territory of the High Contracting Parties’.80 The Court defined this jurisdiction such that ‘the responsibility of a Contracting Party may also arise when as a consequence of military action—whether lawful or unlawful—it exercises effective control of an area outside its national territory’.81 Yet, what has been celebrated (or criticized) as a bold extension of the ECHR’s scope created new problems.

In the first place, the ‘effective control of an area’ test is not always easy to handle. Whereas, for example, some 35,000 Turkish soldiers may be deemed more than enough to secure effective control over Northern Cyprus (with a population of 150,000 to which must be added some 100,000 Turkish settlers), it may be more difficult to ascertain the same as far as Transdniestria is concerned (with just 1,500 Russian soldiers at the time of the Ilaşcu case). This being said, a territory ‘just’ bombed by a State party’s war planes is not, according to the famous Banković decision, a territory under the effective control of that State.82 The control of air space apparently does not amount to territorial control proper for purposes of establishing the State’s jurisdiction. Obviously (and perhaps due to the heavy political consequences?), Banković does not fit with the Court’s Issa judgment, according to which ‘Article 1 of the Convention cannot be interpreted so as to allow a State party to perpetrate violations of the Convention on the territory of another State, which it could not perpetrate on its own territory’.83 Most recently, the Court attempted to reconcile its earlier case law in Al-Skeini and Others v UK. There, the ECtHR interpreted Article 1 to apply to the killing of an Iraqi national in a British military prison in Iraq as well as the killing of five individuals (p. 323) by British troops in the streets of the UK-occupied Basra since the ‘United Kingdom (together with the United States) assumed in Iraq the exercise of some of the public powers normally to be exercised by a sovereign government’.84

The Loizidou test and its progeny bear some similarity to the effective/overall control test used to attribute responsibility to a foreign State for the acts of groups of individuals, a test that has proved controversial in cases ranging from the ICJ’s Nicaragua judgment85 to the International Criminal Tribunal for former Yugoslavia’s (ICTY) Tadić judgment.86 In Loizidou (as well as Al-Skeini), the discussion was not about control over a group whose acts might be attributed to a State, but about a foreign State’s control over territory. From this point of view, Loizidou proves quite restrictive in terms of ascribing to a State jurisdiction.

The Court has emphasized, in any case, it is only ‘in exceptional circumstances [that] the acts of Contracting States performed outside their territory or which produce effects there (“extra-territorial act”) may amount to exercise by them of their jurisdiction within the meaning of Article 1 of the Convention’.87 More often, the Court will find it has no competence to examine complaints under Article 1 even where a State’s acts have extraterritorial effects. A good example of this was the Ben El Mahi and Others v Denmark decision.88 There, the applicants were Moroccan residents and Moroccan associations operating in Morocco with no jurisdictional link whatsoever with Denmark. As a result, the Court found they could not validly challenge Denmark’s decision to allow publication of cartoons on Prophet Muhammad in Denmark. Even if the concept of ‘jurisdiction’ within the meaning of Article 1 is not necessarily restricted to the national territory, the Court reaffirmed in Issa, that:

[F]rom the standpoint of public international law, the words ‘within their jurisdiction’ in Article 1 of the Convention must be understood to mean that a State’s jurisdictional competence is primarily territorial but also that jurisdiction is presumed to be exercised normally throughout the State’s territory.89

Yet, the control of territory test is not always of paramount importance to the Strasbourg Court’s definition of Article 1 jurisdiction. In addition to defining ‘jurisdiction’ on a territorial/extraterritorial basis, the Court has a separate line of cases that envision it in more personal terms. As Issa puts it, ‘moreover, a State may also be held accountable for violation of the Convention rights and freedoms of persons who are in the territory of another State but who are found to be under the former State’s authority and control through its agents operating—whether (p. 324) lawfully or unlawfully—in the latter State’.90 Policemen, diplomats, secret agents, and so on can then fully engage a State’s ECHR commitments through either their acts or even their omissions. A highly sensitive topic, which is not always void of some confusion, the European case law on extraterritoriality will probably need some great clarification in the coming years.91

IV.  The Territorial Scope of Treaties Concluded between States and International Organizations or between International Organizations

Article 29 of the 1986 VCLT is strikingly similar to Article 29 of the 1969 VCLT. It reads: ‘unless a different intention appears from the treaty or is otherwise established, a treaty between one or more States and one or more international organizations is binding upon each State party in respect of its entire territory’.92 The main principle of integral territorial application for States is thus overtly reaffirmed as well as the bases for derogations. In doing so, the article raises many (if not all) of the same issues discussed above. One might thus ask if this provision was necessary at all. Members of the ILC felt a moral obligation to not introduce any substantial changes from the 1969 VCLT in drafting this treaty. Special Rapporteur Paul Reuter had originally reproduced the 1969 formula literally word by word, although that text had little obvious connection to the conclusion of treaties involving international organizations (IOs).93 This oddity was corrected by the ILC, when N Ushakov, proposed to add the words ‘between one or more States and one or more international organizations’ to the draft text.94 This suggestion was easily accepted and the formula suffered no more changes.

(p. 325) Despite the satisfactory resolution of a text for Article 29 of the 1986 VCLT, two lacunae still render it somewhat enigmatic. They correspond to the two hypotheses on which the 1986 VLCT is built: (i) the conclusion of a treaty by an IO with a State and (ii) the conclusion of a treaty by an IO with another IO. In the first case, Article 29 allows for unequal treatment between the two contracting parties since the application ratione loci concerns only the State and not at all IOs. In the second case, no discrimination whatsoever takes place, but if one sticks to the text itself, the ratione loci application issue is irrelevant. Among IOs, territorial scope appears to have nothing to do with their treaty commitments.

A prima facie reaction might logically attribute the wording of Article 29 of the 1986 VCLT to a lack of adequate attention by its authors. Behind this ‘clumsiness’ explanation, however, lies a more nefarious possibility. Article 29 of the 1986 VCLT suggests that what is actually important when two IOs enter into a treaty is not the IOs themselves but rather their member States, to the whole of the territory of which the treaty will normally apply. Under this reading, IOs become thoroughly transparent, almost non-existent, in so far as, with respect to the territorial application of their treaties, the only thing to behold is their member States’ territory. Ultimately, not only the wording (which could have probably been less awkward) but also the global philosophy of Article 29 of the 1986 VCLT are likely to fuel arguments about the very legal existence of IOs. Assuming territorial application of treaties is an important matter (and one can hardly deny this), IOs seem to just be organs of coordination of the international efforts of their respective member States.

In any event, it may be seriously asked why the ILC (and later on the 1986 Conference) took such a stance, especially where it flies in the face of the 1986 Convention’s other provisions. Why confirm IO treaty-making powers95 if one merely wishes to show that IOs are artificial legal constructions (as the Neapolitan school of law around Rolando Quadri or the Soviet one around Grigory Tunkin thought it at a time)? Still, behind this conceptual difficulty lie two additional questions: (1) May an IO ever claim to possess a territory of its own? (2) In the negative, may the territories of IO member States taken together be deemed to constitute the territory of the organization?

The second question is easier to answer. Even IOs that can boast very significant legal powers abstain from considering the territory of their member States as their own. The EU provides a telling example. Article 52 of the Treaty on European Union and Article 355 TFEU only indicate the territories of the Union’s member States to which those treaties apply; they by no means indicate that these territories are territories of the EU itself. Speaking of a Union territory (customs territory, for instance) as some scholars do is only a metaphorical extrapolation.

As regards the other question, one may usually think that territory is closely related to the concept of statehood (along with a population and political power). Just as an organization cannot possess a population of its own, in the same way it (p. 326) generally cannot possess a territory. Nevertheless, there are some marginal cases where an IO possesses territory, albeit temporarily and not in an absolute but rather a functional way. This applies where organizations, most of the time the UN, exercise de jure (or sometimes both de jure and de facto) sovereignty on a given territory. A good historical example was the Council of the United Nations for Namibia, created by General Assembly Resolution 2248 (XXII) of 19 May 1967, which represented (until Namibia’s independence) something between an IO (it was a subsidiary body of the UN) and a State. Some multilateral conventions in fact dealt with the Council as if it were a State.96 One may also wonder if the UN or, more exactly, UN subsidiary organs responsible for State-building in a given area (East Timor, Kosovo) play a similar role and might therefore be considered as having a kind of territory during the transition period.97 Of course, the ephemeral aspect of State-building operations has nothing to do with the permanence of State sovereignty on State territory. We must not forget that IOs are definitely not States and the territory on which they sometimes exercise their jurisdiction cannot generally be equated to a State’s territory.


Altogether, the simplicity of the territorial scope of treaties seems to be only apparent. Such sensitive issues as the application of treaties to territories largely autonomous and lying most of the time overseas or the distribution of power (as far as the application of treaties is concerned) between federal States and sub-federal entities have only slightly and somewhat superficially been dealt with by international law. Definitely, it is not easy to say whether Article 29 of the VCLT really deals with them at all, and at the same time diplomatic practice and international or national case law are far from being homogeneous. The mere definition of the components of the State territory to which treaties apply is likely to become of some inextricable complexity (especially at sea). A careful drafting of territorial clauses of treaties could be a kind of remedy. Still, the most sensitive matter concerns not so much the diminution of the State territory to which a treaty applies but the perspective of considerably enlarging the scope of treaties to what (p. 327) lies outside the national territory, that is, parts of our planet free of any State sovereignty or parts of territories of third States. In particular in the field of modern human rights case law, the perspective of substituting the notion of ‘jurisdiction’ for one of territorial application leads unavoidably to the accountability of States for acts and omissions of their agents taking place outside the borders of national territory. One still cannot see up to what point exactly this rather strong extraterritoriality movement is likely to lead. Despite some precautions they can take (careful drafting of the treaty, appropriate reservations, declarations, and so on), States seem to be somewhat defenceless. It would be disastrous if they come to consider that the only solution is for them to be less and less committed.

Recommended Reading

  • SH Cleveland, ‘Embedded International Law and the Constitution Abroad’ (2010) 110 Columbia L Rev 225
  • F Coomans and M Kamminga (eds), Extraterritorial Application of Human Rights Treaties (Intersentia, Antwerp–Oxford 2004)
  • MJ Dennis, ‘Application of Human Rights Treaties Extraterritorially in Times of Armed Conflict and Military Occupation’ (2005) 99 AJIL 119
  • M Gibney and S Skogly (eds), Universal Human Rights and Extraterritorial Obligations (U Penn Press, Philadelphia 2010)
  • M Gondek, ‘Extraterritorial Application of the European Convention on Human Rights: Territorial Focus in the Age of Globalization?’ (2005) 52 Netherlands Intl L Rev 349
  • DB Hollis, ‘Executive Federalism; Forging New Federalism Constraints on the Treaty Power’ (2006) 79 S Cal L Rev 1327
  • S Karagiannis, ‘L’aménagement des droits de l’homme outre-mer: la clause des “nécessités locales” de la Convention européenne’ (1995) 28 RBDI 224
  • S Karagiannis, ‘Le territoire d’application de la Convention européenne des droits de l’homme. Vetera et nova’ (2005) 16 Rev trim des droits de l’homme 33
  • B Knoll, The Legal Status of Territories Subject to Administration by International Organisations (CUP, New York 2008)
  • E Lagrange, ‘L’application de la Convention de Rome à des actes accomplis par les Etats parties en dehors du territoire national’ (2008) 112 Rev gén de dr int public 521
  • RL Looper, ‘“Federal State” Clauses in Multilateral Instruments’ (1955–56) 32 BYBIL 162
  • M Milanovic, Extraterritorial Application of Human Rights Treaties (OUP, Oxford 2011)
  • M Milanovic, ‘From Compromise to Principle: Clarifying the Concept of State Jurisdiction in Human Rights Treaties’ (2008) 8 Hum Rts L Rev 411
  • L Moor and AB Simpson, ‘Ghosts of Colonialism in the European Convention on Human Rights’ (2006) 76 BYBIL 121


1  Thus, ratione personae asks to whom the legal text applies; ratione materiae asks to what subject matter it applies; ratione temporis asks when it applies, and ratione loci asks where it applies.

2  A similar provision is found in Art 29 of the 1986 Vienna Convention on the Law of Treaties between States and International Organizations or between International Organizations (adopted 21 March 1986, not yet in force) [1986] 25 ILM 543 (‘1986 VCLT’). See n 92 and accompanying text.

3  Indeed, Paul Reuter, Special Rapporteur for the International Law Commission’s (ILC) work on what became the 1986 VCLT observed that ‘the authors of [Article 29] had simply wanted to enunciate a rule for the interpretation of treaties’: [1977] YBILC, vol I, 117 [24].

4  [1966] YBILC, vol II, 65.

5  Ibid 65–6.

6  The ILC draft stated that ‘unless a different intention appears from the treaty or is otherwise established, the application of a treaty extends to the entire territory of each party’. Ibid 213. Article 29’s phrasing—‘a treaty is binding upon each party in respect of its entire territory’—followed a Ukrainian submission suggesting that the ILC draft did not sufficiently respect dualist legal systems where treaty provisions are not directly applicable as national law.

7  [1959] YBILC, vol II, 48.

8  [1964] YBILC, vol II, 12 et seq.

9  Ibid 179.

10  Ibid 12.

11  Ibid 179.

12  The ‘contrary intention’, which served as the basis for the exception, was replaced by a ‘different intention’. ‘Different’ is admittedly weaker than ‘contrary’; a merely different, but not drastically contrary, intention could thus set aside the main principle favouring integral territorial application.

13  International law formulae are rarely innocent, although not all the official versions of the VCLT place the exception before the principle. The Chinese, English, French, and Russian versions do, but the Spanish one continues to give grammatical privilege to the principle (‘Un tratado será obligatorio para cada una de las partes por lo que respecta a la totalidad de su territorio, salvo que una intención diferente se desprenda de él o conste de otro modo’).

14  The equivalent expression of the Spanish version is quite eloquent (‘de otro modo’). The same can be said of the Russian (‘иным образом’) and the Chinese ones (‘另’) [‘lìng’], whereas the French one is rather vague (‘par ailleurs’).

15  Or, ‘Etablir’, ‘constar’, ‘устанавливать’, ‘确定’ [‘què dìng’] in the other versions.

16  [1964] YBILC, vol II, 12.

17  A Aust, Modern Treaty Law & Practice (2nd edn CUP, Cambridge 2007) 206–8.

18  (1982) 28 DR 99.

19  (App no 7456/76) (1978) 13 DR 40 [5]. See also n 32 and accompanying text for the important Tyrer judgment.

20  Mme Smets, 14 May 1993, (1993) Rev gén de dr int pub 1056.

21  Treaty Section of the UN Office of Legal Affairs, Summary of Practice of the Secretary-General as Depositary of Multilateral Treaties (United Nations, New York 1999) UN Doc ST/LEG/7/Rev 1 [284]–[285].

22  Cf JES Fawcett, ‘Treaty Relations of British Overseas Territories’ (1949) 26 BYBIL 86–107.

23  P Lampué, ‘L’application des traités dans les territoires et départements d’outre-mer’ (1960) 6 AFDI 907–24.

24  See eg Law No 99–209 of 19 March 1999, Arts 21, 22, JO, 21 March 1999, 4197, 4198–99 (giving New Caledonia treaty-making authority in specific areas).

25  Convention for the Suppression of the Traffic in Persons and of the Exploitation of the Prostitution of Others (adopted 2 December 1949, entered into force 25 July 1951) 96 UNTS 271, Art 23.

26  Eg Convention on the Prevention and Punishment of the Crime of Genocide (adopted 9 December 1948, entered into force 12 January 1951) 78 UNTS 277, Art XII.

27  International Convention on the Settlement of Investment Disputes between States and Nationals of Other States (adopted 18 March 1965, entered into force 14 October 1966) 575 UNTS 159, Art 70.

28  Treaty Establishing the European Economic Community (adopted 25 March 1857, entered into force 1 January 1958) 294 UNTS 17, Art 227. Examples of each of these approaches may be found in Section VI of this volume, 715 et seq.

29  [1966] YBILC, vol II, 213.

30  See, for example, the reaction of Grigory Tunkin who asked whether it was ‘appropriate for the Commission to act as if the world had stood still and give its approval to colonial institutions’: [1964] YBILC, vol I, 49 [35]. Others, however, viewed the colonial clause as a potential vehicle for aiding States emerging from colonialism in dealing with the inevitable questions of State succession to treaties. The post-colonial Algerian Government, for example, argued the law of treaties should:

limit the application of a treaty to the metropolitan territory of the parties, unless the still subject peoples through a valid expression of opinion decide to accept the treaty and its effects. Otherwise the legitimate representatives of those peoples may have no alternative but to denounce treaties in which they have taken no part and which are, in its view, often detrimental to their interests.

[1966] YBILC vol II, 65.

31  Cf S Karagiannis, ‘L’aménagement des droits de l’homme outre-mer: la clause des “nécessités locales” de la Convention européenne’ (1995) 28 RBDI 224–305; L Moor and AB Simpson, ‘Ghosts of Colonialism in the European Convention on Human Rights’ (2006) 76 BYBIL 121–94.

32  (App no 5856/72) (1978) Series A no 26 [38].

33  (App no 66289/01) ECHR 2005-I.

34  Colonial clauses, of course, are not the only vehicle for doing this. France has used reservations to reserve room for the application of local law in its territories under various Protocols to the ECHR. See eg French reservation to Protocol No 7 (stating that the Protocol ‘shall apply to the whole territory of the Republic, due regard being had where the overseas territories and the territorial collectivity of Mayotte are concerned, to the local requirements referred to’ in what is now Art 56). Of course, the fact that the French Government thought that the Muslim Sharia in application on Mayotte constitutes a ‘local requirement’ is not in itself a sufficient proof to establish the existence of such a ‘requirement’. Moreover, even if this approach may foster the cohesion of the local communities under the menace of individually coloured Western values, it unavoidably leads to a dwindling of the protection of human rights of individual members of these communities, for example of the women’s rights in the case of Mayotte. It is suggested that the ECHR’s ‘colonial clause’ assumes quite boldly such consequences.

35  Other territories excluded from a complete application of the EU treaties and Union secondary law emerged from accession agreements, and include the Channel Islands, Isle of Man, British Sovereign Base Areas of Akrotiri and Dhekelia in Cyprus, and the Åland Islands. Some other territories are not addressed by Art 355, but only by specific provisions of the respective accession agreement or a simple unilateral declaration, creating doubts as to the applicable legal regime (eg areas traditionally inhabited by the Sami people in Finland or Mount Athos in Greece).

36  Said territories have special relations with four member States, namely Denmark, France, the Netherlands, and the United Kingdom. Amusingly, French Southern and Antarctic Territories and British Antarctic Territory are listed in Annex II as ‘overseas countries and territories’, although all the other member States reject any possibility whatsoever to raise territorial claims on the frozen continent.

37  TFEU Art 198.

38  The Court considers these taxes (‘octroi de mer’) as ‘a charge having an effect equivalent to a customs duty on imports, notwithstanding the fact that the charge is also imposed on goods entering that region from another part of the same State’. Case No C-163/90 Administration des Douanes et Droits Indirects v Léopold Legros [1992] ECR I-04625 [18].

39  TFEU Art 349.3.

40  These are French Saint-Barthélemy and Saint-Martin Islands in the Caribbean, the Portuguese Azores and Madeira Islands, and the Spanish Canary Islands. On 31 March 2011 a fifth French overseas department, Mayotte, in the Indian Ocean, was created. Yet, up to now, Mayotte has not been added to the list of TFEU Art 349. For the time being, this island remains listed in Annex II; which means that the national/unilateral change of the status of a territory does not automatically bind the EU. A formal modification of Art 349 will probably be needed. This is not what one could call a flexible solution for such matters.

41  Aust (n 17) 207–8 (describing UK practice of consultation with its dependencies and other overseas territories on the extension of treaties).

42  Determining which areas are outside the federal governments’ purview can be difficult, particularly in light of inconsistent practice by federal governments in joining treaties on matters that might otherwise seem constitutionally delegated to the sub-federal level. See DB Hollis, ‘Executive Federalism: Forging New Federalism Constraints on the Treaty Power’ (2006) 79 S Cal L Rev 1327, 1371–2.

43  Nonetheless, there are several cases where a federal state has declined to join a treaty on federalism grounds, among the most prominent being the US decision not to seek ratification of the International Covenant on Economic, Social and Cultural Rights (adopted 16 November 1966, entered into force 3 January 1976) 993 UNTS 3, or the Convention on the Rights of the Child (adopted 20 November 1989, entered into force 2 September 1990) 1577 UNTS 3. See also Hollis (n 42) 1372–3.

44  These are not, of course, the only available options. A federal State may also seek to issue reservations or understandings on ratification that address obligations that might otherwise be inconsistent with the State’s federal system. The United States has used both approaches, issuing a federalism reservation for its obligations under the UN Transnational Organized Crime Convention (adopted 15 November 2000, entered into force 29 September 2003) 2225 UNTS 209, and an understanding for the International Covenant on Civil and Political Rights (ICCPR) (adopted 16 December 1966, entered into force 23 March 1976) 999 UNTS 171. See also Hollis (n 42) 1361–3, 1379. Alternatively, federal States may seek to adjust the treaty obligations themselves to avoid federalism issues, for example, by limiting implementation requirements to the ‘national level’. Ibid 1377.

45  [1959] YBILC, vol II, 47. Fitzmaurice did, however, anticipate that States might agree to otherwise provide for inclusion of a ‘federal clause’ adjusting the obligations of Federal States. Ibid 75 [130].

46  See generally RL Looper, ‘“Federal State” Clauses in Multilateral Instruments’ (1955–56) 32 BYBIL 162–203.

47  Constitution of the International Labour Organisation (adopted 9 October 1946, entered into force 20 April 1948) 15 UNTS 35. In addition, the federal State is obligated to consult with its sub-federal units on such matters with a view to promoting coordinated action and reporting back to the ILO on its activities and any implementation done at the sub-federal level. Ibid Art 19(7)(ii)–(v).

48  Article 34 of the Convention for the Protection of the World Cultural and Natural Heritage is typical, limiting the responsibility of a federal State to provisions ‘the implementation of which comes under the legal jurisdiction of the federal or central legislative power’ whereas for implementation coming under the jurisdiction of the component units (which have no obligation to legislate), the federal government’s only obligation is to inform these units of the provisions and recommend their adoption. (adopted 16 November 1972, entered into force 15 July 1975) 1037 UNTS 151, 161–2; see also Convention on the Recognition and Enforcement of Foreign Arbitral Awards (adopted 10 June 1958, entered into force 7 June 1959) 330 UNTS 38, Art XI (New York Convention on Arbitration); Convention Relating to the Status of Refugees (adopted 28 July 1951, entered into force 22 April 1954) 189 UNTS 150, Art 51.

49  Council of Europe Convention on Cybercrime (adopted 23 November 2001, entered into force 1 July 2004) CETS No 185, Art 41.

50  CISG (adopted 11 April 1980, entered into force 1 January 1988) 1489 UNTS 3; see also UN Convention on International Bills of Exchange and International Promissory Notes (adopted 9 December 1988, not yet in force) UN Doc A/Res/43/165, Art 87.

51  Canada, for example, on joining the Sale of Goods Convention declared that it would extend the treaty to nine of its provinces and territories. Secretary-General, ‘Canada: Declaration in Accordance with Article 93’ (5 October 1998) C.N. 631.2003.

52  ICCPR (n 44). In response to this provision, the United States issued the following (controversial) understanding:

The United States shall implement all the provisions of the [Covenant] over whose subject matter the Federal Government exercises legislative and judicial administration; with respect to the provisions over whose subject matter constituent units exercise jurisdiction, the Federal Government shall take appropriate measures, to the end that the competent authorities of the constituent units may take appropriate measures for the fulfillment of this Covenant.

ICCPR, 19 December 1966, S Exec Doc E, 95–2 (1978) xiv. Article 50 does not, moreover, prevent the ICCPR’s Human Rights Committee (HRC) from asking States to produce in their reports specific information on how federated state law complies with provisions of the Covenant.

53  (Adopted 22 November 1969, entered into force 18 June 1978) 1144 UNTS 123.

54  Garrido v Argentina (1998) 4 Inter-American Ybk H Rts 3473.

55  Cf J Kalb, ‘Dynamic Federalism in Human Rights Treaty Implementation’ (2010) 84 Tulane L Rev 1025–66.

56  [1964] YBILC, vol II, 12 [1].

57  [1964] YBILC, vol II, 13 [3].

58  [1966] YBILC, vol II, 213 [3]. Given the sensitivities associated with decolonization discussed above, the ILC indicated a preference for this definition in lieu of one that emphasized territories ‘for which the parties are internationally responsible’. Ibid.

59  It is rather curious that the ILC, to which we largely owe the four 1958 Geneva Conventions on the Law of the Sea, did not elaborate on the maritime aspects of a State’s ‘territory’.

60  UNCLOS (adopted 10 December 1982, entered into force 16 November 1994) 1833 UNTS 3. See S Karagiannis, ‘L’article 59 de la Convention des Nations Unies sur le droit de la mer (ou les mystères de la nature juridique de la zone économique exclusive)’ (2004) 37 RBDI 325–418.

61  [1966] YBILC, vol II, 66.

62  Ibid.

63  A similar American proposition was drafted as such: ‘a treaty also applies beyond the territory of each party whenever such wider application is clearly intended’. Ibid.

64  [1966] YBILC, vol I(2), 49 [15].

65  [1966] YBILC, vol II, 213–14 [5].

66  It may, in the future, extend to celestial bodies as well.

67  The specificity of the high seas allows, however, some adaptations of the treaties concerned. See, for instance, in the case of the ECHR, Medvedyev and Others v France (App no 3394/03) ECHR 29 March 2010. See also Hirsi Jamaa and Others v Italy (App no 27765/09) ECHR 23 February 2012.

68  It is possible, of course, for a State to enter into a treaty on behalf not only of itself but also of another State. If the latter agrees, there is not any difficulty. Problems arise when no such agreement can be shown.

69  M Milanovic, ‘From Compromise to Principle: Clarifying the Concept of State Jurisdiction in Human Rights Treaties’ (2008) 8 Hum Rts L Rev 411–48 (noting claims by victims of aerial bombardment, inhabitants of territories under military occupation, suspected terrorists detained in Guantánamo, and the family of a former KGB spy assassinated in London that a State has human rights treaty obligation towards individuals located outside its territory).

70  ICCPR (n 44) Art 2(1) (emphasis added).

71  HRC, ‘General Comment No 31: Nature of the General Legal Obligations Imposed on States Parties to the Covenant’ (29 March 2004) UN Doc CCPR/C/21/Rev.1/Add.13. The ICJ endorsed this position in its Advisory Opinion on Legal Consequences on the Construction of a Wall in the Occupied Palestinian Territory:

The Court would observe that, while the jurisdiction of States is primarily territorial, it may sometimes be exercised outside the national territory. Considering the object and purpose of the International Covenant on Civil and Political Rights, it would seem natural that, even when such is the case, States parties to the Covenant should be bound to comply with its provisions.

[2004] ICJ Rep 179 [109] (Wall Opinion).

72  See KJ Heller, ‘Does the ICCPR Apply Extraterritorially ?’[2006] Opinio Juris <http://opiniojuris.org/2006/07/18/does-the-iccpr-apply-extraterritorially> (describing US objections to extraterritorial application); MJ Dennis, ‘Application of Human Rights Treaties Extraterritorially in Times of Armed Conflict and Military Occupation’ (2005) 99 AJIL 119, 125 (describing the Dutch position). One can also note that the ‘dual’ expression used in Art 2 of the Covenant is the unfortunate result of the concurrent American (‘within its territory’) and French (‘subject to its jurisdiction’) propositions. Unable to choose, the General Assembly preferred to include both of them separated by the conjunction ‘and’. Strictly speaking, Art 2 could now become too narrow a provision. This is what the HRC tried to avoid in its case law and in its General Comment (n 71).

73  Several other treaties—the American Convention on Human Rights and the UN Convention on the Rights of the Child—follow the ECHR in imposing obligations on States with respect to ‘persons subject to their jurisdiction’. Other treaties—such as the Convention on the Elimination of All Forms of Racial Discrimination (adopted 21 December 1965, entered into force 4 January 1969) 660 UNTS 195, and the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment (adopted 10 December 1984, entered into force 26 June 1987) 1465 UNTS 85—provide specific rights and obligations a State must apply to ‘any territories under its jurisdiction’. Milanovic (n 69) 413–14. For a complete review of international bodies’ and US practice see SH Cleveland, ‘Embedded International Law and the Constitution Abroad’ (2010) 110 Columbia L Rev 225–87.

74  Yet, what qualifies as a military occupation is only apparently clear in law. Cf Adam Roberts, ‘What Is A Military Occupation?’ (1984) 55 BYBIL 249–345.

75  George Vearncombe and Others v United Kingdom and Federal Republic of Germany (App no 12816/87) (1989) 59 DR 186 (decision of the European Commission on Human Rights that ‘acts performed by organs of an occupying State (including members of its army) are generally attributable to this State and not to the occupied State’).

76  Ilaşcu and Others v Moldova and Russia (App no 48787/99) ECHR 2004-VII.

77  Ibid 77 [33].

78  See ‘List of Declarations Made with Respect to Treaty 117’ in <http://conventions.coe.int/treaty/Commun/ListeDeclarations.asp?NT=117&CM=8&DF=02/10/2011&CL=ENG&VL=1>. Similarly, Georgia declared upon ratifying Protocol No 12 that it ‘declines its responsibility for the violations of the provisions of the Protocol on the territories of Abkhazia and Tskhinvali region until the full jurisdiction of Georgia is restored over these territories’. See ‘List of Declarations Made with Respect to Treaty 177’ in <http://conventions.coe.int/treaty/Commun/ListeDeclarations.asp?NT=177&CM=8&DF=02/10/2011&CL=ENG&VL=1>.

79  This is especially true where a special legal relationship links the State to the territory even if the State does not exercise sovereignty proper on this territory. Cf ‘Decision of 12 March 2002 of the Inter-American Commission on Human Rights in the case of Guantánamo detainees’ [2002] 41 ILM 532 (Commission considered that measures requested by petitioners were justified and necessary having regard, inter alia, to the fact that the detainees at Guantánamo ‘remain wholly within the authority and control of the United States government’ despite the fact that the military base at Guantánamo is not strictly speaking US territory but is the subject of a long-term lease from Cuba); Wall Opinion (n 71) 179 [109] (finding Israel in breach of its obligations under the ICCPR and the ICESC with respect to the occupied territories).

80  Loizidou v Turkey (App no 15318/89) (1995) Series A no 310. See also Cyprus v Turkey (App no 25781/94) (2001) ECHR 2001-IV.

81  Loizidou (n 80) [62]. In a separate line of cases, the ECtHR has interpreted jurisdiction in personal, rather than territorial terms, to apply where the use of force by a State’s agents operating outside its territory may bring the individual thereby brought under the control of the State’s authorities into the State’s Art 1 jurisdiction. See Al-Skeini and Others v UK (App no 55721/07) ECHR 7 July 2011 [135]–[136].

82  Banković and Others v Belgium and Others (App no 52207/99) ECHR 2001-XII [59]. This said, an absence of jurisdictional link due to a lack of effective control of a territory may be compensated for by judicial actions brought by the tribunals of the respondent State. Cf Marković and Others v Italy (App no 1398/03) ECHR 2006-XIV [55].

83  Issa and Others v Turkey (App no 31821/96) ECHR 16 November 2004 [71].

84  Al-Skeini (n 81) [149] (emphasis added).

85  Case concerning Military and Paramilitary Activities in and against Nicaragua (Merits) [1986] ICJ Rep 65 [115].

86  (Judgment) ICTY-1994-1 (15 July 1999) [120].

87  Issa (n 83) [68]; accord Al-Skeini (n 81) [149].

88  (App no 5853/06) ECHR 2006-XV.

89  Issa (n 83) [67].

90  Ibid [71]. Similar is the construction of the HRC in cases such as Burgos v Uruguay (1981) 1 Selected Decisions of the Human Rights Committee 88, 91 [12.1]–[12.3] and Casariego v Uruguay (1981) 1 Selected Decisions of the Human Rights Committee 92, 94 [10.1]–[10.3].

91  The extraterritorial application of the ECHR has provoked considerable emotion in military and political circles and given rise to much academic writing. See, among many others, F Coomans and M Kamminga (eds), Extraterritorial Application of Human Rights Treaties (Intersentia, Antwerp–Oxford 2004); S Karagiannis, ‘Le territoire d’application de la Convention européenne des droits de l’homme. Vetera et nova’ (2005) 16 Rev trim des droits de l’homme 33–120; M Gondek, ‘Extraterritorial Application of the European Convention on Human Rights: Territorial Focus in the Age of Globalization?’ (2005) Netherlands Intl L Rev 349–87; V Mantouvalou, ‘Extending Judicial Control in International Law: Human Rights Treaties and Extraterritoriality’ (2005) 9 Intl J Human Rts 147–63; S Skogly, Beyond National Borders: States’ Human Rights Obligations in International Cooperation (Intersentia, Antwerp–Oxford 2006); E Lagrange, ‘L’application de la Convention de Rome à des actes accomplis par les Etats parties en dehors du territoire national’ (2008) 112 Rev gén de dr int public 521–65; M Gibney and S Skogly (eds), Universal Human Rights and Extraterritorial Obligations (U Penn Press, Philadelphia 2010); M Milanovic, Extraterritorial Application of Human Rights Treaties (OUP, Oxford 2011).

92  1986 VCLT (n 2) Art 29. For a discussion of the status of the 1986 VCLT in customary international law see Chapter 3.

93  [1977] YBILC, vol I, 117.

94  Ibid 118 [26].

95  1986 VCLT Art 6 (‘The capacity of an international organization to conclude treaties is governed by the rules of that Organization’).

96  Eg UNCLOS (n 60) Art 305(1)(b). A related line of inquiry is whether UNCLOS’s International Sea-Bed Authority is an IO exercising power analogous to ‘territory’, namely the soil and sub-soil of the high seas beyond national continental shelves. At present, such an analogy seems difficult where the sea soil is at a depth of thousands of metres and the Sea-Bed Authority’s power is primarily functional and subsequently quite limited; its authority is limited to exploration and exploitation of natural resources and does not extend, for example, to shipwrecks. Ibid Art 149.

97  Legal writing on this point becomes extensive. See eg B Knoll, The Legal Status of Territories Subject to Administration by International Organisations (CUP, New York 2008); C Stahn, The Law and Practice of International Territorial Administration. Versailles to Iraq and Beyond (CUP, Cambridge 2008); R Wilde, International Territorial Administration: How Trusteeship and the Civilizing Mission Never Went Away (OUP, Oxford 2008); HF Kiderlen, Von Triest nach Osttimor. Der völkerrechtliche Rahmen für die Verwaltung von Krisengebieten durch die Vereinten Nationen (Springer Berlin, Heidelberg 2008).