Article 3(5) TEU a decade on: Revisiting ‘strict observance of international law’ in the text and context of other EU values

1. Introduction

Article 3(5) TEU was a new provision in the Lisbon Treaty, and one to which scholars were immediately drawn. It arrived when the Court’s respect for international law was perceived to be at its lowest ebb and the prevailing view was undoubtedly that Article 3(5) TEU required the Court to respect international law more greatly moving forward. This optimism pertains today – a decade on. Ultimately, though, Article 3(5) TEU does not appear to have impacted the Court’s case law meaningfully.

This article will argue that the lack of impact is eminently justified concerning respect for international law when considering the text of Article 3(5) TEU more fully. This can be further supported by consideration of the manner in which the Court typically interprets EU Treaty provisions and by the demands of international law itself concerning its application in domestic courts. Accordingly, ‘respect for international law’ is found – both within EU law and international law governing domestic application itself – to be unable to bear the ideological strain which continues to be placed upon it in scholarship. Article 3(5) TEU could, though, be better utilized by the Court in expressly detailing the balanced approach to respecting international law and EU values moving forward.

The article is structured to firstly consider the climate of the case law into which Article 3(5) TEU entered. Secondly, the full text of Article 3(5) TEU will be reflected upon. Thirdly, the Court’s interpretative approach to EU Treaty articles generally will be analysed. Fourthly, the application of international law in domestic courts will be considered from the perspective of international law’s demands and State practice. Finally, the limited impact of Article 3(5) TEU concerning greater respect for international law in case law subsequent to the Lisbon Treaty will be considered and endorsed, although the potential for more meaningful judicial and scholarly engagement with Article 3(5) TEU in its full text and context will be highlighted.

2. Prominent pre-Article 3(5) TEU case law

In the time between the Lisbon Treaty’s conclusion and its entry into force a couple of years passed, and this proved to be a significant period for the Court’s application of international law (its ‘external case law’). One of these years – 2008 – has been described as an ‘annus horribilis’. ¹ Whilst such criticism is often more intense precisely because the EU and its Court had previously been viewed as particularly respectful of international law, ² undoubtedly the changes were notable and real. We can briefly chart relevant developments leading to this.

The EU itself is a product of successive international treaties. Bold doctrines such as supremacy and direct effect of EU law emanated from what Alter notes to have been ‘an obscure Court in Luxembourg’. ³ It is safe to assume that greater tension would have arisen concerning internal acceptance of such doctrines had the Court begun from a position of hostility regarding its relationship with international law; its most obvious touchpoint for legitimacy. Therefore, in Haegeman II the Court asserted that international agreements form ‘an integral part of EU law’ ⁴ and in Woodpulp it appeared that customary international law could even operate as a constraint on the EU Treaty itself. ⁵ The Court opted clearly for a continuous thread with its (then) not so distant past, rather than to plough a lonely furrow. On this reading the most obvious initial benefit of being respectful to international law was internal.

But Brockers and Kuijper note that: ⁶

All this has now changed. The Community has become an important factor in international relations…Moreover, the Court’s position in relation to the other institutions within the Community has changed.

The Court’s treatment of the General Agreement on Trade Tariffs (GATT) and, subsequently, World Trade Organization (WTO) law are the first prominent examples of EU self-interest at the expense of applying international law. Concerning the Court’s refusal to grant direct effect (allowing reliance on international treaty law by individuals) to GATT there was, in fact, much scholarly support and relevant arguments were advanced by the Court highlighting GATT’s flexible nature. ⁹ Concerning WTO law, the Court acknowledged it did ‘differ significantly [from GATT]…in particular by reason of the strengthening of the system of safeguards and the mechanism for resolving disputes’. ¹⁰ However, the disinclination of ‘significant trading partners’ to grant direct effect to WTO law within their legal systems meant that to do so in the EU would have, in the Court’s view, deprived the legislature of the ‘scope for manoeuvre’ enjoyed by others. ¹¹ This position is less readily accepted by scholars and the debate continues to be intense. ¹²

Whilst a significant blot, ‘the lone exception of GATT and WTO law’ regarding the refusal of direct effect meant that the Court’s jurisprudence was still viewed as respectful towards international law. ¹³ This rendered the decision in Intertanko ¹⁴ all the more surprising. ¹⁵ In Intertanko the Court asserted that ¹⁶ :

[The United Nations Convention on the Law of the Sea] does not establish rules intended to apply directly and immediately to individuals and to confer upon them rights or freedoms capable of being relied upon against States, irrespective of the attitude of the ship’s flag State.

Kadi I featured a challenge to an EU Regulation ²⁰ on the basis that it breached fundamental rights. Complexity arose as the Regulation implemented without modification a UN Security Council Resolution targeting potential terrorists by freezing their funds. ²¹ The (then) Court of First Instance was very respectful of the international origin of the EU Regulation and so reviewed it on the basis of jus cogens alone, finding no breach. The Court adopted a very different approach; instead of reviewing an EU Regulation on the basis of international law, it reviewed it on the basis of EU fundamental rights law as the ‘autonomous’ nature of the EU legal system ‘cannot be prejudiced by an international agreement’ ²²  – even the UN Charter. ²³ In finding multiple breaches the Court annulled the regulation but, in recognition that ‘immediate effect would be capable of seriously and irreversibly prejudicing the effectiveness of the restrictive measures imposed’, it suspended the judgment’s effect for three months. ²⁴

The nuance of outcome in Kadi I ²⁵ did little to prevent criticisms from some that ‘the judgment seems to lack a thorough engagement with international law issues’ ²⁶ and that the Court was guilty of ‘withdrawing into one’s own constitutional cocoon, isolating the international context and deciding the case exclusively by reference to internal constitutional precepts’ ²⁷ , with ‘important parts of its reasoning [expressed] in chauvinist and parochial tones’. ²⁸

FIAMM – decided less than a week after Kadi I – returned (like Intertanko) to the issue of direct effect and rounded 2008 off with a further blow to the potential for review of the legislature’s actions where WTO law was concerned. FIAMM was a product of the well-known ‘banana wars’, which saw the EU provide preferential treatment to African, Caribbean and Pacific countries to the detriment of Latin American countries in particular. The result was WTO-approved retaliatory action by the US, which impacted directly on certain traders in fields quite separate from the initial dispute. ²⁹ The Court’s disinclination to review the decision of the legislature was based on the risk of hindering action taken ‘in the public interest…which adversely affect the interests of individuals.’ ³⁰ However, without review the presumption appears that in the WTO sphere the legislator, simply by acting, is always acting in the public interest and cannot be questioned. ³¹ Maduro AG thought the Court had all but established in internal case law the principle that it would review ³² and he thought the threshold for review had been met in FIAMM itself. ³³ Aside from weakening review functions in EU law, the disinclination to engage with international law in such a case, whilst fitting within the trend of the Court’s WTO case law, was a further disappointment to many.

Kadi, Intertanko and FIAMM combined to represent a year in which international law had not been readily received within the Court’s case law, but with the entry into force of Article 3(5) TEU many scholars encouraged and anticipated a resurgence in respect for international law.

3. The (over-)emphasis on ‘strict observance’ in Article 3(5) TEU commentary

References to Article 3(5) TEU in the literature have been manifest but partial: ‘[T]he Union…shall contribute to…the strict observance of international law’ is the typical formulation. ³⁴ It has also become a popular title (emphasizing ‘strict observance’) for journal articles. ³⁵ The prevailing view is clearly that Article 3(5) TEU matters, and that strict observance of international law is its pertinent part. The consequence, unsurprisingly, was that scholars believed it to be a vital new tool for the Court to more robustly apply international law, and one which could be used to criticize case law should it not. Therefore, Mendez suggested that Article 3(5) TEU had ‘supplied the EU Courts with a powerful additional anchor with which to hold a more unreserved embrace of [international] treaty enforcement’ ³⁶ , whilst Petersmann asked: ‘Can the EU’s Disregard for “Strict Observance of International Law” (Article 3 TEU) Be Constitutionally Justified?’ ³⁷ More surprising than scholarship’s focus on respect for international law alone in Article 3(5) TEU was that this imbalanced approach was followed by Advocates General ³⁸ and the Court, ³⁹ with partial citation of Article 3(5) TEU in case law omitting key references to EU values. However, as noted above and developed below, more robust enforcement of, and respect for, international law in the Court has not occurred as a result. The following two sections will focus on both the text and context of Article 3(5) TEU and will argue as a result of the analysis that ‘strict observance’ of international law is overemphasized in scholarship. The reason for this is that fuller consideration of Article 3(5) TEU and other relevant EU provisions limit its significance, whilst international law’s own requirements fall far short of this purported ideal.

4. The text

It is submitted that Article 3(5) TEU’s modest impact is justified when it is viewed in its full form:

In its relations with the wider world, the Union shall uphold and promote its values and interests and contribute to the protection of its citizens. It shall contribute to peace, security, the sustainable development of the Earth, solidarity and mutual respect among peoples, free and fair trade, eradication of poverty and the protection of human rights, in particular the rights of the child, as well as to the strict observance and the development of international law, including respect for the principles of the United Nations Charter.

Reading the provision, it also appears questionable to isolate ‘strict observance’ as a stand-alone term; the word ‘contribute’ appears at the start of the long sentence – containing multiple aims – and appears to suggest further conditionality (for example, ‘ensure strict observance’ would be stronger). ‘[C]ontribute’ could even be read as being dependent on others; requiring an existing practice which could be contributed to. Linguistically at least, one could argue that the Article may even anticipate that the practice of others would already be one of strict observance to which the EU could then contribute (otherwise ‘the’ seems superfluous). However, strict observance of international law is a practice which reviews of domestic courts’ application of international law reveal is far from uniform. ⁴¹ Whilst the term ‘development’ perhaps militates against excusing the EU from making some effort in this circumstance, concerning WTO law it is notable that the reluctance of significant trading partners to grant direct effect was considered by the Court as a barrier to doing so in the EU. Were the Court to have unilaterally granted such effect the compliance pull for others (necessarily representing an empirically larger issue) would clearly have been less than highlighting the problem. ⁴² Similarly, that Kadi I subsequently gave rise to improvements to the UN terrorism listing process in regard to international human rights protection could be seen as a ‘development’ resulting from the Court’s (arguably) less than ‘strict observance’ of the UN Charter. ⁴³

Article 3(5) TEU is less than decisive in multiple aspects ⁴⁴ and consideration of Article 21 TEU arguably lends further support to the view that some flexibility was intended. Whilst seeking ‘a high degree of cooperation in all fields of international relations’, it is notable that the Lisbon Treaty’s Working Group on External Action’s proposal for ‘maximum degree of cooperation’ was not adopted. ⁴⁵ Indeed, the first stated aim of any co-operation in Article 21 TEU is the EU-centric aim to ‘safeguard its values, fundamental interests, security, independence and integrity’. ⁴⁶

This gives rise to possible clashes between EU and international law where values do not align or (as is more often the case) where EU law is progressing in a given area at a speed which differs from international law. Accordingly, the ‘rule of law, the universality and indivisibility of human rights’ ⁴⁷ were at stake in Kadi, with EU rules providing for a higher level of individual protection. Concerning airline emissions, after seeking to ‘help develop international measures to preserve and improve the quality of the environment’ through relevant international aviation bodies, the legislature subsequently acted unilaterally and was challenged in Air Transport – but given the challenges of arriving at international consensus it seems that the EU could not, as mandated by Article 21 TEU, ‘ensure sustainable development’ otherwise. This was similarly so for Intertanko, albeit the environmental imperative of ship source pollution appeared less pressing in breadth and magnitude than that at stake in Air Transport.

Accordingly, whilst it is true that under Article 21 TEU the EU ‘shall promote multilateral solutions to common problems, in particular in the framework of the United Nations’, equally it ‘shall ensure consistency between the different areas of its external action…and its other policies’. Both fundamental rights and environmental protection are good examples of the challenges which pertain as a result of these dual demands. Guaranteeing a certain level of fundamental rights protection or pursuing a certain level of environmental protection within the EU will not – and has not – always accorded with that of UN bodies. The choice then becomes a domestic redrawing of that which was hitherto ‘fundamental’ in the case of rights or of making adjustments which may give rise to unfairness for participants and ineffectiveness for environmental protection. ⁴⁸ The question of how far the EU should pursue ‘multilateral solutions’ before turning to unilateralism – if at all – is heavily debated, but it is reinforced rather than resolved by Article 21 TEU.

It is submitted that Articles 3(5) and 21 TEU are not alone in their indeterminacy regarding the application of international law and appropriate approaches to it within the EU. Should multilateral or bilateral negotiations (encouraged in Articles 3(5) and 21 TEU) end in conclusive agreement, Article 216(2) TFEU states that ‘[a]greements concluded by the Union are binding upon the institutions of the Union and on its Member States.’ However, in articulating the hierarchy or interaction of concluded international agreements with the EU Treaty, secondary legislation and the Member States’ legal systems – and even the first step of governing reliance upon international treaties – Cremona observes that whilst this Article articulates ‘an important principle: [it] leaves many of these questions [of effects] unanswered’. ⁴⁹ Mendez notes equally that ‘[a] textually more explicit provision was certainly possible.’ ⁵⁰

Indeed, in deciding that international treaties concluded by the EU ‘form an integral part of [EU] law’, ⁵¹ the Court did not even make reference to Article 216(2) TFEU (then Article 228(2) EEC). ⁵² In subsequent cases the Article was referred to more frequently. ⁵³ However, the Court’s varied approaches to international treaties’ internal effects seem difficult to object to on the basis of Article 216(2) TFEU alone. The most helpful suggestion is that the provision is merely an expression of the pacta sunt servanda principle (that agreements are binding) in international law. ⁵⁴ Otherwise the dramatic statement that the ‘Community constitutes a new legal order of international law’ ⁵⁵ which was ‘to some extent…defined…in opposition to international law’ ⁵⁶ in Van Gend en Loos would have been rather redundant. But the ‘binding’ nature of international agreements within international law, returns us once more to the debate concerning the limited obligations of domestic courts in such instances. ⁵⁷

In addition to the textual limitations of Article 3(5) TEU (and Article 21 TEU and 216(2) TFEU), it is submitted that there are contextual limitations to be considered. From the EU perspective this concerns interpretative approaches common to all EU Treaty provisions, whilst from the international perspective, the requirements concerning strict observance have been few and the practice of States suggests reluctance.

5. The context

6. Prominent post-Article 3(5) TEU case law

As noted above, Article 3(5) TEU has been referred to by Advocates General and the Court in similar, partial, fashion to that of many scholars. However, the most prominent cases since Lisbon have not heralded significant change in the Court’s approach towards international law.

Air Transport arrived in 2011 and saw a Directive expanding the Emissions Trading Scheme to domestic and international flights departing from or arriving in the EU being challenged on the basis of international customary and treaty law. There were encouraging signs for those hoping Article 3(5) TEU would ‘bite’. Firstly, the Advocate General noted that the EU ‘expressly avows its aim of contributing to the strict observance and development of international law (second sentence of Article 3(5) TEU)’ ¹⁰⁰ and pointed out that Member States could ‘contribute in the public interest to judicial review of the legality of acts of EU institutions [and] [u]nder the second sentence of Article 3(5) TEU this includes ensuring the strict observance of international law.’ ¹⁰¹ The Court also made reference to Article 3(5) TEU, and did so in the partial manner familiar in the scholarship, even adding ‘[c]onsequently, when [the EU] adopts an act, it is bound to observe international law in its entirety, including customary international law’. ¹⁰² However, ultimately the Court was only willing to conduct a ‘manifest errors’ review of the legislature’s action due to its questionable claim that customary international law lacked certainty: ¹⁰³ in this regard, the approach to customary international law – at best – reflected the pre-Lisbon case of Racke. ¹⁰⁴ In that case, too, a margin was provided through ‘manifest errors’ review, although in Racke this was due to the specific uncertainty of the relevant provisions of customary international law. In the earlier case of Poulsen no such limitation was applied to the review, meaning ‘strict observance’ of customary international law now appears weaker after Air Transport, as all rules are treated as less certain.

Kokott AG also – ultimately – seemed to place little emphasis on Article 3(5) TEU in meaningfully changing issues of justiciability of international law before the Court or of internalizing a more significant obligation to respect international law, instead relying on pre-Lisbon case law without distinction or modification on the basis of the new Treaty provisions. ¹⁰⁵ Moreover, in answering whether the EU should have acted outside of the International Civil Aviation Organization (a UN body), Article 21 TEU (which calls on the EU ‘promote multilateral solutions to common problems, in particular in the framework of the United Nations’) was not referred to by the Advocate General ¹⁰⁶ nor by the Court. ¹⁰⁷ Instead the lack of justiciability of international provisions on this issue was the focus. ¹⁰⁸

An interesting aspect of the judgment, though, was the Court’s reference to the fact that ‘European Union policy on the environment seeks to ensure a high level of protection in accordance with Article 191(2) TFEU’ and recognition that the legislation was ‘designed to fulfil the environmental protection objectives which [the EU] has set for itself’. ¹⁰⁹ The Court’s inclination to discuss values which the EU pursues accords with Articles 3(5) and 21 TEU (albeit the Court did not make this point expressly).

The opportunity to engage the Articles more fully was desirable, given that appeals to substantive purpose had not previously been of relevance in deciding whether to apply international treaty or customary international law fully ¹¹⁰  – and without fuller articulation this remains the case. However, Article 3(5) TEU seems to open the possibility that there could be a justification available to the legislature where observance of international law is not ‘strict’ but another important EU interest is pursued in the face of international treaty or customary law. To not permit this would be to require the legislature to simultaneously ride two potentially divergent horses (in respect for international law and promotion of EU values), but to do so would herald a significant shift in focus towards a more normative/substance based approach. If carried forward this would harmonize with the approach adopted in reviewing EU legislation more generally, where (proportionate) pursuit of one interest is permitted even where it impacts on another. ¹¹¹

If this was a (very) gentle testing of the waters in the potential to expand this practice by the Court then response to it was decidedly mixed; for some, the environmental credentials of the action could have been more richly argued by the Court but were a noble pursuit, ¹¹² whilst for others the action was flagrant unilateralism, plainly illegal under international law regardless of normative aim. ¹¹³ Overall, it is notable that in a case which featured so many elements of Articles 3(5) and 21 TEU, the Court saw little help in them (or at least had little recourse to them) and appeared reluctant to operationalize promising aspects of them.

Certainly not much had changed concerning the Court’s observance of international law: the Court’s ‘curt’ reasoning for aspects of international treaty law in Intertanko ¹¹⁴ applied once more, this time to customary international law, and the appeal to the importance of the aim – which led many scholars to support the principles espoused in Kadi concerning fundamental rights – could be seen recurring concerning environmental protection in Air Transport. The availability of Article 3(5) TEU in articulating this reasoning was not fully availed of by the Court, but is hard to see how Air Transport would have been required to be decided differently prior to Lisbon, or how Intertanko and Kadi would have been after Lisbon.

Indeed Kadi II reflected this to a large extent. It is notable that the Court did not redeploy its assertions regarding the autonomous nature of the EU legal system from Kadi I, however, it has been argued above that ultimately (far from autonomy) the Court balanced respect for international law against fundamental rights in the remedy. In Kadi II the Court drew upon the new provisions in the Lisbon Treaty to support what was a very similar analysis: ¹¹⁵

…it is necessary to determine whether, in the light of the requirements stated, in particular, in Article 3(1) and (5) TEU and Article 21(1) and (2)(a) and (c) TEU, relating to the maintenance of international peace and security while respecting international law…[the measure] constitutes an infringement of the rights of the defence and the right to effective judicial protection.

Article 3(5) TEU case law remains limited, ¹¹⁷ but the extent to which it marks a new, more liberal, approach to invocation has also suffered a significant blow concerning WTO law. In C & J Clark International and Puma the position that the Court cannot review the legislature’s actions, even – or as the Court put it here ‘merely’ ¹¹⁸  – after a negative WTO finding for which the time limit to rectify has expired, was confirmed as unchanged after Lisbon. Moreover, far from moving away from the 2008 case law, both Intertanko and FIAMM were treated as exemplars for the direct effect test rather than anomalies from a regrettable year. ¹¹⁹

More respect was shown to WTO law in Commission v. Hungary, where measures requiring international agreements to be concluded prior to providing educational services into Hungary were challenged under Article 258 TFEU. ¹²⁰ However, this also reflected pre-Lisbon case law in which the Commission was free to rely upon WTO law should it choose to. ¹²¹ Accordingly, Hungary argued that the Commission had brought the case against it ‘for purely political considerations’ relating to protecting the Central European University, and had ‘seriously undermined the right to good administration’ as a result. ¹²² It was a charge to which the Court could merely respond by reiterating that the Commission’s discretion in bringing an action was not one which it was able to review. ¹²³ For practical purposes coincidental interests of the EU and international law can lead to greater enforcement across a range of issues, ¹²⁴ but this remains solely at the discretion of the Commission and is impacted as a solid basis for consistent (let alone strict) observance of international law as a result. Indeed, Article 3(5) TEU was not cited.

The Western Sahara case is arguably the strongest evidence of Article 3(5) TEU’s limited impact. ¹²⁵ It saw the referring court question validity of an international agreement based on Article 3(5) TEU specifically. The Court did not address the Article in detail as the case could be dealt with based on narrow textual points within the agreements themselves. Whilst it is true that the Court did not exclude the possibility that Article 3(5) TEU could be justiciable, nor did it herald a new dawn. The Court stated relevant rules that would have been applied if necessary, covering quite exhaustively the key aspects of conclusion and judicial review of international agreements. With mechanical familiarity, ‘the Court’s settled case-law’ is recalled throughout and whilst Article 267 TFEU and 216(2) TFEU feature, Article 3(5) TEU does not. ¹²⁶ Far from change, the Court appears at pains to select at least one pre- and one post-Lisbon case for each point, reinforcing continuity. ¹²⁷

7. Conclusion

Ultimately, ‘strict observance of international law’ in Article 3(5) TEU has had little impact, in spite of its prominent place in scholarship. Notwithstanding specific shortcomings in reasoning identified within recent case law ¹²⁸ and the Court’s apparent reluctance to fully utilize Article 3(5) TEU to introduce more substantive review on the merits of aims pursued within challenged EU legislation, it is submitted that the general position by which Article 3(5) TEU has not resulted in a significant lurch towards this supposed ideal is appropriate.

Firstly, even if one seeks to overlook the balance between promoting EU values and respecting international law within the text of Article 3(5) TEU itself, it is not apparent why this Article alone should be exempt from either teleological interpretation, by which other Treaty articles with varying aims harmonize, or the application of proportionality, by which they compete. Article 3(5) TEU sits within a series of provisions of primary law, which require that things ‘shall’ or ‘shall not’ happen a combined 2531 times. ¹²⁹

Secondly, even if we consider Article 3(5) TEU in isolation and take only its most internationally friendly provisions, we would be confronted with a challenge in determining what a stand-alone obligation to strictly observe international law would mean in practice. It seems a significant leap to claim that language denoting respect for international law in Article 3(5) TEU can be taken to overturn decades of practice in which domestic courts – under international law – have retained discretion in their application of treaty and customary rules. There is evidence that this is still being enjoyed by ‘significant trading partners’’ courts, and the Court of Justice of the European Union becoming a conduit of – rather than participant in – international law would weaken the compliance pull for other States and so could limit ‘strict observance’ in concrete terms.

Thirdly, if the above appears as cynical realpolitik, then a postulation in which we consider the relationship between EU and international law exclusively also poses challenges. The possibility that ‘an “illegal” act…contains the seeds of a new legality’ in international law ¹³⁰ may appear fanciful, but (if illegal) the Kadi I decision reflects aspects of this process in a concrete way regarding improvements to the UN terrorism listing system, and Air Transport and Intertanko seek to do similarly concerning environmental protection. The complexity regarding what is required to respect international law – from the perspective of international law itself – is also confounded by the anticipation of ‘development’ of international law in Article 3(5) TEU.

What, then, should the EU’s interaction with the wider world be and how should the Court respond? These are complex questions and the Treaty undoubtedly assists in addressing them, not least Article 3(5) TEU. But the full text and context of the Article need to be engaged with. Unless we dismiss other provisions within the EU Treaty as meaninglessly shallow, ‘strict observance of international law’ cannot be hallow. We need not look far for confirmation of this; Article 3(5) TEU could just as easily be partially cited as ‘[i]n its relations with the wider world, the Union shall uphold and promote its values and interests and contribute to the protection of its citizens…’