Jason Pobjoy explains in The Child in International Refugee Law (CUP 2017) how the rule set out in Article 31 of the Vienna Convention on the Law of Treaties (VCLT), the ‘general rule of interpretation’, in his words, ‘comprises a single holistic “rule” of interpretation’, and that the adoption, by the International Law Commission (ILC),
‘of a “single, closely integrated rule”’ underscores the need to look beyond a literal construction of the text and to consider the terms of the treaty in light of their object and purposes, in their context and taking into account subsequent extrinsic sources (p 34).
He goes on to explain that this ‘rejection of strict literalism has found favour in domestic refugee jurisprudence’ and then sets out how Articles 31–33 mandate a ‘systemic approach’ to the interpretation and application of the Refugee Convention, with the Convention on the Rights of the Child naturally playing a particularly important role (p 34–43). It is extremely well done.
This approach, clearly and convincingly set out in Chapter 1 of the book, is undoubtedly correct. It is of a feather with the dictum of the International Court of Justice in Interpretation of the Agreement of 25 March 1951 between the WHO and Egypt, according to which a treaty ‘does not operate in a vacuum’ but rather ‘in the context of a wider framework of legal rules of which it forms only a part’ (ICJ Rep 1980, p 73, 76, para 10). Against this background, the book in Chapter 6 develops the argument that the Convention on the Rights of the Child is a complementary source of protection for children, going into some detail on state practice to make out the argument.
As the Refugee Convention litigation in domestic courts is becoming increasingly sophisticated over time, we see more and more references to Articles 31–33, counsel and judges drawing nice distinctions between the various means of interpretation and their interplay within the framework of the VCLT rules. As Jason ably demonstrates, such an interpretative approach is not only correct as a matter of the law of treaties: it is also on the whole beneficial for the child in international refugee law.
It is obviously true that, as Jason observes (p 3), children fall within the ambit of the term ‘person’ in the non-child-specific wording of Article 1A(2) of the Refugee Convention; but, as children are different from adults in several regards, particular accommodation must be sought of children for the object and purpose of the Refugee Convention to be given appropriate effect in relation to them. As the book sets out, that type of accommodation can be done through systemic interpretation.
But it is not only arguments on treaty interpretation by counsel for claimants that are becoming more and more sophisticated: as is to be expected, government lawyers, too, increasingly draw on means of interpretation such as ‘subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’ (Article 31(3)(b) of the VCLT) and other ‘relevant rules of international law’ (Article 31(3)(c) of the VCLT).
In this context, it may well be that literalism or textual interpretation is rather better than its reputation.
Other relevant rules of international law
From time to time respondents, whether at the domestic or international level, will attempt to push the systemic integration argument too far, by taking their stand on other treaties or rules of customary international law that are said in effect to cancel out the state’s obligation under the initial treaty. This could be a way in which the sophisticated approach to interpretation of the Refugee Convention that Jason sets out might be relied on, to reduce to a vanishing point the rights that have been given clear expression in the Convention.
Sir Frank Berman has warned that a court that relies on Article 31(3)(c) of the VCLT must be careful not to go too far, as that might entail ‘plac[ing] an unwarranted limitation on the Parties’ freedom of contract’; the court might then ‘substitute for what the Parties had provided in their Treaty [other] rules of international law that applied between them in any case’ (‘Treaty “Interpretation” in a Judicial Context’ (2004) 29 YJIL 315, 320).
International courts and tribunals have been acutely aware of this problem, both within the context of migration and in other fields of international law. As the Grand Chamber of the European Court of Human Rights held in a case about Libyan migrants travelling towards Italy by sea, Hirsi Jamaa v Italy, where Italy argued that, by operation of a different treaty, it was not bound by the European Convention on Human Rights in relation to an issue governed by the other treaty: ‘Italy cannot evade its own responsibility by relying on its obligations arising out of bilateral agreements with Libya’ (Hirsi Jamaa & Others v Italy (2012) 163 ILR 132, para 129). The other ‘relevant rules of international law applicable in the relations between the parties’ cannot replace the law of the tribunal.
The limits of the systemic approach was drawn with particular clarity by a senior tribunal in Indus Waters Kishenganga (Pakistan v India), where the tribunal, having been invited to incorporate into its constituent treaty (the Indus Waters Treaty) other rules of international law (in that case, rules of customary international law), concluded that: ‘if customary international law were applied not to circumscribe, but to negate rights expressly granted in the Treaty, this would no longer be “interpretation or application” of the Treaty but the substitution of customary law in place of the Treaty’ ((2013) 157 ILR 362, 412, para 112; underlined here).
Putting the point at its lowest, authorities such as Hirsi Jamaa and Kishenganga might suggest that other rules of international law—whether customary or conventional—will be more likely to impact on the interpretation when they go with the grain of the treaty provision to be interpreted than when they are at loggerheads with it. This seems to be entirely in keeping with Jason’s argument and use of the systemic approach.
Subsequent practice of the parties
With a miraculous treaty such as the Refugee Convention—a strongly worded treaty instrument that could not have come about had it not been born of the horrors of World War II and of the abject situation of displaced persons in the wake of that war—it may however be correct to rely only on the wording of the treaty itself.
Why? As explained by one of the last century’s greatest international lawyers, Charles de Visscher, a Judge both of the International Court and its predecessor the Permanent Court of International Justice: ‘Il n’est pas demandé au juge de pénétrer les dispositions intimes des contractantes; il lui est demandé de dégager par les moyens à sa disposition cette part des intentions des parties que des signes extérieures révèlent. Or, les termes librement choisis par elle sont par priorité l’instrument de cette extériorisation. … De cette garantie contractuelle fondamentale, le texte, œuvre commune des parties, est l’instrument essentiel’ (‘Remarques sur l’interprétation dite textuelle des traites internationaux’ in Liber Amicorum JPA François (AW Sijthoff 1959) 383, 383–384). The International Court gave its imprimatur to this approach in, e.g., Territorial Dispute between Libya and Chad, where it observed that ‘interpretation must be based above all upon the text of the treaty’ (ICJ Rep 1994, p 6, 21–22, para 41).
Take for example Article 28 of the Refugee Convention, which provides in the first sentence of its first paragraph that:
‘The Contracting States shall issue to refugees lawfully staying in their territory travel documents for the purpose of travel outside their territory unless compelling reasons of national security or public order otherwise require’.
Many refugees are unable to provide documentation that substantiates their identity. This type of travel document will be the only form of identification that they can lay their hands on—the only way in which they can identify themselves in the receiving state. But, displaying buyers’ remorse, some governments take a dim view of this strongly worded treaty obligation, openly rejecting it on its face. In many countries in Europe these days Convention refugees are having to live in a limbo where they cannot lawfully get a job, travel from one country to another to visit family, enrol for education, set up a bank account, receive social security services for the simple reason that governments refuse to issue them Article 28 ‘travel documents’. They become, in Sedley LJ’s arresting phrase from the context of asset freezing, ‘effectively prisoners of the state’ (Ahmed v HM Treasury  EWCA Civ 1187,  2 AC 534, 580, para 125); the children who are unfortunate enough to be made by the government to grow up without an identity become effectively child prisoners of the state.
In cases before domestic courts, government lawyers have increasingly taken the view that, the clear wording notwithstanding, it surely cannot be the case that all refugees lawfully staying in their territory have a right to such documents unless ‘compelling reasons of national security or public order otherwise require’.
In Siamo (No 116618, 8 June 1994) the French Conseil d’État was asked whether an Angolan refugee who had been issued a travel document but then illegally gave it to certain non-nationals who were seeking to enter France irregularly. Did he have a right to a new Article 28 travel document? The government had refused to give him a new one. But in a terse one page judgment—where the conclusions had been given by one Commissaire du gouvernement Ronny Abraham—the Conseil d’État held that, ‘aucune raison impérieuse de sécurité nationale ou d’ordre public ne justifiait que ce soit d’une manière définitive qu’il fût privé de la possibilité d’obtenir la délivrance de ce document’. Although he had breached French law, Siamo was due a new travel document. The straightforward wording called for a straightforward interpretation; in the face of a clear wording there was no need for any other means of interpretation to enter into it.
Similarly the Borgarting Court of Appeal in Norway found recently in Aron & others v The Immigration Appeals Board (No LB-2016-3734, 13 February 2017) that a number of claimants, of whom some were children born in Norway, had unlawfully been denied travel documents, making it clear that the Norwegian government has been breaching the Refugee Convention in a systematic and flagrant manner. The Norwegian government has taken the view that refugees whose identity is in doubt will by definition be ineligible for issuance of Article 28 travel documents. Shoehorning this large group of refugees into what is Article 28’s narrow exception (‘compelling reasons of national security or public order’) is at sharp angles with the clear meaning of the wording of the Convention text.
Aron is currently pending before the Supreme Court, with hearings scheduled for October 2017. Like Ronny Abraham in Siamo, counsel for the refugees in Aron, Professor Mads Andenas, is arguing that the clear wording of the Convention quite simply affords them a strong right, one which must be allowed to apply and not be adulterated by other means of interpretation. The government’s elaborate arguments in that case have largely been about watering down the clear text of the Convention by reference to materials scripted as ‘subsequent practice’ of the parties. The Court of Appeal gave short shrift to the argument according to which the clear text of the treaty—‘l’instrument essentiel’ for the treaty interpreter, to use de Visscher’s phrase—could be said to have been diluted by the practice of states that are quite simply breaching the clear wording of the Convention.
The International Court of Justice, too, has restricted the use of the subsequent practice argument, as is clear from the judgment of the International Court in Whaling in the Antarctic. There the Court found that the functions which the International Convention for the Regulation of Whaling conferred on the International Whaling Commission (IWC) had ‘made the Convention an evolving instrument’ (ICJ Rep 2014, p 226, 247, para 45); crucially, however, the Court went on to observe that ‘amendments to the Schedule and recommendations by the IWC may put an emphasis on one or the other objectives pursued by the Convention, but cannot alter its object and purpose’ (ibid, para 56). Again, in common with the systemic approach, if the subsequent practice of the parties can be shown to go with the grain of the object and purpose of the treaty, then the argument for reliance on it will be stronger than if the practice cuts against that object and purpose.
The French Conseil d’État’s judgment in Siamo and the Norwegian Court of Appeal judgment in Aron tend to bring out the importance in the context of the interpretation of the Refugee Convention—both in cases concerning adults and children—of the signed text of the treaty itself. ‘The signed text is’, as the ILC Special Rapporteur Sir Humphrey Waldock put it, ‘with very few exceptions, the only and the most recent expression of the common will of the parties’ (ILC Ybk 1964 II p 56; ILC Ybk 1966 II p 220).
So long as the systemic approach and reliance on means of interpretation such as subsequent practice is kept within bounds, Jason Pobjoy’s rejection of literalism in the interpretation of the Refugee Convention in cases relating to children is entirely to be commended. But, given the strong and clear wording of the Refugee Convention and the Convention on the Rights of the Child, and the fact that, more often than not, governments simply deny refugees the rights set out in plain terms in the conventions, the consummation of the promise of ‘a literal construction of the text’—even ‘strict literalism’—is one devoutly to be wished!