Practice makes perfect – the High Court of Australia and Articles 31(3)(a) and (b) of the Vienna Convention on the Law of Treaties – Part 2

This post is the second of a two-part series on the High Court of Australia’s use of subsequent agreement and practice under Article 31(3)(a) and (b) of the Vienna Convention on the Law of Treaties (VCLT) to interpret treaties that appear in domestic law (which I refer to broadly as ‘incorporated treaties’).

Part 1 of this series considered the High Court’s approach to subsequent agreement and practice in Minister for Home Affairs v Zentai (2012) 246 CLR 123 (Zentai), Maloney v The Queen (2013) 252 CLR 168 (Maloney) and Commonwealth Minister for Justice v Adamas (2013) 253 CLR 43 (Adamas). These cases indicate that the High Court will not consider subsequent agreement or practice to interpret an incorporated treaty where such recourse would conflict with Australia’s division of powers. However, where no such conflict arises, the High Court remains divided on what recourse may be made to VCLT Articles 31(3)(a) and (b) materials. 

Part 2 in this series first considers the High Court’s receptive use of subsequent practice in Macoun v Commissioner of Taxation (2015) 257 CLR 519) and then examines the Court’s latest decision concerning an incorporated treaty – Addy v Commissioner of Taxation [2021] HCA 34. Finally, I explain why, subject to any conflict with Australia’s division of powers, Australian courts should consider VCLT Articles 31(3)(a) and (b) materials in its interpretative exercise. 

The twist 

Two years after Adamas, the High Court was presented with a further opportunity to consider extrinsic materials in Macoun v Commissioner of Taxation (2015) 257 CLR 519 (Macoun). This case centered on the International Organisations (Privileges and Immunities) Act 1963 (Cth) (IOPI Act) which gives effect to Australia’s obligations under the Convention on the Privileges and Immunities of the Specialised Agencies (Agencies Convention) through the Specialized Agencies (Privileges and Immunities) Regulations 1986 (SAPI Regulations). Importantly, the Agencies Convention was not incorporated into statute. Instead, it was ‘common ground’ that the Court was examining the Agencies Convention ‘to construe the IOPI Act in a manner which accords with Australia’s international obligations if such a construction is open’ (at [67]).  

The Court stated that VCLT Articles 31 and 32, including Article 31(3), were the relevant rules to construe the Agencies Convention. Despite Macoun’s submission to the contrary, the Court considered practice that post-dated the introduction of the IOPI Act (at [80]), specifically, decisions of foreign domestic courts and the United Nations Administrative Tribunal (UNAT).  

The Court appeared to treat foreign domestic court decisions as subsequent practice under VCLT Article 31(3)(b). This is not strictly correct. Foreign court decisions do not evidence agreement of all the parties to a treaty. Consistent with the International Law Commission’s 2018 work on subsequent agreement and practice, I suggest that decisions of foreign domestic courts are better understood as a subsidiary means of interpretation under VCLT Article 32.  

The Court’s examination of the UNAT decision was more consistent with the VCLT. Pronouncements of international tribunals, such as the UNAT, are not themselves subsequent practice but may help identify subsequent practice where they reflect, give rise to, or refer to the practice of the parties. The Court’s examination of the UNAT decision correctly focuses on the states’ position recorded in the decision, for example, points agreed or conceded by all parties. 

What can we make of the Court’s discussion of subsequent practice in Macoun? Wall suggests that Macoun is a marked improvement in the Court’s approach to treaty interpretation, possibly explained by the Convention’s subject matter, the extrinsic materials in dispute, composition of the bench and quality of submissions. However, a salient distinction in Macoun is that the Agencies Convention was not incorporated into statute. I re-emphasise that the Court only examined the Agencies Convention to construe the IOPI Act. The Court was not interpreting the Agencies Convention on the basis that it had been incorporated into statute (indeed, it had not been incorporated into statute). This may explain why the Court was so receptive to subsequent practice that post-dated the SAPI Regulations.  

If this distinction is correct, it raises a puzzling question: why can you rely on subsequent practice to understand your international legal obligations, but not to interpret statutes that incorporate those obligations? I accept there may be circumstances where it is difficult to accommodate VCLT Article 31(3) materials into the interpretative exercise, for example, where there is a conflict with Australia’s division of powers. However, I see no other reason to exclude VCLT Article 31(3) materials from the interpretative process.  

The latest development

On 3 November 2021, the High Court handed down its decision in Addy v Commissioner of Taxation [2021] HCA 34 (Addy). Whilst Addy does not resolve the issues addressed above, it is worth a brief discussion as the latest example of treaty interpretation from the High Court.  

In 2015, Addy (a national of the United Kingdom) entered Australia on a working holiday visa. Whilst in Australia, Addy accrued taxable income through casual work and was issued with a notice of assessment by the Commissioner of Taxation (Commissioner). Addy objected to that notice on the basis that it contravened Article 25(1) of the Australia-UK Double Tax Agreement (Australia-UK DTA). Article 25(1) provides that nationals of a Contracting State shall not be subject to ‘other or more burdensome’ taxation in the other Contracting State than is imposed on its own nationals ‘in the same circumstances’. The issue for the Court was whether the tax rates imposed on Addy were more burdensome than what would have been imposed on Australian nationals in the same circumstances.  

It is important to note that the Australia-UK DTA is not incorporated into statute in the same way as occurred in Zentai, Maloney and Adamas. Specifically, the text of the Australia-UK DTA does not form part of domestic statute. Rather, it is listed in the International Tax Agreements Act 1953 (Cth) as an agreement that has the force of law in Australia. Despite this, the Court acknowledged that the same principles applied to interpret the Australia-UK DTA, that is, VLCT Articles 31 and 32 (at [23]).  

Addy is interesting for two reasons. First, the Commissioner relied heavily on a New Zealand Court of Appeal (NZCA) decision, Commissioner of Inland Revenue v United Dominions Trust Ltd [1973] 2 NZLR 555, in which the NZCA interpreted a largely equivalent non-discrimination clause in the New Zealand-UK DTA. The High Court did not discuss the interpretative value of this decision, but simply noted that the Commissioner’s reliance on it was ‘misplaced’ and it was of ‘no assistance in the present case’ (at [27]). 

Secondly, the Court in Addy raised, but did not answer, the issue of whether commentaries that post-dated the Australia-UK DTA could be relied upon to interpret the agreement. Specifically, the Court queried whether a later version of Organisation for Economic Cooperation and Development (OECD) Model Tax Convention commentary could be used, presumably under VCLT Article 32. On this point, I note Gardiner’s view that a later commentary may be recognised as an aid to interpretation where it ‘shows clearly the collective intention of those who drew up the treaty’ (at 403).

Concluding remarks

Subject to any conflict with Australia’s division of powers, I propose that Australian courts should consider VCLT Articles 31(3)(a) and (b) materials in its interpretative exercise. This is for two reasons. First, recourse to subsequent agreement and practice ensures that courts interpret incorporated treaties consistently with the community of nations and Australia’s international obligations. If courts disregard VCLT Articles 31(3)(a) and (b) materials, they may interpret incorporated treaties inconsistently with Australia’s international obligations. As court decisions are ‘always attributable to the State’ (at p.138), such interpretations may trigger Australia’s international legal responsibility. Secondly, recourse to subsequent agreement and practice ensures incorporated treaties are interpreted in a certain and uniform manner. Certainty and uniformity are particularly important for incorporated treaties (such as double taxation agreements) whose object would be ‘frustrated if “they were to be interpreted in a manner which would permit or foster conflicting outcomes”‘ (at [148]). 

In 1997, the late Professor Ivan Shearer stated that the interpretation of incorporated treaties was an ‘important question’ that would ‘likely grow in importance’. He was right. Yet despite the passage of more than two decades, we are no closer to an answer on whether, and to what extent, subsequent agreement and practice may be used to interpret incorporated treaties. The only point that appears settled is that subsequent agreement and practice cannot be used to interpret an incorporated treaty where reliance on those materials would conflict with Australia’s division of powers.

All views are my own and do not necessarily reflect the views of past or present employers. I am grateful for the comments of Co-Editor-in-Chief, Josephine Dooley and Assistant Editor, Sarah Barrie. 

Samuel Saunders is a Legal Officer at the Attorney-General’s Department Office of International Law. The views expressed in this post do not necessarily represent the views of his past or present employers.