The Jadhav Case and the Legal Effect of Non-Registration of Treaties

Those following the legal tangle of the Jadhav Case closely would have noticed India’s (attempted) coup de grâce in its oral submissions regarding the bilateral Agreement on Consular Access of 21 May 2008 between India and Pakistan (“2008 Agreement”, Annex 10 in India’s Application Instituting Proceedings) – that it is unregistered and thus, incapable of being invoked. Pakistan’s oral submissions indicate that this Agreement will form a large part of its case on merits, which in fact, is far stronger than the Indian or Pakistani media give it credit for. Pakistan claims that, irrespective of guilt, the fact of arrest on “political or security grounds” exempts Jadhav from the right of consular access, as per paragraph (vi) of the Agreement, which reads as follows: “In case of arrest, detention or sentence made on political or security grounds, each side may examine the case on merits.” Pakistan interprets this examination “on merits”, as regarding the grant of consular access itself, making it a matter of discretion rather than right.

India met this contention head on in the oral stage, with a two-pronged argument. First, it argued that the 2008 Agreement does not purport to restrict or reduce consular access rights provided by the Vienna Convention on Consular Relations, 1963 (“VCCR”). According to India, the 2008 Agreement is for the purpose of “confirming or supplementing or extending or amplifying” (Art. 73 VCCR) the VCCR rights, to the extent that the Agreement “further[s] the objective of humane treatment of nationals of either country arrested, detained or imprisoned in the other country” (preamble of the 2008 Agreement). To that extent, the first part of the Indian argument is one of interpretation of paragraph (vi) of the 2008 Agreement. The argument is that the Agreement must not be interpreted as exempting those arrested on political or security grounds from consular access since such an interpretation would be contrary to its preamble, to the VCCR, and to the law of treaties, since Art. 41 of the Vienna Convention on the Law of Treaties, 1969 (“VCLT”) permits subsequent bilateral agreements only when they are harmonious with pre-existing multilateral treaties. India has not yet offered a counter-interpretation of paragraph (vi). However, a fair guess is that it will argue that the envisaged “examin[ation]…on merits” is for determining the grant of additional rights conferred by the Agreement (such as immediate release and repatriation) and not for the grant of basic VCCR rights themselves.

However, the second argument, only hinted at by the Indian side, is that the 2008 Agreement is, in any event, inherently inapplicable for two separate reasons: if interpreted as Pakistan claims, the bilateral agreement derogates from the effective execution of object and purpose of the earlier, multilateral VCCR, rendering it invalid under Article 41 (read with 42) of the VCLT; and, it has not been registered with the United Nations Secretariat, making it incapable of being invoked before the International Court of Justice.

It is this second argument regarding the invalidity of the Agreement that is the subject of scrutiny here. The legal effect of non-registration of a treaty has not been the subject of significant discussion amongst States, within the UN or even within academic circles. The pertinent legal provision is Article 102 of the Charter of the UN, which, in paragraph (1) mandates registration of all treaties, and in paragraph (2) imposes an additional sanction to incentivise registration – that of being unable to invoke an unregistered treaty before any UN organ. To this extent, this is a remarkable article – most of the Charter’s articles impose obligations upon Member States, but very few of them back this with a specific sanction, in addition to the general breach of international law that violating the Charter would entail [See Hans Kelsen, The Law of the United Nations, Chapter 18 – Sanctions (7^th edn., 2008)]. The reason for such additional punitive force can be traced to its drafting at the UN Conference on International Organizations at San Francisco in 1945, where it was observed that “the basic purpose of the obligation was to prevent secret treaties” [UNCIO Documents, vol. 87, p. 26 cited in Brandon, 29 Brit.Y.B.Int’l L. 186, 196 (1952)]. In fact, the equivalent Article 18 of the Covenant of the League of Nations had gone further, stating that “[n]o such Treaty or International Engagement shall be binding until so registered”. This was watered down when the Charter was being drafted, not least because it led to ambiguity as to how such a treaty, otherwise valid, was to be interpreted outside of the League system – was a non-League body, such as, for instance, the Permanent Court of Arbitration (“PCA”) also not to apply such a treaty? The UN system sought to avert such confusion by providing instead for “relative, not an absolute invalidation” of unregistered treaties [Kelsen, p.722]. Article 102 was worded in such a way as to render the handicap operative only before UN organs. This, it was thought, would create a self-contained system – an offence within the UN, penalised within the UN alone.

However, this raised its own set of doubts and ambiguities – would the same treaty now be valid and operative before the PCA, but not before the ICJ? When challenged simultaneously in different courts, could the treaty’s validity, much like Schrodinger’s cat, exist and not exist at the same time? The simple answer to this question is yes. This is exactly what “relative invalidity” envisages. A more nuanced answer would be that the question is not one of validity as much as operability. Even within the UN system, the validity and binding nature of the treaty is not denied – the Court has stated as much on several occasions (in Qatar v. Bahrain, ¶29, for instance); it is merely its enforceability before that forum that is affected. This handicap is considered necessary to incentivise registration while not resulting in a complete denial of rights since other (non-UN) fora remain available. In any event, there is nothing to bar the party from now registering the treaty and then invoking it before a UN organ [there is difference of juristic opinion on whether this can be done while an organ is seised of the matter or if fresh proceedings must be instituted. Compare Brandon, 200 with Robert Kolb, The International Court of Justice, 543 (2013)]. In other words, Article 102(2) is a procedural hurdle (as opposed to a bar).

A last important feature of Article 102(2) is that it only bars the parties to the unregistered treaty themselves from invoking it; neither third parties nor, importantly, the Court itself is barred from relying on an unregistered agreement. In fact, most authors concur that the UN organs themselves are not obliged to enquire into registration. They may, on their own initiative, and they must, if raised as an objection, disallow parties from invoking unregistered treaties; however, nothing bars them from taking cognizance of such treaties irrespective of parties’ invocation. After all, Article 102(2) was meant as a sanction upon parties concluding “secret treaties”, not as a restriction on the organs of the UN in carrying out their functions [Martens, Ch.XVI Miscellaneous Provisions, Article 102, ¶45, 51 in The Charter of the United Nations: A Commentary, Vol. II (Simma et al. eds., 3^rd edn., 2012)].

Thus, given how this provision is only a procedural hurdle and how it exists only to reinforce the substantive law laid down in Article 102(1), it is unsurprising that the organs of the UN have overwhelmingly chosen to disregard the unregistered status of treaties, presumably considering the substantive issues at stake to be more important than reinforcing the importance of registration. The ICJ in Qatar v. Bahrain, allowed invocation of the unregistered 1987 double Exchange of Letters, which was agreed between the parties to be a treaty. In Corfu Channel, the ICJ accepted jurisdiction under an unregistered Special Agreement and even permitted the United Kingdom to cite, in its defence, an unregistered agreement dated 22 November 1945 [Martens, ¶56; Brandon, 199]. In the Aegean Sea Continental Shelf case, the effect of Article 102 upon the unregistered accord verbal of 31 May 1975 was questioned but it was nevertheless considered and rejected on merits. Even the Permanent Court, bound by the more onerous Article 18 of the Covenant, in Eastern Greenland, found the unregistered Ihlen Declaration of 22 July 1919 to be “binding” despite the express wording of Article 18. Even before the General Assembly, unregistered treaties have been invoked without objection –the USSR relied on an unregistered 1950 treaty with China before the First Committee [GAOR (VI), 1^st Committee, 502^nd mtg. 26 January 1952 cited in Martens, ¶53]. It is evident that the unregistered nature of treaties often passes unnoticed. In fact, there is no recorded instance of a UN organ disallowing invocation of a treaty solely on account of non-registration. Authors are not wrong in concluding that Article 102 is “more honoured in the breach than the observance” [Anthony Aust, Modern Treaty Law and Practice (2000), 280; Martens, ¶57]

What sets this case apart from those in the past is, of course, that the unregistered status has not slipped by unnoticed – it has been actively brought up as an objection by one of the parties. In this case, turning a blind eye is out of the question. The Court is in fact mandated to apply the sanction; however, whether it considers itself able (and whether it chooses) to go over and above this sanction to consider the 2008 Agreement on its own initiative remains to be seen. Given the scant notice paid to Article 102 in the past, there is every reason to believe it will do so. In any case, India will do well to formulate an enunciated counter-interpretation of the 2008 Agreement rather than to count upon its inability to be invoked.