Secondary Sanctions: A Weapon Out of Control? Part II: The legality of secondary sanctions under conventional law and the IMF’s tacit approval procedure for payment restrictions inspired by security concerns

The legality of so-called ‘non-UN’ or ‘autonomous’ sanctions has been amply debated in recent years. Discussion has arisen, for instance, on their compliance with the principle of non-intervention (see here), or their potential qualification as ‘third-party countermeasures’ (see e.g. here and here). The legal challenges are further compounded when States resort to ‘secondary’ sanctions that seek to restrict trade relations between the primary sanctions target, on the one hand, and third States or third-country operators, on the other hand.

Most strikingly, secondary sanctions intuitively appear to be at odds with the restrictions on the exercise of State jurisdiction under customary international law. Our first post in this secondary sanctions triptych has argued, however, that a more fine-grained analysis is due, necessitating a distinction between so-called access restrictions – say, a prohibition to enter the US market for an EU company selling airplane parts to Iran – and restrictions that may result in civil or criminal penalties on the part of non-compliant non-US companies.

Secondary sanctions and conventional law

Leaving aside the jurisdictional aspect, how do secondary sanctions fare under conventional law? The EU has repeatedly taken the position that US secondary sanctions breach WTO law (see e.g. here). Following the adoption of the Helms-Burton Act in 1996, the then EC even initiated a procedure before the WTO Dispute Settlement Body, although it pulled the plug after the US agreed not to enforce the Act against EU persons (that is, until the Trump Administration decided to reactivate Title III of the Helms-Burton Act). At first sight, secondary sanctions also seem to contravene various bilateral treaties entered into by the United States, including treaties of Friendship, Commerce and Navigation (FCN) concluded with Western European countries, or the more recent bilateral investment treaties (BITs) concluded with many Eastern European countries. Two reservations are nonetheless in order.

First, while it is a fairly straightforward exercise to establish that primary sanctions of a sectoral nature breach treaty obligations applicable between the sanctions sender and the target State – as illustrated by the WTO Panel report in case DS512 Russia-measures concerning traffic in transit (here) – it is far more difficult to find an infringement of WTO law or of bilateral BIT/FCN treaties with respect to secondary sanctions. Secondary sanctions are generally unlikely to contravene the national treatment or most-favoured nation (MFN) principles (see e.g. here). In a similar vein, construing a secondary boycott as a quantitative restriction in the sense of Article XI(1) GATT feels like trying to fit a square peg into a round hole. That is not to say that no treaty violations can result. A number of ‘menu-based’ sanctions employed by the US (e.g., denials of import permits, denials of entry for corporate officers (see e.g. here), or exclusion from government procurement) effectively appear contrary to specific WTO rules, including Article XI(1) GATT, the Revised Agreement on Government Procurement (GPA), or the GATS ‘mode 4’ commitments on movement of persons. Similar observations can be made mutatis mutandis in respect of bilateral BIT and FCN treaties.

Second, even when specific sanctions are incompatible with WTO agreement and/or bilateral treaties, these instruments all tend to feature a so-called ‘security exception’ (although the exact wording may differ). Such clauses are not purely self-judging, as was asserted by the ICJ in Nicaragua and Oil Platforms, as well as by the WTO panel in case DS512. Yet, they undeniably pose a high bar to contest sanctions supposedly inspired by national security concerns and other foreign policy objectives. Importantly, however, it could be argued that secondary sanctions must fail the test of Article XXI(b) GATT and other conventional security clauses, even where primary sanctions might satisfy such test. The reason is that ‘[the] threat posed by the [primary] target’s trading partners … is necessarily too attenuated and remote’ to satisfy the necessity test inherent to such security exceptions (here, see also here). An analogy with the treatment of self-help in international law more generally supports the idea that the ‘secondary’ dimension of economic sanctions is problematic from a necessity perspective (see e.g. Article 49(1) of the ILC Articles on State Responsibility).

Secondary sanctions and international monetary law 

Another relevant agreement – one that does not feature a security exception – is the International Monetary Fund’s (IMF) Articles of Agreement. Article VIII(2)(a) indeed provides that ‘no member shall, without the approval of the Fund, impose restrictions on the making of payments and transfers for current international transactions’. This includes all ‘payments due in connection with foreign trade’ (Article XXX), while the term restriction presupposes a ‘direct governmental limitation on the availability or use of exchange’ (1960 IMF Decision). The language of Article VIII(2)(a) is broad enough to cover not only targeted sanctions involving the freezing of bank accounts, but also a wide range of financial restrictions that seek to prohibit dollar-denominated transactions passing through a US correspondent bank account where there is some link with the primary sanctions target.

As none of the exceptions provided for in Articles VII(3)(a) and XIV(2) are relevant in the present context, it would be tempting to conclude that secondary sanctions linked to the use of the dollar are prohibited by Article VIII(2)(a). There is, however, more than meets the eye. In particular, in 1952 the IMF Executive Board adopted a Decision introducing a specific procedure for the granting of IMF approval for restrictions imposed on security grounds. The Decision crucially provides that restrictions must be notified to the IMF in advance (or within a period of 30 days in cases of urgency). If the IMF does not object within 30 days, the restrictions are tacitly approved.

In practice, it is clear from the IMF’s Annual Reports on Exchange Arrangements and Exchange Restrictions (here) that numerous States have notified payment restrictions supposedly inspired by security concerns. What is more, although the application of the Decision remains shrouded in mystery, it would seem that no notification was ever objected to by the Fund (here). There thus appears to be general policy of non-objection. Consequently, to the extent the US authorities took the effort to properly notify secondary sanctions, the conclusion must be that their legality cannot be challenged on the basis of Article VIII(2)(a) of the IMF Articles of Agreement.

The lack of scholarly attention on the 1952 IMF Decision and its implementation is all the more striking considering that its impact potentially stretches beyond the IMF domain. Other clauses on payment restrictions such as Article XI(1) GATS or Article 10(2) of the US-Belgium FCN Treaty indeed tend to explicitly leave room for restrictions that are ‘authorized’ or ‘requested’ by the IMF. Whether mere non-objection by the IMF can be equated to an ‘authorization’ or ‘request’ for purposes of the latter treaty provisions remains an open question, which is as yet untested in judicial practice.

Irrespective of this latter question, the escalating use of secondary sanctions in the financial domain, specifically the United States’ increasing ‘weaponization’ of the US dollar, should give IMF Members pause and provide impetus to revisit the obscure 1952 IMF Decision, or at least to end the general policy of non-objection. While a 1981 proposal to admit only measures in line with UN Security Council resolutions was never debated within the IMF Executive Board (here), perhaps the time is ripe to draw the line at secondary sanctions conditioned on the use of a State’s currency.