Antonios Tzanakopoulos is Lecturer in Public International Law at the University of Glasgow. Many thanks are due to Christian Tams, Marko Milanović, and Dapo Akande for their comments. The usual disclaimer applies.
In the aftermath of the ECJ’s Kadi decision, which annulled the EC Regulation implementing the 1267 sanctions regime against Mr Kadi and the Al Barakaat Foundation, Kadi was almost immediately relisted by the Council of the EU in a new Regulation. This subjected him afresh to the restrictive regime of SCRs 1267 (1999) et seq, most recently SCR 1904 (2009). And, as Devika Hovell reported on this blog, almost immediately Kadi brought a fresh challenge against that Regulation before the CFI, now renamed as the ‘General Court of the EU’ after the entry into force of the Lisbon Treaty. On 30 September, the General Court rendered its decision in Kadi II.
EJIL:Talk! regular readers will know that we have consistently reported on challenges to the 1267 regime before national and regional courts on this blog (see eg here, here, here, here, here, and here). In Kadi II, the General Court grudgingly follows the ECJ’s reasoning in Kadi I and confirms a trend of defiance of Security Council sanctions. In this post I will try to situate the Kadi II decision in the context of challenges to Security Council restrictive measures under Article 41 of the UN Charter.
I. The Story So Far
In SCR 1267 (1999) the Security Council imposed targeted sanctions on individuals and legal entities associated with bin Laden, Al-Qaida, and the Taliban. After the fall of the Taliban in Afghanistan in 2001, the sanctions remained in place, evolving into a general anti-terrorist sanctions regime, complemented by the regime imposed under SCR 1373 (2001). One of the main differences between the two regimes is that under 1267 those to be sanctioned are determined by the Security Council itself (through its Sanctions Committee which maintains a Consolidated List to that effect), while under 1373 it is UN member states that draw up lists of those who are to be subjected to the sanctions, notably (in both cases) asset freezes and travel bans.
The 1267 sanctions regime, in its original incarnation, imposed very far-reaching restrictions on individuals and legal entities, and allowed for no humanitarian exemptions. Further, those listed had no way of challenging the restrictive measures imposed against them, except through requesting diplomatic protection and having representations made on their behalf by their State of nationality or residence before the Security Council. This was in part justified by the allegedly ‘temporary’ and ‘preventive’ nature of the restrictions. Following some early reactions to the 1267 regime, which led to challenges being raised against it before domestic (see Nada) and regional courts (see Kadi I), even with some success (see Othman  EWHC Admin 1022 before the English High Court), the Security Council decided to introduce some humanitarian exemptions for basic expenses and subsistence (SCR 1452 (2002)), and then also to introduce a Focal Point, where those targeted by the sanctions regime could petition the Security Council, in their own name and not through their State of residence or nationality, to be removed from the list (SCR 1730 (2006)).
However, this did little to appease both domestic constituencies and domestic and regional courts, which were faced with a sanctions regime over which the Security Council had full control and seemingly unlimited discretion, be it with respect of who was to be listed, the reasons for such listing, whether exemptions would be granted, and what one had to do to be removed from the list. A new wave of attacks against the measures in domestic courts led to annulments of domestic measures implementing the 1267 sanctions (see comments on Hay) or to innovative interpretations stretching the language of the relevant Security Council resolutions (see comments on Abdelrazik).
In response, the Security Council adopted, at the very end of 2009, Resolution 1904. This was clearly an attempt to thwart the ‘challenges, both legal and otherwise, to the measures implemented by Member States’ under the 1267 sanctions regime, as well as to make procedures for listing and delisting by the competent Sanctions Committee ‘fair and clear’, an ongoing effort (at preamb 9). However, the establishment of an Office of the Ombudsperson by SCR 1904 (2009) still did not satisfy domestic and regional courts, which keep pushing for judicial guarantees in the imposition of asset freezes and travel bans on persons identified by the Security Council as being ‘associated with’ the Taliban or Al-Qaida (see comments on the UK Supreme Court decision in HM Treasury v Ahmed and ors). In Kadi II, the General Court remains unconvinced that the procedure at UN level offers the appropriate guarantees, and annuls the new Regulation subjecting Kadi to the restrictive regime.
II. Applying ECJ Kadi (I)
In Kadi II the General Court grudgingly applies the ECJ’s Kadi. In following the language of Tim Stahlberg in this post over at the ECJBlog.com, I say ‘grudgingly’, because it is apparent throughout the judgment that the General Court is not completely comfortable with the ECJ’s reasoning in Kadi I. For example, while the General Court acknowledges the different scope of the international obligations imposed on member states of the UN under the 1267 and 1373 sanctions regime respectively (at paras 32-33), it accepts the ECJ’s obliteration of this distinction later on (at paras 138-139). That it does so without being fully convinced is evident both as implicit in the language used by the Court (eg para 41: ‘Notwithstanding Articles 25 and 103 of the Charter of the United Nations …, and although it observed … that observance [sic] of the undertakings given in the context of the United Nations was required when the Community gave effect to resolutions adopted by the Security Council under Chapter VII of the Charter, the Court of Justice asserted …’) and explicit when the General Court lists the significant criticisms leveled by scholarship with respect to ECJ Kadi (paras 113-120). The Court goes on to acknowledge ‘that those criticisms are not entirely without foundation’ (para 121). But given that Kadi I was rendered by the ECJ in Grand Chamber formation, ie with the intent of setting down certain principles (ibid), ‘in principle it falls not to [the General Court] but to the Court of Justice to reverse the precedent’ (para 123).
Even if it merely—and grudgingly—follows ECJ Kadi, Kadi II is still an important decision rather than a mere application of precedent, and this not just because it confirms a trend while simultaneously leaving its eventual reversal open. Rather, it is important because it elaborates on two points that were detectable as an undercurrent in both ECJ Kadi, and in other decisions of national courts on 1267 sanctions. The first one refers to the adoption of a ‘Solange argument’, and the other to the overlap between certain rights guaranteed both by domestic constitutions and under (general) international law. In this connection, it raises anew a number of issues which I have tried to systematize. In what follows, I will discuss Kadi II from the perspective of attribution of acts to a state implementing the binding decision of an international organization (section III); from the perspective of the relationship between legal orders and the methods for the regulation of their interaction (section IV); and finally from the perspective of the implications of decisions that rely on domestic law to effectively disobey international obligations for the normative hierarchy of rules in international law—if any (section V); all this with apologies for the long post.
III. Normative Control, Discretion, and ‘Judicial Review’
When a binding decision of an international organization leaves to its member states—as the ‘agents of execution’—no margin of discretion as to its implementation, ie when it imposes a strict obligation, member states are under the effective normative control of the organization. Notwithstanding any arguments for direct attribution of such implementing acts to the organization promulgating the decision (and thus controlling the conduct of member states), it remains a possibility that these implementing acts will be concurrently attributed to the member states (Art 4 of the Articles on State Responsibility). This has been accepted in practice by courts (eg the ECtHR in a number of cases such as Bosphorus, the CFI and ECJ in Kadi I, the Canadian Federal Court in Abdelrazik, the UK courts in Hay and HM Treasury v Ahmed and ors, and others) (see further on normative control and attribution chapter 2.II in Disobeying the Security Council [forthcoming]).
However, the fact remains that in such circumstances state conduct is conditioned by the decision of the international organization, so that any review of state conduct will uno actu also constitute review of the conduct of the organization. There are many ways in which courts have purported to deal with this situation: the ECtHR has established a doctrine of equivalence, whereby if the international organization promulgating the act requiring strict compliance protects human rights on a level equivalent to that under the ECHR, the legality of state conduct taken in implementation is presumed. While the presumption can theoretically be rebutted, this has not happened so far. The CFI in Kadi I acknowledged the predicament of having to review Security Council conduct when reviewing member state or Community conduct in implementation, and opted for reviewing against the lowest common denominator, ie law binding both on the Community and on the UN when acting through the Security Council in Chapter VII mode, which it famously considered to be only that part of international law which has reached the status of jus cogens. The ECJ in Kadi I, on the other hand, along with the UK Supreme Court, radically rejected the connection between the international and the domestic/regional implementing measure, and went ahead to fully review the domestic measure for compliance with domestic law (see here for further comment).
In Kadi II, the Community institutions and the intervening states tried to re-invoke the argument that there was nothing they could do except blacklist Kadi, since their conduct was effectively conditioned by the binding Security Council measure, and they had no margin of discretion in the measure’s implementation. This time, they tried to present this as an argument in favour of marginal review of the impugned Regulation. The argument developed along the following lines (paras 82-111): the ECJ says it can review the Regulation implementing 1267 sanctions; but there’s nothing we can do if the Security Council wants Kadi blacklisted, except communicate to him the summary of the reasons for the listing (generalities and allegations for the most part), give him an opportunity to be heard (of no consequence, as the attempt to question generalities is unlikely to yield results anyway), and then go on and blacklist him as per the Security Council’s command. If EU courts are going to review this, they have yet to establish a standard of review; we argue for the most marginal review to account for our lack of discretion (manifest error or abuse of power).
The General Court does acknowledge that the 1267 regime imposes strict obligations, as opposed to the 1373 regime, which does not (paras 32-33), a distinction it had drawn already in OMPI (at paras 100-102) and other cases. But it then concedes that accepting the arguments of the institutions would be tantamount to reiterating its own (CFI) finding in Kadi I, which has already been reversed by the ECJ (paras 121, 123). It then follows the ECJ in rejecting the argument for marginal review, disassociating the international from the domestic measure, and engaging in ‘full review’ as required by ECJ Kadi (paras 126 seq).
IV. Relationship between Legal Orders, Multi-level Governance and Solange
In situations where strictly binding (‘sovereign’ or ‘governmental’) decisions can be made at various levels of governance (international, regional, domestic), there is significant potential for conflict between those various levels (and the partial legal orders they represent). This is the case, in particular, when the power to impose certain restrictive measures on individuals has been conferred (whether delegated or transferred—see generally Dan Sarooshi’s International Organizations and their Exercise of Sovereign Powers) from one level to another, without similar safeguards being put on the exercise of the conferred power.
When a relevant case comes before the court of the legal order that has conferred the relevant power, a reaction is to be expected. This has so far generally taken the shape of the ‘Solange argument’, enunciated by the German Federal Constitutional Court when attempting to define the relationship between German constitutional law and acts of the European (Economic as it then was) Community, first in the 1970s (Solange I) and then modified in the 1980s (Solange II). A (much simplified) version of the argument runs thus: for as long as the exercise of the conferred powers on the (different or higher) level of governance takes place without safeguards similar to those to which it was subject before the conferral, when it was still exercised at the (lower) level, the court of the (lower) level will review the act of the (higher) level for compliance with the safeguards at the (lower) level (Solange I). When such safeguards are adopted at the (higher) level, the court at the (lower) level will desist, presuming conformity (Solange II).
The Solange argument has been adopted by the ECtHR in its Solange II incarnation (through the immediate acceptance of existence of equivalent protection; in this respect one should be cautious and mention that the Solange argument cuts both ways, as it may serve not as a tool for resistance-as Solange I, but also as a method to water down human rights protection-as Solange II. Also, the Solange argument presumes autonomous (if only so self-proclaimed) legal orders, while the ECtHR has not proclaimed the ECHR to constitute such an autonomous legal order. On these points see this paper by Marko). Its impact, this time as Solange I, was also clear both in the UK Supreme Court’s HM Treasury v Ahmed and ors and in the ECJ’s Kadi (see here for further comment). In Kadi II the General Court is as explicit as it could be: it states that it must ensure the ‘full review’ of the domestic implementing measure for compliance with fundamental rights (guaranteed under Community law), ‘without affording [the measure] any immunity from jurisdiction on the ground that it gives effect to resolutions adopted by the Security Council under Chapter VII of the Charter of the United Nations’ (para 126). ‘That must remain the case’, the Court continues, ‘at the very least, so long as (=solange) the re-examination procedure operated by the Sanctions Committee clearly fails to offer guarantees of effective judicial protection’ (para 127).
The Court further engages with the regulation of multi-level governance in paras 186-187. In response to the institutions’ argument that they have adopted with respect to Kadi the very same safeguards that the Community courts have found adequate in the OMPI line of cases (para 185), the General Court draws a distinction between the two regimes of multi-level governance: the OMPI regime (ie the 1373 sanctions regime) is structured in two tiers, one national and one regional (Community): the provision of safeguards at the national levels relieves Community institutions from the obligation to impose fresh safeguards relating to the same subject-matter on Community level (para 186). Conversely, the Kadi regime (ie the 1267 sanctions regime) is structured in two tiers as well, but here one is international (UN) and the other regional (Community): for so long as there are no safeguards on the UN level, the Community will have to provide them (para 187).
There is surely something to be said about the General Court accepting—however half-heartedly—the obliteration of the distinction between 1267 and 1373 when it comes to the standard of review (see section III above), and then re-introducing it, albeit in another guise, when considering the merits of the claim for the violation of the right to effective judicial protection. Obviously the only reason that safeguards can be adopted at any level under the 1373 regime is because the identification of those to be subjected to restrictive measures is left to the discretion of UN member states, unlike in the case of the 1267 regime, where the targets of the measures are designated by the Security Council, leaving UN member states no room for maneuver.
Be that as it may, Kadi II confirms that the EU courts have adopted the medicine they were served by the Bundesverfassungsgericht in Solange I in their relationships towards the United Nations. And it also confirms the ability of the Solange argument to furnish a powerful tool of resistance, and potentially a powerful incentive, for the regulation of the relationship between different levels of governance. However, the most important aspect from an international law perspective, that of the qualification and the implications of the reaction, remains to be broached.
V. Disobedience and Implications for Normative Hierarchy
From the perspective of international law, the annulment of the domestic implementing measures (in casu the Regulation) clearly results in the breach of the international obligation of member states under Article 25 of the UN Charter. This is in fact raised by the EU institutions and the intervening member states in Kadi II (paras 92-93, 100, 109) and implicitly recognized by the General Court (paras 41, 115, 119). The annulment forces member states to disobey the Security Council decision, lest they disobey the decision of their own court(s).
The Solange argument, in and of itself, cannot provide any justification for this disobedience under international law. This is in particular because the decision to disobey is based on domestic law considerations: here on primary EU law on fundamental rights, otherwise (eg in the UK Supreme Court’s Ahmed and ors) on constitutionally protected fundamental rights. However, it can be argued that the fundamental rights customarily relied upon in Solange-type argumentation (leading to disobedience) are not solely guaranteed under domestic law, but are homonymous to internationally protected human rights. This is the case in particular for the right to effective judicial review, otherwise cast as an aspect of the right to a fair trial.
It would be impossible to fully set out all the possible ramifications of this argument here. In Disobeying the Security Council (chapter 7), I argue in part that court-imposed disobedience could qualify as a countermeasure against the Security Council’s (ie the UN’s) wrongful imposition of sanctions. In a paper on normative hierarchy in international law I consider whether this court-imposed practice of disobedience could constitute practice and opinio juris for either considering certain rights heretofore not considered jus cogens as having reached that status, or alternatively as having the potential of introducing a new hierarchical level in the rudimentary hierarchical structure of international law.
For either of these two arguments to have any hope, it must be shown that there is some correspondence between the rules (rights) claimed under domestic law and those existing under international law. And while the argument can be made in abstracto, the General Court in Kadi II puts it in no uncertain terms, even if rather implicitly. At para 150, when discussing the impact of Security Council sanctions on targeted individuals, the Court refers to the UN High Commissioner for Human Rights, who in UN Doc A/HRC/12/22 (2009) at 15 para 42 qualifies these as punitive, requiring either imposition by a judicial decision or judicial review. At paras 176-177 the Court adopts the criteria elaborated by the ECtHR on the application of Article 5(4) ECHR and proceeds to apply them. While this is not groundbreaking with respect to the way EU courts understand fundamental rights under primary EU law, it is worth noting that Article 9(4) ICCPR is virtually identical to Article 5(4) ECHR. In this, the General Court exemplifies the correspondence in substance between (some) human rights that are both domestically protected and internationally guaranteed.
VI. (Interim) Conclusion
Kadi II then further supports the argument that domestic and regional court reactions to Security Council sanctions are not necessarily motivated by (or need not necessarily be understood as being motivated by) some sort of delusion of grandeur or ‘constitutional hegemonic aspirations’ (see this paper by Tridimas and Gutierrez-Fons). Rather, they can also be cast as crucial elements of practice and opinio juris for the further development of international law, so long as (pun intended) one draws the requisite parallelism between the domestic law relied on and the homonymous international law. It remains to be seen how the Security Council will respond to this new challenge; although it has been slow to do so, it has always sought to find a way to accommodate member state concerns, eg by establishing the Focal Point, or more recently the Office of the Ombudsperson. Perhaps it will now be forced by the strong decentralized reactions to establish a judicial process for the determination of designations under the 1267 sanctions regime, or decide to allow decentralized determinations by the member states as in the case of 1373.