In a startling interlocutory decision of February 2011 (‘Decision’), the Appeals Chamber of the United Nations (‘UN’) Special Tribunal for Lebanon purported to identify an extant customary international crime of transnational terrorism, and applied it in interpreting domestic terrorism offences under Lebanese law.Footnote 1 Until that Decision, the prevailing wisdom was that, in expressly stipulating Lebanese criminal law as the applicable substantive law of the Tribunal, the Statute of the Tribunal intended only ‘pure’ Lebanese criminal offences to apply to terrorist acts in Lebanon in 2005.
The decision astonishes not so much because both the prosecution and the defence agreed that terrorism was not a customary crime. After all, one might accept an inquisitorial mode of getting at the truth, in contrast to an adversarial approach, which would have confined the issues to the parties’ submissions. Nor does the decision confound because the majority of states and scholars do not recognize a customary crime of terrorism, with the prominent exception of one eminent jurist, Antonio Cassese, who happened to be President of the Appeals Chamber. One may accept that President Cassese had an open mind on the issue, notwithstanding his frequent extra-curial writings since 2001 arguing that terrorism is already a crime in customary international law.Footnote 2
Rather, the Decision is astounding because its conclusion has scant empirical grounding in state practice, its reasoning is poorly substantiated, and it ultimately plays fast and loose with custom formation. It provides an excellent example of why, as Sir Michael Wood suggests,Footnote 3 the International Law Commission should study the international law on customary law – in the hope of refining the rules to restrain haphazard analyses of custom such as this. In fact, as this article shows, there simply is no crime of transnational terrorism in customary international law.
The Decision is also startling because it is not an instance of the ordinary kind of incremental judicial activism that necessarily tailors the law to novel circumstances. Rather, it is an example of a judiciary transforming itself into a global legislature, creating entirely new law and exceeding the accepted bounds of the judicial function. In a fit of disguised legislative activism, it invented a new and post facto international criminal liability for terrorism, resulting in the radical expansion of liability under Lebanese criminal law as it was understood in 2005.
Along the way, the Appeals Chamber subjugated the human rights of potential defendants (particularly freedom from retroactive criminal punishment) to its own moralizing conception of what the law ought to be. The Appeals Chamber may well be right that the conduct it identified should be an international crime. It may indeed be abhorrent that a wider net of terrorist conspirators in Lebanon could otherwise get away with – on a positivist technicality – assassinating a former Lebanese prime minister or killing many others. One might even argue on grounds of natural law, channelling Nuremberg, that no competent person in Lebanon in 2005 could seriously have believed that only terrorist bombing was a terrorist crime, but using a machine gun or knife was not. Even so, the lex ferenda should not be misrepresented by judges as the lex lata, however compelling the moral case for punishing the wicked, and even where the positive law then in force turns out to be deficient.
This article analyses the Appeals Chamber's assessment of customary international law and critiques both its analysis of material sources and its substantive conclusions. It focuses only on the finding that terrorism is a customary international crime and does not address other issues in the Decision (including extended modes of criminal participation and other crimes under Lebanese law). The article demonstrates that every category of source relied upon by the Appeals Chamber – national legislation, judicial decisions, regional and international treaties, and UN resolutions – was misinterpreted, exaggerated, or erroneously applied. It also shows that key sources that undermine the Appeals Chamber's findings on customary law were hastily dismissed, trivialized, or ignored altogether by it. In short, it exposes a fatally incorrect decision based on poor reasoning and a laissez-faire attitude towards both criminal liability and custom formation. The conclusion of the article explains some of the broader implications of such a flawed decision, including the damage to human rights and to public confidence in international criminal justice.
1. The interlocutory decision
The Appeals Chamber identified a peace-time customary international-law crime of terrorism consisting of three elements:
(i) the perpetration of a criminal act (such as murder, kidnapping, hostage-taking, arson, and so on), or threatening such an act; (ii) the intent to spread fear among the population (which would generally entail the creation of public danger) or directly or indirectly coerce a national or international authority to take some action, or to refrain from taking it; (iii) when the act involves a transnational element.Footnote 4
The requirement of a transnational elementFootnote 5 rules out purely domestic terrorism. While the Tribunal recognized only peace-time terrorism as a crime, it indicated that ‘a broader norm that would outlaw terrorist acts during times of armed conflict may also be emerging’.Footnote 6 Problematically, the Decision does not specify the date at which such a crime supposedly crystallized. Given that the Tribunal is principally dealing with conduct committed in February 2005, it may be assumed that the Tribunal is suggesting that the crime existed by that date.
The Appeals Chamber did not seek to directly apply international criminal liability for terrorism in Lebanon,Footnote 7 but rather sought to interpret and apply Lebanese law to crimes of terrorism consistently with the identified international legal standards on terrorism.Footnote 8 Those standards were held to include both the customary international crime of terrorism as well as the Arab League Convention on the Suppression of Terrorism, to which Lebanon is a party. Such an interpretive approach was not mandated by international law itself, but rather by the Appeals Chamber's understanding of how Lebanese domestic law uses international law in the interpretation of domestic law, including criminal offences.
The effect of the Appeals Chamber's finding is that the scope of criminal liability explicitly imposed by the Lebanese Criminal Code is expanded to encompass wider elements of liability provided under international law. In particular, whereas the Lebanese law, on its face, recognizes only certain stipulated methods of perpetrating terrorism (such as bombs), additional unspecified methods (such as guns or knives and so forth) are now recognized by importing the international customary crime (which is said not to be limited to enumerated means or methods).
Whereas the Statute of the UN Special Tribunal for Lebanon expressly stipulates that the Lebanese Criminal Code is the applicable substantive law, the Appeals Chamber's Decision enables that law to be interpretively supplemented by international criminal liabilities. This no doubt came as a surprise to the drafters of the Statute, given that a decision was taken at that time to confine liability to offences under purely domestic law. It no doubt also surprises people in Lebanon, who, in 2005, could have had no idea that terrorism was a customary international-law crime: no Lebanese or international court had ever said so. While the Appeals Chamber asserted that it was merely ‘interpreting’ Lebanese law in light of international anti-terrorism law, that claim is disingenuous in this criminal context. The effect of the Decision is really to criminalize new conduct and thereby to import a new offence into Lebanese law.
Whether Lebanese domestic law does in fact utilize international law in interpreting domestic law in the manner suggested by the Appeals Chamber is an expert question of Lebanese law on which no informed opinion is expressed here. However, the Appeals Chamber also hinted that this approach:
is in accord with a general principle of interpretation common to most States of the world: the principle that one should construe the national legislation of a State in such a manner as to align it as much as possible to international legal standards binding on the State.Footnote 9
That claim is contestable. It is true that many states draw upon international law to aid in domestic legal interpretation. However, that occurs in a variety of different ways and contexts in different national systems, and there simply does not exist a uniform principle of the kind suggested that would necessarily produce the result found by the Decision. The various usages in national legal orders are more nuanced. Some states use international law in interpreting domestic provisions that implement treaty obligations. Some states draw on international law only where there is ambiguity in domestic law, but not where the domestic law is clear or settled. Some states invoke international law for protective purposes, to read down excessive or invasive domestic laws in the light of human-rights protections, but not to impose burdens (especially new criminal liabilities) on individuals.
Such interpretive uses of international law are quite different from the Appeals Chamber's approach, which: (i) does not concern a domestic provision that implements a treaty; (ii) does not involve ambiguity, for there was none identified in the Lebanese domestic offence; and (iii) does not invoke international law to protect an individual's rights, but rather to widen their criminal liability. In effect, the Appeals Chamber came close to suggesting that most states are presumptively monist, regardless of whether a particular national legal order in fact requires legislative incorporation of a treaty or customary rule for it to have domestic effect (and is therefore closer to being dualist, particularly as regards criminal liability). Various national courts have explicitly recognized that customary crimes especially have no domestic legal effect without legislative action, partly because of policy concerns about vagueness and unfairness to an accused.Footnote 10
The most controversial aspect of the Appeals Chamber's Decision, however, is its assessment of customary international law. It invoked various supposedly converging material sources to support its contention that there now exists a customary international crime of transnational terrorism: anti-terrorism treaties of regional organizations;Footnote 11 a pattern of General Assembly resolutions since 1994 that condemn terrorism as criminal;Footnote 12 UN Security Council Resolution 1566 (2004), which defines terrorism; the draft definition in the UN Draft Comprehensive Convention under negotiation since 2000; the Terrorist Financing Convention 1999, said to provide ‘the UN's clearest definition of terrorism’;Footnote 13 numerous national laws; and nine judicial decisions from various national courts. The remainder of this article scrutinizes the Appeals Chamber's flawed use of these sources.
2. National laws
The Appeals Chamber (correctly) observed that ‘[c]onsistent national legislation can be another important source of law indicative of the emergence of a customary rule’.Footnote 14 A deceptively impressive array of 37 national laws is then cited by the Appeals Chamber in support of its assertion that ‘the national legislation of countries around the world consistently defines terrorism in similar if not identical terms to those used in the international instruments’.Footnote 15
Yet, for at least five reasons, the Appeals Chamber's analysis of the supposed ‘concordance’ of national laws is far too blunt to substantiate its conclusion about customary law. First, the Appeals Chamber made no effort to differentiate between those national laws that address domestic terrorism and those that concern international terrorism. According to the Appeals Chamber itself, only transnational terrorism is a customary crime. It necessarily follows that national laws on purely domestic terrorism cannot be meaningfully relied upon as evidence of widespread agreement on combating international terrorism. Indeed, that would be like suggesting that because every state prohibits murder in its own territory, therefore transnational murder must be a customary international crime – the proposition is ridiculous.
The Appeals Chamber does not establish that the national laws cited only concern international terrorism. In fact, quite a few of the laws cited are territorially limited (that is, they principally address domestic terrorism), or only have a limited extraterritorial application (such as under the nationality or passive personality principles of prescriptive jurisdiction), or impose universal jurisdiction only over limited sectoral offences but not over terrorism generally.Footnote 16 Further, some of the laws cited do not provide general definitions of international terrorism at all, but domestically implement sectoral offences in existing anti-terrorism treaties or otherwise criminalize certain acts without providing a general definition.Footnote 17 Only a portion of the national laws cited cumulatively criminalize ‘terrorism’ generally, require a transnational element, and extend extraterritorial jurisdiction under the universality principle.Footnote 18
Second, just as the Appeals Chamber simplistically aggregated national laws addressing domestic and international terrorism, so too did it crudely conflate national definitions of terrorism used for criminal purposes with definitions used for non-criminal purposes. Globally, many national legal definitions of terrorism do not create criminal offences, but serve a wide range of other purposes. For example, they may enable ‘civil’ anti-terrorism regimes,Footnote 19 activate emergency powers,Footnote 20 or prompt immigration restrictions.Footnote 21 National definitions may also trigger special law-enforcement powers,Footnote 22 specify jurisdictional arrangements,Footnote 23 modify procedural rules,Footnote 24 or enhance criminal penalties for underlying ordinary offences.
Accordingly, as regards the laws cited by the Appeals Chamber, the United States’ definition serves jurisdictional, remedial, and ancillary offence purposes, but does not establish a criminal offence of terrorism. Likewise, in the United Kingdom, there is no crime of terrorism per se, but only various inchoate offences that ‘hang off’ the terrorism definition; just as in the days of combating the Irish Republican Army, terrorist acts may still be prosecuted as ordinary offences (such as murder and so forth). The Russian Federation's law primarily empowers governmental authorities to combat terrorism, while the Pakistani law attends to court jurisdictional matters.
Third, although the Appeals Chamber argued that the many laws cited converge on a common definition of terrorism, with only ‘peripheral variations’,Footnote 25 that assertion conceals the reality of fundamental conceptual disagreements about terrorism that are clearly evident in those laws. Some of the laws cited go far beyond the Appeals Chamber's stated notion of terrorism (as peace-time violence to intimidate civilians or coerce authorities and having a transnational dimension) and the Appeals Chamber ignored, glossed over, or was inattentive to very real and deep substantive differences in legal conceptions of terrorism in the jurisdictions it cites.
For example, the Iraqi law encompasses ‘sectarian strife, civil war or sectarian fighting’ and ‘aggression’; the Peruvian law includes ‘subverting the constitutional order’; the Egyptian law covers any force or threat that harms public order; the Uzbeki law protects against acts that ‘complicate international relations, infringe upon sovereignty and territorial integrity, undermine the security of the state, provoke war, armed conflict, destabilize sociopolitical situation’ and so on; and the Saudi notion even contemplates ‘the violation of honour’. Many of the common-law jurisdictions cited (United Kingdom, Australia, Canada, New Zealand, South Africa) require an ulterior, specific-intent element (or motive requirement) of a political, religious, or ideological purpose or objective underpinning the underlying physical acts.Footnote 26 All of these are hardly peripheral variations; they go to the heart of how states conceive of what is ‘terrorism’.
Fourth, looking beyond the limited number of 37 ‘best-example’ national laws cited by the Appeals Chamber, it is abundantly clear that legal approaches to terrorism in the bulk of national legal systems – including the 160 states or so not mentioned in the Decision – are even more diverse and divergent in approaching terrorism. My earlier analysis of state reports to the UN Security Council's Counter-Terrorism Committee, admittedly current only to 2006, showed that around 87 states lacked special terrorism definitions or offences (and hence used ordinary offences), 46 states had simple generic terrorism definitions, and 48 states had composite generic terrorism definitions.Footnote 27
To give just a few examples from my region (Asia–Pacific), some definitions (for various purposes) capture comparatively trivial harms or common harms, such as the Sri Lankan law, which classes as terrorism merely defacing a road sign,Footnote 28 or the Bangladeshi law, which classes ‘to steal or seize by force from any person any money, jewelry, valuable article or any other vehicle’ as a terrorist offence, or even ‘to outrage the modesty’ of a woman or child.Footnote 29 In other countries, what is categorized as terrorism is indistinguishable from other public-order offences or political crime, as in Bhutan, where treason (including defaming the king or threatening national security) is an offence under anti-terrorism legislation.Footnote 30 Such diversity at the national level cannot sustain any serious argument that there is a converging global consensus on a definition or crystallized customary crime of terrorism based on national practice.
Fifth, the Appeals Chamber blithely relied upon some national terrorism laws that patently violate international human-rights law. Both the UN Human Rights CommitteeFootnote 31 and the UN Special Rapporteur on terrorism and human rights, Martin Scheinin,Footnote 32 have criticized national definitions of terrorism for targeting political opponents or minorities, excessively restricting human rights, or being too vague to satisfy the principle of legality underpinning the freedom from retroactive criminal punishment under Article 15 of the International Covenant on Civil and Political Rights 1966 (‘ICCPR’). The UN High Commissioner for Human Rights has summarized such concerns about national definitions:
many States have adopted national legislation with vague, unclear or overbroad definitions of terrorism. These ambiguous definitions have led to inappropriate restrictions on the legitimate exercise of fundamental liberties, such as association, expression and peaceful political and social opposition . . .. Some States have included non-violent activities in their national definitions of terrorism. This has increased the risk and the practice that individuals are prosecuted for legitimate, non-violent exercise of rights enshrined in international law, or that criminal conduct that does not constitute ‘terrorism’ may be criminalized as such . . .. There are several examples of hastily adopted counter-terrorism laws which introduced definitions that lacked in precision and appeared to contravene the principle of legality . . .. Particular care must be taken . . . in defining offences relating to the support that can be offered to terrorist organizations or offences purporting to prevent the financing of terrorist activities in order to ensure that various non violent conducts are not inadvertently criminalized by vague formulations of the offences in question.Footnote 33
Various national laws cited by the Appeals Chamber raise such human-rights concerns.Footnote 34 It is problematic to say the least for the Appeals Chamber to seek to evidence custom by relying on national laws that violate existing international human-rights law. Such national laws are simply unlawful under international law (to the extent of their inconsistency with the ICCPR). There is no opinio juris suggesting that excessive national terrorism definitions may lawfully derogate from international human-rights standards and thereby be relevant to custom formation. The Appeals Chamber's punitive impulse to enlarge criminal liability for terrorism appears to overshadow its concern for human-rights considerations.
3. National laws and Security Council Resolutions 1373 (2001) and 1566 (2004)
By Resolution 1373 (2001), adopted under Chapter VII of the UN Charter, the UN Security Council directed all states to criminalize terrorist acts in domestic law. The Appeals Chamber argues that ‘the attitude taken’ in national laws adopted in response to that resolution ‘is concordant and not subject to transient national interests’ and therefore ‘evinces a widespread stand on and a shared view of terrorism’.Footnote 35 For the reasons given above – that is, chronic diversity and divergence of national laws – that assertion is not borne out by present legal realities.
In addition, it is worth emphasizing the factors that have produced such divergence under the UN regime. First, Resolution 1373 did not define terrorism for the purpose of national criminalization, resulting in decentralized and inconsistent national implementations of it. Many states utilized the authority of the resolution to define terrorism to suit their own political purposes and to camouflage assaults on fundamental rights, as well as to deviate from procedural protections ordinarily enjoyed by suspects in criminal proceedings or other law-enforcement settings.
Second, while a working definition of terrorism was offered some years later by the Security Council in Resolution 1566 (2004),Footnote 36 that resolution was not binding and has not yet had a discernible quasi-legislative effect on driving a uniform definition of ‘terrorism’ at the national level. By 2004, quite a few states had already modified their criminal law and were hardly going to revisit the issue in light of a mere recommendation by the Council – particularly one that takes a fairly narrow view of terrorism and would diminish national freedom of action.
Third, according to the UN Special Rapporteur, the UN Security Council's Counter-Terrorism Committee (CTC) paid little attention to the human-rights implications of national definitions of terrorism during the early phases of the reporting process under Resolution 1373 (2001). The CTC's emphasis on strong anti-terrorism measures without due regard for human rights sent a problematic ‘message’ of encouragement to states with rights-violating definitions,Footnote 37 again producing differences amongst different jurisdictions.
At most, all that can be presently said about the combined effects of Resolutions 1373 and 1566 is that, over time, they may stimulate increasing convergence at the national level. Resolution 1566 certainly establishes soft-law guideposts on definition in states’ implementation of Resolution 1373. The Security Council's Counter-Terrorism Directorate now utilizes that definition in its dialogue with states, encouraging convergence. The UN Special Rapporteur is satisfied that Resolution 1566 reflects a fairly narrow and rights-respecting concept of terrorism.Footnote 38 Over time, widespread actual national practice in conformity with it may help to evidence a shared customary international-law concept of terrorism, but there is still a considerable way to go.
Paradoxically, although the Appeals Chamber invoked Resolution 1566,Footnote 39 that definition did not in fact support the Appeals Chamber's own definition of terrorism. Resolution 1566 defines harmful conduct as terrorism only (i) when committed to provoke a state of terror, or intimidate a population, or compel a government or an international organization, and (ii) where the conduct also constitutes an offence under existing ‘sectoral’ anti-terrorism treaties. From the early 1960s onwards, numerous sectoral treaties have addressed particular means of methods used by terrorists (such as hijacking, hostage taking, attacks on diplomats, bombings, and so forth), but usually without mentioning the term ‘terrorism’ and never defining a general treaty crime of ‘terrorism’ as such.
Resolution 1566 does not therefore criminalize any additional conduct that is not already criminal under existing transnational anti-terrorism treaties. Rather, it reclassifies as ‘terrorism’ certain existing criminal wrongs where they also aim to terrorize, intimidate, or compel. By contrast, the Appeals Chamber's definition was crafted precisely so as to avoid the limitations in the Lebanese law – which does not criminalize certain unstipulated means (such as guns or knives) – by recognizing any criminal means that cause the requisite harm.
Yet, existing sectoral anti-terrorism treaties do not cover all possible terrorist means or methods (including guns or knives), but are limited to those categories that the international community has reactively regulated on an ad hoc basis – precisely because there was no agreement to cover all methods in a general terrorism offence of the kind suggested by the Appeals Chamber. Necessarily, Resolution 1566 would not cover such means if not codified in an existing treaty. Accordingly, the sources of custom relied upon by the Appeals Chamber do not support the customary definition of ‘terrorism’ identified by it.
4. National judicial decisions
The Appeals Chamber invoked nine national judicial decisions to support its contention that national courts have ‘explicitly’Footnote 40 recognized ‘that there exists a crime of terrorism under customary international law’.Footnote 41 Regrettably, the Appeals Chamber misread, exaggerated, or misrepresented every one of those decisions, many of which, as Brownlie has said of national decisions generally, themselves ‘rest upon a very inadequate use of the sources’.Footnote 42 As shown below, a careful and accurate reading of each decision readily demonstrates that they simply do not support the propositions claimed. In addition, the Appeals Chamber dismissed judicial decisions that explicitly find that terrorism is not a customary crime as ‘far and few between’ and as diminishing every year.Footnote 43 It fleetingly mentions some of these by burying them in a footnote,Footnote 44 but does not engage with their substantive findings and reasons at all.
In the first case invoked – Suresh v. Canada – at no point does the Canadian Supreme Court declare that terrorism is a customary crime. All the Supreme Court does is obliquely indicate that the definition in the Terrorist Financing Convention ‘catches the essence’ of terrorismFootnote 45 for the limited purpose of interpreting and applying a domestic immigration-law statute. In this context, the Court denied that ‘the term “terrorism” is so unsettled that it cannot set the proper boundaries of [domestic] legal adjudication’.Footnote 46 Yet, the Court also conceded that ‘there is no single definition that is accepted internationally’, that ‘[o]ne searches in vain for an authoritative definition’, and that ‘the term is open to politicized manipulation, conjecture, and polemical interpretation’.Footnote 47 Such equivocal statements, and the limited nature of the proceeding, clearly undercut any implicit conclusion that terrorism was viewed as a customary crime by the Canadian Supreme Court.
The other Canadian case invoked – Zrig v. Canada (a refugee exclusion case) – was said by the Appeals Chamber to hold that there was ‘an international consensus at least as to certain forms of terrorism’ from 1997.Footnote 48 Yet, the very next sentence in the same judgment – omitted by the Appeals Chamber – states that ‘it is not necessary . . . to decide the point’ in the case at hand.Footnote 49 Further, that undecided point is rendered even more equivocal given that the Canadian Federal Court suggests only that ‘certain forms of terrorism’ have attracted international consensus – language that falls far short of a finding as to a comprehensive, universal crime of terrorism. It is well accepted, for instance, that certain terrorist acts, such as hijacking or hostage taking, are prohibited by customary law, without there being a wider customary crime of terrorism as such.
In the third decision mentioned – Bouyahia Maher Ben Abdelaziz et al.Footnote 50 – the Italian Supreme Court of Cassation found that certain anti-terrorism treaties indicate a ‘general definition applicable in both peacetime and in wartime’.Footnote 51 Yet, one critical factor undermines the argument that this decision supports a customary crime of terrorism: the Appeals Chamber's stated conception of terrorism departs in a vital respect from the notion identified by the Italian Court. The Court insisted that ‘it is generally held that in order to be regarded as terrorism, acts must at the psychological level meet further requirements as to political, religious or ideological motivation in accordance with generally accepted international standards’.Footnote 52 In contrast, the Appeals Chamber found that such a purpose requirement (or motive or special-intent element) ‘has not yet been so broadly and consistently spelled out and accepted as to rise to the level of customary law’.Footnote 53
The Appeals Chamber's assessment of the customary crime of terrorism is thus directly at odds with a judgment it invokes to support its assessment. The difference of opinion is not peripheral, but fatal to the Appeals Chamber's argument. To use an analogy, it would be like asserting as follows: there is universal agreement on the crime of ‘breaking and entering’; a national decision supports that conclusion by defining the crime to require both the elements of ‘breaking’ and ‘entering’; yet, paradoxically, an international tribunal still finds that an international crime of ‘breaking and entering’ exists even if its elements only include ‘breaking’ but not ‘entering’.
This is not a trivial or marginal difference of opinion. According to the Italian Court, terrorism is simply not terrorism unless it is defined to include the purpose element. That is also the approach of some significant common-law countries (such as the United Kingdom, Australia, Canada, New Zealand, and South Africa, to name a few, some of which are cited by the Appeals Chamber in support of its arguments). But for the purpose element, conduct is not terrorism, because that element supplies the additional, distinctive characteristic that transforms other forms of crime into the special crime of terrorism.Footnote 54 An indispensible element of the Italian Court's definition cannot simply be disregarded by the Appeals Chamber because it is inconvenient and does not fit the Chamber's own idiosyncratic conception of terrorism.
In so far as the Italian Court identifies a customary definition of terrorism, it is also important to query the basis for that conclusion. The relevant Italian domestic legislation defines specific elements of terrorism but also recognizes ‘other conduct defined as terrorism . . . under conventions or other provisions of international law binding upon Italy’.Footnote 55 In interpreting the scope of terrorism under this ‘open’ definition in domestic law, the Italian Court simply drew upon certain anti-terrorism instruments,Footnote 56 as directed by the legislation. Consequently, the judgment principally relied upon two treaties to determine the scope of a domestic definition, and its extrapolation that terrorism is a customary crime was largely unsubstantiated in its reasons.
In the fourth decision invoked by the Appeals Chamber – the Cavallo case – a single Mexican federal judge is quoted as finding that ‘multiple conventions . . . provide that the crimes of genocide, torture and terrorism are internationally wrongful and impose on member states of the world community the obligation to prevent, prosecute and punish those culpable of their commission’.Footnote 57 However, the judgment does not express the point in customary-law terms, but can equally be read as indicating that various treaties criminalize different acts of terrorism. It does not overtly claim that terrorism per se is a distinct international customary crime.
More importantly, when the same case reached the Mexican Supreme Court on appeal, this issue was not addressed at all because it was not essential to the disposal of the case. The Supreme Court instead focused on the crimes of genocide, terrorism, and torture under the relevant national laws of Spain and Mexico, in a bilateral extradition case also involving questions of domestic constitutional law. Where international law is raised, the focus is on the relevance of treaties in interpreting national law,Footnote 58 not arguments about customary law, and there is no consideration of customary universal jurisdiction. As such, the incidental remarks of a single lower-court judge on a question that was not at issue in the proceedings are of peripheral relevance to assessing customary law.
The fifth decision invoked by the Appeals Chamber – Chile v. Clavel – is equally marginal. A single concurring judge of the Argentinean Supreme Court is quoted for the proposition that ‘customary international law and conventional law echo the need for international cooperation on the repression of terrorism’.Footnote 59 Yet, the joint majority opinion does not mention the word ‘terrorism’ even once: the case did not concern terrorism at all, but rather genocide and crimes against humanity. Any incidental reference to terrorism by a single judge in this context is next to irrelevant, given that terrorism was not even an issue argued or decided in that case. In this light, the decision is hardly authoritative or persuasive.Footnote 60
The sixth national decision cited – Almog v. Arab Bank – found that suicide bombings and ‘other murderous acts intended to intimidate or coerce a civilian population . . . violates an established norm of international law’.Footnote 61 A number of considerations limit the significance of that case in assessing customary law. Foremost is that the case concerned civil jurisdiction under the Alien Tort Claims Act and not criminal jurisdiction under the universality principle.Footnote 62 For this reason, the US District Court itself distinguished its findings in this case from an earlier case – US v. Yousef – in which ‘terrorism’ was held not to be a crime of universal jurisdiction.Footnote 63 At its highest, then, Almog is authority only for the proposition that certain terrorist acts violate the ‘law of nations’ in a civil suit, but are not criminal violations under customary law.
Even on that limited point, the assessment of customary law in Almog is suspect. The US District Court relied on two anti-terrorism treaties – the Terrorist Bombings Convention and the Terrorist Financing Convention – plus UN Security Council Resolution 1566 (2004) and the law of armed conflict as the basis for its conclusion that the violence mentioned above violates the law of nations. Yet, the Court itself notes that those treaties had been ratified by only 120 and 130 states, respectively – hardly enough to meet the Court's own standard for custom formation: ‘whether virtually all States recognize its validity.’Footnote 64 Adding to their analysis a non-binding Security Council resolution and norms applicable only in armed conflict hardly makes up the difference or tips the balance in favour of the existence of a customary crime.
Moreover, the US District Court itself narrowly confined its finding about customary law. Far from claiming that ‘terrorism’ as such is prohibited by customary law, the Court was at pains to observe that only certain acts specifically prohibited by two particular treaties are customary. As the Court states: ‘there is no need to resolve any definitional disputes as to the scope of the word “terrorism”, for the Conventions expressing the international norm provide their own specific descriptions of the conduct condemned.’Footnote 65 That specific conduct is then found to comprise ‘organised, systematic suicide bombings and other murderous attacks on innocent civilians intended to intimidate or coerce a civilian population’. The Court does not go further to indicate that there exists a general international crime of terrorism.
For the same reason, the seventh case cited (and only in a footnote) by the Appeals Chamber – United States v. Yunis Footnote 66 – is of equally limited relevance. The Appeals Chamber notes that the US Court of Appeals in that case accepted universal jurisdiction over ‘perhaps certain acts of terrorism’.Footnote 67 Yet, the passage quoted is in the context of the Court of Appeals’ describing the findings and sources relied upon in a lower District Court decision. In fact, the case was decided on a different point: the US Court of Appeals declared that the case must be decided according to US law and it was not required ‘to conform the law of the land to norms of customary international law’ (the latter thus being irrelevant).Footnote 68
At most, then, the Court of Appeals suggested incidentally (not dispositively) that ‘[a]ircraft hijacking may well be one of the few crimes so clearly condemned under the law of nations that states may assert universal jurisdiction’, since that was the relevant conduct at issue on the facts. There is little doubt that aircraft hijacking is indeed a customary crime; but the case simply did not address the existence or otherwise of any broader, general international crime of terrorism.
The final two decisions invoked by the Appeals Chamber (and only in a footnote)Footnote 69 are of marginal significance in evidencing custom. In EHL, the Belgian Court of Cassation merely proffered a certain notion of terrorism for the purpose of interpreting the scope of political offences under a domestic statute of 1833Footnote 70 and appears to say nothing about customary law. In Al-Sirri v. Home Secretary, it is true that the UK Court of Appeal invoked UN Security Council Resolution 1566 to define international terrorism in a refugee exclusion case. However, the significance of that case is limited, for two key reasons.
First, Lord Sedley pointedly noted that ‘[t]here is no present need for an elaborate definition (which may, I accept, be needed in other contexts)’.Footnote 71 A definition for refugee exclusion purposes is one thing; extrapolating the same definition to establish international criminal responsibility is quite another. Criminal law demands far greater specificity and certainty, not least to satisfy the principle of legality and non-retrospectivity.
Second, the UK Court used the Security Council resolution to work out whether terrorism was contrary to the principles and purposes of the UN under Article 1(F)(c) of the Refugee Convention 1951, which UK law incorporated as the relevant test for refugee exclusion.Footnote 72 Yet, in doing so, the UK Court ‘read down’ the wider UK statutory definition of ‘terrorism’ by preferring the narrower Security Council definition for that purpose.Footnote 73 Paradoxically, at different points in its decision, the Appeals Chamber invoked both this UK judgment (which relies on the Security Council definition) and the UK statutory definition to support a common customary standard. The problem is that the UK Court itself distinguishes the different scope of each definition – hardly a vote of confidence in the Appeal Chamber's assertion that the sources converge.
In sum, as the above analysis incontrovertibly demonstrates, the national decisions invoked by the Appeals Chamber simply do not individually or cumulatively support its conclusion that terrorism is a customary crime. The Appeals Chamber misread, exaggerated, or misrepresented the significance of every decision cited and did not take the necessary care to closely examine any judgment. The somewhat cavalier treatment of those decisions produces an unhappily misleading result.
On a fuller assessment of relevant national decisions, national judicial opinion weighs against the existence of a customary crime of terrorism. As alluded to above, in US v. Yousef, a bench of three judges on the US Court of Appeal (in contrast to a single lower-court judge in Almog) stated as recently as 2003 that ‘We regrettably are no closer now than eighteen years ago to an international consensus on the definition of terrorism or even its proscription’.Footnote 74 That case was directly on point: one issue raised by the defence was whether there existed universal jurisdiction over terrorism itself under customary international law.
The Court found that there was no universal jurisdiction over ‘terrorism’, finding that the District Court erred.Footnote 75 Universal jurisdiction could only be exercised over ‘a limited set of crimes that cannot be expanded judicially’.Footnote 76 In contrast, the Court believed that terrorism ‘is a term as loosely deployed as it is powerfully charged’.Footnote 77 The ‘indefinite category’ of terrorism was not condemned by all states, which had failed to achieve consensus on definition.Footnote 78 The Court noted that even US law contains several different definitions of ‘terrorism’ for different purposes.Footnote 79
The conclusion in US v. Yousef reinforces the earlier finding in an (admittedly) civil case, Tel-Oren v. Libya (1984), in which a US federal appeals court found that terrorism was not an offence against the ‘law of nations’ under the Alien Tort Claim Act.Footnote 80 A majority of two judges found that there was insufficient international consensus on the definition of ‘terrorism’ for it to constitute an offence against the law of nations.Footnote 81 As Judge Edwards stated:
While this nation unequivocally condemns all terrorist acts, that sentiment is not universal. Indeed, the nations of the world are so divisively split on the legitimacy of such aggression as to make it impossible to pinpoint an area of harmony or consensus.Footnote 82
Whilst being an older case on terrorism in US civil jurisdiction than Almog, the decision remains arguably superior, given that it was at the appellate level and that, unlike in many other decisions, the arguments about customary law were fully ventilated.
Outside the United States, acts of terrorism have also rarely been prosecuted as crimes of international terrorism, rather than as sectoral or ordinary offences. In Ghaddafi (2001), the French Court of Cassation found that terrorism was not an international crime entailing the lifting of immunity for heads of state and proceedings were quashed.Footnote 83 Again, that case directly addressed the issue at hand, although the reasoning and analysis – perhaps due to the civil-law tradition of slighter reasoning than at common law – are not so full as to be definitive.
In Singh v. State of Bihar, a 2004 appeal against convictions under Indian anti-terrorism legislation,Footnote 84 the Indian Supreme Court observed that the problem of defining ‘terrorism’ had ‘haunted countries for decades’, not least because of the difficulties posed by ‘freedom fighters’ and state-sponsored terrorism. The Court believed that ‘[t]erminology consensus’ would be necessary for agreement on a comprehensive international treaty and suggested that the ‘lack of agreement’ on definition had been ‘a major obstacle to meaningful international countermeasures’. While proffering a tentative definition, the Court does not suggest that terrorism is a customary international crime – a question that was not at issue in the case. As the authoritative superior court in a country wracked by terrorism for many decades, in a democracy of over one billion people, the incidental remarks of the Indian Supreme Court nonetheless carry some significance.
It is also worth mentioning here that there is simply no support in the decisions of the International Court of Justice or other international tribunals for the proposition that terrorism is a customary international crime. Roslyn Higgins observes that neither the Nicaragua (Merits) case nor the Tehran Hostages case addressed those situations of violence through the prism of ‘terrorism’;Footnote 85 nor, for that matter, did the Lockerbie case. Recent cases such as the Israel Wall Advisory Opinion and Congo v. Uganda similarly omitted to recognize or deal with the terrorist-type violence involved in those situations under any general anti-terrorism norms. Those cases, of course, did not concern individual criminal liability, although state responsibility for international crimes could have been raised if the parties or the court itself thought the term ‘terrorism’ held any significance.
5. International treaties
The Appeals Chamber's aggregation of international and regional anti-terrorism treatiesFootnote 86 to support a unified customary crime of terrorism is not at all persuasive. First, numerous efforts by the international community since the 1920s, including negotiations for a UN Comprehensive Terrorism Convention since 2000, have not produced agreement on a general international crime of terrorism.Footnote 87 In the absence of a general crime in treaty law, no parallel customary rule can be said to exist.
Second, while there are numerous sectoral conventions that address specific criminal methods often used by terrorists (as noted earlier), none of those treaties – individually or collectively – contains a comprehensive definition of ‘terrorism’Footnote 88 or establishes a general international crime of ‘terrorism’, as such. Certain terrorist methods necessarily remain unregulated by treaty law (such as attacks by guns or knives – making it all the more strange that the Appeals Chamber recognized an international customary definition not limited to those enumerated means, in contrast to the view of the UN Security Council explained earlier). Again, no customary rules can emerge in parallel with treaty rules that do not exist. Nor can all of those treaties be cumulatively mashed together to somehow fill in the gaps between them individually: international action against terrorism is not more than the sum of its parts. The mere mention of the term ‘terrorism’ in some of the later treaties does not alter the situation.
It is well known that the sectoral approach was adopted precisely because states could not reach agreement on terrorism generally. Those decades of deadlock cannot mean nothing in assessing custom formation, but, to the contrary, are decisive: there plainly lacked consensus amongst the international community on defining and criminalizing terrorism as such. That disagreement is not only historical, but persists: there is an ongoing inability to conclude the UN Draft Comprehensive Terrorism Convention within the UN General Assembly. That alone is a very persuasive reason for believing that there cannot yet be a customary international crime. Even if there seems to be drafting agreement on the definition in the UN Draft Convention, its significance as evidence of custom still remains doubtful until (i) it is adopted, (ii) it enters into force, (iii) there is widespread participation by states in it, and (iv) actual counterterrorism practice generally conforms to it.
At most, as mentioned earlier, specific offences in some treaties may have entered into customary law, such as aircraft hijacking or hostage taking.Footnote 89 There is also a specialized war crime of spreading terror amongst a civilian population during armed conflict under customary humanitarian law,Footnote 90 but the special circumstances of that offence set it apart from any general peace-time crime of terrorism.
It may even be arguable that certain norms of a particularly recent treaty, the Terrorist Financing Convention 1999, have crystallized into parallel customary rules because of a combination of additional supporting practice (although the same probably cannot be said of the Terrorist Bombings Convention 1997).Footnote 91 All that would establish, however, is that there is a customary crime of terrorist financing, but it does not – as the Appeals Chamber suggested – support the broader proposition that terrorism itself is also an offence. The prohibition of terrorist financing is just another sectoral response addressing a particular manifestation or method of terrorism, but that is all.
6. Regional treaties
The Appeals Chamber also invoked regional treaties to support its conclusion on custom. Far from demonstrating a global consensus, an accurate reading of those conventions establishes exactly the opposite: enormous variation in regional conceptions of terrorism and certainly too much diversity to establish a widespread consensus or shared notion of it.Footnote 92
To begin with, some regional treaties merely focus on specific terrorist methods, without defining terrorism generally at all.Footnote 93 Others contain (often wide and conflicting) generic definitionsFootnote 94 or define terrorism only for the purpose of criminalizing ancillary conduct.Footnote 95 Some do not create offences at all, but serve other purposes (such as extradition or law-enforcement co-operation).Footnote 96
In addition, regional treaties do not necessarily enjoy widespread regional participation by members of the relevant regional organizations; for example, more than a decade since its adoption, only one-quarter of the members of the Organisation of the Islamic Conference (‘OIC’) are parties to the OIC treaty, substantially limiting its significance. Even where regional states are parties, that does not automatically establish qualitative practice in accordance with its norms. The Organisation of African Unity (‘OAU’) treaty, for instance, does not appear to have penetrated deeply or even at all into many national legal systems in Africa, where many countries still lack terrorism laws.
Concerning definitions of ‘terrorism’, there is vast diversity among regional instruments. Some of the conventions reclassify ordinary crimes or public-order offences as terrorism,Footnote 97 or even label insurrection as terrorism.Footnote 98 Some criminalize conduct infringing diffuse values such as the ‘stability, territorial integrity, political unity or sovereignty’ of statesFootnote 99 or the ‘honour’ or ‘freedoms’ of individuals,Footnote 100 or even mere threats to such interests. One intermingles terrorism with concepts of ‘separatism’ or ‘extremism’.Footnote 101 Such vague and subjective elements seem hardly capable of satisfying the principle of legality and the proper concerns of many other states in precisely and carefully delimiting the scope of criminal liabilities.
Further, some treaties use very ambiguous language, such as prohibiting jeopardy to a ‘national resource’,Footnote 102 damage to ‘environmental or cultural heritage’,Footnote 103 causing any ‘consequences dangerous to society’,Footnote 104 or merely ‘occupying’ propertyFootnote 105 or disrupting a public service (even in a democratic protest).Footnote 106 Another includes an ill-defined element of ‘seriously destabilising or destroying the fundamental political, constitutional, economic or social structures of a country or an international organization’.Footnote 107 There is also great variation in the specification of underlying physical acts, and differences in the thresholds of ulterior intention (from merely influencing others to seriously intimidating them).
It is simply not accurate for the Appeals Chamber to simplistically identify convergence between regional treaties, let alone suggest their individual congruence with the Appeals Chamber's own definition of ‘terrorism’. The sheer diversity of regional definitions of terrorism weighs against the view that there is any customary international definition of a crime of ‘terrorism’. To apply the dictum of the International Court of Justice in the Asylum case, these regional treaties reflect ‘so much uncertainty and contradiction, so much fluctuation and discrepancy . . . that it is not possible to discern in all this any constant and uniform usage, accepted as law’.Footnote 108 While the ICJ also stated in the Fisheries case that ‘too much importance need not be attached to a few uncertainties or contradictions’,Footnote 109 the differences between the definitions of terrorism in regional treaties are not merely penumbral, but reflect basic disagreements about the notion of terrorism.
The Appeals Chamber relied on the Arab Convention for the Suppression of Terrorism 1998, not only as a source of custom, but directly to ‘identify a persuasive interpretation of the Lebanese Criminal Code as part of the overall context relevant to its interpretation’.Footnote 110 The Arab Convention was considered relevant because ‘it is a well known principle of international law that, as much as possible, a national law shall be so construed as to make it consistent with a State's international obligations’.Footnote 111
The Tribunal's conclusion on this point is also untenable. First, the terrorism offence in the Lebanese Criminal Code does not purport to domestically implement the Arab Convention, but pre-dates it by many decades, so it has no immediate interpretive relevance. Second, and more decisively, the Arab Convention defines terrorism only for the purpose of transnational co-operation but does not require domestic criminalization of the Convention's definition. As such, it is not intended to affect the definition of terrorist offences under national law, and the Lebanese legislature never modified domestic criminal law after its ratification. Third, the Arab Convention's definition of ‘terrorism’ is wide, vague, and arbitrary, raising serious human-rights concernsFootnote 112 that militate against its use in domestic interpretation in a criminal proceeding.
7. United Nations resolutions
Finally, the Appeals Chamber invoked a series of UN General Assembly resolutions, adopted since the 1994 Declaration on Measures against International Terrorism,Footnote 113 to support its findings on customary law. The repetition of apparently normative standards condemning terrorism as criminal over more than a decade and a half appears prima facie significant. The 1994 Declaration is of greater importance than ordinary resolutions. It was adopted without a vote, suggesting consensus among states, and has been reiterated in numerous later resolutions.Footnote 114 The reference to ‘criminal acts’ invokes normative language rather than mere exhortation.
Nonetheless, there is little support for the view that the definition declares a customary crime. The Declaration itself emphasizes the need to progressively develop and codify the law on terrorism,Footnote 115 far from purporting to reflect existing rules. Adoption by consensus does not guarantee unanimity among states, just that there are no formal objections.Footnote 116 States clearly also supported it for non-legal reasons. Indeed, during discussions in the Sixth Committee of the UN General Assembly, a number of states specifically proposed legal definitions, or gave concrete examples, of terrorism at variance with the 1994 Declaration.Footnote 117
Most importantly, many states argued that there was still a need to define ‘terrorism’ and/or to adopt a comprehensive treaty criminalizing it,Footnote 118 and to distinguish self-determination struggles.Footnote 119 Subsequently, the many states of the Non-Aligned Movement (118 member states) and Organization of the Islamic Conference (56 member states) have insisted on the continuing importance of achieving a legal definition and the differentiation of liberation struggles – even while approving of the 1994 Declaration.Footnote 120 There is also little complementary external practice suggesting, for example, that states have criminalized terrorism, prosecuted it, or extradited suspects in reaction to the terms of the 1994 Declaration or subsequent resolutions.
While certain states dismissed the need for a definition and/or a convention,Footnote 121 the 1994 definition must be viewed as a compromise formula, which identifies a minimal political agreement on the scope of terrorism, or, at most, customary agreement that states themselves are prohibited (at the level of state responsibility) from supporting terrorism. Yet, it still falls short of a legal definition for the purposes of establishing criminal liability, which was left to another day. That day began in 2000 with negotiations for a Draft Comprehensive Terrorism Convention.
The UN resolutions must be interpreted cautiously in this light, since the parallel UN treaty negotiations on the Draft Comprehensive Convention since 2000 have been unable to reach agreement on a legal definition. As noted earlier, this is a decisive point. It would be curious if a customary crime of terrorism existed and yet states remained deadlocked on simply codifying that existing agreement in a treaty. It would be surprising if the protracted disputes on a treaty definition could be circumvented by a recent series of non-binding resolutions in a political body (or, for that matter, by invoking strong peripheral national judicial decisions). Against that background of other practice, the resolutions are not sufficient evidence of a customary crime.
Moreover, the 1994 Declaration's definition of ‘terrorism’ is, in any case, different from that in the Draft Comprehensive Convention (which does not require a political motive), different again from that in Security Council Resolution 1566 (which is limited to underlying sectoral offences), and also different from those many different definitions in national laws and regional treaties (as shown earlier). Further, international treaties negotiated since 1994 – including the 1999 Terrorist Financing Convention and regional treaties – have adopted different definitions, suggesting that the legal definition of ‘terrorism’ remains contested.
More specifically, the 1994 Declaration defines ‘terrorism’ as criminal acts intended to provoke terror in certain persons for political purposes. In contrast, the definition in the 1999 Terrorist Financing Convention refers to serious violent acts for the purpose of intimidating a population, or compelling a government or international organization.Footnote 122 The former definition requires a political purpose, whereas the latter focuses on coercion or intimidation for whatever purpose – including non-political ones. There is thus a significant difference in the definitional elements of terrorism between different international legal instruments. As discussed earlier, this is not a peripheral difference, but goes to the core of whether conduct can be properly described as terrorism or not. Further, as mentioned above, the Security Council's conception is narrower again, by also insisting on the presence of an underlying sectoral offence.
The problem also remains of securing agreement on the scope of exceptions to any international definition of terrorist crimes – which is still unresolved after more than a decade of UN treaty negotiations (and despite apparent earlier agreement in the treaties on terrorist bombings and terrorist financing). Jurisprudentially, it must be doubted whether there can exist a customary crime if there remains basic disagreement over the scope of the liability sought to be created.
8. Conclusion
The Appeals Chamber's assessment of custom was correct on a number of points: there is no general customary crime of terrorism applicable in armed conflictFootnote 123 and state practice does not uniformly require a ‘motive’ element (such as a political, religious, or ideological purpose).Footnote 124 In most other respects, however, the Appeals Chamber badly misjudged the available material sources evidencing the international community's attitude to criminal liability for terrorism. National legislation, judicial decisions, international and regional treaties, and UN resolutions do not sustain any credible argument that terrorism is a customary international crime. The world is certainly closer than before; the tide may be turning, but it is not yet in.
Why does it matter? First, incorrect decisions based on poor reasons and loose methodology are not good for international law or public confidence in its institutions and processes.
Second, such decisions rarely remain quarantined in the ghetto of bad decisions, but tend to take on a life of their own: the incantatory power of an international tribunal is such that what it says is the law, is the law, or quickly becomes it – even if the decision itself is a house of cards. States, courts, and scholars will now routinely cite this decision as authority for the proposition it states, but will rarely look behind it or critically re-examine its evidentiary foundations.
Third, inventing post facto liabilities infringes on the human rights of criminal suspects, even if the conduct of such persons is morally abhorrent or falls through the gaps in the law then in force. After all, it was hardly foreseeable to a person in Lebanon in 2005 that terrorism as such was an international customary-law crime or that such crime would be used ostensibly in ‘interpreting’ domestic offences but, in reality, to import a wholly new offence.
Fourth, to the extent that an international tribunal has held up rights-disrespecting national and regional definitions of ‘terrorism’ as global gold standards, encouragement and legitimacy have been given to despots to demonize and convict their opponents as terrorists. This makes the job of human-rights lawyers harder in a field in which constraining excessive counterterrorism measures is hard enough.
The overly punitive instincts of some international criminal tribunals could benefit from a strong dose of human-rights sensitization, in a field in which states are already pushing the limits too far. The risks of overreach and overreaction are particularly acute in a tribunal established by the Security Council under its security powers, on a subject as politicized as terrorism, against the background of a tense and highly charged Lebanese politics, and in circumstances where a majority of the five judges on the Appeals Chamber are not necessarily international lawyers.
The Decision should also, however, be a wake-up call to states to finalize the negotiation of the UN Draft Comprehensive Convention so that further human-rights abuses from over-eager judicial law-making do not ensue. Public confidence in international criminal justice is also undermined if there are gaps in the law big enough for terrorists to attack through with impunity.
