Since 2001, preventive detention has become almost commonplace as a means of “incapacitating” terrorist networks. No other development – with the possible exception of the use of extreme interrogation techniques – has been as controversial, or as uncomfortably reconciled with conventional legal practices. But there are legitimate uses of preventive detention that respect the legal principles of protection of civil liberties.
In this study Craig Forcese proposes lessons for Canada from state practices in the United States, the United Kingdom and Australia since 9/11 that might reasonably guide the development of an appropriate system of preventive detention; specifically, one that is maximally effective within a sphere of tolerable restrictions on civil liberties. He does so in four sections. In the first section, Forcese compares several models that are essentially systems of preventive detention. In the second section, he examines the Canadian legal environment in which any discussion of preventive detention must be situated. He highlights the extent to which Canadian law already empowers the state to preempt terrorist activity. He concludes that while the gap that might reasonably be filled by a separate system of preventive detention is narrow, it does exist. In the third section, Forcese proposes, first, criteria for measuring the public safety effectiveness of such a system and, second, a zone of tolerable civil liberty restrictions. In the last section he draws on these criteria and prior practice to propose a model for the Canadian system of preventive detention.
The author favours a model that balances effectiveness and civil liberties. He identifies as a legitimate concern the narrow circumstances where the state (1) has reason to believe that a terrorist attack will occur; (2) has reason to believe that a particular group is behind the plot and that the suspect is a member of that group; but (3) has no information, other than this belief, to connect that particular individual to the plot. In these circumstances, conventional legal instruments allowing the state to disrupt that threat through detention of the individual may not be available. In that narrow space, there are arguments in favour of preventive detention.
Forcese suggests that section 83.3 of the the 2001 Anti-terrorism Act, the anti-terrorism provision allowing short-term detention in circumstances where conventional arrest powers could not be exercised, was a reasonable approach. He proposes a revamped section-83.3 process that adds certain other civil liberties safeguards and also permits the constrained use of secret evidence and special advocates. The net result is a system of “catch and release” (or catch and release subject to a peace bond) that focuses on disruption of a threat via the short-term detention of persons who are tied to specific threats. Forcese rejects approaches that detain solely on the basis of perceived inherent dangerousness.
“Indefinite imprisonment without charge or trial,” observed the United Kingdom’s highest court in 2004, “is anathema in any country which observes the rule of law.”1 Yet, since 2001, such detention — or analogous limitations on liberty — has become virtually commonplace as a means of “incapacitating” terrorist networks. No other development — with the arguable exception of the (sometimes related) use of extreme interrogation techniques — has been as controversial or as difficult to reconcile with conventional legal practices.
These controversies extend beyond the infamous detention of terrorism suspects in putative legal black holes like Guantanamo Bay. As Barack Obama’s administration appreciates, the question of whether to use indefinite detention without trial to forestall terrorist acts transcends ideological or partisan divisions. It is not happenstance, therefore, that rather than abandoning the idea altogether, the Obama administration is itself contemplating questions of preventive detention, including the issue of replacing the practice — but not the ultimate objective — of detention at Guantanamo Bay (Obama 2009). In so doing, that government, like others around the world, struggles with a seemingly irresolvable dilemma. As Mr. Justice Laws of the English Court of Appeal has asserted:
This grave and present threat [of terrorism] cannot be neutralised by the processes of investigation and trial pursuant to the general criminal law. The reach of those processes is marked by what can be proved beyond reasonable doubt…In these circumstances the state faces a dilemma. If it limits the means by which the citizens are protected against the threat of terrorist outrage to the ordinary measures of the criminal law, it leaves a yawning gap. It exposes its people to the possibility of indiscriminate murder committed by extremists who for want of evidence could not be brought to book in the criminal courts. But if it fills the gap by confining them without trial it affronts “the most fundamental and probably the oldest, most hardly won and the most universally recognised of human rights”: freedom from executive detention.2
Notably, this freedom from executive detention is much more than a quaint practice to be subordinated in some utilitarian calculation of the perils of terrorism. The fact remains that even in the post-9/11 period, violence by states constitutes a much greater peril to the individual than acts of terror. Even the weakest states possess means of coercive force that are only occasionally matched by nonstate actors. Civil libertarians have one eye on history when they express alarm at new measures that unshackle states to act more freely in detaining citizens and noncitizens alike. This concern is compounded by the overlap in the US case between expanded powers of detention and other, even more notorious practices, like the maltreatment of detainees.
In these circumstances, there is no “right” answer to the question of preventive detention. Outside of a handful of situations,3 it is unquestionably an alien concept, most notably in Western common-law jurisdictions. Done poorly — as at Guantanamo Bay — it may amount to a virtual repudiation of the rule of law and may blight a state’s reputation. On the other hand, preventive detention may also save lives if a terrorist attack is forestalled. The key, inevitably, is to reconcile the minimization of risk with the observance of the individual’s rights.
The two objectives do not dovetail perfectly, certainly in the immediate term. It is possible to imagine stark differences in attitudes to preventive detention, depending on a society’s tolerance for risks to security. A risk-minimizing society would permit mass detentions in the expectation that the minimal increase in public safety from the dragnet would outweighed the massive injury to civil liberties. A rights-maximizing society, however, would deny the state the power to detain except through conventional criminal proceedings (for which it would impose demanding standards), even at the risk of leaving people free whose intent and capacity are clear but whose terrorist acts lie in the future.
These polar positions do not, however, represent the inevitable balancing in which a liberal democracy must engage to reconcile security with rights. Somewhere on the spectrum between a system of detention that seeks to eliminate all risk (but at a severe cost to civil rights) and a system that preserves absolutist civil liberties (without responding to risk) is an optimal point that is tolerable in democratic, rights-respecting societies: the Goldilocks point that is neither too hot nor too cold, but “just right.”
This study cannot determine precisely where this Goldilocks point lies; it depends in large measure on the tolerance of a given society for risk versus civil liberties. What it does do, however, is propose lessons from state practice since 9/11 that might guide the development of a reasonable system of preventive detention, specifically, one that is as effective as possible within a zone of tolerable restrictions on civil liberties.
It does so in four sections. In the first section, I briefly survey several comparative models that arguably amount to systems of “preventive detention.” In the second section, I examine the Canadian legal environment in which any discussion of preventive detention must be situated. Specifically, I highlight the extent to which Canadian law already empowers the state to preempt terrorist activity. I conclude that while the zone that might reasonably be filled by a separate system of preventive detention is narrow, it does exist. In the third section, I therefore propose, first, criteria for measuring the public-safety “effectiveness” of such a system and, second, a zone of “tolerable” restrictions on civil liberties. I then draw in the fourth section on these criteria and the prior parts to make proposals for a Canadian system of preventive detention. I argue in favour of a model that focuses on disruption via short-term detention of persons tied to specific threats, and reject approaches that detain on the basis of perceived inherent dangerousness, unconnected to such specific threats. I label this approach a system of “catch and release.”
It is important to understand at the outset what this study means by “preventive detention.” As Elias (2009, 108) argues persuasively, the concept is ill-defined, and different authors and commentators use the term to encompass different concepts. Elias herself examines practices from 32 states, practices that she categorizes into “pre-trial detention,”4 “immigration detention”5 and “national security detention.”6
For the purposes of this study, however, I shall follow Harding and Hatchard, at least in part, in defining “detention” as “detention without trial by the executive acting under specific constitutional, statutory or other legal powers” (1993, 4). Detention, for my purposes, is physical restraint on a person’s liberty, most often through incarceration. A “preventive” detention is one “where a person is detained for reasons…connected with national security or public order or safety” (1993, 4).7
Looked at in this fashion, preventive detention is a form of “administrative” detention that need not culminate in criminal charges against the detainee and that is designed to prevent future behaviour rather than to punish or rehabilitate in response to past actions. By “administrative” detention, I mean detention by the executive in which “courts are responsible only for considering the lawfulness of this decision and/or its proper enforcement, but not for taking the decision [on whether to detain or not] itself” (United Nations 1989, para. 17). That is, this is not a criminal-law process in which a court proceeding results in a conviction and a sentence. However, the government decision may (and indeed, in a system based on the rule of law, must) be subject to a challenge as to its lawfulness in a court.
The motivation for designing or introducing a special system of administrative preventive detention rather than relying on more conventional instruments such as criminal arrest and trial may vary. For instance, one core reason, which is discussed below, is the reluctance of the state to disclose sensitive intelligence information, as would be required by conventional criminal proceedings. However, the underlying rationale for preventive detention is straightforward: preemption. Put simply, it prevents terrorists from engaging in terrorist acts. It does so because the terrorist or some key co-conspirator is detained and unable to carry out his or her intended course of action. An “effective” system of preventive detention is, therefore, one that meets this preemptive objective, whatever the other motivations for its creation.
By definition, meeting such an objective demands a high degree of prognostication. Unlike in classic criminal investigations, the state is charged not with conducting a forensic assessment of past events, but rather with anticipating a person’s future actions. The more preemptive the system — that is, the further back in time a detention is from the actual committing of a terrorist act — the more foresight is required of the state, and the more difficult the task of discerning who poses a risk and who does not. On the other hand, early preemption may be the most effective means of forestalling terrorism — to stop the suicide bomber after the bomb is strapped to his or her body may be an impossible task.
In many modern, developed states, preventive detention on national security grounds has been considered an aberrant practice, in part because of its association with more repressive regimes. Indeed, when Steven Greer reviewed the practice of preventive detention in 17 countries in 1993, he concluded that “[p]reventive detention on the grounds of public or state security is a flimsy and highly suspect justification for the deprivation of liberty. Abuse of power is seemingly widespread throughout the jurisdictions surveyed here…[M]uch more rigorous criteria than generally apply ought to be met if the practice is to be convincingly defended” (1993, 36).
Since 1993, however, preventive detention has become more common, even in those countries generally hostile to the practice. As noted, the most famous example is the long-term detention of suspected terrorists by the US military at various locations outside of (and occasionally within) the United States. However, other countries too, including ones with which Canada shares a common legal heritage, such as the United Kingdom and Australia, have developed highly refined forms of preventive detention. In this first part of the study I provide a brief overview of these systems. In so doing, I canvass a number of different approaches that are, broadly speaking, preemptive in the sense that their use may be causally related to the disruption of terrorist activity. Not all, however, are proper candidates for the concept of “preventive detention” as it is used in this study; instead they respond to concerns sometimes removed from the core, preemptive objective.
United States laws on preventive detention are paradoxical. On the one hand, that country has pursued an aggressive and controversial approach to the detention of terrorism suspects by the military and the Central Intelligence Agency, usually offshore. On the other, US domestic laws generally do not permit security-related preventive detention.
Since 9/11, the United States has relied on the law of armed conflict to justify not just battlefield detentions, but also the indefinite detention of unlawful “enemy combatants.” The latter concept has been defined broadly, so as to include not just unprivileged belligerents — speaking generally, combatants who are not members of armed forces — in conventional theatres of conflict, but also any and all affiliates of al-Qaeda.8 The 2006 Military Commissions Act, for instance, defines “unlawful enemy combatant” as including “a person who has engaged in hostilities or who has purposefully and materially supported hostilities against the United States or its co-belligerents who is not a lawful enemy combatant (including a person who is part of the Taliban, al-Qaeda, or associated forces).”9
These sweeping definitions of unlawful enemy combatant are an outgrowth of the US position on the conflict with al-Qaeda, namely, that it is a global “noninternational armed conflict” — that is, a conflict that spans more than one country’s territory between states (the United States and its allies) and a nonstate actor in the form of al-Qaeda. As such, the laws of war (including those permitting the use of military force and indefinite detention pending the end of the armed conflict) apply in all dimensions of what was once called the “war on terror,” and without geographic bounds.
In these circumstances, the reach of military detentions is potentially vast, extending into areas where other systems of detention would depend on more conventional legal instruments. Obama officials — including Attorney General Eric Holder and Solicitor General (and now US Supreme Court nominee) Elena Kagan — agreed at their Senate confirmation that military detention extended to a situation in which US intelligence agencies “capture someone in the Philippines that is suspected of financing al-Qaeda worldwide”(D. Savage 2009).10 These statements are broadly consistent with positions taken during the Bush administration, including by US Department of Justice lawyers, who once reportedly argued that “it was within the president’s lawful discretion to imprison as an enemy combatant even a ‘little old lady in Switzerland’ who had unwittingly donated to al-Qaeda” (C. Savage 2010).
The US approach to the reach of the laws of war in the campaign against terrorism is best described as aggressive, and a fuller exploration of it is beyond the scope of this study. It suffices to say that the laws-of-war approach to counterterrorism is of modest relevance in countries such as Canada that have not adopted such doctrines.
Perhaps because of the US focus on military detention, civilian security-related preventivedetention regimes are relatively underdeveloped in the United States, compared to some of the other countries discussed in this study. Indeed, US civilian domestic laws generally reject national-security preventive detention per se. The then US Assistant Attorney General, Viet Dinh, argued in 2003 that “we do not engage in preventive detention. In this respect, our detention differs significantly from that of other countries, even our European partners…What we do here is perhaps best described as preventative prosecution” (2003, 223). By “preventative prosecution,” Dinh meant a policy of aggressively enforcing even minor criminal and regulatory laws against those feared to pose a terrorist threat.11
It is not entirely true that US law does not include special terrorism-related preventive-detention provisions. The USA PATRIOT Act, for example, amended US immigration law to allow for mandatory detention of aliens certified a terrorist risk, pending removal, or for successive six-month periods “if the release of the alien will threaten the national security of the United States or the safety of the community or any person.”12 As Viet Dinh’s statement suggests, however, the US government preferred to use stringent enforcement of more conventional immigration laws in the aftermath of 9/11 rather than resort to this special antiterrorism provision (United States 2002, 17; United States 2003).13 These conventional infractions include remaining past the expiration of visas, entering the United States without inspection and entering the country on invalid papers.
This policy — also called “preventive charging” (Chesney 2005, 31) — is benign in principle, but its application produced controversy in the United States. Almost eight hundred people with suspected terrorism connections were detained on conventional immigration-law grounds in the immediate wake of 9/11. A 2003 US Department of Justice report reviewing this record concluded that administrative practices associated with these detentions produced unduly lengthy periods of incarceration and in some instances unsatisfactory conditions of detention (United States 2003, 195). The median length of detention between arrest and release or removal from the United States was in the range of 100 days (United States 2003, 105).14
The United States has also employed so-called material-witness laws to detain suspected terrorists. US federal law allows a judge to order the arrest of a person if the testimony of that person “is material in a criminal proceeding, and if it is shown that it may become impracticable to secure the presence of the person by subpoena.” A person is to be released “if the testimony of such witness can adequately be secured by deposition, and if further detention is not necessary to prevent a failure of justice.” However, “release of a material witness may be delayed for a reasonable period of time until the deposition of the witness can be taken.”15 Like preventive charging, this practice too has been criticized as a form of de facto preventive detention (see, for example, Levenson 2002).16
The post-9/11 US practice of stretching existing legal instruments to allow preventive detention without developing a frank preventive-detention law has attracted criticism. As Alan Dershowitz has argued, laws are contorted to serve the end of preventive detention but without the accountability a true preventive-detention law should have (see, for example, Dershowitz 2006, 118). Not the least of problems is that there is no jurisprudence that carefully examines the propriety and limits of this de facto preventive detention.17
Much current US discussion on “preventive detention” focuses on how to unwind the conundrums created by the Bush administration’s “enemy combatant” policy, discussed above. One core problem in converting military detention into civil incarceration relates to the treatment of detainees in military or paramilitary (that is, CIA) detention. A number of the persons so detained were also abused, and whatever incriminating information was obtained by their interrogation is therefore inadmissible in conventional courts.18
Because of this maltreatment, a subtext of at least some of the “too dangerous to release but cannot be tried” debate that currently undergirds the discussion on preventive detention in the United States is “cannot be tried because were tortured or subjected to cruel, inhuman or degrading treatment” and “cannot be released because were ‘warped’ by their incarceration.”19 Clad in the discourse of detention under the laws of war, the US discussion also seems to focus on inherent dangerousness — the person’s past pattern of behaviour and views renders them a danger, even without focused consideration of whether that dangerousness has become manifest in the form of an actual threat.20
By the time of this writing, the Obama administration seemed to have retreated from earlier statements suggesting it would seek congressional blessing for a new system of preventive detention. Instead, it will persist in detaining the detainees under presidential war powers that were acknowledged by the post-9/11 congressional resolution authorizing use of force against the Taliban and alQaeda (Baker 2009). Whether it will revisit this approach in the future remained to be seen.
Detention without charge
In comparison to the US approach — that is, a mix of ad hoc recourse to existing law and to the law of armed conflict — the United Kingdom has pursued a highly regimented system of detention without criminal charges that, at least in part, can be characterized as preventive.
Provisions enacted in their original form by the UK Terrorism Act 200021 allow a constable to arrest without a warrant a person whom he or she reasonably suspects to be a terrorist (section 41) — that is, a person suspected of having been “concerned in the commission, preparation or instigation of acts of terrorism” or having committed any of several enumerated offences (section 40). The reference to preparation or instigation gives these arrest powers a strong preemptive bent, although without a clear focus on the imminence of a feared harm.
The 2000 Act permitted a maximum detention (with judicial authorization) of 7 days (sch. 8, paras. 29 and 36), which was extended to 14 days in 2003.22 In a highly controversial amendment passed as part of the Terrorism Act 2006, detentions may now be renewed for supplemental periods of 7 days, up to a maximum of 28 days.23 In the 2006 Act, the grounds for the extended detention were also modified. Detention may be renewed, not simply to allow questioning, but also “pending the result of an examination or analysis of any relevant evidence or of anything the examination or analysis of which is to be or is being carried out with a view to obtaining relevant evidence” (section 24). Under these circumstances, the detention amounts to “investigative” rather than truly preventive detention — that is, incarceration motivated by the desire to pursue an investigation and not as a last-ditch effort to forestall an imminent harm. I discuss the concerns with systems that fold investigative and preventive detention into single procedures below.
The UK has also adopted a controversial system of so-called “control orders” that, while not detention per se, impose sometimes sweeping restrictions on a person’s freedom. A control order imposes obligations on a person “for purposes connected with protecting members of the public from a risk of terrorism” (Terrorism Act, section 1). They come in two species: nonderogating control orders — those that do not constitute a violation of the European Convention on Human Rights — and derogating control orders — those that would amount to a violation of the European treaty unless a proper derogation to the treaty is made by the UK government.24
Nonderogating orders may be made by the Home Secretary, subject to limited judicial supervision, and they last for up to 12 months, or longer with extensions. Derogating measures require a more substantial judicial review and blessing; they last for up to 6 months, longer with extensions. Control orders may, in other words, be renewed indefinitely. Moreover, the penalty for disobeying these control orders is up to five years’ imprisonment (Terrorism Act 2006, section 9).
These measures are imposed where “necessary for purposes connected with preventing or restricting involvement by that individual in terrorism-related activity.” Thus, for nonderogating orders there must be reasonable grounds for suspecting that the individual is or has been involved in terrorism-related activity and that the order is necessary to protect members of the public from a risk of terrorism (section 2). Derogating orders, by comparison, require greater certainty concerning the person’s involvement in terrorism-related activity (section 4).
Terrorism-related activity is a broadly defined term that includes actual involvement in the commission and preparation or instigation of acts of terrorism, as well as “conduct which gives encouragement to the commission, preparation or instigation of such acts.” Indeed, it reaches as far as conduct that supports or assists persons believed to be involved in terroristrelated activity, including those who encourage the instigation of such acts (section 1).
The constraints that may be imposed by control orders are extensive. They include limits on possession of certain articles or use of certain services or facilities and the carrying on of specified activities. Restrictions may be imposed on the nature of employment, membership in associations or communications with other persons. Control orders may also regulate a person’s place of residence and those who can have access to it and may place limitations on a person’s presence in certain places or movement within or from the United Kingdom. The person may be required to allow searches of himor herself or his or her residence, to wear electronic monitoring equipment and to report to the authorities (section 1). In their strictest form, these control orders could confine the suspect to a particular place, thus amounting to de facto detention. The last sort of constraint could be achieved only through a derogating control order.
In 2005, Australia introduced its own system of “preventative detention” and control orders with the Anti-Terrorism Act (No. 2).25 Australia also has a separate system of outright investigative detention.
Under the Anti-Terrorism Act (No. 2), a person may be preventively detained if the police have reasonable grounds to suspect that he or she “will engage in a terrorist act” that is imminent (or at least expected to occur within 14 days) or “possesses a thing that is connected with the preparation for, or the engagement of a person in, a terrorist act” or “has done an act in preparation for, or planning, a terrorist act.” The preventive detention must substantially assist “in preventing a terrorist act occurring” or “preserve evidence of a terrorist attack that occurred within the last twenty-eight days” and must be reasonably necessary for these purposes (section 105.4). The reference to preserving evidence permits preventive detention to be used as a form of investigative detention.
An issuing authority (usually a judicial official) may authorize preventive detention for an initial period of up to 24 hours (section 105.8), with renewals possible for up to an additional 24 hours. However, Australian state law may be used to increase the period of detention to a period of up to 14 days.26 The Anti-Terrorism Act (No. 2) is also notable for including severe penalties for the disclosure by the detainee and his or her lawyer, among others, of the existence of the preventive detention in all but limited circumstances (section 105.41).27
The Australian control order regime allows a court, on application by the government, to impose obligations on a person if it is satisfied, on a balance of probabilities, that “making the order would substantially assist in preventing a terrorist act” or that “the person has provided training to, or received training from, a listed terrorist organisation.” Furthermore, the court must be persuaded on a balance of probabilities that “each of the obligations, prohibitions and restrictions to be imposed on the person by the order is reasonably necessary, and reasonably appropriate and adapted, for the purpose of protecting the public from a terrorist act” (section 104.4).
The Act enumerates the restraints that may be imposed by control orders. These include such things as prohibitions or restrictions on the person’s being at (or away from) specified places, communicating or associating with specific individuals, accessing certain telecommunications or technology, possessing specific substances or engaging in specified activities. A control order may also include requirements that the person wear a tracking device, report to specified persons at specified times and places or participate in specified counselling or education. A control order may last for up to 12 months for adults and for a briefer period of time for persons aged 16 or 17 (section 104.5).28
In 2003, Australia amended the Australian Security Intelligence Organization Act (ASIO Act)29 with the Australian Security Intelligence Organization Legislation Amendment (Terrorism) Act, to enhance the anti-terrorism investigation powers of the Australian Security Intelligence Organization (ASIO).30 The ASIO is Australia’s equivalent of the Canadian Security Intelligence Service (CSIS); it is responsible for collecting and analyzing security intelligence (section 17). Under the new provisions, the ASIO, with the authorization of the relevant minister (section 34D), may seek a warrant from an “issuing authority” — that is, a magistrate or judge — for the questioning of a person in relation to a terrorism offence. If the issuing authority is persuaded that reasonable grounds exist for believing that the warrant will substantially assist the collection of intelligence in relation to a terrorism offence, he or she may issue a warrant valid for a period no longer than 28 days (subsection 34E(5)) and subject to renewal.
The person named in the warrant may then be questioned or asked to produce records or things in front of a more senior judge regarding information “that is or may be relevant to intelligence that is important in relation to a terrorism offence” (subsection 34E(4)).
A person may be questioned for up to a total of 24 hours (48 hours if he or she uses an interpreter) (section 34R) and must provide the information requested, if it is in his or her possession, even if it would incriminate him or her. Those statements are not admissible in criminal proceedings against the person (subsection 34L(9)), although there does not appear to be a bar on evidence discovered as a result of these statements being so used.
The warrant may also authorize the detention of the person for up to 168 hours (section 34S) if there are reasonable grounds to believe that otherwise the person may fail to appear for questioning, may “alert a person involved in a terrorism offence that the offence is being investigated” or “may destroy, damage or alter a record or thing the person may be requested in accordance with the warrant to produce” (subsection 34F(4)).
The existence of the warrant is secret while it is in force, and disclosure of the warrant or any fact relating to its existence or content or to the questioning or detention of the person subject to the warrant during this period is prohibited, under pain of criminal sanction (section 34ZS).
In assessing these various models, it is important to distinguish between them. As noted, current discussions in the United States relating to so-called “preventive detention” are less about “preventive detention” in the sense of disrupting imminent threats than about the indefinite, administrative detention of people perceived as inherently dangerous under a law-of-war model. This is, in other words, a very different system from the model I shall propose later in this paper.
The UK system of detention without charge and the Australian preventive-detention mechanism are more obvious candidates as models of preventive detention. The Australian approach, in particular, is clearly intended to disrupt imminent terrorist activity by detaining persons connected to it. Both systems do, however, reach further.
Specifically, to the extent that US, UK and Australian laws permit detention for the purpose of investigation, they are potentially preemptive but only in the broadest sense of the term. That is, the fruits of the investigation may culminate in measures that stop a terrorist activity. Investigative detention is not, however, truly a system of preventive detention as I describe the concept below.
There may be justifications for investigative detention. However, using incarceration to extract information from potential witnesses raises a host of difficult issues very different from the detention of persons to disrupt a feared imminent terrorist activity. For one thing, the right to silence is a key issue, one that might be resolved by precluding the use in court of the fruits of information disclosed during the investigative detention.
More difficult is the potentially broader scope of investigative detention. The ranks of witnesses presumably outstrip the number of persons who pose an imminent security risk — the US use of material-witness detention laws is proof of this. To bundle together preventive detention with investigative detention risks inflating the scope of such measures.
Nor is it clear that the attributes necessary for a system of detention that effectively preempts imminent harm must be replicated in investigative detention — powers to act in an exigent fashion, without prior judicial blessing, for example, are more questionable in the case of investigative detention. Given these differences, it makes little sense to create a single system of investigative and preventive detention, one that contains the strong measures necessary for the latter and extends these to the broader class of people who may be of merely investigative interest.
In short, if there is merit to a system of investigative detention, then that issue should be debated separately and not conflated in discussion or design with preventive detention.
The UK and Australian control-order regime is also potentially preemptive by restricting the freedom of persons with a history of terrorism-related conduct. Notably, the control-order systems existing in the United Kingdom, in particular, cover much of the same conduct that is criminalized in that jurisdiction, especially with the extended number of terrorism offences recently introduced in the United Kingdom (Carlile of Berriew 2009, para. 4).
However, control orders allow the government to restrict the freedom of individuals the government does not wish to prosecute because of the disclosure requirements that such prosecutions would require (Carlile of Berriew 2009, para. 47; United Kingdom 2009, 1 and 6). In some important respects, therefore, control orders are less a system for preventing immediate dangers than a system of “criminal law lite”: that is, they place restrictions on individual freedom for actions that are also crimes, but they reduce the government’s burden of proof and its disclosure obligations. Again, there may be merit to a system that allows freedom to be restricted for lengthy periods of time via secretive administrative proceedings, but if so then the justifications for these measures are likely very different than those advanced for preventive detention and any conflation of the two confuses the issue. A defence of preventive detention should not depend on whether control orders are an appropriate tool of anti-terrorism.
Iturn now to a close analysis of the Canadian legal environment and the legal utility (if any) of a system of preventive detention. This section examines immigration security certificates; tools of surveillance; and criminal-law arrests and prosecutions and recognizance with conditions (“peace bonds”). In discussing these measures, I am not implying that they are tools of preventive detention per se. They are, however, the legal mechanisms at present available to government to forestall terrorism. Assessing their adequacy as a form of de facto preemption is, therefore, necessary for understanding whether Canada requires a franker system of preventive detention. This assessment is particularly important for immigration security certificates, which have been used on a de facto basis to detain terrorist suspects (or otherwise strictly curtail their freedom) without criminal charge for long periods of time.
Immigration law — most notably in the form of “security certificates” — permits the detention of noncitizens under the Immigration and Refugee Protection Act pending their removal from Canada.31 Removal by means of a security certificate is intended to be speedier than deportation via the regular inadmissibility proceedings conducted in front of immigration adjudicators. A security certificate goes directly to a Federal Court judge, whose decision on the reasonableness of the certificate is final and amounts to a removal order. Decisions by immigration adjudicators are subject to more layers of possible review.
However, as a practical matter, security certificates have proved unworkable as a tool of antiterrorism. For one thing, immigration law has one obvious flaw as a tool of preemption: it applies only to noncitizens. As the UK House of Lords observed in invalidating a similar detention regime under UK law,32 there is no reason to assume that foreign nationals (as opposed to UK nationals) present the greatest terrorist threat. This has certainly proved to be the case in the United Kingdom, and increasingly so in Canada.
What is more critical is that the certificate system is triggered by government perceptions of the inherent “dangerousness” of an individual, not necessarily by a close assessment of whether that person is actually engaged in conduct likely to culminate in an injury to public safety or national security. Certificates may be an appropriate system of removing undesirables who have no right to be in Canada in the first place, so long as the injury inflicted on the removed individual is confined to the fact of removal itself and does not include some other disadvantage. It is, for example, an adequate system for ridding the country of sleeper agents from foreign states, that is, persons who are agents (and not opponents) of their countries of origin. However, security certificates are much more problematic — and raise issues well beyond simply immigration control — where government perceptions of dangerousness lead to long detention or efforts to deport people to countries where they may be tortured.
In this last respect, unlike in past cases involving Russian sleeper agents, removal has been vigorously contested by the persons accused of terrorist affiliations, each of whom risks removal to a country with (at best) a dubious record on torture. Effectively, the Canadian security certificate labels (or reveals, depending on the truth of the matter) these individuals as members of organizations whose purposes are invariably at variance with the interests of their countries of origin, to which they may be removed. In these circumstances, it is to be suspected that the authorities in those states will have their own agendas to pursue with these individuals, giving rise to a prospect of torture that does not exist (or at least is not as acute) with the removal of state sleeper agents.
In the event, these individuals have been detained (or, more recently, subject to strict restrictions on their freedom) for very long periods of time as the constitutionality of each element of the security-certificate process is tested up and down the court system. For these people, caught between the rock of removal to torture and the hard place of limitations on their freedom in Canada, security certificates have become a de facto system of outright detention or, more recently, UK-style “control orders.”
This scenario seems likely to recur each time a security certificate is used against a person suspected of being a terrorist or terrorist affiliate, and removal is to a country that has a poor record on torture and whose own national-security interests are triggered by the individual’s suspected acts or affiliations. Even if, instead of security certificates, the government resorts to the regular inadmissibility proceedings under the immigration law, there is the same incentive to contest removal and the same constitutional issues in terms of detention and removal to torture.
On balance, the reliance on security certificates is an entirely unsustainable anti-terrorism strategy. Long-term detention or strict limitations on freedom without a criminal charge in an ultimately fruitless attempt to remove persons may have the effect of “outing” potentially dangerous persons and limiting their freedom of action. It risks, however, being disproportionate to the actual threat posed by these individuals, especially as detention periods lengthen.
By the time of this writing, the government’s case in two of the five certificate matters had foundered. In Charkaoui,33 the government withdrew its case rather than abide by a court order requiring more disclosure to Charkaoui and his lawyers. Even more tellingly, in Almrei,34 the government lost on the merits of its case; that is, the court concluded that the certificate was not reasonable, especially with the passage of time since it was first issued.
Along the way, the court held that the government had committed an abuse of process by failing for years to disclose to the court evidence that was detrimental to its case (not the least being information placing in serious doubt the credibility of informants). Reviewing the handling of the case by CSIS and specifically its conduct in relation to confidential informants, the court commented that the Service’s mediocre performance “suggests a serious lack of analytical capacity in managing the enormous volume of information collected by the Service” (at para. 164). That comment followed similar 11th-hour revelations concerning the credibility of confidential sources in yet another security-certificate case — Harkat.35 In that case, the judge raised concerns about “possible prevarication by CSIS witnesses called to testify concerning the reliability of the information provided by the human source” (at para. 12). After a subsequent hearing on this matter, the court pointed to systemic problems in the manner in which CSIS had handled the case, and concluded that its conduct had undermined the integrity of the court’s process.36
It seems likely that if there are any more security-certificate cases, the government will handle them very differently; certainly it should meet its disclosure obligations more scrupulously, especially given that those disclosure expectations are now clear. On the other hand, the system is now in serious disrepute, and criticism of the device as a heavy-handed anti-terrorism tool is part of the mainstream.
Critics of security certificates — this author included — have argued that better means of preventing terrorism do exist.
The most obvious candidate is prosecution under the criminal law. Canada’s anti-terrorism criminal law now reaches very far in criminalizing conduct associated — perhaps quite distantly — with actual acts of terrorist violence. For one thing, the definition of “terrorist activity” in the Criminal Code includes “inchoate” offences — conspiracies, attempts, threats, accessory or counselling — tied to the underlying acts of violence constituting the terrorist activity.37 When considered alongside the specific new criminal offences created by the 2001 Anti-terrorism Act — many themselves inchoate offences — it layers inchoate offence upon an inchoate activity in a manner that Kent Roach (2002, 159) calls the “piling of inchoate liability on top of inchoate crimes.” Criminalizing actions that are mere preparation is unusual in the Canadian criminal-law tradition, with its focus on reaction rather than preemption. Still, as proponents of the new model argue, “the nature of terrorism requires a different approach to disrupt and disable the terrorist network before it can carry out its design” (Mosley 2002, 152). Put another way, Canadian anti-terrorism criminal law is now preemptive in its design, more so than most (if not all) other areas of the criminal law.
It is true that much of this law postdates the alleged behaviour at issue in relation to those persons currently subject to security certificates. Thus, it cannot be applied retroactively to these individuals. It is, however, a tool that may be deployed in the future in relation to other individuals, a fact reflected by the now notable recent history of successful prosecutions.38 This is a view apparently shared by RCMP Commissioner Elliot, who called recently for a shift in antiterrorism strategy from the intelligence-dependent security-certificate approach to an evidencedriven criminal-law approach (Elliot 2009).39
However, for the government, abandoning administrative proceedings, with their reduced burdens of proof, for the much more demanding criminal process will increase the demands for cogent evidence. Moreover, unlike the security-certificate process, criminal proceedings are open. Accordingly, if the evidence to be used against an individual comes from an allied security service, extreme sensitivity over its disclosure would likely terminate a prosecution — the so-called “third party rule” or “originator control” rules disallow unauthorized disclosure of foreign-provided intelligence. Such disclosure would constitute a damning sin in intelligence circles.40 Likewise, domestic security services are extremely wary of having their own sources and techniques examined in open court.
The Canada Evidence Act does provide a means for protecting this information from disclosure.41 It also allows a trial judge to terminate the case if nondisclosure would produce an unfair trial. Consequently criminal trials are possible in the three following situations:
Option 3 appears to be the situation in the “Toronto 18” proceedings. Option 2, in effect, arose in the Khawaja prosecution (through the operation of the Canada Evidence Act).43 Option 1 presents the thorniest issue. The bottom line is that every state faces the dilemma of the third-party rule in terrorism cases because modern terrorism straddles borders and implicates a whole web of information-sharing networks. To the extent each state treats the third-party rule as absolute and refuses to consent to disclosure in the courts of another state, each pursues a policy that may, applied against it, imperil its own prosecutions. The result is a sort of intelligence “beggar thy neighbour” approach.
In these circumstances, states should consider requests for disclosure of their shared intelligence carefully (and not dismiss them out of hand or demand secrecy for even the most banal information). Indeed, it would seem wise to develop international protocols among intelligence allies, perhaps initially between states with similar legal systems, on how and where information can be disclosed in court proceedings. These protocols would further invalidate CSIS’s complaint in some of the security-certificate cases that even asking for permission to disclose intelligence provided by foreign sources creates the impression that Canada is not to be trusted with confidences.44
On the other hand, even with the most robust protocol and the most earnest effort to secure consent to disclosure from the intelligence-providing state, there will be instances where none of the three above-noted possibilities exist. In these circumstances, a criminal-law prosecution would be impossible.
This impossibility of prosecuting under the criminal law where disclosure of foreign-sourced information critical to the prosecution is withheld may make the criminal law an imperfect instrument of preemption in some cases. It does not, however, seem likely to impede the preemptive potential of that criminal law writ large. Foreign intelligence may be essential to prosecutions for conduct that takes place overseas. However, it is likely of much less significance where the crime is committed in Canada. As noted, terrorist crimes are now defined quite broadly, and they include conduct that may be quite remote from an actual terrorist attack. This means that while the wish to preserve the confidentiality of foreign intelligence may deter the government from arresting and prosecuting a suspected terrorist affiliate immediately on his or her arrival in Canada, any subsequent acts by the person in furtherance of a terrorist plot very quickly amount to separate criminal offences, for which the suspect may be arrested, charged and tried. Evidence of that conduct would arise from the individual’s Canadian activities, as captured by surveillance, monitoring and investigation by Canadian security services themselves. Put another way, option 3, described above, would be available to the state.
In considering whether this prospect is sufficiently preemptive, it is important to understand the scope of these surveillance and search powers and the subsequent powers of arrest.
Surveillance and search powers
Actions taking place in public — or more precisely, in circumstances where there is no reasonable expectation of privacy — may be observed and monitored by state agents at their discretion. Surveillance or search where there is a reasonable expectation of privacy generally requires judicial authorization.45 Thus, full police searches must usually be authorized by a warrant on reasonable and probable grounds to meet the privacy protections offered by section 8 of the Canadian Charter of Rights and Freedoms. For instance, CSIS may apply for such a warrant if it “believes, on reasonable grounds, that a warrant…is required to enable the Service to investigate a threat to the security of Canada.”46 The latter concept includes “activities within or relating to Canada directed toward or in support of the threat or use of acts of serious violence against persons or property for the purpose of achieving a political, religious or ideological objective within Canada or a foreign state” (section 2).47
Likewise, warrants for intercepts of electronic communications are available to lawenforcement agencies under the Criminal Code. As noted, the Criminal Code’s wealth of terrorism-related crimes are categorized as “terrorism offences.” The broad category compounds the piling up of inchoate actions noted above.48 Because of the way “terrorism offences” are employed in the Criminal Code, this compounding of inchoate offences extends the range of circumstances for which electronic intercept authorization may be granted. Furthermore, the rules on intercepts in investigations of terrorism offences are more permissive of monitoring than are those that apply to other crimes. For instance, warrant authorizations may last for longer under the Criminal Code electronic-intercept rules in terrorism cases — for up to 12 months — and the authorizing justice must review fewer considerations prior to issuing the warrant.49
The relaxation of the intercept authorization requirements for these terrorism offences — especially when compounded by the breadth of activity captured by the piling of inchoate activities in the definition of “terrorism offence” — contributes to what Cohen (2005, 203) has described as the post-9/11 state’s “enlarged capacity to peer into the individual’s private life,” including its power to “examine the trappings of simple associations and lifestyle, and ambiguous activities that the jurisprudence has often described as ‘mere preparation.’”
Warrants for intercepts are sought on an ex parte basis — that is, in the absence of the party who will be under surveillance. The issuing of CSIS warrants in particular is very confidential, and presumably sensitive intelligence is employed to satisfy the reasonable-and-probablecause standard required to obtain the warrant, without any fear of disclosure.
It should be noted that law-enforcement agencies have other, limited search powers that may be used even in the absence of the reasonable and probable grounds typically required by section 8 of the Charter. Screening searches may be authorized by statute at public facilities, such as courthouses. These searches are indiscriminate in the sense that they are not based on suspicions of any one individual, but they may still be constitutional if they are reasonable as measured both by the need for the practice and the manner in which the searches are conducted.50
Meanwhile, mere “reasonable suspicion” is the basis for searches in certain special situations, including at border crossings.51 There may also be other circumstances in which “reasonable suspicion” justifies certain sorts of relatively nonintrusive searches — such as by sniffer dogs. The Supreme Court’s jurisprudence on this issue (see R. v. Kang-Brown) is deeply divided in relation to antinarcotics law.52 However, a majority of the court has implied (although never actually so ruled) that searches justified by “reasonable suspicion” may be permissible in “exigent circumstances relating to apprehended terrorist activities” (para. 13 [per LeBel J.] and para. 56 [per Binnie J.]).
Reasonable suspicion is a lower standard than reasonable and probable grounds. It is more than a hunch derived from experience, and it must still be grounded in “objectively ascertainable facts.” However, the likelihood of criminal conduct suggested by those facts must be greater to meet the “reasonable and probable grounds” than with simple “reasonable suspicion” (para. 75 [per Binnie J.).
The Supreme Court has also recently held that “police officers may detain [in the sense of a cursory stop] an individual for investigative purposes if there are reasonable grounds to suspect
in all the circumstances that the individual is connected to a particular crime and that such a detention is necessary. In addition, where a police officer has reasonable grounds to believe that his or her safety or that of others is at risk, the officer may engage in a protective pat-down search of the detained individual. Both the detention and the pat-down search must be conducted in a reasonable manner. In this connection…the investigative detention should be brief in duration and does not impose an obligation on the detained individual to answer questions posed by the police.”53 Put another way, an officer may briefly stop an individual on the basis of a particularized (as opposed to generalized) suspicion connecting the person to a crime to ask questions, but may not then search that person without more demanding reasonable grounds to believe that he or she is a danger. These reasonable grounds might stem, for example, from the circumstances in which the stop takes place (for example, if the crime to which the person is suspected of being connected involves weapons) or the reaction of the person to the initial stop.
Information amassed by the state through the investigation and surveillance described above may justify a full arrest. Under the Criminal Code, peace officers may make arrests if they have obtained a warrant or, in limited circumstances, without a warrant. Warrants are usually issued by a justice in response to a complaint of criminal conduct about a person (section 504 et seq.). In other words, they concern a crime that has already been committed and are designed generally to prevent the accused from failing to appear before a court.
However, warrantless arrest is also possible in limited circumstances. For instance, a peace officer may arrest without warrant a person he or she finds committing a crime or who, on reasonable grounds, the peace officer believes is about to commit an indictable offence (section 495 [emphasis added]). Thus there is a preemptive aspect to this power of arrest. The Supreme Court’s constitutional jurisprudence grafts on an objective test to this subjective belief by the peace officer: “[O]bjectively there must exist reasonable and probable grounds for the warrantless arrest to be legal.”54 As summarized by the Supreme Court: “[A] reasonable person placed in the position of the officer must be able to conclude that there were indeed reasonable and probable grounds for the arrest. On the other hand, the police need not demonstrate anything more than reasonable and probable grounds. Specifically, they are not required to establish a prima facie case for conviction before making the arrest.”55 The standard is therefore more than mere suspicion that an indictable offence has or will be committed by the person, but much less than the certainty required to secure a conviction.
There is no clear authority on how imminent a prospective indictable crime must be to permit this form of warrantless arrest. However, at least one provincial court of appeal has suggested that warrantless arrest under the Criminal Code is available to prevent harm likely to occur in the “immediate future.”56
A person arrested by a peace officer must be brought before a justice, without unreasonable delay and, in any event, within 24 hours57 except when a justice is not available in that period. In the latter instance, the Criminal Code specifies that the “person shall be taken before a justice as soon as possible” (section 503(1)). This last proviso may be designed to deal with circumstances where a person is arrested in a remote area, but it is of little relevance in urban areas (even when arrests are made on weekends).58
Once the accused is before the justice, he or she is under the control of the courts, and the regular criminal law arraignment and bail provisions apply. An accused denied bail is detained pending the outcome of the criminal proceedings. In practice, this may be a lengthy incarceration, but it is associated with a judicial process of determining the guilt or innocence of the person in question.
If the accused is found guilty, he or she is sentenced, often to a period of imprisonment. Because the sentence is for a specified length of time, detention is not indefinite. Indeterminate detention may, however, be imposed by a special process of declaring the offender a “dangerous offender.” This designation — which then justifies an indefinite sentence — is reserved for particularly brutal criminals or for people with a history of repeated and particularly harmful criminal activity.59 It is relatively uncommon: between 1978 and 2009, it was applied to only 488 offenders. As of April 2009, the 395 dangerous offenders then incarcerated represented 3 percent of the total federal inmate population (Public Safety Canada 2009). In R. v. Lyons, the Supreme Court considered the constitutionality of dangerous-offender status. Its reasoning strongly suggests that indeterminate detention under this regime was consistent with section 7 of the Charter only because it was part and parcel of the sentence handed down upon conviction for a criminal offence.60
Criminal charges, arrest and prosecution are not the only tools available to the state to limit the freedom of potentially dangerous individuals. For some time, the Criminal Code has permitted courts to impose recognizance with conditions — so-called peace bonds — if there are reasonable grounds to fear that a person might engage in various criminal acts. For example, a person who fears (on reasonable grounds) that an individual may commit certain personal injury offences (sections 810 and 810.2) sexual offences (section 810.1) certain offences relating to intimidation of the justice system or a journalist, or a criminal organization offence (section 810.01) may bring the matter to a provincial court judge (although in some cases only with permission of the attorney general).
After 2001, this list was expanded to include a terrorism offence (section 810.01). Under section 810.01, a person “who fears on reasonable grounds that another person will commit…a terrorism offence may, with the consent of the Attorney General, lay an information before a provincial court judge.” If the provincial court judge is persuaded that these reasonable grounds for the fear exist, he or she may order the defendant to “enter into a recognizance to keep the peace and be of good behaviour” for up to 12 months and may impose other reasonable conditions. A refusal by the accused to enter into the recognizance is punishable by imprisonment for up to 12 months. A breach of a recognizance is a criminal offence, punishable by up to two years’ imprisonment (section 811).
These peace bonds are obviously preemptive in that they apply to a person who has not committed an offence. On its face, the imposition of a peace bond on a potential terrorist seems a modest, even farcical, reaction to the threat of terrorist activity, especially in an era of suicide bombers. A peace bond is, however, intended to “out” potential threats and disrupt the “preparatory phase of incipient terrorist activity” (Cohen 2005, 218-19). Furthermore, a peace bond may be associated with detention. Unlike the expired section 83.3 discussed below, the regular peace-bond process does not authorize arrest without warrant. However, there may still be detention in the lead-up to the peace bond: once an information has been laid, a warrant may be issued for the person resulting in their arrest pending imposition of the peace bond by the judge.61
The potentially formidable reach of the peace bond should not be underestimated. A peace bond is a government-crafted, judicially imposed set of behavioural standards tailored to individual persons. The onerous conditions that may be imposed as a part of the recognizance can be easily broken, permitting the subsequent prosecution and incarceration of a feared security risk for behaviour that is benign in its own right. Put another way, a peace bond with a hair trigger could allow a state to use prosecution and punishment for easily proved and trivial violations of the peace bond as a means of achieving robust anti-terrorism preventive detention.62
Some sense of the scope of a peace bond might be drawn from the conditions placed on release pending the outcome of immigration security-certificate adjudications. These release agreements highlight the sort of measures a government is likely to view as necessary to hobble feared terrorist risks. For example, the conditions under which Mohammed Harkat was released in 2006 run to five pages. Among other things, any computer with Internet connectivity must be kept in a locked portion of his residence to which he has no access.63
All together, these rules leave law-enforcement authorities with notable powers of surveillance, disruption, detention and search in the area of anti-terrorism. A key question, therefore, is whether there is any remaining gap that necessitates a system of preventive detention. The Canadian government argues that such a gap does exist. In reintroducing a bill to restore the expired section 83.3 “recognizance with conditions” system discussed below, it argued that “[m]uch of existing criminal law is designed to find and punish those responsible for acts that have already occurred. This approach is often inadequate for terrorist crimes, which are aimed at creating fear and instability by targeting the general population and where the perpetrator may commit suicide when carrying out the attack” (Justice Canada 2009).
As the discussion above suggests, this statement exaggerates the reactive, rather than preemptive, qualities of contemporary anti-terrorism criminal law. Law-enforcement officials have offered a more nuanced justification for preventive detention, noting that the proposed power of recognizance with conditions is among the “last best chance” measures that should be available in an environment where “[n]o one can predict the precise nature of the threat that we will face tomorrow or in five years” (McDonell 2008).
Neither of these statements offers a perspective, however, on how large the gap in law-enforcement powers in the anti-terrorism area truly is. It is useful, therefore, to imagine a series of plausible hypotheses to test whether a gap does exists. I summarize these scenarios in table 1. These scenarios range from the indisputably noncriminal to the indisputably criminal.
Scenario A is the most acutely banal. While the individual’s suspicious wardrobe may suffice to attract scrutiny, the bulky coat alone does not rise to the level even of “reasonable grounds to suspect” that the person is involved with a crime. There is no basis for even a brief investigative stop, let alone a full arrest. A system of preventive detention, applied in these circumstances, would be clearly disproportionate — there is no basis to conclude a threat exists, above and beyond the usual security sensitivities associated with Parliament Hill. Of course, the individual would be subject to standard security screenings, depending on where he or she goes on Parliament Hill. He or she would also be subject to ongoing monitoring while visiting the Hill — the person has no reasonable expectation of privacy while wandering the parliamentary precinct.
In scenario G, by contrast, reasonable and probable grounds exist for arresting the person on the basis that he or she is about to commit or is committing an indictable offence. There is no need here for preventive detention; the threshold to actual criminal conduct has been crossed, as has the threshold for a criminal-law arrest. While the information stems from a confidential informant, that tip-off (or at least the identity of the informant) will not likely be fundamental to any ensuing prosecution since evidence proving the crime will (or will not) be found on the person of the accused — that is, the actual explosive device itself.
The scenarios lying between A and G are more complex. Scenarios B and C are variations on scenario A. In both cases, there is more reason to suspect the individual on Parliament Hill than in scenario A. More is known about that person, and his or her associations may create cause for concern. However, in scenario B, there is at best, suspicion by association. There is no criminal or prospective criminal offence. It seems unlikely (or at least very uncertain) that mere presence on Parliament Hill by someone who associates with a radical preacher creates the reasonable suspicion justifying an investigative stop, even if the person is warmly clad on a warm day. As with scenario A, the logical response is continued surveillance and standard security screening as the person enters controlled areas. Preventive detention in these circumstances would be unreasonable or at least premature — there is security sensitivity but no imminent threat of serious harm.
Scenario C alters that variable — there is now reason to fear an imminent terrorist attack of some sort on an unknown Canadian landmark. Parliament Hill obviously constitutes such a landmark. The person’s association with a radical religious leader who openly preaches violence now has a different hue, as does his or her seemingly inappropriate clothing. Even then, however, there are still no reasonable and probable grounds to believe that this person is about to commit an indictable offence. It seems very unlikely that a generic threat of such a crime from an unknown source, coupled with the presence at a potentially targeted place of a person associated with someone who espouses violence, could sustain the reasonable and probable grounds necessary for a full criminal-law arrest, even where the individual wears unusually bulky clothing.
There may, however, be reasonable grounds to suspect the individual, although there may be debate even as to whether that threshold has been met given the absence of any information linking the individual or the preacher to the feared attack. If there is a basis for such a suspicion — and I would argue that the bulky clothing worn on a warm day by this particular individual pushes this scenario into the realm of reasonable suspicion — then a brief investigative detention is permissible, during which nonintrusive searches by sniffer dogs might be proper (although, admittedly, the jurisprudence on this question is divided), and a full search of the individual may be legitimate in the interest of public safety. The latter search requires a reasonable belief — more than a suspicion — that the person poses a risk to public safety. That threshold would likely only be crossed if the individual’s conduct, once stopped, compounded the existing suspicions of the authorities.
Either way, the brief investigative stop likely makes a special power of preventive detention unnecessary. The person’s plans are likely disrupted, and the nonintrusive search by a sniffer dog and/or a full search precipitated by his or her actions in response to the stop may give rise to new information that will constitute the reasonable and probable grounds for an outright arrest.
In scenarios E and F, in comparison, there is no imminent peril. Rather, the individual is fingered as a terrorist affiliate and indeed as someone who has contributed to the activities of a terrorist organization. That involvement, even though it took place overseas, is a crime in Canada. The person could therefore be arrested and charged. The authorities may not, however, wish to pursue this course of action. The evidence concerning the individual’s actions will almost certainly come, at least in part, from foreign intelligence agencies. Canadian authorities will vigorously oppose the disclosure of information supplied by these agencies in open court, for fear of threatening the information-sharing arrangement with the foreign partner. While there are provisions in Canadian law — most notably under the Canada Evidence Act — that protect this information from disclosure, that protection would preclude a conviction where the only information supporting such a conviction was withheld from the defendant.
A system of preventive detention may satisfy this concern (at least in part) if it applies standards on disclosure that are less robust than those in the criminal law and allows detentions even in the absence of an imminent peril. It is not coincidental that scenario E involves a permanent resident; that person would be subject to immigration proceedings in which disclosure rules are not nearly as robust as in criminal proceedings. As part of an immigration security certificate, for example, the person may be detained and his or her removal from Canada considered by a Federal Court in possession of information that was never fully disclosed to the person himor herself. The citizen in scenario F is not subject to these immigration measures.
In either case, however, there are very good reasons not to resort to detention either under immigration measures or as part of a new system of preventive detention. Namely, if the authorities monitor this person surreptitiously, they may obtain useful intelligence that may foil the actions of the terrorist organization. That possibility evaporates as soon as the individual is detained. More important, preventive detention undertaken when there is no concrete and imminent threat is simply detention of perceived dangerous people. For reasons outlined below, I believe that such an approach would produce an inadequately focused system.
Scenario F becomes slightly more complex if instead of monitoring a returning citizen, the government wishes to impede the departure of that citizen — there are now several notorious cases of radicalized youths from Western countries venturing abroad in support of terrorist activities. There may be good reason to notify the security services of the country to which this person is travelling of any Canadian suspicions. These foreign agencies could then conduct their own investigations and might unearth more useful intelligence on a terrorist plot or conspiracy. Depending, however, on the foreign country in question, the Canadian notification may simply lead to the detention and maltreatment of this individual and to another Maher Arar-style scandal for the Canadian security services. In these circumstances, the only viable option may be to stop the person from leaving Canada in the first place. However, there are means of doing that which fall far short of preventive detention. Revoking a passport, for instance, makes international travel impossible without fake documents. And if the person were to attempt to travel on fake documents, that itself would be a crime for which the person could be arrested and preventive detention would not be necessary.
The most difficult scenario is D. Here, there is reason to fear an imminent terrorist attack of some sort on an unknown Canadian landmark. There is, however, less reason than in scenario C to believe that the individual is in fact involved with the actual planned attack; he or she is not at a landmark and is not acting in a particularly suspicious manner. The only connection between the individual and the attack is his or her membership in the suspected criminal organization. Scenario D can be generalized as a situation in which: (1) the state has reason to believe that an attack will occur; (2) it has reason to believe that a particular group is behind the plot and that the suspect is a member of that group; (3) but it has no information, other than this, connecting that particular individual to the plot.
Under these circumstances, there are no reasonable and probable grounds for an arrest — mere membership in a terrorist organization is not itself a crime. Nor is there any purpose to be served by an investigative stop, even if the legal threshold for such a measure were satisfied. The stop would not disrupt whatever plans the person may have for the future and would be unlikely to unearth information or evidence that could precipitate an outright arrest — the person is not near to any prospective target and is unlikely to have on his or her person (or betray in his or her manner) any evidence that would support an arrest.
Arguably, this is a scenario where preventive detention might prove a useful tool. But such detention would be risky. Detaining that person may do nothing more than alert his or her associates to the interest of the authorities without doing anything to preempt the attack. Unless the law-enforcement authorities obtain useful information from the person during questioning — again, an uncertain prospect if the person knows little or is uncooperative — the detention is counterproductive.
In scenario D, the most reasonable course of action might be to monitor the person in the hope that his or actions will lead the authorities to the perpetrators of the planned terrorist attack. Surveillance is, however, fallible, especially if it must strive to be covert. Faced with an imminent danger by a group with which the individual is associated, the investigators may reason that neutralization is to be preferred to surveillance, despite the investigative cost. If they do, there is little in Canadian law that would allow them to make that choice, other than a peace bond, which would fall short of outright detention.
In short, these scenarios suggest that there may be only the narrowest of “gaps” to be filled by a system of preventive detention: specifically, an imminent threat in circumstances where a person is suspected of being connected to that threat (but not to the point of satisfying the requirement for reasonable and probable cause for arrest for criminal activity) and where preemption through outright detention is regarded as less risky to public safety than continued surveillance.
In these exigent circumstances, peace officers could be driven to exercise conventional powers of criminal law arrest even in the absence of reasonable and probable cause, but only at the risk of violating the constitutional standards on arbitrary detention. Such arrests contort the rule of law in the interest of expediency and open the door to claims of abuse of process. They are not the basis for a sustainable system of preemption in a system built on the rule of law.
If there is a plausible argument in favour of a (narrow) system of preventive detention, the obvious next question is “What sort of system?” In this section, I propose a series of criteria for a system of preventive detention that effectively fills the “gap” noted above. I then discuss a number of civil-rights concerns that such a system would have to satisfy.
It should not be assumed that counterterrorism is always served by detention, preventive or otherwise. Much depends on the design of that system.64
On the basis of my definition of preventive detention, any plausible system of anti-terrorism preventive detention must clearly define what dangers the system seeks to prevent. Any ambiguity on this point renders efforts at prediction and preemption difficult, if not impossible.
Requirement of serious harm
First, I believe that the range of actions that the system tries to prevent must be modest. To prevent every conceivable species of harm is impossible and will inevitably strain limited state resources by diverting attention from the truly serious to the less serious. “Serious” in this context could mean that the injury from the terrorist activity would be both dire and widespread. However, a serious harm could be an injury that, although relatively modest for each person affected, is widespread and therefore has a large societal impact, or one that is very significant for a small number of individuals, that is, not widespread.
Requirement of imminence
More than being serious, the harm should also appear imminent. The system need not focus strictly on the individuals who themselves will engineer the imminent and serious harm. Rather the focus is on disrupting the harm, and not (solely or even mostly) on detaining those who are spearheading the action. If disruption is accomplished by detaining others further removed from the actual terrorist activity, the system still meets its objectives.
The imminence threshold, however, is vital. As discussed above, many of the existing approaches to administrative detention do not include this emphasis on imminent harm. In fact, people are detained on the basis of perceived inherent “dangerousness” and not because that dangerousness has manifested itself in serious harm or even will do so in the foreseeable future. However, a system that authorizes preventive detention for less than imminent harm depends on the notoriously fallible capacity of the state to predict harm, thereby risking the misallocation of resources and compounding the prospect of outright errors that undermine the state’s credibility. I return to this issue below. Furthermore, if detention is used in less than pressing circumstances, it interrupts other tools of anti-terrorism that may ultimately be more fruitful — such as surveillance and investigation designed to unearth more of the plot and co-conspirators.
Last best chance
In keeping with these observations on the importance on surveillance, any effective system of preventive detention must take account of other tools available to preempt serious and imminent harm and must be efficiently integrated with these alternatives. In some sense, preventive detention should be used only where the state’s surveillance, investigations, and intelligence sharing and collecting have failed to unearth a plot before it becomes an imminent threat or if those tactics can no longer be deployed to forestall the feared violence. At the same time, that threat cannot be effectively neutralized by using detention under the conventional criminal law. Preventive detention, in other words, is a “last best chance” when surveillance is no longer adequate and arrest on criminal charges not possible. Treating preventive detention as a “gap filler” that supplements conventional criminal law rather than usurps it permits a more persuasive defence of the measure, in political and perhaps in constitutional terms.
Finally, the system must also be based on fact. This must not simply be a dragnet or mechanism of mass detention. A dragnet is severely damaging to civil rights. Even from the perspective of effectiveness, however, an indiscriminate system of detention is undesirable. First, it seems doubtful that even the most sweeping dragnets will capture the right people. Unlike the innocents swept up by an indiscriminate system, the most dangerous people are likely equipped to avoid capture by the blunt mechanism of a dragnet. Except with the best of luck, an indiscriminate dragnet will at best capture only incidental figures in any terrorist plot.
Second, by detaining the innocent as readily as the dangerous (and probably more readily), this system corrodes public support and faith in the authorities. Interning entire classes of people alienates them, and damages relations between the community and the law-enforcement agencies. Whatever minimal gain in public safety comes from mass detentions is almost certainly lost by the hostility and resistance such actions provoke.
Third, a dragnet consumes an enormous amount of resources since incarcerating large numbers of people has a direct, material expense.
Fourth, a dragnet risks making the state lazy because the reliance on dragnets diverts investigative techniques away from the much more effective, fact-based investigations and procedures that are likely to produce real security gains.
For all of these reasons, an effective system of preventive detention depends on good intelligence and police work. State agents must avoid conventional pitfalls of any investigation: dependence on ethnic profiling, for example, or a tunnel vision in which the preoccupation becomes establishing the guilt of an individual already known to the authorities. Rather they must rely on a flinty-eyed assessment of facts to identify persons (known or unknown) who pose unanticipated risks.
In this last regard, an effective system of preventive detention must certainly appreciate that national security has few of what former US defence secretary Donald Rumsfeld famously called “known knowns” — things we know we know. As Rumsfeld said: “We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns — the ones we don’t know we don’t know. And if one looks throughout the history of our country and other free countries, it is the latter category that tend to be the difficult ones” (United States 2002).
In making these statements in 2002, Rumsfeld was dismissing the absence of actual evidence linking Saddam’s Iraq to terrorist organizations. Ironically, the resulting Iraq war is proof of how presumed “known knowns” may create their own “unknown unknowns”; that is, a policy of action built on resolute suspicions and fixed assumptions may precipitate its own crises and errors.
In practice, guarding against investigative myopia requires a system in which assumptions are constantly tested and retested by a “team B,” that is, by persons without a vested interest in the current set of presumptions. Early testing of the state’s case in front of independent judges is one form of “team B” scrutiny, but internal testing within the security services is also desirable.
In short, a system of preventive detention must be narrowly focused on preventing a carefully understood and circumscribed threat and must be triggered by fact-based assessments that are subject to careful scrutiny and review by persons sufficiently removed from the investigation. As a tool of preemption, preventive detention is a desperate last measure and not one that replaces investigative and surveillance techniques or supersedes the conventional criminal-law process.
A system of preventive detention that departed from this approach and permitted detention simply on the basis of the perceived “dangerousness” of the individual, without that dangerousness being linked to a specific and imminent threat, would be deeply troubling, even from an effectiveness (and not just civil-liberties) perspective. In such a system, the focus would inevitably become the person’s past behaviour, associations and beliefs and not a careful assessment of his or her capacity to cause serious harm.
The range of individuals potentially subject to detention under a “dangerousness” approach would inevitably be much greater than with a threat-specific system of preventive detention, and in the worst instances might come close to a dragnet. Even without the inevitable constitutional challenges that such an approach would ignite, the state would be distracted by its effort to establish inherent “dangerousness,” and the system would be prone to mistakes, exaggerated claims, bias and the inevitable controversy these errors precipitate. The new approach would inevitably be tainted by the credibility problems that have afflicted the government in connection with immigration security certificates, where external observers, including the courts, have come to doubt the government’s claims of dangerousness.
The Almrei security-certificate case is at least a partial case in point. The government alleged that “Almrei supports the extremist Islamist ideology espoused by Osama Bin Laden, that he has connections to persons who share that ideology and that, through his involvement in an international document forgery ring, [he] has the ability and capacity to facilitate the movement of those persons in Canada and abroad who would commit terrorist acts.”65 Almrei did forge documents, a crime for which he could have been prosecuted. He had also been present in Afghanistan in the mid-1990s and associated with protagonists in that country’s conflict during the period. Much of the government’s case (in the public proceedings) depended on an analysis of Almrei’s associations with Afghan-related figures that the government said showed him to be a member of a terrorist network. Put another way, the government argued that he was dangerous for whom he knew, what he believed, what he had done and the skills he could lend to a terrorist network, not because he was actually involved in the planning, facilitation or instigation of particular terrorist activity. Indeed, until the events of 9/11, CSIS was content to monitor Almrei, regarding him as a sleeper, while the RCMP investigated his document-forging activities (at para. 5).
But 9/11 changed this dynamic, and Almrei was now perceived as part of “a much greater threat to North American security as someone who had the skills and the contacts to arrange for terrorists to cross borders on forged papers” (at para. 5). This suspicion may have been reasonable in 2001 given what was known at the time, but it became less tenable with the passage of time and the availability of more accurate information. In the end, the court roundly rejected the government’s terrorism assertions, concluding among other things that “[t]he evidence does not provide reasonable grounds to believe that Almrei had any association with Bin Laden or opportunity to meet apart from a brief period of time when their presence in Afghanistan may have coincided…Rather, he went to camps run by Sayyaf and Khattab, neither of whom can be reasonably said to be part of Al Qaeda” (at para. 441).
Overall, the judgment strongly suggested that on top of the implausibility of claims made by secret informants, the government’s case suffered from a deeply flawed understanding of the political environment in Afghanistan, the concept of jihad, and the affiliations and ideologies of persons with whom Almrei associated. Put another way, the government’s assessment of Almrei’s dangerousness was grounded in large measure in misunderstandings and misinterpretations, which fed increasingly untenable suppositions and suspicions. Almrei, in the meantime, was detained for more than seven years, much of it in facilities not designed for long-term detentions. The government spent literally millions of dollars and untold hours of effort to keep him there. This was money and time fruitlessly expended on trying to prove that Almrei was dangerous that was not spent on other matters which were probably much more useful to national security.
I turn now to the civil-liberties considerations that should undergird any discussion of preventive detention.
Effectiveness is contextual. If everyone were detained, there would be no threat. And there would also be no society worth protecting. In other words, there is a civil-liberties baseline beyond which any measure of “effectiveness” is entirely ephemeral. That baseline in democratic societies is necessarily a demanding one. As the Israeli Supreme Court has noted, dealing with security problems is particular difficult for democracies. This, it concluded,
is the destiny of democracy, as not all means are acceptable to it, and not all practices employed by its enemies are open before it. Although a democracy must often fight with one hand tied behind its back, it nonetheless has the upper hand. Preserving the rule of law and recognition of an individual’s liberty constitutes an important component in its understanding of security. At the end of the day, they strengthen its spirit and its strength and allow it to overcome its difficulties.66
This study assumes that a democratic society adheres to the rule of law; that is, it conducts itself in keeping with standards of conduct prescribed by domestic and international law. This assumption exaggerates, of course, what it means to be “democratic.” Not all democratic states adhere to the law in every instance. Moreover, David Dyzenhaus (2006) warns forcefully that in apparent emergencies, democratic states often create “grey holes” — that is, zones in which the law supposedly applies but in such a narrow procedural or substantive manner as to render legal limitations on executive power meaningless. He cautions that in these circumstances, the facade of the rule of law is maintained but not always close observance of fundamental moral values he believes undergird this rule of law.
Nevertheless, it is also true that democratic countries are self-reflective and self-correcting, at least more so than authoritarian countries. In democracies, a repudiation of fundamental values and a hollowing out of the rule of law are, with the passage of time, often regarded as a failure, and not simply an incidental occurrence. Put another way, adherence to the rule of law is a baseline practice in democratic societies, even if it is occasionally honoured only in the breach. Indeed, the Supreme Court of Canada characterizes the rule of law and the related concept of “constitutionalism” as unwritten Canadian constitutional principle.67
The rule of law necessitates, of course, the existence of positive law; there must be law. It follows that one civil-rights essential in any system of preventive detention is legalism: that is, the system must be prescribed by law and not amount to ad hoc or extra-legal measures. More substantively, observance of the rule of law obliges adherence to, at minimum, specific legal rules relating to detentions. Those rules obviously consist of the statutory standards established to govern preventive detention. The standards in turn must be informed by constitutional norms and the international legal rules to which Canada is bound as a matter of international law and which (often and increasingly) influence the interpretation of domestic law.
The section that follows provides an overview of the key international law rules on detention, examined from the perspective of international human-rights law. I do not deal with the separate and distinct standards that exist under the laws of war (and, specifically, international humanitarian law), because this article does not concern the detention of combatants in a situation of armed conflict.68 I then examine other, constitutional-law standards.
A right to liberty and the obligation not to interfere with freedom through arbitrary detention are basic principles of international human rights law69 and of the constitutional traditions of liberal democracies.70 However, international law does not preclude preventive detention, instead imposing on its use limits and conditions that are intended (mostly) to forestall arbitrary detention.
Where detention is provoked by criminal charges, the state must subsequently provide the accused with a fair trial before an independent and impartial tribunal.71 The fair-trial obligation includes important procedural guarantees that give the accused the ability to mount an effective defence and confront the witnesses against him or her.72
On the other hand, when a person is detained in noncriminal proceedings, international law contains fewer emphatic procedural guarantees. International law simply provides that all detentions must be authorized by law and followed by judicial proceedings assessing the legitimacy of the detention.73 Thus, the International Covenant on Civil and Political Rights (ICCPR) states in article 9 that “no one shall be deprived of his liberty except on such grounds and in accordance with such procedure as are established by law.” Furthermore, “anyone who is arrested shall be informed, at the time of arrest, of the reasons for his arrest and shall be promptly informed of any charges against him.” Article 9 invokes a right to habeas corpus: “[A]nyone who is deprived of his liberty by arrest or detention shall be entitled to take proceedings before a court, in order that that court may decide without delay on the lawfulness of his detention and order his release if the detention is not lawful.” Certain procedural guarantees are developed by article 9, as interpreted by the UN Human Rights Committee. First, there must be “review of the substantive justification of detention.”74
Second, this review must “include the possibility of ordering release” where the detention is arbitrary or otherwise violates the ICCPR and must not be limited to a review of “mere formal compliance of the detention with domestic law governing the detention.”75
Third, the review must be by a “court,” even in cases involving military detentions.76 Article 14 of the ICCPR guarantees that any tribunal determining a criminal charge or any “rights and obligations in a suit at law” must be “competent, independent and impartial.” As a determination of a habeas corpus right before a court is obviously a suit at law determining a right, the court must, therefore, be “competent, independent and impartial.”77
Fourth, those court proceedings are presumed to be open, although it is notable that article 14 does anticipate that the court may be closed to the “press and public” in the interest, among other things, of national security in a “democratic society.”
The standards applied during detention are also the subject of international law. Article 10 of the ICCPR provides that “all persons deprived of their liberty shall be treated with humanity and with respect for the inherent dignity of the human person.”78 Meanwhile, article 7 of the ICCPR and the Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment79 ban torture and cruel, inhuman and degrading (CID) treatment.
Persons who are unlawfully detained must have “an enforceable right to compensation” (ICCPR, article 9(5)).
It should be noted that all of these human rights (other than the ban on torture and CID treatment) are subject to derogation under the ICCPR “in time of public emergency which threatens the life of the nation and the existence of which is officially proclaimed” (ICCPR, article 4).
One extremely serious limitation on freedom (and in many cases the right to life) is the practice of “disappearances”: the secret detention of persons without providing information on their fate.
The International Convention for the Protection of All Persons from Enforced Disappearance was opened for signature in 2007 and will come into force once it receives enough ratifications. The treaty will prohibit enforced disappearances in all circumstances, “whether a state of war or a threat of war, internal political instability or any other public emergency” (article 1).80 An “enforced disappearance” is “the arrest, detention, abduction or any other form of deprivation of liberty committed by agents of the State or by persons or groups of persons acting with the authorization, support or acquiescence of the State, followed by a refusal to acknowledge the deprivation of liberty or by concealment of the fate or whereabouts of the disappeared person, which place such a person outside the protection of the law” (article 2).
The Rome Statute of the International Criminal Court, meanwhile, includes “enforced disappearance of persons” as one of the predicate acts of a crime against humanity (that is, the disappearance is part of a widespread or systematic attack against any civilian population). Likewise, the Torture Convention may also bar disappearances. In its 2006 report on United States compliance with the Torture Convention, the treaty body established by that instrument — the Committee against Torture — concluded that disappearances and the maintenance of secret detention centres constituted in themselves a violation of the treaty.81
Like international law, the Canadian Charter of Rights and Freedoms includes protections against arbitrary detentions. Section 7 guarantees everyone the right to liberty, and the right not be deprived of it without “fundamental justice.” Section 9 codifies the right not to be arbitrarily detained or imprisoned, and guards against detentions made at the sole discretion of law-enforcement agencies.82 Section 10 guarantees that upon arrest or detention, everyone has the right “(a) to be informed promptly of the reasons therefor; (b) to retain and instruct counsel without delay and to be informed of that right; and (c) to have the validity of the detention determined by way of habeas corpus and to be released if the detention is not lawful.” Section 12 precludes “any cruel and unusual treatment or punishment.”
Section 7 and procedural protections
Section 7 contains the most important procedural standards that might apply to a system of preventive detention. The closest equivalent to such a system employed to date — security certificates under the Immigration and Refugee Protection Act (IRPA), described above — has been closely scrutinized by the courts, including the Supreme Court. In Charkaoui I83 and Charkaoui II,84 the Supreme Court established a number of baseline constitutional expectations relating to the procedures that might properly be employed in such a system.
Summarized succinctly, the court’s conclusion was that government restrictions on disclosure of its case against the individual — justified on national security grounds — violated section 7. The constitutional violation was not justified by section 1 of the Charter because there were alternatives open to the government.85 In noting these alternatives, the Supreme Court voiced substantial support, in principle, for some form of system that employs special counsel, prompting the ultimate development of the “special advocate.” In the current security-certificate process, a special security-cleared lawyer has access to all the relevant information in the possession of the government and is empowered to represent the interest of the person subject to the certificate during the closed hearings.
Second, the government has a strict obligation to disclose all relevant material — whether it supports or prejudices its case — to the court. To meet this disclosure requirement, the government, including the security services, must retain and preserve information that might subsequently be at issue in a court proceeding.86
That information must also meet certain baseline “quality control” standards. For one thing, it cannot be the product of torture or cruel, inhuman or degrading treatment or punishment.87 Use of evidence obtained by torture is expressly prohibited under Canada’s criminal law. Section 269.1 of the Criminal Code implements Canada’s obligations under the Torture Convention. It makes torture a crime and also prohibits the use of torture evidence in Canada.
It is also likely that evidence extracted via abusive tactics would be rejected as unreliable and thus generally inadmissible in most proceedings. Indeed, in a case involving evidence admissible before an administrative body — the Parole Board — a majority of the Supreme Court of Canada observed that “information extracted by torture could not be considered reliable by the Board. It would be manifestly unfair for the Board to act on this kind of information. As a result, the Board would be under a duty to exclude such information, whether or not the information was relevant to the decision.”88
Most important, the Supreme Court has signalled clearly that torture evidence would violate constitutional standards in criminal trials,89 an opinion it would probably extend to any other administrative proceeding that triggered the application of section 7.
Section 12 and cruel and unusual treatment
The Supreme Court has also considered the question of cruel and unusual treatment, which is barred by section 12 of the Charter, in assessing security certificates. In Charkaoui I, the court commented on the practice that had been developed by lower courts of releasing individuals subject to security certificates on strict conditions as security concerns waned. This prospect of judicial release on strict conditions was viewed by the court as the feature of the security-certificate system that kept it from being declared cruel and unusual treatment in violation of the Charter.90 The court concluded that “extended periods of detention under the certificate provisions of the IRPA do not violate sections 7 and 12 of the Charter if accompanied by a process that provides regular opportunities for review of detention, taking into account all relevant factors” including reasons for the detention, length of the detention, reasons for the delay of deportation, anticipated future length of detention, and the availability of alternatives to detention (at para. 110 et seq).
Section 9 and arbitrary detention
Commenting on the section 9 prohibition on arbitrary detention, the court held in Charkaoui I that aspects of the security-certificate system, as it then existed, which denied court review to some detainees for a prolonged period of time, violated the guarantee against arbitrary detention, “a guarantee which encompasses the right to prompt review of detention under section 10(c) of the Charter” (at para. 91). In a more recent case, the court has held that even a cursory detention — broadly defined to include brief investigative stops — is arbitrary when done in “the absence of at least reasonable suspicion.”91
The matters at issue in Canadian courts to date do not exhaust the range of possible international and constitutional issues that might be involved in a system of preventive detention. An obvious additional issue is the overlap between preventive detention and conventional criminal law. As noted, preemptive detention should address circumstances where either danger is imminent but has not manifested itself in actual criminal conduct or the information supporting the fear of danger is not of a calibre to secure a conviction. The legitimacy of this system depends, however, on the nature and scope of the resulting detention. If preventive detention amounts to long-term or indefinite incarceration, it is difficult to imagine that a court would permit it without also imposing — probably through the vehicle of section 7 of the Charter — all of the constitutional expectations associated with criminal prosecutions, especially those found in section 11. The government should not be able to detain indefinitely or for long periods of time on standards of proof lower than the criminal-law threshold of beyond a reasonable doubt. If that were permitted, then painting a system as one of “preventive detention” rather than plain criminal law would allow the government — by terminological fiat — to do an end run around one of the most fundamental postulates of our constitutional order.
The issue of standard of proof would also be a thorny one to any system of preventive detention and is one that has not been resolved by the courts. In the system of immigration security certificates, detention is justified on a “reasonable grounds to believe” standard, a threshold that falls far short of the criminal “beyond a reasonable doubt” or even the civil “balance of probabilities” concept. In practice, the information supporting the initial reasonable grounds to believe has rarely sufficed to justify continued detention with the passage of time. The Federal Court has developed a firm practice of conditional release on the theory that the passage of time and the notoriety of the case reduce the security risk. To rebut this conclusion, the government would presumably have to marshal more information pointing to greater security concerns than it had initially anticipated. The effect of this pattern is a de facto ratcheting up of the standard required to maintain a person in detention.
An increasing burden of proof on government is also a reasonable and perhaps obligatory approach for a formal system of preventive detention. Since preventive detention should be aimed at disrupting imminent danger, and not at permitting long-term incarceration, the threshold for government action should initially be modest, relative to the standards for, for example, criminal conviction. However, if the detention continued, the burden on government to demonstrate the necessity of the continued detention would increase, thereby vitiating the prospect of the system morphing into “criminal law lite.”
There must necessarily also be a point at which the detention must expire: a system of preemptive last resort simply cannot be reconciled with indefinite detention. It would be incongruous (and almost certainly unconstitutional) for detention under a system of administrative preventive detention to endure for periods in excess of (and indeed, far less than) what would be permissible under the criminal law following a conviction. The model I discuss below results in modest detention periods — a maximum of 72 hours — that far fall far short of the duration I believe would reasonably attract a “criminal law lite” critique.
A system of preventive detention must be firmly prescribed by law, and not be the product of administrative policy. The detention must be subject to prompt review by a court possessed of all the features of judicial independence, and that review must empower the court to examine the substantive basis for the detention and order release if the detention is not warranted on the merits. These proceedings must be presumptively open, although they may be held in camera where the national security or public safety interest protected by confidentiality is significant enough. The government has strict obligations to retain and disclose information, including information prejudicial to its case. While information injurious to national security or public safety may be withheld from the detainee, it must be disclosed to the reviewing judge and the detainee’s interests must, at minimum, be defended by a special counsel who is also supplied with this information. Information that might plausibly be considered the product of torture or other forms of coercive interrogation is inadmissible.
The fact of the detention, the person’s fate and his or her whereabouts must not be denied or concealed by the state in a manner that precludes, limits or impairs access to judicial review. While detained, the detainee may not be mistreated or subjected to torture or cruel, inhuman or degrading treatment.
If the reviewing court upholds the detention in the first instance, there must be periodic reviews of that detention and there must always be the prospect that the detainee will be either released outright or released on conditions following these reviews.
At some point, the detention must end; it must not become a de facto system of longterm or indefinite detention. If “preventive” detention amounts to criminal-law-style incarceration in everything but name, the courts are likely to insist that the government’s evidence meet criminal-law standards of proof and all the other protections of the criminal law. In practice, therefore, the prospect of release increases with the duration of the detention.
Read together, the effectiveness and civil-liberties concerns set out above provide the framework for a system of preventive detention. I reiterate my key point from the effectiveness section: only a narrow gap exists to be filled by preventive detention. From a civil-liberties perspective, a focus on gap-filling also eases (although it does not eliminate) some of the civil-liberties preoccupations that would arise in great number and with great vigour were preventive detention to have a more ambitious and broader reach.
This final section of the study outlines with greater precision a model of preventive detention reflecting the conclusions reached in the discussions above. Its starting point is an assessment of anti-terrorism “preventive detention” as manifest in the “sunsetted” section 83.3 of the Criminal Code. For several years, the current government has sought to resuscitate this provision, in virtually identical form. Its failure to do so to date is due to interruptions in the parliamentary calendar stemming from dissolutions and prorogations, and not apparently to resistance from parliamentarians themselves.
Originally enacted by the 2001 Anti-terrorism Act, section 83.3 of the Criminal Code was entitled “recognizance with conditions” but labelled “preventive detention” in the popular discourse. This provision was clearly designed to foil terrorist plots on the cusp of execution. Cohen (2005, 218) describes its purpose this way:
The whole scheme is designed to disrupt nascent suspected terrorist activity by bringing a person before a judge who would then evaluate the situation and decide whether it would be useful…to impose conditions on the person. The purpose…is not to effectuate an arrest but merely to provide a means of bringing a person before a court for the purposes of judicial supervision.
With the consent of the federal attorney general, a peace officer was authorized to lay an information before a provincial court judge if that peace officer believed on reasonable grounds that a “terrorist activity” would be carried out and suspected on reasonable grounds that the imposition of recognizance with conditions or arrest on the person was needed to prevent that terrorist activity (Criminal Code, section 83.3). The judge could then require that the person named in the information appear in court.
Under urgent circumstances, where the grounds for laying an information existed (or that information had in fact been laid) and a peace officer suspected on reasonable grounds that a person must be detained to prevent a terrorist activity, a person could be arrested without warrant. Subsequently, an information was to be laid, and then the person was to be brought before a provincial judge without delay and within 24 hours unless a judge was unavailable. In the latter instance, the person was to be brought before a judge “as soon as possible.”
Whether arrested with or without warrant, when the person ultimately appeared before the judge, the latter was to order the person’s release unless the peace officer showed cause for the detention, including the likelihood that a terrorist activity would be carried out if the person was released. Alternatively, where the judge declined to order release, he or she could adjourn further proceedings for no longer than 48 hours pending a full hearing on the peace-bond issue.
The effect of these provisions was theoretically to enable a preventive detention on suspicions of terrorist activity for a maximum initial period of 24 hours (and perhaps longer if a judge was not available within that period) in warrantless arrest cases and then, where the judge agreed to an adjournment during the show-cause proceeding but did not release the detainee, detention for another 48 hours. Preventive arrest could last, in other words, for some 72 hours.
When a full hearing was held, the judge was to consider whether the peace officer had reasonable grounds for his or her suspicion. If he or she did, the judge could order the person to enter into a recognizance of up to 12 months’ duration, a limitation on the person’s freedom equivalent to the still existing peace-bond power discussed above.
By 2007, section 83.3 had never been used. Nevertheless, it remained among the most controversial of the provisions introduced by the 2001 Anti-terrorism Act. In particular, the prospect of detention prior to a full hearing before a judge for up to 72 hours on suspicions of terrorist activity generated disquiet.
To ease those concerns, Parliament inserted requirements that the attorney general and the minister of public safety report annually on the use of the section (section 83.31). The section also included an automatic sunset provision, terminating the section in early 2007 unless it was renewed by vote of Parliament (section 83.32). Both the Commons National Security Committee (House of Commons Canada 2006) and the Special Senate Committee on the Anti-terrorism Act (Senate of Canada 2007) recommended in 2006 and 2007 respectively that this provision be extended. However, a motion to renew the provision was defeated by the opposition parties in February 2007.
In his careful critique of the provision, David Paciocco (2002, 203) considered the section 83.3 process “intuitively offensive.” It is a way to “extend the reach of criminal consequences where the requirements of full proof cannot be met.” This objection rests, in some measure, on a rejection of the concept of preventive detention writ large. If one starts from the premise, however, that there is room for a system of preventive detention to fill gaps in existing Canadian law, I believe section 83.3 stands up quite well.
The provision is extremely modest in its reach and impact, as compared to its closest equivalents in the United Kingdom and Australia and the practice employed in the United States. The period of detention is short, judicial consideration occurs promptly and the provision is clearly fully preemptive; it contains none of the investigative-detention aspects of the systems of other jurisdictions. The system does not depend on an assessment of the perceived “dangerousness” of the individual unconnected to anything other than an imagined or at least hypothetical threat, as does the immigration security-certificate process that has produced detentions enduring for years. Instead it requires a connection to a concrete act — a belief “on reasonable grounds that a terrorist activity will be carried out.” In a different context, the Supreme Court has stated that the “belief on reasonable grounds” standard falls short of the civil balance-of-probabilities standard.92 Nevertheless, it requires that there be “an objective basis [for the belief]…which is based on compelling and credible information.”93
The section also requires suspicion “on reasonable grounds” that the arrest or subsequent peace bond “is necessary to prevent the carrying out of the terrorist activity.” As noted, suspicion on reasonable grounds is a less demanding standard than belief on reasonable grounds, but it still must be grounded in fact and not hunches.
Section 83.3 may be used, in other words, where a law-enforcement agency believes — objectively and on the basis of compelling and credible information — that a terrorist activity will (not could or might) be carried out, and suspects — still on the basis of objectively ascertainable facts and not mere hunches — that the detention of the individual is (not might be) necessary to forestall this terrorist activity. These views must generally be tested in front of a judge prior to any arrest. Even where an arrest is made on urgent grounds, judicial scrutiny occurs initially at the show-cause proceedings within 24 hours or as soon as possible thereafter in the unlikely event that a judge is unavailable. Fuller review occurs within 48 hours of that initial judicial examination, assuming that the person remains detained.
As best I know, in all the times section 83.3 has been before parliamentary committees, no government or law-enforcement witness has proposed that the possible 72 hours of “preventive detention” that might result from this system is inadequate to the principal task of the provision, that is, disrupting a terrorist plot.94 Nor have these witnesses, who are charged with advancing the promulgation of this law project, doubted the overall utility of the provision (even though that matter had not yet been decisively proved). Still, there is some reason to doubt whether the section 83.3 model offers much of real benefit: It may be too modest.
Seventy-two hours of potential detention pending adjudication of the peace bond may seem inadequate. Some critics may prefer a system of outright indefinite detention, that is, permanent or at least semipermanent detention. I have already argued that indefinite detention cannot be reconciled with preventive detention: the former would, in my view, inevitably prompt the courts to insist on the sorts of checks, balances and constraints that exist in the criminal process in an effort to avoid the imposition of penalties potentially in excess of what the criminal law would permit, done in the guise of administrative proceedings. I am also not persuaded that without a system with such checks and balances (and indeed, sometimes not even with such a system), the state can expected to single out correctly those who truly should be incarcerated for lengthy periods from those who should not be.
Other commentators will accept the argument against indefinite detention but query the modest duration of the section-83.3 detention. There is, however, no principled basis on which to pick one period of detention over another — the United Kingdom, for example, has drifted from 7 to 28 days of detention without charge over the last few years, and the government has repeatedly sought parliamentary blessing for even longer periods.
According to a report from the UK civil-rights group Liberty, the periods for equivalent forms of detention without charge are 2 days in South Africa, 2 in New Zealand, 2 in Germany, 3 in Denmark and Norway, 4 in Italy, 5 in Spain and Russia, 6 in France, 7 in Ireland and 7.5 in Turkey. Australia has opted for a period of 48 hours in its federal preventive-detention law, as supplemented by state law (Russell 2007). Detention periods are, in other words, variable, but very few have drifted toward the extremes witnessed in the United Kingdom.
It should be recalled that, in my view, the section 83.3 regime is justifiable only as a last-stop gap filler — a measure that allows the police to press “stop” in the most urgent circumstances when all other measures fail or are unavailable. Without a compelling argument that 72 hours is insufficient to allow the police to regain the initiative in foiling a terrorist attack — and no such argument has been advanced during the many times section 83.3 has been considered in Parliament — it remains a reasonable, although admittedly arbitrary, figure.
The more pressing issue is whether a system of preventive detention operating in the same due-process environment as regular criminal proceedings adds much of value. Unlike the security-certificate system or its closest UK and US equivalents, section 83.3 does not allow the use of secret evidence in support of the government position. All is conducted in open court, subject to Canada Evidence Act exceptions on disclosure of material related to national security that would be tested in separate proceedings in front of the Federal Court. The use of section 83.3 is, therefore, potentially confined to the circumstances described above in relation to the regular criminal law: that is, the information comes from a foreign state that consents to disclosure; or a security-sanitized summary of the evidence can be prepared that accommodates national-security confidentiality while not materially impairing the fairness of the proceeding; or the Canadian security services have enough incriminating evidence of their own that does not prejudice their own sources and techniques to maintain their case.
Obviously, less of this potentially sensitive information need be provided in a section-83.3 proceeding than in a regular criminal prosecution. The burden of proof on the government is much lower in the section-83.3 context. Put another way, there is less that has be proved and less sensitive information is required to do it.
Nevertheless, as the security-certificate cases amply show, even where government bears a reduced burden of proof outside of a truly penal case, disclosure requirements can be substantial and can extend far into areas of acute national-security sensitivity. Put bluntly, it is unlikely that a government would ever wish to resort to section 83.3 if doing so would require even security-certificate-like levels of disclosure.
It will be recalled that preventive detention was regarded as useful in addressing only one of the scenarios spelled out above, specifically, scenario D: Information supplied by a reliable confidential informant and/or foreign intelligence agency pursuant to an intelligence-sharing understanding warns of an attack at an unspecified time but in the near future on an unknown Canadian landmark by a terrorist organization. The informant or the foreign intelligence service identifies a citizen as a known member of that terrorist organization. Depending on the cogency and reliability of this information, section 83.3 arguably applies: there is a reasonable belief of a terrorist activity and, given the group’s known membership, a reasonable belief that links the citizen to the activity and that his or her arrest will forestall that activity.
Still, the provenance of the key information leading to these beliefs and suspicions from confidential sources will trigger government concerns about national security and confidentiality. The government may, therefore, be reluctant to resort to a mechanism — a section-83.3 proceeding — requiring the public admission of evidence that supports the government’s belief and suspicion. In such circumstances, section 83.3 becomes a largely theoretical tool for government. If there truly is a gap in Canadian law — a circumstance in which the state should be able to resort to limited preventive detention — it is not desirable to reduce section 83.3 to a theoretical tool.
The question then is whether this national-security-confidentiality constraint on the use of section 83.3 should be addressed through measures allowing use of secret evidence. In the United Kingdom, the government can use secret evidence in “control order” cases and special advocates represent the interests of the “controlee” in the ex parte, secret sessions. Should Canada follow this process and make the proceeding unfair by denying the accused full disclosure? The presence of special advocates would reduce that unfairness but in no way eliminate it.
The compromise solution is to divide the section-83.3 process into two phases: that associated with the initial detention and show-cause proceeding in front of a judge and that associated with the adjudication of the peace bond. During the initial detention and show-cause process, the state should be able to rely on secret evidence, in which case that evidence should also be disclosed to special advocates who are able to advance the interests of the interested party during the ex parte and in camera proceeding when this secret information is used. The whole purpose of preventive detention is to make expeditious state action possible when other legal instruments provide no tool for disrupting a feared terrorist activity. In dire circumstances, the state should not be obliged to choose between preserving its intelligence methods and relationships (and the future harm they may forestall) and averting an imminent and serious terrorist activity.
At issue during this initial phase is the prospect of (at most) 72 hours of detention. This is a significant limitation on freedom. It should be noted, however, that Canadian law already allows secret evidence to be used ex parte and in camera, and without the presence of a special advocate, when warrants are sought to invade privacy (in often significant degrees) for periods of a full year. It is hazardous to compare the relative impact of a lost year of privacy with 72 hours of detention. Still, there is a reasonable case that the invasion of privacy actually constitutes the greater violation of civil liberties. A warrant under the Canadian Security Intelligence Service Act (CSIS Act) will almost never come to the attention of the target, and the person subjected to a persistent, judicially authorized violation of their privacy may be eternally oblivious of this fact. A detained person will obviously be better informed and will have recourse to remedies for abuse of process never available to the person subject to surveillance, a matter to which I return below.
The state’s capacity to use secret evidence should end, however, when the question of the peace bond is decided. That peace bond may be valid for a year, may be renewed and may include substantial constraints on the individual’s freedom and privacy. Moreover, the peace bond is not about — or is at best only loosely about — preemption of imminent harm. If the state wishes to pursue a peace bond, it should be prepared to make its case in open court, relying only on the (substantial) protections already offered by the Canada Evidence Act regarding the disclosure of sensitive national-security information.
This proposal is a compromise; it seeks to preserve the core utility of section 83.3’s disruptive and preemptive potential, but it does not abandon fairness in a subsequent phase of the proceeding that raises less pressing security concerns.
It is entirely possible in a bifurcated system such as that proposed above that section 83.3 will be used to detain and that, then, in order to preserve its secret evidence, the government will abandon its efforts to impose a peace bond. This decision would not, in my view, itself constitute an abuse of process since the secret evidence would still be subject to vetting at the showcause proceeding in the presence of a special advocate. Again, the evidence is likely to be more closely scrutinized in this sort of hearing than would be the case in an entirely ex parte warrant application.
If the evidence is found wanting in the show-cause proceeding, a finding of abuse of process may be warranted and further detention or proceedings under section 83.3 terminated. If the person has already been detained on urgent grounds on the strength of the problematic evidence, the legislation should provide for compensation for the wronged individual. Preventive detention, in my model, permits a person’s freedom to be impinged upon on the basis of secret information that does not satisfy conventional standards of reasonable and probable grounds for arrest. In other words, it is very permissive in regard to state power. The prospect of both overclaiming by the state and false positives is strong, and the very secrecy of the information supporting state action makes it very difficult for wronged individuals to pursue conventional remedies, such as in the civil courts, for the state’s overreaction. Accordingly, erroneous use of preventive detention should be counterbalanced by unusually onerous obligations to remedy errors, built right into the preventive-detention law. The payment of damages may never fully compensate for loss of freedom, but it is an acknowledgement of wrongdoing and, if the payment is large enough, is another prudential deterrent to unwarranted state action.
While confining the use of secret information to the initial show-cause phase of the section 83.3 proceeding may alleviate some civil-liberties concerns, it certainly does not eliminate them. In my view, the use of secret information is warranted only by the seriousness and imminence of the feared harm. However, the present wording of section 83.3 allows preventive detention in more than such serious instances — the provision is tied to the concept of “terrorist activity” in the Criminal Code. That concept includes acts of violence but also certain inchoate offences such as “conspiracy, attempt or threat to commit any such act, or being an accessory after the fact or counselling in relation to any such act.” These inchoate offences may be far removed from actual acts of violence.
To minimize the prospect that preventive detention will be used in a manner inconsistent with the criteria I have proposed, I believe section 83.3 should specifically provide that it applies only where the peace officer believes on reasonable grounds that a terrorist activity will be carried out imminently and will involve a serious threat to life, health, public safety or property.
The model of preventive detention that I have proposed here tries to accommodate and balance concerns about both effectiveness and civil liberties. Certainly, any form of administrative detention sits uncomfortably with the system of justice applied in Canada and similar countries. Not all forms of administrative detention are, however, equal. It is one thing to develop a system of indefinite executive detention tied to prognostication concerning a person’s dangerousness, with dangerousness measured by beliefs, associations and aptitudes. It is quite another to build a system allowing the state to detain — briefly — individuals believed tied to imminent if still uncertain terrorist threats, in an effort at disruption.
There will be voices suggesting that the system I endorse here does not go far enough, that the criminal law is crippled and that indefinite executive detention is the solution. It is for those who wish to advance this perspective to do more than simply assert it, for both principled and practical reasons. First, a legal tradition that stretches back to the Magna Carta in voicing suspicion of executive detention is not to be summarily dismissed. Those who would abandon that suspicion must demonstrate carefully and thoroughly that the gap in the state’s capacity to counter terrorism is bigger than this study has argued, and that more aggressive measures will be effective, necessary and responsive to the fundamental norms according to which the Canadian legal system must operate. To date, no such case has been made.
It is also my belief that more expansive forms of detention based on perceptions of dangerousness test the limits of a state’s competence. If past experiences with Canadian security certificates and the practices of other jurisdictions — most notably Guantanamo Bay — are any indicator, such a system exceeds that competence. It may preempt an act of terrorism but only bluntly, while incarcerating those with the wrong profile who are also innocent of posing any immediate (or even more than purely hypothetical) threat. Such a system risks becoming a form of “criminal law lite,” imprisoning for acts that fall short of crimes or in circumstances where crimes cannot be proved. Lacking the rigour of the criminal law, it will inevitably overreach and ultimately bring the state into disrepute in a manner that may undermine the state’s broader anti-terrorism activities. Moreover, once a detainee is caught in a system of criminal law lite, there is no evident end point — once dangerous, always dangerous — and detention becomes indefinite.
The more targeted approach urged by this study acknowledges the slender gap in Canadian law; that is, a situation in which the state (1) has reason to believe that a terrorist attack will occur; (2) has reason to believe that a particular group is behind the plot and that the suspect is a member of that group; but (3) has no information, other than this, connecting that particular individual to the plot. In these circumstances, conventional legal instruments allowing the state to disrupt that threat through detention of the individual may not be available. In that narrow space, there are arguments in favour of preventive detention.
The model I endorse gives the state a limited but still strong power of detention to address this gap. It amounts to a system of “catch and release” or perhaps “catch and release, subject to conditions on that release.” In essence, it is the system that the framers of the 2001 Anti-terrorism Act selected, and it remains the best approach.
Baker, P. 2009. “Obama to Use Current Law to Support Detentions.” New York Times, September 23. Accessed July 12, 2010. http://www.nytimes.com/2009/09/24/us/politics/24detain.html
Carlile of Berriew. 2009. “Fourth Report of the Independent Reviewer Pursuant to Section 14(3) of the Prevention of Terrorism Act 2005.” Presented to Parliament pursuant to section 14(6) of the Prevention of Terrorism Act 2005. Accessed June 30, 2010. http://www.official- documents.gov.uk/document/other/9780108508141/97801 08508141.pdf
Cassel, D. 2008. “International Human Rights Law and Security Detention.” Case Western Reserve Journal of International Law 40 (3): 383.
Chesney, R. 2005. “The Sleeper Scenario: Terrorism-Support Laws and the Demands of Prevention.” Harvard Journal of Legislation 42 (1): 31.
Cohen, S. 2005. Privacy, Crime and Terror. Markham, Ont.: Lexis Nexis Butterworths.
Cole, D. 2009. “Out of the Shadows: Preventive Detention, Suspected Terrorists, and War.” California Law Review 97: 693.
Daskal, J. 2008. “A New System of Preventive Detention? Let’s Take a Deep Breath.” Case Western Reserve Journal of International Law 40 (3): 561.
Dershowitz, A. 2006. Preemption. New York: W.W. Norton.
Dinh, V. 2003. “Life After 9/11: Issues Affecting the Courts and the Nation, Panel Discussion at U. of Kan. L. Sch.” University of Kansas Law Review 51: 219.
Dyzenhaus, D. 2006. The Constitution of Law: Legality in a Time of Emergency. Cambridge: Cambridge University Press.
Elias, S.B. 2009. “Rethinking ‘Preventive Detention’ from a Comparative Perspective: Three Frameworks for Detaining Terrorist Suspects.” Columbia Human Rights Law Review 41:102.
Elliot, W. 2009. “Closing the Loop on National Security through Law Enforcement.” John Tait Memorial Lecture, presented at the Canadian Association of Security and Intelligence Studies Conference, October 30. Accessed June 30, 2010. http://www.rcmp-grc.gc.ca/news-nouvelles/speeches-stat- discours-decl/20091109-secur-eng.htm
Forcese, C. 2009. “Canada’s National Security ‘Complex’: Assessing the Secrecy Rules.” IRPP Choices 15 (5).
Glaberson, W. 2009. “President’s Detention Plan Tests American Legal Tradition.” New York Times, May 22. Accessed July 12, 2010. http://www.nytimes.com/2009/05/23/us/politics/23de tain.html
Greer, S. 1993. “Preventive Detention and Public Security — Towards a General Model.” In Preventive Detention and Security Law: A Comparative Survey, edited by A. Harding and J. Hatchard. London: Martinus Nijhoff.
Harding, A., and J. Hatchard, eds. 1993. Preventive Detention and Security Law: A Comparative Survey. London: Martinus Nijhoff.
House of Commons Canada. 2006. “Review of the Anti-terrorism Act Investigative Hearings and Recognizance with Conditions Program: Interim Report of the Standing Committee on Public Safety and National Security.” Subcommittee on the Review of the Anti-terrorism Act, Garry Breitkreuz, Chair. October 2006. Accessed June 2010. http://www2.parl.gc.ca/HousePublications/Publication.aspx ?DocId=2422668&Language=E&Mode=1&Parl=39&Ses=1
Justice Canada. 2009. “Backgrounder: ATA — Recognizance with Conditions and Investigative Hearings Provisions.” Ottawa: Justice Canada. Accessed June 30, 2010. http://www.justice. gc.ca/eng/news-nouv/nr-cp/2009/doc_32344.html
Levenson, L.L. 2002. “Detention, Material Witnesses and the War on Terrorism.” Loyola of Los Angeles Law Review 35.
McDonell, M. 2008. “Evidence.” In Proceedings of the Special Senate Committee on the Anti-terrorism Act, 2nd Sess., 39th Parl., Monday, January 28, issue no. 2.
McLoughlin, J., G.P. Noone, and D.C. Noone. 2008. “Security Detention, Terrorism and the Preventive Imperative.” Case Western Reserve Journal of International Law 40 (3): 463.
Mosley, R. 2002. “Preventing Terrorism — Bill C-36: the Anti-ter- rorism Act 2001.” In Terrorism, Law and Democracy, edited by D. Daubney, W. Deisman, D. Jutras, E. Mendes and P. Molinari. Montreal: TheÌmis.
Nicholson, R. 2007. “Evidence.” In Proceedings of the Special Senate Committee on the Anti-terrorism Act, 2nd Sess., 39th Parl., December 3, issue no. 1.
Obama, B. 2009. “Text: Obama’s Speech on National Security.” New York Times, May 21. Accessed June 30, 2010. http://www.nytimes.com/2009/05/21/us/politics/21obama. text.html?_r=1&pagewanted=all
Paciocco, D. 2002. “Constitutional Casualties of September 11: Limiting the Legacy of the Anti-terrorism Act.” Supreme Court Law Review (2d) 203.
Pearlstein, D. 2008. “We’re All Experts Now: A Security Case against Security Detention.” Case Western Reserve Journal of International Law 40 (3).
Public Safety Canada. 2009. “Corrections and Conditional Release Statistical Overview.” Ottawa: Public Safety Canada. Accessed June 30, 2010. http://ww4.sp-ps.gc.ca/res/cor/rep/ 2009-ccrso-eng.aspx#e3
Roach, K. 2002. “The New Terrorism Offences and Criminal Law.” In The Security of Freedom, edited by R. Daniels, P. Macklem and K. Roach. Toronto: University of Toronto Press.
Russell, J. 2007. “Terrorism Pre-charge Detention Comparative Law Study.” National Council for National Liberties, United Kingdom. Accessed July 12, 2010. http://www.liberty- human-rights.org.uk/issues/pdfs/pre-charge-detention-com parative-law-study.pdf .
Savage, C. 2010. “Obama Team Is Divided on Anti-terror Tactics.” New York Times, March 28. Accessed July 12, 2010. http://www.nytimes.com/2010/03/29/us/politics/29force.html
Savage, D. 2009. “Solicitor General Nominee Says ‘Enemy Combatants’ Can Be Held without Trial.” Los Angeles Times, February 11. Accessed July 12, 2010. http://articles.latimes. com/2009/feb/11/nation/na-solicitor-general11
Senate of Canada. 2007. “Fundamental Justice in Extraordinary Times: Main Report of the Special Senate Committee on the Anti-terrorism Act.” Special Senate Committee on the Anti- terrorism Act, David P. Smith, Chair. Accessed July 12, 2010. http://www.parl.gc.ca/39/1/parlbus/commbus/senate/com- e/anti-e/rep-e/rep02feb07-e.htm
United Kingdom. 2009. “The Government Reply to the Report by Lord Carlile of Berriew Q.C. Report on the Operation in 2008 of the Terrorism Act 2000 and of Part 1 of the Terrorism Act 2006.” Presented to Parliament by the Secretary of State for the Home Department by Command of Her Majesty June 2009. London, UK: Home Office. Accessed June 30, 2010. http://www.official-documents. gov.uk/document/cm76/7651/7651.pdf
United Nations. 1989. Subcommission on the Prevention of Discrimination and Protection of Minorities. Report on the Practice of Administrative Detention.” UN Doc. E/CN.4/Sub.2/ 1989/26, July 6.
United States. 2002. Department of Defense. “DoD News Briefing — Secretary Rumsfeld and Gen. Myers.” February 12. Accessed June 30, 2010. http://www.defense.gov/tran scripts/transcript.aspx?transcriptid=2367.
United States. 2003. Department of Justice. The September 11 Detainees: A Review of the Treatment of Aliens Held on Immigration Charges in Connection with the Investigation of The September 11 Attacks. Washington: Office of the Inspector General.
Wittes, B., and C. Peppard. 2009. Designing Detention: A Model Law for Terrorist Incapacitation. Washington, DC: Brookings Institution.
This publication was published as part of the Security and Democracy research program under the direction of Mel Cappe. The manuscript was copy-edited by Freya Goddard, proofreading was by Barbara Czarnecki, editorial coordination was by Francesca Worrall, production and art direction was by Schumacher Design and printing was by AGL Graphiques.
Craig Forcese is an associate professor in the Faculty of Law at the University of Ottawa, where he teaches pub- lic international law, national security law, administrative law, and public law and legislation,
To cite this document:
Forcese, Craig. 2010. Catch and Release: A Role for Preemptive Detention without Charge in Canadian Anti-terror- ism Law. IRPP Study, No. 7. Montreal: Institute for Research on Public Policy.
Copyright belongs to the IRPP. To order or request permission to reprint, contact:
1470 Peel Street, Suite 200
Montreal, Quebec H3A 1T1