The Tory government tabled a bill on Monday, October 22 proposing amendments to Canada's immigration "security certificates" (Bill C-3). On top of making other minor adjustments (reflecting changes in practice undertaken anyway in the wake of court rulings), the new bill's core purpose is to create a system of "special advocates" to be deployed in in camera, ex parte portions of the hearings to represent the interest of the affected person. Unfortunately, the government effort seems to draw more inspiration from the UK experience than from Canada's own Security Intelligence Review Committee model. It fails to preclude a repeat of the problems experiences in the UK (leaving avoidance of such problems to the discretion of the judge on a case by case basis). Lorne Waldman and I have questioned the wisdom of this model in a op-ed published on October 25 in the National Post, and reproduced below. Other problems space and complexity prevented being discussed in that piece also exist with the bill.
First, the information disclosure regime in the immigration law will continue to be different from the information disclosure regime under the Canada Evidence Act: in the latter, the court is to balance the national security interest against the fair trial interest. In the immigration law, there is no such statutorily-mandated balancing. (And, incidentally, the bill doesn't create a special advocate system for Canada Evidence Act s.38 proceedings. At least one Federal Court judge has concluded that these persons are constitutionally obligatory in such proceedings).
Second, it would also have been helpful to underscore the inadmissibility of evidence where reasonable grounds exist to believe it was produced via torture or cruel, inhuman or degrading (CID) treatment. The bill does specify that unreliable evidence is excluded, but that reference does not go far enough. Torture evidence is already precluded by s.269.1 of the Criminal Code. CID treatment evidence might also usefully be statutorily proscribed, along with rules on who bears the onus of proof in establishing the conditions in which such evidence was produced. (There have been Federal Court decisions on the latter issue, in relation to torture, but a statutory rules seems sensible).
Third, immigration detentions (of a potentially indefinite length) and removals (to persecution and possibly torture, if the security threat is deemed significant enough) will continue to be done on government suspicions rather than any standard of proof otherwise recongized in the legal system. We acknowledge the need to protect national security. We also appreciate, however, that government suspicions are sometimes -- perhaps often -- misplaced. That is one of the lessons of the Maher Arar inquiry. Unquestionably, if Maher Arar had been a non-citizen, and thus subject to an immigration security certificate, he would have been detained and removed to Syria, without any outing of the deficiencies of the government position.
To this point should be added another observation, anticipating an objection raised in our op-ed: The best special advocate in the world would not have stood in the way of this certainty -- it took a full airing and probing of the information in the government's position. That included access to information the government would rather not have shared. In bill C-3, a special advocate will be hard pressed to persuade a judge to see information the government has chosen not to submit to a judge already. He or she will be reduced to arguing that he or she suspects that there might be further relevant material, but not having access to it, will have difficulty making this case. Notably, the Federal Court has, in the past, fixated on the fact that security certificates are administrative and not criminal proceedings, and have tailored its position on disclosure accordingly. I fear that these judges will rely on this (unpersuasive) distinction in responding to such requests. (Keep in mind that the outcome of security certificate processes may be more serious that anything the criminal code could impose -- indefinite detention without charge and removal to torture). The better model, as we suggest, is that of the Security Intelligence Review Committee, with its full access to security service files. Ironically, if C-3 goes through unamended, special advocates may find it useful to bring collateral challenges in an effort to probe more deeply security service files, by tabling complaints in front of SIRC and calling on that body to conduct proxy investigations. It would be nice if a more elegant system was developed from the very beginning.
For more information on the UK experience and SIRC, see our August 2007 report.
Our op-ed follows (albeit with the dramatic headline selected by the editors):
Canada doesn't need a Star Chamber
Craig Forcese And Lorne Waldman
Published: Thursday, October 25, 2007
On Monday, the federal government tabled its amendments to Canada's national security immigration law. There were two options in drafting this bill: Ottawa could have learned from the experience of other countries -- most notably the United Kingdom -- and created a system that avoided problems in those jurisdictions. Or it could table a minimalist proposal that risks repeating those problems here in Canada. The government chose option two.
In response to the Supreme Court's Charkaoui decision in February -- the court held that the manner in which secret evidence was used to detain and deport non-nationals using security certificates was unconstitutional -- the new legislation Ottawa proposes will layer special advocates onto a slightly modified security certificate process. These special advocates will be security-cleared lawyers representing the interests of parties excluded from national security-related hearings where the government presents information deemed secret.
The United Kingdom has employed a system of special advocates for almost a decade, one that has been controversial. Until recently, the U.K.'s special advocates lacked a support office creating anything approximating equality of arms between government and special advocate. Other problems persist. Most notably, U.K. special advocates may not continue to communicate (in any meaningful sense) with the affected person they serve after the advocate has reviewed the secret information. Further, special advocates have reported difficulties in obtaining full disclosure of all relevant (and sometimes exculpatory) information the security services have on this person. In July 2007, the U.K. Parliament Joint Committee on Human Rights issued a strongly worded report describing the U.K. special advocate system as "'Kafkaesque' or like the Star Chamber."
Canada's government is proposing legislation that seems bound to create similar problems here. First, the bill specifically authorizes the special advocate to review secret information provided by the government to the judge in the security certificate process. However, it includes no express procedures for the special advocate to reach beyond this information and seek and obtain government records not already disclosed to the court (other than to ask simply the judge to oblige this disclosure).
Second, while the bill does not close the door on continued contact between the special advocate and the interested party subject to the security certificate, nor does it affirmatively guard this right. Instead, this is a matter left to the discretion of the judge.
Third, the bill is silent of the question of resourcing, failing to create a well-equipped special advocate office able to offer the sort of support now (belatedly) enjoyed by U.K. special advocates.
There are better systems, including ones developed in Canada. One such system is the Security Intelligence Review Committee (SIRC) model, employed successfully in Canada for over 20 years. The SIRC model remedies, at least in part, many of the obvious deficiencies in the UK special advocate system that we seemed destined to repeat. It provides for:
Full Access to Information: Under SIRC procedures, SIRC counsel has access to the entire file in the possession of the Canadian Security Intelligence Service (CSIS). As a result, concerns that the security services might either intentionally or inadvertently fail to disclose relevant (and indeed exculpatory) information to counsel are alleviated.
Continued Contact with Affected Person: In addition, under the SIRC model, SIRC counsel can meet with the affected person even after SIRC counsel has reviewed the secret information. Although counsel must take great care not to reveal secret information, experience over many years at SIRC has established that it is possible to have such meetings without risk of inadvertent disclosure.
Canada's minimalist approach is unnecessary. It has proposed a system that could fall far short of the SIRC model, one that will be employed in security certificate cases where the consequences to interested persons (imposed on the basis of suspicions) may include removal to persecution or prolonged detention without trial. It had an alternative. It could simply have grafted the immigration special advocate function onto the existing SIRC system. Indeed, this is, in essence, the approach that existed for permanent residents prior to 2002.
Given SIRC's successful track record, it is very unlikely that a special advocate model less thorough than the SIRC model will survive constitutional scrutiny. The Charkaoui case turned on the court's conclusion that the government had better options in the design of its security certificate system. That conclusion is not fully reflected in the bill presented by the government on Monday. Parliament should now move promptly to add additional precision and detail to the government's sparse effort.
- Craig Forcese teaches national security law at the University of Ottawa. Lorne Waldman is an immigration lawyer in Toronto. They are co-authors of an August, 2007, study on special advocates.