Bill C-51’s Goldilocks problem, and what needs to change in Canada’s anti-terror response
On January 30, 2015, just three months after the October 2014 shooting on Parliament Hill and in the wake of the first Paris attack targeting Charlie Hebdo, Prime Minister Stephen Harper introduced Bill C-51. He did so at an election-style rally and not in Parliament. Canada’s anti-terrorism response was—from its inception—politicized, even radicalized, in a way that did not occur even after 9/11.
It is worth remembering that, although many Canadians would later grow disenchanted with its divisive focus on “barbaric cultural practices” and not re-elect the Harper government, the omnibus Bill C-51 was massively popular at first. The Official Opposition, the NDP, took a few weeks before deciding to oppose it. Justin Trudeau’s Liberals voted for it, later promising to repeal its “problematic” aspects.
Once in office, the Liberal government began the process of making good on that promise. This past September, the government issued a 21-page document accompanied by a 73-page background document on national security entitled Our Security, Our Rights: National Security Green Paper, 2016.1 It is designed to facilitate a public consultation to end this month. Legislation is expected sometime in 2017.
The green paper will not be a bestseller, but it makes fascinating reading. In some respects, it is the public defence of Bill C-51 that the Harper government did not provide. It likely reflects the sincere beliefs of many in the national security bureaucracy that they need more powers to perform their difficult jobs. But as the federal privacy commissioner has observed, the green paper “focuses heavily on challenges for law enforcement and national security agencies” as opposed to “democratic rights and privacy.”
Many Canadians may be suffering C-51 fatigue, although they should have been roused from any slumber by the revelations in early November that the Canadian Security Intelligence Service had not been candid to judges about its broad collection and indefinite storage of metadata or identifying information about all those who had contact with any of their targets as far back as 2006. In any event, it is alarming to realize that since Bill C-51 became law in June 2015, CSIS and other security agencies are already using their new powers. There is a danger that C-51 has already become the new normal, to the detriment of both our liberties and our security.
In fact, many of the powers it outlines are already in use. The 2015/16 annual report by CSIS’s review body—the Security Intelligence Review Committee—revealed that CSIS has engaged in about two dozen threat-reduction disruptions since June 2015. Unfortunately, the report provides no details about these activities other than that they were not conducted with a judicial warrant. This means that CSIS and its lawyers determined that the activities did not infringe on laws or Charter rights or cause bodily harm, legal verdicts with which the independent review body agreed.
The concern is that these judgements constitute simply the latest iteration of disturbing post-9/11 phenomena: the proliferation of secret law, in which vague statutes are given secret interpretations kept far from public discussion.
Meanwhile, the privacy commissioner recently reported that the extremely broad information-sharing provisions in C-51 have been used to send and to receive information about 100 times in the first six months of C-51’s operations. At least three agencies (including the Canadian Border Services Agency and Citizenship Canada) that are subject to no dedicated national security review are both sending and receiving information about individuals on the basis that the information falls within Bill C-51’s astonishingly overbroad definition of security threats.
But how good is the information that is being shared? One hopes these agencies have learned lessons from poor information-sharing practices—the sharing of faulty information with international agencies—that contributed to the torture of Maher Arar and other Canadians in Syria. But without effective review and audits we cannot know for sure.
There have been no prosecutions under the new C-51 speech offence of advocating “terrorism offences in general.” But this new C-51 crime was used twice in the last half of 2015 as the basis for a wiretap warrant. To be sure, a judge would have authorized such surveillance. But in a closed hearing with no adversarial challenge, there is no one but the government side represented. Perhaps the warrants have been granted to respond to real threats of terrorist violence, but the C-51 offence as written would allow them to be granted on the basis of speech that could stop well short of instructions or calls to commit actual violence.
The problems in C-51 are many: the law as it stands both overreacts and underreacts. The overreaction makes us less free and the underreaction makes us less safe. Here are some specifics:
1) C-51’s poorly drafted information-sharing law is overbroad in defining security threats (overreaction), but fails to oblige CSIS to share information about possible terrorism offences, even though information sharing between agencies is a blind spot identified by the Air India commission (underreaction).
2) CSIS’s new disruption powers, to disrupt or “reduce” all security threats is overbroad by allowing all laws and Charter rights to be breached so long as CSIS does not cause bodily harm, violate sexual integrity or obstruct justice, and the use of those powers is authorized by judges. This is a constitutionally radical and dubious idea (overreaction). At the same time, using these new powers could compromise successful criminal prosecutions of terrorists. CSIS illegalities and Charter violations during the course of investigating a terror plot might lead a court to dismiss charges.
3) Bill C-51’s speech crime applies to anyone who “by communicating statements, knowingly advocates or promotes the commission of terrorism offences in general” infringes on freedom of expression (overreaction). At the same time, no consideration was given to the effect of the speech crime on nascent terrorism prevention programs aimed at dissuading people from violence (underreaction). Those programs require people to voice their perhaps extremist beliefs, which will not happen if saying the wrong thing means risking prosecuting for advocating “terrorism offences in general” and perhaps going to jail for five years.
4) Peace bonds are easier to obtain under C-51, but can create a “Goldilocks” problem: too weak for determined terrorists (underreaction, as compared to actual prosecutions) and too tough when applied to those who have merely expressed extremist or radical views (overreaction).
5) Increased use of secret intelligence to suspend passports and place people on no-fly lists can create false positives (overreactions). At the same time, such sanctions by preventing international travel and terrorism could, as occurred in October 2014, turn would-be foreign terrorists into domestic terrorists (possible underreactions).
Aside from these specific concerns, there is an overarching problem with the green paper’s approach to national security.
The green paper examines ten distinct topics sequentially. But the effects of national security activities on both public safety and protection of rights are holistic and cumulative. The green paper examines accountability as a discrete topic, for instance. But accountability and review issues should permeate a discussion of all security powers.
At present, our expert review bodies are limited to a handful of agencies and are prevented from conducting joint reviews even as the agencies they review conduct joint operations. The result is a system of accountability gaps, a fact recognized most emphatically by the 2006 Arar commission and noted periodically by review bodies themselves since that time.
The government is moving ahead with a bill (Bill C-22) that will create a special committee of parliamentarians with partial access to secret information. But partial access is not enough to address the accountability challenge. And even with full access, there will always be the need for expert review bodies that can hear complaints and audit in detail operational activities.
A solution—and one supported by international trends—would be to create a “super SIRC,” well staffed and well resourced, that would allow one full-time expert review body to examine all security activities across government—intelligence, information sharing and more—without the need for complicated choreography between existing bodies, or the discretionary appointment of public inquiries like the Arar, Iacobucci and Air India inquiries (which all had whole-of-government mandates). Legislative reform would be required to create this super SIRC, but is nowhere to be found in C-22.
The Australian version of SIRC has a much broader intelligence-based remit than its siloed Canadian counterpart. In the United States, inspectors general for various security agencies can conduct the joint investigations that are necessary to overcome bureaucratic walls between different agencies. Both the United Kingdom and the United States have created new civil liberties boards with government-wide mandates. Canada has fallen behind.
A major area of concern in the status quo before Bill C-51 was the troubled relationship between secret intelligence and public evidence available for use in prosecutions. Consider the findings of the hearing into the 1985 Air India bombing, Canada’s largest mass murder, which unfortunately only resulted in two manslaughter convictions (of the same person). One of its major recommendations was that CSIS be required to share its secret intelligence about terrorism offences with someone able to make decisions in the public interest about whether that intelligence information should be disclosed for use of criminal proceedings.
But instead of following the United Kingdom’s lead—where CSIS’s counterpart now often collects terrorism intelligence to evidentiary standards to minimize “intelligence-to-evidence” issues—C-51 permits information sharing about everything connected with an astoundingly broad range of security threats, but does not require information sharing about terrorism offences. And C-51 left in place Canada’s unwieldy court apparatus for deciding whether secrets should be disclosed in court proceedings.
In Canada, only specially designated judges of the Federal Court in Ottawa can decide whether secret intelligence should be shielded from disclosure. The criminal trial judge sitting, say, in Toronto or Montreal must accept Federal Court orders that evidence not be disclosed because it will harm national security. But then the criminal trial judge must decide whether the criminal trial should be stopped because the accused needs the secret material to have a fair trial.
In a case that stemmed from the Toronto 18 prosecutions, the Supreme Court decided that this cumbersome approach was constitutional, while at the same time stressing that it was not pronouncing on the wisdom of the approach that encouraged a form of “constitutional chicken.”
One possible solution? Giving trial judges throughout the country powers to make and revise non-disclosure orders on national security grounds. That would give them more tools, and less blunt ones than terminating trials, to reconcile the competing interests between disclosure and secrecy.
There are other secrecy problems. Bill C-51 restricts the information that security-cleared “special advocates” get in security certificate immigration cases. Special advocates are lawyers who are charged with defending the affected person’s interest in the secret hearings—and their very success in doing so is a likely reason the government sought to roll back their access to information. Their role should be restored, and extended to proceedings challenging no-fly listing, passport revocation and terrorist group proscription.
That is only a partial answer to the unfairness of using secret evidence against persons. It is an imperfect proxy, but it would constitute a significant improvement over the status quo.
Bill C-51 has given CSIS new powers to disrupt and even to engage in illegal dirty tricks that violate Charter rights. The 1981 McDonald Commission that preceded the CSIS Act was crystal clear that “noble cause” illegality should not be a feature of Canada’s security intelligence. Parliament resisted a 1983 version of the CSIS bill because of widespread concern, including from all provincial attorneys general, that it would allow CSIS to violate the law as the RCMP had done, most infamously when it burned a barn in Quebec as a disruption tactic.
The Trudeau government’s green paper has a different take from the McDonald Commission—a more hardline one. It argues that the world has changed since 1984 and that when C-51 was enacted “it was felt that there were situations where CSIS was best placed to take timely action to reduce threats.” The paper also argues that allied agencies have threat disruption powers.
It is difficult to think of another country that has enacted legislation like Bill C-51 to empower its intelligence agency to violate every law and every constitutional right, so long as it does not cause bodily harm, invade sexual integrity or intentionally obstruct justice. The United States does not have similar legislation, but one can imagine the new Trump administration would not welcome Canadian legislation restricting CSIS’s powers.
In the United Kingdom, CSIS’s counterpart often works closely with the police when it engages in disruption. In Australia, the disruption powers of the intelligence agency are much more closely defined in legislation than in Canada—such as the power to change data on a computer system. And so even without a codified constitutional bill of rights, Australia’s approach is more transparent and limited than Canada’s C-51 “blank cheque.”
The concern is that the public consultations and legislation to come will not change this blank cheque approach. The green paper repeats the argument made by the previous government that Charter violations could be authorized by judicial warrants, making CSIS’s use of its powers accountable and constitutional. From a constitutional law perspective, this is a perplexing position.
To be sure, judges grant search warrants, but only because section eight of the Charter specifies clearly that Canadians are protected against “unreasonable” searches and seizures. There is no tradition in Canada of limiting other Charter rights by judicial warrants.
In our system, it is Parliament that must take responsibility for placing and justifying reasonable limits on rights. The judicial role is then to determine if legislative limits are justified, not to enable and authorize the violation of rights.
At a minimum, C-51 should be amended to remove any general reference that allows the Charter of Rights and Freedoms to be contravened by a threat reduction measure. If there is a need to limit a particular Charter right, for example the rights of citizens to leave or return to Canada, then that should be specified in the law. In general, the powers that CSIS claims it needs under this section should be more candidly discussed and better articulated in the legislation.
The problems with Bill C-51 are not only matters of legal and constitutional principles. Even if CSIS becomes expert on disrupting terrorist plots, such disruptions will not facilitate terrorist prosecutions. Indeed, they might even taint subsequent attempts to engage in terrorism prosecutions.
The revised protocol between the RCMP and CSIS, called One Vision 2.0, seems aimed at ensuring that CSIS’s new disruption efforts do not also disrupt ongoing criminal investigations. This is a good and necessary start, but we would urge CSIS in its counterterrorism investigations to go further and act in ways that support criminal investigations as its British counterpart, MI5, seems prepared to do. Otherwise, Canada will be drawn into a system of whack-a-mole disruption with no real endgame. Criminal prosecutions are both our fairest and most legitimate response to real terrorist threats.
It is worth remembering that our security agencies are already using the new powers conferred by Bill C-51, and getting used to them. The Trudeau government’s green paper does not suggest that those powers will be repealed and it was issued before Trump’s election.
Moreover, the green paper suggests that the state may not stop at C-51. In fact, it signals an interest in new powers especially with respect to digital investigations and access to so-called metadata produced by internet communication, even though the Supreme Court has recognized a privacy interest in analogous information. November’s Federal Court finding about undisclosed and broad CSIS retention of metadata since 2006 has understandably heightened concerns. Increased access to our digital trails will invade privacy. And that invasion can then be multiplied by C-51’s overbroad information-sharing laws while not being checked by adequate and full-time independent review that can follow the intelligence trail wherever it leads.
It may be that alternatives to the criminal law bolstered in C-51 such as peace bonds and no-fly listings appear to offer a chance to “scare straight” potential security threats. But in truth, all offer less viable and less legitimate solutions than successful criminal prosecutions when the real threat of terrorist violence presents itself.
In sum, C-51 was a cacophony of tactics whose full use would undermine civil liberties without necessarily making us safer. A long list of tactics will never compensate for a strategy. And it will not compensate for effective terrorism prosecutions. In addition, there is need for a better review of our counterterrorism efforts to guard against the constant dangers of overreacting or underreacting to the real threat of terrorism. Canada can do
Public Safety Canada’s Our Security, Our Rights: National Security Green Paper, 2016 is available at http://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/ntnl-scrt-grn-ppr-2016/index-en.aspx. The background document is available at https://www.publicsafety.gc.ca/cnt/rsrcs/pblctns/ntnl-scrt-grn-ppr-2016-bckgrndr/index-en.aspx. ↩