The Prospects for Candour in Canada: Barriers to Disclosure

In previous posts (here and here) I have introduced the concept of the duty of candour in judicial review proceedings and described the centrality of the ‘record’ to judicial review of administrative action. In this post, I will consider several barriers to disclosure of relevant information which further limit the content of the record and thereby prevent or hinder the development of candour in Canada: deliberative secrecy, public interest immunity, and certificates issued under the Canada Evidence Act.

Deliberative Secrecy

Deliberative secrecy shields “[t]he how and the why of decision-making … to protect the decision maker and the decision-making process”: Cherubini Metal Works Ltd. v. Nova Scotia (Attorney General), 2007 NSCA 37, at para. 18, at para. 20. Deliberative secrecy is the rule for administrative tribunals: Tremblay v. Quebec (Commission des affaires sociales), [1992] 1 SCR 952, at pp. 964-965; Agnew v. Ontario Assn. of Architects (1987), 64 OR (2d) 8 (Div Ct); Noble China Inc. v. Lei (1998), 42 OR (3d) 69 (SC). This “core value” of Canadian administrative law (Milner Power Inc. v. Alberta (Energy and Utilities Board), 2007 ABCA 265, at para. 59) prevents courts from examining the internal records of administrative decision-making bodies, or compelling decision-makers to testify about their reasoning processes. Only where there is evidence of a breach of natural justice can a reviewing court pierce the veil of deliberative secrecy (Ellis-Don Ltd. v. Ontario (Labour Relations Board), 2001 SCC 4, [2001] 1 SCR 221, at paras. 51-55).

The Supreme Court of Canada has created some doubt about the application of the deliberative secrecy rule by referring specifically to the deliberative secrecy of “adjudicative” functions in Commission scolaire de Laval v. Syndicat de l’enseignement de la région de Laval, 2016 SCC 8, [2016] 1 SCR 29, at para. 60. This was an unusual case, however, as deliberative secrecy was held not to apply because the Commission scolaire was acting as an employer. The better view is that deliberative secrecy applies generally, wherever it is necessary in order to permit decision-makers to discharge their functions in view of the purposes of the deliberative secrecy rule. These are to promote finality in decision-making, encourage frank and open debate among those appointed to decision-making bodies, prevent decision-makers from having to spend more time testifying about their work than actually doing it and to avoid a chilling effect on decision-makers: Milner Power Inc. v. Alberta (Energy and Utilities Board), 2007 ABCA 265, at para. 59; Taylor v. WSIB, 2017 ONSC 1223, at paras. 59-62; West Moberly First Nations v. British Columbia, 2018 BCSC 1835, at para. 208; Raymond v. Halifax (Regional Municipality), 2018 NSSC 149, at para. 32.

These considerations indeed apply with their greatest force to adjudicative functions, though not necessarily exclusively to adjudication. The post-Commission scolaire de Laval decision in West Moberly First Nations v. British Columbia, 2018 BCSC 1835 is particularly significant: the court found that the information-gathering aspect of an environmental review panel process was subject to deliberative secrecy. Similarly, in Taseko Mines Limited v. Canada (Environment), 2015 FCA 254, a non-adjudicative environmental review panel process was held to be subject to deliberative secrecy and, in Stevens v. Canada (Attorney General), 2003 FC 1259, so too was a commission of inquiry. In each of these instances, deliberative secrecy would foster good decision-making much more than it would hinder external oversight.

Where the functions at issue are discretionary, however, the purposes of the deliberative secrecy rule are not engaged to the same extent, if at all. Hence the conclusion in Apotex Inc. v. Alberta (1996), 38 Alta LR (3d) 153 (QB) that no deliberative secrecy shielded a minister exercising a discretionary power. The fact that ministers benefit from additional privileges against disclosure (as described below) militates against giving deliberative secrecy too wide a scope but, in general, the concerns for good decision-making are not as forceful where a discretionary power is at issue, especially because disclosure of considerations thought to be relevant might facilitate the consistent exercise of the power in the future. And I would also suggest that the fact that ministers are politically accountable for their actions (whereas tribunal members are shielded from popular scrutiny) also tends to favour disclosure.

Public Interest Privilege

Originally known as Crown privilege, public interest privilege (or public interest immunity) “protects information from disclosure in litigation on the ground that the public interest in keeping the information secret outweighs the public interest in revealing the information for the purpose of resolving the litigation” (Stewart, Halsbury’s Laws of Canada — Evidence (2022 reissue), HEV-191). In an earlier era, public interest immunity meant that “documents otherwise relevant and liable to production must not be produced if the public interest requires that they should be withheld”, to be determined by the Crown either by having regard “to the contents of the particular document” or “by the fact that the document belongs to a class which, on grounds of public interest, must as a class be withheld from production” (Duncan v. Cammell, Laird, [1942] AC 624, at p. 636).

This seemingly draconian position was, to be fair, premised on the assumption that the Crown would ordinarily disclose relevant information: “this was up to the executive, and since the Crown usually granted disclosure it was therefore thought that there was no serious problem” (SI Bushnell, “Crown Privilege” (1973) 51 Canadian Bar Review 551, at p. 560; see also Robinson v. State of South Australia (No. 2), [1931] A.C. 704 (P.C.), at p. 714.). With a state more active in regulating and managing economic activity and assuming welfare functions, however, reliance on the good graces of the Crown to ensure meaningful disclosure became inadequate. Both branches of public interest immunity have been pared back over the years.

The House of Lords, which was responsible for the expansive definition just quoted, engaged in a significant pruning exercise just a couple of decades later in Conway v. Rimmer, [1968] AC 910. The Supreme Court of Canada followed suit in Carey v. Ontario, [1986] 2 SCR 637, moving decisively away from the “excessive views” expressed in Duncan (at para. 35). An objection based on public immunity will generally be raised “by means of a certificate by the affidavit of a Minister or … of a senior public servant” (at para. 38). Ultimately, however, “it is for the court and not the Crown to determine the issue” (at para. 39), giving “due consideration” to the Minister’s view, but noting first that “its weight will vary with the nature of the public interest sought to be protected” and second that “it must be weighed against the need of producing it in the particular case” (at para. 38). Accordingly, there is no general protection for a class of documents — even Cabinet-level documents — with a contextual analysis required in all cases:

Cabinet documents like other evidence must be disclosed unless such disclosure would interfere with the public interest. The fact that such documents concern the decision‑making process at the highest level of government cannot, however, be ignored. Courts must proceed with caution in having them produced. But the level of the decision‑making process concerned is only one of many variables to be taken into account. The nature of the policy concerned and the particular contents of the documents are, I would have thought, even more important. So far as the protection of the decision‑making process is concerned, too, the time when a document or information is to be revealed is an extremely important factor. Revelations of Cabinet discussion and planning at the developmental stage or other circumstances when there is keen public interest in the subject matter might seriously inhibit the proper functioning of Cabinet government, but this can scarcely be the case when low level policy that has long become of little public interest is involved. To these considerations, and they are not all, one must, of course, add the importance of producing the documents in the interests of the administration of justice. On the latter question, such issues as the importance of the case and the need or desirability of producing the documents to ensure that it can be adequately and fairly presented are factors to be placed in the balance. In doing this, it is well to remember that only the particular facts relating to the case are revealed. This is not a serious departure from the general regime of secrecy that surrounds high level government decisions (at paras. 79-80; see also British Columbia (Attorney General) v. Provincial Court Judges’ Association of British Columbia, 2020 SCC 20, at para. 101).

The issue of class privilege arose more recently in Vancouver Airport Authority v. Commissioner of Competition, 2018 FCA 24, [2018] 3 FCR 633. Here, the Commissioner unsuccessfully invoked public interest privilege over a class of 1,200 documents to protect the identities of third party sources who had handed over sensitive information. As Stratas JA observed,”it is perhaps not far from the truth to say that it is now practically impossible for a court, acting on its own, to recognize a new class privilege” (at para. 62), as a class privilege “can be blunt, sweeping and indiscriminate in operation and, thus, can work against the truth-seeking purpose of a court or administrative proceeding” (at para. 49). He accordingly refused to recognize a new class of public interest privilege in this case.

Although recent jurisprudence favours a contextual test and is leery of class privilege, it is still entirely possible for public interest privilege to be successfully asserted. In Canada (Attorney General) v. Slansky, 2013 FCA 199, [2015] 1 FCR 81, for example (noted here), the majority of the Federal Court of Appeal recognized public interest privilege in a report prepared for the Canadian Judicial Council to advise it in relation to allegations made against a judge, because information in the report was generated on the basis of undertakings of confidentiality; and in British Columbia (Attorney General) v. Provincial Court Judges’ Association of British Columbia, 2020 SCC 20 (noted here) a unanimous Supreme Court held that a Cabinet submission (prepared by the Attorney General) on the government’s response to the recommendations of an independent commission did not need to be disclosed in judicial review proceedings, because the Association had not provided a sufficiently strong case in favour of disclosure. Where public interest privilege applies, the record for judicial review will therefore be narrower than otherwise it might have been.

The Canada Evidence Act

Sections 37 and 38 of the Canada Evidence Act, RSC 1985, c C-5 codify the common law of public interest immunity: section 37 is a general provision, whilst section 38 puts specific arrangements in place for national security matters.

Section 39, however, goes much further than the common law, setting an absolute bar on the disclosure of federal cabinet-level material where a minister or the Clerk of the Privy Council certifies that the material is covered by cabinet confidentiality. Whereas at common law (and under sections 37 and 38), the judge conducts a balancing act to determine whether disclosure would be in the public interest, where section 39 is invoked, “the Clerk or minister balances the competing interests. If the Clerk or minister validly certifies information as confidential, a judge or tribunal must refuse any application for disclosure, without examining the information” (Babcock v. Canada (Attorney General), 2002 SCC 57, [2002] 3 SCR 3, at para. 17). It is “the most drastic privilege on the books” (Canada (Citizenship and Immigration) v. Canadian Council for Refugees, 2021 FCA 72, at para. 114), as “Courts are simply barred by section 39 from reviewing the documents and thus the ambit of Cabinet secrecy” (Singh v. Canada (Attorney General) (C.A.), [2000] 3 FC 185, at para. 41).

Cabinet-level material is broadly defined in s. 39(2), subject only to limited exceptions set out in s. 39(4). The power to issue a section 39 certificate must also be exercised for proper purposes, that is, to safeguard the public interest, and not, for example, to thwart public inquiry or gain a tactical advantage in litigation; “If it can be shown from the evidence or the circumstances that the power of certification was exercised for purposes outside those contemplated by s. 39, the certification may be set aside as an unauthorized exercise of executive power” (at para. 25). However, the circumstances in which this may happen are likely to be highly unusual, as the certificate itself is the primary basis for determining the lawfulness of the exercise of the power (at para. 40):

As noted, the Clerk must determine two things: (1) that the information is a Cabinet confidence within s. 39; and (2) that it is desirable that confidentiality be retained taking into account the competing interests in disclosure and retaining confidentiality. What formal certification requirements flow from this?  The second, discretionary element may be taken as satisfied by the act of certification.  However, the first element of the Clerk’s decision requires that her certificate bring the information within the ambit of the Act.  This means that the Clerk or minister must provide a description of the information sufficient to establish on its face that the information is a Cabinet confidence and that it falls within the categories of s. 39(2) or an analogous category; the possibility of analogous categories flows from the general language of the introductory portion of s. 39(2).  This follows from the principle that the Clerk or minister must exercise her statutory power properly in accordance with the statute.  The kind of description required for claims of solicitor-client privilege under the civil rules of court will generally suffice. The date, title, author and recipient of the document containing the information should normally be disclosed.  If confidentiality concerns prevent disclosure of any of these preliminary indicia of identification, then the onus falls on the government to establish this, should a challenge ensue.  On the other hand, if the documents containing the information are properly identified, a person seeking production and the court must accept the Clerk’s determination.  The only argument that can be made is that, on the description, they do not fall within s. 39, or that the Clerk has otherwise exceeded the powers conferred upon her (at para. 28).

In Babcock, the Supreme Court of Canada upheld section 39 against a constitutional challenge. The cogency of this decision has been strongly questioned by Professor Campagnolo in Behind Closed Doors: The Law and Politics of Cabinet Secrecy (UBC Press, 2021). In particular, the Supreme Court did not analyze whether section 39 is inconsistent with section 96 of the Constitution Act, 1867 on the basis that it prevents meaningful scrutiny of the lawfulness of government action, part of the core powers of Canada’s superior courts. In particular, certification by the Clerk — who is a participant in cabinet-level discussions and therefore someone with an interest in the question of disclosure — is effectively immune from judicial oversight. As the Supreme Court remarked with a large dose of understatement in Babcock, the “limitations” on what the Clerk is required to make available for scrutiny “may have the practical effect of making it difficult to set aside a s. 39 certification” (at para. 40).

Nonetheless, section 39 remains on the books and given the “limited scope of review” it contemplates, successful applications are few and far between (Canada (Environment) v. Canada (Information Commissioner), 2003 FCA 68, at para. 21, a rare case of a successful application, affirming Canada (Information Commissioner) v. Canada (Minister of the Environment) (T.D.), 2001 FCT 277, [2001] 3 FC 514 but varying the order).

In subsequent posts I will critically analyze the scope of the record in judicial review proceedings and suggest approaches courts might take (and, indeed, have taken) to alleviate the resultant difficulties.

This content has been updated on March 3, 2023 at 17:31.