Judicial Review, Judicial Independence and Cabinet Confidentiality: British Columbia (Attorney General) v Provincial Court Judges’ Assocation of British Columbia, 2020 SCC 20
Last week’s Supreme Court of Canada decision in British Columbia (Attorney General) v. Provincial Court Judges’ Association of British Columbia, 2020 SCC 20 is most important for what it says about some relatively niche issues of Canadian public law but it also touches on broader matters, such as the relationship between procedure and substance in administrative law and disclosure of sensitive material in public law proceedings.
In Canada, judicial independence is constitutionally entrenched. One implication of the principle’s constitutonal status is that judicial salaries can only be altered after a review by an independent commission: Ref re Remuneration of Judges of the Prov. Court of P.E.I.; Ref re Independence and Impartiality of Judges of the Prov. Court of P.E.I.,  3 SCR 3. Executives and legislatures are not bound by the recommendations of these independent commissions but they should not lightly depart from the recommendations and any such departure is subject to judicial review: Provincial Court Judges’ Assn. of New Brunswick v. New Brunswick (Minister of Justice); Ontario Judges’ Assn. v. Ontario (Management Board); Bodner v. Alberta; Conférence des juges du Québec v. Quebec (Attorney General); Minc v. Quebec (Attorney General), 2005 SCC 44,  2 SCR 286 [Bodner].
Bodner-style judicial review is not equivalent to ordinary judicial review but is rather concerned with whether the government articulated a legitimate reason for departing from the commission’s recommendations, whether the reason has a reasonable factual foundation and whether, on the whole, the process was respected and achieved the goals of depoliticizing judicial remuneration and protecting judicial independence.
The question that arose in this litigation (and the companion case of Nova Scotia (Attorney General) v. Judges of the Provincial Court and Family Court of Nova Scotia, 2020 SCC 21) was whether a Cabinet submission (prepared by the Attorney General) on the government’s response to the recommendations of an independent commission can be disclosed in judicial review proceedings. Writing for a unanimous court, Karakatsanis J took the view that disclosure should be the exception, not the rule.
Administrative lawyers can derive two points of general interest from her analysis.
First, Karakatsanis J plainly appreciated the symbiotic relationship between procedure and substance in judicial review proceedings. In Bodner-style judicial review, the record before the judge will consist of ” any submissions made to the commission by the government, judges and others; the commission’s report, including its recommendations; and the government’s response to the recommendations” (at para. 53), “certain forms of additional evidence put in by the government” as long as these do not seek to bootstrap or supplement defective or deficient reasons (at para. 54) and evidence provided by the challengers “aimed at calling into question the reasonableness of the factual foundation relied on by the government, the government’s lack of meaningful engagement with or respect for the commission process or whether the government’s response was grounded in an improper or colourable purpose” (at para. 55). Thus, Karakatsanis J explained, the record is shaped by the issues to be addressed in Bodner-style proceedings:
Generally, what is in issue in a Bodner review is whether a government failed to meet its constitutional obligations flowing from the principle of judicial independence in its response to a commission’s recommendations. The relevance of any proposed additional evidence must therefore be tested in relation to the issues that the court must determine on Bodner review (at para. 59).
As I have suggested in a couple of recent articles (see here and here), if more material is put before a judge, the judge is liable to increase the breadth and depth of her review of the decision. A more expansive record is likely, all things considered, to lead to more expansive judicial review. If, as here, the scope of judicial review is limited, expanding the record is likely to increase the scope of review. Accordingly, Karakatsanis J’s caution was well placed.
She went on to advance another reason mitigating against disclosure in this case. Acknowleding that cabinet documents will generally be relevant to a Bodner-style review, she warned that routine disclosure of materials to a judge (who, after all, is reviewing a decision relating to judicial remuneration) would be problematic: “something more than relevance is needed to strike the appropriate balance between respecting Cabinet confidentiality and maintaining the overall integrity of Bodner review” (at para. 70). In a system of responsible government, Cabinet confidentiality is an important constitutional value. Routine disclosure would undermine that value:
Routine judicial inspection of a confidential Cabinet document would reveal to a member of the judiciary the content of Cabinet deliberations. Although any inspection of a confidential Cabinet document undermines Cabinet confidentiality to some extent, judicial inspection of a document that concerns Cabinet deliberations about the judiciary would undermine it more significantly. That is especially so where the judge is directly affected by the response resulting from those deliberations. As with adjudication of the Bodner review itself, judicial inspection is appropriate in this context only where it is strictly necessary (at para. 72).
Accordingly, Karakatsanis J held, a two-step test is required: a threshold showing “that there is some basis to believe that the Cabinet document in question may contain evidence which tends to show that the government failed to meet a requirement described in Bodner” (at para. 75); and then the judge must examine the document in private to determine if the document would, “taken with the record as a whole and in light of the applicant’s theory of the case, be of assistance in challenging the legitimacy of the government’s reasons, the reasonableness of the factual foundation it relied on, the respect the government has shown the commission process or whether the objectives of the process have been achieved” (at para. 80).
Even at this stage — and here is the second point of general interest — disclosure will not be automatic. Some other evidentiary rule may be invoked to prevent the document being placed in the record. Public interest immunity is one such rule. Karakatsanis J reiterated the factors to be weighed in considering the balance between cabinet confidentiality and disclosure of documents relating to public decision-making:
(1) the level of the “decision‑making process”;
(2) the “nature of the policy concerned”;
(3) the “particular contents of the documents”;
(4) the timing of disclosure;
(5) the “importance of producing the documents in the interests of the administration of justice”; and
(6) whether the party seeking the production of the documents “alleges unconscionable behaviour on the part of the government” (at para. 101, citing Carey v. Ontario,  2 SCR 637).
Where a Bodner-style review is being conducted, “at least three Carey factors — the level of decision‑making process to which the document relates, the nature of the policy on which the document bears and the contents of the document — will often weigh in favour of keeping the document confidential” (at para. 105). Notably, the government does not need to identify a specific harm which would result from disclosure to successfully invoke public interest immunity: “The focus must remain on whether the public interest in the document’s disclosure outweighs the public interest in its remaining confidential” (at para. 111). Of critical importance in this regard will be the judge’s conclusion “on the nature and probative value of the evidence” (at para. 116), and disclosure is only likely to be appropriate where the probative value is high, a judgement to be made by reference to how the evidence bears on the Bodner factors:
Properly applied in the Bodner context, public interest immunity requires a careful balancing of the public interests in confidentiality and disclosure. Since the public interest in the confidentiality of documents concerning Cabinet deliberations is often particularly strong, the public interest in their disclosure will usually need to be stronger still to warrant their disclosure (at para. 120).
This was all obiter, strictly speaking, beause in the BC case, the threshold was not met. The challengers had relied on a decade-old decision in which the provincial government had been held to have failed the Bodner standard, but there was no reason to think that a government many years later must have fallen into the same trap as its predecessor.
Interestingly, in the companion case from Nova Scotia, the challengers were largely successful. Karakatsanis J accepted that the reasons provided by the government, which were highly critical of the independent commission, “may raise some concerns about whether the government failed to meaningfully engage with the commission’s recommendations and about the government’s respect for the commission process” (at para. 43). Having reviewed the underlying documents, Karakatsanis J was concerned that the discussion of government-wide implications of raising judicial salaries and a “communications plan” which warned of likely public dissatisfaction of adopting the commission’s recommendations would not satisfy the limited requirements of Bodner-style review:
In particular, the suggestion that if the government accepts the commission’s recommendations, it will be criticized for not availing itself of the option given to it by the Nova Scotia legislature to vary or reject the commission’s recommendations, is hardly a rational basis for departing from those recommendations. It would undermine the legitimacy of the government’s response if Cabinet relied on these considerations (at para. 54).
Turning to public interest immunity, Karakatsanis J accepted that ministers were entitled to expect that their deliberations would remain confidential for decades into the future but also noted that the documents had a high probative value for a Bodner-style review. In her view, disclosure was appropriate:
The level of decision making, the nature of the policy concerned, the contents of the discussion of government-wide implications and of the communications plan and the timing of the disclosure all weigh in favour of these components of the Attorney General’s report remaining confidential. Because the policy concerns a constitutional requirement relating to the justice system, and, thus, the administration of justice, it also weighs in favour of disclosure. The exclusion of this evidence from the record would undermine the reviewing court’s ability to deal with central issues on Bodner review: whether the government articulated legitimate reasons for departing from the commission’s recommendations; whether the government’s response was grounded in improper considerations and whether the government respected the commission process. The interests of the administration of justice thus strongly favour the disclosure of these parts of the Attorney General’s report. I conclude that the public interest in their disclosure outweighs the public interest in their remaining confidential (at para. 72).
Notably, the whole of the Attorney General’s submission was not disclosed. Parts of it were simply not relevant. Only in respect of the government-wide implications and communications plan sections did the public interest in disclosure outweigh the public interest in protecting cabinet confidentiality.
In summary, then, I think Karakatsanis J was quite right to be cautious about the scope of the record in Bodner-style judicial review and her analysis of public interest immunity is a very useful application of familiar principles.
This content has been updated on August 3, 2020 at 15:53.