Although the Tribunal framed the issue as being one of jurisdiction, its analysis shifted to a merits review based on Mr. Bokhari’s application and submissions in response to the Notice. The Tribunal moved almost immediately from assessing whether Mr. Bokhari’s pleadings fell within the Tribunal’s jurisdiction to determining whether he could succeed in his application. The Tribunal foreclosed Mr. Bokhari’s claim by choosing the narrower of two lines of authority in deciding whether the type of ankle injury sustained by Mr. Bokhari qualified as a disability under the Code. This is not a proper jurisdictional analysis; it is a disguised merits review. This approach impermissibly collapses the merits of the claim—as pleaded—into the threshold question of whether an application is within the Tribunal’s jurisdiction (at paras. 21-22).

There was “no question” that the asserted claim about discrimination in employment was one the Tribunal could hear — the very essence of the Tribunal’s role is to determine, on the merits, if the claim is made out (at para. 25). And the Tribunal can only make this determination after “a contextual, fact-based inquiry” (at para. 26).

Second, the conclusion that B’s claim fell outside the Tribunal’s jurisdiction was also unreasonable, as the required contextual analysis was never carried out:

Examining [the factors that contribute to disability claims] requires a careful contextual analysis in which the alleged discrimination occurred. This is necessary to determine whether an actual or perceived ailment causes the applicant to experience the loss or limitation of opportunities to take part in the life of the community on an equal level with others: Boisbriand, at para. 80. This is a highly fact-based evaluation (at para. 52).

As a remedy, Rees J ordered that B’s file should proceed to the next stage in the adjudicative process (at para. 59): the only reasonable outcome here was that B’s claim was within the Tribunal’s jurisdiction.

Third, it was unreasonable to use a balance of probabilities standard when screening complaints. As Rees J noted, the practice direction itself offered no explanation for the move from the ‘plain and obvious’ standard to the balance of probabilities standard (at para. 33), a breach of the obligation to justify a departure from internal precedent imposed as an aspect of reasonableness review under Vavilov (at paras. 36-39). Indeed, Rees J explained, it is difficult to understand exactly how a balance of probabilities standard can rationally be applied in circumstances where no evidence is yet available, because the very point of the balance of probabilities standard is to make evidence-based determinations (at paras. 34-35). And, here, it is also important to note the quasi-constitutional status of human rights laws in Canada — applications are not to be dismissed lightly (at para. 41).

Ultimately, the balance of probabilities standard was simply incompatible with the Tribunal’s mandate, even if it might have led to efficiency gains:

Applying the balance of probabilities standard to jurisdictional questions inevitably results in screening out, at a threshold stage, applications that appear unlikely to fall within the jurisdiction of the Tribunal but which, with the benefit of a factual record and argument, would ultimately be determined to be within the Tribunal’s jurisdiction. This results in applicants who have been discriminated against being denied the Code’s protection. Managing workload and efficiency cannot justify this result. A higher threshold must be met before the Tribunal can reasonably close its doors to applicants on jurisdictional grounds under r. 13.

This is not to say that every human rights matter must proceed to a full hearing on the merits. Other means exist of the Tribunal managing its workload (see e.g. Gill v. Human Rights Tribunal of Ontario et al., 2014 ONSC 1840, at para. 12). But these other means are designed to get sufficient information before an adjudicator who can make an informed decision, not simply and solely to achieve efficiency gains.

There is a parallel here with procedural fairness cases where judges are faced with unfairness in a particular instance that results from an approach adopted to achieve efficiencies — invariably, the judges will find that the efficient approach is unfair, regardless of the overall utilitarian calculus (see here). Here, too, where efficiency runs into an unreasonable interpretation of a statutory mandate, the courts will intervene. Put very simply, should a tribunal wish to address its backlog, it has to find a way to do so that is consistent with carrying out its statutory mandate.

Benison

The respondent here was the External Review Committee of the Royal Canadian Mounted Police. The applicants were RCMP members who sought orders of mandamus to compel the RCMP to make decisions on their complaints within 30 calendar days (and to publish service standards). The ERC hears internal labour-related grievances within the RCMP. A series of legislative and regulatory amendments in 2014 increased the ERC’s workload significantly, tripling its case load and triggering the ERC to create a triaging system. By the time of the Federal Court hearing, the applicants had been waiting between 3.5 and 4.5 years for findings and recommendations from the ERC. This notwithstanding the ERC’s prioritization system, designed to grapple with the budgetary constraints it was operating under:

Under the prioritization system, rather than reviewing cases on a first-come, first-served basis, cases are first prioritized based on their subject matter, with cases involving more severe disciplinary sanctions given priority to lessen the impact of such sanctions on members whose appeals are successful. Once prioritized by subject matter, cases “in the queue” that pertain to the same subject matter are reviewed in the order they are received. The intent of the prioritization system is to “maximize the ERC’s current resources” (at para. 114).

The Federal Court dismissed the mandamus applications, faithfully applying the demanding criteria set out in Apotex Inc. v. Canada (Attorney General) (C.A.), 1993 CanLII 3004 (FCA), [1994] 1 FC 742 aff’d 1994 CanLII 47 (SCC), [1994] 3 SCR 1100. If I can be permitted a brief digression, the Federal Court judge who stuck closely to precedent on this occasion was Go J, who was the subject of a bizarre article earlier this year claiming, on the basis of a selective reading of a small selection of her judgments, that she is “in charge of waving migrants into the country from a judicial seat” — it would seem, in fact, that Go J is not an activist hellbent on dismantling the immigration system, which is by far and away the most active practice area as far as mandamus is concerned. But I digress.

It is helpful to list the Apotex criteria for mandamus:

1. there is a public duty to act;

2. the duty is owed to the applicant;

3. there is a clear right to performance of that duty, in that the applicant has satisfied all conditions precedent giving rise to the duty and there was a prior demand for its performance, a reasonable time to comply with the demand and a subsequent refusal, either expressed or implied (e.g., by unreasonable delay);

4. certain criteria are satisfied if the duty sought to be enforced is discretionary;

5. no other adequate remedy is available to the applicant;

6. the order will be of some practical value or effect;

7. the court finds no equitable bar to the relief sought; and

8. the balance of convenience favours issuing the order.

These eight criteria are typically thought of as conjunctive, meaning an applicant must meet all of them to convince a court to issue an order of mandamus. Heckman JA acknowledged this, but went on to hold that the first four criteria and the last four play different roles in the analysis:

The Apotex test is conjunctive. If, in the view of the reviewing court, any of the eight criteria are not favourable to the applicant, mandamus will be denied. However, and significantly, the eight Apotex criteria do not serve the same purpose. The first four criteria are considered by the reviewing court to determine whether the applicant has shown that they would be entitled to an order of mandamus. The final four criteria relate to reasons for which the reviewing court should exercise its discretion to deny the remedy to which the applicant would otherwise be entitled (at para. 45).

The first four represent the “traditional requirements” for mandamus (at para. 46); the latter four “set out some of the grounds on which reviewing courts may refuse relief on judicial review proceedings” (at para. 48). Then, Heckman JA says, the “burden of proof shifts for the second set of criteria” (at para. 49), that is, the burden is on the respondent, once the first four criteria have been met, to convince the court not to grant an order of mandamus. The applicant bears a “tactical burden” to address the exercise of discretion, if the point is raised, but the burden remains on the respondent (at para. 51).

Here, the third criterion was key, i.e. whether the applicants had a “clear right” to the performance of the duty to make a decision. The leading Federal Court authority on this point is Conille v. Canada (Minister of Citizenship and Immigration), [1999] 2 FC 33, which sets out a three-part test: “a delay is unreasonable where (1) it has been longer than the nature of the process required, prima facie; (2) the applicant and their counsel are not responsible for it; and (3) the authority responsible for the delay has not provided satisfactory justification” (at para. 76). This, Heckman JA explained (at paras. 77-79), is consistent with more recent Supreme Court of Canada authorities on delay, namely Law Society of Saskatchewan v. Abrametz, 2022 SCC 29, [2022] 2 SCR 220 and Blencoe v. British Columbia (Human Rights Commission), 2000 SCC 44, [2000] 2 SCR 307).

Heckman JA held that the applicants did have a clear right to the performance of a duty to make a decision. To begin with, the purpose of the ERC regime was timely decision-making, analogous to the labour relations context where the goal is to make decisions swiftly in the interests of preserving workplace harmony:

The ERC impartially reviews appeals of conduct measures imposed on RCMP members and of decisions regarding harassment complaints, revocations of appointments, discharges, demotions and ordered stoppages of pay and allowances. In doing so, it contributes to fair and equitable labour relations and accountability within the RCMP. As such, the values and objectives that animate the ERC’s mandate overlap with those that govern the work of labour arbitrators (at para. 85).

This was underpinned by a statutory requirement on the ERC to publish service standards respecting time limits for dealing with grievances (at para. 90). Accordingly, the multi-year delays meant that the ERC’s decisions had been “delayed beyond the inherent time requirements for processing such matters” (at para. 97). The first Conille factor was therefore satisfied. (In relation to the second application for mandamus, about the service standards, Heckman JA held that the statute did not create a clear legal duty (at para. 171).)

What, then, of the ERC’s justification for the delays? The Federal Court relied on “(a) a “dramatic” increase in the ERC’s caseload that occurred due the 2014 amendments; (b) the government’s failure to increase (until 2020) the ERC’s budget and staffing levels to allow it to keep up with the higher caseload, leading to a backlog of appeals; and (c) the system adopted by the ERC in response to this backlog to prioritize appeals according to severity and under which most of the appellants’ appeals were ranked at the lowest priority level for processing” (at para. 93). Heckman JA had a simple answer to these propositions: “The onus rested with the ERC to adduce evidence relating to whether the relative complexity of individual appeals might justify delays in excess of the time requirements inherent to its proceedings. It failed to meet that onus” (at para. 98. Cf the comment at para. 113 that “The Federal Court’s findings on the causes of the delay are supported by the evidence”).

Nonetheless, Heckman JA went on to express “serious reservations” about two other aspects of the Federal Court’s approach (at para. 99).

First, the Federal Court had drawn an analogy with the approach taken to mandamus in immigration cases. There, the courts have consistently held that questions of delay must be assessed against the backdrop of the Minister’s “power to set policies and make decisions to manage the flow of immigrants to Canada, including imposing deliberate delays on the processing of certain classes of applications for the benefit of other classes, so long as the policies and decisions are made in good faith and are consistent with the purpose, objectives, and scheme of the IRPA” (at para. 107). Heckman JA was not persuaded that the analogy was apt:

In my view, before accepting this analogy, the Federal Court was required to determine whether it was apt. The Minister’s power to discriminate between classes of immigrants that justified the processing delays imposed in Jia and Vaziri was found to be necessary to achieve that scheme’s objective of managing Canada’s immigration flow on the basis of social and economic policy considerations and to be either expressly or impliedly set out in the IRPA scheme (Vaziri at para. 36). In accepting the ERC’s explanation as a satisfactory justification, the Federal Court did not ask whether the ERC’s adoption of its prioritization system and its imposition of open-ended delays on the processing of the appeals filed by certain RCMP members furthered the purpose, objectives and scheme of the RCMP Act. To the contrary, as previously noted, it failed to consider factors that were relevant to ascertaining the values and objectives that underlie the ERC’s mandate under this scheme (at para. 109).

Fundamentally, the problem Heckman JA identified with the ERC’s position was that it could justify a delay of any length: “Essentially, the ERC submits that it is doing what it can with the resources it has been given and that, at some point in the future, the appellants’ appeals will be reviewed. The problem with accepting such an explanation as a satisfactory justification for delay, even if it is supported by adequate evidence, is that it is untethered to the actual delay suffered by an applicant for mandamus” (at para. 118). Put another way, resource constraints can never justify delays that defeat the purpose of the statutory scheme.

Second, the Federal Court had required the applicants to demonstrate that they had suffered significant prejudice as a result of the delay. There has been a developing split in the Federal Court on this issue, with some judges taking the view post-Abrametz that prejudice is now a requirement for mandamus applications. Heckman JA found that significant prejudice is not a requirement for mandamus. Abrametz discussed significant prejudice in the context of delays that are so long as to amount to an abuse of process (and lead to remedies such as stays of proceedings or awards of costs) rather than delays under Conille that might justify an order accelerating the decision-making process:

As specified in Abrametz, both inordinate administrative delay and evidence that this delay has directly caused significant prejudice are required for a finding that the delay has resulted in an abuse of process that could warrant a stay of proceedings. However, these prerequisites are distinct. There is no principled basis on which evidence of significant prejudice could be required to establish an unreasonable delay from which a refusal to perform a public duty to act might be implied (at para. 125).

Heckman JA did intimate that the ERC might be able to adduce evidence in a future case to justify delays, but that it had failed to do so here (at para. 129).

Heckman JA then turned to what he had previously characterized as discretionary grounds for refusing the applications. Two were relevant here: a potential equitable bar to relief (Apotex criterion 7) and the balance of convenience (Apotex criterion 8).

In terms of equitable bars, Heckman JA addressed the proposition, set out in Jia v. Canada (Citizenship and Immigration), 2014 FC 596, that permitting applicants for mandamus to ‘skip the line’ ahead of other similarly situated individuals is inequitable and a potential bar to relief, because it violates the ordinary rule that applications should be processed in the order in which they are received. However, Heckman JA expressed the fear that, in practice, this would make mandamus unavailable as a remedy for administrative delay:

Under this logic, only the individual whose application is next in line to be considered by the decision-maker could ever successfully apply for mandamus. Accepting this argument could shield from the reach of the prerogative writ of mandamus unreasonable delays in administrative decision-making simply by virtue of the fact that they affect groups of individuals, some of whom choose not to seek mandamus (at para. 146).

In this case, this concern did not arise in any event, as the ERC had a prioritization system (not a ‘first-come, first-served’ system) in place (at para. 147).

In terms of balance of convenience, Heckman JA noted that withholding a remedy for unlawful administrative action raises serious rule of law concerns and should, therefore, not be done lightly (at paras. 153-154, citing Apotex and MiningWatch Canada v. Canada (Fisheries and Oceans), 2010 SCC 2, [2010] 1 SCR 6). In this case, there was no basis to refuse a mandamus due to the balance of convenience:

while the issuance of the writ of mandamus sought by the appellants would likely increase the delay required to process higher priority appeals, the evidence in the record does not establish the magnitude of any additional delays nor that the impact of such delays on other RCMP members would be disproportionate. Nor does the record give me any reason to believe that “administrative chaos” would ensue from prioritizing the appellants’ appeals through an order of mandamus, or that this application for mandamus otherwise gives rise to the “clearest of circumstances” in which relief ought to be denied at the last step of the Apotex test. I am therefore not convinced that relief should not, on the balance of convenience, be granted (at para. 162).

Overall, this is an extremely important decision on mandamus, by far the most detailed and comprehensive treatment of the remedy in recent years. The analysis of significant prejudice and equitable bars is likely to be most influential going forward, especially in the immigration space.

More generally, Heckman JA’s approach in Benison is consistent with a recent trend away from treating mandamus as a special remedy subject to its own special rules (see also the discussion of the Rogers decision here). In many respects, the Apotex criteria are a hangover from a previous era, in which the law of judicial review revolved around the prerogative writs of certiorari, prohibition, mandamus, quo warranto and habeas corpus. Today, the law of judicial review revolves instead around the concepts of procedural fairness and reasonableness — substantive principles of administrative law — with remedies being used to put the individual in the position they would have been in but for the unlawful administrative action.

Detaching the first four Apotex criteria from the latter four criteria, as Heckman JA did, is consistent with this move away from an administrative law system based on remedies and towards one based on principles. On Heckman JA’s approach, the latter four criteria are considerations relating to discretion to withhold a remedy that apply in all cases, not just mandamus cases. Indeed, Benison might be read as supporting the proposition that there are only four Apotex criteria (criteria 1–4), not eight.

That said, Heckman JA’s approach still retains some of the specialness of mandamus, as the first four Apotex criteria are retained and must be satisfied by any applicant for mandamus. A more radical approach would be a complete decoupling of principles from remedies, whereby the demonstration of unreasonable administrative delay would function as a finding of unlawfulness that a court could remedy in its discretion. Put another way, undue delay or prolonged inaction may be legally unreasonable and, if so, should be remedied (subject to the existence of discretionary bars to relief: see also here). On this approach, the remedy of a mandatory order directing a decision-maker to act follows from the principle that a decision-maker cannot act unreasonably. Benison could be read, in broad terms, as consistent with the more radical approach, because the general delay in processing applications was unreasonable due to inconsistency with the statutory scheme and required a remedy, of which mandamus was the appropriate one to grant. By contrast, the absence of specific service standards was not unreasonable, as it was not inconsistent with the statutory scheme (at para. 171) and no remedy was required.

Conclusion

This is an important pair of decisions, each of which is thought-provoking (and strikingly well-written). Administrative tribunals grappling with delays, and individuals frustrated by them, would do well to read both closely. In this, as in all things, a decision-maker’s statutory mandate is critically important.