Agency Independence in Canada: Structure, Adjudication, and the Limits of Autonomy

Here are my speaking notes for the webinar on ‘Agency Independence in Comparative Perspective’ (see here)

Agency independence is often spoken of as if it were a single, settled concept. In Canadian administrative law, it is nothing of the sort. Independence is contingent, variable, and to a very large extent the product of legislative choices about institutional design.

Constitutional questions

Canadian regulatory agencies are divided along familiar constitutional lines. Some operate at the federal level—telecommunications, broadcasting, nuclear energy—while others are provincial, such as securities commissions, financial services regulators, and energy boards. Beyond that basic division, there is remarkably little uniformity.

There is no standard template, federally or provincially. Some agencies enjoy long, fixed terms and removal only for cause; others are far more exposed to political control. The result is a wide variety of institutional arrangements, reflecting not a coherent theory of independence but rather pragmatic, and sometimes ad hoc, legislative design.

Unlike courts, administrative agencies in Canada have no distinct constitutional status. They are created by statute to implement “policy” and are said to straddle the line between the judicial and executive branches: Ocean Port Hotel Ltd. v. British Columbia (General Manager, Liquor Control and Licensing Branch), 2001 SCC 52. Regulatory agencies exercise executive, legislative, and judicial functions, often within the same institution: Brosseau v. Alberta Securities Commission, [1989] 1 SCR 301. Their degree of independence therefore depends almost entirely on the choices made by the legislature that creates them. As the Supreme Court put it in Ocean Port, at para. 22:

Ultimately, it is Parliament or the legislature that determines the nature of a tribunal’s relationship to the executive. It is not open to a court to apply a common law rule in the face of clear statutory direction. Courts engaged in judicial review of administrative decisions must defer to the legislator’s intention in assessing the degree of independence required of the tribunal in question.

Structural independence versus adjudicative independence

One important exception to this general dependence on statute arises in the context of adjudication. When administrative agencies determine rights and obligations, the demands of procedural fairness impose meaningful constraints on outside interference. In this context, independence is not merely a matter of legislative grace: Iwa v. Consolidated-Bathurst Packaging Ltd., [1990] 1 SCR 282; Shuttleworth v. Ontario (Safety, Licensing Appeals and Standards Tribunals), 2019 ONCA 518. Statutes will be interpreted to comply with the common law protection for adjudicative independence: see e.g. Canadian Association of Refugee Lawyers v. Canada (Immigration, Refugees and Citizenship), 2020 FCA 196.

This gives rise to a crucial distinction between structural independence and adjudicative independence. Structural independence—term length, security of tenure, budgetary control—is entirely a statutory matter (save in Quebec, where structural independence for adjudicative decision-makers has quasi-constitutional status (Québec (Procureure générale) c. Barreau de Montréal, [2001] R.J.Q. 2058), albeit maybe not for regulatory agencies when not exercising adjudicative functions: Procureur général du Québec c. Duquette, 2025 QCCA 616).

Adjudicative independence, by contrast, is a component of procedural fairness. Decision-makers must be free from improper influence, including political direction, in the resolution of individual cases: Innisfil Township v. Vespra Township, [1981] 2 SCR 145. In addition, a decision-maker cannot act under dictation from political masters: Roncarelli v. Duplessis, [1959] SCR 121.

Recent regulatory legislation increasingly makes this distinction explicit, often affirmatively protecting adjudicative independence while preserving other mechanisms of political accountability: see e.g. Ontario Energy Board Act, 1998, SO 1998, c 15, Sch B, s. 4.3(8); Canadian Energy Regulator Act, SC 2019, c 28, s 10, s. 17(1). This reflects a mature recognition that agencies can be both policy-responsive and adjudicatively impartial. Indeed, even beyond the specific protection for adjudication, it is fair to say that governments do not typically intervene in the day-to-day operations of regulatory agencies.

Policy alignment without direct interference

Consistent with the flexible, ad hoc approach, Canadian governments retain numerous tools to ensure that agencies remain aligned with broader policy objectives.

Appeals are one such mechanism. At the federal level, the telecommunications and transportation statutes – two profoundly significant pieces of legislation – permit appeals or references to Cabinet, allowing political actors to retain ultimate responsibility for major policy choices: see e.g. Telecommunications Act, SC 1993, c 38, s. 12.

Policy directions are another. Most regulatory statutes, both federal and provincial, authorize Cabinet to issue binding policy directions: see e.g. Telecommunications Act, SC 1993, c 38, s. 8. Even the Bank of Canada is not immune: Bank of Canada Act, RSC 1985, c B-2, s. 14. These directions must be followed, but the statutory power is generally limited to “broad” policy matters rather than individual matters (and even if it were not, it would be interpreted to exclude interference with adjudication).

Institutional design has also evolved. There is a growing move toward boards of directors responsible for an agency’s operations, overseeing a CEO and senior management, and answerable to Cabinet on policy matters. Carefully drafted memoranda and governance documents are often used to draw lines between operational oversight and adjudicative independence. The Ontario Energy Board is a good example: the Ontario Energy Board Act provides for a governance structure in which responsibility is split between a board of directors (linked to the Minister via a memorandum of understanding) and commissioners charged with adjudication: see also Securities Commission Act, 2021, SO 2021, c 8, Sch 9.

Appointments and removals remain central. Regulatory appointees are typically appointed by Cabinet and removable only for cause. While courts show deference to executive judgments about “cause” (see e.g. Wedge v. Canada (Attorney General), 1997 CanLII 5331 (FC) (removal for suspected criminal activity notwithstanding the absence of criminal charges), they insist on exacting procedural safeguards: Vennat v. Canada (Attorney General), 2006 FC 1008, [2007] 2 FCR 647; Shoan v. Canada (Attorney General), 2017 FC 426. Here again, the balance is one of constrained political control rather than insulation (see further here). There has been at least one instance of the chair of an agency — the Canadian Nuclear Safety Commission — being removed because of a policy disagreement with the government relating to perceived lack of alacrity in reopening a power plant that was producing isotopes, albeit that the appointment here was at pleasure and not subject to a cause requirement: Keen v. Canada (Attorney General), 2009 FC 353.

Finally, there are informal norms. Day-to-day political interference is generally avoided, not because it is legally impossible, but because it is institutionally destabilizing (see further here, on the Bank of Canada, whose independence is defended more by norm than by law). That said, patronage has been a persistent feature of the administrative state, including in tribunals charged with adjudicating rights and obligations. Independence in practice has always been partial and imperfect.

Conclusion

Agency independence in Canada is a set of negotiated arrangements, shaped by statute, constrained by procedural fairness, and moderated by political norms.

This content has been updated on December 29, 2025 at 21:22.