“Policy” in Administrative Law

“Policy” is one of the most poorly defined terms in the social sciences, as Professor Cairney has observed (The Politics of Evidence-Based Policy Making (2016), at p. 4). In law, the situation is little better. My focus here is on public administration, so I will confine my observations to the law of public administration — notice, though, that they could be applied to private law as well, where “policy” is similarly inchoate and nebulous.

At various points in history the courts have said that “policy” decisions are immune from review by judges. However, “policy” is used in a conclusory way in such cases. In fact, a variety of considerations will lead a court to conclude that a particular matter is, or is not, a “policy” matter. A good example is the decision of the Supreme Court of Canada in Nelson (City) v. Marchi, 2021 SCC 41, where Justices Karakatsanis and Martin laid out a series of considerations relevant to determining whether a “policy” decision made by a public authority is immune from tort liability: “(1) the level and responsibilities of the decision-maker; (2) the process by which the decision was made; (3) the nature and extent of budgetary considerations; and (4) the extent to which the decision was based on objective criteria” (at para. 68). The framework is avowedly open-ended, substantive and contextual: “None of the factors is necessarily determinative alone and more factors and hallmarks of core policy decisions may be developed; courts must assess all the circumstances” (at para. 66). Labelling something as “policy” is insufficient: the four factors govern, not the label applied by the public authority (at paras. 59, 77).

Elsewhere, in the law of judicial review, the concept of “policy” has long since been abandoned as a criterion for the reviewability of a decision. As Professor King has observed, “[t]he conclusion that an issue is ‘political’ and thus inappropriate to determine in a court of law mightbe acceptable as a conclusory label, but it does not obviate the need for a method of analysis” (“Institutional Approaches to Judicial Restraint” (2008) 28 Oxford Journal of Legal Studies 509, at p. 515).

Some authors have attempted to give a meaningful, public-administration-oriented account of “policy”. Macauley and Sprague suggest the following sources of “policy”:

  • (1) orders in council
  • (2) legislation presented to legislature
  • (3) statements made in legislature identified as policy
  • (4) documents presented to legislature identified as policy
  • (5) directives issued under statutory authority
  • (6) regulations

Practice and Procedure Before Administrative Tribunals (Looseleaf), s. 3.2(e)(i)

The difficulty with this list is that many of the items are not really “policy” at all but are in fact legal instruments of binding force and effect. Any decision-maker who is bound to follow the law would be bound to apply a regulation, or to abide by a statutory direction from a third party. Describing these as “policy” is unhelpful, as the legal effects of the instruments listed are independent of their “policy” content. Even “policy” direction provisions do not require regulators to take “policy” into account: they simply require regulators to abide by a binding direction, regardless of its content (see e.g. Broadcasting Act, SC 1991 c 11, s. 23(3): “the Minister may … issue to the Commission a written directive with respect to the condition, regulation or order and the Commission shall comply with any such directive issued by the Minister” (emphasis added)).

More generally, the list really provides a series of places where one might find “policy” memorialized but does not actually illuminate what is “policy” and what is not. If it is to mean anything, “policy” must mean something other than the instructions contained in a binding legal instrument.

Similarly, it is also sometimes said that the role of an administrative tribunal is to implement government “policy” (Ocean Port Hotel Ltd. v. British Columbia (General Manager, Liquor Control and Licensing Branch), 2001 SCC 52, [2001] 2 SCR 781, at para. 24). But this is seriously misleading. Administrative tribunals find facts and apply law to the facts as found: they are engaged in adjudication. Administrative tribunals apply the “policy” embedded in binding legal instruments but they have no free-standing “policy” role. It is not open to an adjudicator to conclude that an individual should not, for example, be refused refugee status for which they meet the statutory criteria on the basis of “policy” considerations. That said, there may be other bodies that make binding decisions — like economic regulators setting basic conditions for participants in a protected market — and who might well be expected to make decisions that are coherent with the “policy” of the government of the day. Still, the broad proposition that these are all “policy” bodies is unsustainable.

It might be thought, in some instances, that administrative decision-makers play a “policy” function when exercising broad statutory powers. But I do not think this is correct either. Consider the strongest possible example of this phenomenon: when a regulator looks to broadly worded statutory objects in order how to determine to exercise a power. In the broadcasting sphere, for example, the CRTC may legitimately look to its dozens of statutory objects when setting licence conditions (Broadcasting Act, s. 3). The objects are each vague, so the CRTC must elaborate their meaning and has a wide margin of appreciation in doing so; moreover, where the objects conflict, the CRTC must choose between them. Is the CRTC making “policy” decisions here? I resist this conclusion: the CRTC is working within (admittedly broad) statutory parameters, as the objects have the force of law, and in choosing between alternative courses of action, the CRTC is implementing the statute. The Broadcasting Act contains the “policy” (albeit one that is perhaps confused or internally incoherent) and the CRTC uses its statutory powers to carry the legislative will into effect as best it can.

“Policy”, therefore, must mean something other than “binding legal instrument” and must also be distinct from underlying substantive considerations motivating action. I will come back to this — after a fairly unpromising start! — in future posts.

This content has been updated on November 27, 2023 at 16:35.