The Language of Administrative Law V: Metaphor

(This is the fifth in a series of posts. Read the first, second, third and fourth. The draft can be downloaded in its entirety here.)

One very useful way to conceptualize the administrative state is to perceive it as a spectrum, along which, as Justice LeBel put it in Imperial Oil, the requirements of the law “may vary in order to reflect the context of a decision‑maker’s activities and the nature of its functions”:^[1]

The categories of administrative bodies involved range from administrative tribunals whose adjudicative functions are very similar to those of the courts, such as grievance arbitrators in labour law, to bodies that perform multiple tasks and whose adjudicative functions are merely one aspect of broad duties and powers that sometimes include regulation‑making power. The notion of administrative decision‑maker also includes administrative managers such as ministers or officials who perform policy‑making discretionary functions within the apparatus of government.^[2]

This well-known idea of a “spectrum”^[3] can be roughly mapped as follows: Ministers[4] à Crown Corporations[5] àSocial and Economic Regulation[6] à Quasi-Judicial[7] à Judicial.[8]

Viewing functions on this spectrum is a very useful, intuitive way of grasping many of the issues that arise in terms of political and legal control of the administration.^[9] On the purely political end of the spectrum lies ministerial decision-making, where political control through conventions of accountability to the legislature predominates. Here, procedural protections are diminished – indeed, in Imperial Oil, an argument that the minister was biased because his department would have won a budgetary windfall from penalizing an oil company failed for precisely this reason. In addition, the range of considerations that a minister might take into account in making a decision is often very broad indeed.^[10]

On the purely legal end of the spectrum lies judicial decision-making. By “judicial” in this context, I do not mean decision-making by courts of law. Rather, I am concerned with the application by administrative bodies of objective legal norms to the facts as found. The French term “juridictionnelle”, which does not have a ready English equivalent, captures the idea. Here, political interference is – or at least should be^[11] –frowned upon, for the distribution of costs and benefits set out by the legislature should not be interfered with by executive fiat.^[12] Legal control is heightened, in the sense that the range of considerations that a judicial decision-maker may legitimately take into account is tightly constrained: if the question is whether an applicant has accumulated enough days of work to claim an entitlement, the decision-maker cannot rove into the applicant’s conduct or utility to society. Procedural protections are at their strongest here, where the decision-making process – the application of objective legal norms^[13] to facts – closely resembles that of a court of law.

In between the two extremes, the extent of political and legal control varies as one moves back and forth along the spectrum, more strongly legal towards the judicial end, more strongly political towards the ministerial end. In terms of giving a snapshot of the administrative state and providing a rough idea against which one’s intuitions about the appropriate level of political and legal control can be tested, this spectrum of functions is exceedingly useful.

As useful as it is, however, the metaphor of the spectrum serves to distract courts from their role in ensuring procedural fairness in the administrative process. In procedural fairness, cases are often said to turn on an appreciation of all the relevant context: “Like the principles of natural justice, the concept of procedural fairness is eminently variable and its content is to be decided in the specific context of each case”.[14] An administrative process “need not assume the trappings of a court”.[15] Only “some kind of hearing”[16] in which the conventional list of procedural protections – notice, making submissions, representation by counsel, cross-examination of witnesses and so on – need not be followed to the letter. Depending on the interplay of the five “Baker” factors,[17] a particular decision can be placed on the spectrum somewhere between the two extremes in order to determine the appropriate procedural protections:

Some administrative tribunals are closer to the executive end of the spectrum: their primary purpose is to develop, or supervise the implementation of, particular government policies.   Such tribunals may require little by way of procedural protections. Other tribunals, however, are closer to the judicial end of the spectrum: their primary purpose is to adjudicate disputes through some form of hearing. Tribunals at this end of the spectrum may possess court-like powers and procedures.[18]

However, the use of “may” in this passage is telling. Placing a decision on a spectrum is not enough. Something more is required before a court can satisfy itself that the procedure at issue was fair or unfair. Indeed, the metaphor of the spectrum obscures more than it reveals. The Baker factors emphasize the importance of context and evoke a spectrum of administrative functions but they do not assist courts or decision-makers in determining whether a particular procedural right should be granted. In reality, very specific normative criteria attach to each of the procedural rights that an individual may claim in the administrative process. For example: minimum standards of fairness require that an individual “have full and complete notice of the charges against him and an opportunity to reply thereto”;[19] an oral hearing will be required where it is necessary to determine an issue of credibility;[20] evidence cannot be refused on a critical point;[21] representation by counsel is required only where the proceedings are so important and complex that an individual will be unable to make his or her case effectively;[22] and cross-examination need only be permitted where it is necessary to ensure the decision-maker has the “full story”.[23] In this sense, the Baker factors are fabulous. Some judicial judgement will be required in determining the answers to the more specific questions relating to particular procedural rights, but judges are certainly not roving along a spectrum searching for a mirror image of the type of process presented by the case in hand.

A risk of swallowing bromides about context and relying on the vague, abstract notion of a spectrum to ‘resolve’ concrete cases is that courts may create “a modern version of Sir William Wade’s nightmare of a Tennysonian ‘wilderness of single instances’ in which all the contextual factors will be relevant in considering what the requirements of procedural fairness are in a given situation without any factor or group of factors having decisive weight in shaping what is in practice required”.[24] However, jurists can avoid this danger by understanding that the question in procedural fairness cases is whether a particular right should have been accorded – not whether the procedure was in fact consistent with the decision-maker’s place on the spectrum of administrative procedures. Fairness is a normative matter, not a descriptive one.

The metaphor of the spectrum also causes problems in determining the appropriate degree of administrative independence from outside interference. Take, for instance, a central bank whose mandate is to set monetary policy. On the spectrum of procedural fairness, a central bank would fall more or less in the middle, along with telecommunications and competition regulators and similar bodies. Yet central bank independence from political interference is a cherished norm of most developed liberal democracies.[25] For instance, even though the Bank of Canada is formally subject to the authority of the Minister of Finance, with whom its governor must meet regularly,[26] howls of outrage greet any perceived political interference with the establishment of monetary policy.[27] Howls of outrage that, one might add, are often absent when politicians attempt to influence judicial decision-making by administrative tribunals. Clearly the metaphor of the spectrum fails to capture important information about administrative independence from political interference.

Moreover, the idea of a set of administrative functions that may glide along a spectrum militates against any suggestion that some decision-makers should be given special protection. Ellis has convincingly argued, for instance, that administrative tribunals exercising judicial functions should be protected by the same doctrines of independence that courts of law benefit from: norms that guarantee security of tenure, security of remuneration and institutional autonomy.^[28] Indeed, in the province of Quebec, where administrative independence has a quasi-constitutional status, courts have been able to ensure these protections for administrative tribunals. It might be the case that the metaphor of the spectrum has such a hold on the judicial imagination that it has become impossible to conceive of an administrative tribunal with a dedicated, single function that is so important as to warrant heightened protection. A metaphor that evokes movement may shield from the judicial eye the need for special protection at certain fixed points on the spectrum.[29]

[1] Imperial Oil, at para. 31.

[2] Imperial Oil, at para. 31.

[3] Martineau v. Matsqui Disciplinary Board, [1980] 1 S.C.R. 602, at pp. 628-629 per Dickson J., dissenting:

A purely ministerial decision, on broad grounds of public policy, will typically afford the individual no procedural protection, and any attack upon such a decision will have to be founded upon abuse of discretion. Similarly, public bodies exercising legislative functions may not be amenable to judicial supervision. On the other hand, a function that approaches the judicial end of the spectrum will entail substantial procedural safeguards. Between the judicial decisions and those which are discretionary and policy-oriented will be found a myriad decision-making processes with a flexible gradation of procedural fairness through the administrative spectrum.

See also Apotex Inc. v. Canada (Attorney General), [2000] 4 F.C.R. 264, at para. 104, per Evans J.A., referring to “the spectrum of powers ranging from the legislative, through the administrative, to the judicial”.

[4] I have in mind here government departments headed by a politician who is accountable to a legislature (and perhaps also, as a modern addition to the convention of responsible government, to the public at large) and whose name departmental officials may act.

[5] I have in mind here the various bodies established by the state to achieve commercial goals; these are relatively rarely subject to judicial review because their activities are primarily commercial in nature, but they are subject to governmental control.

[6] I have in mind here bodies charged with regulating complex aspects of society and the economy. Central banks, telecommunications regulators and competition regulators would fall under this rubric.

[7] I have in mind here administrative tribunals that the legislature has established to settle disputes, usually between an administrative body and an individual but sometimes (as with human rights tribunals) between two individuals. Often these tribunals complement the work of social and economic regulators: in some of its functions, for instance, a telecommunications regulator will act quasi-judicially but in others it will act as a developer of general rules of policy for its regulatory sector.

[8] I explain this term further below: the difference between a quasi-judicial and judicial function is simply one of degree, with a quasi-judicial decision-maker entitled to take into account a broader range of factors – such as the overall effect of a decision on achievement of its overall policy goals – than a judicial decision-maker.

[9] For this distinction, see e.g. Adam Tomkins, Our Republican Constitution (Oxford University Press, Oxford, 2005).

[10] See e.g. Canada (Minister of Transport, Infrastructure and Communities) v. Jagjit Singh Farwaha, 2014 FCA 56, at paras. 88-99.

[11] Ron Ellis, Unjust by Design: Canada’s Administrative Justice System (UBC Press, Vancouver, 2013) inveighs most effectively against political interference, largely unremarked by the wider public, with judicial tribunals.

[12] Martin Shapiro, Who Guards the Guardians? Judicial Control of Administration (Athens: University of Georgia Press, 1988), p. 112:

Imagine our outrage if, at the end of a court trial, the president of the United States called up the judge and told her how he wanted the case to come out. Let us suppose an agency has held a rule-making proceeding that involved hundreds of hours of testimony and thousands of pages of written submissions. It has listened at length to every interested group and heard the rebuttal of each group to the testimony of every other. It has compiled a thousand-page-long rule-making record. It has then composed a statement showing that it has acted synoptically to consider every significant issue and arrive at the best possible decision. Those who have been watching and participating in such a process are going to be equally outraged if, just before the agency publishes its final rule, the president calls to tell the agency what rule it should adopt.

[13] An intelligent student once objected to me that not all norms applied by courts of law can truly be described as objective, open-textured constitutional provisions being perhaps the best example. This is a complicated area (see e.g. Matthew Kramer, Objectivity and the Rule of Law (Cambridge: Cambridge University Press, 2007); R. v. Department of Education and Employment, ex parte Begbie, [2000] 1 W.L.R. 1115, at p. xxx (“Fairness and reasonableness (and their contraries) are objective concepts; otherwise there would be no public law, or if there were it would be palm tree justice”). I nonetheless think the basic point about the application of law to facts withstands scrutiny.

[14] Knight v. Indian Head School Division No. 19, [1990] 1 S.C.R. 653, at p. 682.

[15] Knight, at para. 49.

[16] Henry Friendly, “Some Kind of Hearing” (1975), 123 U. Pa. L. Rev. 1267.

[17] Baker v. Canada (Minister of Citizenship and Immigration), [1999] 2 S.C.R. 817. This non-exhaustive list (ibid., at para. 28) was summarized more recently in Canada (Attorney General) v. Mavi, 2011 SCC 30, [2011] 2 S.C.R. 504, at para. 42:

(i) “the nature of the decision being made and the process followed in making it” (para. 23); (ii) “the nature of the statutory scheme and the ‘terms of the statute pursuant to which the body operates’” (para. 24); (iii) “the importance of the decision to the individual or individuals affected” (para. 25); (iv) “the legitimate expectations of the person challenging the decision” (para. 26); and (v) “the choices of procedure made by the agency itself, particularly when the statute leaves to the decision-maker the ability to choose its own procedures, or when the agency has an expertise in determining what procedures are appropriate in the circumstances” (para. 27).

The citations are to Baker.

[18] Bell Canada v. Canadian Telephone Employees Association, [2003] 1 S.C.R. 884, 2003 SCC 36, at para. 21. Emphasis added. See further Colleen Flood and Lorne Sossin, “The Contextual Turn: Iacobucci’s Legacy and the Standard of Review in Administrative Law” (2007), 57 University of Toronto Law Journal 581.

[19] Confederation Broadcasting (Ottawa) Ltd. v. Canadian Radio-Television Commission, [1971] S.C.R. 906, at p. 925.

[20] Khan v. University of Ottawa (1997), 34 OR (3d) 535, at p. xxx.

[21] Université du Québec à Trois-Rivières v. Larocque, [1993] 1 S.C.R. 471, at p. xxx.

[22] New Brunswick (Minister of Health and Community Services) v. G. (J.), [1999] 3 S.C.R. 46, at para. 86.

[23] Irvine v. Canada (Restrictive Trade Practices Commission), [1987] 1 S.C.R. 181, at para. 87.

[24] R. (L.) v. West London Mental Health NHS Trust, [2014] EWCA Civ. 47, at para. 72, per Beatson L.J.

[25] Adrian Vermeule, “Conventions of Agency Independence” (2013), 113 Columbia Law Review 1163

[26] Bank of Canada Act, R.S.C. 1985, c B-2, s. 14.

[27] See e.g. Andrew Coyne, “Bank of Canada, a.k.a the Harper government, needs to be honest about its monetary policy”, National Post, January 22, 2014.

[28] Ron Ellis, Unjust by Design: Canada’s Administrative Justice System (UBC Press, Vancouver, 2013).

[29] This is not to say that administrative tribunals exercising judicial functions should be treated as the functional equivalents of courts of law. In the rigorous enforcement of norms of independence, context may legitimately be taken into account: the Quebec Court of Appeal has been very clear about the importance of following a nuanced approach that is attentive to context. See e.g. Association des juges administratifs de la Commission des lésions professionnelles c. Québec (Procureur général), 2013 QCCA 1690, where the Court of Appeal refused to require that members of the Commission des lésions professionnelles be treated identically to members of the Tribunal administratif du Québec in terms of independence. As Bich J.A. explained at para. 27:

Les garanties conférées en cette matière aux tribunaux administratifs, et qui trouvent leur source dans les règles de la justice naturelle, peuvent donc varier et elles varieront en fonction de la nature précise du pouvoir décisionnel et des modalités d’exercice de ce pouvoir, le tout sous réserve des termes de la loi. Dans cet ordre fluctuant, le tribunal administratif qui exerce des fonctions purement juridictionnelles – et la Cour suprême, là encore, l’a confirmé – doit offrir le niveau de garantie le plus élevé, niveau qui n’a cependant pas, je me permets de le répéter, à être celui des cours supérieures ni même celui des autres cours de justice.

This content has been updated on September 18, 2015 at 12:16.