Toronto (City) v. C.U.P.E., Local 79, [2003] 3 S.C.R. 77, 2003 SCC 63

Canadian Union of Public Employees, Local 79 Appellant

v.

City of Toronto and Douglas C. Stanley Respondents

and

Attorney General of Ontario Intervener

Indexed as: Toronto (City) v. C.U.P.E., Local 79

Neutral citation: 2003 SCC 63.

File No.: 28840.

2003: February 13; 2003: November 6.

Present: McLachlin C.J. and Gonthier, Iacobucci, Major, Bastarache, Binnie, Arbour, LeBel and Deschamps JJ.

on appeal from the court of appeal for ontario

Labour law — Arbitration — Dismissal without just cause — Evidence — Recreation instructor dismissed after being convicted of sexual assault — Conviction upheld on appeal — Arbitrator ruling that instructor had been dismissed without just cause — Whether union entitled to relitigate issue decided against employee in criminal proceedings — Evidence Act, R.S.O. 1990, c. E.23, s. 22.1 — Labour Relations Act, S.O. 1995, c. 1, Sch. A, s. 48.

Judicial review — Standard of review — Labour arbitration — Recreation instructor dismissed after being convicted of sexual assault — Arbitrator ruling that instructor had been dismissed without just cause — Whether arbitrator entitled to revisit conviction — Whether correctness is appropriate standard of review — Evidence Act, R.S.O. 1990, c. E.23, s. 22.1 — Labour Relations Act, S.O. 1995, c. 1, Sch. A, s. 48.

O worked as a recreation instructor for the respondent City. He was charged with sexually assaulting a boy under his supervision. He pleaded not guilty. At trial before a judge alone, he testified and was cross-examined. The trial judge found that the complainant was credible and that O was not. He entered a conviction, which was affirmed on appeal. The City fired O a few days after his conviction. O grieved the dismissal. At the arbitration hearing, the City submitted the complainant’s testimony from the criminal trial and the notes of O’s supervisor, who had spoken to the complainant at the time. The complainant was not called to testify. O testified, claiming that he had never sexually assaulted the boy. The arbitrator ruled that the criminal conviction was admissible evidence, but that it was not conclusive as to whether O had sexually assaulted the boy. No fresh evidence was introduced. The arbitrator held that the presumption raised by the criminal conviction had been rebutted, and that O had been dismissed without just cause. The Divisional Court quashed the arbitrator’s ruling. The Court of Appeal upheld that decision.

Held: The appeal should be dismissed.

Per McLachlin C.J. and Gonthier, Iacobucci, Major, Bastarache, Binnie and Arbour JJ.: When asked to decide whether a criminal conviction, prima facie admissible in a proceeding under s. 22.1 of the Ontario Evidence Act, ought to be rebutted or taken as conclusive, courts will turn to the doctrine of abuse of process to ascertain whether relitigation would be detrimental to the adjudicative process. The doctrine engages the inherent power of the court to prevent the misuse of its procedure, in a way that would bring the administration of justice into disrepute. It has been applied to preclude relitigation in circumstances where the strict requirements of issue estoppel are not met, but where allowing litigation to proceed would nonetheless violate such principles as judicial economy, consistency, finality and the integrity of the administration of justice. The motive of the party who seeks to relitigate, and the capacity in which he or she does so, cannot be decisive factors in the application of the bar against relitigation. What is improper is to attempt to impeach a judicial finding by the impermissible route of relitigation in a different forum. A proper focus on the process, rather than on the interests of a party, will reveal why relitigation should not be permitted. From the system’s point of view, relitigation carries serious detrimental effects and should be avoided unless the circumstances dictate that relitigation is necessary to enhance the credibility and the effectiveness of the adjudicative process as a whole. Casting doubt over the validity of a criminal conviction is a very serious matter. Collateral attacks and relitigation are not appropriate methods of redress since they inordinately tax the adjudicative process while doing nothing to ensure a more trustworthy result. The common law doctrines of issue estoppel, collateral attack and abuse of process adequately capture the concerns that arise when finality in litigation must be balanced against fairness to a particular litigant. There is no need to endorse a self-standing and independent “principle of finality” as either a separate doctrine or as an independent test to preclude relitigation.

The appellant union was not entitled, either at common law or under statute, to relitigate the issue decided against the grievor in the criminal proceedings. The facts in this appeal point to the blatant abuse of process that results when relitigation of this sort is permitted. O was convicted in a criminal court and he exhausted all his avenues of appeal. In law, his conviction must stand, with all its consequent legal effects. There is nothing in this case that militates against the application of the doctrine of abuse of process to bar the relitigation of O’s criminal conviction. The arbitrator was required as a matter of law to give full effect to the conviction. As a result of that error of law, the arbitrator reached a patently unreasonable conclusion. Properly understood in the light of correct legal principles, the evidence before the arbitrator could only lead him to conclude that the respondent City had established just cause for O’s dismissal.

Issue estoppel has no application in this case since the requirement of mutuality of parties has not been met. With respect to the collateral attack doctrine, the appellant does not seek to overturn the sexual abuse conviction itself, but rather contest, for the purposes of a different claim with different legal consequences, whether the conviction was correct.

Per LeBel and Deschamps JJ.: As found by the majority, this case is appropriately decided on the basis of the doctrine of abuse of process, rather than the narrower and more technical doctrines of either collateral attack or issue estoppel. There was also agreement that the appropriate standard of review for the question of whether a criminal conviction may be relitigated in a grievance proceeding is correctness. This is a question of law involving the interpretation of the arbitrator’s constituent statute, an external statute, and a complex body of common law rules and conflicting jurisprudence dealing with relitigation, an issue at the heart of the administration of justice. The arbitrator’s determination in this case that O’s criminal conviction could indeed be relitigated during the grievance proceeding was incorrect. As a matter of law, the arbitrator was required to give full effect to O’s conviction. His failure to do so was sufficient to render his ultimate decision that O had been dismissed without just cause — a decision squarely within the arbitrator’s area of specialized expertise and thus reviewable on a deferential standard — patently unreasonable, according to the jurisprudence of the Court.

Because of growing concerns with the ways in which the standards of review currently available within the pragmatic and functional approach are conceived of and applied, the administrative law aspects of this case require further discussion. The patent unreasonableness standard does not currently provide sufficiently clear parameters for reviewing courts to apply in assessing the decisions of administrative adjudicators. Certain fundamental legal questions — for instance constitutional and human rights questions and those involving civil liberties, as well as other questions that are of central importance to the legal system as a whole, such as the issue of relitigation — typically fall to be decided on the correctness standard. Not all questions of law, however, must be reviewed under a standard of correctness. Resolving general legal questions may be an important component of the work of some administrative adjudicators. In many instances, the appropriate standard of review in respect of the application of general common or civil law rules by specialized adjudicators should not be one of correctness, but rather of reasonableness. If the general question of law is closely connected to the adjudicator’s core area of expertise, the decision will typically be entitled to deference.

In reviewing a decision under the existing standard of patent unreasonableness, the court’s role is not to identify the correct result. To pass a review for patent unreasonableness, a decision must be one that can be rationally supported. It would be wrong for a reviewing court to intervene in decisions that are incorrect, rather than limiting its intervention to those decisions that lack a rational foundation. If this occurs, the line between correctness on the one hand, and patent unreasonableness, on the other, becomes blurred. The boundaries between patent unreasonableness and reasonableness simpliciter are even less clear and approaches to sustain a workable distinction between them raise their own problems. In the end, the essential question remains the same under both standards: was the decision of the adjudicator taken in accordance with reason? In summary, the current framework exhibits several drawbacks. These include the conceptual and practical difficulties that flow from the overlap between patent unreasonableness and reasonableness simpliciter, and the difficulty caused at times by the interplay between patent unreasonableness and correctness.

The role of a court in determining the standard of review is to be faithful to the intent of the legislature that empowered the administrative adjudicator to make the decision, as well as to the animating principle that, in a society governed by the rule of law, power is not to be exercised arbitrarily or capriciously. Judicial review on substantive grounds ensures that the decisions of administrative adjudicators are capable of rational justification; review on procedural grounds ensures that they are fair.

Administrative law has developed considerably over the last 25 years. This evolution, which reflects a strong sense of deference to administrative decision makers and an acknowledgment of the importance of their role, has given rise to some problems or concerns. It remains to be seen, in an appropriate case, what should be the solution to these difficulties. Should courts move to a two standard system of judicial review, correctness and a revised unified standard of reasonableness? Should we attempt to more clearly define the nature and scope of each standard or rethink their relationship and application? This is perhaps some of the work which lies ahead for courts, building on the developments of recent years as well as on the legal tradition which created the framework of the present law of judicial review.

Cases Cited

By Arbour J.

Referred to: Ontario v. O.P.S.E.U., [2003] 3 S.C.R. 149, 2003 SCC 64; Dr. Q v. College of Physicians and Surgeons of British Columbia, [2003] 1 S.C.R. 226, 2003 SCC 19; Law Society of New Brunswick v. Ryan, [2003] 1 S.C.R. 247, 2003 SCC 20; Pushpanathan v. Canada (Minister of Citizenship and Immigration), 1998 CanLII 778 (SCC), [1998] 1 S.C.R. 982; Toronto (City) Board of Education v. O.S.S.T.F., District 15, 1997 CanLII 378 (SCC), [1997] 1 S.C.R. 487; Parry Sound (District) Social Services Administration Board v. O.P.S.E.U., Local 324, [2003] 2 S.C.R. 157, 2003 SCC 42; Demeter v. British Pacific Life Insurance Co. (1983), 1983 CanLII 1838 (ON SC), 150 D.L.R. (3d) 249, aff’d (1984), 1984 CanLII 1996 (ON CA), 48 O.R. (2d) 266; Hunter v. Chief Constable of the West Midlands Police, [1982] A.C. 529, aff’g McIlkenny v. Chief Constable of the West Midlands, [1980] 1 Q.B. 283; Re Del Core and Ontario College of Pharmacists (1985), 1985 CanLII 119 (ON CA), 51 O.R. (2d) 1; Danyluk v. Ainsworth Technologies Inc., [2001] 2 S.C.R. 460, 2001 SCC 44; Parklane Hosiery Co. v. Shore, 439 U.S. 322 (1979); R. v. Regan, [2002] 1 S.C.R. 297, 2002 SCC 12; Lemay v. The King, 1951 CanLII 27 (SCC), [1952] 1 S.C.R. 232; R. v. Banks, [1916] 2 K.B. 621; Wilson v. The Queen, 1983 CanLII 35 (SCC), [1983] 2 S.C.R. 594; R. v. Sarson, 1996 CanLII 200 (SCC), [1996] 2 S.C.R. 223; R. v. Consolidated Maybrun Mines Ltd., 1998 CanLII 820 (SCC), [1998] 1 S.C.R. 706; R. v. Power, 1994 CanLII 126 (SCC), [1994] 1 S.C.R. 601; R. v. Conway, 1989 CanLII 66 (SCC), [1989] 1 S.C.R. 1659; R. v. Scott, 1990 CanLII 27 (SCC), [1990] 3 S.C.R. 979; Blencoe v. British Columbia (Human Rights Commission), [2000] 2 S.C.R. 307, 2000 SCC 44; R. v. O’Connor, 1995 CanLII 51 (SCC), [1995] 4 S.C.R. 411; United States of America v. Shulman, [2001] 1 S.C.R. 616, 2001 SCC 21; Canam Enterprises Inc. v. Coles (2000), 2000 CanLII 8514 (ON CA), 51 O.R. (3d) 481, rev’d [2002] 3 S.C.R. 307, 2002 SCC 63; Franco v. White (2001), 2001 CanLII 24020 (ON CA), 53 O.R. (3d) 391; Bomac Construction Ltd. v. Stevenson, 1986 CanLII 3573 (SK CA), [1986] 5 W.W.R. 21; Bjarnarson v. Government of Manitoba (1987), 1987 CanLII 993 (MB QB), 38 D.L.R. (4th) 32, aff’d (1987), 1987 CanLII 5396 (MB CA), 21 C.P.C. (2d) 302; R. v. McIlkenny (1991), 93 Cr. App. R. 287; United States v. Burns, [2001] 1 S.C.R. 283, 2001 SCC 7; R. v. Bromley (2001), 2001 NFCA 5, 151 C.C.C. (3d) 480; Q. v. Minto Management Ltd. (1984), 1984 CanLII 2118 (ON SC), 46 O.R. (2d) 756; Nigro v. Agnew‑Surpass Shoe Stores Ltd. (1977), 1977 CanLII 3406 (ON SC), 18 O.R. (2d) 215, aff’d (1978), 1976 CanLII 692 (ON CA), 18 O.R. (2d) 714; Germscheid v. Valois (1989), 1989 CanLII 4156 (ON SC), 68 O.R. (2d) 670; Simpson v. Geswein (1995), 1995 CanLII 16110 (MB QB), 25 C.C.L.T. (2d) 49; Roenisch v. Roenisch (1991), 1991 CanLII 6223 (AB QB), 85 D.L.R. (4th) 540; Saskatoon Credit Union, Ltd. v. Central Park Enterprises Ltd. (1988), 1988 CanLII 2941 (BC SC), 47 D.L.R. (4th) 431; Canadian Tire Corp. v. Summers (1995), 1995 CanLII 7183 (ON SC), 23 O.R. (3d) 106.

By LeBel J.

Referred to: Chamberlain v. Surrey School District No. 36, [2002] 4 S.C.R. 710, 2002 SCC 86; Ontario v. O.P.S.E.U., [2003] 3 S.C.R. 149, 2003 SCC 64; C.U.P.E. v. Ontario (Minister of Labour), [2003] 1 S.C.R. 539, 2003 SCC 29; Dr. Q v. College of Physicians and Surgeons of British Columbia, [2003] 1 S.C.R. 226, 2003 SCC 19; Miller v. Workers’ Compensation Commission (Nfld.) (1997), 1997 CanLII 10862 (NL SC), 154 Nfld. & P.E.I.R. 52; Toronto (City) Board of Education v. O.S.S.T.F., District 15, 1997 CanLII 378 (SCC), [1997] 1 S.C.R. 487; Canada (Attorney General) v. Public Service Alliance of Canada, 1993 CanLII 125 (SCC), [1993] 1 S.C.R. 941; Ivanhoe inc. v. UFCW, Local 500, [2001] 2 S.C.R. 565, 2001 SCC 47; Canadian Broadcasting Corp. v. Canada (Labour Relations Board), 1995 CanLII 148 (SCC), [1995] 1 S.C.R. 157; Pushpanathan v. Canada (Minister of Citizenship and Immigration), 1998 CanLII 778 (SCC), [1998] 1 S.C.R. 982; Canada (Director of Investigation and Research) v. Southam Inc., 1997 CanLII 385 (SCC), [1997] 1 S.C.R. 748; Pezim v. British Columbia (Superintendent of Brokers), 1994 CanLII 103 (SCC), [1994] 2 S.C.R. 557; National Corn Growers Assn. v. Canada (Import Tribunal), 1990 CanLII 49 (SCC), [1990] 2 S.C.R. 1324; Canada (Attorney General) v. Mossop, 1993 CanLII 164 (SCC), [1993] 1 S.C.R. 554; Pasiechnyk v. Saskatchewan (Workers’ Compensation Board), 1997 CanLII 316 (SCC), [1997] 2 S.C.R. 890; Macdonell v. Quebec (Commission d’accès à l’information), [2002] 3 S.C.R. 661, 2002 SCC 71; Canadian Union of Public Employees, Local 963 v. New Brunswick Liquor Corp., 1979 CanLII 23 (SCC), [1979] 2 S.C.R. 227; Law Society of New Brunswick v. Ryan, [2003] 1 S.C.R. 247, 2003 SCC 20; Service Employees’ International Union, Local No. 333 v. Nipawin District Staff Nurses Association, 1973 CanLII 191 (SCC), [1975] 1 S.C.R. 382; Anisminic Ltd. v. Foreign Compensation Commission, [1969] 2 A.C. 147; Metropolitan Life Insurance Co. v. International Union of Operating Engineers, Local 796, 1970 CanLII 7 (SCC), [1970] S.C.R. 425; CAIMAW v. Paccar of Canada Ltd., 1989 CanLII 49 (SCC), [1989] 2 S.C.R. 983; Canadian Union of Public Employees, Local 301 v. Montreal (City), 1997 CanLII 386 (SCC), [1997] 1 S.C.R. 793; Domtar Inc. v. Quebec (Commission d’appel en matière de lésions professionnelles), 1993 CanLII 106 (SCC), [1993] 2 S.C.R. 756; Canada Safeway Ltd. v. RWDSU, Local 454, 1998 CanLII 780 (SCC), [1998] 1 S.C.R. 1079; Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, 1990 CanLII 22 (SCC), [1990] 3 S.C.R. 644; Hao v. Canada (Minister of Citizenship and Immigration) (2000), 2000 CanLII 15150 (FC), 184 F.T.R. 246; United Brotherhood of Carpenters and Joiners of America, Local 579 v. Bradco Construction Ltd., 1993 CanLII 88 (SCC), [1993] 2 S.C.R. 316; Baker v. Canada (Minister of Citizenship and Immigration), 1999 CanLII 699 (SCC), [1999] 2 S.C.R. 817; Reference re Resolution to Amend the Constitution, 1981 CanLII 25 (SCC), [1981] 1 S.C.R. 753; Reference re Secession of Quebec, 1998 CanLII 793 (SCC), [1998] 2 S.C.R. 217.

Statutes and Regulations Cited

Canadian Charter of Rights and Freedoms.

Evidence Act, R.S.O. 1990, c. E.23, s. 22.1 [ad. 1995, c. 6, s. 6(3)].

Labour Relations Act, 1995, S.O. 1995, c. 1, Sch. A, s. 48(1).

Authors Cited

Allars, Margaret. “On Deference to Tribunals, With Deference to Dworkin” (1994), 20 Queen’s L.J. 163.

Comtois, Suzanne. Vers la primauté de l’approche pragmatique et fonctionnelle: Précis du contrôle judiciaire des décisions de fond rendues par les organismes administratifs. Cowansville, Qué.: Yvon Blais, 2003.

Cowan, Jeff G. “The Standard of Review: The Common Sense Evolution?”, paper presented to the Administrative Law Section Meeting, Ontario Bar Association, January 21, 2003.

Dyzenhaus, David. “The Politics of Deference: Judicial Review and Democracy”, in Michael Taggart, ed., The Province of Administrative Law. Oxford: Hart Publishing, 1997, 279.

Elliott, David W. “Suresh and the Common Borders of Administrative Law: Time for the Tailor?” (2002), 65 Sask. L. Rev. 469.

Evans, J. M., et al. Administrative Law: Cases, Text, and Materials, 3rd ed. Toronto: Emond Montgomery, 1989.

Falzon, Frank A. V. “Standard of Review on Judicial Review or Appeal”, in Administrative Justice Review Background Papers: Background Papers prepared by Administrative Justice Project for the Attorney General of British Columbia, June 2002.

Garant, Patrice. Droit administratif, vol. 2, 4e éd. Cowansville, Qué.: Yvon Blais, 1996.

Holloway, Ian. “‘A Sacred Right’: Judicial Review of Administrative Action as a Cultural Phenomenon” (1993), 22 Man. L.J. 28.

Howard, M. N., Peter Crane and Daniel A. Hochberg. Phipson on Evidence, 14th ed. London: Sweet & Maxwell, 1990.

Jones, David Phillip. “Notes on Dr. Q and Ryan: Two More Decisions by the Supreme Court of Canada on the Standard of Review in Administrative Law”, paper presented at the Canadian Institute for the Administration of Justice, Western Roundtable, Edmonton, April 25, 2003.

Lange, Donald J. The Doctrine of Res Judicata in Canada. Markham, Ont.: Butterworths, 2000.

Law Society of Upper Canada. Rules of Professional Conduct. Toronto: Law Society of Upper Canada, 2000.

Lovett, Deborah K. “That Curious Curial Deference Just Gets Curiouser and Curiouser — Canada (Director of Investigation and Research) v. Southam Inc.” (1997), 55 Advocate (B.C.) 541.

MacLauchlan, H. Wade. “Transforming Administrative Law: The Didactic Role of the Supreme Court of Canada” (2001), 80 Can. Bar Rev. 281.

McLachlin, Beverley. “The Roles of Administrative Tribunals and Courts in Maintaining the Rule of Law” (1998-1999), 12 C.J.A.L.P. 171.

Mullan, David J. Administrative Law. Toronto: Irwin Law, 2001.

Mullan, David J. “Of Chaff Midst the Corn: American Farm Bureau Federation v. Canada (Canadian Import Tribunal) and Patent Unreasonableness Review” (1991), 45 Admin. L.R. 264.

Mullan, David J. “Recent Developments in Standard of Review”, in Taking the Tribunal to Court: A Practical Guide for Administrative Law Practitioners. Canadian Bar Association (Ontario), October 20, 2000.

Paciocco, David M., and Lee Stuesser. The Law of Evidence, 3rd ed. Toronto: Irwin Law, 2002.

Perell, Paul M. “Res Judicata and Abuse of Process” (2001), 24 Advocates’ Q. 189.

Perrault, Gabrielle. Le contrôle judiciaire des décisions de l’administration: De l’erreur juridictionnelle à la norme de contrôle. Montréal: Wilson & Lafleur, 2002.

Sossin, Lorne. “Developments in Administrative Law: The 1997-98 and 1998-99 Terms” (2000), 11 S.C.L.R. (2d) 37.

Sprague, James L. H. “Another View of Baker” (1999), 7 Reid’s Administrative Law 163.

Sullivan, Ruth. Sullivan and Driedger on the Construction of Statutes, 4th ed. Markham, Ont.: Butterworths, 2002.

Teplitsky, Martin. “Prior Criminal Convictions: Are They Conclusive Proof? An Arbitrator’s Perspective”, in K. Whitaker et al., eds., Labour Arbitration Yearbook 2001-2002, vol. I. Toronto: Lancaster House, 2002, 279.

Watson, Garry D. “Duplicative Litigation: Issue Estoppel, Abuse of Process and the Death of Mutuality” (1990), 69 Can. Bar Rev. 623.

APPEAL from a judgment of the Ontario Court of Appeal (2001), 2001 CanLII 24114 (ON CA), 55 O.R. (3d) 541, 205 D.L.R. (4th) 280, 149 O.A.C. 213, 45 C.R. (5th) 354, 37 Admin. L.R. (3d) 40, 2002 CLLC ¶220-014, [2001] O.J. No. 3239 (QL), affirming a judgment of the Divisional Court (2000), 2000 CanLII 30193 (ON SCDC), 187 D.L.R. (4th) 323, 134 O.A.C. 48, 23 Admin. L.R. (3d) 72, 2000 CLLC ¶220-038, [2000] O.J. No. 1570 (QL). Appeal dismissed.

Douglas J. Wray and Harold F. Caley, for the appellant.

Jason Hanson, Mahmud Jamal and Kari M. Abrams, for the respondent the City of Toronto.

No one appeared for the respondent Douglas C. Stanley.

Sean Kearney, Mary Gersht and Meredith Brown, for the intervener.

The judgment of McLachlin C.J. and Gonthier, Iacobucci, Major, Bastarache, Binnie and Arbour JJ. was delivered by

Arbour J. —

I. Introduction

[1] Can a person convicted of sexual assault, and dismissed from his employment as a result, be reinstated by a labour arbitrator who concludes, on the evidence before him, that the sexual assault did not take place? This is essentially the issue raised in this appeal.

[2] Like the Court of Appeal for Ontario and the Divisional Court, I have come to the conclusion that the arbitrator may not revisit the criminal conviction. Although my reasons differ somewhat from those of the courts below, I would dismiss the appeal.

II. Facts

[3] Glenn Oliver worked as a recreation instructor for the respondent City of Toronto. He was charged with sexually assaulting a boy under his supervision. He pleaded not guilty. At trial before a judge alone, he testified and was cross‑examined. He called several defence witnesses, including character witnesses. The trial judge found that the complainant was credible and that Oliver was not. He entered a conviction, which was later affirmed on appeal. He sentenced Oliver to 15 months in jail, followed by one year of probation.

[4] The respondent City of Toronto fired Oliver a few days after his conviction, and Oliver grieved his dismissal. At the hearing, the City of Toronto submitted the boy’s testimony from the criminal trial and the notes of Oliver’s supervisor, who had spoken to the boy at the time. The City did not call the boy to testify. Oliver again testified on his own behalf and claimed that he had never sexually assaulted the boy.

[5] The arbitrator ruled that the criminal conviction was admissible as prima facie but not conclusive evidence that Oliver had sexually assaulted the boy. No evidence of fraud nor any fresh evidence unavailable at trial was introduced in the arbitration. The arbitrator held that the presumption raised by the criminal conviction had been rebutted, and that Oliver had been dismissed without just cause.

III. Procedural History

A. Superior Court of Justice (Divisional Court) (2000), [2000 CanLII 30193 (ON SCDC)](https://www.canlii.org/en/on/onscdc/doc/2000/2000canlii30193/2000canlii30193.html), 187 D.L.R. (4th) 323

[6] At Divisional Court the application for judicial review was granted and the decision of the arbitrator was quashed. The Divisional Court heard this case and Ontario v. O.P.S.E.U. at the same time. (Ontario v. O.P.S.E.U., [2003] 3 S.C.R. 149, 2003 SCC 64, is being released concurrently by this Court.) O’Driscoll J. found that while s. 22.1 of the Evidence Act, R.S.O. 1990, c. E.23, applied to all the arbitrations, relitigation of the cases was barred by the doctrines of collateral attack, issue estoppel and abuse of process. The court noted that criminal convictions are valid judgments that cannot be collaterally attacked at a later arbitration (paras. 74-79). With respect to issue estoppel, under which an issue decided against a party is protected from collateral attack barring decisive new evidence or a showing of fraud, the court found that relitigation was also prevented, rejecting the appellant’s argument that there had been no privity because the union, and not the grievor, had filed the grievance. The court also held that the doctrine of abuse of process, which denies a collateral attack upon a final decision of another court where the party had “a full opportunity of contesting the decision”, applied (paras. 81 and 90). Finally, O’Driscoll J. found that whether the standard of review was correctness or patent unreasonableness in each case, the standard for judicial review had been met (para. 86).

B. Court of Appeal for Ontario (2001), [2001 CanLII 24114 (ON CA)](https://www.canlii.org/en/on/onca/doc/2001/2001canlii24114/2001canlii24114.html), 55 O.R. (3d) 541

[7] Doherty J.A. for the court held that because the crux of the issue was whether the Canadian Union of Public Employees (CUPE or the union) was permitted to relitigate the issue decided in the criminal trial, and because this analysis “turned on [the arbitrator’s] understanding of the common law rules and principles governing the relitigation of issues finally decided in a previous judicial proceeding”, the appropriate standard of review was correctness (paras. 22 and 38).

[8] Doherty J.A. concluded that issue estoppel did not apply. Even if the union was the employee’s privy, the respondent City of Toronto had played no role in the criminal proceeding and had no relationship to the Crown. He also found that describing the appellant union’s attempt to relitigate the employee’s culpability as a collateral attack on the order of the court did not assist in determining whether relitigation could be permitted. Commenting that the phrase “abuse of process” was perhaps best limited to describe those cases where the plaintiff has instigated litigation for some improper purpose, Doherty J.A. went on to consider what he called “the finality principle” in considerable depth.

[9] Doherty J.A. dismissed the appeal on the basis of this principle. He held that the res judicata jurisprudence required a court to balance the importance of finality, which reduces uncertainty and inconsistency in results, and which serves to conserve the resources of both the parties and the judiciary, with the “search for justice in each individual case” (para. 94). Doherty J.A. held that the following approach should be taken when weighing finality claims against an individual litigant’s claim to access to justice (at para. 100):

• Does the res judicata doctrine apply?

• If the doctrine applies, can the party against whom it applies demonstrate that the justice of the individual case should trump finality concerns?

• If the doctrine does not apply, can the party seeking to preclude relitigation demonstrate that finality concerns should be given paramountcy over the claim that justice requires relitigation?

[10] Ultimately, Doherty J.A. dismissed the appeal, concluding that “finality concerns must be given paramountcy over CUPE’s claim to an entitlement to relitigate Oliver’s culpability” (para. 102). He so concluded because there was no suggestion of fraud at the criminal trial, because the underlying charges were serious enough that the employee was likely to have litigated them to the fullest, and because there was no new evidence presented at arbitration (paras. 103-108).

IV. Relevant Statutory Provisions

[11] Evidence Act, R.S.O. 1990, c. E.23

22.1 (1) Proof that a person has been convicted or discharged anywhere in Canada of a crime is proof, in the absence of evidence to the contrary, that the crime was committed by the person, if,

(a) no appeal of the conviction or discharge was taken and the time for an appeal has expired; or

(b) an appeal of the conviction or discharge was taken but was dismissed or abandoned and no further appeal is available.

(2) Subsection (1) applies whether or not the convicted or discharged person is a party to the proceeding.

(3) For the purposes of subsection (1), a certificate containing the substance and effect only, omitting the formal part, of the charge and of the conviction or discharge, purporting to be signed by the officer having the custody of the records of the court at which the offender was convicted or discharged, or by the deputy of the officer, is, on proof of the identity of the person named as convicted or discharged person in the certificate, sufficient evidence of the conviction or discharge of that person, without proof of the signature or of the official character of the person appearing to have signed the certificate.

Labour Relations Act, 1995, S.O. 1995, c. 1, Sch. A

48. (1) Every collective agreement shall provide for the final and binding settlement by arbitration, without stoppage of work, of all differences between the parties arising from the interpretation, application, administration or alleged violation of the agreement, including any question as to whether a matter is arbitrable.

V. Analysis

A. Standard of Review

[12] My colleague LeBel J. discusses at length our jurisprudence on standards of review. He reviews concerns and criticisms about the three standard system of judicial review. Given that these issues were not argued before us in this case, and without the benefit of a full adversarial debate, I would not wish to comment on the desirability of a departure from our recently affirmed framework for standards of review analysis. (See this Court’s unanimous decisions of Dr. Q v. College of Physicians and Surgeons of British Columbia, [2003] 1 S.C.R. 226, 2003 SCC 19, and Law Society of New Brunswick v. Ryan, [2003] 1 S.C.R. 247, 2003 SCC 20.)

[13] The Court of Appeal properly applied the functional and pragmatic approach as delineated in Pushpanathan v. Canada (Minister of Citizenship and Immigration), 1998 CanLII 778 (SCC), [1998] 1 S.C.R. 982 (see also Dr. Q, supra), to determine the extent to which the legislature intended that courts should review the tribunals’ decisions.

[14] Doherty J.A. was correct to acknowledge patent unreasonableness as the general standard of review of an arbitrator’s decision as to whether just cause has been established in the discharge of an employee. However, and as he noted, the same standard of review does not necessarily apply to every ruling made by the arbitrator in the course of the arbitration. This follows the distinction drawn by Cory J. for the majority in Toronto (City) Board of Education v. O.S.S.T.F., District 15, 1997 CanLII 378 (SCC), [1997] 1 S.C.R. 487, where he said, at para. 39:

It has been held on several occasions that the expert skill and knowledge which an arbitration board exercises in interpreting a collective agreement does not usually extend to the interpretation of “outside” legislation. The findings of a board pertaining to the interpretation of a statute or the common law are generally reviewable on a correctness standard. . . . An exception to this rule may occur where the external statute is intimately connected with the mandate of the tribunal and is encountered frequently as a result. [Emphasis added.]

[15] In this case, the reasonableness of the arbitrator’s decision to reinstate the grievor is predicated on the correctness of his assumption that he was not bound by the criminal conviction. That assumption rested on his analysis of complex common law rules and of conflicting jurisprudence. The body of law dealing with the relitigation of issues finally decided in previous judicial proceedings is not only complex; it is also at the heart of the administration of justice. Properly understood and applied, the doctrines of res judicata and abuse of process govern the interplay between different judicial decision makers. These rules and principles call for a judicial balance between finality, fairness, efficiency and authority of judicial decisions. The application of these rules, doctrines and principles is clearly outside the sphere of expertise of a labour arbitrator who may be called to have recourse to them. In such a case, he or she must correctly answer the question of law raised. An incorrect approach may be sufficient to lead to a patently unreasonable outcome. This was reiterated recently by Iacobucci J. in Parry Sound (District) Social Services Administration Board v. O.P.S.E.U., Local 324, [2003] 2 S.C.R. 157, 2003 SCC 42, at para. 21.

[16] Therefore I agree with the Court of Appeal that the arbitrator had to decide correctly whether CUPE was entitled, either at common law or under a statute, to relitigate the issue decided against the grievor in the criminal proceedings.

B. [Section 22.1](https://www.canlii.org/en/on/laws/stat/rso-1990-c-e23/latest/rso-1990-c-e23.html#sec22.1_smooth) of [Ontario’s Evidence Act](https://www.canlii.org/en/on/laws/stat/rso-1990-c-e23/latest/rso-1990-c-e23.html)

[17] Section 22.1 of the Ontario Evidence Act is of limited assistance to the disposition of this appeal. It provides that proof that a person has been convicted of a crime is proof, “in the absence of evidence to the contrary”, that the crime was committed by that person.

[18] As Doherty J.A. correctly pointed out, at para. 42, s. 22.1 contemplates that the validity of a conviction may be challenged in a subsequent proceeding, but the section says nothing about the circumstances in which such challenge is or is not permissible. That issue is determined by the application of such common law doctrines as res judicata, issue estoppel, collateral attack and abuse of process. Section 22.1 speaks of the admissibility of the fact of the conviction as proof of the truth of its content, and speaks of its conclusive effect if unchallenged. As a rule of evidence, the section addresses in part the hearsay rule, by making the conviction — the finding of another court — admissible for the truth of its content, as an exception to the inadmissibility of hearsay (D. M. Paciocco and L. Stuesser, The Law of Evidence (3rd ed. 2002), at p. 120; Phipson on Evidence (14th ed. 1990), at paras. 33-94 and 33-95).

[19] Here, however, the admissibility of the conviction is not in issue. Section 22.1 renders the proof of the conviction admissible. The question is whether it can be rebutted by “evidence to the contrary”. There are circumstances in which evidence will be admissible to rebut the presumption that the person convicted committed the crime, in particular where the conviction in issue is that of a non-party. There are also circumstances in which no such evidence may be tendered. If either issue estoppel or abuse of process bars the relitigation of the facts essential to the conviction, then no “evidence to the contrary” may be tendered to displace the effect of the conviction. In such a case, the conviction is conclusive that the person convicted committed the crime.

[20] This interpretation is consistent with the rule of interpretation that legislation is presumed not to depart from general principles of law without an express indication to that effect. This presumption was reviewed and applied by Iacobucci J. in Parry Sound, supra, at para 39. Section 22.1 reflected the law established in the leading Canadian case of Demeter v. British Pacific Life Insurance Co. (1983), 1983 CanLII 1838 (ON SC), 150 D.L.R. (3d) 249 (Ont. H.C.), at p. 264, aff’d (1984), 1984 CanLII 1996 (ON CA), 48 O.R. (2d) 266 (C.A.), wherein after a thorough review of Canadian and English jurisprudence, Osler J. held that a criminal conviction is admissible in subsequent civil litigation as prima facie proof that the convicted individual committed the alleged act, “subject to rebuttal by the plaintiff on the merits”. However, the common law also recognized that the presumption of guilt established by a conviction is rebuttable only where the rebuttal does not constitute an abuse of the process of the court (Demeter (H.C.), supra, at p. 265; Hunter v. Chief Constable of the West Midlands Police, [1982] A.C. 529 (H.L.), at p. 541; see also Re Del Core and Ontario College of Pharmacists (1985), 1985 CanLII 119 (ON CA), 51 O.R. (2d) 1 (C.A.), at p. 22, per Blair J.A.). Section 22.1 does not change this; the legislature has not explicitly displaced the common law doctrines and the rebuttal is consequently subject to them.

[21] The question therefore is whether any doctrine precludes in this case the relitigation of the facts upon which the conviction rests.

C. The Common Law Doctrines

[22] Much consideration was given in the decisions below to the three related common law doctrines of issue estoppel, abuse of process and collateral attack. Each of these doctrines was considered as a possible means of preventing the union from relitigating the criminal conviction of the grievor before the arbitrator. Although both the Divisional Court and the Court of Appeal concluded that the union could not relitigate the guilt of the grievor as reflected in his criminal conviction, they took different views of the applicability of the different doctrines advanced in support of that conclusion. While the Divisional Court concluded that relitigation was barred by the collateral attack rule, issue estoppel and abuse of process, the Court of Appeal was of the view that none of these doctrines as they presently stand applied to bar the rebuttal. Rather, it relied on a self-standing “finality principle”. I think it is useful to disentangle these various rules and doctrines before turning to the applicable one here. I stress at the outset that these common law doctrines are interrelated and in many cases more than one doctrine may support a particular outcome. Even though both issue estoppel and collateral attacks may properly be viewed as particular applications of a broader doctrine of abuse of process, the three are not always entirely interchangeable.

(1) Issue Estoppel

[23] Issue estoppel is a branch of res judicata (the other branch being cause of action estoppel), which precludes the relitigation of issues previously decided in court in another proceeding. For issue estoppel to be successfully invoked, three preconditions must be met: (1) the issue must be the same as the one decided in the prior decision; (2) the prior judicial decision must have been final; and (3) the parties to both proceedings must be the same, or their privies (Danyluk v. Ainsworth Technologies Inc., [2001] 2 S.C.R. 460, 2001 SCC 44, at para. 25, per Binnie J.). The final requirement, known as “mutuality”, has been largely abandoned in the United States and has been the subject of much academic and judicial debate there as well as in the United Kingdom and, to some extent, in this country. (See G. D. Watson, “Duplicative Litigation: Issue Estoppel, Abuse of Process and the Death of Mutuality” (1990), 69 Can. Bar Rev. 623, at pp. 648‑51.) In light of the different conclusions reached by the courts below on the applicability of issue estoppel, I think it is useful to examine that debate more closely.

[24] The first two requirements of issue estoppel are met in this case. The final requirement of mutuality of parties has not been met. In the original criminal case, the lis was between Her Majesty the Queen in right of Canada and Glenn Oliver. In the arbitration, the parties were CUPE and the City of Toronto, Oliver’s employer. It is unnecessary to decide whether Oliver and CUPE should reasonably be viewed as privies for the purpose of the application of the mutuality requirement since it is clear that the Crown, acting as prosecutor in the criminal case, is not privy with the City of Toronto, nor would it be with a provincial, rather than a municipal, employer (as in the Ontario v. O.P.S.E.U. case, released concurrently).

[25] There has been much academic criticism of the mutuality requirement of the doctrine of issue estoppel. In his article, Professor Watson, supra, argues that explicitly abolishing the mutuality requirement, as has been done in the United States, would both reduce confusion in the law and remove the possibility that a strict application of issue estoppel may work an injustice. The arguments made by him and others (see also D. J. Lange, The Doctrine of Res Judicata in Canada (2000)), urging Canadian courts to abandon the mutuality requirement have been helpful in articulating a principled approach to the bar against relitigation. In my view, however, appropriate guidance is available in our law without the modification to the mutuality requirement that this case would necessitate.

[26] In his very useful review of the abandonment of the mutuality requirement in the United States, Professor Watson, at p. 631, points out that mutuality was first relaxed when issue estoppel was used defensively:

The defensive use of non-mutual issue estoppel is straight forward. If P, having litigated an issue with D1 and lost, subsequently sues D2 raising the same issue, D2 can rely defensively on the issue estoppel arising from the former action, unless the first action did not provide a full and fair opportunity to litigate or other factors make it unfair or unwise to permit preclusion. The rationale is that P should not be allowed to relitigate an issue already lost by simply changing defendants . . . .

[27] Professor Watson then exposes the additional difficulties that arise if the mutuality requirement is removed when issue estoppel is raised offensively, as was done by the United States Supreme Court in Parklane Hosiery Co. v. Shore, 439 U.S. 322 (1979). He describes the offensive use of non mutual issue estoppel as follows (at p. 631):

The power of this offensive non-mutual issue estoppel doctrine is illustrated by single event disaster cases, such as an airline crash. Assume P1 sues Airline for negligence in the operation of the aircraft and in that action Airline is found to have been negligent. Offensive non-mutual issue estoppel permits P2 through P20, etc., now to sue Airline and successfully plead issue estoppel on the question of the airline’s negligence. The rationale is that if Airline fully and fairly litigated the issue of its negligence in action #1 it has had its day in court; it has had due process and it should not be permitted to re-litigate the negligence issue. However, the court in Parklane realized that in order to ensure fairness in the operation of offensive non-mutual issue estoppel the doctrine has to be subject to qualifications.

[28] Properly understood, our case could be viewed as falling under this second category — what would be described in U.S. law as “non-mutual offensive preclusion”. Although technically speaking the City of Toronto is not the “plaintiff” in the arbitration proceedings, the City wishes to take advantage of the conviction obtained by the Crown against Oliver in a different, prior proceeding to which the City was not a party. It wishes to preclude Oliver from relitigating an issue that he fought and lost in the criminal forum. U.S. law acknowledges the peculiar difficulties with offensive use of non-mutual estoppel. Professor Watson explains, at pp. 632-33:

First, the court acknowledged that the effects of non-mutuality differ depending on whether issue estoppel is used offensively or defensively. While defensive preclusion helps to reduce litigation offensive preclusion, by contrast, encourages potential plaintiffs not to join in the first action. “Since a plaintiff will be able to rely on a previous judgment against a defendant but will not be bound by that judgment if the defendant wins, the plaintiff has every incentive to adopt a ‘wait and see’ attitude, in the hope that the first action by another plaintiff will result in a favorable judgment”. Thus, without some limit, non-mutual offensive preclusion would increase rather than decrease the total amount of litigation. To meet this problem the Parklane court held that preclusion should be denied in action #2 “where a plaintiff could easily have joined in the earlier action”.

Second, the court recognized that in some circumstances to permit non-mutual preclusion “would be unfair to the defendant” and the court referred to specific situations of unfairness: (a) the defendant may have had little incentive to defend vigorously the first action, that is, if she was sued for small or nominal damages, particularly if future suits were not foreseeable; (b) offensive preclusion may be unfair if the judgment relied upon as a basis for estoppel is itself inconsistent with one or more previous judgments in favour of the defendant; or (c) the second action affords to the defendant procedural opportunities unavailable in the first action that could readily result in a different outcome, that is, where the defendant in the first action was forced to defend in an inconvenient forum and was unable to call witnesses, or where in the first action much more limited discovery was available to the defendant than in the second action.

In the final analysis the court declared that the general rule should be that in cases where a plaintiff could easily have joined in the earlier action or where, either for the reasons discussed or for other reasons, the application of offensive estoppel would be unfair to the defendant, a trial judge should not allow the use of offensive collateral estoppel.

[29] It is clear from the above that American non-mutual issue estoppel is not a mechanical, self-applying rule as evidenced by the discretionary elements which may militate against granting the estoppel. What emerges from the American experience with the abandonment of mutuality is a twofold concern: (1) the application of the estoppel must be sufficiently principled and predictable to promote efficiency; and (2) it must contain sufficient flexibility to prevent unfairness. In my view, this is what the doctrine of abuse of process offers, particularly, as here, where the issue involves a conviction in a criminal court for a serious crime. In a case such as this one, the true concerns are not primarily related to mutuality. The true concerns, well reflected in the reasons of the Court of Appeal, are with the integrity and the coherence of the administration of justice. This will often be the case when the estoppel originates from a finding made in a criminal case where many of the traditional concerns related to mutuality lose their significance.

[30] For example, there is little relevance to the concern about the “wait and see” plaintiff, the “free rider” who will deliberately avoid the risk of joining the original litigation, but will later come forward to reap the benefits of the victory obtained by the party who should have been his co-plaintiff. No such concern can ever arise when the original action is in a criminal prosecution. Victims cannot, even if they wanted to, “join in” the prosecution so as to have their civil claim against the accused disposed of in a single trial. Nor can employers “join in” the criminal prosecution to have their employee dismissed for cause.

[31] On the other hand, even though no one can join the prosecution, the prosecutor as a party represents the public interest. He or she represents a collective interest in the just and correct outcome of the case. The prosecutor is said to be a minister of justice who has nothing to win or lose from the outcome of the case but who must ensure that a just and true verdict is rendered. (See Law Society of Upper Canada, Rules of Professional Conduct (2000), Commentary Rule 4.01(3), at p. 61; R. v. Regan, [2002] 1 S.C.R. 297, 2002 SCC 12; Lemay v. The King, 1951 CanLII 27 (SCC), [1952] 1 S.C.R. 232, at pp. 256-57, per Cartwright J.; and R. v. Banks, [1916] 2 K.B. 621 (C.C.A.), at p. 623.) The mutuality requirement of the doctrine of issue estoppel, which insists that only the Crown and its privies be precluded from relitigating the guilt of the accused, is hardly reflective of the true role of the prosecutor.

[32] As the present case illustrates, the primary concerns here are about the integrity of the criminal process and the increased authority of a criminal verdict, rather than some of the more traditional issue estoppel concerns that focus on the interests of the parties, such as costs and multiple “vexation”. For these reasons, I see no need to reverse or relax the long-standing application of the mutuality requirement in this case and I would conclude that issue estoppel has no application. I now turn to the question of whether the decision of the arbitrator amounted to a collateral attack on the verdict of the criminal court.

(2) Collateral Attack

[33] The rule against collateral attack bars actions to overturn convictions when those actions take place in the wrong forum. As stated in Wilson v. The Queen, 1983 CanLII 35 (SCC), [1983] 2 S.C.R. 594, at p. 599, the rule against collateral attack

has long been a fundamental rule that a court order, made by a court having jurisdiction to make it, stands and is binding and conclusive unless it is set aside on appeal or lawfully quashed. It is also well settled in the authorities that such an order may not be attacked collaterally — and a collateral attack may be described as an attack made in proceedings other than those whose specific object is the reversal, variation, or nullification of the order or judgment.

Thus, in Wilson, supra, the Court held that an inferior court judge was without jurisdiction to pass on the validity of a wiretap authorized by a superior court. Other cases that form the basis for this rule similarly involve attempts to overturn decisions in other fora, and not simply to relitigate their facts. In R. v. Sarson, 1996 CanLII 200 (SCC), [1996] 2 S.C.R. 223, at para. 35, this Court held that a prisoner’s habeas corpus attack on a conviction under a law later declared unconstitutional must fail under the rule against collateral attack because the prisoner was no longer “in the system” and because he was “in custody pursuant to the judgment of a court of competent jurisdiction”. Similarly, in R. v. Consolidated Maybrun Mines Ltd., 1998 CanLII 820 (SCC), [1998] 1 S.C.R. 706, this Court held that a mine owner who had chosen to ignore an administrative appeals process for a pollution fine was barred from contesting the validity of that fine in court because the legislation directed appeals to an appellate administrative body, not to the courts. Binnie J. described the rule against collateral attack in Danyluk, supra, at para. 20, as follows: “that a judicial order pronounced by a court of competent jurisdiction should not be brought into question in subsequent proceedings except those provided by law for the express purpose of attacking it” (emphasis added).

[34] Each of these cases concerns the appropriate forum for collateral attacks upon the judgment itself. However, in the case at bar, the union does not seek to overturn the sexual abuse conviction itself, but simply contest, for the purposes of a different claim with different legal consequences, whether the conviction was correct. It is an implicit attack on the correctness of the factual basis of the decision, not a contest about whether that decision has legal force, as clearly it does. Prohibited “collateral attacks” are abuses of the court’s process. However, in light of the focus of the collateral attack rule on attacking the order itself and its legal effect, I believe that the better approach here is to go directly to the doctrine of abuse of process.

(3) Abuse of Process

[35] Judges have an inherent and residual discretion to prevent an abuse of the court’s process. This concept of abuse of process was described at common law as proceedings “unfair to the point that they are contrary to the interest of justice” (R. v. Power, 1994 CanLII 126 (SCC), [1994] 1 S.C.R. 601, at p. 616), and as “oppressive treatment” (R. v. Conway, 1989 CanLII 66 (SCC), [1989] 1 S.C.R. 1659, at p. 1667). McLachlin J. (as she then was) expressed it this way in R. v. Scott, 1990 CanLII 27 (SCC), [1990] 3 S.C.R. 979, at p. 1007:

. . . abuse of process may be established where: (1) the proceedings are oppressive or vexatious; and, (2) violate the fundamental principles of justice underlying the community’s sense of fair play and decency. The concepts of oppressiveness and vexatiousness underline the interest of the accused in a fair trial. But the doctrine evokes as well the public interest in a fair and just trial process and the proper administration of justice.

[36] The doctrine of abuse of process is used in a variety of legal contexts. The unfair or oppressive treatment of an accused may disentitle the Crown to carry on with the prosecution of a charge: Conway, supra, at p. 1667. In Blencoe v. British Columbia (Human Rights Commission), [2000] 2 S.C.R. 307, 2000 SCC 44, this Court held that unreasonable delay causing serious prejudice could amount to an abuse of process. When the Canadian Charter of Rights and Freedoms applies, the common law doctrine of abuse of process is subsumed into the principles of the Charter such that there is often overlap between abuse of process and constitutional remedies (R. v. O’Connor, 1995 CanLII 51 (SCC), [1995] 4 S.C.R. 411). The doctrine nonetheless continues to have application as a non-Charter remedy: United States of America v. Shulman, [2001] 1 S.C.R. 616, 2001 SCC 21, at para. 33.

[37] In the context that interests us here, the doctrine of abuse of process engages “the inherent power of the court to prevent the misuse of its procedure, in a way that would . . . bring the administration of justice into disrepute” (Canam Enterprises Inc. v. Coles (2000), 2000 CanLII 8514 (ON CA), 51 O.R. (3d) 481 (C.A.), at para. 55, per Goudge J.A., dissenting (approved [2002] 3 S.C.R. 307, 2002 SCC 63)). Goudge J.A. expanded on that concept in the following terms at paras. 55-56:

The doctrine of abuse of process engages the inherent power of the court to prevent the misuse of its procedure, in a way that would be manifestly unfair to a party to the litigation before it or would in some other way bring the administration of justice into disrepute. It is a flexible doctrine unencumbered by the specific requirements of concepts such as issue estoppel. See House of Spring Gardens Ltd. v. Waite, [1990] 3 W.L.R. 347 at p. 358, [1990] 2 All E.R. 990 (C.A.).

One circumstance in which abuse of process has been applied is where the litigation before the court is found to be in essence an attempt to relitigate a claim which the court has already determined. [Emphasis added.]

As Goudge J.A.’s comments indicate, Canadian courts have applied the doctrine of abuse of process to preclude relitigation in circumstances where the strict requirements of issue estoppel (typically the privity/mutuality requirements) are not met, but where allowing the litigation to proceed would nonetheless violate such principles as judicial economy, consistency, finality and the integrity of the administration of justice. (See, for example, Franco v. White (2001), 2001 CanLII 24020 (ON CA), 53 O.R. (3d) 391 (C.A.); Bomac Construction Ltd. v. Stevenson, 1986 CanLII 3573 (SK CA), [1986] 5 W.W.R. 21 (Sask. C.A.); and Bjarnarson v. Government of Manitoba (1987), 1987 CanLII 993 (MB QB), 38 D.L.R. (4th) 32 (Man. Q.B.), aff’d (1987), 1987 CanLII 5396 (MB CA), 21 C.P.C. (2d) 302 (Man. C.A.).) This has resulted in some criticism, on the ground that the doctrine of abuse of process by relitigation is in effect non-mutual issue estoppel by another name without the important qualifications recognized by the American courts as part and parcel of the general doctrine of non-mutual issue estoppel (Watson, supra, at pp. 624-25).

[38] It is true that the doctrine of abuse of process has been extended beyond the strict parameters of res judicata while borrowing much of its rationales and some of its constraints. It is said to be more of an adjunct doctrine, defined in reaction to the settled rules of issue estoppel and cause of action estoppel, than an independent one (Lange, supra, at p. 344). The policy grounds supporting abuse of process by relitigation are the same as the essential policy grounds supporting issue estoppel (Lange, supra, at pp. 347-48):

The two policy grounds, namely, that there be an end to litigation and that no one should be twice vexed by the same cause, have been cited as policies in the application of abuse of process by relitigation. Other policy grounds have also been cited, namely, to preserve the courts’ and the litigants’ resources, to uphold the integrity of the legal system in order to avoid inconsistent results, and to protect the principle of finality so crucial to the proper administration of justice.

[39] The locus classicus for the modern doctrine of abuse of process and its relationship to res judicata is Hunter, supra, aff’g McIlkenny v. Chief Constable of the West Midlands, [1980] Q.B. 283 (C.A.). The case involved an action for damages for personal injuries brought by the six men convicted of bombing two pubs in Birmingham. They claimed that they had been beaten by the police during their interrogation. The plaintiffs had raised the same issue at their criminal trial, where it was found by both the judge and jury that the confessions were voluntary and that the police had not used violence. At the Court of Appeal, Lord Denning, M.R., endorsed non‑mutual issue estoppel and held that the question of whether any beatings had taken place was estopped by the earlier determination, although it was raised here against a different opponent. He noted that in analogous cases, courts had sometimes refused to allow a party to raise an issue for a second time because it was an “abuse of the process of the court”, but held that the proper characterization of the matter was through non‑mutual issue estoppel.

[40] On appeal to the House of Lords, Lord Denning’s attempt to reform the law of issue estoppel was overruled, but the higher court reached the same result via the doctrine of abuse of process. Lord Diplock stated, at p. 541:

The abuse of process which the instant case exemplifies is the initiation of proceedings in a court of justice for the purpose of mounting a collateral attack upon a final decision against the intending plaintiff which has been made by another court of competent jurisdiction in previous proceedings in which the intending plaintiff had a full opportunity of contesting the decision in the court by which it was made.

[41] It is important to note that a public inquiry after the civil action of the six accused in Hunter, supra, resulted in the finding that the confessions of the Birmingham six had been extracted through police brutality (see R. v. McIlkenny (1991), 93 Cr. App. R. 287 (C.A.), at pp. 304 et seq.). In my view, this does not support a relaxation of the existing procedural mechanisms designed to ensure finality in criminal proceedings. The danger of wrongful convictions has been acknowledged by this Court and other courts (see United States v. Burns, [2001] 1 S.C.R. 283, 2001 SCC 7, at para. 1; and R. v. Bromley (2001), 2001 NFCA 5, 151 C.C.C. (3d) 480 (Nfld. C.A.), at pp. 517‑18). Although safeguards must be put in place for the protection of the innocent, and, more generally, to ensure the trustworthiness of court findings, continuous re‑litigation is not a guarantee of factual accuracy.

[42] The attraction of the doctrine of abuse of process is that it is unencumbered by the specific requirements of res judicata while offering the discretion to prevent relitigation, essentially for the purpose of preserving the integrity of the court’s process. (See Doherty J.A.’s reasons, at para. 65; see also Demeter (H.C.), supra, at p. 264, and Hunter, supra, at p. 536.)

[43] Critics of that approach have argued that when abuse of process is used as a proxy for issue estoppel, it obscures the true question while adding nothing but a vague sense of discretion. I disagree. At least in the context before us, namely, an attempt to relitigate a criminal conviction, I believe that abuse of process is a doctrine much more responsive to the real concerns at play. In all of its applications, the primary focus of the doctrine of abuse of process is the integrity of the adjudicative functions of courts. Whether it serves to disentitle the Crown from proceeding because of undue delays (see Blencoe, supra), or whether it prevents a civil party from using the courts for an improper purpose (see Hunter, supra, and Demeter, supra), the focus is less on the interest of parties and more on the integrity of judicial decision making as a branch of the administration of justice. In a case such as the present one, it is that concern that compels a bar against relitigation, more than any sense of unfairness to a party being called twice to put its case forward, for example. When that is understood, the parameters of the doctrine become easier to define, and the exercise of discretion is better anchored in principle.

[44] The adjudicative process, and the importance of preserving its integrity, were well described by Doherty J.A. He said, at para. 74:

The adjudicative process in its various manifestations strives to do justice. By the adjudicative process, I mean the various courts and tribunals to which individuals must resort to settle legal disputes. Where the same issues arise in various forums, the quality of justice delivered by the adjudicative process is measured not by reference to the isolated result in each forum, but by the end result produced by the various processes that address the issue. By justice, I refer to procedural fairness, the achieving of the correct result in individual cases and the broader perception that the process as a whole achieves results which are consistent, fair and accurate.

[45] When asked to decide whether a criminal conviction, prima facie admissible in a proceeding under s. 22.1 of the Ontario Evidence Act, ought to be rebutted or taken as conclusive, courts will turn to the doctrine of abuse of process to ascertain whether relitigation would be detrimental to the adjudicative process as defined above. When the focus is thus properly on the integrity of the adjudicative process, the motive of the party who seeks to relitigate, or whether he or she wishes to do so as a defendant rather than as a plaintiff, cannot be decisive factors in the application of the bar against relitigation.

[46] Thus, in the case at bar, it matters little whether Oliver’s motive for relitigation was primarily to secure re‑employment, rather than to challenge his criminal conviction in an attempt to undermine its validity. Reliance on Hunter, supra, and on Demeter (H.C.), supra, for the purpose of enhancing the importance of motive is misplaced. It is true that in both cases the parties wishing to relitigate had made it clear that they were seeking to impeach their earlier convictions. But this is of little significance in the application of the doctrine of abuse of process. A desire to attack a judicial finding is not in itself an improper purpose. The law permits that objective to be pursued through various reviewing mechanisms such as appeals or judicial review. Indeed reviewability is an important aspect of finality. A decision is final and binding on the parties only when all available reviews have been exhausted or abandoned. What is improper is to attempt to impeach a judicial finding by the impermissible route of relitigation in a different forum. Therefore, motive is of little or no import.

[47] There is also no reason to constrain the doctrine of abuse of process only to those cases where the plaintiff has initiated the relitigation. The designation of the parties to the second litigation may mask the reality of the situation. In the present case, for instance, aside from the technical mechanism of the grievance procedures, who should be viewed as the initiator of the employment litigation between the grievor, Oliver, and his union on the one hand, and the City of Toronto on the other? Technically, the union is the “plaintiff” in the arbitration procedure. But the City of Toronto used Oliver’s criminal conviction as a basis for his dismissal. I cannot see what difference it makes, again from the point of view of the integrity of the adjudicative process, whether Oliver is labelled a plaintiff or a defendant when it comes to relitigating his criminal conviction.

[48] The appellant relies on Re Del Core, supra, to suggest that the abuse of process doctrine only applies to plaintiffs. Re Del Core, however, provided no majority opinion as to whether and when public policy would preclude relitigation of issues determined in a criminal proceeding. For one, Blair J.A. did not limit the circumstances in which relitigation would amount to an abuse of process to those cases in which a person convicted sought to relitigate the validity of his conviction in subsequent proceedings which he himself had instituted (at p. 22):

The right to challenge a conviction is subject to an important qualification. A convicted person cannot attempt to prove that the conviction was wrong in circumstances where it would constitute an abuse of process to do so. . . . Courts have rejected attempts to relitigate the very issues dealt with at a criminal trial where the civil proceedings were perceived to be a collateral attack on the criminal conviction. The ambit of this qualification remains to be determined . . . . [Emphasis added.]

[49] While the authorities most often cited in support of a court’s power to prevent relitigation of decided issues in circumstances where issue estoppel does not apply are cases where a convicted person commenced a civil proceeding for the purpose of attacking a finding made in a criminal proceeding against that person (namely Demeter (H.C.), supra, and Hunter, supra; see also Q. v. Minto Management Ltd. (1984), 1984 CanLII 2118 (ON SC), 46 O.R. (2d) 756 (H.C.), Franco, supra, at paras. 29-31), there is no reason in principle why these rules should be limited to such specific circumstances. Several cases have applied the doctrine of abuse of process to preclude defendants from relitigating issues decided against them in a prior proceeding. See for example Nigro v. Agnew‑Surpass Shoe Stores Ltd. (1977), 1977 CanLII 3406 (ON SC), 18 O.R. (2d) 215 (H.C.), at p. 218, aff’d without reference to this point (1978), 1976 CanLII 692 (ON CA), 18 O.R. (2d) 714 (C.A.); Bomac, supra, at pp. 26‑27; Bjarnarson, supra, at p. 39; Germscheid v. Valois (1989), 1989 CanLII 4156 (ON SC), 68 O.R. (2d) 670 (H.C.); Simpson v. Geswein (1995), 1995 CanLII 16110 (MB QB), 25 C.C.L.T. (2d) 49 (Man. Q.B.), at p. 61; Roenisch v. Roenisch (1991), 1991 CanLII 6223 (AB QB), 85 D.L.R. (4th) 540 (Alta. Q.B.), at p. 546; Saskatoon Credit Union, Ltd. v. Central Park Enterprises Ltd. (1988), 1988 CanLII 2941 (BC SC), 47 D.L.R. (4th) 431 (B.C.S.C.), at p. 438; Canadian Tire Corp. v. Summers (1995), 1995 CanLII 7183 (ON SC), 23 O.R. (3d) 106 (Gen. Div.), at p. 115; see also P. M. Perell, “Res Judicata and Abuse of Process” (2001), 24 Advocates’ Q. 189, at pp. 196‑97; and Watson, supra, at pp. 648‑51.

[50] It has been argued that it is difficult to see how mounting a defence can be an abuse of process (see M. Teplitsky, “Prior Criminal Convictions: Are They Conclusive Proof? An Arbitrator’s Perspective”, in K. Whitaker et al., eds., Labour Arbitration Yearbook 2001-2002 (2002), vol. I, 279). A common justification for the doctrine of res judicata is that a party should not be twice vexed in the same cause, that is, the party should not be burdened with having to relitigate the same issue (Watson, supra, at p. 633). Of course, a defendant may be quite pleased to have another opportunity to litigate an issue originally decided against him. A proper focus on the process, rather than on the interests of a party, will reveal why relitigation should not be permitted in such a case.

[51] Rather than focus on the motive or status of the parties, the doctrine of abuse of process concentrates on the integrity of the adjudicative process. Three preliminary observations are useful in that respect. First, there can be no assumption that relitigation will yield a more accurate result than the original proceeding. Second, if the same result is reached in the subsequent proceeding, the relitigation will prove to have been a waste of judicial resources as well as an unnecessary expense for the parties and possibly an additional hardship for some witnesses. Finally, if the result in the subsequent proceeding is different from the conclusion reached in the first on the very same issue, the inconsistency, in and of itself, will undermine the credibility of the entire judicial process, thereby diminishing its authority, its credibility and its aim of finality.

[52] In contrast, proper review by way of appeal increases confidence in the ultimate result and affirms both the authority of the process as well as the finality of the result. It is therefore apparent that from the system’s point of view, relitigation carries serious detrimental effects and should be avoided unless the circumstances dictate that relitigation is in fact necessary to enhance the credibility and the effectiveness of the adjudicative process as a whole. There may be instances where relitigation will enhance, rather than impeach, the integrity of the judicial system, for example: (1) when the first proceeding is tainted by fraud or dishonesty; (2) when fresh, new evidence, previously unavailable, conclusively impeaches the original results; or (3) when fairness dictates that the original result should not be binding in the new context. This was stated unequivocally by this Court in Danyluk, supra, at para. 80.

[53] The discretionary factors that apply to prevent the doctrine of issue estoppel from operating in an unjust or unfair way are equally available to prevent the doctrine of abuse of process from achieving a similar undesirable result. There are many circumstances in which the bar against relitigation, either through the doctrine of res judicata or that of abuse of process, would create unfairness. If, for instance, the stakes in the original proceeding were too minor to generate a full and robust response, while the subsequent stakes were considerable, fairness would dictate that the administration of justice would be better served by permitting the second proceeding to go forward than by insisting that finality should prevail. An inadequate incentive to defend, the discovery of new evidence in appropriate circumstances, or a tainted original process may all overcome the interest in maintaining the finality of the original decision (Danyluk, supra, at para. 51; Franco, supra, at para. 55).

[54] These considerations are particularly apposite when the attempt is to relitigate a criminal conviction. Casting doubt over the validity of a criminal conviction is a very serious matter. Inevitably in a case such as this one, the conclusion of the arbitrator has precisely that effect, whether this was intended or not. The administration of justice must equip itself with all legitimate means to prevent wrongful convictions and to address any real possibility of such an occurrence after the fact. Collateral attacks and relitigation, however, are not in my view appropriate methods of redress since they inordinately tax the adjudicative process while doing nothing to ensure a more trustworthy result.

[55] In light of the above, it is apparent that the common law doctrines of issue estoppel, collateral attack and abuse of process adequately capture the concerns that arise when finality in litigation must be balanced against fairness to a particular litigant. There is therefore no need to endorse, as the Court of Appeal did, a self-standing and independent “finality principle” either as a separate doctrine or as an independent test to preclude relitigation.

D. Application of Abuse of Process to Facts of the Appeal

[56] I am of the view that the facts in this appeal point to the blatant abuse of process that results when relitigation of this sort is permitted. The grievor was convicted in a criminal court and he exhausted all his avenues of appeal. In law, his conviction must stand, with all its consequent legal effects. Yet as pointed out by Doherty J.A. (at para. 84):

Despite the arbitrator’s insistence that he was not passing on the correctness of the decision made by Ferguson J., that is exactly what he did. One cannot read the arbitrator’s reasons without coming to the conclusion that he was convinced that the criminal proceedings were badly flawed and that Oliver was wrongly convicted. This conclusion, reached in proceedings to which the prosecution was not even a party, could only undermine the integrity of the criminal justice system. The reasonable observer would wonder how Oliver could be found guilty beyond a reasonable doubt in one proceeding and after the Court of Appeal had affirmed that finding, be found in a separate proceeding not to have committed the very same assault. That reasonable observer would also not understand how Oliver could be found to be properly convicted of sexually assaulting the complainant and deserving of 15 months in jail and yet also be found in a separate proceeding not to have committed that sexual assault and to be deserving of reinstatement in a job which would place young persons like the complainant under his charge.

[57] As a result of the conflicting decisions, the City of Toronto would find itself in the inevitable position of having a convicted sex offender reinstated to an employment position where he would work with the very vulnerable young people he was convicted of assaulting. An educated and reasonable public would presumably have to assess the likely correctness of one or the other of the adjudicative findings regarding the guilt of the convicted grievor. The authority and finality of judicial decisions are designed precisely to eliminate the need for such an exercise.

[58] In addition, the arbitrator is considerably less well equipped than a judge presiding over a criminal court — or the jury —, guided by rules of evidence that are sensitive to a fair search for the truth, an exacting standard of proof and expertise with the very questions in issue, to come to a correct disposition of the matter. Yet the arbitrator’s conclusions, if challenged, may give rise to a less searching standard of review than that of the criminal court judge. In short, there is nothing in a case like the present one that militates against the application of the doctrine of abuse of process to bar the relitigation of the grievor’s criminal conviction. The arbitrator was required as a matter of law to give full effect to the conviction. As a result of that error of law, the arbitrator reached a patently unreasonable conclusion. Properly understood in the light of correct legal principles, the evidence before the arbitrator could only lead him to conclude that the City of Toronto had established just cause for Oliver’s dismissal.

VI. Disposition

[59] For these reasons, I would dismiss the appeal with costs.

The reasons of LeBel and Deschamps JJ. were delivered by

LeBel J. —

I. Introduction

[60] I have had the benefit of reading Arbour J.’s reasons and I concur with her disposition of the case. I agree that this case is appropriately decided on the basis of the doctrine of abuse of process, rather than the narrower and more technical doctrines of either collateral attack or issue estoppel. I also agree that the appropriate standard of review for the question of whether a criminal conviction may be relitigated in a grievance proceeding is correctness. This is a question of law requiring an arbitrator to interpret not only the Labour Relations Act, 1995, S.O. 1995, c. 1, Sch. A, but also the Evidence Act, R.S.O. 1990, c. E.23, as well as to rule on the applicability of a number of common law doctrines dealing with relitigation, an issue that is, as Arbour J. notes, at the heart of the administration of justice. Finally, I agree that the arbitrator’s determination in this case that Glenn Oliver’s criminal conviction could indeed be relitigated during the grievance proceeding was incorrect. As a matter of law, the arbitrator was required to give full effect to Oliver’s conviction. His failure to do so was sufficient to render his ultimate decision that Oliver had been dismissed without just cause — a decision squarely within the arbitrator’s area of specialized expertise and thus reviewable on a deferential standard — patently unreasonable, according to the jurisprudence of our Court.

[61] While I agree with Arbour J.’s disposition of the appeal, I am of the view that the administrative law aspects of this case require further discussion. In my concurring reasons in Chamberlain v. Surrey School District No. 36, [2002] 4 S.C.R. 710, 2002 SCC 86, I raised concerns about the appropriateness of treating the pragmatic and functional methodology as an overarching analytical framework for substantive judicial review that must be applied, without variation, in all administrative law contexts, including those involving non-adjudicative decision makers. In certain circumstances, such as those at issue in Chamberlain itself, applying this methodological approach in order to determine the appropriate standard of review may in fact obscure the real issue before the reviewing court.

[62] In the instant appeal and the appeal in Ontario v. O.P.S.E.U., [2003] 3 S.C.R. 149, 2003 SCC 64, released concurrently, both of which involve judicial review of adjudicative decision makers, my concern is not with the applicability of the pragmatic and functional approach itself. Having said this, I would note that in a case such as this one, where the question at issue is so clearly a question of law that is both of central importance to the legal system as a whole and outside the adjudicator’s specialized area of expertise, it is unnecessary for the reviewing court to perform a detailed pragmatic and functional analysis in order to reach a standard of review of correctness. Indeed, in such circumstances reviewing courts should avoid adopting a mechanistic approach to the determination of the appropriate standard of review, which risks reducing the pragmatic and functional analysis from a contextual, flexible framework to little more than a pro forma application of a checklist of factors (see C.U.P.E. v. Ontario (Minister of Labour), [2003] 1 S.C.R. 539, 2003 SCC 29, at para. 149; Dr. Q v. College of Physicians and Surgeons of British Columbia, [2003] 1 S.C.R. 226, 2003 SCC 19, at para. 26; Chamberlain, supra, at para. 195, per LeBel J.).

[63] The more particular concern that emerges out of this case and Ontario v. O.P.S.E.U. relates to what in my view is growing criticism with the ways in which the standards of review currently available within the pragmatic and functional framework are conceived of and applied. Academic commentators and practitioners have raised some serious questions as to whether the conceptual basis for each of the existing standards has been delineated with sufficient clarity by this Court, with much of the criticism directed at what has been described as “epistemological” confusion over the relationship between patent unreasonableness and reasonableness simpliciter (see, for example, D. J. Mullan, “Recent Developments in Standard of Review”, in Canadian Bar Association (Ontario), Taking the Tribunal to Court: A Practical Guide for Administrative Law Practitioners (2000), at p. 26; J. G. Cowan, “The Standard of Review: The Common Sense Evolution?”, paper presented to the Administrative Law Section Meeting, Ontario Bar Association, January 21, 2003, at p. 28; F. A. V. Falzon, “Standard of Review on Judicial Review or Appeal”, in Administrative Justice Review Background Papers: Background Papers prepared by Administrative Justice Project for the Attorney General of British Columbia (2002), at pp. 32-33). Reviewing courts too, have occasionally expressed frustration over a perceived lack of clarity in this area, as the comments of Barry J. in Miller v. Workers’ Compensation Commission (Nfld.) (1997), 1997 CanLII 10862 (NL SC), 154 Nfld. & P.E.I.R. 52 (Nfld. S.C.T.D.), at para. 27, illustrate:

In attempting to follow the court’s distinctions between “patently unreasonable”, “reasonable” and “correct”, one feels at times as though one is watching a juggler juggle three transparent objects. Depending on the way the light falls, sometimes one thinks one can see the objects. Other times one cannot and, indeed, wonders whether there are really three distinct objects there at all.

[64] The Court cannot remain unresponsive to sustained concerns or criticism coming from the legal community in relation to the state of Canadian jurisprudence in this important part of the law. It is true that the parties to this appeal made no submissions putting into question the standards of review jurisprudence. Nevertheless, at times, an in-depth discussion or review of the state of the law may become necessary despite the absence of particular representations in a specific case. Given its broad application, the law governing the standards of review must be predictable, workable and coherent. Parties to litigation often have no personal stake in assuring the coherence of our standards of review jurisprudence as a whole and the consistency of their application. Their purpose, understandably, is to show how the positions they advance conform with the law as it stands, rather than to suggest improvements of that law for the benefit of the common good. The task of maintaining a predictable, workable and coherent jurisprudence falls primarily on the judiciary, preferably with, but exceptionally without, the benefit of counsel. I would add that, although the parties made no submissions on the analysis that I propose to undertake in these reasons, they will not be prejudiced by it.

[65] In this context, this case provides an opportunity to reevaluate the contours of the various standards of review, a process that in my view is particularly important with respect to patent unreasonableness. To this end, I review below:

• the interplay between correctness and patent unreasonableness both in the instant case and, more broadly, in the context of judicial review of adjudicative decision makers generally, with a view to elucidating the conflicted relationship between these two standards; and,

• the distinction between patent unreasonableness and reasonableness simpliciter, which, despite a number of attempts at clarification, remains a nebulous one.

[66] As the analysis that follows indicates, the patent unreasonableness standard does not currently provide sufficiently clear parameters for reviewing courts to apply in assessing the decisions of administrative adjudicators. From the beginning, patent unreasonableness at times shaded uncomfortably into what should presumably be its antithesis, the correctness review. Moreover, it is increasingly difficult to distinguish from what is ostensibly its less deferential counterpart, reasonableness simpliciter. It remains to be seen how these difficulties can be addressed.

II. Analysis

A. The Two Standards of Review Applicable in This Case

[67] Two standards of review are at issue in this case, and the use of correctness here requires some preliminary discussion. As I noted in brief above, certain fundamental legal questions — for instance, constitutional and human rights questions and those involving civil liberties, as well as other questions that are of central importance to the legal system as a whole, such as the issue of relitigation — typically fall to be decided on a correctness standard. Indeed, in my view, it will rarely be necessary for reviewing courts to embark on a comprehensive application of the pragmatic and functional approach in order to reach this conclusion. I would not, however, want either my comments in this regard or the majority reasons in this case to be taken as authority for the proposition that correctness is the appropriate standard whenever arbitrators or other specialized administrative adjudicators are required to interpret and apply general common law or civil law rules. Such an approach would constitute a broad expansion of judicial review under a standard of correctness and would significantly impede the ability of administrative adjudicators, particularly in complex and highly specialized fields such as labour law, to develop original solutions to legal problems, uniquely suited to the context in which they operate. In my opinion, in many instances the appropriate standard of review in respect of the application of general common or civil law rules by specialized adjudicators should not be one of correctness, but rather of reasonableness. I now turn to a brief discussion of the rationale behind this view.

(1) The Correctness Standard of Review

[68] This Court has repeatedly stressed the importance of judicial deference in the context of labour law. Labour relations statutes typically bestow broad powers on arbitrators and labour boards to resolve the wide range of problems that may arise in this field and protect the decisions of these adjudicators by privative clauses. Such legislative choices reflect the fact that, as Cory J. noted in Toronto (City) Board of Education v. O.S.S.T.F., District 15, 1997 CanLII 378 (SCC), [1997] 1 S.C.R. 487, at para. 35, the field of labour relations is “sensitive and volatile” and “[i]t is essential that there be a means of providing speedy decisions by experts in the field who are sensitive to the situation, and which can be considered by both sides to be final and binding” (see also Canada (Attorney General) v. Public Service Alliance of Canada, 1993 CanLII 125 (SCC), [1993] 1 S.C.R. 941 (“PSAC”), at pp. 960-61; and Ivanhoe inc. v. UFCW, Local 500, [2001] 2 S.C.R. 565, 2001 SCC 47, at para. 32). The application of a standard of review of correctness in the context of judicial review of labour adjudication is thus rare.

[69] While in this case and in Ontario v. O.P.S.E.U. I agree that correctness is the appropriate standard of review for the arbitrator’s decision on the relitigation question, I think it necessary to sound a number of notes of caution in this regard. It is important to stress, first, that while the arbitrator was required to be correct on this question of law, this did not open his decision as a whole to review on a correctness standard (see Canadian Broadcasting Corp. v. Canada (Labour Relations Board), 1995 CanLII 148 (SCC), [1995] 1 S.C.R. 157, at para. 48). The arbitrator was entitled to deference in the determination of whether Oliver was dismissed without just cause. To say that, in the circumstances of this case, the arbitrator’s incorrect decision on the question of law affected the overall reasonableness of his decision, is very different from saying that the arbitrator’s finding on the ultimate question of just cause had to be correct. To fail to make this distinction would be to risk “substantially expand[ing] the scope of reviewability of administrative decisions, and unjustifiably so” (see Canadian Broadcasting Corp., supra, at para. 48).

[70] Second, it bears repeating that the application of correctness here is very much a product of the nature of this particular legal question: determining whether relitigating an employee’s criminal conviction is permissible in an arbitration proceeding is a question of law involving the interpretation of the arbitrator’s constitutive statute, an external statute, and a complex body of common law rules and conflicting jurisprudence. More than this, it is a question of fundamental importance and broad applicability, with serious implications for the administration of justice as a whole. It is, in other words, a question that engages the expertise and essential role of the courts. It is not a question on which arbitrators may be said to enjoy any degree of relative institutional competence or expertise. As a result, it is a question on which the arbitrator must be correct.

[71] This Court has been very careful to note, however, that not all questions of law must be reviewed under a standard of correctness. As a prefatory matter, as the Court has observed, in many cases it will be difficult to draw a clear line between questions of fact, mixed fact and law, and law; in reality, such questions are often inextricably intertwined (see Pushpanathan v. Canada (Minister of Citizenship and Immigration), 1998 CanLII 778 (SCC), [1998] 1 S.C.R. 982, at para. 37; Canada (Director of Investigation and Research) v. Southam Inc., 1997 CanLII 385 (SCC), [1997] 1 S.C.R. 748, at para. 37). More to the point, as Bastarache J. noted in Pushpanathan, supra, “even pure questions of law may be granted a wide degree of deference where other factors of the pragmatic and functional analysis suggest that such deference is the legislative intention” (para. 37). The critical factor in this respect is expertise.

[72] As Bastarache J. noted in Pushpanathan, supra, at para. 34, once a “broad relative expertise has been established”, this Court has been prepared to show “considerable deference even in cases of highly generalized statutory interpretation where the instrument being interpreted is the tribunal’s constituent legislation”: see, for example, Pezim v. British Columbia (Superintendent of Brokers), 1994 CanLII 103 (SCC), [1994] 2 S.C.R. 557, and National Corn Growers Assn. v. Canada (Import Tribunal), 1990 CanLII 49 (SCC), [1990] 2 S.C.R. 1324. This Court has also held that, while administrative adjudicators’ interpretations of external statutes “are generally reviewable on a correctness standard”, an exception to this general rule may occur, and deference may be appropriate, where “the external statute is intimately connected with the mandate of the tribunal and is encountered frequently as a result”: see Toronto (City) Board of Education, supra, at para. 39; Canadian Broadcasting Corp., supra, at para. 48. And, perhaps most importantly in light of the issues raised by this case, the Court has held that deference may be warranted where an administrative adjudicator has acquired expertise through its experience in the application of a general common or civil law rule in its specialized statutory context: see Ivanhoe, supra, at para. 26; L’Heureux‑Dubé J. (dissenting) in Canada (Attorney General) v. Mossop, 1993 CanLII 164 (SCC), [1993] 1 S.C.R. 554, at pp. 599‑600, endorsed in Pushpanathan, supra, at para. 37.

[73] In the field of labour relations, general common and civil law questions are often closely intertwined with the more specific questions of labour law. Resolving general legal questions may thus be an important component of the work of some administrative adjudicators in this field. To subject all such decisions to correctness review would be to expand the scope of judicial review considerably beyond what the legislature intended, fundamentally undermining the ability of labour adjudicators to develop a body of jurisprudence that is tailored to the specialized context in which they operate.

[74] Where an administrative adjudicator must decide a general question of law in the course of exercising its statutory mandate, that determination will typically be entitled to deference (particularly if the adjudicator’s decisions are protected by a privative clause), inasmuch as the general question of law is closely connected to the adjudicator’s core area of expertise. This was essentiality the holding of this Court in Ivanhoe, supra. In Ivanhoe, after noting the presence of a privative clause, Arbour J. held that, while the question at issue involved both civil and labour law, the labour commissioners and the Labour Court were entitled to deference because “they have developed special expertise in this regard which is adapted to the specific context of labour relations and which is not shared by the courts” (para. 26; see also Pasiechnyk v. Saskatchewan (Workers’ Compensation Board), 1997 CanLII 316 (SCC), [1997] 2 S.C.R. 890). This appeal does not represent a departure from this general principle.

[75] The final note of caution that I think must be sounded here relates to the application of two standards of review in this case. This Court has recognized on a number of occasions that it may, in certain circumstances, be appropriate to apply different standards of deference to different decisions taken by an administrative adjudicator in a single case (see Pushpanathan, supra, at para. 49; Macdonell v. Quebec (Commission d’accès à l'information), [2002] 3 S.C.R. 661, 2002 SCC 71, at para. 58, per Bastarache and LeBel JJ., dissenting). This case provides an example of one type of situation where this may be the proper approach. It involves a fundamental legal question falling outside the arbitrator’s area of expertise. This legal question, though foundational to the decision as a whole, is easily differentiated from a second question on which the arbitrator was entitled to deference: the determination of whether there was just cause for Oliver’s dismissal.

[76] However, as I have noted above, the fact that the question adjudicated by the arbitrator in this case can be separated into two distinct issues, one of which is reviewable on a correctness standard, should not be taken to mean that this will often be the case. Such cases are rare; the various strands that go into a decision are more likely to be inextricably intertwined, particularly in a complex field such as labour relations, such that the reviewing court should view the adjudicator’s decision as an integrated whole.

(2) The Patent Unreasonableness Standard of Review

[77] In these reasons, I explore the way in which patent unreasonableness is currently functioning, having regard to the relationships between this standard and both correctness and reasonableness simpliciter. My comments in this respect are intended to have application in the context of judicial review of adjudicative administrative decision making.

(a) The Definitions of Patent Unreasonableness

[78] This Court has set out a number of definitions of “patent unreasonableness”, each of which is intended to indicate the high degree of deference inherent in this standard of review. There is some overlap between the definitions and they are often used in combination. I would characterize the two main definitional strands as, first, those that emphasize the magnitude of the defect necessary to render a decision patently unreasonable and, second, those that focus on the “immediacy or obviousness” of the defect, and thus the relative invasiveness of the review necessary to find it.

[79] In considering the leading definitions, I would place in the first category Dickson J.’s (as he then was) statement in Canadian Union of Public Employees, Local 963 v. New Brunswick Liquor Corp., 1979 CanLII 23 (SCC), [1979] 2 S.C.R. 227 (“CUPE”), that a decision will only be patently unreasonable if it “cannot be rationally supported by the relevant legislation” (p. 237). Cory J.’s characterization in PSAC, supra, of patent unreasonableness as a “very strict test”, which will only be met where a decision is “clearly irrational, that is to say evidently not in accordance with reason” (pp. 963‑64), would also fit into this category (though it could, depending on how it is read, be placed in the second category as well).

[80] In the second category, I would place Iacobucci J.’s description in Southam, supra, of a patently unreasonable decision as one marred by a defect that is characterized by its “immediacy or obviousness”: “If the defect is apparent on the face of the tribunal’s reasons, then the tribunal’s decision is patently unreasonable. But if it takes some significant searching or testing to find the defect, then the decision is unreasonable but not patently unreasonable” (para. 57).

[81] More recently, in Law Society of New Brunswick v. Ryan, [2003] 1 S.C.R. 247, 2003 SCC 20, Iacobucci J. characterized a patently unreasonable decision as one that is “so flawed that no amount of curial deference can justify letting it stand”, drawing on both of the definitional strands that I have identified in formulating this definition. He wrote, at para. 52:

In Southam, supra, at para. 57, the Court described the difference between an unreasonable decision and a patently unreasonable one as rooted “in the immediacy or obviousness of the defect”. Another way to say this is that a patently unreasonable defect, once identified, can be explained simply and easily, leaving no real possibility of doubting that the decision is defective. A patently unreasonable decision has been described as “clearly irrational” or “evidently not in accordance with reason” (Canada (Attorney General) v. Public Service Alliance of Canada, 1993 CanLII 125 (SCC), [1993] 1 S.C.R. 941, at pp. 963‑64, per Cory J.; Centre communautaire juridique de l'Estrie v. Sherbrooke (City), 1996 CanLII 168 (SCC), [1996] 3 S.C.R. 84, at paras. 9‑12, per Gonthier J.). A decision that is patently unreasonable is so flawed that no amount of curial deference can justify letting it stand.

[82] Similarly, in C.U.P.E. v. Ontario, supra, Binnie J. yoked together the two definitional strands, describing a patently unreasonable decision as “one whose defect is ‘immedia[te] and obviou[s]’ (Southam, supra, at para. 57), and so flawed in terms of implementing the legislative intent that no amount of curial deference can properly justify letting it stand (Ryan, supra, at para. 52)” (para. 165 (emphasis added)).

[83] It has been suggested that the Court’s various formulations of the test for patent unreasonableness are “not independent, alternative tests. They are simply ways of getting at the single question: What makes something patently unreasonable?” (C.U.P.E. v. Ontario, supra, at para. 20, per Bastarache J., dissenting). While this may indeed be the case, I nonetheless think it important to recognize that, because of what are in some ways subtle but nonetheless quite significant differences between the Court’s various answers to this question, the parameters of “patent unreasonableness” are not as clear as they could be. This has contributed to the growing difficulties in the application of this standard that I discuss below.

(b) The Interplay Between the Patent Unreasonableness and Correctness Standards

[84] As I observed in Chamberlain, supra, the difference between review on a standard of correctness and review on a standard of patent unreasonableness is “intuitive and relatively easy to observe” (Chamberlain, supra, at para. 204, per LeBel J.). These standards fall on opposite sides of the existing spectrum of curial deference, with correctness entailing an exacting review and patent unreasonableness leaving the issue in question to the near exclusive determination of the decision maker (see Dr. Q, supra, at para. 22). Despite the clear conceptual boundary between these two standards, however, the distinction between them is not always as readily discernable in practice as one would expect.

(i) Patent Unreasonableness and Correctness in Theory

[85] In terms of understanding the interplay between patent unreasonableness and correctness, it is of interest that, from the beginning, there seems to have been at least some conceptual uncertainty as to the proper breadth of patent unreasonableness review. In CUPE, supra, Dickson J. offered two characterizations of patent unreasonableness that tend to pull in opposite directions (see D. J. Mullan, Administrative Law (2001), at p. 69; see also H. W. MacLauchlan, “Transforming Administrative Law: The Didactic Role of the Supreme Court of Canada” (2001), 80 Can. Bar Rev. 281, at pp. 285-86).

[86] Professor Mullan explains that, on the one hand, Dickson J. rooted review for patent unreasonableness in the recognition that statutory provisions are often ambiguous and thus may allow for multiple interpretations; the question for the reviewing court is whether the adjudicator’s interpretation is one that can be “rationally supported by the relevant legislation” (CUPE, supra, at p. 237). On the other hand, Dickson J. also invoked an idea of patent unreasonableness as a threshold defined by certain nullifying errors, such as those he had previously enumerated in Service Employees’ International Union, Local No. 333 v. Nipawin District Staff Nurses Association, 1973 CanLII 191 (SCC), [1975] 1 S.C.R. 382 (“Nipawin”), at p. 389, and in CUPE, supra, at p. 237:

. . . acting in bad faith, basing the decision on extraneous matters, failing to take relevant factors into account, breaching the provisions of natural justice or misinterpreting provisions of the Act so as to embark on an inquiry or answer a question not remitted to it.

[87] Curiously, as Mullan notes, this list “repeats the list of ‘nullifying’ errors that Lord Reid laid out in the landmark House of Lords’ judgment” in Anisminic Ltd. v. Foreign Compensation Commission, [1969] 2 A.C. 147. Anisminic “is usually treated as the foundation case in establishing in English law the reviewability of all issues of law on a correctness basis” (emphasis added), and, indeed, the Court “had cited with approval this portion of Lord Reid’s judgment and deployed it to justify judicial intervention in a case described as the ‘high water mark of activist’ review in Canada: Metropolitan Life Insurance Co. v. International Union of Operating Engineers, Local 796”, 1970 CanLII 7 (SCC), [1970] S.C.R. 425 (see Mullan, Administrative Law, supra, at pp. 69-70; see also National Corn Growers, supra, at p. 1335, per Wilson J.).

[88] In characterizing patent unreasonableness in CUPE, then, Dickson J. simultaneously invoked a highly deferential standard (choice among a range of reasonable alternatives) and a historically interventionist one (based on the presence of nullifying errors). For this reason, as Mullan acknowledges, “it is easy to see why Dickson J.’s use of [the quotation from Anisminic] is problematic” (Mullan, Administrative Law, supra, at p. 70).

[89] If Dickson J.’s reference to Anisminic in CUPE, supra, suggests some ambiguity as to the intended scope of “patent unreasonableness” review, later judgments also evidence a somewhat unclear relationship between patent unreasonableness and correctness in terms of establishing and, particularly, applying the methodology for review under the patent unreasonableness standard. The tension in this respect is rooted, in part, in differing views of the premise from which patent unreasonableness review should begin. A useful example is provided by CAIMAW v. Paccar of Canada Ltd., 1989 CanLII 49 (SCC), [1989] 2 S.C.R. 983 (“Paccar”).

[90] In Paccar, Sopinka J. (Lamer J. (as he then was) concurring) described the proper approach under the patent unreasonableness standard as one in which the reviewing court first queries whether the administrative adjudicator’s decision is correct: “curial deference does not enter the picture until the court finds itself in disagreement with the tribunal. Only then is it necessary to consider whether the error (so found) is within or outside the boundaries of reasonableness” (p. 1018). As Mullan has observed, this approach to patent unreasonableness raises concerns in that it not only conflicts “with the whole notion espoused by Dickson J. in [CUPE, supra] of there often being no single correct answer to statutory interpretation problems but it also assumes the primacy of the reviewing court over the agency or tribunal in the delineation of the meaning of the relevant statute” (Mullan, “Recent Developments in Standard of Review”, supra, at p. 20).

[91] In my view, this approach presents additional problems as well. Reviewing courts may have difficulty ruling that “an error has been committed but . . . then do[ing] nothing to correct that error on the basis that it was not as big an error as it could or might have been” (see Mullan, “Recent Developments in Standard of Review”, supra, at p. 20; see also D. J. Mullan, “Of Chaff Midst the Corn: American Farm Bureau Federation v. Canada (Canadian Import Tribunal) and Patent Unreasonableness Review” (1991), 45 Admin. L.R. 264, at pp. 269-70). Furthermore, starting from a finding that the adjudicator’s decision is incorrect may colour the reviewing court’s subsequent assessment of the reasonableness of competing interpretations (see M. Allars, “On Deference to Tribunals, With Deference to Dworkin” (1994), 20 Queen’s L.J. 163, at p. 187). The result is that the critical distinction between that which is, in the court’s eyes, “incorrect” and that which is “not rationally supportable” is undermined.

[92] The alternative approach is to leave the “correctness” of the adjudicator’s decision undecided (see Allars, supra, at p. 197). This is essentially the approach that La Forest J. (Dickson C.J. concurring) took to patent unreasonableness in Paccar, supra. He wrote, at pp. 1004 and 1005:

The courts must be careful to focus their inquiry on the existence of a rational basis for the decision of the tribunal, and not on their agreement with it.

I do not find it necessary to conclusively determine whether the decision of the Labour Relations Board is “correct” in the sense that it is the decision I would have reached had the proceedings been before this Court on their merits. It is sufficient to say that the result arrived at by the Board is not patently unreasonable.

[93] It is this theoretical view that has, at least for the most part, prevailed. As L’Heureux-Dubé J. observed in Canadian Union of Public Employees, Local 301 v. Montreal (City), 1997 CanLII 386 (SCC), [1997] 1 S.C.R. 793 (“CUPE, Local 301”), “this Court has stated repeatedly, in assessing whether administrative action is patently unreasonable, the goal is not to review the decision or action on its merits but rather to determine whether it is patently unreasonable, given the statutory provisions governing the particular body and the evidence before it” (para. 53). Patent unreasonableness review, in other words, should not “become an avenue for the court’s substitution of its own view” (CUPE, Local 301, supra, at para. 59; see also Domtar Inc. v. Quebec (Commission d’appel en matière de lésions professionnelles), 1993 CanLII 106 (SCC), [1993] 2 S.C.R. 756, at pp. 771 and 774-75).

[94] This view was recently forcefully rearticulated in Ryan, supra. Iacobucci J. wrote, at paras. 50-51:

[W]hen deciding whether an administrative action was unreasonable, a court should not at any point ask itself what the correct decision would have been. . . . The standard of reasonableness does not imply that a decision maker is merely afforded a “margin of error” around what the court believes is the correct result.

. . . Unlike a review for correctness, there will often be no single right answer to the questions that are under review against the standard of reasonableness. . . . Even if there could be, notionally, a single best answer, it is not the court’s role to seek this out when deciding if the decision was unreasonable.

Though Iacobucci J.’s comments here were made in relation to reasonableness simpliciter, they are also applicable to the more deferential standard of patent unreasonableness.

[95] I think it important to emphasize that neither the case at bar, nor the companion case of Ontario v. O.P.S.E.U., should be misinterpreted as a retreat from the position that in reviewing a decision under the existing standard of patent unreasonableness, the court’s role is not to identify the “correct” result. In each of these cases, there were two standards of review in play: there was a fundamental legal question on which the adjudicators were subject to a standard of correctness — whether the employees’ criminal convictions could be relitigated — and there was a question at the core of the adjudicators’ expertise on which they were subject to a standard of patent unreasonableness — whether the employees had been dismissed for just cause. As Arbour J. has outlined, the adjudicators’ failure to decide the fundamental relitigation question correctly was sufficient to lead to a patently unreasonable outcome. Indeed, in circumstances such as those at issue in the case at bar, this cannot but be the case: the adjudicators’ incorrect decisions on the fundamental legal question provided the entire foundation on which their legal analyses, and their conclusions as to whether the employees were dismissed with just cause, were based. To pass a review for patent unreasonableness, a decision must be one that can be “rationally supported”; this standard cannot be met where, as here, what supports the adjudicator’s decision — indeed, what that decision is wholly premised on — is a legal determination that the adjudicator was required, but failed, to decide correctly. To say, however, that in such circumstances a decision will be patently unreasonable — a conclusion that flows from the applicability of two separate standards of review — is very different from suggesting that a reviewing court, before applying the standard of patent unreasonableness, must first determine whether the adjudicator’s decision is (in)correct or that in applying patent unreasonableness the court should ask itself at any point in the analysis what the correct decision would be. In other words, the application of patent unreasonableness itself is not, and should not be, understood to be predicated on a finding of incorrectness, for the reasons that I discussed above.

(ii) Patent Unreasonableness and Correctness in Practice

[96] While the Court now tends toward the view that La Forest J. articulated in Paccar, at p. 1004 — “courts must be careful [under a standard of patent unreasonableness] to focus their inquiry on the existence of a rational basis for the decision of the tribunal, and not on their agreement with it” — the tension between patent unreasonableness and correctness has not been completely resolved. Slippage between the two standards is still evident at times in the way in which patent unreasonableness is applied.

[97] In analyzing a number of recent cases, commentators have pointed to both the intensity and the underlying character of the review in questioning whether the Court is applying patent unreasonableness in a manner that is in fact deferential. In this regard, the comments of Professor Lorne Sossin on the application of patent unreasonableness in Canada Safeway Ltd. v. RWDSU, Local 454, 1998 CanLII 780 (SCC), [1998] 1 S.C.R. 1079, are illustrative:

Having established that deference was owed to the statutory interpretation of the Board, the Court proceeded to dissect its interpretation. The majority was of the view that the Board had misconstrued the term “constructive lay-off” and had failed to place sufficient emphasis on the terms of the collective agreement. The majority reasons convey clearly why the Court would adopt a different approach to the Board. They are less clear as to why the Board’s approach lacked a rational foundation. Indeed, there is very little evidence of the Court according deference to the Board’s interpretation of its own statute, or to its choice as to how much weight to place on the terms of the collective agreement. Canada Safeway raises the familiar question of how a court should demonstrate its deference, particularly in the labour relations context.

(L. Sossin, “Developments in Administrative Law: The 1997-98 and 1998‑99 Terms” (2000), 11 S.C.L.R. (2d) 37, at p. 49)

[98] Professor Ian Holloway makes a similar observation with regard to Lester (W.W.) (1978) Ltd. v. United Association of Journeymen and Apprentices of the Plumbing and Pipefitting Industry, Local 740, 1990 CanLII 22 (SCC), [1990] 3 S.C.R. 644:

In her judgment, [McLachlin J. (as she then was)] quoted from the familiar passages of CUPE, yet she . . . reached her decision on the basis of a review of the case law. She did not ask whether, despite the fact that it differed from holdings in other jurisdictions, the conclusion of the Newfoundland Labour Relations Board could be “rationally supported” on the basis of the wording of the successorship provisions of the Labour Relations Act. Instead, she looked at whether the Board had reached the correct legal interpretation of the Act in the same manner that a court of appeal would determine whether a trial judge had made a correct interpretation of the law. In other words, she effectively equated patent unreasonability with correctness at law.

(I. Holloway, “‘A Sacred Right’: Judicial Review of Administrative Action as a Cultural Phenomenon” (1993), 22 Man. L.J. 28, at pp. 64-65 (emphasis in original); see also Allars, supra, at p. 178.)

[99] At times the Court’s application of the standard of patent unreasonableness may leave it vulnerable to criticism that it may in fact be doing implicitly what it has rejected explicitly: intervening in decisions that are, in its view, incorrect, rather than limiting any intervention to those decisions that lack a rational foundation. In the process, what should be an indelible line between correctness, on the one hand, and patent unreasonableness, on the other, becomes blurred. It may very well be that review under any standard of reasonableness, given the nature of the intellectual process it involves, entails such a risk. Nevertheless, the existence of two standards of reasonableness appears to have magnified the underlying tension between the two standards of reasonableness and correctness.

(c) The Relationship Between the Patent Unreasonableness and Reasonableness Simpliciter Standards

[100] While the conceptual difference between review on a correctness standard and review on a patent unreasonableness standard may be intuitive and relatively easy to observe (though in practice elements of correctness at times encroach uncomfortably into patent unreasonableness review), the boundaries between patent unreasonableness and reasonableness simpliciter are far less clear, even at the theoretical level.

(i) The Theoretical Foundation for Patent Unreasonableness and Reasonableness Simpliciter

[101] The lack of sufficiently clear boundaries between patent unreasonableness and reasonableness simpliciter has its origins in the fact that patent unreasonableness was developed prior to the birth of the pragmatic and functional approach (see C.U.P.E. v. Ontario, supra, at para. 161) and, more particularly, prior to (rather than in conjunction with) the formulation of reasonableness simpliciter in Southam, supra. Because patent unreasonableness, as a posture of curial deference, was conceived in opposition only to a correctness standard of review, it was sufficient for the Court to emphasize in defining its scope the principle that there will often be no one interpretation that can be said to be correct in interpreting a statute or otherwise resolving a legal dispute, and that specialized administrative adjudicators may, in many circumstances, be better equipped than courts to choose between the possible interpretations. Where this is the case, provided that the adjudicator’s decision is one that can be “rationally supported on a construction which the relevant legislation may reasonably be considered to bear”, the reviewing court should not intervene (Nipawin, supra, at p. 389).

[102] Upon the advent of reasonableness simpliciter, however, the validity of multiple interpretations became the underlying premise for this new variant of reasonableness review as well. Consider, for instance, the discussion of reasonableness simpliciter in Ryan, that I cited above:

Unlike a review for correctness, there will often be no single right answer to the questions that are under review against the standard of reasonableness. . . . Even if there could be, notionally, a single best answer, it is not the court’s role to seek this out when deciding if the decision was unreasonable.

(Ryan, supra, at para. 51; see also para. 55.)

It is difficult to distinguish this language from that used to describe patent unreasonableness not only in the foundational judgments establishing that standard, such as Nipawin, supra, and CUPE, supra, but also in this Court’s more contemporary jurisprudence applying it. In Ivanhoe, supra, for instance, Arbour J. stated that “the recognition by the legislature and the courts that there are many potential solutions to a dispute is the very essence of the patent unreasonableness standard of review, which would be meaningless if it was found that there is only one acceptable solution” (para. 116).

[103] Because patent unreasonableness and reasonableness simpliciter are both rooted in this guiding principle, it has been difficult to frame the standards as analytically, rather than merely semantically, distinct. The efforts to sustain a workable distinction between them have taken, in the main, two forms, which mirror the two definitional strands of patent unreasonableness that I identified above. One of these forms distinguishes between patent unreasonableness and reasonableness simpliciter on the basis of the relative magnitude of the defect. The other looks to the “immediacy or obviousness” of the defect, and thus the relative invasiveness of the review necessary to find it. Both approaches raise their own problems.

(ii) The Magnitude of the Defect

[104] In PSAC, supra, at pp. 963-64, Cory J. described a patently unreasonable decision in these terms:

In the Shorter Oxford English Dictionary “patently”, an adverb, is defined as “openly, evidently, clearly”. “Unreasonable” is defined as “[n]ot having the faculty of reason; irrational. . . . Not acting in accordance with reason or good sense”. Thus, based on the dictionary definition of the words “patently unreasonable”, it is apparent that if the decision the Board reached, acting within its jurisdiction, is not clearly irrational, that is to say evidently not in accordance with reason, then it cannot be said that there was a loss of jurisdiction.

While this definition may not be inherently problematic, it has become so with the emergence of reasonableness simpliciter, in part because of what commentators have described as the “tautological difficulty of distinguishing standards of rationality on the basis of the term ‘clearly’” (see Cowan, supra, at pp. 27-28; see also G. Perrault, Le contrôle judiciaire des décisions de l’administration: De l’erreur juridictionnelle à la norme de contrôle (2002), at p. 116; S. Comtois, Vers la primauté de l’approche pragmatique et fonctionnelle: Précis du contrôle judiciaire des décisions de fond rendues par les organismes administratifs (2003), at pp. 34-35; P. Garant, Droit administratif (4th ed. 1996), vol. 2, at p. 193).

[105] Mullan alludes to both the practical and the theoretical difficulties of maintaining a distinction based on the magnitude of the defect, i.e., the degree of irrationality, that characterizes a decision:

. . . admittedly in his judgment in PSAC, Cory J. did attach the epithet “clearly” to the word “irrational” in delineating a particular species of patent unreasonableness. However, I would be most surprised if, in so doing, he was using the term “clearly” for other than rhetorical effect. Indeed, I want to suggest . . . that to maintain a position that it is only the “clearly irrational” that will cross the threshold of patent unreasonableness while irrationality simpliciter will not is to make a nonsense of the law. Attaching the adjective “clearly” to irrational is surely a tautology. Like “uniqueness”, irrationality either exists or it does not. There cannot be shades of irrationality. In other words, I defy any judge or lawyer to provide a concrete example of the difference between the merely irrational and the clearly irrational! In any event, there have to be concerns with a regime of judicial review which would allow any irrational decision to escape rebuke even under the most deferential standard of scrutiny.

(Mullan, “Recent Developments in Standard of Review”, supra, at pp. 24‑25)

Also relevant in this respect are the comments of Reed J. in Hao v. Canada (Minister of Citizenship and Immigration) (2000), 2000 CanLII 15150 (FC), 184 F.T.R. 246, at para. 9:

I note that I have never been convinced that “patently unreasonable” differs in a significant way from “unreasonable”. The word “patently” means clearly or obviously. If the unreasonableness of a decision is not clear or obvious, I do not see how that decision can be said to be unreasonable.

[106] Even a brief review of this Court’s descriptions of the defining characteristics of patently unreasonable and unreasonable decisions demonstrates that it is difficult to sustain a meaningful distinction between two forms of reasonableness on the basis of the magnitude of the defect, and the extent of the decision’s resulting deviation from the realm of the reasonable. Under both standards, the reviewing court’s inquiry is focussed on “the existence of a rational basis for the [adjudicator’s] decision” (see, for example, Paccar, supra, at p. 1004, per La Forest J.; Ryan, supra, at paras. 55-56). A patently unreasonable decision has been described as one that “cannot be sustained on any reasonable interpretation of the facts or of the law” (National Corn Growers, supra, at pp. 1369-70, per Gonthier J.), or “rationally supported on a construction which the relevant legislation may reasonably be considered to bear” (Nipawin, supra, at p. 389). An unreasonable decision has been described as one for which there are “no lines of reasoning supporting the decision which could reasonably lead that tribunal to reach the decision it did” (Ryan, supra, at para. 53).

[107] Under both patent unreasonableness and reasonableness simpliciter, mere disagreement with the adjudicator’s decision is insufficient to warrant intervention (see, for example, Paccar, supra, at pp. 1003-4, per La Forest J., and Chamberlain, supra, at para. 15, per McLachlin C.J.). Applying the patent unreasonableness standard, “the court will defer even if the interpretation given by the tribunal . . . is not the ‘right’ interpretation in the court’s view nor even the ‘best’ of two possible interpretations, so long as it is an interpretation reasonably attributable to the words of the agreement” (United Brotherhood of Carpenters and Joiners of America, Local 579 v. Bradco Construction Ltd., 1993 CanLII 88 (SCC), [1993] 2 S.C.R. 316, at p. 341). In the case of reasonableness simpliciter, “a decision may satisfy the . . . standard if it is supported by a tenable explanation even if this explanation is not one that the reviewing court finds compelling” (Ryan, supra, at para. 55). There seems to me to be no qualitative basis on which to differentiate effectively between these various characterizations of a rationality analysis; how, for instance, would a decision that is not “tenably supported” (and is thus “merely” unreasonable) differ from a decision that is not “rationally supported” (and is thus patently unreasonable)?

[108] In the end, the essential question remains the same under both standards: was the decision of the adjudicator taken in accordance with reason? Where the answer is no, for instance because the legislation in question cannot rationally support the adjudicator’s interpretation, the error will invalidate the decision, regardless of whether the standard applied is reasonableness simpliciter or patent unreasonableness (see D. K. Lovett, “That Curious Curial Deference Just Gets Curiouser and Curiouser — Canada (Director of Investigation and Research) v. Southam Inc.” (1997), 55 Advocate (B.C.) 541, at p. 545). Because the two variants of reasonableness are united at their theoretical source, the imperative for the reviewing court to intervene will turn on the conclusion that the adjudicator’s decision deviates from what falls within the ambit of the reasonable, not on “fine distinctions” between the test for patent unreasonableness and reasonableness simpliciter (see Falzon, supra, at p. 33).

[109] The existence of these two variants of reasonableness review forces reviewing courts to continue to grapple with the significant practical problems inherent in distinguishing meaningfully between the two standards. To the extent that a distinction is advanced on the basis of the relative severity of the defect, this poses not only practical difficulties but also difficulties in principle, as this approach implies that patent unreasonableness, in requiring “clear” rather than “mere” irrationality, allows for a margin of appreciation for decisions that are not in accordance with reason. In this respect, I would echo Mullan’s comments that there would “have to be concerns with a regime of judicial review which would allow any irrational decision to escape rebuke even under the most deferential standard of scrutiny” (Mullan, “Recent Developments in Standard of Review”, supra, at p. 25).

(iii) The “Immediacy or Obviousness” of the Defect

[110] There is a second approach to distinguishing between patent unreasonableness and reasonableness simpliciter that requires discussion. Southam, supra, at para. 57, emphasized the “immediacy or obviousness” of the defect:

The difference between “unreasonable” and “patently unreasonable” lies in the immediacy or obviousness of the defect. If the defect is apparent on the face of the tribunal’s reasons, then the tribunal’s decision is patently unreasonable. But if it takes some significant searching or testing to find the defect, then the decision is unreasonable but not patently unreasonable.

[111] In my view, two lines of difficulty have emerged from emphasizing the “immediacy or obviousness” of the defect, and thus the relative invasiveness of the review necessary to find it, as a means of distinguishing between patent unreasonableness and reasonableness simpliciter. The first is the difficulty of determining how invasive a review is invasive enough, but not too invasive, in each case. The second is the difficulty that flows from ambiguity as to the intended meaning of “immediacy or obviousness” in this context: is it the obviousness of the defect in the sense of its transparency on the face of the decision that is the defining characteristic of patent unreasonableness review (see J. L. H. Sprague, “Another View of Baker” (1999), 7 Reid’s Administrative Law 163, at pp. 163 and 165, note 5), or is it rather the obviousness of the defect in terms of the ease with which, once found, it can be identified as severe? The latter interpretation may bring with it difficulties of the sort I referred to above — i.e., attempting to qualify degrees of irrationality. The former interpretation, it seems to me, presents problems of its own, which I discuss below.

[112] Turning first to the difficulty of actually applying a distinction based on the “immediacy or obviousness” of the defect, we are confronted with the criticism that the “somewhat probing examination” criterion (see Southam, supra, at para. 56) is not clear enough (see D. W. Elliott, “Suresh and the Common Borders of Administrative Law: Time for the Tailor?” (2002), 65 Sask. L. Rev. 469, at pp. 486-87). As Elliott notes: “[t]he distinction between a ‘somewhat probing examination’ and those which are simply probing, or are less than probing, is a fine one. It is too fine to permit courts to differentiate clearly among the three standards.”

[113] This Court has itself experienced some difficulty in consistently performing patent unreasonableness review in a way that is less probing than the “somewhat probing” analysis that is the hallmark of reasonableness simpliciter. Despite the fact that a less invasive review has been described as a defining characteristic of the standard of patent unreasonableness, in a number of the Court’s recent decisions, including Toronto (City) Board of Education, supra, and Ivanhoe, supra, one could fairly characterize the Court’s analysis under this standard as at least “somewhat” probing in nature.

[114] Even prior to Southam and the development of reasonableness simpliciter, there was some uncertainty as to how intensely patent unreasonableness review is to be performed. This is particularly evident in National Corn Growers, supra (see generally Mullan, “Of Chaff Midst the Corn”, supra; Mullan, Administrative Law, supra, at pp. 72-73). In that case, while Wilson J. counselled restraint on the basis of her reading of CUPE, supra, Gonthier J., for the majority, performed quite a searching review of the decision of the Canadian Import Tribunal. He reasoned, at p. 1370, that “[i]n some cases, the unreasonableness of a decision may be apparent without detailed examination of the record. In others, it may be no less unreasonable but this can only be understood upon an in‑depth analysis.”

[115] Southam itself did not definitively resolve the question of how invasively review for patent unreasonableness should be performed. An intense review would seem to be precluded by the statement that, “if it takes some significant searching or testing to find the defect, then the decision is unreasonable but not patently unreasonable” (para. 57). The possibility that, in certain circumstances, quite a thorough review for patent unreasonableness will be appropriate, however, is left open: “[i]f the decision under review is sufficiently difficult, then perhaps a great deal of reading and thinking will be required before the judge will be able to grasp the dimensions of the problem” (para. 57).

[116] This brings me to the second problem: in what sense is the defect immediate or obvious? Southam left some ambiguity on this point. As I have outlined, on the one hand, a patently unreasonable decision is understood as one that is flawed by a defect that is evident on the face of the decision, while an unreasonable decision is one that is marred by a defect that it takes significant searching or testing to find. In other places, however, Southam suggests that the “immediacy or obviousness” of a patently unreasonable defect refers not to the ease of its detection, but rather to the ease with which, once detected, it can be identified as severe. Particularly relevant in this respect is the statement that “once the lines of the problem have come into focus, if the decision is patently unreasonable, then the unreasonableness will be evident” (para. 57). It is the (admittedly sometimes only tacit) recognition that what must in fact be evident — i.e., clear, obvious, or immediate — is the defect’s magnitude upon detection that allows for the possibility that in certain circumstances “it will simply not be possible to understand and respond to a patent unreasonableness argument without a thorough examination and appreciation of the tribunal’s record and reasoning process” (see Mullan, Administrative Law, supra, at p. 72; see also Ivanhoe, supra, at para. 34).

[117] Our recent decision in Ryan has brought more clarity to Southam, but still reflects a degree of ambiguity on this issue. In Ryan, at para. 52, the Court held:

In Southam, supra, at para. 57, the Court described the difference between an unreasonable decision and a patently unreasonable one as rooted “in the immediacy or obviousness of the defect”. Another way to say this is that a patently unreasonable defect, once identified, can be explained simply and easily, leaving no real possibility of doubting that the decision is defective. A patently unreasonable decision has been described as “clearly irrational” or “evidently not in accordance with reason” (Canada (Attorney General) v. Public Service Alliance of Canada, 1993 CanLII 125 (SCC), [1993] 1 S.C.R. 941, at pp. 963‑64, per Cory J.; Centre communautaire juridique de l’Estrie v. Sherbrooke (City), 1996 CanLII 168 (SCC), [1996] 3 S.C.R. 84, at paras. 9‑12, per Gonthier J.). A decision that is patently unreasonable is so flawed that no amount of curial deference can justify letting it stand. [Emphasis added.]

This passage moves the focus away from the obviousness of the defect in the sense of its transparency “on the face of the decision”, to the obviousness of its magnitude once it has been identified. At other points, however, the relative invasiveness of the review required to identify the defect is emphasized as the means of distinguishing between patent unreasonableness and reasonableness simpliciter:

A decision may be unreasonable without being patently unreasonable when the defect in the decision is less obvious and might only be discovered after “significant searching or testing” (Southam, supra, at para. 57). Explaining the defect may require a detailed exposition to show that there are no lines of reasoning supporting the decision which could reasonably lead that tribunal to reach the decision it did.

(Ryan, supra, at para. 53)

[118] Such ambiguity led commentators such as David Phillip Jones to continue to question in light of Ryan whether

whatever it is that makes the decision “patently unreasonable” [must] appear on the face of the record . . . Or can one go beyond the record to demonstrate — “identify” — why the decision is patently unreasonable? Is it the “immediacy and obviousness of the defect” which makes it patently unreasonable, or does patently unreasonable require outrageousness so that the decision is so flawed that no amount of curial deference can justify letting it stand?

(D. P. Jones, “Notes on Dr. Q and Ryan: Two More Decisions by the Supreme Court of Canada on the Standard of Review in Administrative Law”, paper originally presented at the Canadian Institute for the Administration of Justice, Western Roundtable, Edmonton, April 25, 2003, at p. 10.)

[119] As we have seen, the answers to such questions are far from self-evident, even at the level of theoretical abstraction. How much more difficult must they be for reviewing courts and counsel struggling to apply not only patent unreasonableness, but also reasonableness simpliciter? (See, in this regard, the comments of Mullan in “Recent Developments in Standard of Review”, supra, at p. 4.)

[120] Absent reform in this area or a further clarification of the standards, the “epistemological” confusion over the relationship between patent unreasonableness and reasonableness simpliciter will continue. As a result, both the types of errors that the two variants of reasonableness are likely to catch — i.e., interpretations that fall outside the range of those that can be “reasonably”, “rationally” or “tenably” supported by the statutory language — and the way in which the two standards are applied will in practice, if not necessarily in theory, be much the same.

[121] There is no easy way out of this conundrum. Whatever attempts are made to clarify the contours of, or the relationship between, the existing definitional strands of patent unreasonableness, this standard and reasonableness simpliciter will continue to be rooted in a shared rationale: statutory language is often ambiguous and “admits of more than one possible meaning”; provided that the expert administrative adjudicator’s interpretation “does not move outside the bounds of reasonably permissible visions of the appropriate interpretation, there is no justification for court intervention” (Mullan, “Recent Developments in Standard of Review”, supra, at p. 18). It will thus remain difficult to keep these standards conceptually distinct, and I query whether, in the end, the theoretical efforts necessary to do so are productive. Obviously any decision that fails the test of patent unreasonableness must also fall on a standard of reasonableness simpliciter, but it seems hard to imagine situations where the converse is not also true: if a decision is not supported by a tenable explanation (and is thus unreasonable) (Ryan, supra, at para. 55), how likely is it that it could be sustained on “any reasonable interpretation of the facts or of the law” (and thus not be patently unreasonable) (National Corn Growers, supra, at pp. 1369‑70, per Gonthier J.)?

[122] Thus, both patent unreasonableness and reasonableness simpliciter require that reviewing courts pay “respectful attention” to the reasons of adjudicators in assessing the rationality of administrative decisions (see Baker v. Canada (Minister of Citizenship and Immigration), 1999 CanLII 699 (SCC), [1999] 2 S.C.R. 817, at para. 65, per L’Heureux‑Dubé J., citing D. Dyzenhaus, “The Politics of Deference: Judicial Review and Democracy”, in M. Taggart, ed., The Province of Administrative Law (1997), 279, at p. 286, and Ryan, supra, at para. 49).

[123] Attempting to differentiate between these two variants of curial deference by classifying one as “somewhat more probing” in its attentiveness than the other is unlikely to prove any more successful in practice than it has proven in the past. Basing the distinction on the relative ease with which a defect may be detected also raises a more theoretical quandary: the difficulty of articulating why a defect that is obvious on the face of a decision should present more of an imperative for court intervention than a latent defect. While a defect may be readily apparent because it is severe, a severe defect will not necessarily be readily apparent; by the same token, a flaw in a decision may be immediately evident, or obvious, but relatively inconsequential in nature.

[124] On the other hand, the effect of clarifying that the language of “immediacy or obviousness” goes not to ease of detection, but rather to the ease with which, once detected (on either a superficial or a probing review), a defect may be identified as severe might well be to increase the regularity with which reviewing courts subject decisions to as intense a review on a standard of patent unreasonableness as on a standard of reasonableness simpliciter, thereby further eliding any difference between the two.

[125] An additional effect of clarifying that the “immediacy or obviousness” of the defect refers not to its transparency on the face of the decision but rather to its magnitude upon detection is to suggest that it is feasible and appropriate for reviewing courts to attempt to qualify degrees of irrationality in assessing the decisions of administrative adjudicators: i.e., this decision is irrational enough to be unreasonable, but not so irrational as to be overturned on a standard of patent unreasonableness. Such an outcome raises questions as to whether the legislative intent could ever be to let irrational decisions stand. In any event, such an approach would seem difficult to reconcile with the rule of law.

[126] I acknowledge that there are certain advantages to the framework to which this Court has adhered since its adoption in Southam, supra, of a third standard of review. The inclusion of an intermediate standard does appear to provide reviewing courts with an enhanced ability to tailor the degree of deference to the particular situation. In my view, however, the lesson to be drawn from our experience since then is that those advantages appear to be outweighed by the current framework’s drawbacks, which include the conceptual and practical difficulties that flow from the overlap between patent unreasonableness and reasonableness simpliciter, and the difficultly caused at times by the interplay between patent unreasonableness and correctness.

[127] In particular, the inability to sustain a viable analytical distinction between the two variants of reasonableness has impeded their application in practice in a way that fulfils the theoretical promise of a more precise reflection of the legislature’s intent. In the end, attempting to distinguish between the unreasonable and the patently unreasonable may be as unproductive as attempting to differentiate between the “illegible” and the “patently illegible”. While it may be possible to posit, in the abstract, some kind of conceptual distinction, the functional reality is that once a text is illegible — whether its illegibility is evident on a cursory glance or only after a close examination — the result is the same. There is little to be gained from debating as to whether the text is illegible simpliciter or patently illegible; in either case it cannot be read.

[128] It is also necessary to keep in mind the theoretical foundations for judicial review and its ultimate purpose. The purpose of judicial review is to uphold the normative legal order by ensuring that the decisions of administrative decision makers are both procedurally sound and substantively defensible. As McLachlin C.J. explained in Dr. Q, supra, at para. 21, the two touchstones of judicial review are legislative intent and the rule of law:

[In Pushpanathan,] Bastarache J. affirmed that “[t]he central inquiry in determining the standard of review exercisable by a court of law is the legislative intent of the statute creating the tribunal whose decision is being reviewed” (para. 26). However, this approach also gives due regard to “the consequences that flow from a grant of powers” (Bibeault, supra, at p. 1089) and, while safeguarding “[t]he role of the superior courts in maintaining the rule of law” (p. 1090), reinforces that this reviewing power should not be employed unnecessarily. In this way, the pragmatic and functional approach inquires into legislative intent, but does so against the backdrop of the courts’ constitutional duty to protect the rule of law.

In short, the role of a court in determining the standard of review is to be faithful to the intent of the legislature that empowered the administrative adjudicator to make the decision, as well as to the animating principle that, in a society governed by the rule of law, power is not to be exercised arbitrarily or capriciously.

[129] As this Court has observed, the rule of law is a “highly textured expression, importing many things which are beyond the need of these reasons to explore but conveying, for example, a sense of orderliness, of subjection to known legal rules and of executive accountability to legal authority” (Reference re Resolution to Amend the Constitution, 1981 CanLII 25 (SCC), [1981] 1 S.C.R. 753, at pp. 805‑6). As the Court elaborated in Reference re Secession of Quebec, 1998 CanLII 793 (SCC), [1998] 2 S.C.R. 217, at para. 71:

In the Manitoba Language Rights Reference, supra, at pp. 747‑52, this Court outlined the elements of the rule of law. We emphasized, first, that the rule of law provides that the law is supreme over the acts of both government and private persons. There is, in short, one law for all. Second, we explained, at p. 749, that “the rule of law requires the creation and maintenance of an actual order of positive laws which preserves and embodies the more general principle of normative order”. . . . A third aspect of the rule of law is . . . that “the exercise of all public power must find its ultimate source in a legal rule”. Put another way, the relationship between the state and the individual must be regulated by law. Taken together, these three considerations make up a principle of profound constitutional and political significance.

“At its most basic level”, as the Court affirmed, at para. 70, “the rule of law vouchsafes to the citizens and residents of the country a stable, predictable and ordered society in which to conduct their affairs. It provides a shield for individuals from arbitrary state action.”

[130] Because arbitrary state action is not permissible, the exercise of power must be justifiable. As the Chief Justice has noted,

. . . societies governed by the Rule of Law are marked by a certain ethos of justification. In a democratic society, this may well be the general characteristic of the Rule of Law within which the more specific ideals . . . are subsumed. Where a society is marked by a culture of justification, an exercise of public power is only appropriate where it can be justified to citizens in terms of rationality and fairness.

(See the Honourable Madam Justice B. McLachlin, “The Roles of Administrative Tribunals and Courts in Maintaining the Rule of Law” (1998-1999), 12 C.J.A.L.P. 171, at p. 174 (emphasis in original); see also MacLauchlan, supra, at pp. 289-91.)

Judicial review on substantive grounds ensures that the decisions of administrative adjudicators are capable of rational justification; review on procedural grounds (i.e., does the decision meet the requirements of procedural fairness?) ensures that they are fair.

[131] In recent years, this Court has recognized that both courts and administrative adjudicators have an important role to play in upholding and applying the rule of law. As Wilson J. outlined in National Corn Growers, supra, courts have come to accept that “statutory provisions often do not yield a single, uniquely correct interpretation” and that an expert administrative adjudicator may be “better equipped than a reviewing court to resolve the ambiguities and fill the voids in the statutory language” in a way that makes sense in the specialized context in which that adjudicator operates (p. 1336, citing J. M. Evans et al., Administrative Law (3rd ed. 1989), at p. 414). The interpretation and application of the law is thus no longer seen as exclusively the province of the courts. Administrative adjudicators play a vital and increasing role. As McLachlin J. helpfully put it in a recent speech on the roles of courts and administrative tribunals in maintaining the rule of law: “A culture of justification shifts the analysis from the institutions themselves to, more subtly, what those institutions are capable of doing for the rational advancement of civil society. The Rule of Law, in short, can speak in several voices so long as the resulting chorus echoes its underlying values of fairness and rationality” (McLachlin, supra, at p. 175).

[132] In affirming the place for administrative adjudicators in the interpretation and application of the law, however, there is an important distinction that must be maintained: to say that the administrative state is a legitimate player in resolving legal disputes is properly to say that administrative adjudicators are capable (and perhaps more capable) of choosing among reasonable decisions. It is not to say that unreasonable decision making is a legitimate presence in the legal system. Is this not the effect of a standard of patent unreasonableness informed by an intermediate standard of reasonableness simpliciter?

[133] On the assumption that we can distinguish effectively between an unreasonable and a patently unreasonable decision, there are situations where an unreasonable (i.e., irrational) decision must be allowed to stand. This would be the case where the standard of review is patent unreasonableness and the decision under review is unreasonable, but not patently so. As I have noted, I doubt that such an outcome could be reconciled with the intent of the legislature which, in theory, the pragmatic and functional analysis aims to reflect as faithfully as possible. As a matter of statutory interpretation, courts should always be very hesitant to impute to the legislature any intent to let irrational administrative acts stand, absent the most unequivocal statement of such an intent (see Sullivan and Driedger on the Construction of Statutes (4th ed. 2002), at pp. 367-68). As a matter of theory, the constitutional principle of the primacy of the rule of law, which is an ever-present background principle of interpretation in this context, reinforces the point: if a court concludes that the legislature intended that there be no recourse from an irrational decision, it seems highly likely that the court has misconstrued the intent of the legislature.

[134] Administrative law has developed considerably over the last 25 years since CUPE. This evolution, which reflects a strong sense of deference to administrative decision makers and an acknowledgment of the importance of their role, has given rise to some problems or concerns. It remains to be seen, in an appropriate case, what should be the solution to these difficulties. Should courts move to a two standard system of judicial review, correctness and a revised unified standard of reasonableness? Should we attempt to more clearly define the nature and scope of each standard or rethink their relationship and application? This is perhaps some of the work which lies ahead for courts, building on the developments of recent years as well as on the legal tradition which created the framework of the present law of judicial review.

III. Disposition

[135] Subject to my comments in these reasons, I concur with Arbour J.’s disposition of the appeal.

Appeal dismissed with costs.

Solicitors for the appellant: Caley & Wray, Toronto.

Solicitors for the respondent the City of Toronto: Osler, Hoskin & Harcourt, Toronto.

Solicitor for the intervener: Attorney General of Ontario, Toronto.